21.6 Chronic kidney disease 4830 Alastair Hutchiso

21.6 Chronic kidney disease 4830 Alastair Hutchison

ESSENTIALS Definition Chronic kidney disease (CKD) is defined as kidney damage lasting for more than 3 months characterized by structural or functional abnor- malities of the kidney, with or without decreased glomerular filtration rate (GFR). Staging CKD has been subdivided into six stages depending on the esti- mated GFR (eGFR) and degree of proteinuria: CKD stage 1 is eGFR greater than 90 ml/​min (per 1.73 m2) with other evidence of renal disease; CKD stage 2 is eGFR 60 to 89 ml/​min, with other evidence of renal disease; CKD stage 3a is eGFR 45 to 59 ml/​min; CKD stage 3b is eGFR 30 to 44 ml/​min; CKD stage 4 is eGFR 15 to 29 ml/​ min; and CKD stage 5 is eGFR less than 15 ml/​min. At each stage the CKD is further categorized according to the degree of pro- teinuria based on the albumin:creatinine ratio (ACR), from A1 (no increase in protein excretion) to A3 (severe proteinuria). The eGFR is least accurate when the serum creatinine is within or near the normal range. Epidemiology Mild CKD is common, with about 10% of the population of the United States of America having CKD stage 1, 2, or 3 (combined), but advanced CKD is relatively rare (about 0.2% are receiving renal re- placement therapy). Patients with CKD stage 1, 2, or 3 are at relatively low risk of progressing to require renal replacement therapy, but are at high risk of death from cardiovascular disease. Aetiology The causes of chronic renal failure recorded in various national registries are diabetes mellitus (22–​45%), glomerulonephritis (10–​23%), hypertension (5–​25%), chronic pyelonephritis (0.5 to 7%), adult polycystic kidney disease (2–​7%), renal vascular disease (2–​7%), other recognized conditions (13–​15%), and un- known causes (4–​26%). However, these data are flawed for many reasons: diagnoses are often allocated as ‘best guesses’ by clin- icians; there is no universal agreement on the meaning of terms such as ‘pyelonephritis’; glomerulonephritis may be diagnosed without histological proof; hypertension is often cited when it may be no more than a consequence of whatever caused the renal failure. Pathophysiology Compensatory mechanisms and their consequences—​as kidney function gradually fails, these generally maintain acceptable health until the GFR is about 10 to 15 ml/​min, and patients will not usually die of renal failure until the GFR is less than 5 ml/​min. Despite a widened range of single-​nephron GFR in damaged or diseased kidneys, glomerular and tubular function remains closely integrated in all individual nephrons (the ‘intact nephron hypothesis’). However, the functional adaptations required to maintain overall homeostasis come at a price (the ‘trade-​off hy- pothesis’), with the ‘hyperfiltration hypothesis’ most clearly ar- ticulating how these adaptive changes lead, in the long run, to glomerulosclerosis and tubulointerstitial fibrosis and progressive decrease in GFR. Pathophysiological changes—​these include impairment in (1) con- centration and/​or dilution of the urine; (2) excretion and/​or con- servation of sodium; (3) excretion of potassium, with hyperkalaemia often the immediate life-​threatening consideration in the manage- ment of patients with renal failure; (4) excretion of acid; (5) calcium/​ phosphate/​vitamin D/​bone homeostasis; (6) erythropoietin produc- tion, leading to renal anaemia; (7) excretion of many substances and metabolites that act as ‘uraemic toxins’; and (8) a wide range of endo- crine functions. The clinical presentation of CKD is discussed in Chapter 21.3 and investigation of patients with renal disease in Chapter 21.4. 21.6 Chronic kidney disease Alastair Hutchison Acknowledgement: Professor E. Ritz, Professor T. Drueke, and Dr J. Firth wrote on chronic kidney disease for the fifth edition of this textbook: some parts of their text and some figures are retained here. Dr C. Winearls wrote on chronic renal failure in the fourth edition of this textbook: a small part of his chapter is also retained.

21.6  Chronic kidney disease 4831 Prevention of progression Specific and general measures—​in some patients, measures to con- serve renal function may be specific to the cause of renal impair- ment (e.g. relief of obstruction), but it is probable that all patients will benefit from good blood pressure control and (when relevant) measures to reduce proteinuria, which is not only a marker but a promoter of progression of CKD. Blood pressure and proteinuria—​there is limited information on
the target blood pressure to be achieved in patients with chronic kidney disease, but the consensus is that the lower the blood pres- sure, the better—​as long as this can be achieved without unaccept- able side effects. The 2013 European Society of Hypertension/​ European Society of Cardiology guidelines recommend a target systolic blood pressure less than 140 mmHg, and state that if there is significant proteinuria, less than 130  mmHg ‘may be pursued’. Proteinuria should be lowered as much as possible, preferably to less than 1 g/​24 h (roughly equivalent to an ACR of <60 mg/​mmol or a protein:creatinine ratio of <100 mg/​mmol). Combination therapy with several antihypertensive agents (including loop diuretics) is usu- ally required, but there is good evidence that the regimen should contain an angiotensin-​converting enzyme inhibitor or angiotensin receptor blocker, which have antiproteinuric effects, if they are tol- erated (hyperkalaemia being the most common reason why they cannot be used in this context). Medical management of the consequences of CKD Diet—​only patients with oliguric endstage renal failure need to re- strict their fluid intake precisely. It is sensible to recommend modest dietary sodium restriction (100 mmol/​day) in most cases. Patients with a tendency to hyperkalaemia should be offered advice re- garding a low-​potassium diet (with particular care taken if they are given medications that induce hyperkalaemia). Chronic acidosis will benefit from treatment with alkali. Malnutrition is common in ad- vanced CKD, can be detected by serial monitoring of body weight and serum albumin concentration, and is best treated by initiating renal replacement therapy. CKD mineral and bone disorders—​these include hyperparathyroid bone disease, mixed lesions, osteomalacia, adynamic bone, and osteopenia/​porosis, the impact of which extends beyond the bones to cardiovascular structure and function, with increased mortality. Pathogenesis is complex but includes phosphate retention, defi- ciency of active forms of vitamin D, hypocalcaemia, and the devel- opment of hyperparathyroidism. Secondary hyperparathyroidism can be prevented by giving (1) cholecalciferol 2000 U/​day if serum 25-​(OH)D3 is low; (2) calcium carbonate 0.5 to 1.0 g with each meal if plasma calcium is decreased and/​or plasma phosphate is increased; (3)  calcium-​free phosphate binder (e.g. sevelamer or lanthanum carbonate) if serum phosphate is increased and plasma calcium is normal or high; or (4) calcitriol 0.25 µg/​day, or equivalent doses of alfacalcidol or other active vitamin D analogues, if serum parathy- roid hormone (PTH) is consistently above target ranges and serum calcium/​phosphate is normal (spontaneously or after intervention). Advanced hyperparathyroidism can be treated by (1) normalizing serum calcium and phosphate levels if serum intact PTH is con- stantly above target range; (2)  reducing serum phosphate, if this is elevated, by using phosphate binders, dietary restriction, and increased dialysis; (3) reducing serum calcium if this is elevated by reducing/​withdrawing calcium-​containing phosphate binders and active vitamin D sterols, and by reducing dialysate calcium concen- tration; (4) if serum calcium and phosphate have been normalized and elevated intact PTH persists, by increasing the dose or fre- quency of calcitriol or other active vitamin D sterols (e.g. alfacalcidol, paricalcitol, and doxercalciferol), or alternatively administering the calcimimetic cinacalcet, which renders the calcium receptor more sensitive to calcium; and (5) if serum intact PTH fails to decrease and/​ or hypercalcaemia/​hyperphosphataemia develops or persist, then consider surgical parathyroidectomy or cinacalcet. Anaemia—​this is common in CKD and is particularly marked in patients with diabetes. Partial correction of such anaemia by use of erythropoiesis-​stimulating agents (ESAs) improves patients’ physio- logical and clinical status, as well as quality of life. If a patient with CKD has a haemoglobin concentration of less than 11 g/​dl and symp- toms that might be attributable to anaemia, then treatment to re- store haemoglobin to the range 11 to 12 g/​dl is warranted, but it has been convincingly shown in randomized studies that correction to a higher level (‘normal or near normal’) is associated with poorer out- comes and should be avoided. Treatment involves (1) exclusion of other causes of anaemia; (2) optimization of iron status, which usually requires administration of intravenous iron; and (3) initiation and ad- justment of dosage/​frequency of administration of ESAs, with regular monitoring to achieve haemoglobin in the target range 11 to 12 g/​dl. Preparation for renal replacement therapy or conservative management of uraemia Once endstage renal failure is inevitable, the patient must be pre- pared physically and psychologically for renal replacement therapy (see Chapters 21.7.1–​21.7.3). In many cases, it is possible to predict approximately when the endstage will be reached from consider- ation of the rate of renal deterioration, most easily demonstrated by plotting the declining eGFR against time. There are patients for whom dialysis may appear inappropriate, or who either choose not to start or to discontinue treatment. In frail patients, usually elderly and with multiple comorbidities, it is not likely that dialysis will greatly prolong life, although it can certainly reduce the quality of it. The ethical and legal issues are complex and require that the patient makes the decision not to start or to discon- tinue treatment when fully informed and able to do so. They must be given a realistic account of what dialysis can achieve, what it cannot achieve, and at what cost—​access, travel, restrictions, and complica- tions. These conversations can be difficult and cannot be hurried, it being critically important that the patient (and their relatives/​friends) does not get the entirely erroneous impression that dialysis means that ‘the doctors care and I’ll live for ever’, whereas no dialysis means that ‘the doctors don’t care and I’ll die soon’. Properly managed, death from uraemia is peaceful and free of suffering. Definition Chronic kidney disease (CKD) is defined as an abnormality of kidney structure or function lasting for more than 3 months. The current classification is presented in the Kidney Disease: Improving

section 21  Disorders of the kidney and urinary tract 4832 Global Outcomes (KDIGO) CKD Work Group (2013). The classi- fication, prognosis, and risk of adverse health-​related outcomes in CKD are assessed using glomerular filtration rate (GFR) category and albuminuria category (Fig. 21.6.1). A decline in GFR and an in- crease in albumin:creatinine ratio increases the risk of adverse renal and cardiovascular outcomes with a combination of the two multi- plying the risk (Fig. 21.6.2). The current classification is based on six stages of kidney func- tion based on GFR and termed G1 to G5 with three categories of albumin:creatinine ratio, A1 to A3. Stage G5 A3 would reflect those with the poorest function, that is, GFR less than 15 and kidney failure with severely increased albuminuria greater than 30 mg/​ mmol (Fig. 21.6.1). The assessment of kidney function and definition of impair- ment should be based on measurements of serum creatinine using standardized assays. Clinical laboratories should report for every serum creatinine result (measured in an ambulatory context) a glomerular filtration rate estimation (eGFR) using a prediction equation such as the Chronic Kidney Disease Epidemiology Collaboration (CKD-​EPI) creatinine equation or an update of the formula used in the Modification of Diet in Renal Disease (MDRD) study. This is based on standardized measurement of serum creatinine (a method which is easily disturbed by numerous confounders) and considers, in addition to serum creatinine, the age, sex, and ethnicity of the patient (see Chapter 21.4 for further discussion). There are pitfalls in the assessment of kidney function and diag- nosis of CKD when using GFR. CKD should not be confirmed if the duration of the impaired function is less than 3 months or unclear. In individuals with extremes of muscle mass, GFR calculated from serum creatinine will be underestimated in those with high muscle mass such as bodybuilders and overestimated in those with low muscle mass, including amputees and patients with muscle wastage. A correction factor, multiplication of the eGFR by 1.14, is also re- quired for those of African or Afro-​Caribbean ethnicity when using the CKD-​EPI creatinine equation. A meal including meat (protein rich) should be avoided in the 12 h before a blood sample is obtained for creatinine and GFR estimation. The estimation of GFR is also inaccurate for those with preserved function and true GFR greater than 60 ml/​min per 1.73 m2. Caution in diagnosing CKD is also needed where eGFR based on serum cre- atinine is 45 to 59 ml/​min per 1.73 m2 and where there are no other markers of structural kidney disease or evidence of albuminuria. In such cases, serum cystatin C may be used to calculate eGFR, and if the value is greater than 60 ml/​min per 1.73 m2, a diagnosis of CKD should not be given (see Chapter 21.4). Prognosis of CKD by GFR and Albuminuria Categories: KDIGO 2012 Normal or high ≥90 Normal to mildly increased <30 mg/g <3 mg/mmol

300 mg/g 30 mg/mmol 30–300 mg/g 3–30 mg/mmol Moderately Increased Severely increased 60–89 45–59 30–44 15–29 <15 Mildly decreased Mildly to moderately decreased Moderately to severely decreased Severely decreased Kidney failure Persistent albuminuria categories Description and range GFR categories (mI/min/ 1.73m2) Description and range A1 G1 G2 G3a G3b G4 G5 A2 A3 Fig. 21.6.1  Stages of CKD and risk of adverse outcomes. Prognosis of CKD by GFR and albuminuria category. Green, low risk (if no other markers of kidney disease, no CKD); yellow, moderately increased risk; orange, high risk; red, very high risk. CKD, chronic kidney disease; GFR, glomerular filtration rate; KDIGO, Kidney Disease: Improving Global Outcomes. Reprinted from Kidney Disease: Improving Global Outcomes (KDIGO) CKD work group (2013). KDIGO 2012 clinical practice guideline for the evaluation and management of chronic kidney disease. Kidney International Supplements, 3, 1–​150. Copyright © 2013, with permission from Elsevier.

21.6  Chronic kidney disease 4833 Prevalence and incidence of chronic kidney disease The high frequency of impaired renal function in the general population has been recognized only recently. Worldwide preva- lence of CKD is estimated to be greater than 10%, with estimates of greater than 50% in some higher-​risk groups. Particularly at risk are prediabetic patients with metabolic syndrome, patients with diabetes mellitus, and individuals with many other conditions, for example, female sex, smokers, or patients receiving potentially nephrotoxic medications such as nonsteroidal anti-​inflammatory drugs (NSAIDs). Advanced age is another risk factor, with preva- lence over 20% of those aged greater than 60 and 35% of those older than 70 in the United States of America. Based on data from the National Health and Nutrition Examination Survey (NHANES) reported in 2003, it has become apparent that—​ in contrast to the relatively modest number of patients on renal re- placement therapy (RRT) in the United States of America (300 000, i.e. <1% of the population of 175 million) —​the estimated number of adults with CKD stages 1, 2, and 3 (combined) was 18.8 million (i.e. 10–​11% of the population). This observation is not only of aca- demic interest, but also of major public health importance. The reason is shown in Table 21.6.1: in individuals with CKD stage 2, the risk of them living long enough to require RRT is 20 times lower than their risk of dying from cardiovascular causes. Even patients in the more advanced stage 4 of CKD are twice as likely to die from cardiovascular causes as they are to end up on RRT. From a public health perspective, there is therefore a great need for early recogni- tion of CKD to allow opportunity for therapeutic interventions, in particular to improve cardiovascular prognosis. These public health implications for CKD are illustrated further in relation to risk of car- diovascular death in Fig. 21.6.3 and loss of life expectancy in relation to age in Table 21.6.1. The prevalence and incidence of endstage renal disease is shown in Table 21.6.2, which gives the total number of patients on RRT worldwide at the end of 2004 (i.e. 1.8 million), of whom 1.4 million were treated by haemodialysis or continuous ambulatory peritoneal dialysis (or other modalities of peritoneal dialysis) and 412 000 were alive with a functioning renal transplant. In Europe, an estimated total of 324 000 patients were on RRT and 149 000 were alive with a functioning renal graft. This gives a prevalence of 400 per million population for patients on dialysis (haemodialysis and peritoneal dialysis combined) and 185 per million population for kidney trans- plant recipients. Between 1990 and 1999, there was a dramatic increase in the adjusted incidence of RRT in Europe, rising from 73 per mil- lion population (range 58–​101  per million population) in 1990 to 1991 to 117 per million population (range 92–​145 per million population) in 1998 to 1999, that is, by 4.8% per year (range 3.1–​ 6.4%). The increase was greater in men than in women and did not flatten out at the end of the decade, except in the Netherlands. Since 2000, annual rises in RRT incidence rates have slowed being 120 per million population in 2000, 125 per million population in 2006, and 133 per million population in 2012. Rates appear to have stabilized for females aged 65 to 74 years, males aged 75 to 84 years, as well as in patients with RRT causes of hypertension and renovascular disease. It is of particular note that the incidence of endstage renal disease due to diabetic nephropathy (the com- monest single cause of requiring RRT) has begun to flatten out in 4 2 1 0.5 4.0 (b) (a) 2.0 1.0 0.5 2.5 5 10 30 300 ACR (mg/g) 1000 15 30 45 60 75 90 105 120 HR for CVD mortality (ACR studies) HR for CVD mortality (ACR studies) Significant values Non-significant values Fig. 21.6.2  Independent associations of kidney function and proteinuria with cardiovascular mortality. (a) Kidney function
(eGFR); reference value of 95 ml/​min per 1.73 m² is shown with a diamond. (b) Albuminuria (ACR); the reference value of 5 mg/​g
is shown with a diamond. Hazard ratios are adjusted for each
other (eGFR or ACR), age, sex, ethnic origin, and traditional cardiovascular risk factors. ACR, albumin-​to-​creatinine ratio; CVD, cardiovascular disease; eGFR=estimated glomerular filtration rate;
HR, hazard ratio. Reprinted from The Lancet, 382, 339–​52. Gansevoort R, et al. (2013). Chronic kidney disease and cardiovascular risk: epidemiology, mechanisms, and prevention. Copyright © 2013, with permission from Elsevier. Table 21.6.1  Risk (percentage chance over 5 years) of death from CKD progression towards endstage renal disease (defined as need for renal replacement therapy, RRT) and risk of death from cardiovascular disease in 27 998 American patients with glomerular filtration rate (GFR) less than 90 ml/​min per 1.73 m2
(for the period 1996–​2001) GFR (ml/​min per 1.73 m2) Risk over 5 years (%) RRT Cardiovascular death CKD stage 2 (60–​89) 1.1 19.5 CKD stage 3 (30–​59) 1.3 24.3 CKD stage 4 (15–​29) 19.9 45.7 Reproduced from Keith DS, et al. (2004). Longitudinal follow-​up and outcomes among a population with chronic kidney disease in a large managed care organization. Arch Intern Med, 164, 659–​63, with permission.

section 21  Disorders of the kidney and urinary tract 4834 the general population of the United States of America. It has also done so in Pima Indians, in whom the cardiovascular risk is less than in whites people, which excludes the possibility that failure to observe more diabetic patients reaching endstage renal disease was due to their more frequent death from cardiovascular causes before reaching endstage renal disease. A similar trend with regard to diabetic nephropathy as a cause of endstage renal failure has also been observed in Denmark, but there have been large differences in overall incidence and prevalence of RRT between European countries. Specifically, in the United Kingdom, the annual accept- ance rate for RRT increased from 20 per million population in 1982 to 101 per million population in 2002, largely due to higher acceptance rates of patients over 65 years of age and of those with comorbidities. Most recent data from the United Kingdom shows RRT incidence rates of 109 per million population in 2013, a rate which appears to have stabilized since 2006. It has long been known that endstage renal disease is much more frequent in older people than in young people. The European Renal Association/​European Dialysis and Transplant Association (ERA/​ EDTA) Registry data showed a four-​ to fivefold increase in the in- cidence and prevalence of RRT over the period 1985 to 1999, and in 1999 no fewer than 48% of new patients were above age 65 years, leading the registry authors to refer to RRT as an epidemic of ageing. In this context, it is also of interest that the long-​term function of kidney grafts obtained from elderly donors, irrespective of whether they were live donors or cadaveric, is inferior to the function of kidney grafts from younger donors. As typical signs of senescence, telomere length is reduced, repair capacity is curtailed, and the kid- neys of older people are generally hypoperfused as a consequence of vascular sclerosis. It is of interest that the incidence of new patients requiring RRT does not parallel the prevalence of CKD stages 3 and 4. Despite similar prevalence of CKD stages 3 and 4, the incidence of CKD pa- tients requiring RRT in Norway is significantly lower than in white patients in the United States of America. This is explained not by high mortality prior to endstage renal disease, but by more effective intervention to prevent progression of CKD to stage 5, an observa- tion that illustrates the great benefit of dedicated nephrological care for patients with advanced CKD. Causes of chronic kidney disease The usual sources for descriptions of the causes of CKD are endstage renal failure databases. These are flawed for many reasons: diagnoses are often allocated as ‘best guesses’ by clinicians; there is no uni- versal agreement on the meaning of terms such as ‘pyelonephritis’; glomerulonephritis may be diagnosed without histological proof; hypertension is often cited when it may be no more than a conse- quence of whatever caused the renal failure. With these important caveats in mind, Table 21.6.3 lists the given causes of endstage renal failure in recent reports from the United Kingdom Renal Registry, the Australia and New Zealand Dialysis and Transplant Registry, and the United States Renal Data Systems. Less common causes of chronic renal failure are given in Box 21.6.1. Note that some causes of chronic renal failure that are uncommon from a global perspec- tive (e.g. Balkan nephropathy and HIV nephropathy) may be very common in some populations. In most countries, diabetic nephropathy has become the single most frequent cause of endstage renal disease. In comparison with previous years, the relative contribution of glomerulonephritis has substantially decreased, and the diagnosis of ‘pyelonephritis’, that is, the concept that chronic bacterial colonization of the kidney and of the urinary tract causes chronic loss of renal function even in the absence of malformation or urological disease, has increas- ingly been abandoned on the basis of follow-​up studies in cohorts of subjects with uncomplicated chronic urinary tract infection who failed to develop CKD in the absence of additional pathologies. It is of note, however, that in aboriginal populations, the spectrum of renal disease is strikingly different, and this is also true for nonwhite immigrant populations, including in the United Kingdom, who Fig. 21.6.3  Loss of life expectancy due to cardiovascular disease, by chronic kidney disease stage (Canadian data). (a) eGFR stages, (b) albuminuria stages. Loss is compared with life expectancy in people with normal or mildly impaired kidney function (stage 1–​2, eGFR 60 ml/​min per 1.73 m² or higher) and normal or mildly increased albuminuria (stage 1, albumin-​to-​creatinine ratio <30 mg/​g). Data are adjusted for sex. eGFR, estimated glomerular filtration rate; RRT, renal replacement therapy. Reprinted from The Lancet, 382, 339–​52. Gansevoort R, et al. (2013). Chronic kidney disease and cardiovascular risk: epidemiology, mechanisms, and prevention. Copyright © 2013, with permission from Elsevier.

21.6  Chronic kidney disease 4835 have an excess of diabetic nephropathy and a variety of renal dis- eases which are less frequently or rarely seen in white patients, for example, renal tuberculosis, sickle cell anaemia nephropathy, and HIV nephropathy. Detailed discussion of particular causes of CKD can be found in other chapters. Pathophysiology of chronic kidney disease As kidney function gradually fails, compensatory mechanisms gener- ally maintain acceptable health until the GFR is about 10 to 15 ml/​min per 1.73 m2, and patients will not usually die of renal failure until the GFR is less than 5 ml/​min per 1.73 m2. The remarkable capacity of renal function to adapt to maintain overall homeostasis in the face of such dramatic reduction in glomerular filtration is best understood on the basis of the ‘intact nephron’ hypothesis and the ‘trade-​off’ hypothesis. Table 21.6.2  Global and regional overview of endstage renal disease (ESRD), dialysis, and transplant patient numbers, and prevalence values at year-​end 2004 (numbers rounded) Region Patient numbers Prevalence values (per million population) ESRD Dialysis (HD + PD) Transplant ESRD Dialysis (HD + PD) Transplant Global 1 783 000 1 371 000 412 000 280 215 65 North America 492 000 337 000 154 000 1505 1030 470 Europe 473 000 324 000 149 000 585 400 185 (thereof EU) (387 000) (252 000) (135 000) (850) (550) (295) Japan 261 000 248 000 13 000 2045 1945 100 Asia (excluding Japan) 237 000 196 000 41 000 70 60 10 Latin America 205 000 170 000 35 000 380 320 65 Africa 61 000 57 000 5000 70 65 5 Middle East 54 000 39 000 15 000 190 140 55 HD, haemodialysis; PD, peritoneal dialysis. Reproduced from Grassmann A, et al. (2005). ESRD patients in 2004: global overview of patient numbers, treatment modalities and associated trends. Nephrol Dial Transplant, 20, 2587–​93, with permission. Table 21.6.3  Percentage distribution of primary renal diagnosis in patients starting on renal replacement therapy Diagnosis UKRRa ANZDATAb USRDSc Uncertain 14.5d 5 Not specified Diabetese 25.4 35 43.8 Glomerulonephritis 14.4f 19g 7.5 Chronic pyelonephritish 6.9 2 Not specified Renal vascular disease 5.1 Not specified Not specified Adult polycystic kidney disease 7.6 6 2.1 Hypertensionj 7.6 14 28.7 Otherk 18.3 15 18l ANZDATA, Australia and New Zealand Dialysis and Transplant Registry (2014 report); UKRR, United Kingdom Renal Registry (December 2014 report); USRDS, United States Renal Data Systems (2015 report). a Data from 6353 patients beginning dialysis in United Kingdom renal units in 2013 for whom a primary renal diagnosis was reported. b Data from all 2544 patients beginning dialysis in Australian renal units in 2013. c Data from116 990 incident patients from 2013. d Also includes ‘glomerulonephritis not proven by biopsy’. e It can be impossible to know whether diabetes is a cause of endstage renal failure or merely an association. f 100% biopsy proven. g 76% biopsy proven. h Patients diagnosed as having ‘reflux nephropathy’ are included under this heading. i Classified as a subgroup of ‘hypertension’. j Reporting practices varied widely and variation in attribution of hypertension as a cause almost certainly does not reflect real differences in causation of endstage renal failure. k Includes urinary obstruction. l Includes ‘uncertain’. Box 21.6.1  Less common causes of chronic renal failure Metabolic Cystinosis, cystinuria (stones), oxalosis, nephrocalcinosis, and urate nephropathy. Hereditary Alport’s syndrome, Fabry’s disease, tuberous sclerosis, sickle cell disease, medullary cystic disease (and the metabolic conditions listed previously). Vasculitides and other multisystem disorders Granulomatosis with polyangiitis (formerly referred to as Wegener’s granulomatosis), microscopic polyangiitis, polyarteritis nodosa, Henoch–​ Schönlein purpura, systemic lupus erythematosus, scleroderma, and sarcoidosis. Malignancy Renal cell cancer, von Hippel–​Lindau disease, and lymphoma. Dysproteinaemias Myeloma, primary (AL) and secondary (AA) amyloid, and cryoglobulinaemia. Structural/​infection/​interstitial Cystic disease other than autosomal dominant polycystic kidney dis- ease, congenital and acquired abnormalities of the lower urinary tract, tuberculosis, schistosomiasis, Balkan nephropathy, Chinese herb neph- ropathy, analgesic nephropathy, nephrotoxic metals, and radiation nephropathy. Other Haemolytic uraemic syndrome, postpartum renal failure, acute cortical necrosis, accelerated (‘malignant’) hypertension, and HIV nephropathy.

section 21  Disorders of the kidney and urinary tract 4836 The intact nephron hypothesis, first articulated by Bricker, states that despite a widened range of single-​nephron GFR in damaged or diseased kidneys, glomerular and tubular function remains closely integrated in all individual nephrons, both normal and damaged. As the GFR of the whole kidney falls, those nephrons that are still functioning produce an increased volume of filtrate (hyperfiltration), and their tubules respond appropriately for overall homeostasis by excreting fluid and solutes in amounts that maintain external balance, although the capacity for adaptation is variable. For sodium and potassium, compensation can occur down to a GFR as low as 5 ml/​min per 1.73 m2, but this cannot be achieved for phosphate and urate, plasma concentrations of which may be elevated when the GFR falls below 20 ml/​min per 1.73 m2 in some patients. The trade-​off hypothesis recognizes that the functional adapta- tions required to maintain overall homeostasis come at a price and that they contribute to changes characteristic of the syndrome of uraemia. The best example of such a trade-​off is the generation of hyperparathyroidism. As GFR falls, leading (if there were no com- pensation) to a reduction in phosphate excretion and a rise in serum phosphate, the serum parathyroid hormone (PTH) concentration rises and serves to maintain homeostasis by reducing tubular re- absorption of phosphate. However, a consequence is secondary (and sometimes tertiary) hyperparathyroidism, with adverse effects on blood vessels and bones. Electrolyte, water, and acid–​base homeostasis An inability to concentrate the urine in the face of dehydration is sometimes the first symptom of chronic renal failure, manifesting as polyuria, nocturia, and thirst. This is particularly likely in condi- tions that predominantly affect the renal medulla (e.g. obstructive uropathy, interstitial nephritides, and medullary cystic disease). By contrast, urinary diluting capacity is preserved until renal failure is advanced, at which time urinary osmolality becomes fixed at around 300 mOsm/​kg (roughly isotonic with plasma) and there is obligatory polyuria. It should be noted, however, that although urinary diluting capacity is maintained until late in chronic renal failure, large water loads are excreted more slowly than in normal subjects and excessive intake (by drinking or ill-​advised iatrogenic infusion of dextrose-​ containing solutions) can lead to symptomatic hyponatraemia (see Chapter 21.2.1). As renal function decreases, sodium balance and extracellular fluid volume are maintained until GFR is less than about 10 ml/​ min per 1.73 m2 by an increase in the fractional excretion of sodium (the amount excreted in the urine as a fraction of that filtered at the glomerulus) from 1% (normal) to 30%. This capacity for adaptation is overwhelmed in advanced chronic renal failure, with sodium re- tention manifest as hypertension and/​or oedema (peripheral and/​or pulmonary). Such patients can also be at increased risk of sodium depletion as they are unable to restrict sodium excretion promptly in response to stimuli that would normally be expected to lead to such restriction (e.g. diarrhoea, vomiting). A  few patients with modest impairment of GFR (e.g. CKD stage 3), usually with diseases affecting the medulla, may present with low blood pressure (often with a postural drop) due to sodium depletion caused by a urinary sodium leak: sodium supplements may be required. Most patients maintain normal external potassium balance until their GFR is less than 5 ml/​min per 1.73 m2, but their capacity to excrete potassium is limited and severe hyperkalaemia may follow a sudden reduction in GFR such as might be caused by intercur- rent illness, excess dietary intake, or prescription of drugs that im- pair potassium excretion (angiotensin-​converting enzyme (ACE) inhibitors, angiotensin receptor blockers, potassium-​sparing di- uretics). Some patients, particularly those with diabetes mellitus and/​or interstitial nephritis, may develop hyperkalaemia due to hyporeninaemic hypoaldosteronism at levels of GFR that would not otherwise be expected to cause problems with potassium homeo- stasis. Hyperkalaemia is often the immediate life-​threatening con- sideration in the management of patients with renal failure. See Chapters 21.2.2 and 21.5 for further discussion. The kidney normally maintains acid–​base homeostasis by re- absorbing filtered bicarbonate, acidifying urinary buffers, and excreting ammonia. Increasing acidosis tends to occur at a GFR of less than 10 ml/​min per 1.73 m2, and is more likely to be an early feature in diseases that primarily affect the tubules and interstitium (aside from the renal tubular acidoses, where there are specific de- fects in acid–​base homeostasis—​see Chapter 21.15). Abnormalities of calcium and phosphate homeostasis are dis- cussed later in this chapter. Endocrine dysfunction In CKD, hormone concentrations may be elevated as a result of re- duced degradation (e.g. insulin) or increased secretion in response to metabolic changes (e.g. PTH), or reduced by impaired production (e.g. 1,25-​dihydroxyvitamin D, erythropoietin, oestrogen, or tes- tosterone). Reductions in hormone-​binding proteins (e.g. through protein losses in patients with nephrotic syndrome or on peritoneal dialysis) are common and may affect levels of free hormones cir- culating in the blood. Effects on vitamin D and erythropoietin are discussed later in this chapter. Total thyroxine (T4) and T3 (tri-​iodothyronine) may be low, with impaired peripheral deiodination of T4 to T3 and preferential di- version to inactive metabolites. However, patients are not clinically hypothyroid, and levels of thyroid-​stimulating hormone are gener- ally normal and can be used in the usual way as a diagnostic test for hypothyroidism. The kidney is the main site of growth hormone degradation and plasma levels of growth hormone are abnormally high in patients with renal failure because of this, and also because of alterations in hypothalamic–​pituitary control. It is not clear whether or not this has any clinical impact in adults with renal failure, but in children with renal failure and growth restriction the impaired production of insulin-​like growth factor 1 can be overcome by treatment with supraphysiological doses of recombinant growth hormone. Decreased insulin clearances seems to be balanced by increased peripheral resistance to the effects of insulin, hence patients with renal failure are not prone to hypoglycaemia or diabetes, but there is a reduced requirement for exogenous insulin in people with diabetes as renal function declines. Prolactin levels are high in renal failure and may contribute to gynaecomastia and sexual dysfunction in men and to infertility in women. Men with CKD may also have testosterone levels that are low to normal, with raised gonadotropins implying that testicular failure is the cause. The pituitary–​ovarian access is disturbed in ad- vanced renal failure, with many cycles anovulatory, causing oes- trogen deficiency.

21.6  Chronic kidney disease 4837 Middle molecules and the uraemic syndrome Many of the clinical manifestations that are seen with progressive decline in renal function are attributable to derangements in fluid balance, electrolyte handling, and endocrine function as previously described. However, these derangements do not provide an ad- equate explanation for all clinical features, and it is assumed that those which are unexplained are due to retention of substances and metabolites that the failing kidney is unable to excrete. The nature of these ‘uraemic toxins’ is uncertain. Accumulation of urea itself has modest effects, and failure to excrete a variety of small water-​ soluble compounds, protein-​bound compounds, and ‘middle mol- ecules’ (meaning those whose molecular weight is in the range 500–​12 000 Da) is held responsible, although incriminating spe- cific toxins has proved difficult. Among the small water-​soluble compounds, those thought to have a role in the uraemic syn- drome include various guanidine compounds, oxalate, phosphates, and polyamines; and among the protein-​bound compounds, p-​cresol and p-​cresylsulphate, homocysteine, various indoles, and furanproprionic acid. The best-​characterized example of a uraemic middle molecule is β2-​microglobulin, which is normally excreted by the kidneys, reaches a blood concentration 30 times higher than normal in dialysis patients, and accumulates as β2-​microglobulin amyloid in joints and bone. Many of the other middle molecules that have been incriminated have a proinflammatory effect, including a variety of advanced glycosylation end products that are probably at- tributable to increased concentrations of small carbonyl precursors in uraemia (and not to hyperglycaemia). Progression of chronic kidney disease Brenner advanced the concept that the adaptive changes (de- scribed in ‘Pathophysiology of chronic kidney disease’) aimed at increasing the excretory capacity of the kidney are maladaptive in the long run, causing deterioration of renal function mainly as a re- sult of glomerulosclerosis and tubulointerstitial fibrosis, the final result being endstage renal disease. The adaptive changes leading to single-​nephron hyperfiltration, which sustains whole-​kidney GFR in the short term, are mediated by a reduction of the resist- ance of the afferent preglomerular arterioles and an increase of the angiotensin-​dependent resistance of the postglomerular efferent ar- terioles, which combine to raise intraglomerular capillary pressure (glomerular hypertension). Over time, however, GFR will decrease, driven by several pathogenic mechanisms, including proteinuria and oxidative stress. This hypothesis led not only to the introduction of remarkably effective therapeutic and preventive approaches aimed at inter­ fering with progression (see later in this section), it also had re- percussions concerning the relation between GFR and metabolic abnormalities in CKD. The Mild and Moderate Kidney Disease (MMKD) study found an increase in the plasma concentration of asymmetric dimethyl l-​arginine (ADMA), even when whole-​ kidney GFR was still normal. The same was true for apolipoprotein abnormalities and sympathetic overactivity. A  plausible explan- ation for these findings, despite normal whole-​kidney GFR, is loss of nephrons with associated loss of metabolic capacity while whole-​kidney GFR is still normal because of single-​nephron hyperfiltration; hence, having a normal GFR does not exclude functional renal abnormalities. The rate of progression of CKD varies considerably depending on age, sex, type of underlying renal disease, genetics (family history of renal disease and cardiovascular disease), and many other fac- tors. For a given primary kidney disease, the risk of progression is lower in premenopausal women than in men. Such a sex difference is not found in children or postmenopausal women. This finding points to the role of sex hormones, which has also been documented in experimental studies. Testosterone aggravates renal disease, pos- sibly by raising blood pressure and activating the renin–​angiotensin system (RAS), whereas oestrogens ameliorate the evolution of renal disease. Some types of renal disease tend to progress slowly and others rapidly:  adult polycystic kidney disease, renal dysplasia, IgA-​ glomerulonephritis, and membranous glomerulonephritis usually progress slowly; rapid progression is anticipated in antiglomerular basement membrane glomerulonephritis and vasculitis; and inter- mediate rates of progression are typical in diabetic nephropathy. Ethnicity also determines the renal risk for currently poorly understood reasons: in Australian aboriginals, Maoris, American Indians, and particularly individuals of African descent, the fre- quency of CKD and the rate of its progression are substantially higher than in white people. This has been particularly well docu- mented in the immigrant population from south Asia living in the United Kingdom. A powerful predictor of the renal risk is a positive family history, not only of CKD but also of cardiovascular events in first-​degree relatives, which is associated with a substantial increase in the risk of developing progressive kidney disease. A history of pre-​eclampsia is associated with a dramatic increase in the risk of developing overt kidney disease later in life. Similarly, faulty prenatal programming in utero plays a role in the onset and progression of CKD. This is part and parcel of the Barker hypothesis, according to which interference with organogenesis in the prenatal period predisposes an individual at adult age to CKD, hypertension, and the metabolic syndrome. Although such nonmodifiable predictors are of some interest, more important clinically are modifiable predictors associated with a higher rate of progression (Box 21.6.2). It has been well documented that blood pressure has an adverse effect on progression of CKD, and this is true even for values below the former definition of hyperten- sion (>140/​90 mmHg). This finding has greatly increased the ability to interfere with progression, and the same is true for measures to interfere with the activation of the RAS to reduce proteinuria. Of major clinical importance is the fact that while renal function tends to deteriorate progressively in many (but not all) patients with CKD, almost exclusively this is true only in patients with albumin- uria/​proteinuria. The increased risk of cardiovascular events and cardiovascular death in patients with albuminuria/​proteinuria has already been emphasized (Table 21.6.1). Clinical presentation The various presentations of renal disease are discussed in Chapter 21.3, but in brief, the clinical presentation of CKD depends on the degree of renal dysfunction at the time that the condition is recognized.

section 21  Disorders of the kidney and urinary tract 4838 Asymptomatic patients At one extreme are patients with no symptoms whatsoever in whom an abnormal eGFR and/​or proteinuria are detected on routine examination, such as a medical examination for insurance purposes. Such patients may be shocked when hearing for the first time that they have lost what might be a substantial amount of their kidney function. Counselling and persuading them to comply with follow-​ up and to take medications occasionally proves difficult. Illnesses known to cause CKD, such as autosomal dominant polycystic kidney disease, other hereditary kidney diseases, or diabetic neph- ropathy, are generally easier to manage because these patients are more likely to understand the progressive nature of CKD and the benefit of treatment to prevent or slow progression. Patients with associated disease The presence of CKD may be diagnosed in many medical contexts in the absence of any symptoms pointing to the kidneys, for example, outpatient clinics for hypertension, diabetes, cardiac disorders, or urological diseases. Symptomatic presentation The diagnosis of CKD is made in relatively few patients on the basis of symptoms or signs pointing to kidney disease, such as nocturia, foaming of the urine, facial or pedal oedema, or haematuria. Symptoms or signs pointing to advanced CKD include (among others) lethargy, personality change, changes in mentation, loss of appetite, nausea (often in the morning), and vomiting. These may or may not be accompanied by evidence of hypervolaemia, such as dyspnoea on exertion or at rest, swollen ankles/​legs, elevated venous pressure, cardiomegaly, or basal crackles, and often only fluid re- moval can decide whether such manifestations are the result of fluid retention caused by renal failure, or by cardiac disease, and not in- frequently they are the result of both. The final stage of uraemia is heralded by bleeding tendency, pericarditis, obtundation, and coma. See Chapter 21.3 for further discussion. A relatively common clinical challenge to renal services is patients who present as an acute emergency requiring urgent RRT. This is not infrequently due to late referral, but it is not uncommon for there to be an acute deterioration of renal function in patients with pre-​existing CKD as a result of intercurrent illness or medical inter- ventions, for example, cardiac decompensation from myocardial infarction or arrhythmia, hypovolaemia, exposure to radiocontrast media, or prescription of particular drugs. Regarding the latter, common culprits would be the new prescription of ACE inhibitors or angiotensin receptor blockers, particularly in patients with ad- vanced CKD in the presence of hypovolaemia, heart failure, or other conditions activating the RAS. NSAIDs are also commonly incrim- inated, with other possibilities being nephrotoxic antibiotics (e.g. aminoglycosides) and cis-​platinum. When a patient arrives in hospital as an ‘acute uraemic emer- gency’, it is important to determine whether the problem is acute kidney injury (acute renal failure, which may be entirely revers- ible), acute-​on-​chronic kidney failure (when recovery to previous baseline renal function may be possible), or the endstage of pro- gressive chronic kidney failure (which by definition is not revers- ible). The only infallible method of determining whether a patient has CKD is to find documentation of previously reduced GFR in- dicated by past measurement of serum creatinine. When this is not available it is sometimes simply not possible to be sure of the situ- ation at the time of presentation, although ultrasonography with documentation of small kidneys is often of great help in indicating chronicity. Features that suggest the presence of chronic renal failure are shown in Box 21.6.3, and causes of acute deterioration of chronic renal failure are shown in Box 21.6.4. Clinical assessment Personal history When taking the personal history, it is important to ask about nocturia (which may be the first, although nonspecific, sign of renal disease), foaming of the urine, past or present periorbital or pedal oedema, episodes of macroscopic haematuria, and whether or not urinary abnormalities (proteinuria, haematuria) have been detected previously, for example, in medical examinations performed before military service or for occupational or insurance purposes. The in- formation that the patient was told that they had ‘a bit of protein/​ blood in the urine’ many years ago (and almost certainly ‘not to Box 21.6.2  Modifiable and nonmodifiable predictors of the rate of progression of CKD Modifiable predictors • Elevated (systolic) blood pressure (particularly during the night) • Activation of the RAS • Proteinuria • Smoking • Salt intake • Obesity/​metabolic syndrome • Protein intake? Nonmodifiable predictors • Age • Sex • History of pre-​eclampsia • Ethnicity • Genetics (family history—​renal and cardiovascular) • Type of renal disease • Prenatal programming (e.g. maternal hyperglycaemia, malnutrition, pre-​eclampsia, low birthweight) Box 21.6.3  Indications of chronic renal failure History More than 6 months of ill health, long-​standing hypertension, protein- uria, nocturia for more than 6 months, sexual dysfunction, abnormalities previously detected during routine medicals and/​or pregnancies, recur- rent illness during childhood. Examination Pallor, pigmentation, pruritus, brown nails, evidence of long-​standing hypertension; the patient often appears ‘well’ for their very abnormal biochemistry. Investigations Normochromic anaemia, small kidneys on ultrasound examination (except in adult polycystic kidney disease, myeloma, amyloid), renal osteodystrophy on radiography.

21.6  Chronic kidney disease 4839 worry about it’) clearly indicates long-​standing renal pathology in the context of a patient presenting with renal impairment. In women, information on proteinuria in pregnancy or pre-​ eclampsia is important. A history of pre-​eclampsia greatly increases the risk of CKD later in life, and pregnancy often leads to the first manifestation or aggravation of pre-​existing renal disease. In elderly men, a history of prostatic disease or related lower urinary tract symptoms (problems with bladder emptying, dribbling, etc.) should raise the suspicion of obstructive nephropathy. Also important when evaluating a patient with CKD are episodes of urinary tract infection before and after puberty, as well as a history of urolithiasis. A history of urinary stones may point to a urological cause of renal failure, and may also very occasionally be an indica- tion of hyperparathyroidism. In proteinuric patients with CKD of unknown origin, one should enquire for a history of chronic bacterial infection (e.g. bronchiec- tasis or osteomyelitis) and for a history of rheumatoid arthritis, all of which are potential causes of secondary amyloidosis. Particularly in older people with CKD of unknown origin, it is also important to ask about symptoms that might indicate multiple myeloma, which can cause renal impairment via a number of mechanisms (see Chapter 21.10.5). The patient should be asked about the use of potentially nephro- toxic drugs, such as NSAIDs, analgesics (if taken regularly in high dose for prolonged periods, particularly compound preparations or those containing phenacetin, which has now been banned in many countries), nephrotoxic antibiotics (e.g. aminoglycosides), antineoplastic agents (cis-​platinum), and herbal/​alternative/​homeo- pathic treatments (e.g. Chinese herbs; see Chapter 21.9.2). Covert drug intake, including laxatives and diuretics with or without sur- reptitious vomiting, may also cause CKD. Since many drugs are ex- creted via the kidney, it is also important to ask patients with CKD for a history of drug side effects. Women with CKD may have menorrhagia or (in advanced CKD) amenorrhoea, and men with CKD not infrequently have erectile failure/​impotence, but few will volunteer this information and it will only be discovered by the physician who asks directly. Symptoms of advanced renal failure (uraemia) With advancing renal failure to CKD stage 5, patients may develop anaemia and fluid retention manifested as breathlessness on exer- cise or even at rest, particularly at night-​time, and they may notice swelling of their ankles and legs. The first symptoms pointing to the uraemic syndrome are anorexia with attendant loss of body weight and nausea and vomiting, frequently in the morning when brushing the teeth. Further symptoms are superimposed if a patient enters the preterminal phase of uraemia, including severe dyspnoea as the combined result of metabolic acidosis and pulmonary congestion, pericarditic and pleuritic chest pains, bleeding tendency (bleeding mouth ulcers, gastrointestinal haemorrhage), numbness of the feet (polyneuropathy), and insomnia, headache, myoclonic jerks, and impairment of consciousness/​coma. Family history It is important to ask for a family history of renal disease, the most common causes of familial nephropathy being autosomal dominant polycystic kidney disease and reflux nephropathy; see Chapter 21.12 for further discussion. If the patient has a history of hypertension, then concomitant proteinuria or diabetes should be sought, because very often these will be undiagnosed. Patients with type 2 diabetes who have a family history of diabetic nephropathy are at higher risk of nephropathy themselves. A patient’s risk of progressive kidney disease is greater if a first-​ degree relative has had any kind of CKD, even if this was not caused by the same kidney disease as that in the patient, suggesting that hereditary mechanisms are involved in the genesis of progression. Furthermore, the renal risk is higher if a first-​degree relative has had essential hypertension or cardiovascular events. Physical examination Except in areas of the world with very poor access to medical services, or in individuals who are severely neglected, the physician will rarely see the desperate case of terminal uraemia with coma, blindness from retinopathy, pericardial rub from uraemic pericar- ditis, massive gastrointestinal ulceration and bleeding, and urea frost covering the face. Patients with CKD stage 5 who are typically seen in Western countries may be pale (anaemic), have severe hyper- tension, and have fluid retention manifested as peripheral oedema and crackles/​effusions in the lungs. As uraemia progresses, patients usually develop progressive hyperpigmentation of the skin, pruritus with excoriations and prurigo, xerosis, brown nails, and evidence of peripheral polyneuropathy. Box 21.6.4  Causes of acute deterioration in patients with chronic renal failure • Renal hypoperfusion:

—​ Dehydration—​diarrhoea, vomiting, excessive diuretics, inadequate fluid replacement (e.g. postsurgical)

—​ Cardiac failure

—​ Drugs—​especially ACE inhibitors, angiotensin receptor blockers, NSAIDs

—​ Systemic infection

—​ Renal vascular disease

—​ Pericardial tamponade (rare) • Obstruction and infection of the urinary tract:

—​ Benign prostatic hypertrophy

—​ Urinary stones

—​ Cancer—​particularly of prostate or bladder

—​ Clot in the ureter

—​ Papillary necrosis and sloughing—​to be considered in patients with diabetes, sickle cell disease, and analgesic nephropathy.

—​ Polycystic cysts (rare) • Metabolic and toxic:

—​ Hypercalcaemia

—​ Hyperuricaemia

—​ Contrast media—​especially in diabetes

—​ Drugs—​especially aminoglycosides • Progression/​relapse of underlying diseases:

—​ Various nephritides and autoimmune/​vasculitic conditions—​look for an active urinary sediment with proteinuria, haematuria, and cellular casts, and also for serological evidence of disease activity • Development of accelerated-​phase hypertension • Renal vein thrombosis—​usually in severely nephrotic patients • Pregnancy: • At the end of the pregnancy or after delivery (e.g. in patients with reflux nephropathy)

section 21  Disorders of the kidney and urinary tract 4840 Further signs are superimposed if a patient enters the preterminal phase of uraemia, including severe tachypnoea (before terminal cessation of breathing) with an acidotic sighing respiratory pat- tern and/​or widespread crackles of pulmonary oedema, jerking movements (metabolic asterixis/​flap/​twitching), pericarditic and/​ or pleuritic rubs, impairment of consciousness/​coma, red eyes, and evidence of bleeding (from mouth or rectum). In routine outpatient practice, all patients presenting with CKD require a thorough general physical examination with particular at- tention to the features described in Table 21.6.4. Rarely there will be manifestations of systemic or genetic disorders associated with renal disease (see Chapters 21.10.1–​21.10.10 and 21.12). Investigation Investigations required in the assessment of patients with CKD are shown in Table 21.6.5. See Chapter 21.4 for further discussion. Management to prevent progression
of chronic kidney disease In some patients, measures to conserve renal function may be spe- cific to the cause of renal impairment, for example, relief of ob- struction, cessation of nephrotoxic drugs (e.g. lithium, ciclosporin, or NSAIDs), relief of renal artery stenosis, and treatment of active autoimmune/​vasculitic disorders. For further details, see the rele- vant specific chapters. However, the mechanisms of progression de- scribed briefly earlier are common to all forms of CKD, and insight into them has provided effective strategies aimed at slowing or—​ when started early—​even halting progression of CKD. Since not all patients with primary kidney disease inevitably progress to endstage renal disease, it is important to identify factors predisposing to an adverse renal outcome in individual patients—​the main ones being high blood pressure (particularly at night-​time), proteinuria, and reduced baseline eGFR—​and to direct particular attention towards such cases. It is recommended that all patients with CKD stages 4 and 5 (also those with CKD stage 3 whose GFR is deteriorating rapidly—​ meaning by >5 ml/​min per 1.73 m2 per year—​or who have signifi- cant proteinuria/​haematuria), excepting those who are terminally ill or for whom it would otherwise be inappropriate, are referred to and managed by or in conjunction with a nephrologist. This is important because appropriate timely intervention can reduce the rate of progression and also because the results of eventual dialysis and/​or transplantation are better if the patient has been properly prepared (including blood pressure normalization, correction of anaemia, hepatitis B vaccination, creation of vascular access, and the search for a living kidney donor that might render pre-​emptive transplantation possible). Interventions to prevent progression of CKD should be initiated as early as possible, which justifies efforts to diagnose renal dis- ease sooner rather than later. The best evidence comes from studies of patients with diabetes. Controlled trials in advanced diabetic nephropathy achieved reduction of progression by no more than about 30%. By contrast, in diabetic patients with early nephrop- athy, mostly at the stage of microalbuminuria, the DETAIL study had achieved by its fifth year a rate of loss of GFR that was virtually indistinguishable from that seen with advancing age. Even if pro- gression is not completely halted, but only attenuated, the gain in years free of RRT is substantially greater if treatment is started early rather than late. Blood pressure control Although the adverse effects of high blood pressure on progres- sion of renal disease had been recognized for many years, the first solid documentation of the importance of achieved blood pressure Table 21.6.4  Features to look for on physical examination of patients with CKD System Feature Comment General Pallor Anaemia is a common feature of advanced CKD Pigmentation, scratch marks, brown nails Indicate long-​standing CKD Cardiovascular/​respiratory Blood pressure Hypertension is a common association (but uncommon cause) of CKD Elevated jugular venous pressure, enlarged heart, gallop rhythm, mitral regurgitant murmur, pulmonary crackles, peripheral oedema Manifestations of fluid overload caused by CKD and/​or cardiac failure associated with hypertension or ‘uraemia’ Aortic systolic murmur Valvular calcification is more common in CKD Peripheral pulses, vascular bruits Absence of pulses and/​or presence of bruits increase the likelihood of renovascular disease as the cause of CKD Abdominal Palpable kidneys Likely to indicate adult polycystic kidney disease in this context Palpable bladder, malignant-​feeling prostate on digital examination Consider urinary obstruction as cause of CKD Hernias Require repair if peritoneal dialysis otherwise preferred as RRT Neurological Peripheral neuropathy Indicates long-​standing CKD Proximal myopathy Indicates long-​standing CKD Rheumatological Arthritis Carpal tunnel syndrome Gout and pseudogout are associated with CKD Ocular fundi Features of hypertension CKD, chronic kidney disease; RRT, renal replacement therapy.

21.6  Chronic kidney disease 4841 Table 21.6.5  Investigations required in the assessment of patients with CKD Type of investigation/​test Comment All patients Urine Dipstick Screening for proteinuria, haematuria, glycosuria, nitrite, pyuria. Heavy proteinuria suggests glomerular disease Microscopy If abnormality on stick testing to detect bacteriuria, quantitate pyuria, look for cellular casts (indicative of renal inflammatory process) Albumin:creatinine ratio (ACR) To quantitate proteinuria Culture To detect urinary tract infection. For tuberculosis if clinically indicated (sterile pyuria) Blood Creatinine To quantitate level of renal function by eGFR Electrolytes Hyperkalaemia is the obvious concern Urea Elevated out of proportion to creatinine in dehydration and catabolism Glucose Exclusion of diabetes mellitus Lipids Assessment of cardiovascular risk Full blood count To detect anaemia Liver and bone profile Routine screen Cardiac assessment ECG Look for changes of ischaemia and/​or left ventricular hypertrophy. Changes in morphology due to hyperkalaemia indicate the level of risk of cardiac arrhythmia and dictate treatment (see Chapter 21.5) Chest radiograph Assessment of heart size and evidence of fluid overload (pulmonary oedema, pleural effusion) Echocardiogram To assess left ventricular function. To exclude pericardial effusion when clinically appropriate Tests for coronary heart disease—​exercise tolerance test, radionuclide stress test, coronary angiography If symptomatic. To assess risk of and fitness for renal transplantation Renal imaging Ultrasonography of urinary tract Expect to find two small and echogenic kidneys in CKD; exclude urinary obstruction; diagnose presence of renal cysts In selected patients to diagnose particular causes of CKD Urine Bence Jones proteinuria To detect myeloma when clinically appropriate Blood Serum protein electrophoresis or free light chains To detect myeloma when clinically appropriate Autoimmune/​vasculitis screen—​ANA, complement, ANCA, anti-​GBM, cryoglobulins To detect autoimmune (systemic lupus erythematosus) and vasculitic (granulomatosis with polyangiitis, microscopic polyangiitis) conditions, also Goodpasture’s disease and cryoglobulinaemia, when clinically relevant Renal imaging Plain abdominal film (KUB) To look for urinary stones and/​or renal calcification CT scan To determine cause of urinary obstruction Renal angiography (direct, CT or MR) If renal artery stenosis is likely DMSA/​MAG3 scan To detect renal scarring Renal biopsy To be performed only in cases where there is diagnostic doubt and there is a reasonable likelihood that the result will affect the management of the particular patient. Should not be done otherwise In patients with significantly impaired renal function (CKD stages 3B, 4, or 5) to look for complications of renal impairment and/​or guide management Blood PTH To diagnose hyperparathyroidism and monitor response to treatment Bicarbonate To detect acidosis and monitor response to treatment Uric acid In patients with gout associated with CKD Haematinics—​iron studies, vitamin B12, folate In patients with anaemia. Optimal response to erythropoietin (see later) requires plentiful iron stores (high normal/​high ferritin)

section 21  Disorders of the kidney and urinary tract 4842 on progression was provided only in 1989, when a study of patients of the Kaiser Permanente cohort in the United States of America who had normal baseline urinary findings found that even within the range of normotensive values the risk of reaching endstage renal disease increased progressively with progressively higher blood pressure values at baseline (more so in patients with than without diabetes). There is limited information on the optimal target blood pres- sure and the role of different antihypertensive agents. One recent high-​quality study in African American patients with hypertensive kidney disease found no benefit of intensive (mean 130/​78 mmHg) compared to standard (mean 141/​86 mmHg) blood pressure con- trol in preventing decline of renal function, excepting in those who had significant proteinuria. Nevertheless, the consensus remains that the lower the blood pressure, the better, as long as this can be achieved without unacceptable side effects. The 2013 guidelines of the European Society of Hypertension/​European Society of Cardiology on treatment of hypertension recommend ‘in patients with diabetic or non-​diabetic renal disease, systolic blood pressure should be lowered to less than 140 mmHg and when overt protein- uria is present values less than 130 mmHg may be pursued, pro- vided that changes in eGFR are monitored’. To achieve the blood pressure goal, combination therapy of several antihypertensive agents (including loop diuretics when the eGFR is <30 ml/​min per 1.73 m2) is usually required, with an ACE inhibitor or angio- tensin receptor blocker used whenever possible to reduce protein- uria (but a combination of these two classes of drug is no longer recommended). It obviously needs to be considered whether such aggressive blood pressure lowering is safe. No excess of cardiovascular events or of mortality has been reported in nondiabetic patients and patients without coronary disease, although in older people limited toler- ance and safety (orthostatic hypotension, falls) of intensive blood pressure lowering have to be taken into account. By contrast, higher mortality in patients with advanced diabetic nephropathy has been found with an achieved systolic blood pressure below 120 mmHg, and in nonrenal patients with coronary heart disease the rate of myo- cardial infarction has been reported to be increased if diastolic blood pressure is lowered to below 70 mm/​Hg, hence caution in pursuit of very low blood pressure targets is indicated in populations at high cardiovascular risk, including those with CKD. The most important predictor of progressive loss of renal func- tion is the systolic blood pressure, not the mean or the diastolic blood pressure. Of particular concern in renal patients is the ten- dency for an attenuated decrease of blood pressure—​or even a para- doxical increase—​at night-​time, recognition of which requires the use of ambulatory blood pressure measurements. Because of the complexity of blood pressure analysis and the limited accuracy of office blood pressure measurements, it is bad advice to stick only to fixed blood pressure targets—​particularly since almost all studies have shown that the proportion of renal patients reaching target blood pressure values is disappointingly low, despite the use of multidrug regimens. Selection of antihypertensive agents There is good evidence from experimental studies that activation of the RAS and of the sympathetic nervous system play a major causal role in the genesis of progression. There has been some controversy over whether the effect of antihypertensive agents interfering with the activity of the RAS is exclusively explained by lowering of blood pressure. Considering controlled interven- tion trials where similar blood pressure lowering was achieved with antihypertensive agents that did and did not block the RAS, it is obvious that a definite but limited contribution to the slowing of progression is achieved by RAS blockade. An example of the finding of such a trial is shown in Fig. 21.6.4. A head-​to-​head comparison, although somewhat underpow- ered, suggested that the effect of ACE inhibition and angiotensin re- ceptor blockade is comparable in magnitude with respect to halting progression of renal disease. The blockade of the RAS is effective and safe, if properly supervised, but the risk of hyperkalaemia has to be taken into consideration, which is caused most frequently by excessive potassium intake and prescription of other drugs interfering with potassium excretion (β-​blockers, potassium-​ sparing diuretics). Patients with CKD are at high cardiovascular risk and, aside from renal benefit, ACE inhibition (if tolerated) has been shown in some trials to reduce cardiovascular mortality more effectively in those with impaired renal function than in those with normal renal function. Reduction of proteinuria Proteinuria is not only a marker, but also an extremely important promoter of CKD progression, and it has therefore become an im- portant additional treatment target. In patients with primary renal disease, the higher the baseline proteinuria, the greater the subse- quent decline in GFR, and also the greater the risk of cardiovascular events. Even a very modest increase in proteinuria within the normal range is associated with increased mortality. A recent meta-​analysis of data collected over more than 700 000 patient years reported that compared with an ACR of 0.6 mg/​mmol (low normal range), an ACR of 1.1 mg/​mmol was associated with a mortality hazard ratio of 1.20, and an ACR of 3.4 mg/​mmol (upper limit of normal range for men) with a mortality hazard ratio of 1.63. Type of investigation/​test Comment Virology—​screen for hepatitis C, hepatitis B, HIV Relevant to RRT with dialysis or by transplantation Virology—​screen for CMV, HSV, HZV, EBV Relevant to RRT by transplantation Blood group Relevant to RRT by transplantation ANA, antinuclear antibodies; ANCA, antineutrophil cytoplasmic antibodies; anti-​GBM, antiglomerular basement membrane antibody; CKD, chronic kidney disease; CMV, cytomegalovirus; DMSA, dimercaptosuccinic acid; EBV, Epstein–​Barr virus; ECG, electrocardiogram; HSV, herpes simplex virus; HZV, herpes zoster virus; KUB, kidneys, ureter, and bladder; MAG3, mercaptoacetyltriglycine; PTH, parathyroid hormone; RRT, renal replacement therapy. Table 21.6.5  Continued

21.6  Chronic kidney disease 4843 Current guidelines recommend that proteinuria should be lowered as much as possible, preferably to less than 1 g/​24 h (roughly equivalent to an ACR <60 mg/​mmol or a protein:creatinine ratio <100 mg/​mmol). When this goal is not achieved, it is wise to check whether modifiable factors explain the limited efficacy of blood pres- sure lowering with RAS blockers, such as patient noncompliance, high sodium intake (which abrogates the antiproteinuric action of RAS blockade), high protein intake (which increases proteinuria), smoking, and poor glycaemic control in people with diabetes. If these factors have been excluded, there are two strategies. First, one can attempt dose escalation of the ACE inhibitor or angiotensin re- ceptor blocker above the dose licensed for blood pressure lowering. This strategy has been shown to be effective in slowing the progres- sion of CKD, with angiotensin receptor blockers often preferred in this situation because of their better side effect profile. Second, consider blockade of the mineralocorticoid receptor: aldosterone levels decrease immediately after the start of RAS blockade, but in many patients they subsequently rise again, a phenomenon known as ‘aldosterone escape’, which can be accompanied by failure to restrain proteinuria. Several studies have shown that, without further lowering of blood pressure, blockade of the mineralo- corticoid receptor by low doses of spironolactone (25 mg/​day) or eplerenone can prevent such relapse. There is, however, a danger of precipitating hyperkalaemia, a risk that is justifiable only in patients who have received and accept appropriate dietary education and agree to close follow-​up. Giving a combination of an ACE inhibi- tors and an angiotensin receptor blocker is no longer recommended following studies that demonstrated risks including worsening of kidney function. Other interventions Dietary protein restriction Reducing dietary protein intake may protect against progression of CKD by haemodynamically mediated reduction in intraglomerular pressure, by changes in cytokine profile, and/​or by changes in matrix synthesis in the renal interstitium. It has been shown to be effective in attenuating progression of renal failure in several well-​conducted animal studies. In clinical trials, numerous uncon- trolled studies have also shown an apparently beneficial effect of low dietary protein intake on CKD progression, but the only major controlled prospective trial (the MDRD study) failed to docu- ment a significant effect. Although subsequent post hoc analyses were consistent with some benefit from dietary protein restriction, its magnitude does not compare with what can be achieved by lowering blood pressure and RAS blockade, and it is important to recognize that drastic lowering of protein intake carries the risk of malnutrition. In routine practice, most nephrologists do no more than to advise against high-​protein diets, recommending limita- tion of protein (but not energy) intake only in CKD patients who have a gross excess of protein intake, for example, more than 2 g/​kg body weight per day, documented by urinary urea measurements. Those of a more interventionist persuasion may recommend a daily protein intake of 0.8 to 1 g/​kg per day (and some even 0.7 g/​ kg per day), with careful monitoring to ensure that protein mal- nutrition does not develop and that there is an adequate intake of calories. Other dietary and lifestyle measures As shown in Box 21.6.2, other modifiable progression promoters are smoking, high salt intake, obesity, and the metabolic syn- drome. In diabetic patients, the risk of onset of microalbuminuria, and the progression from microalbuminuria to overt diabetic nephropathy, is higher in smokers, as is the rate of loss of GFR in smokers with advanced diabetic nephropathy. At least in dia- betic patients, some evidence suggests that cessation of smoking slows the progression of CKD. Thus, in addition to the aim of re- ducing cardiovascular risk, this finding is another strong reason for patients with renal disease to stop smoking. Recent evidence even indicates that passive smoking increases albuminuria and is injurious to the kidney. There is scant evidence in humans, in contrast to convincing evi- dence in animals, for an adverse effect of high salt intake on the progression of primary renal diseases. It is, however, important to recognize that high salt intake militates against control of hyper- tension and interferes with the antiproteinuric action of ACE inhibition, hence it seems sensible—​in line with general recom- mendations accepted for the treatment of hypertension—​to advise a reduction in sodium intake from the usual (in developed countries) 150 to 200 mmol/​day to 100 mmol/​day (6 g of sodium chloride). In practice, this means the patient reducing the amount of processed food that they consume. A further risk factor for the onset and promotion of CKD progres- sion is visceral obesity and the metabolic syndrome. As shown in Fig. 21.6.5, the worse the metabolic syndrome, the greater the preva- lence of CKD and of microalbuminuria in the general (American) population, and the degree of obesity at which renal risk increases Placebo Treatment < 134 < 149 134–140 Quartile of ave systolic BP 141–149 Irbesartan Amlodipine Relative risk of renal endpoint 0.600 1.200 1.800 Fig. 21.6.4  The contribution of achieved blood pressure (BP) lowering and renin–​angiotensin system (RAS) inhibition on the relative risk of reaching a renal endpoint (doubling of serum creatinine or endstage renal disease). At any given blood pressure level, patients on irbesartan (an angiotensin receptor blocker) had a lower renal risk than those on amlodipine (which does not block the RAS). Reproduced with permission from Pohl MA, et al. (2005). Independent and additive impact of blood pressure control and angiotensin-​II receptor blockade on renal outcomes in the irbesartan diabetic nephropathy trial: clinical implications and limitations. J Am Soc Nephrol, 16, 3027–​37. Copyright © 2005 American Society of Nephrology.

section 21  Disorders of the kidney and urinary tract 4844 is surprisingly low (Fig. 21.6.6). Overweight patients with early CKD should therefore try to reduce their body weight, since this has proven to be effective in several studies. However, in more ad- vanced stages of CKD—​namely at an eGFR of approximately 30 to 50 ml/​min—​the situation is more complex, with higher body weight (even in the range of morbid obesity) lowering the risk of death for unknown reasons. Other pharmacological interventions Other interventions in addition to ACE inhibition or angiotensin receptor blockade to attenuate CKD progression that are under evaluation include renin inhibitors, endothelin I  blockers, glyco- saminoglycans (increasing the electronegativity of the charge of the glomerular basement membrane), erythropoietin (effective in experimental studies), NSAIDs (potentially dangerous because of overshooting reduction of GFR), vitamin D analogues, and statins. Cochrane meta-​analysis data suggests the use of statins in CKD may reduce the risk of death and cardiovascular events in nondialysis pa- tients. Evidence in support of statin therapy delaying progression of CKD is not yet convincing although recent data suggest high-​ intensity statins may have benefit. The latest KDIGO guidelines on lipid management now recommend use of statins in all nondialysis patients with CKD over the age of 50, and treatment for younger patients with coronary disease, diabetes, previous stroke, or risk of cardiovascular events greater than 10% over 10 years. Medical management of the consequences of chronic kidney disease Water, electrolytes, acidosis, and nutrition Only patients with oliguric endstage renal failure need to restrict their fluid intake precisely, when the usual recommendation is that the patient’s daily intake should be a volume equal to their daily urinary output plus 500 ml for insensible losses. This is difficult to adhere to unless sodium intake is carefully discussed and min- imized, and a diet with ‘no added salt’ is essential. However, most   Prevalence CKD (%) Prevalence Microalbuminuria (%) Metabolic syndrome risk factors Metabolic syndrome risk factors 0 0 5 10 20 15 2 4 6 8 10 0 1 2 3 4 5 0 1 2 3 4 5 Fig. 21.6.5  The prevalence of CKD (GFR <60 ml/​min per 1.73 m2) and of microalbuminuria (ACR 30–​300 mg/​g, approximately 3–​30 mg/​mmol) in relation to the number of metabolic risk factors that a patient has. The five metabolic risk factors considered were (1) waist circumference greater than 102 cm (men); (2) fasting glucose greater than 110 mg/​dl (>6.1 mmol/​litre); (3) high-​density lipoprotein cholesterol (HDL-​C) less than 40 mg/​dl (<1.03 mmol/​litre); (4) triglycerides greater than 140 mg/​dl (>1.58  mmol/​litre); and (5) blood pressure greater than 130/​80 mmHg. Reproduced with permission from Chen et al. Ann Int Med 2004; 140; 167–​74. http://​www.annals.org/​content/​vol140/​issue3/​ Copyright © 2004, The American College of Physicians. BMI (kg/m2) <18.5

40.0 35.0–39.9 18.5–24.9 25.0–29.9 RR 0.44 1.00 1.87 3.57 6.12 7.07 Adjusted relative risk ESRD 0.1 1.0 10.0 30.0–34.9 Fig. 21.6.6  Body mass index (BMI) is an independent predictor of endstage renal disease in the United States of America. Reproduced with permission from Hsu et al. Ann Int Med 2006; 144: 21–​28 http://​www.annals.org/​cgi/​content/​ full/​144/​1/​21 Copyright © 2006, The American College of Physicians.

21.6  Chronic kidney disease 4845 patients with CKD pass a normal volume of urine, but they need to avoid binge drinking because their ability to excrete free water is impaired. They should be aware that they will need to drink more if they have other significant fluid losses, for example, vomiting, diarrhoea, or excessive sweating, because their kidneys will not be able to produce appropriately concentrated urine. They should be informed of the importance of temporarily stopping diuretics and possibly antihypertensive medication if increased oral intake is not possible during intercurrent illness to minimize the likelihood of developing acute (on chronic) kidney injury. Dietary sodium restriction (100 mmol/​day) is important to min- imize thirst and fluid retention in patients with CKD, with no salt added to food after cooking. Reduction of sodium intake to around 60 mmol/​day can be helpful in patients with fluid retention in the context of advanced CKD, but many patients find this unacceptable. Using loop diuretics to encourage sodium excretion is more effective in most cases, but excessive salt intake will negate the diuretic effect. Hyperkalaemia is most commonly seen in the context of renal failure (acute or chronic) and can be life-​threatening. Emergency management is discussed in Chapter 21.5, but an important aim of medical management of patients with CKD is to avoid such prob- lems. Monitoring of the serum potassium must be routine whenever the creatinine is checked, and especially after introduction of drugs that are known to cause hyperkalaemia (including ACE inhibitors, angiotensin receptor blockers, and potassium-​sparing diuretics). The patient should be offered dietary advice (see Chapter 21.2.2) if serum potassium is found to be in the range 5.5 to 6.5 mmol/​litre, and the measurement should be repeated a few days later (unless they are known to have a stable potassium in this range). Dietary advice combined with stopping of all medications that might pre- cipitate hyperkalaemia is appropriate if potassium is in the range 6.5 to 7 mmol/​litre, again with repeated measurement a few days later. If the potassium is above 7 mmol/​litre, the patient should be reviewed in hospital, with checking of the ECG for hyperkalaemic mani- festations being an immediate priority, followed by consideration of possible precipitants, which include dietary indiscretion (fruit, chocolate, coffee) and gastrointestinal haemorrhage as well as more obvious intercurrent illness. Chronic acidosis with serum bicarbonate in the range 12 to 20 mmol/​litre is most common in patients with interstitial renal dis- ease and will aggravate hyperkalaemia, inhibit protein anabolism, and accelerate calcium and phosphate loss from bone. Treatment with alkali to maintain the serum bicarbonate above 22 mmol/​litre is recommended. Sodium bicarbonate (1.0–​2.0 g twice day) is the first-​line treatment, with sodium citrate an alternative for those who cannot tolerate bicarbonate (usually because of abdominal bloating) as long as they are not taking aluminium-​containing antacids (cit- rate enhances aluminium absorption). Malnutrition is common in advanced chronic renal failure because of anorexia, impaired gut function, and acidosis. The most practical ways of detecting its insidious development is by serial monitoring of body weight and serum albumin concentration. Standard advice is that patients with CKD should have a diet containing about 30 to 35 kcal/​kg body weight per day. Dietary supplements may be helpful in achieving this, but they are not a cure for the problem: a patient that is becoming malnourished probably needs to start RRT sooner rather than later. Mineral and bone disorder The mineral and bone disorder (MBD) associated with CKD is a major cause of disability in patients with endstage renal failure (Box 21.6.5). The term was coined to describe the threefold elements of biochemical, vascular, and skeletal abnormalities which develop with progressive CKD. The term renal osteodystrophy refers solely to the skeletal abnormalities, but it has become clear that the bio- chemical, skeletal, and vascular problems are all intimately linked. This is mainly, but not exclusively, the consequence of abnormalities in the metabolism of calcium, phosphorus, PTH, and vitamin D, and also the more recently discovered phosphaturic hormone FGF23, produced by osteocytes in bone. The biochemical and metabolic abnormalities that make up CKD-​MBD are complex and incompletely understood, but begin to develop as early as CKD stage 3. The initial sequence of events is unclear but involves progressive reduction in the kidneys’ ability to excrete phosphate, to produce the active form of vitamin D3 (1,25-​ (OH)2D3 or calcitriol), to maintain normal calcium levels, and therefore to prevent PTH rising. Consequently, secondary hyper- parathyroidism is common in unmanaged CKD patients, but now that oral vitamin D replacement, phosphate binders, and calcium supplements are widely prescribed to control parathyroid over- activity, patients with advanced stages of CKD are now more likely to develop relative hypoparathyroidism. This results in low bone turnover (adynamic bone on iliac crest biopsy) in which the skeleton is unable to absorb excess calcium and phosphate resulting in epi- sodes of hypercalcaemia, hyperphosphataemia, and an association with increased likelihood of vascular calcification for reasons which are uncertain. A degree of controlled secondary hyperparathyroidism appears to be required to maintain normal bone turnover, whilst trying to pre- vent PTH rising beyond nine times the upper limit of normal (ULN), which is associated with an increased risk of arterial and valvular calcification, and increased cardiovascular and all-​cause mor- tality. Current guidelines (based on very limited evidence) suggest maintaining serum PTH in the range between two and nine times the ULN, which is a difficult task attempted by control of serum cal- cium, phosphate, and use of oral vitamin D supplements, with use of an oral calcimimetic (cinacalcet) if PTH appears to be rising uncon- trollably. This drug blocks production of PTH by mimicking a state of extreme hypercalcaemia at the calcium-​sensing receptor (CaR) on parathyroid cells. Box 21.6.5  Definition of CKD–​MBD A systemic disorder of mineral and bone metabolism due to CKD mani- fested by either one or a combination of the following: • Abnormalities of calcium, phosphorus, PTH, or vitamin D metabolism • Abnormalities in bone turnover, mineralization, volume, linear growth, or strength • Vascular or other soft-​tissue calcification Definition of renal osteodystrophy • Renal osteodystrophy is an alteration of bone morphology in patients with CKD • It is one measure of the skeletal component of the systemic disorder of CKD–​MBD that is quantifiable by histomorphometry of bone biopsy.

section 21  Disorders of the kidney and urinary tract 4846 Pathogenesis Phosphate excess In the early stages of CKD, the plasma phosphate concentration re- mains normal or may be even low, which is achieved at the price of increased fractional clearance of phosphate. Hyperphosphataemia begins to develop once the GFR has decreased to between 60 and 30 ml/​min per 1.73 m2 (CKD stage 3). Hyperphosphataemia ag- gravates secondary hyperparathyroidism by indirect mechanisms, such as inhibition of the synthesis of the active vitamin D metabolite 1,25 dihydroxyvitamin D3 (1,25-​(OH)2D3 or calcitriol) in tubular epithelial cells, and possibly also by inducing a tendency for hypo- calcaemia. Phosphate also directly stimulates PTH synthesis and se- cretion as well as parathyroid cell proliferation, independent of low serum 1,25-​(OH)2D3 and hypocalcaemia. Increased PTH secretion reduces tubular phosphate reabsorption and therefore increases urinary phosphate excretion. In parallel, plasma FGF23 increases as it attempts to increase phosphate excretion, but as renal impairment progresses, this eventually becomes an impossible task because the feedback loop is broken, so that FGF23 and PTH rise uncontrollably. The relative roles of PTH and FGF23 in the control of phosphate re- absorption are currently unclear. 1,25-​dihydroxyvitamin D3 deficiency The hepatic vitamin D metabolite, 25-​hydroxyvitamin D3 (25-​(OH) D3 or cholecalciferol), is further hydroxylated, mainly in renal tubular epithelial cells, to the most active vitamin D metabolite, 1,25-​dihydroxyvitamin D3 (1,25-​(OH)2D3 or calcitriol), which acts as a circulating hormone (apart from paracrine actions of locally synthesized 1,25-​dihydroxyvitamin D3 in tissues such as activated macrophages, vascular cells, and others). Its synthesis is stimulated by PTH but inhibited by phosphate and FGF23. Even in early stages of CKD there is a tendency for 1,25-​dihydroxyvitamin D3 concen- tration to decrease, although this is counteracted to some extent by simultaneous increases in PTH. Nevertheless, average concentra- tion of 1,25-​dihydroxyvitamin D3 tends to fall progressively, unless treated, as CKD advances (Fig. 21.6.7). The renal 1α-​hydroxylase reaction is normally substrate inde- pendent, but with progression of CKD it becomes increasingly dependent on the availability of the substrate 25-​hydroxyvitamin D3, with deficiency of this form of vitamin D further aggravating the impairment of synthesis of 1,25-​dihydroxyvitamin D3. For this reason some renal units prescribe weekly oral 25-​hydroxyvitamin D3 supplements (cholecalciferol) routinely to all their CKD stage 5 and stage 5d patients. 1,25-​dihydroxyvitamin D3 increases calcium (and phosphate) uptake from the gastrointestinal tract and reduces calcium loss in urine by increasing renal tubular calcium reabsorption, both effects raising serum calcium levels. In times of dietary calcium deficiency it can also stimulate calcium release from bone to correct hypocal- caemia. In patients with heavy proteinuria or nephrotic syndrome, circulating vitamin D metabolites bound to vitamin D-​binding pro- tein may be lost in the urine, so that deficiency of vitamin D me- tabolites, such as 25-​hydroxyvitamin D3 and 1,25-​dihydroxyvitamin D3, may ensue. As also shown in Fig. 21.6.7, average serum levels of PTH increase progressively as average serum 1,25-​dihydroxyvitamin D3 levels de- crease with decreasing GFR in patients with various stages of CKD. Hypocalcaemia On average, plasma (total and ionized) calcium concentrations are maintained in the normal range until CKD stage 5. Nevertheless, in the untreated state there is a tendency towards hypocalcaemia, mainly because of reduced active intestinal calcium absorption as a result of insufficient active vitamin D generation and diminished release of calcium (skeletal resistance) in response to PTH and 1,25-​dihydroxyvitamin D3. Skeletal resistance to normal serum levels of PTH is not under- stood at all, but may well play an important role in the pathogen- esis of secondary hyperparathyroidism, with another mechanism being impaired inhibition of the parathyroid gland, which senses this calcium levels via the CaR, expression of which is decreased GFR(ml/min) 60–90 20–40 Normal 40–60 GFR(ml/min) 60–90 20–40 Normal 40–60 60 50 40 30 20 10 (a) (b) 70 80 1,25 (OH)2D3 (ng/L) 60 50 40 30 20 10 1-84 iPTH (pmol/l) Fig. 21.6.7  1,25-​(OH)2vitamin D3 (a) and intact plasma parathyroid hormone (b) values as a function of glomerular filtration rate (GFR) in patients with various stages of CKD. Reproduced with permission from Reichel et al. (1991). Calcium metabolism in early chronic renal failure: implications for the pathogenesis of hyperparathyroidism. NDT, 6, 162–​9. Copyright © 1991, Oxford University Press.

21.6  Chronic kidney disease 4847 in parathyroid tissue in CKD-​MBD. Some reports have showed abnormal Ca2+ sensing even in early stages of CKD. CaR down-​ regulation in the parathyroid gland can be reversed by a low phos- phate diet, calcimimetics, and other compounds. Types of bone lesions (renal osteodystrophy) in CKD The bone lesions discussed in the following sections, collectively known under the term renal osteodystrophy, are found in the skel- eton of patients with CKD, in isolation or often in combination (Box 21.6.6). They are the result of the variable effect of PTH on uraemic bone, and in particular on the two main types of bone cells—​osteoblasts and osteoclasts. Osteoblasts are responsible for deposition of bone matrix, which is then calcified under the influ- ence of vitamin D3. During this process, some osteoblasts become incorporated into the calcified matrix within lacunae, and become osteocytes able to produce FGF23. Osteoclasts are multinucleate giant cells which are able to resorb mineralized bone, thereby starting a remodelling process. Severe hyperparathyroidism This is characterized histologically by an enormous increase in both osteoclastic bone resorption and osteoblastic bone apposition rates with (1) intense and disorganized remodelling of bone trabeculae in the spongiosa, and (2) uncontrolled resorption and tunnelization of cortical and cancellous bone by groups of osteoclasts, with or without deposition of fibrous tissue (endosteal fibrosis) (Fig. 21.6.8). The de- position of disorganized fibrous tissue gave rise to the descriptive name of osteitis fibrosa, but in reality there is no inflammatory pro- cess at work, just excessive stimulation of bone turnover by very high levels of PTH. The end result of this process is significant weakening of the bone’s mechanical strength. Typical radiological appearances are shown in Fig. 21.6.9, but it should be noted that it is uncommon to see such extreme changes since the introduction of accurate PTH monitoring. Osteomalacia Osteomalacia is characterized by a disparity between the rate of bone matrix synthesis and bone matrix mineralization, leading to widening of the seam of unmineralized bone matrix (osteoid), usually associated with signs of diminished numbers and activities of cells at the bone surface. Pure osteomalacia is rarely seen now- adays: 30 years ago it was mainly due to aluminium intoxication, and before that to overt vitamin D (cholecalciferol) deficiency. Mixed lesions In many patients with advanced stages of CKD, a combination of os- teitis fibrosa and osteomalacia is present, which is commonly called mixed lesions or mixed renal osteodystrophy. This is the commonest lesion found in recent series of bone biopsies, present in up to 50% of prevalent dialysis patients. Adynamic bone In patients with low or normal serum intact PTH concentrations (less than twice the ULN), the number and activity of cells on the bone surface is strikingly reduced, as is bone turnover. This condi- tion is most common in patients with CKD due to diabetes, those with poor nutritional status, and those who have been overtreated with active vitamin D and/​or calcium-​containing phosphate binders. Box 21.6.6  Bone lesions in patients with CKD Bone lesions that may be found either in isolation or in combination • High turnover Severe hyperparathyroidism (PTH usually > 15 × ULN) Moderate hyperparathyroidism (PTH usually 2–​9 × ULN) • Normal turnover Relatively uncommon in CKD stage 5 and stage 5d patients • Mixed lesions High and low turnover lesions evident in same biopsy • Low turnover Adynamic bone, oversuppression of PTH with
calcium/​vitamin D3 Osteomalacia, uncommon since advent of oral vitamin D3 Further pathologies that must be considered in atypical cases • β2-​Microglobulin-​related amyloidosis • Sequelae of preceding corticosteroid therapy—​fractures, femoral osteonecrosis • Osteopenia and osteoporosis, particularly in postmenopausal patients • Reflex sympathetic dystrophy • Bony problems caused by primary disease leading to CKD (e.g. oxalosis) (b) (a) Fig. 21.6.8  Light microscopic images of severe secondary hyperparathyroidism in a patient with CKD stage 5 (transiliac bone biopsy, Masson trichrome stain). (a) Increased osteoclastic bone resorption (arrows) and osteoblastic bone apposition with deposition of fibrous tissue (endosteal fibrosis). Widened osteoid seams (between large arrow heads). (b) Greatly enhanced mineralization activity, as evidenced by double fluorescent tetracycline bands (tetracycline stain). Courtesy of ALM de Francisco MD.

section 21  Disorders of the kidney and urinary tract 4848 It predisposes to hypercalcaemia, hyperphosphataemia, and soft tissue calcification because the capacity of the skeleton to sequester calcium and phosphate is reduced because most bone turnover has ceased. There is some evidence that adynamic bone may contribute to skeletal fractures in patients on dialysis through accumulation of unrepaired microfractures, but it is not known whether there are more far-​reaching clinical implications. Although often referred to as adynamic bone ‘disease’ there is no reason for this lesion to be considered in this way. Osteopenia and osteoporosis Diminished bone mass is possibly more common in patients with CKD stages 4, 5, and 5d than in the general population, and frac- ture rates per 1000 patient years are around four times higher. Furthermore, death rates in the days following a major fracture are at least three times higher for such patients. Contributory factors are underlying renal osteodystrophy, a history of treatment with ster- oids, and (premature) menopause. It is not known whether the risk is aggravated by smoking and low-​calcium diets, or whether it can be prevented by substitution of oestrogens/​gestagens or selective oes- trogen receptor modulation. In CKD stage 3, osteoporosis can be treated in the same way as the general population, including bisphosphonates where neces- sary. However, in CKD stages 5 and 5d it is important to consider the possibility of underlying renal osteodystrophy. In particular, the combination of osteoporosis and adynamic bone could result in bisphosphonates making matters worse, by reducing bone turnover to zero and preventing any improvement. Bisphosphonates prob- ably remain within such bone for many years. Bone biopsy is re- quired before commencing any treatment if there is doubt about the underlying renal osteodystrophy. In CKD stage 4, there are so few published data that treatment re- commendations are not possible, but anecdotal reports of treatment with denosumab are published. Other bone-​related pathologies In patients with CKD, several bone pathologies unrelated to calcium metabolism may coexist with CKD-​MBD (Box 21.6.5). Specifically, dialysis-​associated amyloidosis with preferential osteoarticular involvement, called β2-​microglobulin-​related amyloidosis, must be considered in the differential diagnosis of bone pain and osteoarticular destruction. Symptoms and signs MBD in CKD is usually asymptomatic. Bone pain is not common, even in advanced osteitis fibrosa, although bones subjected to mech- anical stress (spine, calcaneus, foot) may be painful. While fractures are uncommon, skeletal deformity, facial leontiasis, and avulsion of the patella may occur. By contrast, osteomalacia, particularly that secondary to aluminium intoxication, may be very painful, espe- cially when Looser’s zones (fatigue fractures) are present. Again, it is important to exclude alternative causes of bone pain, particularly myeloma and metastases (Box 21.6.7). Severe extraosseous calcifications (Figs. 21.6.9–​21.6.11), specif- ically periarticular, bursal, and visceral calcifications, are usually the consequence of severe hyperphosphataemia and high-​normal serum calcium, with either very high or low serum PTH concentra- tions. Tumoural tissue calcification is often triggered by trauma with haematoma formation and favoured by low bone turnover, which diminishes the capacity of the bone to sequester calcium and phos- phate from the extracellular space. Slowly progressing arterial and cardiac valvular calcifications may be associated with clinical evidence of cardiovascular disease, indeed cardiovascular mortality in dialysis patients is strongly pre- dicted by the presence of coronary artery calcification detected by cardiac CT scanning. (b) (a) Fig. 21.6.9  Radiographs of the hand. (a) Reduced mineral density as well as fluffy and mottled texture of the bones. Note (1) subperiosteal resorption zones at the radial side of the middle phalanges; (2) erosion cavities at the periosteal surface with overlying areas of calcification, so-​called periosteal neostosis; (3) longitudinal striation of cortical bone (corresponding to enlarged Haversian channels); (4) thinning of cortical bone by endosteal bone resorption; (5) loss of the terminal lamella of the terminal phalanx; and (6) vascular calcification above the first digit and along the side of the radius. (b) A detail of the second digit, the terminal phalanx of which has collapsed such that the patient had ‘pseudo-​clubbing’. Box 21.6.7  Differential diagnosis of bone pain in patients with advanced CKD or on dialysis Mineral and bone disorders related to CKD • Severe hyperparathyroidism—​pain relatively rare • Osteomalacia—​secondary to vitamin D deficiency, or exceptionally aluminium accumulation • β2-​Amyloidosis Bone pathologies not directly caused by CKD • Skeletal metastases, myeloma • Osteomyelitis, mostly spondylodiscitis—​may be related to infected vascular access • Neuromelic pain after creation of arteriovenous fistula • Bone infarction, osteonecrosis—​related to corticosteroid treatment, sickle cell anaemia • Osteopenia-​associated infarctions and fractures

21.6  Chronic kidney disease 4849 Calciphylaxis, more correctly called calcific uraemic arteriolopathy, is a rare medical emergency characterized by is- chaemic eschars of the skin secondary to calcification of cuta- neous arterial vessels. Predisposing factors, apart from secondary hyperparathyroidism, include diabetes, obesity, female sex, and (probably) treatment with warfarin. It can produce gangrene and may be fatal in up to 40% of cases. It may respond to tight control of serum phosphate levels with a noncalcium-​containing phos- phate binder, and parathyroidectomy (or cinacalcet treatment) if serum PTH levels are elevated. Biochemical abnormalities While patients with advanced stages of CKD left untreated usu- ally have hypocalcaemia and hyperphosphataemia, patients with advanced secondary hyperparathyroidism are characterized by hypercalcaemia and hyperphosphataemia associated with an in- crease in serum total alkaline phosphatase and its bone-​specific isoenzyme. The findings on serum biochemistry in patients with CKD-​MBD are shown in Table 21.6.6, with the findings in the two main contrasting forms of MBD compared in Table 21.6.7. In pa- tients with hypercalcaemia, it is important to consider causes other than secondary hyperparathyroidism which necessitate specific treatment, as listed in Box 21.6.8. Prophylaxis and management of secondary hyperparathyroidism Secondary hyperparathyroidism is the combined result of failing ex- cretory function of the kidney (leading to phosphate retention) and failing endocrine function of the kidney (leading to calcitriol de- ficiency). Appropriate management requires prevention (whenever possible) and treatment of both abnormalities. Approach to serum phosphate control It is usually recommended that phosphate-​lowering interventions should begin once plasma phosphate concentrations exceed the upper limit of the normal range, that is, 1.45 mmol/​litre. This is generally the case when GFR has decreased to less than 30 ml/​min per 1.73 m2 and persists even when patients with CKD are on dia- lysis, which cannot (on a conventional 3 × 4 h weekly haemodialysis regimen, or by peritoneal dialysis) clear the phosphate consumed in a standard Western diet (50–​100 mmol/​day, of which 50–​70% is absorbed). However, slow continuous therapies (e.g. peritoneal dialysis) will clear on average around 30% more than conventional haemodialysis. Phosphate is present in virtually all foods, hence reduction of dietary intake is difficult without incurring the risk of malnutrition, especially in the elderly. Patients should be reviewed by a renal diet- ician and may need to avoid food items with very high phosphate content (e.g. some dairy products) and those to which phosphate is added, such as sausages and phosphate-​rich soft drinks. A protein-​ restricted diet will reduce phosphate intake, but the merit or other- wise of protein restriction is debatable (see previous discussion). However, given that sufficient dietary restriction of phosphate is usually not feasible, patients with advanced CKD remain in positive phosphate balance unless oral phosphate binders are administered. All oral phosphate binders trap phosphate in the gastric and small intestinal lumen by forming insoluble complexes, hence they must be taken at meal times, after which the phosphate concen- tration in the gut is highest. The agents most commonly used in Europe are calcium carbonate and calcium acetate, with magne- sium carbonate sometimes used if total calcium intake needs to be reduced. Although very effective in controlling serum phos- phate, the prescription of aluminium-​containing compounds has been abandoned by almost all nephrologists because of the risk of aluminium intoxication, except for short-​term use. However, calcium-​containing phosphate binders cause positive calcium balance and perhaps promote vascular calcification, although this remains controversial. Over the past decade, the calcium-​free and Fig. 21.6.10  Calcification of the popliteal artery in a patient with diabetes, CKD, and severe secondary hyperparathyroidism. Fig. 21.6.11  Tumoural calcification around the left shoulder in a chronic haemodialysis patient with severe low-​turnover bone disease.

section 21  Disorders of the kidney and urinary tract 4850 aluminium-​free phosphate binders sevelamer (an anion-​exchange resin) and lanthanum carbonate have been introduced into clinical practice. These allow similar control of plasma phosphate, while avoiding calcium and aluminium overload. Use of these binders may be associated with reduced progression of vascular calcifica- tion in animal models and a few small randomized trials. However, no randomized trial has shown that selecting a particular phos- phate binder improves any clinically relevant outcome, and the more important question remains—​does controlling phosphate to any specific target result in improved patient outcomes? If hyperphosphataemia does not respond to medical intervention, then issues to consider include nonadherence to prescribed binders or dietary measures, increased phosphate release from the skeleton (e.g. in severe hyperparathyroidism), stimulation of intestinal phos- phate absorption by excessive amounts of active vitamin D sterols, and (in patients on RRT) inadequate dialysis. Approach to serum calcium control The target range for serum total calcium (corrected for albumin) is 2.20 to 2.60 mmol/​litre (8.8–​10.4 mg/​litre). If the corrected total serum calcium level exceeds 2.60 mmol/​litre (10.4 mg/​litre), then therapies that cause a rise in serum calcium should be adjusted and other potential causes of hypercalcaemia should be considered, in particular an inappropriately high dialysate calcium concentration and immobilization (Box 21.6.8). Reversal of  deficiency of  native vitamin D3  Deficiency of the parent compound vitamin D3 (cholecalciferol) is common among patients with CKD stages 3, 4, 5, and 5d as a result of al- tered lifestyle, with insufficient sun exposure, dark skin, and loss of protein-​bound vitamin D (metabolites) into proteinuric urine or peritoneal dialysis fluid. Vitamin D deficiency can be diagnosed using the same criteria as in the general population, when plasma 25-​hydroxyvitamin D3 concentrations are less than 30 nmol/​litre (12 ng/​ml), and vitamin D insufficiency when they are between 30 and 50 nmol/​litre (12–​20 ng/​ml). However, current practice in patients with CKD is hindered by uncertainties regarding the Table 21.6.6  Serum biochemistry in the evaluation of CKD-​MBD Analyte Comments Normal range Calcium Low, normal, or elevated (elevated in severe HPT, vitamin D excess, therapy with calcium-​ containing phosphate binders, inappropriately high dialysate calcium, immobilization) 2.2–​2.6  mmol/​litre or 8.8–​10.4 mg/​dl Phosphate Elevated in advanced CKD (GFR <30 ml/​min) 0.8–​1.4  mmol/​litre or 2.4–​4.2 mg/​dl Intact PTH Elevated in HPT; can be normal or even low (mostly in cases of aluminium intoxication, adynamic bone disease, overtreatment with calcitriol, after parathyroidectomy); beware of interassay variations of intact PTH measurements 1–​7 pmol/​litre or 10–​65 pg/​ml 25-​(OH)D3 (calcidiol) Often low because of reduced sun exposure or low dietary intake; seasonal variation; if increased, check for exogenous source 50–​200 nmol/​litre or 20–​80 ng/​mla 1,25-​(OH)2D3 (calcitriol) Usually low (if increased, check for calcitriol ingestion; rarely endogenous overproduction (granulomatous disease)) 50–​120 pmol/​litre or 25–​50 pg/​mla Total alkaline phosphatases (AP) Normal or increased; elevated in severe HPT (exclude concomitant liver disease by determination of γ-​GT or bone-​specific AP isoenzyme) 60–​170 IU/​litre Osteocalcin Diagnostic information analogous to AP; fragments accumulate in advanced CKD; probably no extra information in addition to intact PTH and bone isoenzyme of AP c.3–​8 µg/​litreb (may depend on assay) Magnesium Normal or elevated (decreased renal excretion) 0.8–​1.3  mmol/​litre Aluminium Normal; elevated if aluminium-​containing phosphate binders are taken or if dialysate is aluminium contaminated <10 µg/​litre AP, alkaline phosphatase; CKD, chronic kidney disease; GFR, glomerular filtration rate; γ-​GT, γ-​glutamyl transferase; HPT, hyperparathyroidism; MBD, mineral and bone disorder; PTH, parathyroid hormone. a Normal range varies depending on season (consult local laboratory guidance). b In individual with normal renal function. Table 21.6.7  Typical serum biochemistry in the two main contrasting forms of renal bone disease Analyte Severe hyperparathyroidism Adynamic bone Calcium Variable, high normal or elevated in advanced secondary hyperparathyroidism Tendency to hypercalcaemia Phosphate Marked increase (dissolution of bone mineral) No typical pattern, often elevated Intact PTH Markedly elevated Normal or low Total alkaline phosphatases Usually elevated Tend to be low Box 21.6.8  Differential diagnosis of hypercalcaemia in patients with advanced CKD or on dialysis Related to CKD • Severe hyperparathyroidism • Intoxication with vitamin D sterols—​cholecalciferol, ergocalciferol, calcidiol, calcitriol or active vitamin D derivatives • Excessive dose of calcium-​containing phosphate binders • Inappropriately high dialysate calcium concentration Not directly related to CKD • Immobilization • Cancer with bone metastases • Myeloma • Granulomatous disease (e.g. sarcoidosis, tuberculosis) • Other rare causes of hypercalcaemia—​see Chapter 13.4 • Pseudohypercalcaemia—​elevated of total protein concentration

21.6  Chronic kidney disease 4851 appropriate threshold for intervention and by a lack of evidence that supplementation will impact on endpoints beyond the serum 25-​hydroxyvitamin D3 level. Furthermore, it is known that there is significant variation between different assay systems, adding fur- ther uncertainties to published guidelines. In the presence of symptoms and very low plasma levels, a loading regimen to provide a total of approximately 300 000 IU vitamin D (given either as separate weekly or daily doses over 6 to 10 weeks) can be used, with the exact regimen depending on the local availability of vitamin D preparations and guidelines. For example, 20 000-​IU capsules, two given weekly for 7 weeks (280  000 IU), or 800-​IU capsules, five a day given for 10 weeks (280 000 IU). Maintenance therapy then comprises vitamin D in doses equivalent to 800 to 2000 IU daily (occasionally up to 4000 IU daily), given daily or alterna- tively weekly at higher doses. In CKD, the synthesis of 1,25-​dihydroxyvitamin D3 depends on the concentration of the precursor substance 25-​hydroxyvitamin D3. This explains why administration of up to 2000 IU cholecalcif- erol per day (which is two to three times the average daily intake with vitamin D-​fortified food) leads to an increase of serum calcitriol and a decrease of serum intact PTH in many patients with CKD stages 3 and 4. Serum 25-​hydroxyvitamin D3 levels can usually be raised to a target of at least 50 nmol/​litre (20 ng/​ml) by supplementation with cholecalciferol or (less reliably) by sufficient sun exposure. It is important to note, however, that treatment with pharmaco- logical doses of cholecalciferol (‘native’ vitamin D) is potentially haz- ardous in CKD, especially if a patient confuses a weekly for a daily prescription. Cholecalciferol is less effective than its dihydroxylated and more active metabolites in terms of PTH reduction (see fol- lowing section on active vitamin D sterols), and carries a substantial risk of inducing prolonged increases in serum calcium and phos- phate that may damage residual renal function and possibly exacer- bate soft tissue calcification. Administration of  active vitamin D sterols  Current practice is hindered by uncertainties regarding the appropriate threshold for intervention and by a lack of evidence that supplementation will im- pact on endpoints beyond the serum 25-​hydroxyvitamin D level. If serum 25-​hydroxyvitamin D3 levels are higher than 50 nmol/​ litre (20 ng/​ml) and intact PTH levels remain elevated, treat- ment with low doses of active vitamin D sterols calcitriol and alfacalcidol should be considered (paricalcitol and doxercalciferol can also be used at greater expense). In patients with CKD stage 5, whether on dialysis or not, complete return of intact PTH concentrations to normal is not desirable. In advanced CKD, maintenance of normal bone turnover requires measured intact PTH concentrations to be around twice the ULN. It is uncertain whether this reflects PTH resistance of the skeleton, or problems with the second-​generation intact PTH assay which also meas- ures probably inactive fragments of PTH. However, in reflection of these uncertainties, the current KDIGO guidelines advise that in patients with CKD stages 3–​5 not on dialysis, the optimal PTH level is not known, but suggest that patients with levels of in- tact PTH above the ULN of the local assay are first evaluated for hyperphosphataemia, hypocalcaemia and vitamin D deficiency. Any abnormalities found should be corrected and the effect on PTH assessed. If PTH continues to rise, then oral alfacalcidol or calcitriol should be introduced. Box 21.6.9 provides an algorithm for the prophylaxis of secondary hyperparathyroidism. In experimental studies, daily administration of calcitriol or alfacalcidol lowers intact PTH concentration and prevents parathyroid hyperplasia. Higher-​dose intermittent therapy (1–​2 µg twice weekly) may decrease the likelihood of inducing hypercalcaemia, but there is no published evidence that this is more effective in the long term. Medical management of advanced hyperparathyroidism Hypocalcaemia is the most potent driver of PTH in both normal physiology and in CKD, hence this must be corrected before any other measures are likely to be effective. Once this has been done, other measures such as higher-​dose vitamin D sterols can be introduced. The main side effects of treatment with active vitamin D sterols are hypercalcaemia and hyperphosphataemia. There has therefore been an intense search for new analogues that suppress parathyroid activity while causing less hypercalcaemia and hyperphosphataemia. Some of these—​including paricalcitol (19-​nor-​1α,25-​dihydroxyvitamin D2) and doxercalciferol (1α-​hydroxyvitamin D2)—​are available in some countries, but no clinical trials have provided convincing evidence that they are better than the parent compound, calcitriol. Observational studies in large dialysis patient cohorts suggest that treatment with any active vitamin D compound is associated with better outcome than no vitamin D treatment, and that treatment with novel active vitamin D derivatives may lead to better patient outcome than treatment with calcitriol. However, in observational studies it is impossible to know whether such associations are due to bias, confounding, chance, or cause, and hence such results cannot be relied on. Another important method of influencing calcium balance in patients on dialysis is to manipulate the concentration of ionized calcium in the dialysate. Normal serum ionized calcium is around 1.1 to 1.2  mmol/​litre, and in the past a dialysate calcium con- centration of 1.75 mmol/​litre (7 mg/​100 ml) was recommended, such that net uptake of calcium occurred during the dialysis ses- sion. However, if calcium-​containing phosphate binders or active Box 21.6.9  Algorithm for prophylaxis of secondary hyperparathyroidism Monitoring (serum biochemistry) • Calcium, albumin, phosphate, intact PTH, 25-​hydroxyvitamin D3, aluminium Prophylactic measures If serum 25-​hydroxyvitamin D3 is low, i.e. below 50 nmol/​litre (20 ng/​ml): • Consider cholecalciferol 2000 U/​day or 20 000 U/​week If plasma calcium is decreased and/​or plasma phosphate is increased: • Give calcium acetate 0.5–​1.0 g with each meal If serum phosphate is increased and plasma calcium is normal or high: • Consider calcium-​free phosphate binder (e.g. lanthanum or sevelamer carbonate) If serum intact PTH is consistently above target ranges (see text) and serum calcium/​phosphate is normal (spontaneously or after intervention): • Give alfacalcidol/​calcitriol 0.25 µg/​day or equivalent doses of other analogues

section 21  Disorders of the kidney and urinary tract 4852 vitamin D sterols are administered, then intestinal uptake of cal- cium is increased and patients may develop a strongly positive calcium balance, with or without hypercalcaemia. The standard concentration of calcium in dialysate currently in use is usually 1.50 mmol/​litre (6 mg/​100 ml). This may be reduced further to 1.25 mmol/​litre (5 mg/​100 ml) if hypercalcaemia develops, but it is important to ensure that careful monitoring is in place since if calcium carbonate and/​or active vitamin D derivatives are omitted, then a negative calcium balance may result with exacerbation of secondary hyperparathyroidism. In peritoneal dialysis fluids, a calcium level of 1.25 mmol/​litre is widely used since the risk of inducing hypercalcaemia is much less because the volumes of fluid in use are so much lower. Another effective treatment for hyperparathyroidism is to use a calcimimetic, only one of which—​cinacalcet—​is in clinical use at present. This renders the CaR more sensitive to extracellular Ca2+ and has been shown to reduce elevated serum intact PTH concen- trations in dialysis patients with moderate or severe hyperparathyr- oidism. Its beneficial effect on serum biochemistry is maintained over prolonged time periods, but it is not licensed for use in patients with CKD who are not on dialysis, and some experts—​including the KDIGO working group—​do not recommend its use in such cases because of lack of data on long-​term efficacy and safety in this population. In any patient with significant residual renal function, including a transplanted patient, the risk of inducing symptomatic hypocalcaemia is significant and therefore initial weekly monitoring of serum calcium is essential. Tumour-​like parathyroid growth and parathyroidectomy Uncontrolled severe secondary hyperparathyroidism can become an adenomatous process that bears similarities to tumour growth. Nodular hyperplasia is usually found in patients whose estimated parathyroid mass exceeds 1 to 1.5 g, with the nodules exhibiting clonal growth, with microsatellite analysis showing loss of hetero- zygosity for many alleles, including putative tumour suppressor genes. The parathyroid tissue within these nodules is also charac- terized by reduced expression of vitamin D receptors and CaRs, which could explain—​at least in part—​the eventual lack of response to medical management, at which point the condition is referred to as tertiary hyperparathyroidism. It appears that continuous stimu- lation of the parathyroid gland selectively favours cells with higher proliferative potential so that the gland progressively escapes from growth-​inhibitory control mechanisms. This is illustrated by the fact that regrowth, including locally invasive regrowth, occurs in many patients after subtotal parathyroidectomy or autotransplantation of parathyroid tissue. Tertiary hyperparathyroidism is by definition uncontrollable by manipulation of calcium and phosphate levels, and it tends to be associated with high levels of both, plus increased serum alkaline phosphatase. It is also generally unresponsive to vitamin D therapy, which may be impossible to deploy because of high calcium levels. The patient may have a variety of symptoms including general mal- aise, pruritus, muscle and bone aches, depression, and anaemia due to PTH-​induced unresponsiveness to erythropoietin. Before the introduction of cinacalcet treatment, parathyroidectomy was generally considered in symptomatic patients with marked ele- vation of serum intact PTH (>100 pmol/​litre or 950 pg/​ml) who failed to respond to medical treatment within 8 to 12 weeks. Such patients almost certainly have tertiary hyperparathyroidism, and imaging will usually show significant parathyroid enlargement (es- timated mass >1.0–​1.5 g). Over the past decade, oral cinacalcet has been used more commonly in this situation, and will effectively re- duce PTH levels if carefully monitored and the dose appropriately titrated according to response. The National Institute for Health and Care Excellence (NICE) published guidance on its use in this situ- ation (Box 21.6.10). An indication for urgent parathyroidectomy is calcific uraemic arteriolopathy (also called calciphylaxis), namely ischaemic skin ne- crosis secondary to calcification of skin arteries, but only if associ- ated with elevated serum intact PTH levels. There has been a long-​standing debate as to whether total parathyroidectomy or subtotal parathyroidectomy should be pre- ferred, the latter with a remnant left in situ or autotransplanted into the subcutaneous abdominal fat or forearm musculature. There are no trial data to inform decision-​making, and autotransplantation has become generally less usual over recent years. Leaving parathy- roid tissue behind is associated with a relatively high risk of local recurrence, presumably because of the increased growth potential of parathyroid cells, although the risk can be reduced if only non-​ nodular parts of the gland are autotransplanted. Alcohol injection into the enlarged parathyroids under ultrasonographic guidance has been tried as an alternative to surgery, but this procedure has not found wide application, except in Japan. Box 21.6.11 summarizes the approach to the management of pa- tients with advanced renal secondary hyperparathyroidism. Anaemia Pathogenesis The physiological mechanisms controlling red cell mass are shown in Fig. 21.6.12. The maintenance of a normal red cell mass requires an appropriate rate of red cell production by the bone marrow, with no substrate limitations and under the influence of an adequate amount of erythropoietin. The production rate should balance red cell loss and destruction. All elements are disturbed in uraemia, and anaemia is one of the most obvious manifestations of the uraemic syndrome. Red cell lifespan is shortened by accelerated destruction. Erythropoietin secretion is enhanced, but not to a sufficient level (Fig. 21.6.13). Box 21.6.10  NICE guidance on the use of oral cinacalcet 1 Cinacalcet is not recommended for the routine treatment of sec- ondary hyperparathyroidism in patients with endstage renal disease on maintenance dialysis therapy 2 Cinacalcet is recommended for the treatment of refractory secondary hyperparathyroidism in patients with endstage renal disease (including those with calciphylaxis) only in those:

—​ who have ‘very uncontrolled’ plasma levels of intact parathyroid hormone (defined as greater than 85 pmol/​litre (800 pg/​ml)) that are refractory to standard therapy, and a normal or high adjusted serum calcium level, and

—​ in whom surgical parathyroidectomy is contraindicated, in that the risks of surgery are considered to outweigh the benefits 3 Response to treatment should be monitored regularly and treatment should be continued only if a reduction in the plasma levels of intact parathyroid hormone of 30% or more is seen within 4 months of treat- ment, including dose escalation as appropriate

21.6  Chronic kidney disease 4853 Epidemiology and clinical significance Anaemia (defined as a haemoglobin concentration <130 g/​litre in adult men and postmenopausal women, and <120 g/​litre in pre- menopausal women) is common in CKD, particularly in those with diabetes, and affects nearly 90% of all with CKD stages 4 and 5, many of whom will have haemoglobin concentrations less than 100 g/​litre. Renal anaemia, which is normochromic and normocytic, accounts for many of the symptoms that previously were attributed to uraemia, including lethargy, cold intolerance, and general fatigue. Population studies and registry data generally report higher mor- tality for dialysis patients with haematocrit 30 to 33% compared with those with haematocrit 33 to 36%, or for haemoglobin levels less than 110 g/​litre or 115 g/​litre than above. Management As a result of many trials of erythropoietin and other erythropoiesis-​ stimulating agents (ESAs), there is now general agreement that partial correction of anaemia in patients with CKD improves physiological and clinical status, as well as quality of life. This agree- ment has manifested itself in a plethora of clinical practice guide- lines, including those produced by the National Kidney Foundation Dialysis Outcomes Quality Initiative (DOQI) in the United States of America, the ERA/​EDTA, the Canadian Society of Nephrology, and the Japanese Society for Dialysis Therapy. There are minor differences between the particular guidelines, but they are all essentially similar. There is no particular haemoglobin concentration at which symptoms become manifest in all patients, hence the decision to start treatment in a particular patient is always a matter of judge- ment. As a rule of thumb, if a patient with CKD has a haemoglobin concentration of less than 110 g/​litre and symptoms that might be attributable to anaemia, then treatment to restore haemoglobin to the range 110 to 120 g/​litre is warranted if available, but it has been convincingly shown in randomized studies that full correction to a higher level (‘normal or near normal’) is associated with poorer outcomes and should be avoided. Before starting treatment with ESAs it is important to exclude other causes of anaemia: serum vitamin B12, red cell folate, and in- dices of iron status should be assessed in all patients, with other tests ( O 2 de li ve ry in u ra e mi a) All tissues Bone marrow (Shortened red cell survival in uraemia) Red cell mass Sense O2 delivery (abnormal in uraemia) Hypoxia-inducible factor (HIF) Epo gene in interstitial fibroblasts Epo synthesis O x y g e n d el i v e r y EPO ( in uraemia) ( Red cell production in uraemia) Kidney Fig. 21.6.12  The relationship between red cell mass, oxygen delivery, erythropoietin synthesis, and red cell production in the bone marrow. Erythropoietin production is reduced in uraemia because of defective sensing, reduced synthesis, or both. 1000 10 Haematocrit (%) Log Epo concentration Spread of values in renal anaemia Spread of values in nonrenal anaemia 10 20 30 40 50 Fig. 21.6.13  In renal anaemia, the erythropoietin concentration rises in response to anaemia, but to a much lower level than in comparably severe nonrenal anaemia. Box 21.6.11  Treatment of advanced hyperparathyroidism If serum intact parathyroid hormone (PTH) is constantly above target range (see text): • normalize serum calcium and phosphate levels If serum phosphate is elevated: • give calcium carbonate, calcium acetate, or calcium-​free phosphate binders with meals • reduce excessive intake of dietary phosphate • increase efficacy of dialysis (higher blood flow, longer dialysis sessions, more frequent dialysis sessions) If serum calcium is elevated: • reduce dialysate calcium to 1.5 mmol/​litre (6 mg/​dl) or—​transiently—​to 1.25 mmol/​litre (5 mg/​dl) • reduce or withdraw calcium-​containing oral phosphate binders or active vitamin D sterols If serum calcium and phosphate have been normalized and elevated intact PTH persists: • increase dose of calcitriol (0.5–​3 µg) or alternative active vitamin D sterols (e.g. alfacalcidol, paricalcitol, doxercalciferol)—​these can be given one to three times per week, or daily, with dose and time interval depending on degree of elevation of serum intact PTH; alternatively, administer cinacalcet (initial dose 30 mg/​day) • monitor serum calcium, phosphate, intact PTH, and total alkaline phosphatases If serum intact PTH decreases below approximately 16.5 pmol/​litre (150 pg/​ml): • interrupt administration of calcitriol, measure intact PTH again, and decide whether low-​dose, long-​term prophylaxis is necessary If serum intact PTH fails to decrease and/​or hypercalcaemia/​ hyperphosphataemia develop or persist: • monitor parathyroid gland size (ultrasonography; MIBI scan before surgery to localize ectopic glands) • consider cinacalcet (initial dose 30 mg/​day) or surgical parathyroidectomy Note: increased active vitamin D dose is contraindicated as long as plasma phosphate or calcium is elevated.

section 21  Disorders of the kidney and urinary tract 4854 on the basis of clinical suspicion, for example, an elderly patient presenting with renal impairment and marked anaemia may have myeloma. If a patient is significantly iron deficient, then standard clinical methods of history and examination followed by appropriate investigation may be required to determine the cause, but otherwise it is important to recognize that an optimal response to ESAs re- quires plentiful iron (ferritin >200 μg/​litre), not simply a level that is not deficient. It is also important to ensure that blood pressure is reasonably controlled before ESAs are given. In the early days of erythropoietin treatment, rapid increases in haemoglobin concen- tration in combination with poorly controlled blood pressure pre- cipitated hypertensive encephalopathy in some patients. Therefore ESAs should not be started (or a dose should be omitted) if blood pressure is higher than 160/​100 mmHg. Box 21.6.12 shows an algorithm whereby the patient’s iron status is optimized before ESAs are administered. A variety of ESAs are available:  all are clinically effective, many can be administered intravenously or subcutaneously, patients may prefer one rather than another because of the particular method of delivery and fre- quency of administration (variable from once or twice a week to monthly), and those paying may wish to choose the cheapest. If the haemoglobin fails to respond, or falls after initially responding, then causes given in Table 21.6.8 need to be considered. An orally active inhibitor of hypoxia-inducible factor (HIF, see Fig. 21.6.12) prolyl hydoxylase, roxadustat, has recently been shown to be effective in correcting anaemia in patients with chronic kidney disease (both on and not on dialysis). The place of this and other similar agents in the management of renal anaemia has not yet been established. Complications of chronic renal failure Chronic renal failure affects all parts of the body. Many of its compli- cations have been discussed in this chapter, but a more complete—​ although not exhaustive—​list is given in Table 21.6.9. Preparation for renal replacement therapy In any patient with progressive CKD who is likely to require fu- ture dialysis, it is important to preserve the superficial veins of the forearm for creation of an arteriovenous fistula. Whenever possible, blood should only be taken from the veins on the dorsal surface of the hand, or—​if veins in the forearm or elbow must be punctured or cannulated—​the nondominant arm must be kept free from assault for later formation of an arteriovenous dialysis fistula. Once endstage kidney failure is inevitable, and progression of the underlying disease process cannot be halted, the patient must be prepared physically and psychologically for RRT or conserva- tive management. The length of time needed for this preparation process, which requires contact with a variety of members of the renal multidisciplinary team, will vary from patient to patient, but usually needs 9–​12 months in total, hence an approximate predic- tion of the point at which RRT will become necessary, and when symptoms are likely to begin to become troublesome, is essen- tial (Box 21.6.13). This can be established by consideration of the Box 21.6.12  Treatment of renal anaemia Is anaemia due to CKD? • Consider other causes Determine iron status: • Iron deficiency is defined by serum ferritin less than 100 µg/​litrea • Functional iron deficiency is defined by serum ferritin greater than 100 µg/​litre with hypochromic red cellsb greater than 6% or transferrin saturation less than 20% Optimize iron status: • Aim to maintain serum ferritin greater than 200 µg/​litre with hypochromic red cells less than 6% (unless ferritin >800 µg/​litre) or transferrin saturation greater than 20% (unless ferritin >800 µg/​litre) • It is likely that this will require intravenous iron (usually 600–​1000 mg for adults) Initiate ESAs and adjust dose and frequency: • To maintain stable haemoglobin (Hb) in range 100–​120 g/​litre (adjustments to ESA doses should be considered when Hb is <105 or >115 g/​litre) • To keep the rate of increase of Hb between 1 and 2 g/​litre per month Maintain adequate iron levels: • Keep serum ferritin in the range 200–​500 µg/​litre with hypochromic red cells less than 6% (unless ferritin >800 µg/​litre) or transferrin satur- ation greater than 20% (unless ferritin >800 µg/​litre) • It is likely that this will require regular but infrequent intravenous iron Monitor: • Hb every 2 to 4 weeks (induction phase, or after ESA dosage change) or every 1 to 3 months (maintenance phase) • Iron status every 1 to 3 months (but not within a week of receiving intravenous iron) • Keep serum ferritin less than 800 ug/​litre • Aim to increase haemoglobin by 10 to 20 g/​litre per month Review: • Patient’s clinical response • If there is any unexpected change in Hb level a The ‘normal’ range for ferritin is usually quoted as 15–​200 µg/​litre. b Percentage of hypochromic red cells directly reflects the number of red cells with suboptimal haemoglobin content and may be determined by some automated analysers: less than 2.5% is normal and greater than 10% indicates definite iron deficiency. Table 21.6.8  Causes of failure to respond to
erythropoiesis-​stimulating agents Cause Comment Is the patient receiving the injections? Absolute or functional iron deficiency Hypochromic red cells, reticulocyte haemoglobin, or transferrin saturation with serum ferritin Acute or chronic inflammatory states These reduce the efficacy of ESAs Other haematological conditions Consider myeloma, other malignant diseases affecting the bone marrow, thalassaemia, vitamin
B12 or folate deficiency Chloramine in dialysis water Can cause haemolysis presenting as apparent resistance to ESAs Aluminium overload Rare Antierythropoietin antibodies Rare, but a significant concern with one ESA preparation that led to its temporary withdrawal

21.6  Chronic kidney disease 4855 Table 21.6.9  Complications of chronic renal failure Complication Comment Cardiovascular system Hypertension Discussed in text Left ventricular hypertrophy Found in 75% of dialysis patients Coronary atherosclerosis Cardiovascular disease is responsible for about 50% of deaths of patients receiving RRT. High risk of acute myocardial infarction, but sudden arrhythmic death is the most common fatal cardiac event Pericarditis A feature of neglected uraemia, including inadequate dialysis; can lead to tamponade and death Calcific valvular disease Mitral valve calcification found in one-​third of dialysis patients. Calcific aortic stenosis can progress very rapidly Respiratory system Pulmonary oedema Feature of fluid retention Pleural effusion Feature of fluid retention Gastrointestinal system Anorexia, nausea and vomiting Poor oral hygiene Haemorrhage Due to nonspecific gastric ulceration and/​or angiodysplasia anywhere in the gastrointestinal tract; CKD renders normal bone marrow compensatory mechanisms less effective Pancreatitis Can be provoked by hypercalcaemia; long-​term dialysis patients develop pancreatic fibrosis, but this does not seem to affect pancreatic function Nervous system Encephalopathy Typically presents with confusion, myoclonic muscular twitching, and impairment of consciousness; seizures are rare unless there is accelerated hypertension Sensorimotor peripheral polyneuropathy Presents as dysaesthesias, restless legs, eventually weakness with foot drop; dialysis leads to slow improvement, but patients are often left with motor disability Autonomic neuropathy Manifests as abnormal cardiovascular reflexes, particularly on dialysis Carpal tunnel compression Caused by β2-​microglobulin amyloid deposition; a feature in almost all patients who have been on dialysis for more than 10 years Dialysis dementia Presents as gradual deterioration in intellectual performance, progressing to dementia with abnormal movements. Due to aluminium intoxication. Should be of historical interest only Musculoskeletal system Mineral and bone disorder Discussed in text Proximal myopathy Crystal arthropathy Gout and pseudogout (pyrophosphate arthropathy) are common. Management of gout can be difficult: NSAIDs are best avoided if possible in patients with advanced CKD who are not on dialysis, although very short-​term use is acceptable; diarrhoea caused by colchicine can lead to acute deterioration of CKD; a short course of oral prednisolone (20 mg/​day) may be the best treatment for an acute attack; reduced dose of allopurinol required Skin Pigmentation Pruritus Can be a cause of significant distress. Associated with dry skin (xerosis), and worse when the skin is warm. Cause is uncertain—​raised calcium × phosphate product, histamine sensitivity, and ‘uraemia’ have been blamed. Scratching can lead to infection and nodular prurigo. Treatments include starting/​increasing dialysis, emollient lotions/​creams, controlling plasma phosphate, keeping cool, antihistamines (e.g. chlorphenamine 4 mg at night), naltrexone, and ultraviolet phototherapy Calciphylaxis Discussed in text Bullous eruptions Pseudoporphyria, affecting sun-​exposed areas. Thought to be due to accumulation of porphyrins in high molecular weight protein-​bound complexes that are not removed by haemodialysis. Treatment is by avoidance of sun exposure, phlebotomy (for patients who are not anaemic and who have increased iron stores) and ESAs (which remove iron from stores by enhancing production of red blood cells) Sexual function Men Loss of libido and erectile impotence are common and of multifactorial cause. Sperm counts may be low leading to reduced fertility. Priapism is a rare complication of haemodialysis treatment Women Most women with severe CKD develop irregular periods or amenorrhoea and are infertile, with rare pregnancies almost always ending in miscarriage. For general discussion of pregnancy in women with kidney disease, see Chapter 14.5 (continued)

section 21  Disorders of the kidney and urinary tract 4856 rate of renal deterioration, most easily demonstrated by plotting the reciprocal creatinine or eGFR against time and extrapolating forward to the month when eGFR is likely to reach 10 ml/​min or less (Fig. 21.6.14). This information is useful for the patient and those planning care, providing a guide for the timing of the cre- ation of permanent vascular access (thereby avoiding the perils of temporary lines), placement of peritoneal dialysis catheters, or activating the patient on to a transplant waiting list. If the process is planned sufficiently well then the transition to RRT should be possible without requiring hospital admission. Complication Comment Psychological Anxiety and depression are predictable and understandable consequences of loss of health, control, and pleasure. They tend to be most prominent in young patients. The best treatment is by sympathetic support of the dialysis multidisciplinary team. Psychotherapy/​counselling can be helpful, but psychiatrists and/​or medication have little to offer unless there is a specific mental illness Metabolic Glucose intolerance CKD causes resistance to insulin-​mediated glucose uptake in skeletal muscle Complex effects on lipids Increased very-​low-​density lipoproteins; increased high-​density lipoproteins Enhanced protein catabolism Risk of malnutrition discussed in text Haematological Anaemia Discussed in text Impaired platelet function Platelet numbers are normal, but function impaired at the level of endothelial contact Impaired T-​cell immunity Mechanism uncertain, but puts patients with advanced CKD at higher risk of reactivation of tuberculosis and herpes zoster, of failure to clear some viral infections (e.g. hepatitis B), and of failure to generate normal responses to immunization (e.g. hepatitis B vaccine) Impaired neutrophil function Mechanism uncertain, but may in part explain high incidence and severity of bacterial infections Infective Blood-​borne viruses Dialysis is a risk factor for hepatitis C, hepatitis B, and HIV Methicillin-​resistant Staphylococcus aureus (MRSA) Dialysis patients are at high risk of acquiring MRSA because of their frequent contact with medical services, common requirement for invasive procedures/​indwelling lines, and (perhaps) susceptibility to infection Clostridium difficile infection A problem on many renal units Endocarditis Bacteraemias in dialysis patients are often attributable to infection of vascular access sites, which, combined with high prevalence of calcific valvular disease, creates high risk of endocarditis, usually (70%) due to S. aureus CKD, chronic kidney disease; ESA, erythropoiesis-​stimulating agent; NSAID, nonsteroidal anti-​inflammatory drug; RRT, renal replacement therapy. Table 21.6.9  Continued Box 21.6.13  Recommended standards to improve choice and quality in dialysis and kidney transplantation (from UK Renal National Service Framework (2004)) Standard one: a patient-​centred service All children, young people, and adults with CKD are to have access to information that enables them with their carers to make informed de- cisions and encourages partnership in decision-​making, with an agreed care plan that supports them in managing their condition to achieve the best possible quality of life. Standard two: preparation and choice All children, young people, and adults approaching established renal failure are to receive timely preparation for RRT so the complications and progression of their disease are minimized, and their choice of clinically appropriate treatment options is maximized. Standard three: elective dialysis access surgery All children, young people, and adults with established renal failure are to have timely and appropriate surgery for permanent vascular or peri- toneal dialysis access, which is monitored and maintained to achieve its maximum longevity. Standard four: dialysis Renal services are to ensure the delivery of high-​quality, clinically appro- priate forms of dialysis which are designed around individual needs and preferences and are available to patients of all ages throughout their lives. Standard five: transplantation All children, young people, and adults likely to benefit from a kidney transplant are to receive a high-​quality service which supports them in managing their transplant and enables them to achieve the best possible quality of life. 10.0 100 15 9.0 8.0 7.0 6.0 Nephrotic syndrome Bx: MCGN Transplant from mother UNIVERSITY Starts Epo eGFR Established career CAPD 1/creatinine (μmol/l) × 1000 5.0 4.0 3.0 2.0 1.0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 eGFR ml/min Fig. 21.6.14  A graph showing the progressive decline in renal function in a patient with glomerulonephritis. The timing of the need to start dialysis could be predicted sufficiently well to allow advance planning of treatment.

21.6  Chronic kidney disease 4857 The temptation to delay starting dialysis for as long as possible should be avoided: severe uraemia puts the patient at risk and is likely to result in a prolonged inpatient admission to deal with un- foreseen problems and potentially life-​threatening complications. However, there is no evidence that starting dialysis early in an asymptomatic patient (eGFR 10–​14 ml/​min) is better than starting when they begin to develop their first uraemic symptoms. The absolute indications for dialysis, other than in patients for whom such treatment would be inappropriate, are the develop- ment of complications that cannot be treated by conservative and pharmacological means. These are hyperkalaemia, fluid overload, acidosis, severe hypertension, pericarditis, encephalopathy, and neuropathy. To wait for these to develop puts the patient at risk, and makes hospital admission more likely. However, if plans for RRT are in place and dialysis access is prepared, then many patients and their nephrologists will opt to wait until some uraemic symptoms such as anorexia, lassitude, and pruritus develop, if only because their relief reinforces the need to adjust to regular dialysis. Apart from the serum potassium concentration and the degree of acidosis, blood tests such as urea and creatinine do not provide a safe guide as to when to RRT should commence. Nevertheless, it is advis- able to start dialysis, even in the absence of symptoms, when eGFR falls below about 7–​8 ml/​min. In small patients with little muscle bulk, the urea concentration is often between 25 and 35 mmol/​litre and the creatinine concentration between 650 and 800 µmol/​litre; in larger patients, the blood urea concentration is typically 40 to 50 mmol/​litre and that of creatinine above 800 µmol/​litre. Initiation of dialysis at lower blood levels of urea and creatinine is recom- mended in patients with diabetes as the primary cause of CKD. Pre-​emptive transplantation In choosing a modality of RRT, the first question for discussion should be whether the patient is suitable for kidney transplantation and whether there are any suitable live donors within the family. Pre-​ emptive kidney transplant is the treatment of choice for CKD stage 5 as successful transplantation provides increased patient survival time, improved quality of life, and reduced costs compared to dia- lysis. Activation on the transplant waiting list should occur at least 6 months prior to need for dialysis, typically when eGFR falls below 15 ml/​min/​1.73 m2, and in view of both better transplant outcomes and the shortage of deceased donors, the possibility of a living donor should be explored early with all potential recipients. The treatment of choice for all patients with a living donor is a pre-​emptive transplant occurring before dialysis commences. This strategy results not only in avoidance of dialysis but improved pa- tient survival and also improved graft survival. The pre-​emptive approach shows survival benefits in those being considered for their first transplant and also in those with failing grafts requiring retransplantation. Current pre-​emptive transplant rates in the United Kingdom (2014) are 37% for living donors and 16% for de- ceased donors, with rates in the United States of America (2012) being 18% and 11% respectively. There is potential to improve pre-​ emptive rates as 40% of all transplant recipients are predialysis at the time of wait-​listing (United Kingdom), while recent data from the United States of America suggest a third of nonpre-​emptive living donor recipients had dialysis for less than one year prior to trans- plantation. See Chapter 21.7.3 for further discussion. Home dialysis The second question to discuss is does the patient wish to dialyse at home (home haemodialysis or peritoneal dialysis) or in-​centre (haemodialysis)? The answers to these two questions depend on many factors, but availability of all treatment modalities should not be a factor (see Chapters 21.7.1 and 21.7.2 for further discussion). If haemodialysis (either at home or in-​centre) is chosen, vascular access should be created at least 2 months before it is needed. If con- tinuous ambulatory peritoneal dialysis is to be used, the Tenckhoff catheter should be placed 2 to 3 weeks before dialysis needs to be started to allow it to seal. Conservative kidney management of terminal uraemia Few treatments demand as much of a patient’s endurance and deter- mination as peritoneal or haemodialysis, with hospital admissions, dietary restrictions, multiple invasive procedures, and polypharmacy. It is therefore particularly important that patients with progressive CKD are enabled to understand all that will be involved, and the im- plications for both their quantity and quality of life. For those over the age of 75 years with significant co-​morbidities, survival on dia- lysis is relatively short (50% survival, <2 years), and 3 days in every week are consumed by travel, dialysis, and recovery. Consequently, many such people may opt for reduced quantity but better quality of life, and reject dialysis treatment. Because intuitively many people expect that instituting dialysis in a patient with kidney failure and other comorbid conditions will result in some improvement of their clinical condition, it is common to find that friends and relatives find it hard to understand why someone may decide not to have dialysis. It is therefore important that the multidisciplinary team supports the patient to make their own decision wherever possible, and to understand fully its implications. These difficult decisions can only be made once the patient and relatives or carers have a full understanding of both dialysis and con- servative kidney management. This takes time, effort and patience from both the renal multidisciplinary team and the patient. The patient’s capacity to determine their own treatment must be con- sidered, and any advance treatment directives taken into account. In the United Kingdom, many of these aspects are now legally en- shrined within the Mental Capacity Act. Equally there are patients for whom dialysis may appear inappro- priate to the clinical team, but who wish to commence for a variety of personal and family reasons, and others who choose to discon- tinue dialysis after a certain point. Under occasional circumstances, a 3-​month trial of dialysis might be appropriate, but this may sug- gest a failure of the planning process. In-​centre haemodialysis is usually the treatment modality offered in such circumstances, but establishing access and the requirement for thrice-​weekly transport to a dialysis facility which the frail body may find difficult to tolerate, recovering only in time for the next dialysis session, can be truly miserable for the patient and all others concerned. Home dialysis by means of assisted automated peritoneal dialysis may be more appro- priate and less intrusive. In frail patients, it is not likely that dialysis will greatly prolong life, although it can certainly reduce its quality. There have been no

section 21  Disorders of the kidney and urinary tract 4858 trials that have randomized such patients to treatment with RRT or to conservative (palliative) management, but one observational study reported the outcome of 63 patients who were recommended to receive palliative care after multidisciplinary assessment and counselling about treatment options. Ten of these patients opted for and received dialysis treatment, but their median survival after initiation of dialysis (8.3 months) was not significantly longer than survival beyond the putative date of dialysis initiation in the pal- liatively treated patients (6.3 months), and 65% of those treated with dialysis died in hospital, compared with 27% of those re- ceiving palliative care. A study of 3702 nursing home residents in the United States of America who started on dialysis treatment be- tween June 1998 and October 2000 revealed that 12 months later most (58%) had died and only 13% had maintained their predialysis functional state. The process of dialysis withdrawal and dying can be traumatic for the patient, the patient’s family and friends, and for staff if careful discussion and planning is not in place. With appropriate prep- aration it should be a peaceful, painless, and dignified end to life. Around 10% of deaths in dialysis programmes follow withdrawal of treatment. If one takes the view that dialysis is a treatment to allow a patient to continue living with a reasonable quality of life, as opposed to delaying death in the short term, then it will rarely be taken up by patients with other immediately life-​limiting conditions. However, the ethical and legal issues are complex and it is essential that the patient makes the decision for conservative management when fully informed and able to do so. The various members of the multidis- ciplinary team should discuss with the patient the option of conser- vative management well before dialysis is actually needed. If this is not possible because of an acute presentation, the discussion process may need to be rapid but must not be avoided. The patient must be given a realistic account of what dialysis can achieve, what it cannot achieve, and at what personal cost—​ access, travel, restrictions, and complications. One could argue that it should not be started when survival beyond 3 months outside hospital is unlikely, but a patient who is looking forward to a par- ticular event (birth of a grandchild, a graduation ceremony, a final holiday) may wish to commence dialysis (perhaps on a temporary basis) against all expectations. These conversations can be difficult and ideally should not be hurried. They should routinely be aided by offering arrangements for the patient (and relative/​friend) to visit the dialysis unit, and in particular to meet other patients and carers outside the hospital environment. Without careful and timely dis- cussions it is possible for the patient (and their relatives/​friends) to be left with the impression that dialysis means that ‘the staff care, and I’ll be well looked after’, whereas no dialysis means that ‘the staff don’t care, and I’ll be left at home to fend for myself’. Properly managed, death from uraemia is peaceful and free of pain or suffering. It is important to ensure that the patient has peace of mind, that they are comfortable with their decision, and that their family members are understanding and supportive. They will be comforted to know that their doctor respects their decision. A late change of mind can occasionally occur and must similarly be respected. The preferred place of death should be discussed and decided. Clearly a hospital ward is a possibility, but the patient may never have been an inpatient and therefore have no relationship with the staff. Hospices will admit terminally ill uraemic patients, but most people prefer to die at home if at all possible. This requires close liaison with appropriate primary care staff and continuing sup- port from the renal multidisciplinary team. Several distressing symptoms may need to be controlled in the final 2 or 3 weeks of life. Breathlessness from pulmonary oedema and acidosis is best controlled with a subcutaneous morphine in- fusion. Nausea and anorexia can be helped with regular chlorpro- mazine 25 mg four times daily, and ondansetron 8 mg twice daily can also be effective. Food and fluid should be offered in small, palatable helpings, with no pressure to eat or drink exerted on the patient. The mouth can become dry and crusted from mouth breathing and will smell foul from uraemic saliva, for which regular mouth washes and gum care will help. Pruritus is managed by keeping the skin cool, and soft with emollients. The patient may not be aware of myoclonic jerks, but these may distress the family—​benzodiazepines, such as clonazepam, can be prescribed if needed. FURTHER READING Combs SA, Davison SN (2015). Palliative and end-​of-​life care issues in chronic kidney disease. Curr Opin Support Palliat Care, 9,
14–​19. Cooper BA, et al. (2010). A randomized, controlled trial of early versus late initiation of dialysis. N Engl J Med, 363, 609–​19. Coresh J, et  al. (2003). Prevalence of chronic kidney disease and decreased kidney function in the adult US population:  Third National Health and Nutrition Examination Survey. Am J Kidney Dis, 41, 1–​12. Davison R, Sheerin NS (2014). Prognosis and management of chronic kidney disease at the end of life. Postgrad Med J, 90, 98–​105. Eckardt KU, et al. (2013). Evolving importance of kidney disease: from subspecialty to global health burden. Lancet, 382, 158–​69. Hallan SI, et al. (2006). International comparison of the relationship of chronic kidney disease prevalence and ESRD risk. J Am Soc Nephrol, 17, 2275–​84. Hallan S, et al. (2007). Association of kidney function and albuminuria with cardiovascular mortality in older vs younger individuals: the HUNT II Study. Arch Intern Med, 167, 2490–​6. Hsu CY, et al. (2005). Elevated blood pressure and risk of end-​stage renal disease in subjects without baseline kidney disease. Arch Intern Med, 165, 923–​8. Hsu CY, et al. (2006). Body mass index and risk for end-​stage renal disease. Ann Intern Med, 144, 21–​8. Jha V, et al. (2013). Chronic kidney disease: global dimensions and perspectives. Lancet, 382, 260–​72. Keith DS, et al. (2004). Longitudinal follow-​up and outcomes among a population with chronic kidney disease in a large managed care organization. Arch Intern Med, 164, 659–​63. Kheder-​Elfekih R, et al. (2015). Hypertension and chronic kidney dis- ease: respective contribution of mean and pulse pressure and arterial stiffness. J Hypertens, 33, 2010–​15. Kurella-​Tamura M, et al. (2009). Functional status of elderly adults be- fore and after initiation of dialysis. N Engl J Med, 361, 1539–​47. Matsushita K, et al. (2010). Association of estimated glomerular filtra- tion rate and albuminuria with all-​cause and cardiovascular mor- tality in general population cohorts: a collaborative meta-​analysis. Lancet, 375, 2073–​81. Pippias M, et al. (2016). The changing trends and outcomes in renal replacement therapy: data from the ERA-​EDTA Registry. Nephrol Dial Transplant, 31, 831–​41.

21.6  Chronic kidney disease 4859 Rossing P (2006). Diabetic nephropathy:  worldwide epidemic and effects of current treatment on natural history. Curr Diab Rep, 6, 479–​83. Singh AK, Kari JA (2013). Metabolic syndrome and chronic kidney disease. Curr Opin Nephrol Hypertens, 22, 198–​203. Turin TC, et al. (2013). Proteinuria and life expectancy. Am J Kidney Dis, 61, 646–​68. Van der Velde M, et al. (2011) Lower estimated glomerular filtration rate and higher albuminuria are associated with all-​cause and car- diovascular mortality. A  collaborative meta-​analysis of high-​risk population cohorts. Kidney Int, 79, 1341–​52. Progression of CKD and its prevention Appel LJ, et al. (2010). Intensive blood-​pressure control in hyperten- sive chronic kidney disease. N Engl J Med, 363, 918–​29. Baigent C, et al. (2011). The effects of lowering LDL cholesterol with simvastatin plus ezetimibe in patients with chronic kidney disease (Study of Heart and Renal Protection):  a randomised placebo-​ controlled trial. Lancet, 377, 2181–​92. Barnett AH, et  al. (2004). Angiotensin-​receptor blockade versus converting-​enzyme inhibition in type 2 diabetes and nephropathy. N Engl J Med, 351, 1952–​61. Brenner BM, Meyer TW, Hostetter TH (1982). Dietary protein in- take and the progressive nature of kidney disease: the role of hemo- dynamically mediated glomerular injury in the pathogenesis of progressive glomerular sclerosis in aging, renal ablation, and in- trinsic renal disease. N Engl J Med, 307, 652–​9. Casas JP, et al. (2005). Effect of inhibitors of the renin–​angiotensin system and other antihypertensive drugs on renal outcomes: sys- tematic review and meta-​analysis. Lancet, 366, 2026–​33. Chronic Kidney Disease Progression Consortium (2010). Association of estimated glomerular filtration rate and albuminuria with all-​ cause and cardiovascular mortality in general population cohorts: a collaborative meta-​analysis. Lancet, 375, 2073–​81. Gansevoort RT, et al. (2013). Chronic kidney disease and cardiovas- cular risk: epidemiology, mechanisms, and prevention. Lancet, 383, 339–​52. Jay CL, et al. (2016) Reassessing pre-​emptive kidney transplantation in the United States: are we making progress. Transplantation, 100, 1120–​7. Kasiske BL, et al. (2002). Pre-​emptive kidney transplantation: the ad- vantage and the advantaged. J Am Soc Nephrol, 5, 1358–​64. Kidney Disease:  Improving Global Outcomes (KDIGO) Lipid Work Group (2013). KDIGO Clinical Practice Guideline for Lipid Management in Chronic Kidney Disease. Kidney Int Suppl, 3, 259–​305. Klahr S, et al. (1994). The effects of dietary protein restriction and blood-​pressure control on the progression of chronic renal disease. Modification of Diet in Renal Disease Study Group. N Engl J Med, 330, 877–​84. Lewis EJ, et  al. (2001). Renoprotective effect of the angiotensin-​ receptor antagonist irbesartan in patients with nephropathy due to type 2 diabetes. N Engl J Med, 345, 851–​60. Muntner P, et al. (2009). Hypertension awareness, treatment, and con- trol in adults with CKD: results from the chronic renal insufficiency cohort (CRIC) study. Am J Kidney Dis, 55, 441–​51. National Institute for Health and Care Excellence (2014). Cardiovascular disease:  risk assessment and reduction, including lipid modification. Clinical guideline [CG181]. NICE, London. Palmer SC, et al. (2014). HMG CoA reducatase inhibitors (statins) for people with chronic kidney disease not receiving dialysis. Cochrane Database Syst Rev, 5, CD007784. Peterson JC, et  al. (1995). Blood pressure control, proteinuria, and the progression of renal disease. The Modification of Diet in Renal Disease Study. Ann Intern Med, 123, 754–​62. Pohl MA, et  al. (2005). Independent and additive impact of blood pressure control and angiotensin-​II receptor blockade on renal out- comes in the irbesartan diabetic nephropathy trial: clinical implica- tions and limitations. J Am Soc Nephrol, 16, 3027–​37. Remuzzi G, Chiurchiu C, Ruggenenti P (2004). Proteinuria predicting outcome in renal disease:  nondiabetic nephropathies (REIN). Kidney Int Suppl, 92, S90–​6. Sarnak MJ, et al. (2005). The effect of a lower target blood pressure on the progression of kidney disease: long-​term follow-​up of the modi- fication of diet in renal disease study. Ann Intern Med, 142, 342–​51. Wanner C, Tonelli M (2014). KDIGO Clinical Practice Guideline for Lipid Management in CKD: summary of recommendation state- ments and clinical approach to the patient. Kidney Int, 85, 1303–​9. Mineral and bone disorder Arnold A, et  al. (1995). Monoclonality of parathyroid tumors in chronic renal failure and in primary parathyroid hyperplasia. J Clin Invest, 95, 2047–​54. Block GA, et al. (2007). Mortality effect of coronary calcification and phosphate binder choice in incident hemodialysis patients. Kidney Int, 71, 438–​41. Bover J, et al. (2016). Clinical and practical use of calcimimetics in dia- lysis patients with secondary hyperparathyroidism. Clin J Am Soc Nephrol, 11, 161–​74. Briet M, et al. (2012). Arterial stiffness and pulse pressure in CKD and ESRD. Kidney Int, 82, 388–​400. Cunningham J, Locatelli F, Rodriguez M (2011). Secondary hyper- parathyroidism: Pathogenesis, disease progression, and therapeutic options. Clin J Am Soc Nephrol, 6, 913–​21. Eknoyan G, Levin A, Levin NW (2003). Bone metabolism and disease in chronic kidney disease. Am J Kidney Dis, 42, 1–​201. EVOLVE Trial Investigators (2012). Effect of cinacalcet on cardio- vascular disease in patients undergoing dialysis. N Engl J Med, 367, 2482–​94. Greeviroj P, et al. (2018). Cinacalcet for treatment of chronic kidney disease-mineral and bone disorder: a meta-analysis of randomised controlled trials. Nephron, 139, 197–210. Hollis BW (2005). Circulating 25-​hydroxyvitamin D levels indicative of vitamin D sufficiency: implications for establishing a new effective dietary intake recommendation for vitamin D. J Nutr, 135, 317–​22. Hutchison AJ, Smith CP, Brenchley PE (2011). Pharmacology, efficacy and safety of oral phosphate binders. Nat Rev Nephrol, 7, 578–​89. Kazama JJ, et al. (2015). Chronic kidney disease and bone metabolism. J Bone Miner Metab, 33, 245–​52. Ketteler M, Gross ML, Ritz E (2005). Calcification and cardiovascular problems in renal failure. Kidney Int, 67, S120–​7. Kidney Disease: Improving Global Outcomes (KDIGO) CKD-​MBD Work Group (2017). KDIGO 2017 clinical practice guideline update for the diagnosis, evaluation, prevention, and treatment of chronic kidney disease—​mineral and bone disorder (CKD-​MBD). Kidney Int Suppl, 7, 1–59. National Institute for Health and Care Excellence (2007). Cinacalcet for the treatment of secondary hyperparathyroidism in patients with end-​stage renal disease on maintenance dialysis therapy. Technology appraisal guidance [TA117]. NICE, London. Navaneethan SD, et  al. (2009). Benefits and harms of phosphate binders in CKD: a systematic review of randomized controlled trials. Am J Kidney Dis, 54, 619–​37.

section 21  Disorders of the kidney and urinary tract 4860 Razzaque MS (2009). The FGF23-​Klotho axis: endocrine regulation of phosphate homeostasis. Nat Rev Endocrinol, 5, 611–​19. Rodriguez M, et al. (2015). The use of calcimimetics for the treatment of secondary hyperparathyroidism: a 10 year evidence review. Semin Dial, 28, 497–​507. Schmitt CP, Odenwald T, Ritz E (2006). Calcium, calcium regulatory hormones, and calcimimetics: impact on cardiovascular mortality. J Am Soc Nephrol, 17, S78–​80. Tonelli M, Pannu N, Manns B (2010). Oral phosphate binders in pa- tients with kidney failure. N Engl J Med, 362, 1312–​24. Urena-​Torres P, et al. (2007). Klotho: an antiaging protein involved in mineral and vitamin D metabolism. Kidney Int, 71, 730–​7. Renal anaemia Besarab A, et al. (2000). Optimization of epoetin therapy with intra- venous iron therapy in hemodialysis patients. J Am Soc Nephrol, 11, 530–​8. Chen N, et al. (2019). Roxadustat treatment for anemia in patients with kidney disease not receiving dialysis. N Eng J Med, 381, 1001–10. Chen N, et al. (2019). Roxadustat treatment for anemia in patients undergoing long-term dialysis. N Eng J Med, 381, 1011–22. Drueke TB, et al. (2006). Normalization of hemoglobin level in patients with chronic kidney disease and anemia. N Engl J Med, 355, 2071–​84. Hsu CY, Mcculloch CE, Curhan GC (2002). Iron status and hemo- globin level in chronic renal insufficiency. J Am Soc Nephrol, 13, 2783–​6. Kaufman JS, et al. (2001). Diagnostic value of iron indices in hemodi- alysis patients receiving epoetin. Kidney Int, 60, 300–​8. Kazmi WH, et al. (2001). Anemia: an early complication of chronic renal insufficiency. Am J Kidney Dis, 38, 803–​12. Li S, Collins AJ (2004). Association of hematocrit value with cardio- vascular morbidity and mortality in incident hemodialysis patients. Kidney Int, 65, 626–​33. MacDougall IC, et  al. (2009). A peptide-​based erythropoietin-​ receptor agonist for pure red-​cell aplasia. N Engl J Med, 361, 1848–​55. National Collaborating Centre for Chronic Conditions (2006). Anaemia management in chronic kidney disease:  national clinical guideline for management in adults and children. Royal College of Physicians, London. National Institute for Health and Care Excellence (2014). Chronic kidney disease. Managing anaemia. NICE guideline [NG8]. NICE, London. Van Wyck D, et al. (2005). A randomized, controlled trial comparing IV iron sucrose to oral iron in anaemic patients with nondialysis-​ dependent CKD. Kidney Int, 68, 2846–​56.