19.3 Clinical investigation 4395 Michael Doherty a

19.3 Clinical investigation 4395 Michael Doherty and Peter C. Lanyon

ESSENTIALS Laboratory and imaging markers are an adjunct to competent clin- ical assessment and should not be used as a substitute. Tests should only be ordered if the results will alter diagnosis, prognosis, or clinical management. Synovial fluid examination—​this is the key investigation to confirm the diagnosis of either acute crystal or septic arthritis. Fluid can usu- ally be obtained by direct aspiration from any peripheral joint, or alternatively under ultrasound guidance. The identification of crystals requires compensated polarized light microscopy. Plain radiographs—​these remain the single most useful imaging technique, enabling the detection of soft-​tissue swelling, changes in bone density, and cartilage and bone erosion or remodelling, which in conjunction with the pattern of joint sites involved can aid con- firmation of diagnosis and assessment of disease extent. The car- dinal radiological features of rheumatoid arthritis are osteopenia and cartilage/​bone erosion: the features of osteoarthritis are preserved bone density, joint-​space narrowing, osteophyte formation, and bone cysts. Other imaging techniques—​(1) magnetic resonance imaging pro- vides additional benefit with plain radiographs in the assessment of the anatomy and biochemistry of soft tissues as well as bone; (2)  ultrasound is emerging as an effective bedside technique for detecting joint effusions (particularly at clinically occult sites), crystal deposits and tophi, and to assess joint erosions and neovascularity. Blood tests—​(1) inflammatory markers:  C-​reactive protein is the single most useful measure of the acute phase response, having greater reproducibility and greater sensitivity to change than the erythrocyte sedimentation rate; (2) specific antibodies: antibodies to cyclic citrullinated peptides are a marker associated with rheuma- toid arthritis, with similar sensitivity to rheumatoid factor but higher specificity for distinguishing between rheumatoid arthritis and other rheumatic diseases; antibodies detected against nuclear components have high sensitivity for connective tissue diseases (e.g. systemic lupus erythematosus) but low specificity, and hence a positive result does not confirm the diagnosis unless appropriate clinical features are present. Compared with antinuclear antibodies, antibodies to ex- tractable nuclear antigens have higher specificity and associate with different patterns of system involvement within the same disease. Introduction Disease markers are pathological or physiological characteris- tics of an individual that assist in determining the diagnosis, the current activity of disease, or the expected prognosis of the condition in that individual (Fig. 19.3.1). Some markers relate to just one of these elements; others may relate to two, or occasion- ally all three. Clinical markers are derived from enquiry and examination of the patient. For many common rheumatic disorders, clinical assessment alone gives sufficient information for patient diag- nosis and management. In some situations, however, particularly with inflammatory, metabolic, or multisystem disease, a search for additional investigational markers may be warranted. It is im- portant to emphasize that the requirement for and selection of investigations, as well as their subsequent interpretation, is prin- cipally determined by the clinical assessment. Investigations are an adjunct, never a substitute, for competent clinical assessment. There is no place for a battery of screening tests. Investigational markers may include: • Laboratory markers (biochemical, haematological, microbio- logical, histological) sought through investigation of body fluids and tissues. 19.3 Clinical investigation Michael Doherty and Peter C. Lanyon Diagnosis Disease markers Clinical Laboratory (e.g. biochemical, histological) Structural, physiological (imaging) Genetic Prognosis Activity Fig. 19.3.1  Markers may be used for diagnosis, assessment of disease activity, or prognosis.

section 19  Rheumatological disorders 4396 • Structural and physiological markers, mainly assessed by imaging (radiography, scintigraphy, magnetic resonance imaging (MRI), ultrasonography). • Genetic disease susceptibility and prognostic markers—​these hold promise for the future but at present have clinical application only to rare monogenic disorders When considering any investigation, the following deliberations are pertinent. • ‘Is this the most appropriate investigation to answer the clinical question?’ This may depend on various factors, for example, the sensitivity and specificity of the marker being sought, its predictive value (which takes into account disease prevalence as well as the sensitivity and specificity), the cost and availability of the investi- gation, and the pros and cons of invasive versus non​invasive tests. • ‘Will the result of this test alter the diagnosis or clinical manage- ment of the patient?’ It is easy to initiate more investigations than are really required. • ‘Will I be able to interpret and act on the results of this test?’ Tests should only be ordered if the implications of either a normal or abnormal result are understood. In common rheumatological practice, the investigations that are of most use in diagnosis are synovial fluid analysis and the plain radiograph. Confirmation of clinically assessed inflammatory dis- ease activity and its response to treatment is mainly by full blood count and either direct or indirect measures of the acute phase re- sponse. These investigations are therefore given special prominence in this chapter: the usefulness of other investigations will be dis- cussed in the context of specific clinical scenarios. Synovial fluid analysis This is the key investigation to confirm the diagnosis of the two curable rheumatic diseases: septic arthritis and gout. Other crystal-​ associated arthropathies and intra-​articular bleeding are also diagnosed in this way. Synovial fluid analysis is thus the pivotal investigation for an acute monoarthritis, especially with overlying erythema. Synovial fluid can be obtained from almost any peripheral joint, with only a small volume required for diagnostic purposes. Aspiration of large joints should be no more uncomfortable than venepuncture. The patient should be informed of the purpose and nature of the procedure and positioned on a couch in a comfortable and relaxed position, with full exposure of the relevant joint. The risk of introducing sepsis is negligible, as long as sterile equipment and the same sensible precautions used for venepuncture are employed. Macroscopic appearance Normal synovial fluid is present in small volume, contains very few cells, is clear, colourless to pale yellow, and has high viscosity due to macromolecular hyaluronan (Fig. 19.3.2). In general, with increasing joint inflammation the volume increases, the total cell count and proportion of neutrophils rises (causing turbidity), and the viscosity lowers (due to degradation of hyaluronan by pro- tease). However, there is such overlap between arthropathies that these features are of little diagnostic value. Frank pus or pyarthrosis due to very high neutrophil counts should always lead to exclu- sion of sepsis, but can occur with any florid synovitis such as acute crystal synovitis or rheumatoid. High concentrations of urate or cholesterol crystals may result in white synovial fluid, known as joint ‘milk’. Non​uniform bloodstaining of synovial fluid is common and re- flects inconsequential needle trauma to synovial vessels. Uniform bloodstaining (haemarthrosis) most commonly occurs in associ- ation with florid synovitis but may also result from a bleeding di- athesis, trauma, or pigmented villonodular synovitis. A lipid layer floating above bloodstained fluid is diagnostic of intra-​articular fracture. Gram stain and culture Synovial fluid should be sent for urgent Gram stain and culture if sepsis is suspected. Placement in blood culture bottles in addition to a sterile universal container may be indicated if there is likely to be a delay in processing the sample. If gonococcal sepsis or uncommon organisms are suspected, es- pecially in immunocompromized patients, it is advisable to discuss this with the microbiologist so that the optimal cultures can be estab- lished and molecular techniques of antigen detection used if appro- priate. Although a positive result on Gram staining is found in over 50% of cases of adult septic arthritis (predominantly Staphylococcus aureus), a negative result does not exclude infection. If there is a strong clinical suspicion of sepsis, the patient should be given intra- venous antibiotics pending the results of synovial fluid, blood, and other culture results. Crystal identification Accurate identification of common synovial fluid crystals requires a compensated polarized light microscope and an experienced ob- server. Monosodium urate and calcium pyrophosphate crystals may be seen by ordinary light microscopy, but confident identifi- cation resides in their light characteristics as well as their morph- ology. Analysis is best performed on fresh unrefrigerated synovial fluid taken into a plain container to avoid problems of crystal Fig. 19.3.2  Different macroscopic appearances of synovial fluids: (a) on the left, clear straw-​coloured fluid from an osteoarthritic knee (easy to read writing behind it); (b) less viscous, turbid (high cell count) ‘inflammatory’ fluid from a rheumatoid knee; and (c) uniform bloodstaining (haemarthrosis) due to acute pseudogout.

19.3  Clinical investigation 4397 dissolution, post-​aspiration crystallization, and artefacts from tube additives. If only a few drops are obtained, these should be placed straight onto a clean microscope slide and a second slide or cover- slip placed on top. Even with an apparently ‘dry tap’, it is worth ex- pelling the contents of the needle on to a slide: a very small amount of fluid is sometimes obtained and may be diagnostic. Urate crys- tals are long and needle-​shaped and show a strong intensity with negative birefringence (Fig. 19.3.3). Pyrophosphate crystals are smaller, rhomboid in shape, usually less numerous than urate, are often non​birefringent but can also show weak intensity and posi- tive birefringence (Fig. 19.3.4). Although usually identified in the setting of acute synovitis, crystals are also often present in fluid aspirated from the joint after the attack has settled. Aspiration of an asymptomatic first metatarsophalangeal joint (gout) or knee (gout, pseudogout) may therefore permit confirmation of a suspected diagnosis. This is par- ticularly important in gout because of the possible implications of lifelong urate-​lowering therapy. The diagnosis can also be made by analysis of a tophus aspirate. Plain radiography In conjunction with a full history and examination, plain radiog- raphy remains the single most useful and widely available imaging technique for assessment of rheumatic disease. Although a radio- graph is a static record of predominantly past events, it can dem- onstrate visually alterations that reflect the underlying pathological processes of rheumatic disease (e.g. cartilage and bone erosion, bone remodelling, calcification). The abnormalities that may be seen on a plain film include: • soft-​tissue swelling—​seen as altered skin contours and displaced fat planes and intracapsular fat pads (fat appears dark on a radiograph) • decreased or increased bone density (localized or generalized) (Table 19.3.1) • joint erosion (non​proliferative or proliferative marginal erosion, central erosion) • joint-​space narrowing (osteoarthritis—​focal; inflammatory arthritis—​generalized) • new bone formation (osteophyte, enthesophyte, syndesmophyte) • periosteal reaction (Table 19.3.2) • calcification (cartilage (chondrocalcinosis), synovium, capsule, ligament, tendon, muscle, fat, vascular, skin) • bone cysts and radiolucent lesions (Box 19.3.1) • intra-​articular osteochondral bodies • deformity Fig. 19.3.3  Monosodium urate crystals viewed by compensated polarized light microscopy (×400) showing bright birefringence (negative sign) and needle-​shaped morphology. Fig. 19.3.4  Calcium pyrophosphate crystals viewed by polarized light microscopy (×400) showing weak birefringence (positive sign), scant numbers, and a predominantly rhomboid morphology. These are clearly more difficult to detect than urate crystals. Table 19.3.1  Some causes of changes in bone density Causes of increased bone density Causes of decreased bone density Generalized/​multiple regional Myelofibrosis Osteoporosis Osteopetrosis Myeloma, leukaemia Osteomalacia, rickets Hyperparathyroidism Vitamin C deficiency Osteogenesis imperfecta Localized Paget’s disease (with altered trabecular pattern and radiolucent areas) Inflammatory arthritis (juxta-​articular) Metastases (especially prostate, breast) Infection Osteoid osteoma (sometimes with a central radiolucency) Complex regional pain syndrome (algodystrophy) Extreme disuse Bone islands Table 19.3.2  Some causes of periosteal reaction Localized Infection Trauma Tumour Multiple sites Hypertrophic osteoarthropathy Seronegative spondyloarthropathy Scurvy

section 19  Rheumatological disorders 4398 Although most of these abnormalities taken individually have low specificity, various combinations of some of these features, to- gether with their targeting of certain joint sites (Fig. 19.3.5), result in characteristic patterns of abnormality and distribution that have high diagnostic specificity. The distribution of joint involvement, of course, is usually apparent following clinical assessment of the pa- tient, and joints to be investigated by radiography will usually be selected on this basis. However, an important exception is seronega- tive spondyloarthropathy, in which sacroiliac involvement is often asymptomatic and is difficult to detect clinically. When this condi- tion is suspected, an anteroposterior view of the pelvis and a lateral thoracolumbar spine view (i.e. two films) are usually sufficient to show sacroiliitis and syndesmophytes if these are present. Radiographs should be selected to answer specific questions. For example, to address the question of whether a patient with chronic inflammatory polyarthritis affecting hands, elbows, neck, knees, and ankles has erosive disease typical of rheumatoid, posteroanterior views of hands and feet (i.e. two films), but not radiographs of all symptomatic joints, are appropriate. This is because rheumatoid ero- sions appear first in wrists and the small joints of hands and feet, and may first affect the metatarsophalangeal joints, even if they are rela- tively asymptomatic. However, if the degree of structural damage in one large joint is a principal cause for concern, then a radiograph of that particular joint should obviously be taken. For most joints, a single (two-​dimensional) view is sufficient (e.g. anteroposterior view of pelvis, posteroanterior view of both hands, posteroanterior view of both feet), although two views are required for some (e.g. posteroanterior weight-​bearing (semi-​flexed) view of both knees, plus individual lateral or bilateral skyline patello-​femoral view). Thus selection of radiographs will often differ for purposes of diag- nosis or disease assessment. Erosions An important hallmark of inflammatory arthropathies is cartilage and bone erosion. Intracapsular bone erosion first occurs at the ‘bare areas’ of the joint margin (marginal erosion) where bone is ex- posed directly to inflammatory synovium without the protection of overlying cartilage. Loss of the sharp cortical line, the ‘dot-​dash’ ap- pearance, is the first radiographic sign that precedes more definite scalloping of the bony contour (Fig. 19.3.6). Cartilage erosion also commences at the joint margin and slowly works centrally, resulting in relatively late loss of interosseous distance or ‘joint space’. Both rheumatoid disease and the seronegative spondyloar­ thropathies (especially psoriatic and chronic reactive arthritis) cause marginal erosions. In rheumatoid disease, however, the ag- gressive synovitis overwhelms any reparative response, presenting a very atrophic appearance (‘non​proliferative erosions’) (Figs. 19.3.6 and 19.3.7) with no new bone or periosteal reaction and Box 19.3.1  Some causes of radiolucent lesions • Bone cysts (isolated or in association with osteoarthritis) • Inflammatory arthritis (erosions) • Infection • Metastases (especially breast, lung, kidney, thyroid) • Myeloma • Osteochondromata • Osteogenic sarcoma • Histiocytosis X (a) (b) (c) Fig. 19.3.5  Diagram to show different target sites of involvement in the forefoot for (a) rheumatoid arthritis, (b) psoriatic arthritis, and (c) osteoarthritis. (d) (c) (b) (a) Fig. 19.3.6  Diagram of metacarpophalangeal joint showing (a) early dot-​dash erosion, (b) the later definite nonproliferative erosion of rheumatoid arthritis, (c) the proliferative erosion of psoriatic arthritis, and (d) the intra and extracapsular pressure erosions of gout. Fig. 19.3.7  Radiograph of metacarpophalangeal joint showing late
non​proliferative marginal erosions of rheumatoid arthritis, more obvious proximally than distally (reflecting the more proximal than distal distribution of synovium in small finger joints) and eventual global loss of cartilage.

19.3  Clinical investigation 4399 only juxta-​articular osteopenia (a sign of inflammation) and soft-​ tissue swelling as accompanying early radiographic features. By con- trast, the seronegative spondyloarthropathies are characterized by a degree of low-​grade inflammation that permits some reparative response. Such inflammation results in a tendency to fibrosis, calci- fication, and ossification. Marginal erosions in these arthropathies are therefore commonly accompanied by fluffy new bone forma- tion (‘proliferative erosions’) (Figs. 19.3.6 and 19.3.8) with normal or increased periosteal and bone density rather than osteopenia. The fact that different joints are targeted in these conditions, and the common accompanying involvement of entheses—​fibrous in- sertions of tendons, ligaments, or capsule into bone—​further assists differentiation in most cases. In early septic arthritis the radiograph is often normal, apart from osteopenia and soft-​tissue swelling, for one to two weeks. However, erosion proceeds rapidly and results in generalized loss of joint space with loss of cortical integrity centrally (central erosion) as well as marginally. In chronic gout, bony defects develop slowly as massive crystal concretions (tophi) causing pressure necrosis to surrounding bone; such pressure erosions (Fig. 19.3.6) occur at extracapsular as well as intracapsular sites and are unaccompanied by osteopenia. Osteoarthritis The features of osteoarthritis, by far the most common joint disease, are highly characteristic and contrast with those of inflammatory arthropathy. The two cardinal features are narrowing of the joint space and osteophytes. By contrast to inflammatory arthropathies, joint-​space narrowing is focal rather than widespread within the joint, mainly targeting the maximum load-​bearing region (Fig. 19.3.9). Bony osteophyte is most noticeable at the margins of the joint but also occurs centrally and as periosteal osteophyte (‘buttressing’) at sites such as the femoral neck. Subchondral scler- osis, or increased density of bone is also common, principally below the site of maximal narrowing. Additional features include subchondral ‘cysts’, osteochondral (‘loose’) bodies within the synovium, and an increased association with chondrocalcinosis. In contrast to inflammatory arthritis, the bone density is normal or in- creased and marginal erosions are not a feature. Calcification Calcification can affect any locomotor tissue. Calcification of fibro and hyaline cartilage (chondrocalcinosis) is most commonly due to calcium pyrophosphate crystals, less commonly to apatite or other basic calcium phosphates. This can occur as an isolated phenom- enon (mainly age associated, rarely as a result of metabolic or fa- milial disease predisposition) or in association with structural changes of osteoarthritis (chronic pyrophosphate arthropathy). Less commonly, pyrophosphate crystals also cause calcification of the synovium and capsule, and linear tendon calcification (mainly hip adductors, Achilles, triceps). Periarticular calcification is usually apatite. Isolated periarticular calcification mainly affects central sites such as the shoulder (supraspinatus tendons) or hip (abductor tendons), appearing as single dense concretions with rounded contours, as opposed to the linear calcification of pyrophosphate. Shedding of these crystal deposits can result in severe, self-​limiting inflammation (acute calcific periarthritis) with reduction or loss of the radiographic calcification. Spotty, multiple calcification of soft tissues (calcinosis) mainly targets peripheral and intermediate sites such as the finger pulps, wrists, and forearms and is a feature of connective tissue disease, most commonly CREST syndrome (calcinosis, Raynaud’s, oesopha- geal dysmotility, sclerodactyly, telangiectasia). Calcinosis requires distinction from small blood vessel calcification, which has a thin, meandering tramline appearance (and is increased in diabetes and chronic renal failure), sesamoids, and solitary dense calcified phleboliths. Myositis ossificans is rare and appears as dense sheets of calcification mainly at proximal sites such as the hip. Fine reticular or linear calcification of subcutaneous fat and muscle may follow young onset dermatomyositis. Fig. 19.3.8  Radiograph of the hallux showing proliferative erosions and cartilage loss of the interphalangeal joint, and associated increased bone density (ivory phalanx) typical of psoriatic arthropathy. Fig. 19.3.9  Radiograph of the hip to show changes of osteoarthritis, specifically superior joint-​space narrowing, subchondral sclerosis, marginal osteophyte, and cysts.

section 19  Rheumatological disorders 4400 Other imaging Scintigraphy Scintigraphy is a cheap, readily available technique that delivers only a very small amount of radiation. It involves gamma camera imaging following an intravenous injection of radioisotope, usually technetium-​99m diphosphonate. Early ‘flow’ images obtained im- mediately post-​injection, or a little later when the isotope is in the soft tissues (‘blood pool’ phase), reflect vascularity and will show, for example, the increased perfusion of inflamed synovium, Pagetic bone, or hypervascular primary or secondary bone tumour (Fig. 19.3.10). ‘Delayed’ images, taken a few hours after injection, indicate bone remodelling due to localization of the diphosphonate to sites of active bone turnover. Although non​specific and lacking high spatial resolution, the major advantage of scintigraphy is its high sensitivity for detecting important bone and joint pathology that may not be apparent on plain radiographs. However, although this technique has predominately been superseded by MRI in the assessment of a single region, it can still be used to detect: • bone metastases at clinically occult sites • bone or joint sepsis at clinically occult sites • early osteonecrosis at the presenting site and at clinically occult sites • The extent and current activity of Paget’s disease of bone • complex regional pain syndrome (reflex sympathetic dystrophy, algodystrophy) • hypertrophic osteoarthropathy Computed tomography (CT) This can give detailed information on anatomy, especially of bone, allowing three-​dimensional visualization of structures such as the spinal canal and facet joints. Its principal use is therefore in assessing areas of complex anatomy such as the spine or pelvis, where plain radiographs may be inadequate (e.g. to investigate stenosis of the spinal canal). Drawbacks, however, include limited soft-​tissue reso- lution and exposure to a considerable radiation dose; in many situ- ations it is has now been superseded by MRI. It still has advantages to MRI in specific situations; for example, visualization of complex fractures including detection of osseous fragments, anatomical mapping prior to complex bone surgery, and visualizing joints in the the thorax where MRI would be degraded by respiration artefacts (costovertebral and sternoclavicular). Magnetic resonance imaging (MRI) The ability of MRI to image the anatomy and biochemistry of soft tissue as well as bone means that it provides detailed information not only on structure but also on the pathophysiology of all locomotor tissues. Further advantages include its capacity for multiplanar imaging (e.g. coronal, axial, sagittal, oblique) and its safety, without radiation exposure. The physics of MRI is complex. When a pa- tient is placed in the magnetic field of the scanner, the protons in the body align along the central axis of the field. Application of a radiofrequency pulse, or sequence, causes the protons to spin in phase with each other. When the pulse is stopped, the protons re- turn to random spinning and dephase. As they do so, they emit a signal that is converted to an image by computer manipulation. In general, T1-​weighted short sequences are useful for defining anatomy, and T2-​weighted long sequences are useful for assessing pathology. Other sequences are selected for special purposes (e.g. the short tau inversion recovery sequence is used to image marrow as it suppresses fat and makes the marrow appear dark). MRI, with or without enhancement with gadolinium, is particularly useful in detecting and assessing the following: • early osteonecrosis at the presenting site and the contralateral clinically occult site • intervertebral disc disease, root entrapment, and spinal cord compression • osteoarticular and soft-​tissue sepsis • osteoarticular and soft-​tissue malignancy • internal mechanical derangement of joints (particularly the knee) • assessment of soft-​tissue and periarticular pathology (e.g. early synovitis, rotator cuff tears bursitis, tenosynovitis) • stress fracture • complex regional pain syndrome (reflex sympathetic dystrophy, algodystrophy) The choice between three-​phase scintigraphy and MRI for de- tection of conditions such as early osteonecrosis, where both have excellent sensitivity (scintigraphy 90%, MRI 100%), will depend on practical issues such as ease of access, musculoskeletal reporting ex- pertise, radiation risk, and local cost. Ultrasonography This is a safe, accessible technique that can be used at the bedside and is increasingly used to detect synovitis/​effusions at both clin- ically occult joints (e.g. hip) and small peripheral joints (e.g. prox- imal interphalyngeal joints) when there is clinical uncertainty. It can also be used to assess for neovascularity (a positive Doppler signal Fig. 19.3.10  Bone scan demonstrating secondary deposits of prostate cancer. The presenting painful lesion was in the right hemipelvis and the plain pelvic radiograph was normal. The spinal lesion (and two others not shown on this photograph) were asymptomatic.

19.3  Clinical investigation 4401 indicates increased blood perfusion) and erosions, and hence is im- portant in assessing response to treatment. Crystal deposits in car- tilage (urate at the surface, pyrophosphate in the mid-​zone) may also show as a specific hyperechoic double contour sign, and tophi (often mobile on pressure) may be evident in synovium and extracapsular sites. Ultrasound examination is also very useful to assess soft-​tissue changes such as popliteal cysts, or Achilles tendon pathology, and can also be used to guide synovial fluid aspiration and intra-​articular and soft-​tissue injection. Arthrography The main use of this technique, which requires injection of positive (iodinated) or negative (air) contrast, or a combination of both, is in addition to MRI to delineate labral tears within the shoulder or hip capsule. Blood tests for inflammation and systemic disease The full blood count, erythrocyte sedimentation rate (ESR), and C-​reactive protein may show changes that indicate the presence of inflammation somewhere in the body. These changes are very sensi- tive but are non​specific: they are mainly used as a semiquantitative measure to complement the clinical assessment of inflammatory disease and its response to treatment. Detection of an acute phase response The systemic response to injury that results in these changes is summarized in Fig. 19.3.11. At any site of injury or inflammation, macrophages and monocytes release soluble intercellular signalling polypeptides (cytokines) including interleukin 1, interleukin 6, and tumour necrosis factor-​α. Some of these enter the systemic circula- tion and exert effects on the hypothalamus, bone marrow, and liver. These combined systemic effects are called the acute phase response, even though they accompany chronic as well as acute inflammation. Interleukin 6 is the main cytokine to influence the liver, causing increased production of certain acute phase proteins (including fi- brinogen and C-​reactive protein) and decreased production of other negative acute phase reactants (such as albumin and transferrin). Much of the acute phase response is beneficial for body defence and adaptation to injury, especially for dealing with the two major complications of injury that threaten life: haemorrhage and sepsis. For example, the thrombocytosis and increased serum levels of clot- ting factors facilitate haemostasis; neutrophilia and the increased serum levels of complement, immunoglobulin, and C-​reactive pro- tein (an opsonin) combat infection; and the anaemia and low serum transferrin levels result in diminished delivery of iron to bacteria and parasites. C-​reactive protein Of all the acute phase proteins, C-​reactive protein shows the greatest shift from very low to very high levels, varying over a several-​ hundredfold range in concentration. C-​reactive protein also closely mirrors the current degree of inflammation, rising rapidly at its onset and falling as inflammation subsides, such that it is therefore the single most useful direct measure of the acute phase response. Interestingly, some rheumatic diseases—​specifically lupus, sys- temic sclerosis, and dermatomyositis—​associate with only modest or no elevation of C-​reactive protein, despite unequivocal patho- logical evidence of inflammation and tissue damage. The reason for this remains unclear, but patients with such disease are capable of mounting a typical acute phase response (e.g. in response to infec- tion). In a patient with lupus or scleroderma, gross elevation of C-​ reactive protein should therefore suggest an incidental cause such as sepsis. Some clinical features of active systemic lupus and infection overlap; hence, in this situation the C-​reactive protein can prove a useful test. Erythrocyte sedimentation rate The ESR is a long-​established indirect measure of the acute phase re- sponse. It mainly reflects the degree of rouleaux formation. Normally our circulating erythrocytes do not clump together because of the net balance of three electrical forces (Fig. 19.3.12), namely weak at- tractant van der Waal’s forces resulting from red cells being bodies; a strong repellent net negative surface charge, or zeta potential, due mainly to membrane sialic acid residues; and an attractant dielec- tric constant resulting from the charge characteristics of the plasma constituents. In health the zeta potential far exceeds the sum of the two at- tractant forces, so that erythrocytes electrostatically repel each other and remain single. However, during the acute phase response the change in plasma protein concentrations leads to an increase in dielectric constant. Fibrinogen is particularly important in this re- spect. Although its increase in concentration is relatively modest, fibrinogen is a very asymmetric molecule that exerts a major elec- trical charge effect. The resulting increase in dielectric constant is sufficient to overcome the zeta potential, so that rouleaux form more readily. Rouleaux have a higher ratio of mass per surface area and (Rise in ESR) Liver Brain IL-1, IL-6 IL-1, TNF Acute phase reactants, e.g. CRP Reduced albumin synthesis Fibrinogen Fever Slow wave sleep Reduced appetite Inflamed joint IL-8, IL-1 TNF Bone marrow Muscle Wasting IL-1, TNF Neutrophilia Reduced red blood cell production (anaemia) Thrombocytosis Immune system Lymphocyte adherence Platelet activation B,T-cell activation Fig. 19.3.11  Diagram to show the important elements of the acute phase response.

section 19  Rheumatological disorders 4402 so sediment faster than single red cells, which is the property meas- ured in the ESR. In the Westergren test system a 200 mm capillary tube is filled with the patient’s blood. After 1 h, the clearance of red cells from the top is measured. If there is little rouleaux formation, the discrete red cells sediment only slowly and the clearance is small (<5–​10 mm). However, if there is significant rouleaux formation, the clearance is greater and the ESR in the first hour is elevated. Therefore, in a patient with an acute phase response, the ESR and C-​reactive protein are both elevated, with the ESR lagging behind the C-​reactive protein in terms of speed of change. It is important to recognize, however, that the ESR may be elevated for reasons other than the acute phase response. Immunoglobulins are very symmetrical molecules, and their modest increase in con- centration during the acute phase response has relatively little ef- fect, compared with fibrinogen, on the dielectric constant. However, large increases in immunoglobulin concentration (e.g. in multiple myeloma or associated with autoimmune diseases such as Sjögren’s syndrome) will increase the dielectric constant and lead to rouleaux formation. In this situation, the patient may have a high ESR but normal or relatively low C-​reactive protein. Such discordance be- tween the ESR and C-​reactive protein should lead to consideration of hypergammaglobulinaemia and myeloproliferative disease and to direct measurement of serum immunoglobulins and paraprotein electrophoresis. In addition to the changes reflecting an acute phase response, the full blood count may show other alterations that are nonspecific in themselves but which, taken in the context of the clinical features, may be characteristic of certain rheumatic diseases or their com- plications (Fig. 19.3.13). For example, neutrophilia may be seen in systemic vasculitis and neutropenia in lupus. Furthermore, many of the slow-​acting drugs used to control chronic inflammation have toxicity on the bone marrow, such that the full blood count is often included in the routine monitoring of such treatment. Taken together, therefore, the full blood count, ESR, and C-​reactive protein can be a useful complement of tests in the major rheumatic diseases, to: • assess inflammatory disease activity • assess response to disease-​suppressing treatment • detect certain disease complications • screen for drug toxicity Van der Waal’s forces Zeta potential Dielectric constant Rouleaux Mass/surface.area Fig. 19.3.12  Diagram showing the balance of three electrical forces that influence clumping of erythrocytes. Rouleaux sediment faster than individual erythrocytes. Erythrocytes Chronic inflammation Lupus (haemolysis) Drugs Neutrophils Eosinophils Monocytes Lymphocytes Platelets Lupus Felty’s syndrome Drugs Infection Lupus Steroid therapy Drugs Infection Steroid therapy Vasculitis Drugs Psoriatic arthropathy Adult onset Still’s disease Polyarteritis nodosa Sepsis Steroid therapy Felty’s syndrome Lupus Drugs Steroid therapy Infection Inflammation Fig. 19.3.13  Diagram showing some of the nonspecific changes that may occur in individual elements of the full blood count in patients with systemic rheumatic disease.

19.3  Clinical investigation 4403 However, it is important to appreciate that although an elevated acute phase response is consistent with inflammatory rheumatic disease, it is non​specific; also that the degree of elevation is often proportional to the amount or ‘burden’ of inflammatory tissue (e.g. isolated small joint synovitis in rheumatoid arthritis may not be suf- ficient to cause a detectable acute phase response, but this does not mean that inflammatory disease is not present). Immunological tests There are an increasing variety of autoantibodies that can be detected in a serum sample by clinical laboratory services. Production of some of these is a common, age-​related phenomenon that may be exagger- ated by the presence of chronic inflammation. The isolated presence of some autoantibodies therefore has low diagnostic specificity and may not imply the presence of any disease at all. If present in high concen- trations, however, their disease specificity usually increases; hence it is always important to know how much antibody is present (the titre or concentration in units), rather than just whether it is detectable. The titre of some autoantibodies (e.g. antineutrophil cytoplasmic antibody (ANCA)) can reflect the activity of the associated disease and predict relapse risk after treatment; only a few antibodies (e.g. anti-​dsDNA) have high diagnostic specificity. Again, the correct choice and inter- pretation of tests will depend on detailed knowledge of the patient. There are different detection and assay systems for many of these auto- antibodies: close liaison with the local immunology service is required. Rheumatoid factor The definition of a rheumatoid factor is an antibody directed against a specific region of the Fc (crystallizable) fragment of human IgG. The antibody itself may be of any immunoglobulin class, although IgM anti-​IgG is the rheumatoid factor that is most commonly meas- ured in the first instance. One of the traditional methods used to detect IgM rheumatoid factor is to coat latex beads with human IgG. If the patient’s serum is then added to the test system the penta- meric IgM antibody binds to the IgG, causing the latex particles to flocculate, producing a positive latex fixation test. The amount the patient’s serum must be diluted before this flocculation is lost is then determined; the higher this ‘titre’, the higher the concentration of antibody present. Although ‘rheumatoid factor’ was so named be- cause it was first detected in the sera of patients with rheumatoid arthritis, it also occurs in association with a variety of other condi- tions, as well as in some normal adults (Box 19.3.2). It thus has low diagnostic specificity, particularly in older people, and is not a ‘test for rheumatoid arthritis’. In terms of sensitivity, it is present in most patients with erosive rheumatoid disease but may only appear after many months or years of disease, once the diagnosis is beyond dis- pute. It is therefore of little value in making a diagnosis of rheuma- toid arthritis, being neither sufficient nor necessary:  rheumatoid arthritis is predominately a clinical diagnosis, based on detecting the presence of synovitis (capsular swelling, joint line tenderness, stress pain). However, if present in high titre at the onset of rheuma- toid arthritis it associates with a poorer prognosis. IgG rheumatoid factor has greater specificity for major rheumatic disease, but these caveats still remain. One situation in which a negative rheumatoid factor is of diagnostic significance is in a patient with arthritis and nodules. As a general rule all patients with nodular rheumatoid arth- ritis are seropositive (as are those patients who have extra-​articular disease manifestations), so that in this situation, other causes of ‘arthritis plus nodules’ must be considered (e.g. tophaceous gout or hypercholesterolaemia). Anti-​ cyclic citrullinated peptide antibodies A highly specific autoantibody system has been described in rheuma- toid arthritis in which antibodies are detected against modified (citrullinated) arginine residues. The detection of antibodies to cyclic citrullinated peptides (anti-​CCPs) is a relatively new test, with similiar sensitivity for rheumatoid factor (65–​80%) but higher specificity (89–​ 98%). It can also be positive in patients who are negative for rheuma- toid factor, and this is one of the clinical scenarios in which the test is most useful. Anti-​CCP positivity in the setting of a patient with possible early inflammatory arthritis has prognostic significance as it is strongly associated with progression to persistent rheumatoid dis- ease, higher inflammatory markers, and greater erosive change. Unlike rheumatoid factor, anti-​CCP status is unlikely to be influenced by age, the presence of other autoimmune diseases, infection or smoking, and it therefore has a much lower false positivity rate. Antinuclear antibody An antinuclear antibody (ANA) is any autoantibody directed against one or more components of the nucleus. The standard method of detection is immunofluorescence microscopy after serum has been applied to a nucleated tissue substrate, usually either rodent liver, kidney, or a human cell line (e.g. HEp-​2). Four main patterns of staining are reported. As with rheumatoid factor, the higher the titre of antinuclear antibody, the greater its significance, but a high titre does not necessarily imply more severe disease. The specificity and sensitivity also vary according to the antigen preparation used in the test system, but tests are not universally standardized and (again) liaison with the laboratory is important in order to determine the cut-​off titre that is considered to be ‘abnormal’. The many causes of a positive ANA are outlined in Box 19.3.3. The commonest reason to undertake an antinuclear antibody test is in a patient with suspected lupus. For lupus, the antinuclear anti- body has high sensitivity (97–​100%), but because the specificity is very low (10–​40%), a positive result does not make the diagnosis; by contrast, a negative antinuclear antibody virtually excludes it, par- ticularly if the extractable nuclear antigen (ENA) test is negative. If a screening serum antinuclear antibody test is positive, most la- boratories will then attempt to determine the specific antigenic deter- minants. Some of these determinants are soluble and can be extracted Box 19.3.2  Associations with a positive rheumatoid factor • Rheumatoid arthritis (about 75%) • Lupus, scleroderma, Sjögren’s syndrome, dermatomyositis • Chronic infection: —​ Bacterial endocarditis — Viruses (rubella, cytomegalovirus, infectious mononucleosis) • Parasites • Neoplasms—​after irradiation or chemotherapy • Hyperglobulinaemic states: — Hypergammaglobulinaemic purpura — Cryoglobulinaemia — Chronic liver disease Note that normal subjects can be seropositive.

section 19  Rheumatological disorders 4404 from the nucleus, hence ‘extractable nuclear antigens’, although many of the antigen–​antibody specificities in human disease remain to be discovered. Compared with the antinuclear antibodies, antibodies against specific nuclear antigens have higher specificity for certain diagnoses or for certain patterns of system involvement within the same disease. For example, antinuclear antibodies directed against double-​stranded DNA are highly specific for lupus, but present in only a minority of patients, usually those who have more severe disease with a greater likelihood of renal involvement, in whom the diagnosis is usually already clear. Antibodies to Sm antigen occur almost exclu- sively in lupus and are associated with a higher risk of renal disease. Antitopoisomerase 1 and anticentromere antibodies are found in dif- fuse cutaneous and limited cutaneous systemic sclerosis, respectively. Antibodies to Ro antigen occur predominantly in Sjögren’s syndrome (often in combination with anti-​La) and in lupus and are associated with a high frequency of photosensitive rashes and a risk of neo- natal heart block. Antibodies to ribonucleoprotein (RNP) are found in lupus but are often associated with the presence of overlapping connective tissue disorders that may have features of scleroderma, myositis, and rheumatoid arthritis. Antisynthetase autoantibodies (e.g. Jo-​1) are associated with myositis (see Chapter 19.11.5). The antiphospholipid syndrome, defined by the occurrence of ar- terial and venous thromboses, recurrent fetal losses, and thrombo- cytopenia in the presence of antiphospholipid antibodies, occurs in lupus (secondary antiphospholipid syndrome (APS)) and other autoimmune diseases, and also in subjects with no other underlying disease (primary antiphospholipid syndrome). Antiphospholipid antibodies can be detected in assays for anticardiolipin and anti-​β-2 glycoprotein 1 antibodies and in phospholipid-​dependent coagu- lation studies to detect lupus anticoagulants (prolonged activated partial thromboplastin time (APTT) which fails to correct with the addition of normal serum). Antiphospholipid antibodies also occur in a wide variety of rheumatic, infectious (bacterial, viral, proto- zoal) and malignant conditions, and can also be drug induced, but in these situations they are not usually associated with thromboses. Further information on these tests can be found in Chapters 14.14 and 19.11.2. Tests for specific clinical situations Chronic inflammatory disease at a single site Patients with unexplained inflammatory disease at a single loco- motor site (monoarthritis, bursitis, tenosynovitis, osteitis) should be considered for biopsy. The timing for this will vary according to how florid the lesion appears, but in general this should be undertaken for any undiagnosed lesion that has persisted for six months. The reason is to determine or exclude specific disease that can only be diagnosed by this means (Box 19.3.4). Although these conditions are uncommon or rare, they require a specific treatment approach, rather than a continuing empirical symptomatic one. The com- monest site for unexplained inflammatory monoarthritis is the knee, followed by the wrist and small hand joints. Arthroscopic biopsy is ideally used for larger joints (for additional information from direct visualization and guided biopsy), open biopsy for smaller joints, and periarticular lesions. Tissue should be examined histologically and sent for culture, including mycobacteria. Apart from the specific conditions in Box 19.3.4, histopathology has no role in the diagnosis or management of most common rheumatic disease. Investigation of suspected muscle disease There are five principal investigations for the diagnosis and moni- toring of muscle disease: serum creatine kinase, myositis-​specific autoantibodies, electromyography, MRI, and muscle histology. None are 100% sensitive, hence each may be normal despite ab- normality detected by one or both of the others. Although creatine kinase is the most indirect measure, it is readily available and com- monly measured in the first instance, alongside autoantibodies if the test result is raised. It is important to realize that elevation of the creatine kinase may result from a variety of causes (Box 19.3.5), and certain racial groups (e.g. Afro-​Caribbean) have higher ‘normal range’ values. MRI is usually performed next as it is non​invasive and useful for demonstrating soft tissue and muscle changes, including oedema within and around muscles (which although not specific for myositis, is a typical finding in the acute phase), muscle calcification and fatty infiltration/​atrophy (often seen in the chronic phase). Muscle involvement can also be patchy and asymmetric; hence MRI can be used to select muscle groups for electromyog- raphy (EMG) or biopsy. Box 19.3.4  Causes of chronic single-​site synovitis • Foreign body (e.g. plant thorn) • Infection, including tuberculosis, fungi • Sarcoidosis • Amyloidosis • Pigmented villonodular synovitis • Synovial chondromatosis • Synovial sarcoma Box 19.3.5  Causes of elevation of serum creatine kinase • Inflammatory myositis ± vasculitis • Inclusion body myositis • Muscular dystrophy • Motor neuron disease • Alcohol, drugs • Trauma, strenuous exercise • Myocardial infarction • Hypothyroidism, metabolic myopathy Note that rhabdomyolysis is associated with massive elevation of serum creatine kinase. Box 19.3.3  Associations with a positive antinuclear antibody • Systemic lupus erythematosus • Rheumatoid arthritis • Sjögren’s syndrome • Polymyositis • Polyarteritis nodosa • Juvenile idiopathic arthritis • Chronic active hepatitis • Autoimmune thyroid disease • Myaesthenia gravis • Extensive burns Note that normal subjects may be antinuclear antibody positive.

19.3  Clinical investigation 4405 Electromyography or muscle biopsy will be undertaken next, the choice depending on local availability and expertise. How much in- formation on diagnosis and disease activity has been gained from the first of these tests will then often determine whether the second is also undertaken. Electromyography measures the action potentials produced at rest and during voluntary contraction. Normal muscle is electric- ally silent at rest. On slight contraction, motor-​unit potentials of 500–​1000 µV in amplitude and 4–​8 ms in duration are recorded. On maximal contraction, as many motor units as possible are recruited and an interference pattern develops. With inflammatory polymyo- sitis the electromyography may show a diagnostic triad of spontan- eous fibrillation, short-​duration action potentials in a polyphasic disorganized outline, and repetitive bouts of high-​voltage oscilla- tions produced by contact of diseased muscle with the needle. Muscle histology can readily be obtained from a needle muscle biopsy sample. This is a relatively simple procedure requiring a local anaesthetic, small skin incision (no stitches required), an appropriate muscle biopsy needle, and no subsequent limitation of activity: it can easily be repeated serially for subsequent monitoring of response to treatment. The quadriceps is usually chosen, although the deltoid or other muscles can also be biopsied this way. The two or more small cores of tissue obtained need to be transported rapidly to the laboratory and correctly orientated before freezing and sectioning. Immunohistochemical staining in conjunction with plain histology gives considerable information concerning primary and secondary muscle and neuromuscular disease. Although open biopsy will yield more tissue than needle biopsy, only a small amount of muscle is ac- tually required and serial open biopsy is clearly problematic. Investigation of suspected vasculitis The presenting features of vasculitis are usually those of vessel wall inflammation leading to ischaemia in the affected organ (e.g. upper airway, kidney, peripheral nerve, abdomen, skin), in combination with the non​specific systemic effects of inflammation (malaise, weight loss, night sweats). These symptoms can also be produced both by severe infection and by malignancy, and in view of this—​ and the potential toxicity of appropriate treatment—​further in- vestigation is always required. In terms of laboratory tests, simple measures such as a urine dipstick test and microscopy should not be overlooked, as the prognosis of many of these diseases is dictated by renal involvement. Antineutrophil cytoplasmic antibodies, initially detected in pa- tients with glomerulonephritis, are antibodies directed against enzymes present in neutrophil granules. Two main patterns of im- munofluorescence are detected: cytoplasmic and perinuclear. Most c-​ANCAs and p-​ANCAs are specific for the neutrophil enzymes proteinase 3 (PR3) and myeloperoxidase (MPO), respectively. These two antibodies associate with particular vasculitis subtypes, for ex- ample, anti-​PR3 with granulomatosis with polyangiitis (Wegener’s), and anti-​MPO with microscopic polyangiitis. However, positive antineutrophil cytoplasmic antibodies occur in a variety of other settings, including malignancy and infections (bacterial and HIV) as well as other autoimmune diseases (inflammatory bowel disease, rheumatoid arthritis, thyroid disease, pulmonary fibrosis). Equally, some patients with definite systemic vasculitis are antineutrophil cytoplasmic antibody negative, and thus diagnosis of these con- ditions should not be made or refuted on the antineutrophil cytoplasmic antibody test alone. Other supporting evidence should be obtained whenever possible by biopsy of an appropriate organ (nose, kidney, muscle, skin). For further information, see Chapter 19.11.9. Investigation of multiple regional pain In most patients who present with widespread musculoskeletal pain, the diagnosis is made from clinical examination alone (e.g. wide- spread rheumatoid disease). In some cases, however, there may be little to detect on clinical examination to explain the widespread pain. In most cases the diagnosis will be fibromyalgia. This is con- firmed clinically by the appropriate symptoms (e.g. widespread pain, non​restorative sleep, marked fatigue, tension headache, irritable bowel symptoms, anxiety and depression, poor memory and concen- tration, urinary frequency) and the presence of widespread hyper- algesic tender sites and negative control sites (see Chapters 19.2, 26.3.3, and 26.5.12). However, several other conditions may present similarly with multiple regional symptoms and few, if any, physical findings, hence a limited screen (Table 19.3.3) is justified to detect conditions for which there are specific treatments, either as the pri- mary cause or as a comorbidity in people with fibromyalgia. FURTHER READING Berden A, et al. (2012). Diagnosis and management of ANCA associ- ated vasculitis. BMJ, 344, e26. Catrina A, et al. (2014). Lungs, joints and immunity against citrullinated proteins in rheumatoid arthritis. Nat Rev Rheumatol, 10, 645–​53. Lisnevskaia L, Murphy G, Isenberg D (2014). Systemic lupus erythematosus. Lancet, 384, 1878–​88. Marinos C, Dalakas MD (2015). Inflammatory muscle diseases. N Engl J Med, 372, 1734–​47. Mariz HA, et al. (2011). Pattern on the antinuclear antibody–​HEp-​2 test is a critical parameter for discriminating antinuclear antibody–​ positive healthy individuals and patients with autoimmune rheum- atic diseases. Arthritis Rheum, 63, 191–​200. McQueen FM, Doyle A, Dalbeth N (2011). Imaging in gout—​what can we learn from MRI, CT, DECT and US? Arthritis Res Ther, 13, 246. Rowbotham EL, Grainger AJ (2011). Rheumatoid arthritis: ultrasound versus MRI. AJR Am J Roentgenol, 197, 541–​6. Ten Cate DF, Luime JJ, Swen N (2013). Role of ultrasonography in diagnosing early rheumatoid arthritis and remission of rheumatoid arthritis—​a systematic review of the literature. Arthritis Res Ther, 15, R4. van Venrooij WJ1, van Beers JJ, Pruijn GJ (2011). Anti-​CCP anti- bodies: the past, the present and the future. Nat Rev Rheumatol, 7, 391–​8. Table 19.3.3  Investigations to perform in multiple regional pain Investigation Condition ESR/​CRP Systemic inflammation Creatine kinase Myositis Antinuclear antibody Lupus Calcium Parathyroid disease, osteomalacia Thyroid function tests Thyroid disease CRP, C-​reactive protein; ESR, erythrocyte sedimentation rate.