8.6.3 Pneumococcal infections 975

8.6.3 Pneumococcal infections 975

8.6.3  Pneumococcal infections 975 Murray BE (1990). The life and times of the Enterococcus. Clin Microbiol Rev, 3, 46–​65. Royal College of Obstetricians and Gynaecologists (2003). Green Top Guideline 36. Prevention of early onset neonatal group B streptococcal diseases. http://​www.rcog.org.uk/​files/​rcog-​ corp/​uploaded-​files/​GT36GroupBStrep2003.pdf Stevens DL (1992). Invasive group A streptococcus infections. Clin Infect Dis, 14, 2–​13. Stevens DL (1995). Streptococcal toxic shock syndrome: spectrum of disease, pathogenesis and new concepts of treatment. Emerg Infect Dis, 1, 69–​78. Stevens DL (2004). Streptococcal infections. In: Goldman L, Ausiello D (eds) Cecil textbook of medicine, 22nd edition, pp. 1782–​7. Saunders, Philadelphia, PA. Stevens DL, et  al. (2000). Molecular epidemiology of nga and NAD glucohydrolase/​ADP-​ribosyltransferase activity among Streptococcus pyogenes causing streptococcal toxic shock syndrome. J Infect Dis, 182, 1117–​28. Stevens DL, et al. (2005). Practice guidelines for the diagnosis and management of skin and soft-​tissue infections. Clin Infect Dis, 41, 1373–​406. Stevens DL, Bryant AE (2017). Necrotising soft tissue infections.
N Engl J Med, 377, 2253–65. Verani JR, McGee L, Schrag SJ, Division of Bacterial Diseases, National Center for Immunization and Respiratory Diseases, Centers for Disease Control and Prevention (CDC) (2010). Prevention of peri- natal group B streptococcal disease—​revised guidelines from CDC. MMWR Recomm Rep, 59(RR-​10), 1–​36. Woodford N (1998). Glycopeptide-​resistant enterococci: a decade of experience. J Med Microbiol, 47, 849–​62. 8.6.3  Pneumococcal infections Anthony Scott ESSENTIALS Streptococcus pneumoniae is an encapsulated Gram-​positive bac- terium that lives almost exclusively in the human nasopharynx. Each pneumococcus expresses one of more than 90 immunologically dis- tinguishable capsular polysaccharides that are the principal target of systemic human immunity and define its serotype. Epidemiology Pneumococci are transmitted through contact with infected nasal se- cretions or by airborne dissemination, and most preschool children carry them in their nasopharynx. The risk of acquisition is increased by contact with other children, crowded environments, and cold weather. The incidence of pneumococcal disease is highest in young children and elderly people, and also increased in males, certain in- digenous populations, smokers, alcoholics, and patients with chronic medical illnesses or immune susceptibility, including HIV infection, sickle cell disease, and splenectomy. Clinical features Pneumonia—​before introduction of Pneumococcal conjugate vac- cines, pneumococci were the most common cause of severe community-​acquired pneumonia at all ages in the developed and developing world, though their role is now diminishing. Typical pres- entation of pneumococcal lobar pneumonia is with abrupt onset of fever, followed by cough, difficulty breathing, pleuritic chest pain, haemoptysis, and purulent sputum. Physical signs include high pyr- exia, raised respiratory rate, cyanosis, and chest features of lobar con- solidation; namely reduced chest movement, dullness on percussion, fine crepitations, and bronchial breathing over the affected area. The chest radiograph shows a lobar opacity, often with a pleural effusion. Other diseases—​pneumococci cause significant morbidity in adults and children through meningitis and septicaemia, and they can also cause bronchopneumonia and multiple disease syndromes sim- ultaneously (e.g. meningitis and pneumonia). In children, the most common pneumococcal disease is otitis media. Other less common presentations include sinusitis, pleural empyema, pericarditis, endo- carditis, septic arthritis, osteomyelitis, peritonitis, and conjunctivitis. Diagnosis S. pneumoniae is a fastidious organism that grows successfully on blood agar, producing α-​haemolysis. Blood culture is the principal aetiological tool to diagnose pneumococcal pneumonia, but cul- tures are positive in only 15–​30% of cases. The capsular serotype is identified by a positive Quellung reaction with specific rabbit anti- sera. In addition: (1) pneumococci can be observed on microscopy as Gram-​positive diplococci in sputum or, in cases of meningitis, in cerebrospinal fluid, and can be cultured from both specimens; (2) a urinary antigen test for the common pneumococcal constituent C-​polysaccharide is sensitive and specific for pneumococcal pneu- monia in adults, but not in children; (3) polymerase chain reaction is useful in cerebrospinal fluid, especially when the patient is partially treated and cultures are sterile. Treatment and prognosis Most pneumococci are sensitive to β-​lactam antibiotics, but some are resistant. (1) Pneumonia—​when caused by sensitive or inter- mediately resistant pneumococci, this should be treated with high-​dose oral amoxicillin or intravenous cefotaxime, the latter being effective against pneumococci with cephalosporin min- imum inhibitory concentrations up to 1–​2 µg/​ml. Macrolides and newer fluoroquinolones can be used to treat infections that are fully resistant to β-​lactam antibiotics. (2) Meningitis—​when caused by susceptible pneumococci, ceftriaxone is effective; vancomycin should be added as empirical meningitis therapy in areas with penicillin-​resistant pneumococci; dexamethasone is an effective adjunctive treatment for pneumococcal meningitis where HIV prevalence is low. The case fatality of pneumococcal pneumonia is 5%, but in bacter- aemic pneumonia and pneumococcal meningitis it is 30%. Prevention A single dose of 23-​valent capsular polysaccharide vaccine prevents invasive pneumococcal disease in elderly or high-​risk populations. In infants and young children, 10-​ or 13-​valent pneumococcal conju- gate vaccine is highly effective in preventing invasive pneumococcal disease as well as pneumococcal pneumonia, meningitis, and otitis media. It is given routinely as two or three doses in infancy, with a booster dose at 12–​15 months of age. Immunization of children re- duces pneumococcal transmission and prevents pneumococcal dis- ease in older family members.

section 8  Infectious diseases 976 Introduction Streptococcus pneumoniae (the pneumococcus) is a ubiquitous yet potentially fatal human pathogen. Its only viable habitat is the human nasopharynx. Throughout the world, most children and a significant minority of adults carry it at any one time. Most people are exposed to the pneumococcus several times a year but only rarely does this result in illness. When it invades it causes a diverse range of dis- ease syndromes of which pneumonia, meningitis, and septicaemia have high case fatalities, and yet its most common disease mani- festation, otitis media, is relatively benign. In old age it affects the healthy, but throughout life it is a burden to those with chronic med- ical illnesses. Pneumococcal disease is common in temperate and tropical climates; however, because the pneumococcus is fastidious in culture, it is rarely identified, and the disease burden is frequently underestimated. Its differentiation into more than 90 serotypes indi- cates the complexity of its immunological interaction with humans and its need for adaptability in this long-​enduring host–​pathogen relationship. Microbiologists and physicians have regarded the pneumococcus as a formidable opponent for over 135 years. They have fought it with antibiotics and, more recently, with efficacious vaccines, each of which renders its nasopharyngeal niche a hostile home. Yet, through its capacity to combine DNA from other bacteria into its own chromosome, it has evolved and survived. It has all the fascination of the esoteric, yet a busy doctor will not pass a week in practice without seeing a case. Historical perspective S. pneumoniae is a Gram-​positive bacterium first isolated in 1881 by Sternberg in the United States of America and, simultaneously, by Pasteur in France through experiments inoculating rabbits with human saliva. Serotypes of pneumococcus are defined by the rabbit immune response to its variable capsular polysaccharide. In 1910, Neufeld and Händel described two serotypes; there are now over 90 serotypes. Convalescent sera from surviving pneumonia patients were shown to be protective against pneumococcal disease in rabbit models in 1891. The protective substance was identified as homolo- gous anticapsular antibody and this underpinned the development of serum therapy in the early years of the 20th century. Although successful, serum therapy required determination of the serotype of the infecting pneumococcus from sputum or lung aspirate cultures, leading to a delay in treatment. With the introduction of sulphona- mide antibiotics in 1938, serum therapy was abandoned. Antibiotic chemotherapy has been the mainstay of management of pneumococcal disease ever since, but the rapid evolution of re- sistant strains in the 1990s reactivated interest in vaccine develop- ment. Efficacy of a polyvalent capsular polysaccharide vaccine was demonstrated against pneumococcal pneumonia in South African miners in 1976. It was poorly immunogenic in infants, among whom most episodes of pneumococcal disease occur, but conjugation of the polysaccharide to immunogenic proteins overcame this limi- tation and a pneumococcal conjugate vaccine (PCV) against seven serotypes was introduced into the childhood immunization pro- gramme in the United States of America in 2000. In 1928, Griffith inoculated rabbits with a suspension of live avirulent unencapsulated pneumococci and heat-​killed serotype 3 pneumococci. The rabbits subsequently succumbed to serotype 3 septicaemia. The avirulent isolate was derived from a serotype 2 strain suggesting that it acquired the type 3 capsule, and virulence, from the heat-​killed organisms. In 1944, Avery isolated and puri- fied the ‘active principle’ that brought about this transformation and characterized it chemically as DNA. The sequence of the pneumo- coccal genome was first described in 2000 and thousands of strains have now been fully annotated. These sequences have been used to identify conserved surface-​expressed proteins that may serve as antigens in non​capsular vaccines. Epidemiology Incidence The risk of pneumococcal disease is highest in infancy and declines throughout the first 5 years of life. The lowest risk is in older children and young adults and from the age of 50 years onwards disease risk rises progressively (Fig. 8.6.3.1). Among children younger than 5 years, the incidence of culture-​ proven pneumococcal disease was 70–​100 per 100 000 population in the United States of America before vaccine introduction; in Africa it was 110–​430 per 100 000 population. In developed countries, the in- cidence was 15–​20 per 100 000 population among adults of all ages and at least 50 per 100 000 population among adults aged 65 or over. In the prevaccine era, Native Australians and Alaskans and White Mountain Apaches had incidence rates of 200–​1000 per 100 000 popu- lation among children, 50–​180 per 100 000 population among adults 18–​59 years old, and 120–​170 per 100 000 population among older adults. The epidemiology of pneumococcal disease is markedly different in the meningitis belt of West Africa. Here, infants and working-​ age adults are at highest risk, the disease is strongly associated with the dry season, and the case fatality rate is greater than 60% among adults aged 40 years or more. Serotype 1 accounts for more than half of all infections and the burden of meningitis caused by pneumo- coccus, with an incidence of 8–​12 per 100 000 population, rivals that caused by Neisseria menigitidis. The total burden of pneumococcal disease is frequently under- estimated because it is difficult to detect. Studies of ‘invasive pneumococcal disease’, which rely on cultures of S.  pneumoniae from specimens of blood, cerebrospinal fluid, and pleural fluid, fail to identify most of the cases of pneumococcal pneumonia that are not bacteraemic. Lung aspirates obtained by percutaneous fine needle puncture significantly increase the yield of pneumococci from pneumonia cases at all ages but are now rarely undertaken. A World Health Organization (WHO) model of the incidence of pneumo- coccal pneumonia, meningitis, and other serious manifestations esti- mated the annual global burden of disease in children aged less than 5 years as 14.5 million cases in 2000 leading to 826 000 deaths; the es- timate for 2015 was 317 300 deaths. Most of these deaths take place in Africa and Asia. The global burden of disease in adults is not known. Carriage, transmission, and serotypes Viable S.  pneumoniae have been described in collections of dust and in epizootics of some mammals, but the principal habitat and

8.6.3  Pneumococcal infections 977 critical ecological niche of the pneumococcus is the human naso- pharynx (Fig. 8.6.3.2). Infants can acquire infection within hours of birth and most infants in developing countries become infected in the first 3 months of life. In The Gambia, more than 90% of children aged less than 5 years old are colonized by the pneumococcus at any one time. In the United Kingdom, carriage prevalence among chil- dren is approximately 50%. Adults also carry S. pneumoniae in the nasopharynx but at lower prevalence. Although the pneumococcus has evolved more than 90 capsular serotypes, prior to vaccine introduction 58 to 66% of disease in chil- dren was caused by just seven (serotypes 1, 5, 6A, 6B, 14, 19F, and 23F), depending on region. The ratio of the incidence of invasive disease to the incidence of nasopharyngeal acquisition provides an index of the invasiveness of pneumococcal serotypes and can be used to group them. In the prevaccine era, serotypes 1, 5, 12F, and 46 were found among series of invasive isolates but were rarely isolated in the nasopharynx. These are labelled ‘adult’ types because they are associ- ated with disease in adults. Other serotypes, including 10A, 11A, 15B, 15C, 16F, and 33F, were found among series of colonizing isolates but were uncommon causes of disease. A third group, which includes serotypes 6A, 6B, 14, 19A, 19F, and 23F, were found very commonly in the nasopharynx but also caused invasive disease. These are la- belled ‘paediatric’ types because they caused most of the invasive disease episodes among young children before vaccine introduction. Serotypes that are highly prevalent in the nasopharynx tend to be less invasive than others but when they do cause disease the case fatality 300 White Black Other Race 400 200 100 0 Age group, y Incidence, cases per 100 000 <2 2–4 5–17 18–34 35–49 50–64 65–79 ≥80 Fig. 8.6.3.1  Incidence of invasive pneumococcal disease in the United States of America between 1995 and 1998 by age and race. The data are taken from the Active Bacterial Core Surveillance of the Centers for Disease Control and Prevention. From Robinson KA, et al. (2001). Epidemiology of invasive Streptococcus pneumoniae infections in the United States, 1995–​1998: opportunities for prevention in the conjugate vaccine era. JAMA, 285, 1729–​35. Fig. 8.6.3.2  Electron micrographs of pneumococci (strains TIGR4 and G54) adherent to D562 human pharyngeal epithelial cells in culture. From Kimaro Mlacha SZC, et al. (2013). Phenotypic, genomic, and transcriptional characterization of Streptococcus pneumoniae interacting with human pharyngeal cells. BMC Genomics, 14, 383.

section 8  Infectious diseases 978 ratios are higher. Success in colonization is a function of evading the host defences (neutrophil-​mediated killing) and outcompeting other serotypes for the ecological niche of the nasopharynx. The duration of carriage can be as short as 3 h or as long as 3 years; in most instances it is between 1 and 6 months. It declines with the age of the host, probably as a result of CD4+ T-​cell acquired im- munity mediated by IL-​17A. For some serotypes, circulating anticapsular IgG appears to reduce the acquisition of carriage and vaccine-​induced anticapsular antibodies are also highly effective in reducing carriage prevalence. The pneumococcus is transmitted by carriers, particularly pre- school children, through direct contact with nasal secretions or by infected fomites. The rapid spread of pneumococcal pneumonia in outbreaks in adults suggests that airborne dissemination, facilitated by cough, is another mechanism of spread. Risk factors Risk of pneumococcal disease is a function of exposure to the bac- terium, leading to colonization, and of host resistance to invasion. Exposure is increased in crowded environments at home and in in- stitutions (e.g. military barracks, homeless shelters, jails, miners’ compounds) and by contact with preschool children. In temperate climates pneumococcal disease follows a consistent seasonal variation with winter peaks and summer troughs. Risk is especially high at New Year when families gather and generations intermingle (Fig. 8.6.3.3). Throughout life, males have an incidence of pneumococcal disease 1.2 to 1.5 times greater than females. Chronic medical conditions predispose to pneumococcal invasion (see Box 8.6.3.1). Alcoholism is consistently associated with pneumococcal disease and may act directly on macrophage function or, like seizure disorders, by com- promising laryngeal defences leading to aspiration. Untreated HIV infection increases the risk of invasive pneumococcal disease by ap- proximately 50-​fold. Where HIV prevalence is greater than 2%, as in much of Africa, most cases of pneumococcal disease occur among HIV-​positive patients. Recurrent pneumococcal disease is especially common in this group. Respiratory viral infections, especially influenza, increase the risk of invasive pneumococcal disease. Influenza enhances the acquisition of pneumococci in the nasopharynx, in animal models, and increases the density of colonization in children. The virus facilitates pneumo- coccal binding by damaging respiratory epithelial cells and synergy between the two pathogens enhances the action of their separate neuraminidases and stimulates type 1 interferons, impairing macro- phage function. Antibiotic resistance Epidemiology Laboratory isolates of penicillin-​resistant pneumococci were first reported in 1967 and by the early 1970s they were being isolated in clinical specimens in Australia and Papua New Guinea. During the 1990s, penicillin resistance spread widely throughout the world reaching a prevalence of 35% in several countries (e.g. France, South Africa, Japan, Hong Kong). In Europe pneumococci are classified as sensitive (minimum inhibitory concentration (MIC) of 0.06 µg/​ml or below) or resistant to benzylpenicillin (MIC 2 µg/​ml and above). For pneumonia treatment, the classification between these cut-​offs depends on the dose given; for example, when treating with high doses of benzylpenicillin (2.4 g every 4 hours), MICs up to 2 µg/​ml are considered susceptible. Because of the difficulty in achieving high concentrations of penicillin in cerebrospinal fluid, 90 80 70 60 50 40 30 20 10 0 Rate (cases/100 000 population/year) 1998 1997 1996 Jan Jan Apr Jul Oct Jan Apr Jul Oct Jan Apr Jul Oct Fig. 8.6.3.3  Annualized weekly incidence of pneumococcal disease among adults in the United States of America between 1996 and 1998 showing a consistent increase in incidence in the winter and a sharp increase in incidence during the Christmas/​New Year holiday season. From Dowell SF, et al. (2003). Seasonal patterns of invasive pneumococcal disease. Emerg Infect Dis, 9, 573–​9.

8.6.3  Pneumococcal infections 979 pneumococci isolated in a case of meningitis are classified as fully resistant if the MIC is greater than 0.06 µg/​ml. In the United States of America, the classification of isolates from non​meningitis inva- sive infections (sensitive ≤2 µg/​ml; intermediate = 4 µg/​ml; resistant ≥8 g/​ml) identifies a higher proportion of isolates as sensitive and therefore treatable with benzylpenicillin. This reflects the higher doses of penicillin used in the United States of America as well as results from epidemiological surveillance. For meningitis, the sen- sitive/​resistant cut-​off is the same as in Europe. Resistance to other antibiotic classes, including macrolides, also increased during the 1990s. The use of long-​acting macrolides appears to be responsible for the observed increase in resistance to erythromycin. Multidrug-​resistant pneumococci were first observed in South Africa in 1978. Their presence suggests that the use of one antibiotic can select for resistance against another. Penicillin-​resistant strains are transmitted more successfully than penicillin-​susceptible strains but multiresistant strains spread most successfully within popu- lations. Much of this spread is driven by the expansion of a small number of clones. In the United States of America, 78% of all re- sistant isolates are represented by just 12 clonal groups. In the United Kingdom, antimicrobial prescriptions have fallen since 1995 and at the same time there has been a reduction in both penicillin-​resistant and macrolide-​resistant pneumococci. Introduction of 7-​valent PCV in the United States of America in 2000 resulted in a marked decrease in the incidence of disease with penicillin-​resistant strains in children, followed by a gradual rise be- tween 2005 to 2010; introduction of PCV13 then led to a further decline in the incidence of resistant invasive infections in children. Pneumococcal resistance to fluoroquinolones has increased in prevalence but remains less than 5%. Resistance mechanisms Pneumococcal resistance to penicillin is entirely due to the accumu- lation of genetic variations among the six penicillin-​binding proteins (PBPs) that normally catalyse cross-​linkage of the bacterial cell wall. By binding to PBPs, β-​lactam antibiotics inhibit cell wall synthesis and promote cell lysis. Resistance to penicillin occurs when a PBP variant arises which has low binding affinity for β-​lactams. Sensitive pneumococci have MICs for benzylpenicillin and cefotaxime that are approximately 0.02 µg/​ml. Mutations in PBP 2b or 2x genes lead to a 2-​ to 30-​fold increase in MIC. However, isolates that are fully resistant to penicillin usually contain alterations at three PBP genes, 2b, 2x, and 1a. High-​level resistance to cefotaxime may be observed with a combination of changes in only two (PBP 2x and 1a). Resistant strains come about through horizontal transfer and re- combination into chromosomal DNA of large sequence blocks of mosaic genes acquired from other streptococci. This transfer can in- clude capsular and resistance genes simultaneously. Pneumococci colonizing the nasopharynx are then exposed to antibiotics, which are commonly prescribed for community-​acquired respiratory tract infections, and this selects resistant strains. Pneumococcal genes cat, erm(B), and tet(M), which confer re- sistance to chloramphenicol, macrolides, and tetracyclines, respect- ively, have been found together on DNA elements (conjugative transposons) that spread between pneumococci without involving recombination, thus facilitating multidrug resistance. Resistance to fluoroquinolones and trimethoprim/​sulfamethoxazole is acquired by point mutations in topoisomerase genes and folate synthesis genes, respectively. Exposure to low levels of antibiotic selects single gene mutants, which then acquire higher resistance through add- itional mutations. Pathogenesis Pneumococci exist in two morphologically distinct phenotypes; in the opaque phase they have abundant capsular expression and in the transparent phase they have little (Fig. 8.6.3.4). In the nasopharynx, transparent-​phase pneumococci predominate as abundant capsule prevents attachment of pneumococcal cell Box 8.6.3.1  Risk factors for pneumococcal disease Social and demographic • Older age • Male sex • Black race • Indigenous populations • Lower level of education • Unemployment • Excess alcohol use • Occupational exposure to metal fumes (Welding) Exposure to pneumococci • Contact with preschool children • Day care attendancea • Crowding in the home • Crowded adult environments (homeless shelters, military, or occupa- tional barracks) • Institutionalized care • Winter season • Hospital admission Respiratory tract damage • Currently smoking • Passive smoking • Indoor air pollution • Chronic obstructive pulmonary disease • Recent viral respiratory tract infection Preexisting medical conditions • Chronic renal failure • Congestive heart failure • Cirrhosis • Cerebrovascular disease • Dementia • Seizure disorder • Asthma • Diabetes • Malignancies of the lung Immune susceptibility • HIV • Hypogammaglobulinaemia • Sickle cell disease • Asplenia/​splenectomy • Pregnancy • Not breastfeedinga • Previous pneumococcal disease a These apply only to infants or children.

section 8  Infectious diseases 980 wall structures to epithelial cells. This attachment is mediated by binding of pneumococcal phosphorylcholine and choline-​binding protein A (CbpA) to human platelet activating factor receptors and polymeric immunoglobulin receptors. Once attached, pneumo- cocci can cause disease by local spread to the middle ear or sinuses, by aerosol inhalation to the lung, or by blood stream invasion to the meninges, joint spaces, or heart valves. Blood stream invasion begins with endocytosis across the mucosal barrier although the components of this pathway are not well understood. In the blood stream pneumococci are found in the opaque phase since capsule is effective in evading opsonophagocytosis. Pneumococci colonizing the nasopharynx cannot bind to the ciliated epithelium of the bronchi and therefore make their way to the lung in aerosols. However, if the ciliated epithelium is damaged, by antecedent viral infection or by cigarette smoke, it reveals a basement membrane to which pneumococci can ad- here easily. Pneumolysin released from pneumococci causes further epithelial damage by direct cytotoxicity and encour- ages inflammation of the larger bronchioles, which leads to bronchopneumonia. In the alveoli, pneumococci multiply in serous fluid and spread from one alveolus to another through the pores of Kohn. They ad- here to the alveolar type 2 cells, through expression of CbpA, and stimulate production of the inflammatory mediators tumour ne- crosis factor-​α, nitric oxide, and interleukins IL-​1, IL-​6, and IL-​10, which initiates oedema. This creates the first pathological phase of pulmonary consolidation—​engorgement. In the second phase—​red hepatization—​erythrocytes leak into the alveolar spaces, reducing the compliance of the lung and leading to a liver-​like appearance of the gross lung specimen (Fig. 8.6.3.5). Fibrin deposition creates a mesh of erythrocytes, leucocytes, and damaged epithelial cells and the lymphatics become dilated with cells and fibrin. Without ventilation, perfusion declines, and the lung becomes maximally consolidated. CbpA binding stimulates epithelial cells to release chemokines that attract leucocytes to the lung which initiates the third phase of consolidation—​grey hepatization. Neutrophils trap pneumo- cocci against the alveolar wall and engulf them by surface phago- cytosis. C-​reactive protein enhances this process by binding to choline residues on pneumococcal surfaces. The chemokines activate complement that, together with anticapsular antibody, facilitates opsonophagocytosis. Thereafter, lung inflammation be- gins to decline simultaneously with neutrophil apoptosis, fever declines, and macrophages are recruited to the lung to absorb the debris. Over a period of weeks this leads to complete resolution of the pathology. Fig. 8.6.3.4  Immunoelectron microscopy of pneumococcal capsules showing an increased zone of capsular material in opaque (a) and (b) compared to transparent (c) and (d) variants of type 6B pneumococcal strain P324. Reproduced from Kim JO, et al. (1999). Relationship between cell surface carbohydrates and intrastrain variation on opsonophagocytosis of Streptococcus pneumoniae. Infect Immun, 67, 2327–​33.

8.6.3  Pneumococcal infections 981 Immunity to pneumococcal disease Historically, the role of anticapsular antibody in recovery from pneu- monia and the efficacy of serotype-​specific serum therapy suggested that anticapsular IgG was the primary mechanism of immunity to pneumococcal disease. The age groups at highest risk of pneumo- coccal disease, infants, and elderly people, have little anticapsular antibody or have antibody that lacks avidity. The genetic diver- sity of capsular expression into more than 90 variants suggests the antigen is under considerable immune selection and the success of polysaccharide antigens as vaccines further reinforces the import- ance of anticapsular immunity. Capsular polysaccharides are complex molecules with repeating epitopes that create cross-​linkage of antigen receptors on B lympho- cytes and which stimulate antibodies of the IgM and IgG2 isotypes. Antibody production can occur in the absence of T lymphocytes (T-​ independent) but it does not induce memory responses and can lead to antigen tolerance following repeated stimulation. Pneumococcal capsular and cell wall components both activate the human com- plement system leading to deposition of C3b and C3d on capsular polysaccharide. In the presence of both anticapsular antibody and complement, encapsulated pneumococci are opsonized and taken up by phagocytes expressing the receptor Fcγ-​RIIa (CD32). In contrast to natural immunity, conjugates of highly immuno- genic proteins (e.g. diphtheria toxoid) with polysaccharides induce T-​cell-​dependent immunity with a predominance of IgG1 antibody and a memory response. These responses are inducible even in very young infants. Anticapsular antibody responses are measured by IgG enzyme-​linked immunosorbent assay (ELISA) and a serum concentration of 0.35 µg/​ml correlates with vaccine-​induced protec- tion in infancy. In addition to adaptive immunity, innate mechanisms (e.g. lipoteichoic acid stimulation of Toll-​like receptor (TLR) 2, or pneumolysin stimulation of TLR4) in response to pneumococcal virulence factors (Fig. 8.6.3.6) appear to be important in shaping the inflammatory response to pneumococcal disease and in determining host survival. Furthermore, several lines of evidence suggest that (a) (b) Fig. 8.6.3.5  (a and b) Red hepatization in fatal pneumococcal pneumonia. Copyright D. A. Warrell. ATP-binding cassette transporter Pneumolysin PsaA PiaA PiuA Metal- binding proteins Choline- binding proteins LPXTG-anchored neuraminindase proteins Hyl Sortases PavA Eno LytA Capsule Cell wall Cell membrane PspA PspC Fig. 8.6.3.6  Pneumococcal virulence factors. Important pneumococcal virulence factors include: the capsule; the cell wall; choline-​binding proteins; pneumococcal surface proteins A and C (PspA and PspC); the LPXTG-​anchored neuraminidase proteins; hyaluronate lyase (Hyl); pneumococcal adhesion and virulence A (PavA); enolase (Eno); pneumolysin; autolysin A (LytA); and the metal-​binding proteins pneumococcal surface antigen A (PsaA), pneumococcal iron acquisition A (PiaA) and pneumococcal iron uptake A (PiuA). Reprinted by permission from Macmillan Publishers Ltd: Nature Reviews Microbiology. Kadioglu A, et al. (2008). The role of Streptococcus pneumoniae virulence factors in host respiratory colonization and disease. Nat Rev Microbiol, 6, 288–​301, copyright © 2008.

section 8  Infectious diseases 982 CD4 T cells play an important role in nasopharyngeal immunity that could be exploited by vaccines consisting of pneumococcal proteins or even whole cell killed pneumococci. Prevention Pneumococcal polysaccharide vaccine In 1976, Austrian reported a trial of a 13-​valent vaccine among South African gold miners in which the efficacy against putative pneumococcal pneumonia was 78%. This led to the commercial- ization of a pneumococcal polysaccharide vaccine, initially with 14 serotypes and later extended to 23 serotypes. Trials of these vaccines in older people or in high-​risk populations do not pro- vide consistent evidence of protection against pneumococcal pneumonia. Observational studies using case–​control designs or the indirect cohort method have more consistently indicated pro- tection against bacteraemic pneumococcal disease. The evidence of effect is greater for older people than for those with chronic disease. On the basis of meta-​analyses of the observational studies, pneumococcal polysaccharide vaccination is recommended in the United Kingdom for all adults aged 65 years or more and for all persons aged 5 or more years who belong to an at-​risk group (e.g. with asplenia, splenectomy, chronic respiratory disease, chronic heart, liver, or renal disease, diabetes, or immunosuppression, including HIV infection at all stages). Among patients having planned splenectomy, vaccination should take place well before the operation. For all adults with HIV, PCV is recommended; pneumococcal polysaccharide vaccine is recommended for HIV adults who are elderly or have other comorbidities, although the use of pneumococcal polysaccharide vaccine remains controver- sial. In a study of the vaccine in HIV-​positive individuals from Uganda, the vaccinated group had an elevated risk of pneumonia. The vaccine is not recommended for HIV-​positive populations in the developing world. Pneumococcal conjugate vaccine A 7-​valent PCV, consisting of separate protein–​polysaccharide con- jugates for serotypes 4, 6B, 9V, 14, 18C, 19F, and 23F, was licensed in the United States of America in 2000 following successful trials in Californian infants and Native American children. The seven serotypes in the vaccine accounted for 83% of invasive disease in American children less than 2 years old. The efficacy against inva- sive pneumococcal disease caused by these serotypes was 97% after a four-​dose schedule given at 2, 4, 6, and 15 months of age. The vaccine was also shown to protect against pneumococcal meningitis, bacter- aemia, pneumonia, and otitis media. PCV reduces nasopharyngeal colonization by pneumococci of the serotypes included in the vaccine and increases colonization by other serotypes commensurately. In routine immunization, this has produced two effects. First, the transmission of vaccine-​serotype pneumococci has declined providing ‘herd protection’ for older children and adults whose pneumococcal disease rates have fallen substantially (Fig. 8.6.3.7). Second, the incidence of disease caused by serotypes not included in the vaccine has increased slightly. So far this ‘serotype replacement disease’ has been much smaller in mag- nitude than the substantial reductions in vaccine-​serotype disease. However, to mitigate the effects of serotype replacement disease new vaccines with 10 serotypes (including 1, 5 and 7F) or 13 sero- types (also including 3, 6A and 19A) were developed and licensed. A 13-​valent PCV was introduced into childhood immunization pro- grammes in the United Kingdom and United States in 2010. Vaccine trials in children in South Africa and The Gambia have shown that 9-​valent PCV, which includes the common African sero- types 1 and 5, can protect against invasive pneumococcal disease among HIV-​infected children and can reduce childhood mortality and admissions to hospital with pneumonia among young children. 1998 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 1999 2000 2001 2002 2003 2004 2005 2006 2007 90 (a) (b) 80 70 60 50 40 30 20 10 0 40 35 30 25 20 15 10 5 0 Year Year Cases/100 000 population Cases/100 000 population Serotype group PCV7 type non-PCV7 type 19A Fig. 8.6.3.7  Changes in invasive pneumococcal disease incidence by serotype group among American children aged less than 5 years (a) and adults aged ≥65 years (b), 1998–​2007, illustrating herd protection in adults and serotype replacement disease in both groups. From Pilishvili T et al. (2011) Sustained reductions in invasive pneumococcal disease in the era of conjugate vaccine. J Infect Dis, 201, 32–​41, by permission of Oxford University Press.

8.6.3  Pneumococcal infections 983 In 2007, WHO recommended introduction of PCV into childhood immunization programmes in developing countries and, by 2017, over 50 developing countries had introduced PCV with financial support from GAVI, The Vaccine Alliance. Among HIV-​infected Malawian adults, with a previous history of invasive pneumococcal disease, 7-​valent PCV reduced recurrent invasive pneumococcal disease by 74%. Pneumococcal disease is a significant problem among African adults, including both HIV sero- positive adults in Southern Africa and otherwise healthy adults in the meningitis belt of West Africa. However arguments in favour of adult vaccination have not yet been translated into health policy. The protective efficacy of 13-​valent PCV against pneumococcal pneumonia has been evaluated among Dutch adults aged 65 years or older. The serotype responsible for the pneumonia episode was determined by detecting capsular polysaccharides in urine using an immunodiagnostic assay based on monoclonal antibodies. For dis- eases caused by serotypes in the vaccine, the efficacy was 46% against pneumonia and 75% against invasive disease. The vaccine was sub- sequently recommended for use in adults aged 65 years or older in the United States of America but not in the United Kingdom, where disease is effectively controlled by indirect protection through the childhood immunization programme. In the United Kingdom, where herd protection is now well es- tablished, the recommended immunization schedule for infants is two doses of 13-​valent PCV given at 12 weeks and 12 months of age. For children 2 to 5 years of age who belong to an at-​risk group, the recommendation is for a single dose of PCV followed 2 months later by a single dose of pneumococcal polysaccharide vaccine. Other forms of prevention For children who are at high risk of invasive pneumococcal disease, including those with sickle cell disease or nephrotic syndrome, or following splenectomy, daily prophylaxis with oral penicillin re- duces risk by over 80%. It should be continued until at least 5 years of age. In developing countries, simple measures such as reducing indoor smoke from cooking stoves and improving nutrition are likely to be effective in prevention. Zinc supplementation can reduce the inci- dence of pneumonia in children by 40%. In Pakistan, a community intervention to promote hand washing reduced pneumonia inci- dence in children by one-​half. Diagnosis Culture of S. pneumoniae A specific diagnosis of pneumococcal disease is made by culture of S. pneumoniae from a normally sterile site in a patient with a compatible illness. Pneumococci are fastidious organisms but they grow readily on 5% blood agar incubated in 5% CO2. Colonies are small and grey with a draughtsman like central indentation and are surrounded by a greenish zone of α-​haemolysis. Species iden- tity is confirmed by sensitivity to optochin (ethylhydroxycupreine), bile solubility, and serotyping. The capsular type of S. pneumoniae is differentiated by a change in the refractive index around the cell seen on microscopy in the presence of specific rabbit antisera, the (Neufeld) Quellung reaction (Fig. 8.6.3.8). In pneumococcal meningitis, cerebrospinal fluid frequently yields a positive culture. Pneumococci are also cultured from pleural and joint fluid in thoracic empyema and septic arthritis, respectively. Diagnosis of pneumococcal pneumonia by culture is, however, highly insensitive. Blood culture is a poor diagnostic test for several reasons; infection can be confined to the lungs; episodes of bacter- aemia are only intermittent; the density of bacteraemia is too low, especially in children; or the patient can have taken antibiotics that inhibit growth. Sputum culture lacks specificity, since the pharynx is colonized by pneumococci even in healthy individuals, and also has relatively poor sensitivity. Most cases of pneumococcal pneumonia are, therefore, not formally diagnosed. A measure of this insensi- tivity is obtained from the Gambian trial of PCV where 15 cases of radiographic pneumonia were prevented by vaccination for every two cases of detectable bacteraemic disease prevented. Antigen detection Patients with pneumococcal pneumonia excrete C-​polysaccharide, a universal component of pneumococcal cell walls, and capsular polysaccharides. Detection of C-​polysaccharide in urine has been commercialized in a rapid immunochromatographic test that has a sensitivity of approximately 80% for bacteraemic pneumococcal pneumonia in adults and is highly specific. In children it lacks spe- cificity as positive results may be obtained from healthy individuals who are merely colonized with pneumococci. A sensitive and spe- cific Luminex-​based serotype-​specific antigen detection assay has been developed for analysis of urine in adults but, at present, this is restricted to the serotypes contained in the 13-​valent PCV. Testing for C-​polysaccharide antigen in cerebrospinal fluid is a useful ad- junct to the diagnosis of meningitis. Polymerase chain reaction Primers targeting genes encoding the pneumococcal proteins pneumolysin, autolysin, pneumococcal surface adhesin A, and PBPs have been used for polymerase chain reaction (PCR) diagnosis. These Fig. 8.6.3.8  The (Neufeld) Quellung reaction. Pneumococci show an apparent increase in the thickness of capsule when mixed with homologous anticapsular antibodies. The negative control is shown on the left and the positive reaction on the right. From Werno AM, Murdoch DR (2008). Medical microbiology: laboratory diagnosis of invasive pneumococcal disease. Clin Infect Dis, 46, 926–​32, by permission of Oxford University Press.

section 8  Infectious diseases 984 assays have the same limitations as culture-​based detection. PCR of respiratory specimens does not distinguish colonization from lung infection, and PCR of blood has poor sensitivity for pneumococcal pneumonia. Conversely, PCR of cerebrospinal fluid is sensitive and specific and has proven useful in the investigation of epidemic men- ingitis. Quantitative real-​time PCR for the autolysin gene, lytA, has shown high specificity and good sensitivity in validation studies in adults and children in high-​income settings although it lacks speci- ficity in children in low-​income settings where carriage prevalence and carriage densities are high in health children. Clinical features S.  pneumoniae causes pneumonia, meningitis, septicaemia, otitis media, endocarditis, peritonitis, sinusitis, conjunctivitis, and puru- lent infections of the pleura, joints, and bone. These conditions do not necessarily occur in isolation; pneumococcal meningitis is quite frequently accompanied by pneumonia. Pneumococcal pneumonia Symptoms Typically, the illness starts suddenly although there may be an ante- cedent upper respiratory tract infection. Fever is usually the first symptom and it is frequently accompanied by rigors. The patient feels weak and anorexic and may have severe headache and myalgia. Cough develops within 24–​72 h and becomes a prominent symptom. At first the cough is non​productive but it becomes productive of blood-​tinged (‘rusty’) sputum and later of purulent sputum. Stabbing pleuritic chest pain also develops during the course of the illness. The patient might try to obtain relief by splinting the af- fected side of the chest or lying on the affected side. Involvement of the diaphragmatic pleura leads to misleading abdominal pain or referred pain in the shoulder. Among young children, a history of cough and difficulty breathing should raise suspicion of pneumonia. Elderly and immunocompromised patients might present with general malaise and delirium, with few respiratory symptoms and no fever. Prior antibiotic treatment also modifies the classic presentation. Physical signs On general examination, adults have tachycardia and pyrexia, with a rectal temperature as high as 40°C. With early presentation the respiratory system can appear normal, but pneumonia patients go on to develop rapid and difficult breathing (of which flaring of the alae nasi is a subtle early sign), cyanosis, and signs of lobar consoli- dation including reduced chest movement, dullness on percussion, fine crepitations, and, occasionally, bronchial breathing over the af- fected area. A pleural rub is sometimes audible. Abdominal distension, upper abdominal tenderness, and guarding suggest involvement of the diaphragmatic pleura. Mild jaundice occurs in a minority. Concomitant herpes labialis (‘cold sores’) is common. Delirium is a sign of severity and is frequently observed in elderly patients. In infants, the signs of pneumonia are non​specific; most will have a raised respiratory rate and nasal flaring but only a minority will have crepitations. In developing countries, most cases of pneumonia are diagnosed and treated by non​medical health workers. To facili- tate diagnosis and promote early treatment, the WHO has designed a simple diagnostic algorithm as part of its Integrated Management of Childhood Illness (IMCI), which defines pneumonia on the basis of respiratory rate and lower chest wall indrawing (Table 8.6.3.1). Severe pneumonia, requiring admission to hospital, is indicated by hypoxia or generic ‘danger signs’ (Fig. 8.6.3.9). Investigations The pathological process of pneumonia is confirmed by the chest radiograph. This typically shows a homogenous area of opacification confined within the lobar structure (Fig. 8.6.3.10). The lower lobes are affected more frequently than the upper lobes. The area of path- ology may be localized to a single lobule or extend over several lobes; early in the presentation there may be no abnormality at all. Table 8.6.3.1  The WHO classification of pneumonia in children in developing countries Sign or symptom Classification Treatment Cough or difficulty in breathing with: • Oxygen saturation <90% or central cyanosis • Severe respiratory distress (e.g. grunting, very
severe chest indrawing) • Signs of pneumonia with a general danger sign (inability to breastfeed or drink, lethargy or
reduced level of consciousness, convulsions) Severe pneumonia •​ Admit to hospital •​ Give oxygen if saturation <90% •​ Manage airway as appropriate •​ Give recommended antibiotic •​ Treat high fever if present Fast breathing: •​ ≥50 breaths/​min in a child aged 2–​11 months •​ ≥40 breaths/​min in a child aged 1–​5 years Chest indrawing Pneumonia •​ Home care •​ Give appropriate antibiotic •​ Advise the mother when to return immediately if symptoms of severe pneumonia •​ Follow-​up after 3 days No signs of pneumonia or severe pneumonia No pneumonia: cough or cold •​ Home care •​ Soothe the throat and relieve cough with safe remedy •​ Advise the mother when to return •​ Follow-​up after 5 days if not improving •​ If coughing for more than 14 days, refer to chronic cough Adapted from World Health Organization (2013). Pocket book of hospital care for children: guidelines for the management of common illnesses with limited resources. World Health Organization, Geneva, copyright © 2013.

8.6.3  Pneumococcal infections 985 In adults this might also reveal an underlying risk factor (e.g. lung cancer). In children, widespread patchy opacification (broncho- pneumonia) is common. Lateral radiographs add to the sensitivity of posteroanterior projections particularly for lower lobe disease hidden beneath the dome of the diaphragm. In adults, pneumococcal aetiology is defined by the C-​ polysaccharide antigen test in urine. Many patients are severely dehydrated on admission and cannot readily produce a urine spe- cimen. Blood culture is positive for S. pneumoniae in about 10 to 30% of adults and about 15% of children. Genuine sputum samples should be differentiated from upper respiratory tract secretions by a high ratio of pus cells to epithelial cells on microscopy. The appear- ance of large numbers of Gram-​positive diplococci on microscopy together with culture of S. pneumoniae is diagnostic. However, be- cause prior antibiotic use is common, sputum microscopy is positive in only about one-​quarter of patients and sputum culture is positive in only one-​half. Young children cannot normally produce a sputum specimen. Differential diagnosis The abrupt onset of symptoms often leads the patient to seek care before focal signs become established and it is not possible to differ- entiate pneumonia from other causes of acute febrile illness. In trop- ical countries, malaria is the main differential at this stage. When localizing symptoms and signs are established, pneumonia must be distinguished from pulmonary infarction. Both conditions lead to chest pain and haemoptysis and are accompanied by tachycardia. Pyrexia and rigors favour a diagnosis of pneumonia, while a very sudden history of chest pain and frank haemoptysis favour pul- monary embolism. Pulmonary oedema (secondary to heart failure), pulmonary atelectasis, pleurisy, lung abscess, tuberculosis, and acute bronchitis should also be considered in the differential diagnosis. Outside the chest, subdiaphragmatic lesions such as cholecystitis, a subphrenic abscess, or an amoebic liver abscess can mimic the clin- ical picture of lower lobe pneumonia. Bacterial pneumonia is differentiated from viral or mycoplasma pneumonia by its abrupt onset, severity of symptoms and systemic illness, raised peripheral white blood cell count, and C-​reactive protein level exceeding 125 mg/​litre in serum. Confusion, signs of multiorgan involvement, lymphopenia, or a low serum sodium should raise the possibility of legionnaires’ disease. Tuberculosis occasionally presents with an acute pneumonia in adults. In HIV-​ infected patients, the differential diagnosis also includes infection by Pneumocystis jirovecii, mycobacteria, and cytomegalovirus. Treatment Management of pneumonia first requires an assessment of severity to determine whether the patient should be treated at home, ad- mitted to hospital, or admitted to the intensive care unit. The British Thoracic Society (BTS) recommendations define pneumonia as se- vere if there are three or more CURB-​65 features: Confusion, Urea exceeding 7 mmol/​litre (Fig. 8.6.3.11), Respiratory rate equal to or exceeding 30 breaths/​min, abnormal Blood pressure, either systolic (<90 mm Hg) or diastolic (≤60 mm Hg) hypotension, and age equal to or exceeding 65 years. Additional features that might influence this assessment include the presence of coexisting disease, hypox- aemia (Pao2 less than 8 kPa or Sao2 less than 94%), and bilateral or multilobe involvement on the chest radiograph. Bacteraemia is itself an indicator of severity and increased risk of death. Supportive care includes analgesia for chest pain, ample hydration and nutrition, ad- vice to stop smoking, and oxygen for inpatients with hypoxaemia. Empirical guidelines for pneumonia treatment focus on treatment of pneumococcal pneumonia. High-​dose penicillin or amoxicillin Fig. 8.6.3.9  Kenyan child with very severe pneumonia, as defined by the WHO, receiving high-​flow oxygen therapy. Taken in the clinical service of Kilifi District Hospital; supplied by Dr Mike English. The patient’s parents gave written consent for the taking of this photograph and for its use for educational purposes. Fig. 8.6.3.10  Chest radiograph of an adult with clinical signs of left lower lobe pneumonia illustrating a well-​demarcated area of alveolar consolidation.

section 8  Infectious diseases 986 therapy will provide serum concentrations sufficiently high to treat pneumonia that is caused by pneumococci with MICs up to 2 µg/​ml. Based on efficacy, cost, and acceptability, the optimum antibiotic is amoxicillin. In the UK oral amoxicillin 500 mg three times daily for 5 days is recommended for non​severe cases of community-​acquired pneumonia treated at home. Doxycycline 100  mg daily (initial 200 mg loading dose) or clarithromycin 500 mg twice daily are ac- ceptable alternatives. A fluoroquinolone with enhanced pneumo- coccal activity (e.g. levofloxacin, moxifloxacin) can be considered in outbreaks of resistant pneumococcal disease or in patients unre- sponsive to first-​line antibiotics. Cases of pneumonia with high severity (CURB65 score 3–​4) should be treated empirically with a broad-​spectrum intra- venous antibiotic such as co-​amoxiclav (1.2 g three times daily), cefuroxime (1.5 g three times daily), cefotaxime (1 g three times daily), or ceftriaxone (2 g once daily) for 10 days. Cefotaxime or ceftriaxone are most active against pneumococci and should be ef- fective against pneumonia caused by pneumococci with high ceph- alosporin MICs of 1–​2 µg/​ml. Empiric therapy with clarithromycin (500 mg twice daily) should also be given to cover other causes of pneumonia. After 3 days of intravenous antibiotics, clinically stable patients can be safely switched to oral therapy. Pneumonia patients admitted to hospital with pneumonia of moderate se- verity (CURB65 score 2) should be treated with oral amoxicillin (500–​1000 mg three times per day) plus clarithromycin 500 mg twice daily. Course and prognosis Historically, untreated patients who survived long enough to make specific anticapsular polysaccharide antibody recovered spontan- eously by crisis, or by a more gradual lysis, 7 to 10 days after the onset of illness. However, mortality from pneumococcal pneu- monia in the preantibiotic era was 20–​40%. With antibiotic treat- ment, mortality is about 5% overall but is 30% among the subset of patients with bacteraemia. Mortality is higher among elderly and very young patients, and among those with an underlying illness such as cirrhosis, alcoholism, or heart disease. Most deaths occur within the first few days of admission to hospital. The causes of death are difficult to establish but include shock, cardiac arrhyth- mias, and respiratory failure. Pneumococcal pleural effusion and empyema A large pleural effusion or an empyema develops during treat- ment in 2 to 5% of patients with established pneumococcal pneumonia. Symptoms Some patients with pneumococcal empyema give a history of re- cent lung infection but others develop the disease without any previous illness. Hectic fever, rigors, sweats, malaise, anorexia, and marked weight loss are characteristic symptoms, often going back several weeks. Patients with a large pleural collection are breathless and may complain of dull pain on the affected side. A productive cough is unusual unless a bronchopleural fistula has developed. Physical signs General examination reveals pyrexia, tachycardia, and evidence of recent weight loss. Examination of the chest usually shows the char- acteristic signs of a pleural effusion: diminished chest movement, stony dullness on percussion, and diminished breath sounds over the accumulated fluid. The chest wall overlying an empyema may be tender. Investigations The effusion will usually be visible on the chest radiograph but loculated effusions may require localization by ultrasonography. On aspiration, turbid fluid or thick pus is obtained which contains pneumococci and degenerate white cells. If antibiotics have been given it might not be possible to culture pneumococci, but the fluid contains detectable pneumococcal antigens. The peripheral white blood cell count is raised predominantly with neutrophils. Differential diagnosis The principal differential diagnosis is pulmonary tuberculosis, and pleural biopsy may be required if the pleural fluid is sterile. The ab- sence of copious, purulent sputum differentiates pleural empyema from a lung abscess. Treatment Successful treatment requires both intravenous antibiotics and pleural drainage. Appropriate antibiotic treatment for pneumo- coccal empyema follows the recommendations for pneumococcal pneumonia, with intravenous amoxicillin or a cephalosporin. (a) (b) Fig. 8.6.3.11  Urea frost in two patients with uraemia complicating pneumococcal pneumonia. Copyright D. A. Warrell.

8.6.3  Pneumococcal infections 987 Because of the frequent coexistence of penicillin-​resistant aerobes and anaerobes, a β-​lactamase inhibitor or metronidazole should also be given. Antibiotics should be continued for 4 to 6 weeks. Course and prognosis If untreated, an empyema might rupture through the chest wall (em- pyema necessitatis) or into a bronchus causing a bronchopleural fis- tula. Even when pus is aspirated and healing achieved, subsequent fibrosis and calcification may seriously restrict expansion of the underlying lung. Pneumococcal meningitis Pneumococci colonizing the nasopharynx can gain access to the subarachnoid space either by direct spread (from paranasal sinusitis or otitis media), following damage to the base of the skull, or, more commonly, via the bloodstream where they cross the blood–​brain barrier at the choroid plexus and cerebral capillaries. Symptoms Adults with pneumococcal meningitis usually have fever, head- ache, neck stiffness, and impaired consciousness. At presentation, one-​half of all patients have been ill for less than 24 h. Nausea and photophobia are common and seizures occur before diagnosis in 5 to 10% of patients. Among elderly patients, confusion may be the only symptom. Deterioration in the psychological or neurological state of an elderly patient with community-​acquired pneumonia should be investigated with lumbar puncture. The presentation of meningitis in infants can be subtle, beginning with inability to feed and followed by irritability or lethargy. Physical signs Patients with pneumococcal meningitis are pyrexial and toxaemic. Classic signs such as nuchal rigidity, Kernig’s sign, and Brudzinski’s sign are absent in many patients with pyogenic meningitis. Bulging of the anterior fontanelle may be present in infants. Consciousness is often impaired, varying from drowsiness and confusion to deep coma. Raised intracranial pressure due to cerebral oedema or a cere- bral abscess may be indicated by bradycardia and hypertension, but papilloedema is rarely seen. A cranial CT or MRI is mandatory be- fore lumbar puncture in the presence of signs of cerebral or cranial nerve damage including a dilated pupil, ocular palsies, hemiparesis, history of focal seizures, decreased or rapidly falling level of con- sciousness, irregular respiration, tonic seizures, and decerebrate or decorticate posturing. An associated pneumococcal condition, such as otitis media or pneumonia, might be detected. Investigations Lumbar puncture should be undertaken whenever meningitis is suspected. In pneumococcal meningitis the cerebrospinal fluid is usually turbid and the leucocyte count is equal to or exceeds 1000 × 106/​litre. Most of the leucocytes are neutrophils. A few pa- tients have a low leucocyte count (<100 × 106/​litre) and in patients who present very early the leucocyte count can be normal; a repeat lumbar puncture several hours later will confirm the diagnosis. In pneumococcal meningitis, the concentration of protein in cere- brospinal fluid is increased and the ratio of glucose concentrations in cerebrospinal fluid and plasma is usually less than one-​third. In untreated cases, culture of cerebrospinal fluid is usually positive and pneumococci are visible following Gram’s staining. In patients who have received less than 48 h of antibiotic therapy the leucocyte count remains high and pneumococcal antigen may be detectable in cerebrospinal fluid. In developing countries, particularly in Asia, the use of immunochromatographic tests for C-​polysaccharide antigen in cerebrospinal fluid has increased the number of cases of pneumococcal meningitis diagnosed. Culture of blood also fre- quently reveals the pneumococcus. The peripheral white cell count is usually elevated. Differential diagnosis Pneumococcal meningitis cannot be differentiated clinically from other forms of meningitis and the aetiology must be defined by in- vestigation of the cerebrospinal fluid. An associated ear infection or pneumonia, or a history of head trauma favours pneumococcal infection. Conversely, rashes are rarely found in pneumococcal meningitis and petechiae or purpura on skin or mucosae strongly suggest meningococcal disease. Widespread use of pneumococcal conjugate vaccines has reduced the likelihood of S. pneumoniae aeti- ology in children with meninigitis. Treatment Standard empirical therapy for meningitis in adults is cefotaxime (300 mg/​kg per day divided into three or four doses) or ceftriaxone (100 mg/​kg per day divided into two doses) for 10 to 14 days. In parts of the world where strains with intermediate or full resistance to cefotaxime or ceftriaxone have emerged, vancomycin (60 mg/​ kg per day divided into four doses) should be added to the empiric therapy. Meropenem is a useful alternative to cefotaxime and is ac- tive against pneumococci of intermediate but not full cefotaxime resistance. Imipenem increases susceptibility to seizures. Penicillin or ampicillin are effective therapy for culture-​proven pneumococcal meningitis caused by penicillin-​sensitive strains, but intermediately resistant strains are not adequately treated by these drugs. In children in developing countries, outside of epidemics, the WHO recommends treatment with ceftriaxone 100 mg/​kg once daily for 5 to 7 days with a maximum daily dose of 2 g. A multicountry trial of 5 days versus 10 days of ceftriaxone therapy in children has provided empiric support for short-​course therapy for 5 days, pro- vided there is good clinical evidence of response. Treatment with antibiotics should be started as soon as a clin- ical diagnosis of bacterial meningitis is made. Delay in treatment until after hospitalization is associated with increased mortality. Other supportive therapies include adequate oxygenation, mainten- ance of normal blood pressure, prevention of hypoglycaemia and hyponatraemia, and control of seizures (which may be covert and unsuspected in an unconscious patient) with anticonvulsants. The use of dexamethasone in bacterial meningitis in adults has been controversial for many years. In a Cochrane review of five ran- domized controlled trials, the summary mortality reduction attribut- able to dexamethasone adjunctive treatment was 43%; the effect was greatest in meningitis caused by S. pneumoniae. The summary find- ings were influenced to a large extent by a single study of European patients. In a study of HIV-​infected adults in Africa, among whom case fatality rates were very high, dexamethasone was not beneficial, suggesting that dexamethasone might only be useful in populations with low HIV prevalence. The recommended dose is 10 mg every

section 8  Infectious diseases 988 6 h for 4 days. Meningeal inflammation facilitates diffusion of vanco- mycin into the cerebrospinal fluid and the anti-​inflammatory action of dexamethasone might lead to suboptimal antibiotic concentra- tions. Patients on treatment for cephalosporin-​resistant pneumo- coccal meningitis should therefore be monitored both clinically and by repeat lumbar puncture. In children, dexamethasone is highly protective against hearing loss in H. influenzae meningitis but also provides some protection in pneumococcal meningitis if given with or before administration of antibiotics. For children in developing countries, however, the evidence suggests there is no benefit to ad- junctive dexamethasone. Course and prognosis The prognosis of patients with pneumococcal meningitis is poor. Most patients develop complications of which the most important are seizures, brain infarction, brain swelling, hydrocephalus, and cranial nerve palsies. Subdural collections are commonly seen on brain imaging and may require needle puncture to exclude subdural empyema, especially if there is persistence of fever, irritability, neck stiffness, or continued cerebrospinal fluid leucocytosis detected by repeat lumbar puncture. Over one-​third of patients also develop systemic complications such as shock, cardiorespiratory failure, and disseminated intra- vascular coagulation, and these are frequently the final cause of death among older patients. Among children, supportive therapy to sustain adequate blood pressure is important to maintain cerebral blood flow against the resistance of raised intracranial pressure. The mortality from pneumococcal meningitis in industri- alized countries varies between 10 and 40%, being lower in children than in adults. In developing countries, the mortality range is higher (30–​60%). Features on admission that are asso- ciated with a poor outcome include advanced age, seizures, cra- nial nerve palsies, deep coma, low cerebrospinal fluid leucocyte count (below 1000×106/​litre), low glucose concentration in cere- brospinal fluid, and associated pneumonia. Death is almost in- evitable in patients who are in deep coma at the time they are admitted to hospital. Survivors are frequently affected by neuro- logical sequelae:  hearing loss occurs in one in five; cerebral damage is common leading to hemiparesis, ataxia, and aphasia; and cranial nerve palsies, particularly of the oculomotor nerve, occur in a small percentage. Among those who appear to make a good recovery from pneumococcal meningitis, one-​quarter have residual cognitive slowness. Otitis media In children aged below 2 years, otitis media is one of the most common reasons for seeking medical advice. The pneumococcus causes a significant fraction of all cases. Following conjugate pneumococcal vaccine introduction, the incidence of pneumo- coccal otitis media has fallen in several countries. Beyond child- hood, otitis media is uncommon. Symptoms Acute otitis media starts suddenly, although there may be a history of a recent upper respiratory tract infection. Fever, crying, and ex- treme irritability are the usual features in young children, in whom febrile convulsions may also occur. Fever and severe pain in the ear are the usual presenting complaints in older children and adults, and patients may also complain of deafness and tinnitus. Physical signs On otoscopic examination of the affected ear, the tympanic mem- brane is red and swollen and lacks the normal light reflection. It may bulge outwards into the external ear and there may be an air–​fluid level indicating a middle ear effusion. Pus or blood in the external auditory canal suggests a perforation that is confirmed by observing a ragged hole in the tympanic membrane. The affected ear is usually partially deaf. In children, meningism may be present; if so, menin- gitis must be excluded by lumbar puncture. Investigations Fine needle puncture of the tympanic membrane (tympanocentesis) and aspiration of middle ear fluid is used with variable frequency in different countries but is of most value where antibiotic resist- ance is prevalent. In complicated cases or in those not responding to initial antibiotics, culture of middle ear fluid may guide therapy. A tympanogram can identify increased middle ear pressure and ac- cumulation of fluid. Treatment Most episodes of otitis media are diagnosed clinically without micro- biological confirmation. Randomized controlled trials show that ot- itis media resolves in most otherwise healthy children whether or not they take antibiotics. This evidence underpins a policy of ob- servation without treatment. Immediate antibiotics are indicated in young infants (<6 months) and in older children (>2 years) with a clear bacteriological diagnosis of otitis media or symptoms of se- verity. The antibiotic of choice for pneumococcal otitis media is oral amoxicillin (90 mg/​kg per day) for 5 days. For penicillin-​resistant pneumococcal infection the appropriate antibiotic is intravenous ceftriaxone for 3 days. Course and prognosis Pneumococcal otitis media normally resolves rapidly and com- pletely. However, rupture of the drum can lead to partial conductive deafness and pneumococcal otitis media can give rise to a chronic discharging ear requiring prolonged or complicated treatment. The infection can spread to cause acute mastoiditis, meningitis, or a cere- bral abscess. Other clinical syndromes The pneumococcus is an important cause of bacterial sinusitis resulting from direct spread from the nasopharynx. Sinusitis that does not resolve within 5 to 7 days may require treatment with an antibiotic effective against pneumococcus. A mild form of pneumococcal bacteraemia, variously labelled as ‘occult bacteraemia’ or ‘walk-​in bacteraemia’, was encountered rela- tively commonly in children before the introduction of PCV. Significant bloodstream infection is less common but can lead to septicaemia or purulent localization in meninges, vertebrae, joints, orbits, or testes. Pneumococcal conjunctivitis has been observed in outbreaks among college students and has two un- usual features:  (1) the causative strain is unencapsulated and

8.6.3  Pneumococcal infections 989 (2) the attack rates are high, suggesting little pre-​existing im- munity. Pneumococci can also cause endophthalmitis, and are the commonest pathogen observed in case series of bacterial keratitis. Septicaemia Acute septicaemia is a less common form of pneumococcal infec- tion and is encountered most frequently in immunocomprom- ised patients or those without a spleen. Sudden fever, peripheral circulatory collapse, and bleeding (purpura fulminans) are the usual presenting features of this condition, which is indistinguish- able from other forms of overwhelming bacterial septicaemia. Leucopenia is usually found. Bleeding is due to disseminated intra- vascular coagulation. The mortality from septicaemia is very high, even when treatment is started promptly. The pneumococcus has been rarely associated with toxic shock syndrome and with haemo- lytic uraemic syndrome. Endocarditis and pericarditis Cardiac manifestations of pneumococcal infection are well de- scribed but where there is good access to antibiotics they are now rare, occurring in less than 1% of all pneumococcal infections. Acute endocarditis may complicate pneumococcal septicaemia to affect healthy heart valves, especially the aortic valve, which may rupture and cause severe aortic incompetence leading to valve replacement in those who survive the initial episode. Pneumococci may spread directly from the lower lobes of the lung to produce pericarditis which is clinically silent in some pa- tients or may be manifest only as a transient pericardial rub or an abnormal electrocardiogram. Patients with a pericardial empyema usually complain of dull or pleuritic central chest pain and give a history of persistent fever, malaise, anorexia, and weight loss over several days or weeks. Many patients with a pneumococcal pericar- dial empyema are critically ill by the time they reach hospital and have pericardial tamponade: a rapid small-​volume pulse, pulsus paradoxus, a low blood pressure, elevation of the jugular venous pressure with a further increase during inspiration, peripheral oe- dema, and ascites. The heart sounds are usually faint and a chest radiograph may show globular enlargement of the heart together with evidence of an associated lung infection. Ultrasound can help to define the best sites for diagnostic and therapeutic drainage. The electrocardiogram shows low-​voltage potentials and ST elevation or depression may be present. Pneumococcal aetiology can be con- firmed by culture or antigen detection of drained pus. Mortality is high, and among patients who survive the initial episode con- strictive pericarditis may develop within weeks or months of their acute illness. Peritonitis Pneumococcal peritonitis is an uncommon condition that is en- countered among three risk groups: (1) patients with cirrhosis of the liver or nephrotic syndrome; (2) patients with gastrointestinal disease (e.g. appendicitis), intra-​abdominal surgery, or peritoneal dialysis; and (3) otherwise healthy young girls, possibly as a com- plication of pelvic infection. The condition is characterized by sudden fever and abdominal pain and tenderness. The ascitic fluid is turbid and contains neutrophils and pneumococci. The prognosis of pneumococcal peritonitis is determined principally by the severity of the underlying illness. Future developments Childhood pneumococcal disease has been controlled in high-​ income settings and this is likely to be replicated across low-​income settings. Nonetheless, much pneumococcal disease remains. In many countries, particularly in Asia, the current PCV formulations do not target most disease-​causing serotypes. Among older adults in industrialized countries the incidence of pneumococcal infections caused by non​vaccine serotypes is significant and rising. Given that conjugate vaccines can prevent pneumonia in older adults, a conju- gate vaccine comprising the commoner serotypes not included in the current infant PCV formulations would make a significant con- tribution to health. Conjugate vaccines are expensive to produce and several lines of research are examining alternative methods to sustain vac- cine control in low-​income countries. These include novel in- expensive manufacturing technologies, strategies to sustain indirect protection with fewer doses of vaccine, and serotype-​ independent third generation vaccines or with fractional doses of vaccine. Several vaccines, consisting of either combinations of surface-​expressed proteins and virulence factors, or inactivated whole pneumococcal cells, are in phase I/​II studies. The perturbation of pneumococcal ecology by conjugate vac- cines has highlighted our ignorance of the natural habitat of Streptococcus pneumoniae. Current studies of colonizing pneumo- cocci are investigating inter-​ and intraspecies competition, bio- film development, interactions with respiratory viruses, and the mechanism of clearance induced by PCV. Studies in genomics, transcription and phylogenetic, are tackling the same problems at pathogen-​population level. As persistent colonizers of children, pneumococci are exposed repeatedly to antibiotics, encouraging the development and dis- semination of antibiotic resistance. Resistance to new agents such as fluoroquinolones is established in adult disease and has been re- ported among paediatric cases in South Africa. The problem is only likely to worsen unless antibiotic usage in low-​income countries can be brought under control through judicious and rational pre- scribing. At the same time, the clinical relevance of in vitro resist- ance to antibiotics, especially intermediate β-​lactam resistance in the treatment of pneumonia, remains an interesting area requiring further investigation. FURTHER READING Austrian R, Gold J (1964). Pneumococcal bacteremia with especial reference to bacteremic pneumococcal pneumonia. Ann Intern Med, 60, 759–​76. Austrian R, et al. (1976). Prevention of pneumococcal pneumonia by vaccination. Trans Assoc Am Physicians, 89, 184–​94. Bentley SD, et al. (2006). Genetic analysis of the capsular biosyn- thetic locus from all 90 pneumococcal serotypes. PLoS Genet, 2, e31, 1–​8.

section 8  Infectious diseases 990 Berkley JA, et al. (2005). Bacteremia among children admitted to a rural hospital in Kenya. N Engl J Med, 352, 39–​47. Black S, et al. (2000). Efficacy, safety and immunogenicity of hepta- valent pneumococcal conjugate vaccine in children. Northern California Kaiser Permanente Vaccine Study Center Group. Pediatr Infect Dis J, 19, 187–​95. Bonten MJ, et  al. (2015). Polysaccharide conjugate vaccine
against pneumococcal pneumonia in adults. N Engl J Med, 372, 1114–​25. Briles DE, et al. (2000). Immunization of humans with recombinant pneumococcal surface protein A (rPspA) elicits antibodies that passively protect mice from fatal infection with Streptococcus pneumoniae bearing heterologous PspA. J Infect Dis, 182, 1694–​701. Brouwer MC, et al. (2015). Corticosteroids for acute bacterial menin- gitis. Cochrane Database Syst Rev, 5, CD004405. Castelblanco RL, et al. (2014). Epidemiology of bacterial meningitis in the USA from 1997 to 2010: a population-​based observational study. Lancet Infect Dis, 14, 813–​9. Cobey S, Lipsitch M (2012). Niche and neutral effects of acquired im- munity permit coexistence of pneumococcal serotypes. Science, 335, 1376–​80. Croucher NJ, et al. (2011). Rapid pneumococcal evolution in response to clinical interventions. Science, 331, 430–​4. Cutts FT, et al. (2005). Efficacy of nine-​valent pneumococcal conju- gate vaccine against pneumonia and invasive pneumococcal disease in The Gambia: randomised, double-​blind, placebo-​controlled trial. Lancet, 365, 1139–​46. Dowell SF, et al. (2003). Seasonal patterns of invasive pneumococcal disease. Emerg Infect Dis, 9, 573–​9. Dowson CG, et al. (1989). Horizontal transfer of penicillin-​binding protein genes in penicillin-​resistant clinical isolates of Streptococcus pneumoniae. Proc Natl Acad Sci U S A, 86, 8842–​6. French N, et al. (2010). A trial of a 7-​valent pneumococcal conjugate vaccine in HIV-​infected adults. N Engl J Med, 362, 812–​22. Geno KA et al. (2015). Pneumococcal capsules and their types: past, present, and future. Clin Microbiol Rev, 28, 871–​99. Gessner BD, et al. (2010). African meningitis belt pneumococcal dis- ease epidemiology indicates a need for an effective serotype 1 con- taining vaccine, including for older children and adults. BMC Infect Dis, 10, 22. Ginsburg AS, et al. (2012). Issues and challenges in the develop- ment of pneumococcal protein vaccines. Expert Rev Vaccines, 11, 279–​85. Gordon SB, et  al. (2000). Bacterial meningitis in Malawian adults: pneumococcal disease is common, severe, and seasonal. Clin Infect Dis, 31, 53–​7. Griffin MR et  al. (2013). U.S.  hospitalizations for pneumonia after a decade of pneumococcal vaccination. N Engl J Med, 369,
155–​63 Heffron R (1939). Pneumonia with special reference to pneumo- coccus lobar pneumonia. The Commonwealth Fund, New York, NY. (Second printing 1979, Harvard University Press, Cambridge, MA) Johnson HL, et al. (2010). Systematic evaluation of serotypes causing invasive pneumococcal disease among children under five:  the pneumococcal global serotype project. PLoS Med, 7, e1000348. Kadioglu A, et al. (2008). The role of Streptococcus pneumoniae viru- lence factors in host respiratory colonization and disease. Nat Rev Microbiol, 6, 288–​301. Klugman KP, et al. (2003). A trial of a 9-​valent pneumococcal con- jugate vaccine in children with and those without HIV infection.
N Engl J Med, 349, 1341–​8. Lim WS, et al. (2009). British Thoracic Society guidelines for the man- agement of community acquired pneumonia in adults: update 2009. Thorax, 64, Suppl 3, iii, 1–​55. Malley R, et  al. (2001). Intranasal immunization with killed unencapsulated whole cells prevents colonization and invasive dis- ease by capsulated pneumococci. Infect Immun, 69, 4870–​3. Moberley S et al. (2013). Vaccines for preventing pneumococcal infec- tion in adults. Cochrane Database Syst Rev, 1, CD000422. Musher DM (1992). Infections caused by Streptococcus pneumo- niae: clinical spectrum, pathogenesis, immunity, and treatment. Clin Infect Dis, 14, 801–​7. Pallares R, et al. (1995). Resistance to penicillin and cephalosporin and mortality from severe pneumococcal pneumonia in Barcelona, Spain. N Engl J Med, 333, 474–​80. Palmu AA, et al. (2013). Effectiveness of the ten-​valent pneumococcal Haemophilus influenzae protein D conjugate vaccine (PHiD-​CV10) against invasive pneumococcal disease: a cluster randomised trial. Lancet, 381, 214–​22. Pilishvili T, et al. (2010). Sustained reductions in invasive pneumo- coccal disease in the era of conjugate vaccine. J Infect Dis, 201, 32–​41. Roca A, et  al. (2011). Effects of community-​wide vaccination with PCV-​7 on pneumococcal nasopharyngeal carriage in the Gambia: a cluster-​randomized trial. PLoS Med, 8, e1001107. Scott JA, et al. (2000). Aetiology, outcome, and risk factors for mor- tality among adults with acute pneumonia in Kenya. Lancet, 355, 1225–​30. Tettelin H, et al. (2001). Complete genome sequence of a virulent iso- late of Streptococcus pneumoniae. Science, 293, 498–​506. Trzcinski K, Thompson CM, Lipsitch M (2004). Single-​step cap- sular transformation and acquisition of penicillin resistance in Streptococcus pneumoniae. J Bacteriol, 186, 3447–​52. Tuomanen EI, et  al. (eds) (2004). The pneumococcus. ASM Press, Washington, DC. van der Poll T, Opal SM (2009). Pathogenesis, treatment, and preven- tion of pneumococcal pneumonia. Lancet, 374, 1543–​56. von Gottberg A et  al. (2014). Effects of vaccination on invasive pneumococcal disease in South Africa. N Engl J Med, 371, 1889–​99. Wahl B, et  al. (2018). Burden of Streptococcus pneumoniae and Haemophilus influenzae type b disease in children in the era of
conjugate vaccines: global, regional, and national estimates for 2000-2015. Lancet Global Hlth, 6, e744–757. Waight PA, et al. (2015). Effect of the 13-​valent pneumococcal con- jugate vaccine on invasive pneumococcal disease in England and Wales 4 years after its introduction: an observational cohort study. Lancet Infect Dis, 15, 535–​43. Watera C, et al. (2004). 23-​Valent pneumococcal polysaccharide vac- cine in HIV-​infected Ugandan adults: 6-​year follow-​up of a clinical trial cohort. AIDS, 18, 1210–​3. Watson DA, et al. (1993). A brief history of the pneumococcus in bio- medical research: a panoply of scientific discovery. Clin Infect Dis, 17, 913–​24. Weinberger DM, et al. (2011). Serotype replacement in disease after pneumococcal vaccination. Lancet, 378, 1962–​73. Weiser JN, et al. (1994). Phase variation in pneumococcal opacity: re- lationship between colonial morphology and nasopharyngeal col- onization. Infect Immun, 62, 2582–​9.