gkHub

Streptococcus pneumoniae (the pneumococcus) was recognized as a major cause of pneumonia in the 1880s. Although the name Diplococcus pneumoniae was originally assigned to the pneumococcus, the organism was renamed Streptococcus pneumoniae because, like other streptococci, it grows in chains in liquid medium.Widespread vaccination has reduced the incidence of pneumococcal infection, but this organism remains the principal bacterial cause of otitis media, acute purulent rhinosinusitis, pneumonia, and meningitis.


MICROBIOLOGY

Pneumococci are identified in the clinical laboratory as catalase-negative, gram-positive cocci that grow in pairs or chains and cause á-hemolysis on blood agar. More than 98% of pneumococcal isolates are susceptible to ethylhydrocupreine (optochin), and virtually all pneumococcal colonies are dissolved by bile salts. Peptidoglycan and teichoic acid are the principal constituents of the pneumococcal cell wall, whose integrity depends on the presence of numerous peptide side chains cross-linked by the activity of enzymes such as trans- and carboxypeptidases. á-Lactam antibiotics inactivate these enzymes by covalently binding their active site. Unique to S. pneumoniae and present in all strains is C-substance (“cell-wall” substance), a polysaccharide consisting of teichoic acid with a phosphorylcholine residue. Surface-exposed choline residues serve as a site of attachment for potential virulence factors, such as pneumococcal surface protein A (PspA) and pneumococcal surface adhesin A (PsaA), which may prevent phagocytosis. Except for strains that cause conjunctivitis, nearly every clinical isolate of S. pneumoniae has a polysaccharide capsule, a structure that renders the bacteria virulent by preventing phagocytosis. All strains produce pneumolysin, a toxin that may cause many of the manifestations of pneumococcal infection. There are 90 serologically distinct capsules of S. pneumoniae. Serotyping remains clinically relevant because the activity of available vaccines is based on stimulating antibody to specific capsular polysaccharides.


EPIDEMIOLOGY

S. pneumoniae colonizes the nasopharynx and, on any single occasion, can be isolated from 5–10% of healthy adults and from 20–40% of healthy children. Once adults are colonized, organisms are likely to persist for 4–6 weeks but may be present for as long as 6 months. Pneumococci spread from one individual to another by direct or droplet transmission as a result of close contact; transmission may be enhanced by crowding or poor ventilation. Day-care centers have been a site of spread, especially of penicillin-resistant strains of serotypes 6B, 14, 19F, and 23F. Outbreaks of pneumococcal disease occur among adults in crowded living conditions—e.g., in military barracks, prisons, and shelters for the homeless—as well as among susceptible populations in settings such as nursing homes. The risk of pneumococcal pneumonia is generally not increased by contact in schools or workplaces (including hospitals). The incidence data provided below were obtained before widespread administration of pneumococcal conjugate vaccine to infants and children. (For the impact of widespread vaccination, see “Prevention” later in the chapter.) In the absence of vaccination (which alters natural history), invasive pneumococcal disease is, by far, most prevalent among children <2 years old. The incidence is low among older children and adults <65 years of age but then rises in older adults.


The fatality rate is also highest at the extremes of age. One surveillance study in the late 1980s found incidences of pneumococcal bacteremia among infants, young adults, and persons ≥70 years of age to be 160, 5, and 70 cases per 100,000 population, respectively. Most cases of pneumococcal bacteremia in adults are due to pneumonia, and there are 3–4 cases of nonbacteremic pneumonia for every 375 bacteremic case. Thus an estimated 20 cases of pneumococcal pneumonia per 100,000 young adults and 280 cases per 100,000 persons over the age of 70 occur annually. The disease is more frequent among men than among women. The incidence of pneumococcal bacteremia among adults exhibits a distinct midwinter peak and a striking dip in summer; in children, the incidence is relatively constant throughout the year except for a marked dip in midsummer. For reasons that are unclear but probably multifactorial, Native Americans, Native Alaskans, and African Americans are more susceptible to invasive pneumococcal disease than are Caucasians. Natives of the Pacific Rim region are likewise more susceptible.


PATHOGENETIC MECHANISMS

Infection results when pneumococci colonizing the nasopharynx are carried into anatomically contiguous areas (e.g., the eustachian tubes, the nasal sinuses) and bacterial clearance is hindered (e.g., by mucosal edema due to allergy or viral infection). Clearly, the resistance of pneumococci to phagocytosis is central to their capacity to cause infection. Pneumonia ensues when organisms are inhaled or aspirated into the bronchioles or alveoli and are not cleared—especially, for example, if mucus production is increased and/or ciliary action is damaged by viral infection or by cigarette smoke or other toxic substances. Viral infection may also inhibit clearance by upregulating pneumocyte receptors that bind pneumococci. In normally sterile sites, such as the sinuses or the lungs, pneumococci activate complement, stimulating the production of cytokines that attract polymorphonuclear leukocytes (PMNs). The polysaccharide capsule, however, renders the pneumococci resistant to phagocytosis. In the absence of anticapsular antibody, a large bacterial inoculum and/or a compromise of phagocytic function allows the initiation of infection. Infection of the meninges, joints, bones, and peritoneal cavity may result from pneumococcal spread through the bloodstream, usually from a respiratory tract focus of infection.


Unencapsulated pneumococci virtually never cause invasive disease, although they can cause conjunctivitis. Symptoms of disease are largely attributable to the inflammatory response, which may cause pain by increasing pressure (as in sinusitis or otitis media) or may interfere with vital bodily functions by preventing oxygenation of blood (as in pneumonia) or by inhibiting blood flow (as in vasculitis due to meningitis). Cell-wall constituents of S. pneumoniae, especially peptidoglycan, activate complement by the alternative pathway; the reaction between cell-wall structures and antibody (present in all humans) also activates the classic complement pathway.The result is the release of C5a, a potent attractant for PMNs, into the surrounding medium. Peptidoglycan can also directly stimulate the release of proinflammatory cytokines such as interleukin (IL) 1â, tumor necrosis factor (TNF) á, and IL-6. All pneumococci generate pneumolysin, a toxin that damages ciliary cells and PMNs and also activates the classic complement pathway. Injection of pneumolysin into the lungs of experimental animals produces the histologic features of pneumonia; in mice, immunization with this substance or challenge with genetically engineered mutants that do not produce it is associated with a significant reduction in virulence. HOST


DEFENSE MECHANISMS

Mechanisms of host defense may be nonimmunologic or immunologic. Immunologic mechanisms may be natural (innate) or specific (humoral). Nonimmunologic Mechanisms Nonimmunologic mechanisms that protect against pneumonia include filtration of air as it passes through the nasopharynx, the glottal reflex, laryngeal closure, the cough reflex, clearance of organisms from the lower airways by ciliated cells, and ingestion by pulmonary macrophages and PMNs of small bacterial inocula that manage to reach alveolar spaces. Respiratory virus infection, chronic pulmonary disease, or heart failure compromises these mechanisms, predisposing to the development of pneumococcal pneumonia. Immunologic Mechanisms Innate Immunity Innate immune mechanisms participate in clearance of pneumococci from the nasopharynx as well as in phagocytosis by PMNs and macrophages via the microbial pattern recognition receptor Toll-like receptor 2 (TLR2).


Humoral Immunity Immunologically specific humoral mechanisms provide the best protection against pneumococcal infection. Most healthy adults have antibody to constituents of S. pneumoniae, such as PspA, PsaA, and the cell wall; however, there is no convincing evidence for an opsonic role of these antibodies, especially at their usual concentrations. Most healthy adults lack IgG antibody to the majority of pneumococcal capsular polysaccharides. Antibody appears after colonization, infection, or vaccination. In the first few weeks after colonization, nonspecific mechanisms probably protect the host from infection. Thereafter, newly developed anticapsular antibody provides a high degree of specific protection. Adults who are at risk of aspirating pharyngeal contents and/or who have diminished mechanisms of lower airway clearance are at risk of developing pneumonia before antibody is produced. Persons with a diminished capacity to form antibody probably remain susceptible as long as they are colonized. The risk of serious pneumococcal infection is greatly increased in persons with conditions that compromise IgG synthesis and/or the phagocytic function of PMNs and macrophages. Most patients hospitalized for pneumococcal pneumonia have one or more of these conditions (Table 34-1). Once a pneumococcal infection has been initiated, the absence of a spleen predisposes to fulminant disease. The liver can remove opsonized (antibody-coated) pneumococci from the circulation; in the absence of antibody, however, only the slow passage of blood through the splenic sinuses and prolonged contact with reticuloendothelial cells in the cords of Billroth can result in bacterial clearance. Patients without spleens tend to develop overwhelming pneumococcal disease that rapidly progresses to death.


SPECIFIC INFECTIONS CAUSED BY S. PNEUMONIAE

S. pneumoniae causes infections of the middle ear, sinuses, trachea, bronchi, and lungs (Table 34-2) by direct spread from the nasopharyngeal site of colonization. Infections of the central nervous system (CNS), heart valves, bones, joints, and peritoneal cavity usually arise by hematogenous spread. Peritoneal infection may also result from ascent via the fallopian tubes. The CNS may also be infected by drainage from nasopharyngeal lymphatics or veins or by contiguous spread of organisms (e.g., through a tear in the dura). Primary pneumococcal bacteremia—i.e., the presence of pneumococci in the blood with no apparent source—occurs commonly in children <2 years of age and accounts for a small percentage of all cases of pneumococcal bacteremia in adults; if no therapy is given, a source and/or a secondary site of infection may become apparent. Pleural infection results either from direct extension of pneumonia to the visceral pleura or from hematogenous bacterial spread from a pulmonary or extrapulmonary focus to the pleural space; the route usually cannot be determined in any individual case. Infections listed after meningitis in Table 34-2 are uncommon or rare.


Otitis Media and Sinusitis Otitis media and acute rhinosinusitis are similar in terms of pathogenesis. Bacteria are trapped in a normally sterile site when drainage is impaired, often as a result of viral infection, allergies, or exposure to pollutants (including cigarette smoke). In both disease states, S. pneumoniae is the most common or second most common isolate (after nontypable Haemophilus influenzae) from cultures of the infected site. Pneumonia The distinctive symptoms and signs of pneumonia, whether due to the pneumococcus or to other bacteria, are (1) cough and sputum production, which reflect bacterial proliferation and the resulting inflammatory response in the alveoli; (2) fever; and (3) radiographic detection of an infiltrate. Predisposing Conditions Pneumococcal pneumonia is most common at the extremes of age. Despite the undisputed role of S. pneumoniae as a major pathogenic bacterium for humans, the great majority of adults with pneumococcal pneumonia have underlying diseases that predispose them to infection. Otherwise-healthy military recruits involved in outbreaks of infection may be an exception to this rule; however, many of these individuals have been under extreme physical and/or psychological stress and/or have had an antecedent viral-type illness that may have reduced their normal host resistance. Infections with respiratory viruses, especially influenza virus, predispose to pneumococcal pneumonia.


Other common predisposing conditions are alcoholism, malnutrition, chronic pulmonary disease of any kind (including asthma), cigarette smoking, HIV infection, diabetes mellitus, cirrhosis of the liver, anemia, prior hospitalization for any reason, renal insufficiency, and coronary artery disease (with or without recognized congestive heart failure). In elderly subjects, the predisposition is generally multifactorial. Presenting Symptoms Patients often present with a clear exacerbation of a preexisting respiratory condition.They may have felt unwell for several days, with coryza or a nonproductive cough and low-grade fever, but they feel distinctly worse at the time of onset of pneumonia. Coughing, often productive of purulent sputum, becomes prominent. The temperature may rise to 38.9°–39.4°C (102°–103°F), although a substantial proportion of patients are afebrile at admission. In a small proportion of cases, the onset of disease follows a hyperacute pattern in which the patient suddenly has a single episode of shaking chills followed by sustained fever and a cough productive of blood-tinged sputum.


In the elderly, the onset of disease may be especially insidious and may not suggest pneumonia at all. Such persons may have minimal cough, no sputum production, and no fever, instead appearing tired or confused. Nausea and vomiting or diarrhea occurs in up to 20% of cases of pneumococcal pneumonia. Symptoms of a new cardiac arrhythmia, myocardial ischemia, or an actual infarction occur in 10% of patients at a veterans’ hospital who are admitted for pneumonia, and these manifestations may even predominate.The pneumonia may precipitate cardiogenic or noncardiogenic pulmonary edema. Pleuritic chest pain may result from extension of the inflammatory process to the visceral pleura; persistence of this pain, especially after the first day or two of treatment, raises concern about empyema (see “Complications” later in the chapter). Clearly, the range of symptoms is sufficiently broad that no characteristic presentation distinguishes pneumococcal pneumonia from other types of bacterial pneumonia or from some types of nonbacterial pneumonia. Physical Findings Patients with pneumococcal pneumonia usually appear ill and have a grayish, anxious appearance that differs from that of persons with viral or mycoplasmal pneumonia. Temperature, pulse, and respiratory rate are typically elevated. Elderly patients may have only a slight temperature elevation or may be afebrile.


Hypothermia may be documented instead of fever and is associated with increased morbidity and mortality. Pleuritic chest pain may cause diminished respiratory excursion (splinting) on the affected side. Dullness to percussion is noted in about half of cases, and vocal fremitus is increased over the area of consolidation. Breath sounds may be bronchial or tubular, and crackles are heard in most cases if enough air is being moved to generate them. Flatness to percussion at the lung base, absent fremitus, and lack of the expected degree of diaphragmatic motion suggest the presence of pleural fluid, which raises the possibility of empyema. The finding of a heart murmur—certainly if new—raises concern about endocarditis, a rare but serious complication. Hypoxia or the generalized response to pneumonia may cause the patient to be confused, but the appearance of confusion should also raise concern about meningitis. Obtundation or neck stiffness should lead to an immediate consideration of this complication. Radiographic Findings In patients sick enough to be hospitalized, pneumococcal pneumonia is limited to one lung segment in onefourth of cases and to one lobe in another one-fourth, with multilobar disease in the remaining one-half. Air-space consolidation is the predominant finding and is detected in 80% of cases (Fig. 34-1). Air bronchogram (visualization of the air-filled bronchus against a background of alveolar consolidation) is evident in fewer than half of cases and is more common in bacteremic than in nonbacteremic disease.


Rarely, pneumococcal pneumonia leads to a lung abscess.Although some pleural fluid may actually be present in half of cases, ≤20% of patients have a sufficient volume of fluid to allow aspiration, and in only a minority of these patients is empyema documented. General Laboratory Findings Anemia (hemoglobin level, <10 g/dL) is documented in 25% of cases. The peripheral-blood white blood cell (WBC) count exceeds 12,000/ìL in the great majority of patients with pneumococcal pneumonia. A low WBC count (<6000/ìL) is found in 5–10% of persons hospitalized for pneumococcal pneumonia and is strongly associated with fatal disease.The serum bilirubin level is modestly elevated in one-third of cases; hypoxia, inflammatory changes in the liver, and breakdown of red blood cells in the lung are all thought to contribute to this increase.A serum albumin level of <2.5 g/dL in 30% of cases may indicate predisposing malnutrition or may be the result of sepsis. About 20% of patients have serum sodium concentrations of ≤130 meq/L, and another 20% have serum creatinine concentrations of ≥2 mg/dL. Abnormalities of pleural fluid in empyema are reviewed in Chap. 17. Differential Diagnosis S. pneumoniae is the most common cause of so-called community-acquired pneumonia, but patients who present with this syndrome may actually have infection due to a broad array of microorganisms.


The extensive list includes (but is not limited to) the following: H. influenzae or Moraxella catarrhalis in persons with little to predispose them other than chronic or acute inflammation of the airways; Staphylococcus aureus, especially in persons who take glucocorticoids, who have influenza, or who have major anatomic disruption of the airways; Streptococcus pyogenes; Neisseria meningitidis; anaerobic and microaerophilic bacteria in persons who may have aspirated oropharyngeal contents; Legionella; Pasteurella multocida in dog or cat owners; gram-negative bacilli, especially in persons who have severely damaged lungs and are taking glucocorticoids; viruses, especially influenza virus (in season), adenovirus, or respiratory syncytial virus; Mycobacterium tuberculosis; fungi, including Pneumocystis (depending on epidemiologic factors and HIV infection status); Mycoplasma; Chlamydia pneumoniae, especially in older adults; and Chlamydia psittaci in bird owners.


Many older men with lung cancer present with pneumonia, as do persons who have acute-onset inflammatory pulmonary conditions of uncertain etiology or those with pulmonary embolus and infarction. The breadth of this list vividly illustrates the deficiency of empirical therapy for community-acquired pneumonia (Table 34-3). Many of these diseases require evaluation, and the increasing availability of specific therapy makes a precise etiologic diagnosis desirable. Diagnostic Microbiology In patients with community-acquired pneumonia, a pneumococcal etiology is strongly suggested by the microscopic demonstration of large numbers of PMNs and slightly elongated gram-positive cocci in pairs and chains in the sputum. A sample such as the one shown in Fig. 34-2 is highly specific for pneumococcal infection of the lower airways. In the absence of such microscopic findings, the identification of pneumococci by culture is less specific, possibly reflecting colonization of the upper airways. Prior treatment with antibiotics can rapidly clear pneumococci from sputum. These factors need to be considered when sputum cultures from patients who appear to have pneumococcal pneumonia are said to yield only “normal mouth flora” and when the medical literature describes what appear to be poor results of sputum culture.


A study of sputum Gram’s stain and culture in patients with proven (bacteremic) pneumococcal pneumonia showed that about half of patients could not provide a sputum sample, provided a sample of poor quality, or had received antibiotics for >18 h; results in the remaining cases showed >80% sensitivity of microscopic examination of a Gram-stained sputum sample and 90% sensitivity of a sputum culture. Blood cultures yield S. pneumoniae in ∼25% of patients hospitalized for pneumococcal pneumonia. Complications Empyema is the most common complication of pneumococcal pneumonia, occurring in ∼2% of cases. Some fluid appears in the pleural space in a substantial proportion of cases of pneumococcal pneumonia, but this parapneumonic effusion usually reflects an inflammatory response to infection that has been contained within the lung, and its presence is self-limited. When bacteria reach the pleural space—either hematogenously or as a result of contiguous spread, possibly across lymphatics of the visceral pleura—empyema results. The finding of frank pus, bacteria (by microscopic examination), or fluid with a pH of ≤7.1 indicates the need for aggressive and complete drainage, preferably by prompt insertion of a chest tube, with verification by CT that fluid has been removed. Failure to drain most or all of the fluid indicates the need for additional treatment, including placement of other tube(s) (thoracostomy) or thoracotomy. Empyema is likely if fluid is present and fever and leukocytosis (even low-grade) persist after 4–5 days of appropriate antibiotic treatment for pneumococcal pneumonia.


At this stage, thoracotomy is often needed for cure. Aggressive drainage is likely to reduce morbidity and mortality from empyema. Meningitis Except during outbreaks of meningococcal infection, S. pneumoniae is the most common cause of bacterial meningitis in adults. Because of the remarkable success of H. influenzae type b vaccine, S. pneumoniae now predominates among cases in infants and toddlers as well (but not among those in newborns); nevertheless, the incidence of pneumococcal meningitis among children has been dramatically reduced by use of the pediatric pneumococcal conjugate vaccine (see “Prevention” later in the chapter). No distinctive clinical or laboratory features differentiate pneumococcal meningitis from other bacterial meningitides. Patients note the sudden onset of fever, headache, and stiffness or pain in the neck.Without treatment, there is a progression over 24–48 h to confusion and then obtundation. On physical examination, the patient looks acutely ill and has a rigid neck. In such cases, lumbar puncture should not be delayed for CT of the head unless papilledema or focal neurologic signs are evident.Typical findings in cerebrospinal fluid (CSF) consist of an increased WBC count (500–10,000 cells/ìL) with ≥85% PMNs, an elevated protein level (100–500 mg/dL), and a decreased glucose level (<30 mg/dL).


If antibiotics have not been given, large numbers of pneumococci are seen in Gram-stained CSF in virtually all cases, and specific therapy can be administered, although, because of its similar appearance, Listeria may be misidentified as the pneumococcus. If an effective antibiotic has already been given, the number of bacteria may be greatly decreased and microscopic examination of a Gram-stained specimen may yield negative results. In this situation, immunologic methods may detect pneumococcal capsule in the CSF in up to two-thirds of cases. Other Syndromes The appearance of pneumococcal infection at other, ordinarily sterile body sites indicates hematogenous spread, usually during frank pneumonia or, in a small proportion of cases, from an inapparent focus of infection. A case of pneumococcal endocarditis is seen every few years at large tertiary-care hospitals. Purulent pericarditis, occurring as a separate entity or together with endocarditis, is even rarer. The name Austrian’s syndrome is given to the concurrence of pneumococcal pneumonia, endocarditis, and meningitis. Septic arthritis can arise spontaneously in a natural or prosthetic joint or as a complication of rheumatoid arthritis. Osteomyelitis in adults tends to involve vertebral bones. Pneumococcal peritonitis occurs by one of three pathogenetic pathways: (1) hematogenous spread when ascites or other preexisting peritoneal disease is present; (2) local spread from a perforated viscus (usually appendicitis or perforated ulcer); or (3) transit via the fallopian tubes. Salpingitis may be recognized with or without accompanying peritonitis. Epidural and brain abscesses arise as a complication of sinusitis or mastoiditis. Cellulitis is also uncommon, developing most often in persons who have connective tissue diseases or HIV infection. The appearance of any of these unusual pneumococcal infections may suggest that tests for HIV infection should be undertaken. Finally, for reasons that are unclear, unencapsulated (but not encapsulated) pneumococci may cause sporadic or epidemic conjunctivitis.


PNEUMOCOCCAL INFECTIONS ANTIBIOTIC SUSCEPTIBILITY

Beta-Lactam antibiotics, the cornerstone of therapy for serious pneumococcal infection, bind covalently to the active site and thereby block the action of enzymes (endo-, trans-, and carboxypeptidases) needed for cell-wall synthesis. Because these enzymes were identified by their reaction with radiolabeled penicillin, they are called penicillinbinding proteins. Until the late 1970s, virtually all clinical isolates of S. pneumoniae were susceptible to penicillin (i.e., were inhibited in vitro by concentrations of <0.06 ìg/mL). Since then, an increasing number of isolates have shown some degree of resistance to penicillin. Resistance results when spontaneous mutation or acquisition of new genetic material alters penicillinbinding proteins in a manner that reduces their affinity for penicillin, thereby necessitating a higher concentration of penicillin for their saturation. The genetic information that renders pneumococci resistant to penicillin is acquired from oral streptococci and is transmitted along with genes that convey resistance to other antibiotics as well. Selection of antibiotic-resistant strains worldwide—especially in countries where antibiotics are available without prescription and in loci of high antibiotic use, such as day-care centers—greatly contributes to the prevalence of multidrug resistance. At present, ∼20% of pneumococcal isolates in the United States exhibit intermediate resistance to penicillin [minimal inhibitory concentration (MIC) 0.1–1.0 ìg/mL], and 15% are resistant (MIC ≥2.0 ìg/mL; Fig. 34-3).


The rate of resistance is lower in countries that, by tradition, are conservative in their antibiotic use (e.g., Holland and Germany) and higher in countries where usage is more liberal (e.g., France). In Hong Kong and Korea, resistance rates approach 80%. These definitions of resistance, however, were based on drug levels achievable in CSF during treatment of meningitis, whereas levels reached in the bloodstream, lungs, and sinuses are actually much higher. Thus the MIC needs to be interpreted in light of the infection being treated. Pneumonia caused by a penicillin-resistant strain is likely to respond to conventional doses of â-lactam antibiotics, whereas meningitis may not. The recently revised definition of amoxicillin resistance (susceptible, MIC ≤2 ìg/mL; intermediately resistant, MIC = 4 ìg/mL; and resistant, MIC ≥8 ìg/mL) is based on susceptibility to serum levels, with the assumption that no physician would knowingly treat meningitis with this oral medication.


Pneumonia due to a pneumococcal strain with intermediate amoxicillin resistance is still likely to respond to treatment with this drug, whereas that due to a resistant strain may not. On the assumption that antibiotic concentrations in middle-ear fluid or sinus cavities approach those in serum, similar inferences can be made about the treatment of otitis or sinusitis. Penicillin-susceptible pneumococci are susceptible to all commonly used cephalosporins.Penicillin-intermediate strains tend to be resistant to all first- and many secondgeneration cephalosporins (of which cefuroxime retains the best efficacy), but most are susceptible to certain third-generation cephalosporins, including cefotaxime, ceftriaxone, cefepime, and the oral cefpodoxime. Onehalf of highly penicillin-resistant pneumococci are also resistant to cefotaxime, ceftriaxone, and cefepime, and nearly all are resistant to cefpodoxime. Just as in the case of penicillin, susceptibility to cefotaxime and ceftriaxone is defined on the basis of achievable CSF levels. Thus pneumonia caused by intermediately resistant strains (MIC = 2 ìg/mL) still responds well to usual doses of these drugs, and pneumonia due to a resistant organism (MIC ≥4 ìg/mL) is likely to respond.


Meningitis due to intermediately resistant strains may not respond, and meningitis due to a resistant strain is likely not to respond to treatment with cefotaxime or ceftriaxone. About one-quarter of all pneumococcal isolates in the United States are resistant to erythromycin and the newer macrolides, including azithromycin and clarithromycin, with much higher rates of resistance among penicillin-resistant strains. This resistance will certainly affect empirical therapy for bronchitis, sinusitis, and pneumonia. In the United States, the majority of macrolide-resistant pneumococci bear the so-called M phenotype (erythromycin MIC = 1–8 ìg/mL) and are susceptible to clindamycin. In this case, resistance is mediated by an efflux pump mechanism; to some extent, M-type resistance can be overcome by clinically achievable levels of macrolides. In Europe, most macrolide resistance is due to a mutation in ermB, which confers highlevel resistance not only to macrolides but also to clindamycin; >90% of pneumococcal isolates in the United States are susceptible to clindamycin. Rates of doxycycline resistance are similar to those observed for macrolides. One-third of pneumococcal isolates are resistant to trimethoprim-sulfamethoxazole. The newer fluoroquinolones remain effective against pneumococci; the rate of resistance is generally <2–3% in the United States but is higher elsewhere and may be much higher in closed environments where these drugs are heavily prescribed, such as nursing homes and assisted-living facilities. Ketolides (such as telithromycin) appear to be uniformly effective against pneumococci, as does vancomycin.


ANTIBIOTIC REGIMENS

Otitis Media (Table 34-4) Current treatment recommendations for otitis media are based on the following points: (1) Acute otitis media is the most common diagnosis leading to an antibiotic prescription in the United States. (2) The diagnosis is often based on inadequate evidence for true middle-ear infection. (3) In proven cases, S. pneumoniae and H. influenzae are the most likely causes. (4) Because penetration into a closed space may be reduced, high serum levels of an effective antibiotic are required to treat otitis caused by intermediately or fully resistant pneumococci. (5) S. pneumoniae is more likely than Haemophilus and much more likely than Moraxella to cause progression to serious complications without specific therapy. (6) Antibiotics that are effective against pneumococci and yet resist â-lactamases tend to be very expensive compared with amoxicillin. As a result of these considerations, the American Academies of Pediatrics and Family Practice recommend that clinicians apply due diligence in diagnosing otitis. In children 6 months to 2 years of age with nonsevere illness and an uncertain diagnosis and in children >2 years of age with nonsevere illness (even if the diagnosis seems certain), symptom-based therapy and observation may be used instead of antimicrobial therapy. When parents of children with otitis are given a prescription for an antibiotic but are instructed not to fill it unless the disease progresses, no antibiotic is given in many cases, yet rates of patient satisfaction are high. If otitis media is clearly diagnosed, high-dose amoxicillin is recommended (Table 34-4). If this regimen fails, highly penicillinresistant pneumococci or â-lactamase–producing Haemophilus or Moraxella may be responsible; amoxicillin may be given at the same total dosage but with one-half of the dose in the form of amoxicillin/clavulanic acid. If this regimen fails, three doses of ceftriaxone at daily intervals are likely to be curative.


A quinolone or ketolide may also be tried in adults. Patients must be monitored closely for a response. An otolaryngology consultation is recommended if all these treatments fail. Despite the detection (by molecular analysis) of pneumococcal DNA in middle-ear fluid, chronic serous otitis (“glue ear”) is probably not due to active infection and does not require antibiotic therapy. Treatment for otitis is recommended for a total of 10 days in children <2 years of age but for only 5 days in children ≥2 years old who do not have complicated infections. A recent study reported identical rates of clinical and bacteriologic cure with a 10-day course of amoxicillin and a single dose of azithromycin (30 mg/kg). Acute Sinusitis (Table 34-4) Just as the pathogenesis and microbial etiology of acute rhinosinusitis are similar to those of otitis media, so are the principles of diagnosis and treatment. The diagnosis is often empirical, and the less rigorously it is made, the more irrelevant antibiotics are likely to be.


The estimated efficacy rate for amoxicillin/clavulanic acid, fluoroquinolones, and ceftriaxone (available for parenteral use only) is 90–92%, as opposed to 83–88% for amoxicillin, trimethoprimsulfamethoxazole, and oral second- or third-generation cephalosporins and 71–81% for macrolides and doxycycline. Treatment should be given for longer periods than are recommended for otitis media (perhaps 10–14 days), but the optimal duration is uncertain. Pneumonia (Table 34-5) This section will deal primarily with the treatment of pneumococcal pneumonia.The broader issue of empirical therapy for communityacquired pneumonia is covered elsewhere (Chap. 17). Unless epidemiologic, clinical, and radiologic findings strongly favor another etiology, empirical therapy for pneumonia must include an agent that will be effective against S. pneumoniae, which remains the most likely causative agent of community-acquired pneumonia. Outpatient Therapy Amoxicillin (1 g three times daily) effectively treats virtually all cases of pneumococcal pneumonia. Neither cefuroxime nor cefpodoxime offers any advantages over amoxicillin, and they are far more expensive.


Telithromycin is likely to be equally effective.Moxifloxacin is also highly likely to be effective in the United States except in patients who come from a closed population where these drugs are used widely or who have themselves been treated recently with a quinolone. Clindamycin is effective in 90% of cases and doxycycline, azithromycin, or clarithromycin in 80%. Treatment failure resulting in bacteremic disease due to macrolide-resistant isolates has been amply documented in patients treated empirically with azithromycin. As noted above, rates of resistance to all these antibiotics are lower in some countries and much higher in others; high-dose amoxicillin remains the best option worldwide. Inpatient Therapy Pneumococcal pneumonia is readily treatable with â-lactam antibiotics. The conventional dosages shown in Table 34-5 are acceptable against intermediately resistant strains and against many or most fully resistant isolates. Recommended agents include ceftriaxone and cefotaxime. Ampicillin is also widely used, usually in the form of ampicillin/sulbactam. The likely efficacy of newer quinolones such as moxifloxacin, macrolides such as azithromycin, and clindamycin is discussed above. On the basis of in vitro considerations, vancomycin is likely to be uniformly effective against pneumococci; this drug or a quinolone should be used together with a third-generation cephalosporin for initial therapy in a patient who is likely to be infected with a highly antibiotic-resistant strain.


Patients who have had a severe allergic reaction to penicillins or cephalosporins may be treated with a carbapenem (e.g., imipenemcilastatin), a quinolone, or vancomycin. The failure of a patient to respond promptly should at least prompt consideration of drug resistance. Evidence for loculated infections (such as empyema) and/or other causes of fever should be sought and addressed appropriately. Duration of Therapy The optimal duration of treatment for pneumococcal pneumonia is uncertain. Pneumococci begin to disappear from the sputum within several hours after the first dose of an effective antibiotic, and a single dose of procaine penicillin, which produces an effective antimicrobial level for 24 h, was curative in otherwise-healthy young adults in an era when all isolates were susceptible. Early in the antibiotic era, most physicians treated pneumococcal pneumonia for 5–7 days. In the absence of data suggesting a need for longer treatment, younger physicians tend to treat the infection for 10–14 days.


In the opinion of this author, a few days of close observation and parenteral therapy followed by an oral antibiotic—with the entire course of treatment continuing for no more than 5 days after the patient becomes afebrile—may be the best approach for treating pneumococcal pneumonia, even in the presence of bacteremia. Cases with a second focus of infection (e.g., empyema or septic arthritis) require longer therapy. Meningitis (Table 34-6) Pneumococcal meningitis should be treated initially with ceftriaxone plus vancomycin. Equivalent doses of cefotaxime or cefepime may be used in place of ceftriaxone. The cephalosporin will be effective against most—but not all—isolates and will readily penetrate the blood-brain barrier; all isolates will be susceptible to vancomycin, but this drug has a somewhat unpredictable capacity to cross the blood-brain barrier. If the isolate is shown to be susceptible or intermediately resistant, treatment can be continued with ceftriaxone, and vancomycin can be discontinued. If the organism is resistant, treatment with both drugs should be continued. A very few studies of experimental animals suggest benefits of the addition of rifampin, but in vitro studies indicate antagonism between this drug and ceftriaxone or vancomycin; in the absence of data to support the practice in humans, this author does not recommend that rifampin be added.


Imipenem may be used in place of the cephalosporin in patients who have had life-threatening allergic reactions to â-lactam antibiotics. The total duration of therapy for pneumococcal meningitis is 10 days. A recent study demonstrated clear benefit from the addition of glucocorticoids (Chap. 29). Endocarditis Pneumococcal endocarditis is associated with rapid destruction of heart valves. Pending results of susceptibility studies, treatment should be initiated with ceftriaxone or cefotaxime; if the prevalence of highly resistant strains increases, it might be prudent to add vancomycin until results of susceptibility studies are available. In vitro, aminoglycosides are somewhat synergistic and rifampin or quinolones are antagonistic with â-lactams against pneumococci; there is no clear evidence from in vivo studies that adding any of these antibiotics to the regimen is beneficial.


OTHER THERAPEUTIC MODALITIES

Addition of drotrecogin, an activated protein C preparation, may be beneficial in treating patients with severe pneumococcal sepsis. Glucocorticoids and agents that block the action of TNF-á, IL-1, or platelet-activating factor have conferred no benefit. PREVENTION Capsular Polysaccharide Vaccine The pneumococcal capsular polysaccharide vaccine administered to adults since the early 1980s contains 25 ìg per dose of capsular polysaccharide from each of the 23 most prevalent serotypes of S. pneumoniae.Vaccination stimulates antibody to most serotypes in most recipients. One case-control study showed a protection rate of 85% lasting ≥5 years in adults <55 years old (Table 34-7). The level and duration of protection decreased with advancing age. Other studies have suggested an overall protection rate in the adult population of 50–70%. In high-risk subgroups (e.g., debilitated elderly persons and individuals with severe chronic lung disease), vaccine has not been shown conclusively to be effective. Persons who most need the vaccine because of poor IgG responses (e.g., those with lymphoma or AIDS) are likely not to respond at all. Nevertheless, the Advisory Committee on Immunization Practices of the Centers for Disease Control and Prevention has broadened its recommendations for pneumococcal vaccination to include all persons >2 years of age who are at substantially increased risk of developing pneumococcal infection and/or having a serious complication of such an infection. Perhaps most in need of vaccination are persons with anatomic or functional asplenia, who are at risk for overwhelming, life-threatening infections.


Others who might fall within these recommendations are persons who (1) are over the age of 65; (2) have a CSF leak, diabetes mellitus, alcoholism, cirrhosis, chronic renal insufficiency, chronic pulmonary disease, or advanced cardiovascular disease; (3) have an immunocompromising condition associated with increased risk of pneumococcal disease (e.g.,multiple myeloma, lymphoma, Hodgkin’s disease, HIV infection, organ transplantation, or chronic glucocorticoid use); (4) are genetically at increased risk (e.g., Native Americans and Native Alaskans); or (5) live in environments where outbreaks are particularly likely to occur (e.g., nursing homes). Recommendations regarding revaccination seem somewhat inconsistent. A single revaccination is advocated for persons over the age of 65 if >5 years have elapsed since the first vaccination. Since antibody levels decline and there is no anamnestic response, it seems more reasonable simply to recommend revaccination at 5-year intervals, especially in persons over the age of 65, who tend to have almost no adverse reaction to vaccination, and in splenectomized patients, who are most in need. Protein-Conjugate Pneumococcal Vaccine Pneumococcal polysaccharide vaccine is not useful in children <2 years of age, whose immune system does not respond well to polysaccharide antigens.


Conjugating the polysaccharide to a protein yields an immunogen that is effective in infants and young children. Initial studies of a protein-conjugate pneumococcal vaccine consisting of capsular material from the seven serotypes most likely to cause disease in children (Prevnar) showed a 98% reduction in rates of bacteremia and meningitis and a 67% reduction in rates of otitis media due to vaccine serotypes. Since it was marketed in 2000, widespread use of this vaccine has caused a dramatic decline in the incidence of invasive pneumococcal disease among infants and children (Fig. 34-4). Colonization rates have also greatly decreased. In an Alaskan village, rates of carriage of vaccine strains decreased in children from 55 to 10% and in adults from 15 to 5%. Studies of protein conjugate vaccines that contain antigen from more than seven common infecting serotypes are nearing completion, with favorable results. The incidence of invasive pneumococcal disease has also declined among unvaccinated children and among adults, to whom this vaccine is not even offered (Fig. 34-4).


This decrease illustrates the “herd effect”—i.e., the impact of widespread vaccination on unvaccinated members of the population—and is probably attributable to the effects of the conjugate vaccine on nasopharyngeal carriage of vaccine serotypes. Another effect of the widespread use of this vaccine is the decreasing proportion of all pneumococcal disease that is due to antibiotic-resistant isolates, a trend that reflects the targeting of antibiotic-resistant strains by the vaccine.An unwanted effect of vaccination has been an increase in infections caused by serotypes that are not included in the vaccine (replacement serotypes), which, in fact, are increasingly expressing antibiotic resistance. Still, as noted above, the overall incidence of pneumococcal disease in all segments of the population has steadily declined. For further information, the reader is referred to the American Academy of Pediatrics Red Book Online (http://aapredbook.aappublications.org).  

affiliate_link