Isolates in Fulminant Pneumococcal Infections

Transcription

1 INFECllON AND IMMUNITY, Jan. 7, p. -6 Vol., No. Copyright 7 American Society for Microbiology Printed in U.S.A. Host Factors and Capsular Typing of Body Fluid Isolates in Fulminant Pneumococcal Infections RONALD M. SHAPERA AND JOHN M. MATSEN Departments of Laboratory Medicine and Pediatrics, University of Minnesota School of Medicine, Minneapolis, Minnesota Received for publication September 7 A 0-year retrospective study was made to determine the spectrum of underlying disease in, and the capsular types of pneumococci isolated from, blood and body fluids of 0 hospitalized patients. Fifteen serotypes were found among 8 typed isolates, % of which were types 8,, or. There was a predominance of males and a high incidence of underlying pneumonia, chronic liver disease, nephrotic syndrome, cerebrospinal fluid rhinorrhea, and malignancy. Mortality rates were higher among the very young and aged, and the immunologically comprised host at either extreme of life was more susceptible to disseminated pneumococcal infection. Pneumococci of the higher capsular types were prominent in children, and those of the lower types in adults, a finding in agreement with other recent studies. Overwhelming systemic infections due to Diplococcus pnewnoniae are frequent despite this organism's continued susceptibility to penicillin, and thus pneumococcus remains an important cause of morbidity and mortality. Renewed interest in the epidemiology, pathogenesis, and prophylaxis of pneumococcal infections has prompted us to review the past 0 years of experience with serious infections caused by this agent at the University of Minnesota Hospitals. The purpose of this study was to (i) determine the types of pneumococci isolated from blood and body fluids of hospitalized patients from 6 to 70, for purposes of comparison with similar data from studies done in different localities and in different years; (ii) ascertain if the capsular types of pneumococci isolated from blood and body fluids are similar to types reported to have been isolated from the sputum of patients with pneumonia and from the upper respiratory tract of asymptomatic carriers; (iii) confirm the role of host factors such as age and underlying disease in the pathogenesis of disseminated pneumococcal infection. MATERIALS AND METHODS The patients chosen for study were those hospitalized at the University of Minnesota Hospitals from January 6 to December 70 from whom pneumococci were isolated from blood, spinal, pleural, or peritoneal fluids. Pneumococci were identified by colonial morphology, hemolysis, Gram stain, optochin sensitivity, and the Quellung (capsular swelling) reaction by using antisera (Lederle and Difco) to types to. Isolates from the respiratory tract were not typed. Patients with pneumococci obtained at autopsy or from mixed infections were excluded from analysis. Charts of patients from whom pneumococci were isolated from body fluids were reviewed to obtain the following data: age and sex incidence, presence of pneumonia with pneumococci in the sputum preceding or during dissemination, mortality rate in relation to age and to specific pneumococcal types, distribution of pneumococci according to age, predilection of certain types to localize in certain body cavities, and the spectrum of underlying disease in this hospitalized population. An attempt was also made to assess the role of pneumococcal infection as a cause of death, by correlating the clinical picture with X ray, autopsy, and bacteriological culture reports. RESULTS During this retrospective study covering a span of 0 years, 0 patients had pneumococci isolated from their blood or body fluids. Some patients harbored pneumococci in several body sites, or possibly on several occasions from a single site, so that a total of 68 isolates was recovered, 8 of which were typed (Table ). The remaining isolates included 6 which were not typed and which were untypable by our battery of types to antisera. All 8 typed isolates fell among types, none higher than type (Table ); of the 68 isolates were from blood (Table ). Of the 6 typed-blood isolates, (7%) were types 8,, or. No predominant types were noted among cere- Downloaded from on September, 08 by guest

2 VOL., 7 FULMINANT PNEUMOCOCCAL INFECTIONS TABLE. Sources ofppneumococci isolated from body fluids at the University of Minnesota Hospitals Source Total isolates No. typed Blood... 6 CSFa... Pleural... 7 Peritoneal... Total TABLE. Typintg of pneumococci isolated at the Uniiversity of Minnesota Hospitals from blood and other body fluids (6-70) Type Total Blood 0 6 csfa Source Pleural Peritoneal brospinal fluid (CSF), peritoneal, and pleural isolates typed (Table ). The distribution of capsular types isolated from blood according to age is illustrated in Table. There was a predominance of lower types, especially and 8 from adults, and of higher types including, 8, and from children. Eighty-four (60%) of the patients were male and 6 (0%/) were female. Pneumonia was present in77 (%) and was the most frequent focus of infection prior to dissemination. Fifty of the 0 patients died, most likely related to their pneumococcal infection (Table ). All the pneumococci from patients who died were typed, and all deaths occurred in patients who were infected with the 8 most common of the types found in this study (Table ). Notably, 0 of the deaths were attributable to just three types (, 8, and ), and, in 0 of these, pneumonia was present and had probably been the primary focus of infection (Table ). Eighty per cent of all the patients and % of those who died were under age 0 or over age 0 (Table ). Ascites associated with chronic liver disease or the nephrotic syndrome was present in of 0 patients with pneumococcal peritonitis (Table 6). CSF rhinorrhea was present in 0 of patients with meningitis (Table 7). One hundred and nine patients had serious TABLE. Distribution of pnteumococcal types isolated from blood according to age at the University of Minnesota Hospitals (6-70) Type No. under age No. age and older ~~~~~~~~~~ Total TABLE. Relationship of age antd mortality in patients with pneumococcal body-fluid isolates at the University of Minnesota Hospitals (6-70) Age patients who patients duet e died mococcus No. Total Death likely Less than month month to years to 0 years.. 6 to years. 0 6 to 0 years. 8. to 6 years.. 7 Over 6 years. Totala a Sixty-five (6.%) were less than 0 years; 7 (.%), greater than 0 years; 8 (0.0%), between and 0 years. Fifty-six of 0 were females; 8 were males. Downloaded from on September, 08 by guest

3 SHAPERA AND MATSEN INFECT. IMMUNITY TABLE. Deaths probably related to differeni pneumococcal types at the University of Minniesota Hospitals (6-70) deaths patients Type directly related to dying who had pneumococcus pneumonia Total 0 TABLE 6. Anialysis of cases of pneuimococcal peritonitis at the Unziversity of Minlinesota Hospitals (6-70) Pneumococcal U d A type Underlying disease Age Biliary atresia 8 Cirrhosis of liver Biliary atresia Nephrotic syndrome Nephrotic syndrome Biliary atresia Autoimmune hepatitis Alcoholic cirrhosis 6 8 Chronic renal failure Nephrotic syndrome 8 underlying systemic disease, notably leukemia, reticuloendothelial or solid malignancies, autoimmune disorders, splenic dysfunction, uremia, and diabetes mellitus (Table 8). DISCUSSION Our data focus on the continued prevalence of severe pneumococcal infection in the antibiotic era. It has been suggested that the incidence of such disease will not further decline without a change in approach to prophylaxis or therapy (). Previous clinical trials with polyvalent pneumococcal vaccines have indicated their protective efficacy and ability to induce longlasting immunity (). Austrian () has advocated their use in certain high-risk groups. He concluded that an appropriate vaccine should be protective against the small number of pneumococcal types that he () and others (6) have found to be frequently associated with bacteremia and high mortality rates. In his experience, bacteremia and death, particularly with types and 8, was more common in elderly males and in those with underlying systemic diseases. The overall incidence of bacteremia accompanying pneumococcal pneumonia in adults was found to be as high as % (). In a multiinstitutional study, 60% of the capsular types causing bacteremia were types to 8, and 0 capsular types were responsible for 7 % of,008 bacteremic infections. These same capsular types were also responsible for 7%, of the deaths in a series of 0 such infections (). In children, the incidence of bacteremia accompanying pneumococcal pneumonia is TABLE 7. Anialysis of cases of pneumococcal menin7gitis at the Unziversity of Minnesota Hospitals (6-70) Underlyinig disease Otitis media. CSF- rhinorrhea... Anemia... Biliary atresia... No disease.... Total... patients isolates typed Capsular types 0 0 Of type Of type 7 Of type Of type Of type 8 Of type TABLE 8. Antalysis of disease processes associated wtith pnteumococcal body fluid isolates Disease processes Cerebrospinal fluid leak.... Collagen vascular diseases... Chronic liver disease... Nephrotic syndrome.... Hypersplenism or previous splenectomy Reticuloendothelial or hematopoetic malignancy Chronic pulmonary disease... Chronic renal failure Solid organ malignancy Otitis media and mastoiditis... Diabetes mellitus No underlying disease... patients 0 8 Downloaded from on September, 08 by guest

4 VOL., 7 FULMINANT PNEUMOCOCCAL INFECTIONS said to be less than %c (). Recently, a rather obscure variety of pneumococcal bacteremia of the so called "cryptogenic" type has been reported to occur in children in whom no obvious focus of infection could be found (,, 8). In many, bacteremia was thought to have originated from a mild focus of infection in the respiratory tract or middle ear which had escaped notice. When typing was done, pneumococci above type 8, but still among the lower types, were most often responsible (). Some of these patients seemed to recover spontaneously before the results of cultures were known and treatment could be initiated. In our study, 6 of the deaths probably attributable to pneumococcal infection involved types higher than type 8, notably types, 8, and. Burke et al. also noted a high incidence of pneumococci above type 8 in a recent study of pneumococcal bacteremia in children (). Thirty per cent of their isolates were type and only 8 of the isolates were among types to 8. Type was also a major cause of pneumonia in children in the pre-antibiotic era (0). The higher incidence of and mortality rate from types to in some recent studies (, 6), as compared to our experience, may be largely attributable to the age of the patients included for analysis. More than half of our typed isolates came from patients under years of age, whereas all of Austrian's () patients were over that age. That age probably was significant in this regard is strikingly illustrated in Table, showing the paucity of lower types from children under age and the infrequency of higher types among adults, in agreement with previous studies. An additional factor of importance in the comparison of the distribution of types may have been the incidence and nature of underlying diseases in the different studies. Eighty-one per cent of our patients had complicating systemic diseases in contrast to %r in Austrian's study (). In addition, considerably more of our patients had malignancies being treated with drugs which inhibit host immune responses and which might have made them more susceptible to some of the less virulent higher capsular types. Lund () noted differences in pneumococcal types isolated from different body sites, indicating a possible propensity for certain capsular types to infect certain organs. We could not confirm this possibility among our smaller number of typed isolates. The possible predominance of certain pneumococcal types in different geographic locations could obviate the widespread use of a single polyvalent pneumococcal vaccine. Fortunately, this does not seem to be the case on comparing our data with that of others. Our blood isolates (7%) were types 8,, or and these were among the most common types in large studies in other locales (, ). A rapid shift over a short period of time in the types of pneumococci isolated from clinical specimens does not seem likely, in view of the relative constancy of the predominant types noted from the pre-antibiotic era () until recently (, 6). Apparently there is a similarity between the types isolated from throats of nonhospitalized carriers (7), from the sputum of patients with pneumococcal pneumonia (, 6), and from the blood and body fluids of patients with disseminated infection (, ). Calder et al. (6) found types and 8 to comprise 6 and %, respectively, of pneumococcal isolates from the sputum of patients with pneumonia. Lund () found these same types to be among the commonest isolated from blood. Feldman (7) showed among a group of asymptomatic pneumococcal carriers that types and 8 comprised.% of isolates from the throat. Because of this similarity among the capsular types which are generally carried asymptomatically and those which are isolated from clinical infections, a vaccine aimed at preventing colonization by the types most often found in the respiratory flora might be expected to lower the incidence of systemic infection caused by these same types. The increased susceptibility to infection of patients with autoimmune disorders, diabetes mellitus, and immune deficiency states is well known. Absence of normal splenic function due to splenectomy, congenital absence, hypoplasia, or massive infiltration of the spleen has been noted to increase the risk of overwhelming pneumococcal infection (8). In sickle cell anemia, when splenic infarction leads to functional autosplenectomy, a deficiency of pneumococcal opsonic activity results in a greater susceptibility to pneumococcal septicemia and meningitis (6). Recurrent meningitis in the presence of occult or obvious CSF rhinorrhea is also well documented. The frequency of pneumococcal peritonitis complicating the nephrotic syndrome () and chronic liver disease (7, ) is incompletely understood. Pneumococci which are commonly present in the respiratory tract may cause transient bacteremia, and the presence of ascitic fluid in these two conditions may provide an optimum environment for growth of organisms which seed the peritoneal cavity. Our study is not strictly comparable to those in which the patients were chosen on thebasis of a pneumococcal respiratory infection. Our Downloaded from on September, 08 by guest

5 6 A - _v SHAPERA ) x. - -' MATSEN INFECT. IMMUNITY selection of patients on the basis of positive body fluid cultures tends to include many with severe underlying disease, thus tending to obscure the exact contribution of pneumococcal infection as a cause of death. However, all of our patients showed clinical deterioration at the onset of pneumococcal infection, and postmortem data tended to confirm the role of such infection in the ultimate demise of. Thus, several points seem to emerge from our analysis. Life-threatening pneumococcal infections tend to occur in infancy, in the elderly, and in those with underlying systemic diseases affecting the immune system. The high incidence of deaths from higher pneumococcal types in our study seems to be related to the inclusion of a large number of children, who tend to become infected with these types more often than adults. In the latter, morbidity and mortality are most often due to lower types. Deaths among both adults and children seem attributable to the higher pneumococcal types in patients not in the agonal state at the time of acquisition of these organisms, indicating a probable etiological role for pneumococci in their demise. These factors indicate that an appropriate vaccine should contain multiple polysaccharide antigens including some of the higher types, if it is to be widely applicable. Such a vaccine has recently been undergoing field trials and may offer the best means of controlling the persistently high mortality rates from pneumococcal infections. LITERATURE CITED. Austrian, R. 68. Current status of bacterial pneumonia with especial reference to pneumococcal infection. J. Clin. Pathol. (Suppl ) :-7.. Austrian, R., and J. Gold. 6. Pneumococcal bacteremia kl with especial reference to bacteremic pneumococcal pneumonia. Ann. Int. Med. 60: Belsey, M. A. 67. Pneumococcal bacteremia. Amer. J. Dis. Child. : Bullowa, J. G. M. 7. The management of the pneumonias. Oxford University Press, New York. S. Burke, J. P., J. 0. Klein, H. M. Gezon, and M. Finland, 7. Pneumococcal bacteremia. Amer. J. Dis. Child. :-. 6. Calder, M. A., V. M. Valentine, and M. E. Schonell. 70. Importance of pneumococcal typing in pneumonia. Lancet : Epstein, M., F. M. Calia, and G. J. Gabuzda. 68. Pneumococcal peritonitis in patients with postnecrotic cirrhosis. N. Engl. J. Med. 78: Eraklis, A. J., S. V. Kevy, L. K. Diamond, and R. E. Gross. 67. Hazards of overwhelming infection after splenectomy in childhood. N. Engl. J. Med. 76:-.. Finland, M., and R. C. Tilghman. 7. Clinical significance of bacteremia in pneumococcal pneumonia. Arch. Intern. Med. : Heffron, R.. Pneumonia with special reference to pneumococcus lobar pneumonia. Commonwealth Fund, New York.. Heldich, F. J. 70. Diplococcus pneumoniae bacteremia. Amer. J. Dis. Child. :-7.. Kerr, D. N. S., D. T. Pearson, and A. E. Read. 6. Infection of ascitic fluid in patients with hepatic cirrhosis. Gut :-8.. Lund, E. 70. Types of pneumococci found in blood, spinal fluid and pleural exudate during a period of years (- 6). Acta Pathol. Microbiol. Scand. 78:-6.. MacLeod, C. M., R. G. Hodges, M. Heidelberger, and W. G. Bernard.. Prevenition of pneumococcal pneumonia by immunization with specific capsular polysaccharides. J. Exp. Med. 8:-6.. Pahmer, M. 0. Pneumococcus peritonitis in nephrotic and non-nephrotic children. J. Pediat. 7: Pearson, H. A., R. P. Spencer, and E. A. Cornelius. 6 Functional asplenia in sickle-cell anemia. N. Engl. J. Med. 8: Suhs, R. H., and H. A. Feldman. 6. Pneumococcal types detected in throat cultures from a population of "normal families". Amer. J. Med. Sci. 0: Torphy, D. E., and C. G. Ray. 70. Occult pneumococcal bacteremia. Amer. J. Dis. Child. :6-8. Downloaded from on September, 08 by guest