Introduction

Disease due to Streptococcus pneumoniae (pneumococcus), such as bacteremia, pneumonia and meningitis, is a major cause of morbidity and mortality in the general population and may pose a greater risk to immunosuppressed persons. Previous studies have reported a significantly high incidence of invasive pneumococcal disease (IPD) in hematopoietic stem cell transplant (HSCT) recipients. A European multicenter survey estimated the incidence to be 5.97 per 1000 transplantations.¹ Incidence is also significantly increased in the presence of GVHD and allogeneic transplants appear to be at greater risk than autologous SCT.^{2, 3} The incidence of late infections (as defined as 100 days post-SCT) also appears to be greater than early invasive pneumococcal infections. The American Committee on Immunization Practices and Centers for Disease Control recommend two doses of pneumococcal polysaccharide vaccine at 12 and 24 months post transplant for adult HSCT recipients.^{4, 5} In addition, the European Bone Marrow Transplant group has recommended a single dose of polysaccharide vaccine at 12 months post transplant and suggested following pneumococcal serology to determine protection if chronic GVHD exists.⁶ Although data on incidence of infections exist, there have been no studies that quantify the risk as compared to the general population.

The Toronto Invasive Bacterial Diseases Network (TIBDN) is a prospective study of bacterial infectious diseases epidemiology serving a large geographic area, which includes prospective data on bacterial infections occurring in all HSCT patients living in the geographic area captured by the TIBDN. The primary objective of this study was to prospectively determine the incidence of IPD in a population of adult HSCT recipients over a period of 10 years. We evaluated the clinical impact, serotypes and antimicrobial resistance patterns of IPD and compared this to the general population in the same geographic area.

Methods

Population-based surveillance

Pneumococcal isolates were collected from 1 January 1995 to 31 December 2004 through the TIBDN. The TIBDN has conducted prospective, population-based surveillance of IPD in metropolitan Toronto and the adjacent regional municipality of Peel, in Ontario, Canada (population, 5 million), since 1 January 1995. The surveillance network includes all microbiology laboratories in hospitals to which residents of the population area may be admitted and consists of 25 licensed microbiology laboratories that serve 27 hospitals, long-term care facilities and outpatient offices. Personnel from these laboratories telephone the central TIBDN study office at the Mount Sinai Hospital in Toronto whenever S. pneumoniae is isolated from a sterile site specimen. For each case, initial demographic data and the pneumococcal isolate are forwarded to the central TIBDN study office. Additional clinical data, including patient comorbidities, clinical course and outcome, antimicrobial therapy in the 3 months before presentation and outpatient therapy for the current episode before the blood sample was obtained for culture, are acquired by chart review, patient interview and by contacting the patient's attending physicians. Annual audits are conducted in each laboratory to ensure complete reporting. Surveillance and associated studies are approved by the research ethics boards of all participating institutions. This surveillance network encompasses the geographic catchment area for the transplant patient population that was included in this study.

Transplant population

Adult HSCT recipients (age 18) living in the catchment area for TIBDN are followed at the Princess Margaret Hospital, Toronto, Canada. Trimethoprim-sulfamethoxazole (TMP/SMX) prophylaxis for Pneumocystis jiroveci (one double-strength tablet three times per week) was used in all patients unless a specific contraindication such as allergy was present. Pneumococcal polysaccharide vaccine is routinely recommended to all transplant recipients at 12 and 24 months post transplant. A comprehensive database of HSCT recipients that includes information such as demographics, dates of transplant, allograft failure and death is maintained and updated every 3 months. This database was used for calculation of disease incidence and analysis of variables contributing to IPD.

Definitions

IPD was defined as isolation of S. pneumoniae from a sterile body fluid with a compatible clinical syndrome. Sterile sites included blood, cerebrospinal fluid, peritoneal fluid, pleural fluid or needle aspiration of a collection. Sputum or bronchoalveolar lavage (BAL) isolates are not considered sterile site isolates and therefore were not included in the estimate of confirmed IPD.⁷ Nonbacteremic pneumococcal pneumonia was defined as: (1) a clinical presentation including symptoms (cough, sputum and fever) and physical findings consistent with pneumonia; (2) an infiltrate on chest radiograph; (3) microscopic examination of a Gram-stained sputum showing 20 WBC per high-power field and a predominance of Gram-positive cocci in pairs or chains and a sputum culture that grows only S. pneumoniae and (4) no positive blood cultures for S. pneumoniae.⁸ All patients with a pneumococcal infection were followed for 30 days from the onset of illness and the following variables were assessed: clinical symptoms, length of hospitalization, 7 day and 30 day mortality. Early or late IPD was defined as that occurring at <100 days or 100 days post transplant, respectively. All isolates were serotyped by the quellung reaction at the National Center for Streptococcus, Edmonton, Canada using commercial anti-sera (Statens Seruminstitut, Copenhagen, Denmark; http://www.provlab.ab.ca/ncs/ncs.htm) and underwent antimicrobial susceptibility testing by broth microdilution to CLSI (Clinical and Laboratory Standards Institute) standard.⁹ Susceptible and nonsusceptible (intermediate or resistant) isolates were defined as per the breakpoints set by CLSI methods.¹⁰

Statistical methods

All patients were included in the analysis. Incidence, antimicrobial susceptibility rates and serotypes were compared between the HSCT and general TIBDN population by using a ²-test. Factors contributing to IPD in allogeneic HSCT were compared between patients with IPD and those without by Fisher's exact test. All statistical analysis was done using SPSS version 13.0 (SPSS Inc., Chicago, IL, USA).

Results

Incidence

There were 1238 adult HSCT recipients residing in the TIBDN surveillance area during the study period including both allogeneic (n=431) and autologous (n=807) transplant recipients. Taking into account the time of transplant, graft-failures and deaths, this translated into 4037 total person years of follow-up in the transplant population (that is, total at-risk population years). Median follow-up period for the transplant population was 847 days. There were 14 cases of identified invasive pneumococcal infection (crude incidence of 14 per 4036 person years). Based on the number of at-risk individuals, the incidence was calculated as 347 per 100 000 transplanted patients per year or 11.3 per 1000 transplants. Specific incidences for allogeneic and autologous HSCT were 590 and 199 per 100 000 transplanted patients per year, respectively (P=0.076).

Incidence was compared to the adult population followed by TIBDN during the same time period. During the 10-year period of active surveillance, 4073 episodes of IPD were detected, for an overall incidence of 11.5 cases per 100 000 population per year. No clear trends were detected in the incidence over time. The annual incidence in the general population did not vary significantly over the study period. Compared to the general population, the HSCT population had a 30.2-fold (95% confidence interval (CI) 17.8–50.8; P<0.00001) greater incidence of IPD.

Clinical characteristics

The baseline characteristics of the 14 cases of IPD in adult HSCT recipients are shown in Table 1. There were nine allogeneic and five autologous transplant patients with IPD over the 10-year period. Median time to presentation was 2.28 years post transplant (range 0.02–7.47 years). The majority of subjects (12 of 14; 85.7% ) had late onset IPD. In those that had undergone allogeneic transplant, chronic GVHD, prior receipt of total body irradiation, and transplant subtype (peripheral blood vs bone marrow) were analyzed as potential risk factors for development of IPD. These factors were compared to all allogeneic transplant patients residing in the TIBDN area during the study period. The presence of chronic GVHD was the only significant factor in those that developed IPD compared to the remainder of the allogeneic transplant population (77.8 vs 41.3% , respectively, P=0.045). Total body irradiation irrespective of dose (66.7 vs 72.6% , P=0.70) and transplant subtype (22.2 vs 32.1% , P=0.72) was not a significant factor.

Table 1 - Baseline characteristics of HSCT patients with pneumococcal disease from 1995 to 2004.

Full table

Clinical presentations and outcomes are shown in Table 2. All patients had bacteremia and the majority also had pneumonia (n=10). Two patients (2 of 14; 14.3% ) died during hospitalization; both deaths were attributable to IPD. In patients of the same age range (18–65 years of age), the 30-day mortality was 10.3% (P=0.64 compared to mortality in transplant patients) and the in-hospital mortality was 12.2% .

Table 2 - Clinical presentations of pneumococcal disease in HSCT patients.

Full table

In HSCT patients, the rate of vaccination with the 23-valent pneumococcal polysaccharide vaccine prior to the onset of IPD was 21.4% (3 of 14). Of the three patients who had previously ever been vaccinated, all had been vaccinated within the past 5 years. The remaining 11 of 14 (78.6% ) patients had never been vaccinated although 4 of these patients did receive vaccine once they recovered from the IPD episode. IPD occurred at a mean of 10.59.4 months (range 0.5–19.1 months) post-vaccination. Serotyping was available for 13 of 14 subjects. The most frequent serotypes were 23F (3 of 13, 23.1% ), 6B (3 of 13, 23.1% ) and 9V (2 of 13, 15.4% ). The remainder of the patients had one of the following serotypes: 4, 7F, 18C, 19A and 22F. Transplant patients (9 of 13, 69.2% ) had disease due to a serotype that is included in the 7-valent pneumococcal conjugate vaccine and 13 of 13 (100% ) patients had disease due to serotypes covered by the 23-valent pneumococcal polysaccharide vaccine. For IPD, the most frequent serotypes in the general population were 14 (20% ), 6B (8.9% ), 4 (7.8% ), 23F (7.4% ), 19F (7.2% ) and 3 (7.1% ).

Antimicrobial susceptibility

Pneumococcal isolates underwent antibiotic susceptibility testing against penicillin, erythromycin, levofloxacin and TMP/SMX. At the time of their infection 2 of 14 patients were receiving TMP/SMX prophylaxis. The pneumococcal isolate from one of these patients was resistant to TMP/SMX. The nonsusceptibility rates (tested intermediate or resistant in vitro) were available for 13 of 14 transplant patient isolates and were 7.7% for penicillin, 23.1% for erythromycin, 0% for levofloxacin and 46.2% for TMP/SMX. Resistance rates of all other sterile site pneumococcal isolates from the TIBDN population for the matched time period and age group are as follows: penicillin (12% , P>0.05), erythromycin (11% , P>0.05), levofloxacin (0.9% , P>0.05 and TMP/SMX (20% , P=0.03).

Nonsterile site isolates

From April 2002 to the end of the study period, nonsterile site isolates from HSCT recipients were also collected as part of the TIBDN surveillance program. There were four additional patients (two allogeneic and two autologous transplants) identified with nonsterile site isolates of S. pneumoniae, all from the respiratory tract (sputum (n=3) and BAL (n=1)). Although all four patients were treated clinically for pneumococcal pneumonia, only three of four met strict TIBDN criteria for nonbacteremic pneumococcal pneumonia. To estimate an upper limit of the incidence of microbiologically proven IPD in the HSCT population, data from sterile site and nonsterile site isolates were combined. The calculated incidence of IPD for all transplants was 446 per 100 000 person-years (721 per 100 000 for allogeneic vs 279 per 100 000 for autologous HSCT).

Discussion

Our study prospectively quantifies the risk over a period of 10 years and directly compares it to the non-HSCT population over the same time period and living in the same geographic area. In our primary analysis (sterile site isolates only), we estimated the incidence of IPD in the HSCT population to be 347 per 100 000 persons per year compared to the general population incidence of 11.5 per 100 000 persons per year. The incidence was highest in allogeneic transplant recipients (590 per 100 000 persons per year), but was higher than the general population for both autologous and allogeneic combined. The higher rate specifically observed in allogeneic transplant recipients may be due to multiple factors including post transplant immunosuppression, immunomodulatory effects of GVHD, impairment of splenic function, as well as the absence of prior vaccination or a suboptimal vaccine response. The incidence in adult HSCT patients is also greater than in solid organ transplant recipients residing in the same geographic region where the relative risk was 12.8 compared to the general population.¹¹

The majority of patients were diagnosed late in their post transplant course. This was also reported by Engelhard et al.¹ in a European survey. Isolation and reduced social interaction early post transplant may protect patients from acquiring virulent pneumococcal strains, while persistent deficits in humoral and cellular immunity later post transplant may increase their risk of illness on exposure to the same strains in the community. Two (14.3% ) patients died during the hospitalization for pneumococcal disease and both cases could be directly attributed to IPD. However, this was not significantly different than the in hospital mortality for the TIBDN general population of 19.5% . This is also similar to the published literature where mortality for bacteremic pneumococcal pneumonia is 14–17% .^{8, 12, 13} Antimicrobial resistance rates were significantly higher especially to TMP/SMX and this is likely due to the use of TMP/SMX prophylaxis. The routine strategy at our centers is to discontinue TMP/SMX prophylaxis at 1-year post transplant in the absence of GVHD. All of our allogeneic transplant patients were given penicillin prophylaxis after recovery from IPD regardless of the presence of GVHD. Trials for the use of penicillin in adult HSCT patients are lacking; however, this practice may be an extrapolation from the pediatric post-splenectomy literature where pencillin prophylaxis is recommended.¹⁴ Failure of pencillin with breakthrough penicillin-resistant IPD has been described.¹⁵ This is an area that requires further study.

Vaccination is another recommended strategy to prevent IPD.^{4, 5, 6} Only 21.4% of our patients with IPD had received the pneumococcal vaccine, and vaccination rates are not routinely recorded for our adult HSCT cohort. In a survey of vaccination done in US transplantation centers, 80% of responding centers administered pneumococcal vaccine post-allogeneic transplant.¹⁶ However, only 26% gave more than one dose and 63% gave the first dose within the first year of transplant. In addition, 11 different dosing schedules were reported in the centers surveyed. The low vaccination rate at our center is likely attributable to the lack of a comprehensive pneumococcal vaccination program for transplant patients in our area and our results suggest that all transplant programs should offer vaccination.

Limitations of our study include the fact that only patients who had a positive culture were captured by the TIBDN case-finding process. It is possible that cases of pneumococcal disease were missed because a culture was not done or was negative (for example, if cultures were performed after administration of antimicrobials). Therefore, this study provides a minimal estimate of the risk of IPD and the actual risk may be greater. We also did not have information available as to whether immunoglobulin levels or the presence of children in the household were risk factors in this population for development of IPD. This study only provides data for adults and did not include pediatric HSCT patients. Strengths of our study include the use of a large prospective multiyear database to capture detailed information on IPD both in the transplant and general population. This allowed accurate comparison of incidence, serotypes and antimicrobial resistance rates with the general population residing in the same geographic area.

In summary, we have estimated the incidence of pneumococcal disease in HSCT recipients to be significantly greater than in the general population. Disease contributes to morbidity and mortality. There are greater rates of resistance to TMP/SMX than in the general population and serotypes 23F and 6B appear to predominate. The development of more comprehensive and optimal vaccination strategies is needed to decrease the risk of invasive disease.

Notes

Disclosure/conflict of interest

No author has a conflict of interest relevant to this article.

References

1. Engelhard D, Cordonnier C, Shaw PJ, Parkalli T, Guenther C, Martino R et al. Early and late invasive pneumococcal infection following stem cell transplantation: a European bone marrow transplantation survey. Br J Hematol 2002; 117: 444–450. | Article |
2. Kulkarni S, Powles R, Treleaven J, Riley U, Singhal S, Horton C et al. Chronic graft versus host disease is associated with long-term risk for pneumococcal infections in recipients of bone marrow transplants. Blood 2000; 95: 3683–3686. | PubMed | ISI | ChemPort |
3. Youssef S, Rodriguez G, Rolston KV, Champlin RE, Raad II, Safdar A. Streptococcus pneumoniae infections in 47 hematopoietic stem cell transplantation recipients: clinical characteristics of infections and vaccine-breakthrough infections, 1989–2005. Medicine (Baltimore) 2007; 86: 69–77. | Article | PubMed |
4. Centers for Disease Control and Prevention. Prevention of pneumococcal disease: recommendations of the Advisory Committee on Immunization Practices (ACIP). MMWR 1997; 46: 1–24.
5. Centers for Disease Control and Prevention. Guidelines for preventing opportunistic infections among hematopoietic stem cell transplant. MMWR 2000; 49: 1–128.
6. Ljungman P, Engelhard D, de la Camara R, Einsele H, Locasciulli A, Martino R et al. Vaccination of stem cell transplant recipients: recommendations of the infectious diseases working party of the EBMT. Bone Marrow Transplant 2005; 35: 737–746. | Article | PubMed | ISI | ChemPort |
7. Gill VJ, Fedorko DP, Witebsky FG. The clinician and the microbiology laboratory In: Mandel GL, Bennett JE, Dolin R (eds). Principles and Practice of Infectious Diseases. Churchill Livingstone: Philadelphia, PA, USA, 2005, pp 203–241.
8. Musher DM, Alexandraki I, Graviss EA, Yanbeiy N, Eid A, Inderias LA et al. Bacteremic and nonbacteremic pneumococcal pneumonia: a prospective study. Medicine (Baltimore) 2000; 79: 210–221. | Article | PubMed | ChemPort |
9. National Committee for Clinical Laboratory Standards (NCCLS). Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically, 5th edn. Approved standard M7-A5 National Committee for Clinical Laboratory Standards. NCCLS: Wayne, PA, 2000.
10. National Committee for Clinical Laboratory Standards (NCCLS). Performance standards for antimicrobial testing. 13th informational supplement. NCCLS Document M100-S14 NCCLS: Wayne, PA, 2003.
11. Kumar D, Humar A, Plevneshi A, Green K, Prasad GV, Siegal D et al. Invasive pneumococcal disease in solid organ transplant recipients—10-year prospective population surveillance. Am J Trans 2007; 7: 1209–1214. | Article | ChemPort |
12. Metlay JP, Hofmann J, Cetron MS, Fine MJ, Farley MM, Whitney C et al. Impact of penicillin susceptibility on medical outcomes for adult patients with bacteremic pneumococcal pneumonia. Clin Infect Dis 2000; 30: 520–528. | Article | PubMed | ChemPort |
13. Yu VL, Chiou CC, Feldman C, Ortqvist A, Rello J, Morris AJ et al. An international prospective study of pneumococcal bacteremia: correlation with in vitro resistance, antibiotics administered, and clinical outcome. Clin Infect Dis 2003; 37: 230–237. | Article | PubMed | ChemPort |
14. Castagnola E, Fioredda F. Prevention of life-threatening infections due to encapsulated bacteria in children hyposplenia or asplenia: a brief review of current recommendations for practical purposes. Eur J Haematol 2003; 71: 319–326. | Article | PubMed |
15. Tauro S, Dobie D, Richardson G, Hastings M, Mahendra P. Recurrent penicillin-resistant pneumococcal sepsis after matched unrelated donor (MUD) transplantation for refractory T cell lymphoma. Bone Marrow Transplant 2000; 26: 1017–1019. | Article | PubMed | ISI | ChemPort |
16. Henning KJ, White MH, Sepkowitz KA, Armstrong D. A national survey of immunization practices following allogeneic bone marrow transplantation. JAMA 1997; 277: 1148–1151. | Article | PubMed | ISI | ChemPort |

Acknowledgements

Collaborating investigators in the Toronto Invasive Bacterial Disease Network are as follows: P Da Camara and J Downey, Toronto East General Hospital (Toronto, Canada); HR Devlin, St Michael's Hospital (Toronto); H Dick, Vita-Tech Laboratories (Toronto); IN Gaid and I Kitai, Rouge Valley Health System (Toronto); P Garrod and N Rau, Halton Healthcare Services (Oakville, Canada); R Lovinsky, D Noria, D Rose and J Braithwaite, The Scarborough Hospital (Toronto); F Jamieson, Ontario Public Health Laboratory (Toronto); R Grossman, Credit Valley Hospital (Mississauga, Canada); J Kapala, Gamma Dynacare Laboratories (Toronto); S Krajden, St Joseph's Health Centre (Toronto); KS Lee and M Baqi, Humber River Regional Hospital (Toronto); M Loeb and F Smaill, Hamilton Health Sciences Center (Hamilton, Canada); M Lovgren and G Tyrrell, National Centre for Streptococcus (Edmonton, Canada); AG Matlow, Hospital for Sick Children (Toronto); R McKeown, Peel Region Health Department (Brampton, Canada); B Mederski and K Katz, North York General Hospital (North York, Canada); Z Moloo, D Richardson and C Quan, William Osler Health Care Centre (Brampton); M Naus, British Columbia Centers for Disease Control (Vancouver, Canada); K Ostrowska and A Sarabia, Trillium Health Centre (Mississauga); P Shokry and I Ephtimios, Markham Stouffville Hospital (Markham, Canada); AE Simor and M Vearncombe, Sunnybrook and Women's College Health Science Centre (Toronto); D Sturman, Bridgepoint Hospital (Toronto); P Van Nostrand, The Rehabilitation Institute of Toronto (Toronto); S Walmsley, University Health Network (Toronto); D Low, B Willey and S Pong-Porter, Toronto Medical Labs/Mount Sinai Hospital (Toronto); B Yaffe, City of Toronto Public Health (Toronto); D Yamamura, MDS Laboratories (Toronto); M Silverman, Lakeridge Health (Oshawa, Canada); R Robertson, Royal Victoria Hospital (Barrie, Canada) and G Volkening, Southlake Regional Health Center (Newmarket, Canada).

Dr Kumar had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.