The combined DNA sequence set for P. jirovecii recovered from other contemporaneous PJP patients residing in the Sydney region demonstrated substantial genetic diversity and harbored dissimilar genotypes from the outbreak cluster. From the 20 P. jirovecii isolates analyzed, a single DHPS genotype, 2 β-TUB genotypes, 2 mtLSU, and 4 ITS1/2 genotypes, resulted in seven identified MLST sequence types (Table 3).
Within the case cluster, a case cohort study comprising the 14 patients and 324 unaffected control patients was undertaken to determine risk factors for PJP. Univariate analysis identified several risk factors (Table 2), including impaired graft function (P<0.001). CMV disease within the preceding 12 months occurred in five patients (35.7%, P<0.001 vs. control, with gastrointestinal cytopathic effect confirmed by immunoperoxidase in 4, clinical viremia and response to specific treatment in one). None of the five patients had received antiviral medication at PJP presentation and the interval between CMV and PJP was 20.3±16.0 months [range, 2–40 months]. Underlying pulmonary disease was present in four PJP patients (28.6% from previous H1N1 and later respiratory syncytial virus infections, nonspecific pneumonitis, Wegeners's granulomatosis, and bronchiectasis) compared with 2.6% of controls (P<0.001).
By multivariate analysis, significant risk factors for PJP infection were underlying pulmonary disease (odds ratio [OR], 10.1; 95% CI, 2.3–45.0; P=0.002), previous CMV infection (OR, 65.9; 95% CI, 7.9–550; P<0.001), and impaired estimated glomerular filtration rate (OR, 1.61 per 10 mL/min/1.73 m2; 95% CI, 1.15–2.25; P=0.006) compared with unaffected clinic patients.
Other transplant factors such as human leukocyte antigen (HLA) mismatching, biopsy-proven rejection, previous rejection, and recent (<12 months) pulse corticosteroids and current dose, calcineurin inhibitor type, dose or levels, antimetabolite dose or type, previous OKT3 or antithymocyte globulin therapy, and smoking status were not significantly different between groups. Initial TMP-SMZ prophylaxis rates of 93% of PJP patients for 6.3±9.5 months were comparable with controls. Patients without prophylaxis were transplanted before 1989 by other units. The number of clinic visits (a surrogate for exposure) was comparable between groups, but dominated by recent recipients seen frequently but still protected by routine prophylaxis.
Outcomes and Infection Control Measures
All cases presented with severe pneumonia and respiratory failure. Ten required ventilation in intensive care for 22.0±7.2 days, and two were managed with noninvasive respiratory support by high-dependency units. Empirical high-dose intravenous TMP-SMZ was commenced on admission in 13 cases (and delayed in two patients), accompanied by increased corticosteroids and supportive measures. Patients were isolated in single rooms until complete clinical response. Two patients (including the index case) died from high pulmonary compliance and failed mechanical ventilation. Six patients displayed severe preexisting transplant dysfunction (serum creatinine, 337±112 μmol/L) and renal failure followed in seven during illness from acute kidney injury. Six patients returned to dialysis and one recovered to baseline creatinine.
PCR testing on oropharyngeal rinses from 18 healthcare workers from the outpatient facility (and 16 nonclinical controls) yielded negative results for P. jirovecii DNA, as did sampling of ambient air from the consulting rooms, corridor, and waiting rooms of outpatient clinic using standard solid supports for trapping of potential organisms tested after outbreak control had commenced.
A policy of extended duration of PJP prophylaxis from 6 to 12 months after transplantation for all recent recipients and preemptive prophylaxis (TMP-SMZ, 800/160 mg daily or alternatives for allergy) was restarted in all potentially exposed asymptomatic prevalent clinic patients. Two late cases of PJP missed TMP-SMZ as their clinic exposure was not recognized, but presented with milder disease.
Although the exact mode of transmission of P. jirovecii infection is unclear, data from molecular genotyping studies of nosocomial clusters have recently suggested interhuman transmission among immunocompromised renal transplant patients (5–9, 12, 13, 15). Earlier studies using single-locus genotyping were inconclusive (16, 17); however, four-locus-based MLST analysis in the present study, incorporating discriminatory loci identified a predominant MLST genotype from our cluster, similar to previously observed results (7, 13, 15). In one study, human-to-human transmission was traceable in 22 of 33 PJP cases using ITS1/2 sequencing (12). Molecular and contact tracing data support the notion of patient-to-patient transmission.
Interhuman PJP transmission is likely mediated by airborne droplets (1, 2). P. jirovecii colonizes the respiratory tract in 0% to 20% healthy adults (1, 18), 65% in postmortem lungs (19), and 31% to 68% of HIV-infected people (1). Colonization frequencies of kidney transplant recipients are poorly documented. Surveillance of this patient group to estimate prevalence of colonization would be of interest yet the clinical significance of such colonization is not known. Although negative P. jirovecii PCR results from oropharyngeal rinses of outpatient health care workers mitigate against asymptomatic a staff reservoir of infection, this possibility is not excluded because the sensitivity of this method in healthy individuals is poorly defined (18); collection of induced sputum samples for PCR analysis was not feasible. Sensitivity of various PCR assays in the detection of P. jirovecii in oropharyngeal washes and other upper airway specimens (nasopharyngeal aspirates) have been reported to range from 78% to 90% in HIV/AIDS and other immunocompromised patients (20, 21). We were unable to detect P. jirovecii in ambient air using standard solid supports for trapping of potential organisms tested after outbreak control was underway, yet others have reported (albeit uncommonly) the presence of P. jirovecii DNA in air filters placed in hospital rooms of patients with PJP (22, 23). More recently, using new liquid impactor air sampling devices P. jirovecii was recovered from exhaled air in 79.8% at 1 m from infected patient's heads and in one third in the adjacent corridor at 8 m, (24) comparable most transplant outpatient clinic dimensions, including ours. Genotyping studies examining the link between patient P. jirovecii isolates with putative source isolates are required to more definitely determine mode of transmission.
Only 4.2% of exposed patients developed PJP, and risk factors derived from sporadic cases collected by registries may not extrapolate to nosocomial outbreaks. Our cohort analysis identified pulmonary disease, previous CMV infection, and transplant dysfunction as independent predictors of PJP; consistent with host vulnerability, but not previous rejection or corticosteroids seen in older studies—where early routine PJP prophylaxis was usually absent and rejection rates higher (11, 14). CMV is an immunosuppressive infection and has been associated with sporadic PJP from old registry series (11, 14) and retrospective United States Renal Data System data (25), and was a strong risk in this nosocomial cluster (OR=65.9).
The new finding of impaired graft function independently associated with PJP, may be also explained by reduced host resistance, as uremia impairs monocyte and T-cell function (needed for PJP clearance) and downregulates costimulatory B7–2 (CD86) molecules (1, 26).
Preexisting lung disease was another novel risk factor identified this our study (OR=10.1). Previous tuberculosis was associated with sporadic PJP in a case-control series from Argentina (11). Asymptomatic Pneumocystis colonization occurs in cystic fibrosis (1.3%–21.6% prevalence) (27), chronic obstructive pulmonary disease (28), or interstitial lung disease (33.8% by BAL sampling) (29). Hence, abnormal lung architecture may be predisposed to P. jirovecii colonization, which along with impaired respiratory clearance may explain the association between underlying lung disease and development of overt disease.
This nosocomial PJP cluster was extremely costly: with a 14% death rate, one-half returning to dialysis with transplant failure (albeit with initial poor function), and substantial healthcare costs from prolonged ICU ventilation as $31,854 (mean±SD $26,048) per patient. Early epidemic control is desirable. Our outbreak was belatedly controlled by extension of the duration of TMP-SMZ prophylaxis from 6 months to 1 year in recent recipients (costing AUD $0.86 per day), and reintroduction of TMP-SMZ to all asymptomatic exposed prevalent clinic patients.
We would propose a preemptive strategy of (1) early recognition of a Pneumocystis cluster (defined by two PJP patients within 1 month) and (2) prompt intervention after the third confirmed case with blanket prophylaxis for all asymptomatic exposed patients—to limit outbreak escalation. Just two cases statistically exceeded our low baseline prevalence rates, which interestingly, is the same number epidemiologists use to define a cluster. Blanket anti-PJP prophylaxis was successful in limiting the number of cases. An alternative strategy would be to extend or reintroduce prophylaxis for only those patients with significant risk factors for PJP, that is, CMV infection, previous lung disease, transplant dysfunction. This approach however is contingent on timely physician and laboratory assessment and adequate resources, and would require consensus for when to institute prophylaxis.
The average incubation period of Pneumocystis ranged from 124 to 172 days (similar to Yazaki's absolute range of 1–24 weeks) (12), means that once outbreaks become established, new cases present regularly (months to years in one outbreak), and can form larger clusters if uncontrolled (6, 12). The strategy of universal blanket prophylaxis and extension from 6 to 12 months TMP-SMZ treatment duration in recent patients to eliminate active Pneumocystis from the unit's transplant population is supported by our economic modeling where prophylaxis with inexpensive generic TMP-SMZ generated 17 days of life and $2634 saved over the lifetime of a transplant recipient, and was cost-saving (data not presented).
In summary, nosocomial PJP clusters in solid organ transplantation are likely due to interhuman transmission by airborne droplets to hosts rendered susceptible from previous CMV infection, pulmonary disease, and uremia. Early recognition of an outbreak and a strategy of prompt preemptive intervention with blanket TMP-SMZ prophylaxis to all exposed transplant clinic patients is recommended to limit its potential escalation.
MATERIALS AND METHODS
Patients and Case Definitions
During March 2010 to January 2011, 14 cases of PJP occurred in the Renal Transplant Unit of Westmead Hospital, Sydney, beginning with an index case (patient 1) transferred from interstate had not received anti-Pneumocystis prophylaxis. Our unit routinely uses universal TMP-SMZ (800/160 mg) prophylaxis for 6 months after transplantation (or dapsone, pyrimethamine, or trimethoprim combinations for sulfa allergy) (30).
Case definitions used for this cluster included all kidney transplant recipients at Westmead Hospital from March 1, 2010, onward (date chosen from expected incubation period before presentation of index case until the last case in January 2011), who presented with typical symptoms of PJP (including cough and dyspnea), characteristic radiographic findings (pulmonary interstitial abnormalities on imaging), and response to TMP-SMZ treatment. Two patient groups were defined as follows:
- Confirmed cases: clinically typical PJP cases where P. jirovecii was detected by a P. jirovecii-specific PCR-based assay. P. jirovecii DNA was detected in induced sputum and/or BAL specimens using an “in-house” real-time PCR assay adapted from Brancart et al. (31) (see Materials and Methods, SDC 2, http://links.lww.com/TP/A552).
- Probable cases: clinically typical cases who responded to specific cotrimoxazole therapy occurring within the outbreak time period—but where induced sputum or BAL specimens were unavailable for PCR testing, and alternate causes of lung infection were excluded. All patients had induced sputum/BAL fluid examined for the presence of bacteria (including Nocardia and Legionella culture), mycobacteria, fungi, and respiratory viruses and CMV. In addition, urine was collected for testing for the presence of pneumococcal and Legionella antigen using standard laboratory methods.
Contact tracing was undertaken by review of outpatient clinic visits, hospitalizations, and other potential contact points in all PJP cases. Oropharyngeal rinses from health care workers from the outpatient facility were tested for the presence of P. jirovecii DNA by real-time PCR. Sampling of ambient air from consulting rooms, corridor, and waiting rooms of outpatient clinic used an air sampler placed 1 m above the floor (Model MAS-100, Merck, Whitehorse, NJ) with testing for P. jirovecii DNA within 48 hr (31).
Genotyping of P. jirovecii was performed by MLST based on sequence analysis of four genetic loci: the internal transcribed spacer 1 and 2 (ITS1/2) regions including the 5.8S rRNA gene of the nuclear rRNA gene cluster, and the P. jirovecii-specific β-TUB, mtLSU, and DHPS genes (6, 13). Upper case italics as (β-TUB), Using DNA extracts from 11 confirmed PJP cases, the ITS1/2, β-TUB, β-TUB, mtLSU, and DHPS loci were amplified using primers and amplification conditions as specified (see Table, SDC 3, http://links.lww.com/TP/A553) (32–35). Amplicons were purified and then sequenced commercially (Macrogen Inc., Seoul, Korea). Sequences were aligned (Bioedit Sequence Alignment Editor program, version 18.104.22.168 and Sequencer version 4.7), concatenated, and further aligned with published PJP reference sequences for each of the four MLST loci, archived in the GenBank database (β-TUB; AF170964, DHPS; AJ586567, mtLSU: M58605; and ITS1/2: U07220). Genotypic comparison used DNA extracts from nine outbreak-unrelated patient samples, “the outgroup.” Phylogenetic analyses were performed using the program PAUP* version 4.0b10.
An unpaired Student t test or Wilcoxon test was used for nominal data, conditional binomial tests examined categorical data, and Poisson regression modeled the PJP incidence. Multiple logistic regression identified risk after backward elimination. The incubation period was calculated by the mean of the minimum and maximal clinic cross-over days between prodromal asymptomatic PJP patient before clinical PJP presentation. ORION guidelines for cluster outbreak reporting were used (36). The institutional ethics approval was HREC 2011/4/5.3 (3301). Data are expressed as mean±SD unless stated, and a probability less than 0.05 is considered significant.
The authors thank Karen Byth-Wilson for statistics; Elaine Chan for costing; Kathy Dempsey and Allison Sutor for collection of samples; Carmen Paterson and Natalie Perkins for contact tracing; the CIDMLS molecular biology staff for DNA extractions; and Sue Sleiman, Vishal Ahujah, and Rady Kim for support in the Mycology Laboratory. The authors acknowledge clinical support from Drs. Jeremy Chapman, Philip O'Connell, Angela Webster, Kamal Sud, Lukas Karaitis, and Grahame Elder.
1.Catherinot E, Lanternier F, Bougnoux ME, et al. Pneumocystis jirovecii pneumonia
. Infect Dis Clin North Am
2010; 24: 107.
2.Cushion MT. Are members of the fungal genus Pneumocystis
(a) commensals; (b) opportunists; (c) pathogens; or (d) all of the above? PLoS Pathog
2010; 6: e1001009.
3.Monnet X, Vidal-Petiot E, Osman D, et al. Critical care management and outcome of severe Pneumocystis pneumonia
in patients with and without HIV infection. Crit Care
2008; 12: R28.
4.de Boer MG, de Fijter JW, Kroon FP. Outbreaks and clustering of Pneumocystis pneumonia
in kidney transplant recipients: A systematic review. Med Mycol
2011; 49: 673.
5.Arichi N, Kishikawa H, Mitsui Y, et al. Cluster
outbreak of Pneumocystis pneumonia
among kidney transplant patients within a single center. Transplant Proc
2009; 41: 170.
6.de Boer MG, Bruijnesteijn van Coppenraet LE, Gaasbeek A, et al. An outbreak of Pneumocystis jiroveci pneumonia
with 1 predominant genotype among renal transplant recipients: Interhuman transmission or a common environmental source? Clin Infect Dis
2007; 44: 1143.
7.Gianella S, Haeberli L, Joos B, et al. Molecular evidence of interhuman transmission in an outbreak of Pneumocystis jirovecii pneumonia
among renal transplant recipients. Transpl Infect Dis
2010; 12: 1.
8.Hennequin C, Page B, Roux P, et al. Outbreak of Pneumocystis carinii pneumonia
in a renal transplant unit. Eur J Clin Microbiol Infect Dis
1995; 14: 122.
9.Hocker B, Wendt C, Nahimana A, et al. Molecular evidence of Pneumocystis
transmission in pediatric transplant unit. Emerg Infect Dis
2005; 11: 330.
10.Kumar D, Gourishankar S, Mueller T, et al. Pneumocystis jirovecii pneumonia
after rituximab therapy for antibody-mediated rejection in a renal transplant recipient. Transpl Infect Dis
2009; 11: 167.
11.Radisic M, Lattes R, Chapman JF, et al. Risk factors for Pneumocystis carinii pneumonia
in kidney transplant recipients: A case-control study. Transpl Infect Dis
2003; 5: 84.
12.Yazaki H, Goto N, Uchida K, et al. Outbreak of Pneumocystis jiroveci pneumonia
in renal transplant recipients: P. jiroveci
is contagious to the susceptible host. Transplantation
2009; 88: 380.
13.Schmoldt S, Schuhegger R, Wendler T, et al. Molecular evidence of nosocomial Pneumocystis jirovecii
transmission among 16 patients after kidney transplantation. J Clin Microbiol
2008; 46: 966.
14.Arend SM, Westendorp RG, Kroon FP, et al. Rejection treatment and cytomegalovirus infection as risk factors for Pneumocystis carinii pneumonia
in renal transplant recipients. Clin Infect Dis
1996; 22: 920.
15.Rabodonirina M, Vanhems P, Couray-Targe S, et al. Molecular evidence of interhuman transmission of Pneumocystis pneumonia
among renal transplant recipients hospitalized with HIV-infected patients. Emerg Infect Dis
2004; 10: 1766.
16.Helweg-Larsen J, Tsolaki AG, Miller RF, et al. Clusters of Pneumocystis carinii pneumonia
: Analysis of person-to-person transmission by genotyping. QJM
1998; 91: 813.
17.Olsson M, Eriksson BM, Elvin K, et al. Genotypes of clustered cases of Pneumocystis carinii pneumonia
. Scand J Infect Dis
2001; 33: 285.
18.Vargas SL, Pizarro P, Lopez-Vieyra M, et al. Pneumocystis
colonization in older adults and diagnostic yield of single versus paired noninvasive respiratory sampling. Clin Infect Dis
2010; 50: e19.
19.Ponce CA, Gallo M, Bustamante R, et al. Pneumocystis
colonization is highly prevalent in the autopsied lungs of the general population. Clin Infect Dis
2010; 50: 347.
20.Durand-Joly I, Chabe M, Soula F, et al. Molecular diagnosis of Pneumocystis pneumonia
. FEMS Immunol Med Microbiol
2005; 45: 405.
21.Tsolaki AG, Miller RF, Wakefield AE. Oropharyngeal samples for genotyping and monitoring response to treatment in AIDS patients with Pneumocystis carinii pneumonia
. J Med Microbiol
1999; 48: 897.
22.Bartlett MS, Vermund SH, Jacobs R, et al. Detection of Pneumocystis carinii
DNA in air samples: Likely environmental risk to susceptible persons. J Clin Microbiol
1997; 35: 2511.
23.Olsson M, Sukura A, Lindberg LA, et al. Detection of Pneumocystis carinii
DNA by filtration of air. Scand J Infect Dis
1996; 28: 279.
24.Choukri F, Menotti J, Sarfati C, et al. Quantification and spread of Pneumocystis jirovecii
in the surrounding air of patients with Pneumocystis pneumonia
. Clin Infect Dis
2010; 51: 259.
25.Neff RT, Jindal RM, Yoo DY, et al. Analysis of USRDS: Incidence and risk factors for Pneumocystis jirove
2009; 88: 135.
26.Girndt M, Sester M, Sester U, et al. Molecular aspects of T- and B-cell function in uremia. Kidney Int Suppl
2001; 78: S206.
27.Calderon EJ, Friaza V, Dapena FJ, et al. Pneumocystis jirovecii
and cystic fibrosis. Med Mycol
2010; 48(suppl 1): S17.
28.Calderon EJ, Rivero L, Respaldiza N, et al. Systemic inflammation in patients with chronic obstructive pulmonary disease who are colonized with Pneumocystis jiroveci
. Clin Infect Dis
2007; 45: e17.
29.Vidal S, de la Horra C, Martin J, et al. Pneumocystis jirovecii
colonisation in patients with interstitial lung disease. Clin Microbiol Infect
2006; 12: 231.
30.Green H, Paul M, Vidal L, et al. Prophylaxis of Pneumocystis pneumonia
in immunocompromised non-HIV-infected patients: Systematic review and meta-analysis of randomized controlled trials. Mayo Clin Proc
2007; 82: 1052.
31.Brancart F, Rodriguez-Villalobos H, Fonteyne PA, et al. Quantitative TaqMan PCR for detection of Pneumocystis jiroveci
. J Microbiol Methods
2005; 61: 381.
32.Montes-Cano MA, de la Horra C, Martin-Juan J, et al. Pneumocystis jiroveci
genotypes in the Spanish population. Clin Infect Dis
2004; 39: 123.
33.Beser J, Hagblom P, Fernandez V. Frequent in vitro recombination in internal transcribed spacers 1 and 2 during genotyping of Pneumocystis jirovecii
. J Clin Microbiol
2007; 45: 881.
34.Lee CH, Helweg-Larsen J, Tang X, et al. Update on Pneumocystis carinii
f. sp. hominis typing based on nucleotide sequence variations in internal transcribed spacer regions of rRNA genes. J Clin Microbiol
1998; 36: 734.
35.van Hal SJ, Gilgado F, Doyle T, et al. Clinical significance and phylogenetic relationship of novel Australian Pneumocystis jirovecii
genotypes. J Clin Microbiol
2009; 47: 1818.
36.Stone SP, Cooper BS, Kibbler CC, et al. The ORION statement: Guidelines for transparent reporting of outbreak reports and intervention studies of nosocomial
infection. Lancet Infect Dis
2007; 7: 282.
Nosocomial; cluster; Pneumocystis jirovecii; Pneumocystis carinii; pneumonia
Supplemental Digital Content
© 2011 Lippincott Williams & Wilkins, Inc.