Pneumocystis pneumonia (PCP) remains a significant cause of morbidity and mortality in HIV-positive patients [1–3]. Widespread trimethoprim-sulfamethoxazole (TMP-SMX) use as PCP prophylaxis has raised concerns regarding the development of sulfa-resistant Pneumocystis jirovecii (formerly P. carinii) [4,5]. Sulfa medications (SMX and dapsone, a sulfone) interfere with folate synthesis by competitively inhibiting the enzyme dihydropteroate synthase (DHPS). These medications exert apparent selective pressure on P. jirovecii, as more DHPS gene mutations are observed in patients previously exposed to sulfa medications [7–11].
Although similar DHPS mutations confer resistance to sulfa medications in other organisms, [12,13], the association between DHPS mutations and clinical resistance in P. jirovecii is unclear. Without in vitro culture systems for direct sensitivity testing of P. jirovecii, inferences regarding this putative association are drawn primarily from epidemiological studies, which offer conflicting results. Although one study demonstrated increased mortality and a second found increased treatment failure , a third found neither . In all three studies, the majority of patients with Pneumocystis DHPS mutations responded favorably to TMP-SMX treatment [15,16]. The explanation for these results is unclear. Prior studies included relatively few patients with Pneumocystis DHPS mutations who were treated with TMP-SMX, limiting their ability to control for independent risk factors associated with mortality or treatment failure [14–16]. A 5-year prospective study of PCP in HIV-positive patients admitted to San Francisco General Hospital has been conducted to determine whether severity and outcome of PCP differ according to Pneumocystis DHPS genotype. The study also examined the impact of TMP-SMX dosing on outcome. This report describes the largest number of PCP episodes containing mutant Pneumocystis DHPS genotypes to date, substantially expanding upon our previous report .
Patients were consecutive HIV-infected adults with microscopically confirmed PCP admitted to the San Francisco General Hospital between 1 May 1997 and 31 July 2002. Patients provided written informed consent; the protocol was approved by the institutional review boards of the San Francisco General Hospital and the Centers for Disease Control and Prevention (CDC). The analysis included 215 episodes of PCP, including 70 episodes previously reported .
PCP infections were microscopically confirmed by sputum induction or bronchoalveolar lavage as part of routine diagnostic procedures. After clinical use, the remainder of the specimen was shipped to the CDC where the DHPS locus was amplified by the polymerase chain reaction and genotyped by direct DNA sequencing, as previously described .
Dihydropteroate synthase genotype
Laboratory investigators classified P. jirovecii DHPS genotypes blinded to clinical information. As in previous studies, the wild-type genotype was defined as the DNA sequence observed in Pneumocystis specimens from other mammalian species and humans before 1995. A mutant genotype was defined as any DNA sequence that differed from the wild-type sequence. The observed mutations were all non-synonymous changes that resulted in amino acid substitutions at amino acid positions 55 or 57, identical to mutations reported elsewhere . Specimens containing multiple Pneumocystis DHPS genotypes that included at least one mutant genotype were classified as mutant in the analyses.
Outcomes and trimethoprim-sulfamethoxazole dosing
Clinical investigators performed chart abstraction using standardized data abstraction forms and classified patient outcomes blinded to genotype information. The primary end-point was all-cause mortality within 60 days after PCP diagnosis. Secondary end-points were death caused by PCP, TMP-SMX treatment failure, and the need for mechanical ventilation. Death was classified as caused by PCP when the physician recorded PCP as the primary cause of death in the medical record and death certificate. A TMP-SMX treatment failure was classified when therapy was changed owing to lack of clinical response documented in the medical record. A therapy change owing to adverse effects was not classified as a TMP-SMX treatment failure.
Dose and frequency of all anti-PCP medications administered during hospitalization were recorded. TMP-SMX was administered in fixed-dose combinations (1 mg TMP/5 mg SMX). Daily dose was calculated (mg/kg) based on the trimethoprim component. Patients were excluded in analyses of dosing if TMP-SMX was reduced because of acute renal failure.
Statistical analysis was performed using Stata version 7.0 (StataCorp, College Station, Texas, USA). Univariate analyses used two-tailed Student's t-test, Wilcoxon's rank sum, chi-square, or Fisher's exact tests. Odds ratios (OR) and 95% confidence intervals (CI) were calculated to assess univariate risk factors for poor outcome. Multivariate logistic regression was performed to assess independent predictors of death. Clinically relevant variables were included in the model, and those found significant at different P values (0.05, 0.10, and 0.20) were entered into the model in a step-up fashion in order of increasing univariate P values. Statistical significance was defined as P < 0.05.
From May 1997 through July 2002, 244 HIV-infected patients with 267 episodes of PCP were enrolled. The DHPS locus was sequenced in 197 patients with 215 PCP episodes. There were no significant differences in patient characteristics, PCP severity, or outcomes between these 215 episodes and the 47 episodes excluded because genotyping was unsuccessful. Recurrent episodes of PCP were included in the analyses; their exclusion had no significant effect on the multivariate model.
Prevalence and predictors of mutations in the gene for dihydropteroate synthase
During the five-year study, 81% (175 of 215) of the Pneumocystis DHPS genotypes were mutant, without significant change over time (P = 0.25). Of the mutant genotypes, 65% (113) contained amino acid substitutions at both positions 55 and 57. Patients receiving sulfa or sulfone PCP prophylaxis within 3 months preceding admission were more likely to present with PCP containing Pneumocystis DHPS gene mutations compared with patients without this history (OR, 2.79; 95% CI, 1.17–6.66; P = 0.021). In contrast, patients with newly diagnosed HIV within 30 days preceding PCP were less likely to present with mutant DHPS genotypes (OR, 0.48; 95% CI, 0.23–0.99; P = 0.048).
Severity and outcome of Pneumocystis pneumonia according to dihydropteroate synthase genotype
There were no significant differences in baseline patient characteristics according to Pneumocystis DHPS genotype (Table 1). There were no significant differences in initial PCP severity, hospital course or complications. Despite comparable patient characteristics and severity of PCP, there was a trend for a higher proportion of patients with PCP containing Pneumocystis DHPS mutations to have worse outcomes compared with patients with wild-type DHPS. Overall, 14.3% of patients with PCP containing a mutant Pneumocystis DHPS genotype died, compared with 7.5% of patients with PCP containing wild-type (P = 0.31), and 14.3% compared with 2.5% required mechanical ventilation (P = 0.056).
Univariate predictors of mortality included a serum albumin < 30 g/l and the requirement for early admission to an intensive care unit (ICU) (within 72 h of hospitalization) (Table 2). This window was chosen because patients who would normally qualify for ICU admission were occasionally denied immediate admission through lack of beds. Serum albumin (OR, 4.62; 95% CI, 1.63–13.1; P = 0.004) and early ICU admission (OR, 5.06; 95% CI, 1.43–18.0; P = 0.012) remained independent predictors of mortality in all multivariate models tested (Table 2). The presence of Pneumocystis DHPS mutations was not associated with mortality (OR, 2.58; 95% CI, 0.70–9.53; P = 0.16).
Impact of dihydropteroate synthase genotypes and trimethoprim-sulfamethoxazole dose on outcome
To examine the combined impact of Pneumocystis DHPS genotype and TMP-SMX treatment on outcome, the outcomes of PCP were compared according to DHPS genotype (wild-type or mutant) and PCP treatment (TMP-SMX or non-TMP-SMX). Overall, 85% of patients were treated with TMP-SMX. Of patients with mutant Pneumocystis DHPS genotypes treated with TMP-SMX, 15.1% died compared with only 8.3% of patients with wild-type DHPS genotypes treated with TMP-SMX and 7.7% of those with mutant DHPS treated with non-TMP-SMX (P = 0.49 for the overall comparison). Likewise, more patients with mutant Pneumocystis DHPS genotypes treated with TMP-SMX tended to die from PCP, to fail TMP-SMX treatment, or to require mechanical ventilation (not significant for all comparisons).
To examine the impact of TMP-SMX dosing on outcome, the overall mortality of patients who received ≤ 15.0 mg/kg daily (‘low-dose’) was compared with patients who received > 15.0 mg/kg daily (‘high-dose’) according to DHPS genotype. This cut-off was chosen because it represents the lower TMP-SMX dose of the recommended treatment range for PCP in HIV-infected patients, and we were interested in examining potential low-level sulfa drug resistance. Duration of in-hospital TMP-SMX therapy and the average daily dose of TMP-SMX were similar in patients with PCP containing mutant versus wild-type DHPS. Although not statistically significant, patients treated with low-dose versus high-dose TMP-SMX tended to be more likely to die if they had mutant DHPS (OR, 1.98; 95% CI, 0.76–5.19; P = 0.16) than if they had wild-type DHPS (OR, 1.06; 95% CI, 0.06–18.5; P = 1.0).
This study represents the largest cohort of HIV-infected patients with PCP containing Pneumocystis DHPS gene mutations to date. We observed a mutant DHPS genotype in 81% of the 215 episodes of PCP. Overall, we found no significant differences in the patient characteristics or PCP severity according to DHPS genotype. Despite these similarities, there was a trend for a higher proportion of patients with PCP containing Pneumocystis DHPS mutations to have worse outcomes compared with patients with wild-type DHPS.
In our analyses, the presence of a mutant DHPS genotype was not an independent predictor of increased mortality. Overall, 86% of patients with PCP containing Pneumocystis DHPS gene mutations survived. When we examined only episodes of PCP treated with TMP-SMX, 85% of these patients survived. Independent predictors of mortality at presentation were a low serum albumin and the requirement for early ICU admission. These results are consistent with prior studies [18–20] and suggest that mortality in most HIV-infected patients with PCP is related primarily to the underlying severity of illness and the initial severity of PCP.
Although we were unable to demonstrate that Pneumocystis DHPS mutations were an independent predictor of mortality, as found in another study , several results combine to suggest that there may be an association between DHPS mutations and poor outcomes. First, patients with mutant Pneumocystis DHPS genotypes tended to be more likely to die and to fail therapy with TMP-SMX. These results contrasted with our prior study where patients with mutant DHPS genotypes were less likely to die and to fail TMP-SMX therapy when compared with patients with wild-type DHPS . The explanation for this apparent qualitative reversal remains unclear as the Pneumocystis DHPS genotypes observed and the clinical practice at our institution are unchanged. Next, more patients with mutant Pneumocystis DHPS genotypes required mechanical ventilation. Finally, there was a trend for patients with Pneumocystis DHPS mutations to do less well if they receive low-dose TMP-SMX, raising the possibility that DHPS mutations may confer a low-level of resistance to TMP-SMX that is overcome by daily doses of > 15.0 mg/kg.
Our study had limitations. Although we have seen trends towards worse outcomes in patients with PCP containing Pneumocystis DHPS gene mutations and have reported on the largest series with DHPS mutations to date, we had insufficient power to detect differences of these magnitudes with statistical significance. Potentially, these differences may be significant with a larger sample. Since this was an observational study, patients were not randomized to low-dose or high-dose TMP-SMX to allow full investigation of the question of low-level TMP-SMX drug resistance – a study design that cannot be carried out in humans. Also, our analysis was based on the dose of TMP-SMX per kilogram body weight, as is the current clinical practice,  rather than on serum drug levels. Despite similar dosing, drug levels could have varied between patients.
In conclusion, we demonstrated that the majority of HIV-infected patients with PCP containing Pneumocystis DHPS gene mutations survived. However, our data also raised the possibility that mutant Pneumocystis DHPS genotypes may be associated with increased mortality and increased TMP-SMX failure, highlighting the importance of continuing studies on the question of sulfa drug resistance in Pneumocystis spp. The present mutations may be only the first in a series, and highly resistant P. jirovecii may emerge in the future. Therefore, continued work on a reliable culture system for P. jirovecii and on new medications to treat PCP is imperative.
Sponsorship: C. B. Beard was supported in part by funds from the CDC Opportunistic Infections Working Group. L. Huang was supported in part by NIH K23 HL072117, and A. Morris in part by NIH K23 HL072837.
1. Jones JL, Hanson DL, Dworkin MS, Alderton DL, Fleming PL, Kaplan JE, et al
. Surveillance for AIDS-defining opportunistic illnesses, 1992–1997. MMWR Surveill Summ 1999; 48:1–22.
2. Aderaye G, Bruchfeld J, Olsson M, Lindquist L. Occurrence of Pneumocystis carinii
in HIV-positive patients with suspected pulmonary tuberculosis in Ethiopia. AIDS 2003; 17:435–440.
3. Worodria W, Okot-Nwang M, Yoo SD, Aisu T. Causes of lower respiratory infection in HIV-infected Ugandan adults who are sputum AFB smear-negative. Int J Tuberc Lung Dis 2003; 7:117–123.
4. Stringer JR, Beard CB, Miller RF, Wakefield AE. A new name (Pneumocystis jiroveci
) for Pneumocystis
from humans. Emerg Infect Dis 2002; 8:891–896.
5. Armstrong W, Meshnick S, Kazanjian P. Pneumocystis carinii
mutations associated with sulfa and sulfone prophylaxis failures in immunocompromised patients. Microbes Infect 2000; 2:61–67.
6. Lane BR, Ast JC, Hossler PA, Mindell DP, Bartlett MS, Smith JW, et al
. Dihydropteroate synthase polymorphisms in Pneumocystis carinii
. J Infect Dis 1997; 175:482–485.
7. Kazanjian P, Locke AB, Hossler PA, Lane BR, Bartlett MS, Smith JW, et al
. Pneumocystis carinii
mutations associated with sulfa and sulfone prophylaxis failures in AIDS patients. AIDS 1998; 12:873–878.
8. Ma L, Borio L, Masur H, Kovacs JA. Pneumocystis carinii
dihydropteroate synthase but not dihydrofolate reductase gene mutations correlate with prior trimethoprim-sulfamethoxazole or dapsone use. J Infect Dis 1999; 180:1969–1978.
9. Ma L, Kovacs JA, Cargnel A, Valerio A, Fantoni G, Atzori C. Mutations in the dihydropteroate synthase gene of human-derived Pneumocystis carinii
isolates from Italy are infrequent but correlate with prior sulfa prophylaxis. J Infect Dis 2002; 185:1530–1532.
10. Huang L, Beard CB, Creasman J, Levy D, Duchin JS, Lee S, et al
. Sulfa or sulfone prophylaxis and geographic region predict mutations in the Pneumocystis carinii
dihydropteroate synthase gene. J Infect Dis 2000; 182:1192–1198.
11. Miller RF, Lindley AR, Ambrose HE, Malin AS, Wakefield AE. Genotypes of Pneumocystis jiroveci
isolates obtained in Harare, Zimbabwe, and London, United Kingdom. Antimicrob Agents Chemother 2003; 47:3979–3981.
12. Dallas WS, Gowen JE, Ray PH, Cox MJ, Dev IK. Cloning, sequencing, and enhanced expression of the dihydropteroate synthase gene of Escherichia coli
MC4100. J Bacteriol 1992; 174:5961–5970.
13. Brooks DR, Wang P, Read M, Watkins WM, Sims PF, Hyde JE. Sequence variation of the hydroxymethyldihydropterin pyrophosphokinase: dihydropteroate synthase gene in lines of the human malaria parasite, Plasmodium falciparum
, with differing resistance to sulfadoxine. Eur J Biochem 1994; 224:397–405.
14. Helweg-Larsen J, Benfield TL, Eugen-Olsen J, Lundgren JD, Lundgren B. Effects of mutations in Pneumocystis carinii
dihydropteroate synthase gene on outcome of AIDS-associated P. carinii
pneumonia. Lancet 1999; 354:1347–1351.
15. Kazanjian P, Armstrong W, Hossler PA, Burman W, Richardson J, Lee CH, et al
. Pneumocystis carinii
mutations are associated with duration of sulfa or sulfone prophylaxis exposure in AIDS patients. J Infect Dis 2000; 182:551–557.
16. Navin TR, Beard CB, Huang L, del Rio C, Lee S, Pieniazek NJ, et al
. Effect of mutations in Pneumocystis carinii
dihydropteroate synthase gene on outcome of P. carinii
pneumonia in patients with HIV-1: a prospective study. Lancet 2001; 358:545–549.
17. Beard CB, Carter JL, Keely SP, Huang L, Pieniazek NJ, Moura IN, et al
. Genetic variation in Pneumocystis carinii
isolates from different geographic regions: implications for transmission. Emerg Infect Dis 2000; 6:265–272.
18. Bedos JP, Dumoulin JL, Gachot B, Veber B, Wolff M, Regnier B, et al
. Pneumocystis carinii
pneumonia requiring intensive care management: survival and prognostic study in 110 patients with human immunodeficiency virus. Crit Care Med 1999; 27:1109–1115.
19. Forrest DM, Zala C, Djurdjev O, Singer J, Craib KJ, Lawson L, et al
. Determinants of short- and long-term outcome in patients with respiratory failure caused by AIDS-related Pneumocystis carinii
pneumonia. Arch Intern Med 1999; 159:741–747.
20. Morris A, Wachter RM, Luce J, Turner J, Huang L. Improved survival with highly active antiretroviral therapy in HIV-infected patients with severe Pneumocystis carinii
pneumonia. AIDS 2003; 17:73–80.
21. Kovacs JA, Gill VJ, Meshnick S, Masur H. New insights into transmission, diagnosis, and drug treatment of Pneumocystis carinii
pneumonia. JAMA 2001; 286:2450–2460.