Résumé
Background: Pneumocystis pneumonia (PCP) case clusters involving solid organ transplant (SOT) recipients have been reported worldwide. Mycophenolic acid, an immunosuppressant used to prevent rejection in SOT recipients that targets the inosine 5 ′ -monophosphate dehydrogenase (IMPDH) protein, has been hypothesised to exert selective pressure on Pneumocystis jirovecii, with the missense mutation G1020A (Ala261Thr) in the impdh gene a possible marker of such selective pressure. The aim of this study was to test the hypothesis that SOT recipients harbour P jirovecii with mutations in the impdh gene and are infected with specific P jirovecii strains.Methods: In this retrospective, multicentre, cross-sectional study of nationally representative, individual-level data, we included SOT recipients, regardless of the organ transplanted, involved or not in a PCP case cluster, and non-SOT recipients (control group) without mycophenolic acid exposure, diagnosed with PCP from 26 French and one Swiss secondary health-care centres. We included patients aged 18 years or older for whom archival P jirovecii DNA extracts were available in sufficient quantity and quality. P jirovecii specimens were characterised using a multilocus sequence typing method including impdh gene analysis. The primary outcome of this study was the detection of the G1020A (Ala261Thr) mutation in the impdh gene. A multivariable logistic regression was done to assess the relation between this mutation and the following variables retrieved from medical records: age, mycophenolic acid exposure at the time of PCP diagnosis, involvement in a PCP case cluster, PCP prophylaxis, and clinical outcome.Findings: 58 SOT recipients (44 treated with mycophenolic acid) and 59 non-SOT recipients (control group; not treated with mycophenolic acid) diagnosed with PCP between Jan 1, 2001, and Dec 31, 2021, were enrolled. The G1020A (Ala261Thr) mutation was detected in P jirovecii specimens from 40 (68⋅9%) SOT recipients (37 treated with mycophenolic acid) and in none of the P jirovecii specimens from the patients in the control group. The multivariable analysis showed that the allele characterised by the G1020A mutation was associated with mycophenolic acid exposure at the time of PCP diagnosis (adjusted odds ratio 73⋅61 [95% CI 17⋅41-455⋅70]; p<0⋅0001) and involvement in a PCP case cluster (12⋅77 [1⋅58-171⋅90]; p=0⋅029), whereas it was not associated with age, PCP prophylaxis, and clinical deterioration. A second missense mutation, G1020T (Ala261Ser) was identified in P jirovecii specimens from three SOT recipients (two treated with mycophenolic acid). Two specific multilocus genotypes (MLG-31 and MLG-34) of P jirovecii associated with Ala261Thr and Ala261Ser substitutions in IMPDH, respectively, were detected only in SOT recipients (38 patients with MLG-31 and three patients with MLG-34).Interpretation SOT recipients in this study were primarily infected with specific P jirovecii strains with mutations in the impdh gene, which might confer a selective advantage as both the G1020A (Ala261Thr) and G1020T (Ala261Ser) are associated with mycophenolic acid resistance in other fungi. Mycophenolic acid selective pressure might explain the maintenance and circulation of these P jirovecii strains within this patient population, and consequently their potential involvement in PCP case clusters.