Pneumocystis jirovecii prophylaxis with trimethoprim-sulfamethoxazole and resistance in ESBL-producing Enterobacterales bloodstream infections among kidney transplant recipients

Abstract Background Kidney transplant (KT) recipients are at high risk of bloodstream infection (BSI) caused by extended-spectrum beta-lactamase-producing Enterobacterales (ESBL-E). Trimethoprim-sulfamethoxazole (TMP-SMX) is widely used for post-transplant prophylaxis, but its relationship with subsequent TMP-SMX resistance in this setting remains poorly defined. Methods We conducted a retrospective multicentre cohort study of KT recipients with monomicrobial ESBL-E BSI from the INCREMENT-SOT cohort. The primary outcome was TMP-SMX resistance; the secondary outcome was co-resistance to TMP-SMX and ciprofloxacin. The association with TMP-SMX prophylaxis during the preceding 30 days was assessed using inverse probability of treatment weighting and modified Poisson regression with centre-clustered robust variance. Additionally, exploratory Classification and Regression Tree (CART) and stratified analyses were performed. Results Among 323 KT recipients with ESBL-E BSI, 160 (49.5%) had received TMP-SMX prophylaxis within 30 days before BSI onset. Overall, TMP-SMX resistance was detected in 267/323 cases (82.7%). Resistance was more frequent among prophylaxis recipients than among non-recipients (91.3% vs. 74.2%; P < 0.001). Recent prophylaxis was associated with TMP-SMX resistance in the IPTW analysis (RR, 1.18; 95% CI 1.02–1.37; P = 0.03) and after additional adjustment for lymphopenia, chronic liver disease, and microbiological aetiology (aRR, 1.16; 95% CI 1.00–1.35; P = 0.05). Exploratory CART analysis identified lymphopenia as the main splitting variable among prophylaxis recipients. In patients with lymphopenia, resistance occurred in 94.4% of prophylaxis recipients and 74.5% of non-recipients, with an IPTW-adjusted RR of 1.23 (95% CI 1.00–1.51; P = 0.048). TMP-SMX and ciprofloxacin co-resistance occurred in 70.3% of patients and was more frequent among prophylaxis recipients than among non-recipients (75.6% vs. 65.0%; P = 0.04), although the association did not reach statistical significance after IPTW and multivariable adjustment (aRR, 1.11; 95% CI 0.97–1.28; P = 0.10). Conclusions In KT recipients with ESBL-E BSI, recent TMP-SMX prophylaxis was associated with a higher probability that the causative isolate was resistant to TMP-SMX. Exploratory analyses identified a particularly high prevalence of resistance among prophylaxis recipients with lymphopenia. These findings should be interpreted cautiously given the observational design and the specific study population, but they may be useful for risk stratification and for anticipating limited oral treatment options in certain clinical scenarios.