Abstract Rationale The prognostic impact of underlying collagen disease in patients with non-HIV pneumocystis pneumonia (PCP) remains unclear. Collagen diseases, including connective tissue diseases, vasculitis, chronic inflammatory diseases, and autoinflammatory syndromes, have shown conflicting associations with prognosis in non-HIV PCP, with some studies suggesting improved outcomes and others indicating poor survival. This inconsistency may reflect the heterogeneity among collagen diseases because different phenotypes and treatment backgrounds can influence clinical outcomes. Among these, rheumatoid arthritis (RA), a prevalent and well-characterized form, is typically managed with long-term immunosuppressive therapy. In this multicenter, retrospective study, we aimed to clarify the relationship between collagen-disease heterogeneity and prognosis in patients with non-HIV PCP. Methods We retrospectively analyzed patients diagnosed with and treated for non-HIV PCP at Kameda Medical Center, Seirei Hamamatsu General Hospital, and Seirei Mikatahara General Hospital between June 2006 and March 2021. Patients were categorized into collagen- and non-collagen-disease groups, with the former further subdivided into RA and non-RA collagen-disease groups. The primary and secondary endpoints were 30- and 180-day mortalities, respectively. Propensity-score adjustment balanced baseline characteristics, and logistic regression, and Kaplan-Meier analyses identified prognostic factors. Results This study included 159 patients: 116 with collagen disease (84 with RA) and 43 without RA. After propensity-score adjustment, the collagen-disease group showed significantly lower 30- (7.7% vs. 26.5%, p = 0.020) and 180-day mortalities (21.2% vs. 45.1%, p = 0.026) than the non-collagen-disease group. In the subgroup analysis, patients with RA had lower mortality rates than those with other collagen diseases (30-day, 4.8% vs. 9.4%; 180-day, 7.1% vs. 28.1%). Multivariate logistic regression analysis identified non-RA collagen disease (odds ratio OR 0.10, p = 0.026), RA (OR 0.09, p = 0.011) and hematologic malignancy (OR 0.10, p = 0.036) as significantly associated with prognosis. Conclusions Patients with non-HIV PCP and underlying collagen disease had significantly better short- and mid-term outcomes than those without collagen disease. Individuals with RA exhibited particularly favorable prognoses. Multivariate analysis suggested that RA, non-RA collagen disease, and hematologic malignancy were the major underlying conditions influencing prognosis. These results highlight the importance of considering disease phenotypes and treatment backgrounds in managing this heterogeneous population. This abstract is funded by: None
Ito et al. (Fri,) studied this question.