Gut Faecalibacterium abundance in patients with plasma cell disorders is associated with survival after autologous HSCT

Abstract

The gut microbiota (GM) has been linked to the development, progression, and response to therapy in plasma cell neoplasms (PCNs). The primary goal of this study was to investigate the relationship between the composition of the GM before and during autologous hematopoietic stem cell transplant (HSCT) with clinical outcomes of patients with PCNs. We focused on the genus Faecalibacterium, which includes the most abundant anaerobic commensal bacterium in the GM. Fecal samples were collected prospectively before, mid (at 1 week from the start of intervention), and end (at engraftment) of intervention (liberalized vs neutropenic diet) and subjected to 16S ribosomal DNA sequencing. Eighty-three patients were enrolled. Their median age was 64 (range, 31-79) years. Fifty-four patients received HSCT as part of frontline therapy and 29 for relapsed/refractory disease. With median follow-up time for survivors (n = 82) of 32 (range, 0.7-61) months, the median progression-free survival (PFS) was 40 months. Higher preintervention Faecalibacterium abundance was associated with improved PFS (hazard ratio [HR], 0.92; 95% confidence interval [CI], 0.86-0.99; P = .02). Faecalibacterium abundance was found to decrease early after transplant (P < .01). Although the administration of high-dose melphalan (200 mg/m²) was significantly associated with PFS in both univariable (HR, 0.38, 95% CI, 0.19-0.75; P = .006) and multivariable (HR, 0.42; 95% CI, 0.20-0.87; P = .02) analyses, preintervention Faecalibacterium abundance remained independently associated with PFS (HR, 0.93; 95% CI, 0.86-0.99; P = .04) on multivariable analysis. In conclusion, lower preintervention Faecalibacterium abundance was associated with inferior PFS.

Graphical abstract

Figure 1.

Swimmer plot of all patients in this study. The plot illustrates the relative timing of the start of intervention and stool sample collection in relation to the autologous HSCT and exposure to nonprophylactic antibiotics for each patient. ID, identity.

Figure 2.

GM throughout transplant. (A) GM α-diversity using Chao1. (B) Faecalibacterium abundance before, mid, and end intervention in patients with PCNs. Both show decreases throughout transplant, with greater decreases from preintervention to mid intervention than from mid intervention to end intervention. The P values were derived by fitting a linear mixed model for each case with patient-specific random intercepts, in which the intervention was added as a factor variable. The reported P values are associated with the intervention variable.

Figure 3.

PFS by preintervention Faecalibacterium abundance percentage. Higher preintervention level Faecalibacterium is associated with improved PFS in patients with PCN undergoing autologous transplant.

Figure 4.

Forest plot of PFS subgroup analysis. (A) Univariable analysis showed the dose of melphalan conditioning chemotherapy predicted PFS; age, sex, transplant as part of frontline therapy, melphalan dose, nonprophylactic antibiotic exposure (n = 83); race (n = 81); Durie-Salmon stage (n = 76); and high-risk cytogenetics (n =74). (B) On multivariable analyses, the dose of melphalan again predicted PFS and preintervention Faecalibacterium abundance remained prognostic; all variables (n = 79). HR and 95% CI were calculated via Cox regression analysis.

Figure 5.

Correlations between the gm and nonprophylactic antibiotics (administered for treatment of suspected or documented infection) at different time points in relation to the intervention. α-Diversity (A-D). Faecalibacterium abundance (E-H). For the mid intervention and the change from preintervention to mid-intervention analyses, exposed patients were defined as those who received nonprophylactic antibiotic treatment at any point before mid intervention. For the end intervention and the change from preintervention to end-intervention analyses, exposed patients were defined as anyone who received nonprophylactic antibiotics by end of intervention.