Combination of GM CSF and carbapenem is superior to carbapenem monotherapy in difficult-to-treat spontaneous bacterial peritonitis: A randomized controlled trial

Background: Patients with cirrhosis and treatment non-responsive spontaneous bacterial peritonitis (SBP) have high mortality. We aimed to investigate whether GM-CSF can improve SBP response rates.

Patients and methods: In this open-label RCT, 131 cirrhosis patients with difficult-to-treat SBP (DTT SBP) were randomized to receive meropenem alone (1 g IV thrice daily for 5 days) (MERO Group, n = 66) or in combination with GM-CSF (1.5 mcg/Kg daily IV till resolution or till 5d) (MEROGM Group, n = 65). The primary end-point was SBP early-response (reduction in absolute neutrophil count (ANC) by >25% after 48 h). Secondary end-points included SBP resolution at day 5.

Results: Patients in MEROGM group in comparison to MERO group had higher SBP early-response (60% vs. 31.8%; p = .001) and SBP resolution rates (55.4% vs. 24.2%; p = .0003). Patients in the combination arm also had better resolution of pneumonia {8/17 (47.05%) vs. 2/19 (10.5%), p = .02} and lower incidence of new-onset AKI (15.4% vs. 31.8%, p = .02), HE (18.5% vs. 34.8%, p = .04) and infections (21.5% vs. 37.9%, p = .05). In comparison to MERO group, 7-day survival was higher in MEROGM group (89.2% vs. 78.7%, p = .03), though the 28-day survival was comparable (78.4% vs. 71.2%; p = .66). None of the patients developed treatment-related severe adverse effects requiring discontinuation of therapy.

Conclusions: The addition of GM-CSF to meropenem significantly improves response rates in DTT SBP patients within 48 h. Early use of GMCSF modulates host immune response, and enhances antibiotic response with higher SBP resolution. The use of GMCSF needs to be considered in combating difficult SBP in cirrhosis patients.