In this retrospective study of critically ill patients admitted with severe or fulminant CDI, we observed that FMT had a significant mortality benefit over SOC with an absolute risk reduction of 31.2% corresponding to an NNT of 3.2 to prevent one death. With an average CCI score of 7.4, our population represented a group of severely ill patients in critical care or step-down settings. By utilising both the CCI and IDSA/SHEA guidelines, we were able to objectively classify patients regarding the severity of their CDI as well as their global illness severity. One prior study observed a similar mortality benefit for early FMT in patients with severe CDI but did not include patients with fulminant CDI and had a median CCI of 2, indicating a less severe cohort of patients.
There is very little published data on the safety outcomes of FMT in patients with severe or fulminant CDI with rates of serious adverse events ranging from 0% to 28%.[13,14,19] In our cohort, we found a small number of serious adverse events in patients who underwent FMT including one occurrence of bacteraemia (6.3%) and one perforation (6.3%). Our reported rates of severe adverse events do not differ from what has been reported to date in patients undergoing FMT. In the SOC group, there was a higher rate of bacteraemia (37.5%) and similar rate of perforation (6.3%) compared to the FMT group. Given that our study population was comprised of critically ill patients, at high risk of morbidity and mortality, the low rate of serious adverse events with FMT is encouraging. Further studies specifically looking at safety outcomes are imperative, especially with the recent report of two immunocompromised patients who developed infections with multidrug-resistant organisms traced back to their FMT donor.
There are several limitations to this study. This study was conducted using a retrospective matched cohort design which may have introduced selection bias. We attempted to control for confounding factors that could have impacted mortality other than CDI by matching for CCI and CDI severity, which classifies patients based on objective markers during hospitalisation. We also utilised the CARDS tool to assess the predicted mortality of each group. We found that patients who received FMT had a higher predicted rate of mortality compared to patients who received SOC treatment, with a CARDS score of 10.9 and 9.3 respectively (P = .03). This difference potentially strengthens the findings of our study as the FMT group had a statistically significant decrease in mortality compared to SOC, despite a higher CARDS score.
The decision to pursue FMT was primarily decided by the treatment team. Though it is possible that there was selection bias in who received FMT, we found no significant difference between the FMT and SOC groups with regard to demographics, severity indices, percentage of patients who had systolic blood pressures < 100, required pressor support or required intubation. Only 4 of the 32 patients in the SOC group were evaluated for FMT. We suspect that the majority of SOC patients were not offered FMT due to lack of knowledge about FMT as an option for severe or fulminant CDI as it is not part of current treatment guidelines. Given our findings, we believe that there should be increased awareness across specialties regarding the utility of FMT in severe and fulminant CDI. In fact, some groups have reported decrease in CDI-related colectomy rates when initiating an inpatient FMT programme.
The small number of patients included may have affected secondary outcomes including colectomy and readmission rates. For the subanalysis of fulminant cases alone, we were likely underpowered to show statistical significance. Prior studies have shown that hospitalised patients with severe and fulminant CDI have a significant rate of recurrence, which likely accounts for the high rate of recurrent CDI in both cohorts. While colectomy is part of standard of care guidelines for fulminant CDI, very few patients in our cohort underwent colectomy. It is possible that the inclusion of severe patients reduced the absolute number of colectomies as typically these patients do not proceed to surgical intervention. In our fulminant cohort, 15.1% went to surgery with no difference between FMT and SOC. While there was no difference in mortality rates at 30 and 90 days post hospitalisation, we were unable to derive any statistically significant conclusions based on small number of patients with follow up information. Further studies on long-term outcomes in this cohort are warranted.
Many studies have reported improved mortality for patients with fulminant CDI who underwent early operative intervention.[22–25] While the average time to FMT in our study was 9 days, it is possible that earlier time to FMT could have improved survival rates. In one retrospective study, early FMT (defined as 2–4 days after initiating antibiotics) was associated with improved mortality in severe CDI.
In conclusion, FMT was associated with a 77% decrease in mortality compared to SOC in hospitalised patients with severe or fulminant CDI requiring critical care. As evidence for FMT in this high-risk cohort is still emerging, selection criteria for who receives FMT has yet to be definitively determined. Our study contributes to the growing body of literature that argues for the use of FMT in severe and fulminant patients either in place of or in conjunction with surgical intervention.[11,18,19] Most importantly, we firmly believe that early intervention in this high risk cohort is key. Further prospective studies are needed to confirm our findings, assess the overall safety of FMT, and ensure that results are applicable to larger cohorts of patients with severe and fulminant CDI.
RU is supported by an NIH K23 Career Development Award (K23KD111995-01A1). AG is supported by the SUCCESS (Sinai Ulcerative Colitis Clinical Experimental and System Studies) grant from the Bacchetta Foundation. This work was supported by the Digestive Disease Research Foundation (DDRF) Fellowship award to ET.
Guarantor of the article
Ari Grinspan is the submission's guarantor and takes responsibility for the integrity of the work as a whole, from inception to published article.
Aliment Pharmacol Ther. 2019;50(10):1094-1099. © 2019 Blackwell Publishing