Abstract
Background
Cancer has been associated with an increased risk of in-hospital mortality in CDI patients. However, data on delayed mortality in cancer patients with CDI are scarce.
Aim/Objective
The aim of the present study was to compare outcomes between oncological patients and the general population with Clostridioides difficile infection (CDI) after 90 days of follow-up.
Methods
A multicenter prospective cohort study was conducted in 28 hospitals participating in the VINCat program. Cases were all consecutive adult patients who met the case definition of CDI. Sociodemographic, clinical, and epidemiological variables and evolution at discharge and after 90 days were recorded for each case.
Findings/results
The mortality rate was higher in oncological patients (OR = 1.70, 95% CI: 1.08–2.67). In addition, oncological patients receiving chemotherapy (CT) presented higher recurrence rates (18.5% vs 9.8%, p = 0.049). Among oncological patients treated with metronidazole, those with active CT showed a higher rate of recurrence (35.3% vs 8.0% p = 0.04).
Discussion
Oncological patients presented a higher risk of poor outcomes after CDI. Their early and late mortality rates were higher than in the general population, and in parallel, those undergoing chemotherapy (especially those receiving metronidazole) had higher rates of recurrence
Keywords: Outcome, Clostridioides difficile , oncological patients, chemotherapy, vancomycin, metronidazole
Background
Malignancy is a well-established risk factor for primary and recurrent Clostridioides difficile infection (CDI) (Delgado et al., 2017). This association is likely due to alterations in gut microbiota caused by frequent exposure to broad-spectrum antibiotics, chemotherapy, the health care environment, or simply immunosuppression (Cornely et al., 2020). Cancer has also been associated with an increased risk of in-hospital mortality in CDI patients (Gupta et al., 2017). However, data on delayed mortality in cancer patients with CDI are scarce. The aim of the present study was to compare outcomes between oncological patients and the general population with CDI after 90 days of follow-up.
Methods
A multicenter prospective cohort study was conducted in 28 hospitals participating in the VINCat program (Infection Control and Antimicrobial Stewardship Catalan Program), which represent 40.9% of all adult acute-care hospital beds in Catalonia. Cases were all consecutive adult patients (≥18 years old) treated at the participating hospitals in 2018 who met the case definition of CDI: presence of diarrhea or toxic megacolon without any other known cause, and at least one of the following criteria: (a) presence of toxin-producing strain or positive C. difficile toxins A or B in a laboratory stool specimen; and (b) pseudomembranous colitis confirmed by endoscopic, surgical, or histological examination. Oncological patients were defined as patients with a diagnosis of hematological malignancies or solid tumors in the previous 5 years. Colonized and asymptomatic patients (even if they carried a toxin-producing strain), patients with previous CDI episodes, or those admitted to specific convalescence and palliative care units were excluded. Sociodemographic, clinical, and epidemiological variables and evolution at discharge and after 90 days were recorded for each case. Recurrence was defined as a second episode occurring between eight weeks and 90 days after the onset of a previous episode, provided that CDI symptoms from the first episode had resolved. Poor outcome was defined as recurrence or mortality within the 90-day follow-up period.
Community-acquired CDI was defined as CDI starting in the community or within 48 hours of admission, identified in patients with no history of admission to a healthcare facility or who have been discharged >4 weeks prior to symptom onset. Hospital-acquired CDI was defined as CDI identified >48 h after admission and before discharge. The stool sample that identified the CDI must have been collected >48 h after hospital admission. Healthcare-related CDI was defined as CDI starting in the community or within 48 hours of admission identified in patients who have been discharged from a healthcare facility ≤4 weeks prior to symptom onset. Differences between categorical variables were calculated by Pearson χ2 test or Fisher exact test and continuous variables using a Student T-test. Associations of demographic, clinical, and epidemiological variables were calculated by the Mantel–Haenszel test. The 95% confidence intervals (CIs) were calculated. p-Values of <0.05 were considered statistically significant. All reported p-values were two-sided. The analyses were performed using IBM SPSS Statistics 20 software.
Results
During the study period, 682 consecutive CDI cases were identified. The 90-day follow-up visits were achieved in 635 patients (93.1%), who were finally included in the study. Overall, 127 cases out of 635 (20.0%) were oncological patients.
The mortality rate in oncological patients was higher than in the general population: overall (OR = 1.70, 95% CI: 1.08–2.67), in the first 4–7 days after diagnosis (OR = 3.15, 95% CI: 1.29–7.66) and in the long-term follow-up at 31–90 days after diagnosis (OR = 2.47, 95% CI: 1.29–4.72) (Table1).
Table 1.
Demographic data, comorbidities, and clinical outcomes of CDI in oncological patients and the general population.
| Oncological patients n = 127 (%) | General population n = 508 (%) | OR (95% CI) | p-Value | |
|---|---|---|---|---|
| Gender (male) | 76 (59.8%) | 221 (43.5%) | 1.93 (1.30 to 2.88) | p = 0.001 |
| Age, median (SD) | 70.8 (12.9) | 69.9 (17.9) | (−3.61 to 1.87) | p = 0.533 |
| Place of acquisition | ||||
| Community-acquired | 21 (16.5%) | 154 (30.3%) | 0.45 (0.27 to 0.75) | p = 0.002 |
| Hospital-acquired | 51 (40.2%) | 218 (42.9%) | 0.89 (0.60–1.33) | p = 0.574 |
| Healthcare-related | 55 (43.3%) | 136 (26.8%) | 2.09 (1.39 to 3.12) | p < 0.001 |
| Antibiotic exposure* | 91 (71.7%) | 359 (70.7%) | 1.05 (0.68–1.61) | p = 0.827 |
| Diabetes mellitus | 29 (22.8%) | 149 (29.3%) | 0.71 (0.45–1.12) | p = 0.146 |
| HIV infection | 2 (1.6%) | 4 (0.8%) | 2.02 (0.36–11.13) | p = 0.412 |
| Solid organ transplant | 2 (1.6%) | 16 (3.1%) | 0.49 (0.11–2.17) | p = 0.549 |
| Monotherapy** | 90 (76.3%) | 363 (74.8%) | 1.08 (0.67–1.73) | p = 0.748 |
| Vancomycin | 29 (32.2%) | 158 (43.5%) | 0.62 (0.38–1.01) | p = 0.052 |
| Metronidazole | 57 (63.3%) | 204 (56.2%) | 1.35 (0.84–2.17) | p = 0.221 |
| Fidaxomicin | 4 (4.4%) | 1 (0.3%) | 16.84 (1.86 to 152.55) | p = 0.012 |
| Fecal microbiota transplantation | 2 (1.6%) | 4 (0.8%) | 2.02 (0.36–11.13) | p = 0.412 |
| Proton pump inhibitors | 85 (66.9%) | 328 (64.6%) | 1.11 (0.74–1.68) | p = 0.618 |
| Parenteral nutrition | 12 (9.4%) | 33 (6.5%) | 1.50 (0.75–2.99) | p = 0.249 |
| ICU admission*** | 3 (2.4%) | 37 (7.3%) | 0.31 (0.09–1.01) | p = 0.053 |
| Previous surgery**** | 6 (4.7%) | 57 (11.2%) | 0.39 (0.16 to 0.93) | p = 0.034 |
| Megacolon | 3 (2.4%) | 6 (1.2%) | 2.02 (0.49–8.21) | p = 0.323 |
| Outcome | ||||
| Clinical cure | 77 (60.6%) | 368 (72.4%) | 0.59 (0.39 to 0.88) | p = 0.010 |
| Recurrence | 16 (12.6%) | 50 (9.8%) | 1.32 (0.72–2.41) | p = 0.364 |
| Mortality | 34 (26.8%) | 90 (17.7%) | 1.70 (1.08 to 2.67) | p = 0.021 |
| Mortality within 72 h | 3 (2.4%) | 16 (3.1%) | 0.74 (0.21–2.59) | p = 0.778 |
| Mortality within 4–7 days | 9 (7.1%) | 12 (2.4%) | 3.15 (1.29 to 7.66) | p = 0.008 |
| Mortality within 8–30 days | 6 (4.7%) | 34 (6.7%) | 0.69 (0.28–1.68) | p = 0.541 |
| Mortality within 31–90 days | 16 (12.6%) | 28 (5.5%) | 2.47 (1.29 to 4.72) | p = 0.005 |
*Antibiotic exposure in the 30 days before inclusion. ** refereed to specific antibiotic to treat CDI ***ICU: intensive care unit; **** surgery performed in the 30 days before inclusion.
Malignancy itself was not found to be an independent risk factor for recurrent CDI; however, oncological patients undergoing chemotherapy presented a higher recurrence rate than the general population (18.5% vs 9.8%, p = 0.049).
Among oncological patients treated with metronidazole, those receiving active chemotherapy showed a higher risk of recurrence than those not receiving this treatment (35.3% vs 8.0%; OR = 6.273 95% CI: 1.08 to 36.25; p = 0.04). No impact on the risk of poor outcome was identified among cases treated with vancomycin and patients without and with active CT (70.0% vs 68.4%; OR = 1.80, CI: 0.20 to 5.68; p = 0.930). Oncological patients receiving metronidazole showed delayed mortality compared with those treated with vancomycin (mean mortality days after diagnosis: 36.4 days (SD: 32.4) vs 5.0 days (SD: 1.4), respectively, (95%CI: −61.40 to −1.46; p = 0.043).
No risks factors for poor outcomes were identified among oncological patients (Table2).
Table 2.
Risk factors associated with poor outcome in oncological patients with CDI.
| Poor outcome n = 50 (%) | Clinical cure n = 77 (%) | OR (95% CI) | p-Value | |
|---|---|---|---|---|
| Gender (male) | 31 (62.0%) | 45 (58.4%) | 1.16 (0.56–2.40) | p = 0.689 |
| Age, median (SD) | 73.3 (14.2) | 69.1 (11.7) | (−8.71 to 0.45) | p = 0.077 |
| Place of acquisition | ||||
| Community-acquired | 8 (16.0%) | 13 (16.9%) | 0.94 (0.36–2.46) | p = 0.896 |
| Hospital-acquired | 18 (36.0%) | 33 (42.9%) | 0.75 (0.36–1.56) | p = 0.442 |
| Health-care related | 24 (48.0%) | 31 (40.3%) | 1.37 (0.67–2.80) | p = 0.390 |
| Previous antibiotic exposure* | 37 (74.0%) | 54 (70.1%) | 1.21 (0.55–2.69) | p = 0.637 |
| Diabetes mellitus | 9 (18.0%) | 20 (26.0%) | 0.63 (0.26–1.15) | p = 0.298 |
| HIV* infection | 0 | 2 (2.6%) | - | - |
| Solid organ transplant | 0 | 2 (2.6%) | - | - |
| Ongoing chemotherapy | 23 (46.0%) | 31 (40.3%) | 1.26 (0.62–2.59) | p = 0.523 |
| Monotherapy** | 33 (73.3%) | 57 (78.1%) | 0.77 (0.33–1.83) | p = 0.556 |
| Vancomycin | 9 (27.3%) | 20 (35.1%) | 0.69 (0.27–1.78) | p = 0.446 |
| Metronidazole | 23 (69.7%) | 34 (59.6%) | 1.56 (0.62–3.87) | p = 0.342 |
| Fidaxomicin | 1 (3.0%) | 3 (5.3%) | 0.56 (0.06–5.64) | p = 0.625 |
| Fecal microbiota transplantation | 0 (0.0%) | 2 (2.6%) | - | - |
| Proton pump inhibitors | 34 (68.0%) | 51 (66.2%) | 1.08 (0.51–2.31) | p = 0.836 |
| Parenteral nutrition | 5 (10.0%) | 7 (9.1%) | 1.11 (0.33–3.72) | p = 0.864 |
| ICU admission*** | 1 (2.0%) | 2 (2.6%) | 0.76 (0.07–8.67) | p = 0.829 |
| Surgery**** | 2 (4.0%) | 4 (5.2%) | 0.76 (0.13–4.32) | p = 0.757 |
| Megacolon | 1 (2.0%) | 2 (2.6%) | 0.76 (0.07–8.67) | p = 0.829 |
*Antibiotic exposure in the 30 days before inclusion. ** refereed to specific antibiotic to treat CDI ***ICU: intensive care unit; **** surgery performed in the 30 days before inclusion.
Discussion
High rates of unrelated and delayed mortality in CDI patients (up to 40% at 1 year follow-up) have been reported (Shorr et al., 2016). Our study shows that this late mortality is even higher in cancer patients. CDI is a devastating illness, causing long hospitalizations, immobilization, long-term dysbiosis, and potential delay in scheduled chemotherapy cycles (Cózar et al., 2019) (Dutta and Lim, 2020). All these features may explain the higher late mortality in this group.
In our study, oncological patients undergoing chemotherapy had statistically significant higher recurrence rate, and it is widely recognized that chemotherapeutic agents can alter the normal gut microbiota composition by causing mucosal inflammation, decreasing the repair capacity of the mucosal epithelium, or promoting an anaerobic environment, which favor CDI (Dutta and Lim, 2020) (Chung et al., 2016). Probably, the profound immunological and gut changes associated with the use of chemotherapy itself may underlie the increased risk of recurrence.
Metronidazole has been shown to be less effective than vancomycin in observational studies and in randomized clinical trials (McDonald et al., 2018) (Vardakas et al., 2012), especially in the subgroup of patients with highly impaired gut microbiota. These studies have demonstrated that vancomycin is superior to metronidazole in terms of clinical cure, risk of recurrence in mild-moderate episodes, and for mortality in severe episodes. Our results are in line with this previous reported finding. This fact has motivated that in the last guidelines, metronidazole is no longer recommended for treatment of CDI when fidaxomicin or vancomycin is available (Van Prehn et al., 2021). On the other hand, patients treated with metronidazole showed a significant delayed mortality compared with those treated with vancomycin. These results may be due to a selection bias. Vancomycin is the treatment of choice in patients with severe forms of CDI (McDonald et al., 2018), which would explain the earlier mortality, directly caused by the severity of infection. The higher risk of treatment failure associated with the use of metronidazole (McDonald et al., 2018) (Vardakas et al., 2012) may explain the late mortality found in this group.
The main strengths of our study are its prospective multicenter design and the thorough collection of clinical and laboratory data from a large cohort of consecutive patients. The study also has a number of limitations. First, despite the multicenter design, all the participating hospitals were located in Catalonia and the results may not be generalizable to other settings. Second, this was an observational study and treatment decisions were made at the discretion of the treating physician; there may have been a selection bias in relation to severity and use of vancomycin. Third, the use of fidaxomicin and bezlotoxumab is underrepresented in the present cohort and no conclusions can be drawn regarding the role of these treatments in preventing poor outcomes. Fourth, ribotypes of C.difficile isolates were not studied. Finally, the sample size did not allow us to identify strong correlations between clinical factors and the risk of poor outcomes.
Conclusion
In conclusion, patients with hematological malignancies and solid tumors presented a higher risk of poor outcomes after CDI. Their early and late mortality rates were higher than in the general population, and in parallel, those undergoing chemotherapy (especially those receiving metronidazole) had higher rates of recurrence.
Acknowledgments
On behalf the VINCat program (Infection Control and Antimicrobial Stewardship Catalonian Program). VINCat program group: Alex Smithson (Fundació Hospital de l'Esperit Sant), Àngels García (HC Sant Jaume Calella i HC de Blanes), Anna Martinez (Hospital de Campdevànol), Antoni Castro (Hospital Universitari Sant Joan de Reus), David Blancas (Hospital Residència Sant Camil), David Castander (Hospital Sant Pau i Santa Tecla), Esther Moreno (Hospital Residència Sant Camil), Graciano Garcia (Hospital Univ. Joan XXIII de Tarragona), Josep Maria Tricas (Hospital de Mollet), Lourdes Gabarró (Hospital Comarcal de l'Alt Penedès), Ma Rosa Coll (Hospital Universitari Sagrat Cor), Maria de la Roca Toda (Hospital de Palamós), Maria Pilar Barrufet (Hospital de Mataró), Marisa Jofré (Hospital Santa Caterina), Montse Brugues (Hospital d'Igualada), Naiara Villalba (Hospital Comarcal de Sant Bernabé), Rosa Benítez (Hospital Municipal de Badalona), Sonia Tortajada (Centro Médico Teknon, Grupo QuironSalud), Vicens Diaz (Parc Sanitari Sant Joan Déu – HG), and Yolanda Meije (Hospital de Barcelona).
Footnotes
Author contributions: EC, EL, and GCR designed the study. EL obtained the required funding, and EL and EC acquired the necessary resources. SH, NS, LC, JLC JC, MLS, RP, CG, AC, MM, JE, MA, PM, and GCR contributed to the data preparation. EC, SH, and EL performed the data analysis. NS, LC, JLC JC, MLS, RP, CG, AC, MM, JE, MA, PM, and GCR validated the data. SH performed the statistical analysis. EC, SH, and EC wrote the original draft, and NS, LC, JLC JC, MLS, RP, CG, AC, MM, JE, MA, PM, and GCR reviewed and edited all subsequent versions of the manuscript. All authors read and approved the final manuscript.
The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Esther Calbo has accepted grants, speaking engagements, and conference invitations from Astellas, AstraZeneca, Novartis, Pfizer, and MSD. Joaquín Lopez-Contreras has funded by MSD, Pfizer, Actelion, Summit Therapeutics, GSK, Shionogi, Angelini, Marato TV3, and ISCIII to conduct clinical research. Travel support was obtained from Pfizer, MSD, and Guerbet. Education support was obtained from MSD and Pfizer. Advisory board fees from Pfizer, MSD, Hartmann, and Astra-Zeneca. Enric Limón has been a speaker in a symposium organized by MSD. All other authors declare no conflicting interests.
Funding: The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the MSD through Investigator Initiated Studies Program.
Ethical approval: No diagnostic tests were made or samples taken from any participant in addition to those required by routine care. This study complies with the principles of the Declaration of Helsinki and the legal structure according to international human rights and biomedicine and personal data protection legislation. The Ethics Committee of Hospital Universitari de Bellvitge approved the study (ref: PR066/18). All data were treated as confidential, and records were accessed anonymously.
ORCID iDs
Sergi Hernández https://orcid.org/0000-0002-0461-6190
Enric Limón https://orcid.org/0000-0002-5396-1521
References
- Chung MS, Kim J, Kang JO, et al. (2016) Impact of malignancy on Clostridium difficile infection. European Journal of Clinical Microbiology and Infectious Diseases 35(11): 1771–1776. DOI: 10.1007/s10096-016-2725-6 [DOI] [PubMed] [Google Scholar]
- Cornely OA, Mullane KM, Birch T, et al. (2020) exploratory evaluation of bezlotoxumab on outcomes associated with clostridioides difficile infection in modify I/II participants with cancer. Open Forum Infectious Diseases 7(2): ofaa038. DOI: 10.1093/ofid/ofaa038 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Cózar A, Ramos-Martinez A, Merino E, et al. (2019) High delayed mortality after the first episode of Clostridium difficile infection. Anaerobe 57: 93–98. DOI: 10.1016/j.anaerobe.2019.04.004 [DOI] [PubMed] [Google Scholar]
- Delgado A, Reveles IA, Cabello FT, et al. (2017) Poorer outcomes among cancer patients diagnosed with Clostridium difficile infections in United States community hospitals. BMC Infectious Diseases 17(1): 448. DOI: 10.1186/s12879-017-2553-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- Dutta D, Lim SH. (2020) Bidirectional interaction between intestinal microbiome and cancer: opportunities for therapeutic interventions. Biomarker Research 8(1): 31. DOI: 10.1186/s40364-020-00211-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Gupta A, Tariq R, Frank RD, et al. (2017) Trends in the incidence and outcomes of hospitalized cancer patients with clostridium difficile infection: a nationwide analysis. Journal of the National Comprehensive Cancer Network 15(4): 466–472. DOI: 10.6004/jnccn.2017.0046 [DOI] [PubMed] [Google Scholar]
- McDonald LC, Gerding DN, Johnson S, et al. (2018) Clinical practice guidelines for clostridium difficile infection in adults and children: 2017 update by the infectious diseases society of America (IDSA) and society for healthcare epidemiology of America (SHEA). Clinical Infectious Diseases 66(7): e1–e48. DOI: 10.1093/cid/cix1085 [DOI] [PMC free article] [PubMed] [Google Scholar]
- van Prehn J, Reigadas E, Vogelzang EH, et al. (2021) ‘european society of clinical microbiology and infectious diseases: 2021 update on the treatment guidance document for Clostridioides difficile infection in adults’, Clinical Microbiology and Infection, 27, pp. S1–S21. doi: 10.1016/j.cmi.2021.09.038 [DOI] [PubMed] [Google Scholar]
- Shorr AF, Zilberberg MD, Wang L, et al. (2016) mortality and costs in clostridium difficile infection among the elderly in the United States. Infection Control and Hospital Epidemiology 37(11): 1331–1336. DOI: 10.1017/ice.2016.188 [DOI] [PubMed] [Google Scholar]
- Vardakas KZ, Polyzos KA, Patouni K, et al. (2012) Treatment failure and recurrence of Clostridium difficile infection following treatment with vancomycin or metronidazole: a systematic review of the evidence. International Journal of Antimicrobial Agents 40(1): 1–8. DOI: 10.1016/j.ijantimicag.2012.01.004 [DOI] [PubMed] [Google Scholar]