Further evidence to support judicious use of antibiotics in patients with cancer

Cancer immunotherapy centered on blockade of the programmed death receptor (PD1) and ligand (PDL1) has become an essential therapeutic approach for many advanced cancers. A chorus of research reports have focused on tumor-intrinsic factors associated with response to therapy, such as expression of interferon-γ associated genes including PDL1 or tumor mutational burden, as well as mechanisms of resistance such as genomic deletion of class I major histocompatibility complex [1]. Other factors impacting systemic immunity and immunotherapy efficacy are additionally of growing interest including patient germline genetics and environmental or microbiota based factors.

The role of the microbiota, particularly the fecal microbiome, in human disease is increasingly being recognized to play a major etiologic and treatment modification role. In diseases ranging from neurologic and endocrine to autoimmune or cancer a growing number reports are detailing an essential role. The human fecal microbiome is the largest microbiome reservoir in the body and is naturally transferred vertically from mother to fetus at birth. While commonly thought of in terms of bacterial populations, it is important to note that this additionally includes archaea, viruses, fungi and meiobenthos such as protozoa and helminths. This diverse ecosystem is exquisitely sensitive to many factors of daily human life but particularly the intake of medications [2]. A growing literature describing the microbiome as an etiologic agent in driving cancer growth is emerging with multiple reports suggesting an impact on treatment outcomes for chemotherapy [3, 4] as well as immunotherapy [5].

Mouse models and human studies suggest that modulation of the fecal microbiome has major impact on outcomes of cancer immunotherapy both regarding toxicity and efficacy. In hematologic malignancies, reports from patients who have undergone allogenic bone marrow transplant suggest that the diversity of the fecal microbiome at baseline is associated with relapse following treatment while antibiotic use during the transplant course is associated with increased frequency of graft versus host disease and inferior overall survival [6, 7]. Along similar rationale, multiple groups have detailed associations of lower microbial diversity and the presence of certain bacteria in association with efficacy and toxicity from the anti-CTLA4 antibody ipilimumab in melanoma [8, 9]. More recently, a series of studies from patients with advanced melanoma, renal cell carcinoma (RCC) or non-small-cell lung cancer (NSCLC) has suggested associations of distinct bacterial populations with improved efficacy of PD1/L1 blockade [10–12].

Within one of these reports investigating PD1 outcomes, the group from Gustave Roussy in Paris made an initial observation for a deleterious impact of antibiotic use on outcomes for PD1 based immunotherapy in patients with RCC, NSCLC and urothelial bladder cancer. In the current issue of Annals of Oncology, Derosa et al. expand upon this observation reporting on a more robust sample now including patients treated in Europe and the United States [13]. In cohorts of patients including 121 RCC and 239 NSCLC they observed antibiotic use within 30 days of first dose of checkpoint blockade in 13% and 20%, respectively. The outcomes of these patients were inferior to the non-antibiotic treated patients with multivariate analysis suggesting a significantly worse progression-free survival in RCC and overall survival in NSCLC. Interestingly this effect did not extend to antibiotic usage beyond 3 months as there was no effect in the 3–6 month pre-treatment time window. This work was retrospective in nature and results need to be taken with caution. That being said, the results are conceptually striking and consistent with prior basic and translational research.

The implications of this work raise many immediately pressing questions. A short non-exclusive list of such questions might include the following. Are certain antibiotics potentially more immunosuppressive than others? What is the mechanism whereby the microbiome communicates to the tumor microenvironment? Can we supplement the microbiome with a probiotic or perhaps identify circulating factors made by a healthy microbiome to promote antitumor immunity? What about the impact of other antibiotics such as antiviral or antifungal agents? In addition to efficacy, does antibiosis impact on toxicity, either in terms of incidence or level/length of immunosuppression required to manage immune-related adverse events?

Fortunately, and excitingly, several groups have demonstrated the feasibility of microbiome transfer from human patients into murine systems and replicated the clinical phenotype in the mice [10–12]. This approach may facilitate the ability to study some modulation of the microbiome via in vivo preclinical models that can then be reverse translated to supplement current treatment paradigms. Conversely an important point to raise may also be the identification of tumor-resident bacteria in pancreas and colon cancers that may limit the efficacy chemotherapy [14, 15] and immunotherapy [16] as well as likely other tumors or treatment modalities [17]. In this setting then depletion of pathogenic bacteria will be a priority in addition to promotion of a healthy commensal environment.

As quickly as cancer immunotherapy has arrived to change standards of care across tumor types, our understanding of the complexity of the tumor microenvironment and perhaps equally important the systemic host response is rapidly changing. The importance of the fecal microbiome in cancer is clearly an emerging area requiring a major focus. Multiple prospective interventional trials are on-going to investigate microbiome modulation in conjunction with standard therapies and it should be emphasized that collection of at least baseline fecal microbiome biomarker samples in clinical trials will be important moving forward. Clinically, it must be emphasized that the cancer community is only at the beginning of our understanding of the microbiome and no evidence based clinical recommendations can yet be given to patients surrounding the use of probiotics. Limiting the use of antibiotics only where necessary should be a priority, however, given the data of Derosa et al., the precarious emergence of drug-resistant pathogens and general medical best practices.

Funding

JJL is supported by Department of Defense Career Development Award (W81XWH-17-1-0265), the National Cancer Institute (P30CA014599-43S), Team Science Award from the Prostate Cancer Foundation (16CHAL12), the Arthur J Schreiner Family Melanoma Research Fund, J. Edward Mahoney Foundation Research Fund and Brush Family Immunotherapy Fund as well as support from Center for Research Informatics of The University of Chicago Biological Science Division and The Institute for Translational Medicine/CTSA (NIH UL1 RR024999).

Disclosure

Consultancies to: 7 Hills, Actym, Amgen, Array, AstraZeneca, BeneVir, Bristol-Myers Squibb, Castle, CheckMate, Compugen, EMD Serono, Gilead, Janssen, Merck, NewLink, Nimbus, Novartis, Palleon, RefleXion, Syndax, Tempest, WntRx with research support to from AbbVie, Array, Boston Biomedical, Bristol-Myers Squibb, Celldex, CheckMate, Corvus, Delcath, Five Prime, Genentech, Immunocore, Incyte, MedImmune, Macrogenics, Novartis, Pharmacyclics, Palleon, Merck, Tesaro, Xencor and travel reimbursement from Amgen, Array, AstraZeneca, BeneVir, Bristol-Myers Squibb, Castle, CheckMate, EMD Serono, Gilead, Janssen, Merck, NewLink, Novartis, RefleXion. JJL is a co-inventor on a patent submitted by the University of Chicago covering use of microbiota to improve cancer immunotherapy outcomes. PS: Consulting: Genentech, Aveo, Eisai, Roche, Pfizer, Novartis, Exelixis, Ipsen, BMS, Astellas; Honoraria: Genentech.