Microbiome¶

Confidently known¶

mouse Gut microbiota modulate ICI efficacy in mouse models (Sivan 2015 Bifidobacterium¹; Vétizou 2015 Bacteroides/CTLA-4²). Gnotobiotic/vendor-swap designs, cohousing controls, and FMT demonstrate causality in mice.
human Antibiotic exposure around ICI initiation is associated with worse outcomes (Routy 2018⁴). The antibiotic-harm signal has replicated across tumor types and is the most consistently reproducible microbiome observation in this field. Clinically actionable: avoid non-essential antibiotics in the weeks surrounding ICI start.
human Gut microbiota composition differs between responders and non-responders in multiple independent human cohorts (Gopalakrishnan 2018³, Routy 2018⁴, Matson 2018). The existence of an association is well established.

Study	N	Feature	Effect	95% CI / p	Method
Routy 2017	n=249 (NSCLC, RCC, urothelial on anti-PD-1/PD-L1)	peri-ICI antibiotic exposure / A. muciniphila abundance	OS HR antibiotics → shorter PFS/OS (NSCLC mOS ~8 vs ~20 mo; HR ~3.5)	—	clinical-record audit + shotgun metagenomics
Gopalakrishnan 2017	n=112 (melanoma on anti-PD-1; shotgun-metagenomics subset n=43 (30 R, 13 NR))	gut alpha diversity / Faecalibacterium enrichment	response association higher diversity & Faecalibacterium in responders	P<0.01	16S / shotgun metagenomics + responder→GF-mouse FMT

human mouse Specific responder-associated taxa do not replicate across geographies. Gopalakrishnan highlighted Faecalibacterium / Ruminococcaceae; Routy highlighted Akkermansia muciniphila; Matson highlighted Bifidobacterium longum. Meta-analyses (Lee 2022, McCulloch 2022 Nat Med) found no single taxon reliably predicts anti-PD-1 response across cohorts. Signatures are cohort-specific.
human Akkermansia relationship is non-monotonic. Derosa 2022 Nat Med showed patients with very high A. muciniphila relative abundance had worse outcomes than those with intermediate levels. A simple "more Akkermansia = better" reading is wrong.
human mouse Bacteroidales sign-flipped across checkpoint targets. Vétizou 2015 highlighted Bacteroides fragilis as beneficial for anti-CTLA-4²; Gopalakrishnan 2018 reported Bacteroidales enrichment as associated with anti-PD-1 non-response³. The discordance is not fully reconciled.
mouse Bifidobacterium's dramatic preclinical effect (Sivan 2015¹) has not cleanly translated to humans. The mouse vendor-effect mechanism may be too context-dependent to survive the heterogeneity of human microbiomes.

Study	N	Feature	Effect	95% CI / p	Method
Gopalakrishnan 2017	n=112 (melanoma on anti-PD-1; shotgun-metagenomics subset n=43 (30 R, 13 NR))	gut alpha diversity / Faecalibacterium enrichment	response association higher diversity & Faecalibacterium in responders	P<0.01	16S / shotgun metagenomics + responder→GF-mouse FMT

human mouse No strong new mechanism nominations in this window. The current state is that the microbiome does modulate ICI biology, but which specific taxa, how, and how to manipulate them clinically remain open.

No human-study citations in this section.

human FMT from anti-PD-1 responders can rescue anti-PD-1-refractory melanoma. Baruch 2021 (phase I, 3/10 objective responses)⁵ and Davar 2021 (phase II, 6/15 clinical benefit including ~20% ORR)⁶ converge independently. This is the most surprising and actionable microbiome finding to date — it implies that some proportion of anti-PD-1 refractoriness reflects a reversible host-microbiome state rather than tumor-intrinsic escape.
human The Davar study showed a mechanistic signature: reduced intratumoral IL-8+ myeloid cells, increased CD8+ T cell activation, consistent with microbiome-driven TME reprogramming. This is the first plausible mechanistic link between gut state and tumor immune state beyond correlative signatures.

Study	N	Feature	Effect	95% CI / p	Method
Baruch 2020	n=10 (phase I FMT in anti-PD-1-refractory metastatic melanoma)	FMT from anti-PD-1-responder donors + anti-PD-1 reinduction	ORR 3/10 (1 CR, 2 PR)	—	phase I FMT trial
Davar 2021	n=15 (single-arm phase II FMT in anti-PD-1-refractory melanoma)	responder-FMT + pembrolizumab	clinical benefit / ORR 6/15 benefit; ORR ~20%	—	phase II single-arm trial

Antibiotics: avoid unnecessary antibiotic exposure in the 1–2 months around ICI initiation. This is the most robust clinical implication of the microbiome literature.
Probiotics / specific supplementation: not evidence-based as ICI adjuncts outside a clinical trial.
Diet: high-fiber diet shows associations with better ICI outcomes in observational data but is not rigorously established as causal in humans.
FMT: investigational. 20–30% response rates in small refractory trials are promising enough to support ongoing trials but not standard of care.