Ipilimumab improves clinical outcomes when combined with nivolumab in early-stage, operable non-small cell lung cancer. In a first for a neoadjuvant trial, the study also found that responsiveness to the dual checkpoint blockade strategy may be influenced by a patient's gut microbiome.

Two immune checkpoint inhibitors (ICI) are better than one at eliminating tumor cells ahead of surgery for patients with early-stage non–small cell lung cancer (NSCLC)—and gut bacteria seem to play a major role in shaping the likelihood of achieving a major pathologic response (MPR).

That's according to initial results from the NEOSTAR trial, the first randomized comparison of immunotherapy regimens in patients with operable NSCLC (Nat Med 2021 Feb 18 [Epub ahead of print]). The benefit of adding the CTLA4 inhibitor ipilimumab (Yervoy; Bristol Myers Squibb) to nivolumab (Opdivo; Bristol Myers Squibb), a PD-1 inhibitor, came at the expense of somewhat greater toxicity, as well as a longer wait before surgery. However, study authors identified a putative biomarker—the presence of certain bacteria in stool samples—that could help clinicians identify the best candidates for the dual ICI strategy.

“That's potentially very interesting,” says Patrick Forde, MD, of Johns Hopkins University in Baltimore, MD, who was not involved in the research. “The bacteria they have identified do appear to predict efficacy in this population.”

The study is the first to show that intestinal microbiota, known mediators of tumor response to checkpoint blockade in metastatic disease, can affect neoadjuvant outcomes as well.

In this part of NEOSTAR, 44 patients were randomly assigned to receive nivolumab alone or with ipilimumab, followed by surgery. Those who received dual therapy were more likely to have an MPR, defined as 10% or fewer viable cancer cells in the resected tumor: The MPR rate was 38%, compared with 22% among those who received nivolumab alone. The rates of pathologic complete response—no viable cancer cells detected—were 29% and 9%, respectively.

Exploratory analyses of resected tumor specimens showed that the ICI combination enhanced tumor immune infiltration and immunologic memory, compared with monotherapy. “Interestingly, we see this increased immune infiltrate after combination checkpoint blockade despite having administered ipilimumab only once to patients and at low dose,” says lead investigator Tina Cascone, MD, PhD, of The University of Texas MD Anderson Cancer Center in Houston.

Pretreatment tumor PD-L1 expression levels tended to be higher among responders, consistent with the results of another small neoadjuvant study, led by Forde, evaluating the same combination in NSCLC (Journal for ImmunoTherapy of Cancer 2020;8:e001282). Additionally, Cascone and her colleagues found discriminative fecal microbiome signatures associated with pathologic response: In stool samples collected before treatment, a relative abundance of certain bacteria was linked to greater likelihood of MPR following combination checkpoint inhibition, as well as reduced toxicity and favorable T-cell profiles. Other bacteria were indicative of poor outcomes.

The results warrant further study to interrogate mechanisms between microbial signatures and favorable T-cell profiles in a neoadjuvant setting. But for now, they suggest that “the gut microbiome is a pretty darn good biomarker,” says co-author Jennifer Wargo, MD, of MD Anderson.

More work is also needed to better define functional indicators of a favorable gut ecosystem, notes co-author Nadim Ajami, PhD, also of MD Anderson. “It's more about the function than the composition,” he says.

At least one other study has documented an association between microbiome metrics and the success of neoadjuvant immunotherapy. A team led by Georgina Long, PhD, MBBS, from the Melanoma Institute Australia in Sydney, found that patients with stage III melanoma given preoperative nivolumab and ipilimumab were more likely to respond and less likely to experience immune-related side effects if they had diverse gut microbes (Cancer Res 2020;80:5734). “It's really an ecosystem effect,” Long says.

Any of these putative microbiome biomarkers still require independent validation. Unlike other response-associated signatures, such as levels of circulating IFNγ and tumor mutation burden, which show up repeatedly across a range of malignancies, reports of potential bacterial biomarkers in neoadjuvant immunotherapy trials are few, notes Christian Blank, MD, of the Netherlands Cancer Institute in Amsterdam. “At the moment, I am very careful with microbiome correlations,” he says. –Elie Dolgin