PREDICTIVE FACTORS OF RESPONSE TO IMMUNOTHERAPY

A big issue with the use of immunotherapy for lung cancer, as well as for other tumor types, is that only some patients respond — but how to predict who will?  New research suggests that looking at a patient's immune profile could give some clues.

The findings were presented here at the European Lung Cancer Conference (ELCC) 2017  by Italian researchers reporting on their experience in patients with non-small cell lung cancer (NSCLC) treated with the programmed cell death (PD) inhibitor nivolumab (Opdivo, Bristol-Myers Squibb).

Patients who have high baseline levels of, among others, natural killer (NK) cells or a low neutrophil-to-lymphocyte ratio may be significantly more likely to have a complete or partial response or stable disease with the drug, they found.  

Lead researcher, Marcello Tiseo, MD, PhD, coordinator of DMT Thoracic Oncology, University Hospital of Parma, Italy, said in a release: "The identification of a panel of blood predictive biomarkers would enable the early identification of patients most likely to benefit from anti-PD-1 and anti-PD-L1 [anti-PD-ligand 1] treatment."

Stefan Zimmermann, MD, senior consultant, Medical Oncology Department, HFR–Hôpital Cantonal, Fribourg, Switzerland, commented on the research, saying: "In the current era of precision medicine and increasing healthcare costs, we urgently need proper predictive biomarkers to select patients that will benefit from a specific therapy."

He added: "This study found that baseline levels of certain white blood cells do have a role in predicting response to immunotherapy in patients with lung cancer.

"These new factors should be investigated in future clinical trials, together with tumor PD-L1 expression and other markers that constitute the cancer immunogram to predict whether or not patients will benefit from treatment."

Study Details 

The Italian group reports on a series of 54 consecutively treated patients with NSCLC who received nivolumab 3 mg/kg every 14 days.

The median age of the patients was 69 years, and 45 (83%) were male. Most patients (72%)  had a Eastern Cooperative Oncology Group (ECOG) performance status of 0 to 1, and 92% were current smokers or ex-smokers. There was an even split between squamous (52%) and adenocarcinoma (48%) tumors, and most (89%) patients had stage IV disease.

The researchers examined the patients' peripheral blood immune profiles at baseline and after two and four treatment cycles, with CD3, CD4, CD8, NK (CD56), Treg (FOXP3), and myeloid-derived suppressor cell lymphocyte subpopulations assessed by using fluorescence-activated cell sorting analysis.The results showed that after treatment with nivolumab, 20 patients had a complete or partial response or stable disease, while 34 had progressive disease.

Patients who responded to therapy or had stable disease were significantly more likely than those with progressive disease to have an ECOG performance status of 0, at 76.9% vs 23.1% (P < .001), and to have had their last chemotherapy cycle at least 4 months before assessment, at 61.9% vs 38.1% (P < .004).

Looking at leukocyte phenotypes, the team found that patients who responded to therapy or had stable disease were significantly more likely than those who progressed to have a higher baseline CD3+ count (P = .046) and higher NK cell count (P < .001).

Higher overall lymphocyte counts at baseline had borderline significant association with a response to nivolumab or stable disease (P = .052), while a lower neutrophil count (P = .002) and lower neutrophil-to-lymphocyte ratio (P < .001) were also significantly associated with a response or stable disease following treatment.

Further analysis in 31 patients with data available for all three assessments indicated that significant increases in NK cells and a significant decrease in the proportion of CD4 cells from T0 to T2 were associated with treatment response. Responders also had a greater number and concentration of CD8+ expressing PD-1. 

Exploratory Study 

Discussing the findings, Rolf A. Stahel, MD, Department of Hematology and Oncology, Clinic of Oncology, University Hospital Zurich, Switzerland, cautioned that the study was "exploratory in nature" because it was not conducted "within the framework of a randomized trial." Furthermore, he highlighted that while there was "a lot of analysis," this was "not a multivariate analysis."

Dr Stahel also emphasized that the "curse of multiplicity" has to be borne in mind, as the probability of a false-positive finding increases as the number of subgroup analyses increases. For example, for every 10 analyses conducted, there is a 40% chance that there will be at least one significant association, a figure that rises to 50% with 14 subgroup analyses.

He concluded by noting that this is a "hugely complex area." Because there are already data on the effect of PD-L1 expression and total tumor load on treatment outcomes, he asked, "How are we ever going to explore additional markers and really identify patients who profit from treatment or avoid treatment in patients who do not profit?"

"I think the community still has a huge task to do," he said.

The authors have disclosed no relevant financial relationships . 

European Lung Cancer Conference (ELCC) 2017. Abstract 30PD. Presented May 6, 2017.