A Randomized Phase II Trial Comparing the Efficacy and Safety of Pioglitazone, Clarithromycin and Metronomic Low-Dose Chemotherapy with Single-Agent Nivolumab Therapy in Patients with Advanced Non-small Cell Lung Cancer Treated in Second or Further Line (ModuLung).

ABSTRACT

Background: Most non-small cell lung cancers occur in elderly and frequently comorbid patients. Therefore, it is necessary to evaluate the efficacy of biomodulatory active therapy regimen, concertedly interfering with tumor-associated homeostatic pathways to achieve tumor control paralleled by modest toxicity profiles. Patients and Methods: The ModuLung trial is a national, multicentre, prospective, open-label, randomized phase II trial in patients with histologically confirmed stage IIIB/IV squamous (n = 11) and non-squamous non-small cell (n = 26) lung cancer who failed first-line platinum-based chemotherapy. Patients were randomly assigned on a 11 ratio to the biomodulatory or control group, treated with nivolumab. Patients randomized to the biomodulatory group received an all-oral therapy consisting of treosulfan 250 mg twice daily, pioglitazone 45 mg once daily, clarithromycin 250 mg twice daily, until disease progression or unacceptable toxicity. Results: The study had to be closed pre-maturely due to approval of immune checkpoint inhibitors (ICi) in first-line treatment. Thirty-seven patients, available for analysis, were treated in second to forth-line. Progression-free survival (PFS) was significantly inferior for biomodulation (N = 20) vs. nivolumab (N = 17) with a median PFS (95% confidence interval) of 1.4 (1.2-2.0) months vs. 1.6 (1.4-6.2), respectively; with a hazard ratio (95% confidence interval) of 1.908 [0.962; 3.788]; p = 0.0483. Objective response rate was 11.8% with nivolumab vs. 5% with biomodulation, median follow-up 8.25 months. The frequency of grade 3-5 treatment related adverse events was 29% with nivolumab and 10% with biomodulation. Overall survival (OS), the secondary endpoint, was comparable in both treatment arms; biomodulation with a median OS (95% confidence interval) of 9.4 (6.0-33.0) months vs. nivolumab 6.9 (4.6-24.0), respectively; hazard ratio (95% confidence interval) of 0.733 [0.334; 1.610]; p = 0.4368. Seventy-five percent of patients in the biomodulation arm received rescue therapy with checkpoint inhibitors. Conclusions: This trial shows that the biomodulatory therapy was inferior to nivolumab on PFS. However, the fact that OS was similar between groups gives rise to the hypothesis that the well-tolerable biomodulatory therapy may prime tumor tissues for efficacious checkpoint inhibitor therapy, even in very advanced treatment lines where poor response to ICi might be expected with increasing line of therapy.