Hyperprogressive disease determinants upon immunecheckpoint blockade in non-small cell lung cancer

Immunecheckpoint inhibitors that target programmed cell death/programmed cell death ligand-1 (PD-1/PD-L1) axis (ICI) have significantly shifted the treatment paradigm in advanced non-small cell lung cancer (NSCLC), however an acceleration of tumor growth upon ICI, known as hyperprogressive disease (HPD), has been reported in 14%-26% of NSCLC pretreated patients, suggesting the existence of an ICI related “dark-side effect”. HPD has never been assessed in patients with NSCLC treated with ICI in 1^st line and, up to now, no clear validated biomarker exists to pre-identify NSCLC patients at high risk of HPD upon single-agent ICI (SA-ICI), who could deserve escalating treatment strategies, such as platinum-based chemotherapy (PCT) and ICI combinations. To better characterize HPD determinants we performed several translational analyses on PBMCs, plasma and tissue samples of NSCLC patients treated with single agent ICI. To define the effect of ICI-PCT treatment on HPD determinants, we performed similar analysis in patients receiving 1st line ICI+PCT. From clinical data, we found a correlation between HPD and derived neutrophil lymphocyte ratio (dNLR), therefore we hypothesized that circulating neutrophils could be a predictor of HPD. From flow cytometry data, we identified >29.6% of circulating immature LDNs, where immaturity was defined as loss of CD10, as a biomarker of HPD in patients treated with single-agent ICI. We also found that the addition of platinum-based chemotherapy was able to prevent HPD in patients defined at high risk, potentially by reducing CD10^-LDNs level. We next explored both in vitro and from blood and tissue samples of patients with NSCLC, the effect on T-cells and the life-span of LDNs subsets. Although we were not able to dissect the exact mechanism, we found that elevated CD10^-LDNs are associated with downregulation of costimulatory CD28 signal on T-cells, which acquired a senescent phenotype. Furthermore, we found that CD10^-LDNs could be a subset of highlander neutrophils, able to survive up to 5 days in vitro. This prolonged life span was coupled with increased sensitivity to cisplatin induce cell death. .Altogether these data indicate that in HPD a opposite maturation drift occurs in circulating neutrophils compared to T-cells. In particular, immature neutrophils and senescent T-cells characterize patients at high risk of HPD, who should therefore be addressed to 1^st line ICI-PCT combinations. We subsequently modeled HPD in vivo, using immunocompetent Lewis Lung Carcinoma tumor bearing mice. The in vivo results, showed that HPD in mice could occur also upon ICI + cisplatin combinations. HPD in mice was associated with reduce expression of Ox40 by T-regulatory cells (Tregs), suggesting a potential switch of Tregs in CD4 T-cells with an IL-17 secreting phenotype. Furthermore, anti-PD-1 treatment significantly increased circulating granulocytes with an immature phenotype (cKIT⁺) (although no difference according to HPD status was observed), suggesting a potential crosstalk between adaptive and innate immunity with increased IL-17 mediated emergency granulopoiesis. Although no difference according to HPD status was found, cisplatin chemotherapy was able to avoid the increase of circulating immature granulocytic subsets. The identification of HPD biomarker is of great importance in NSCLC, where immunotherapy has become one of the most relevant and expensive treatment in the metastatic setting. In Italy, only in 2022 ~43.900 new diagnoses of NSCLC have been reported; considering that HPD, could occur in approximately 15% of cases, identifying these patients early and addressing them to more effective treatment combinations could potentially save hundreds of lives.