ESMO 2017-CHEMOTHERAPY FOR LARGE CELL NSCLC

Patients with specific genomic subtypes of pulmonary large cell neuroendocrine carcinoma (LCNEC) demonstrated significantly improved survival when treated with a chemotherapy regimen designed for non-small cell lung cancer (NSCLC) over chemotherapy for small-cell lung cancer (SCLC), according to findings reported at ESMO 2017, the Annual Congress of the European Society for Medical Oncology in Madrid, Spain. Only patients with LCNEC tumours having RB1 wild-type were seen to benefit equally from NSCLC treatment type.

The debate of which type of chemotherapy provides the most clinical benefit to patients with LCNEC may have been resolved by this analysis of genomic profiling of archival LCNEC samples.

Citing the controversy surrounding whether LCNEC is better treated with NSCLC-type chemotherapy, such as platinum-gemcitabine/taxanes or by the platinum-etoposide therapy used for SCLC, Prof. Anne-Marie Dingemans of the GROW School for Oncology and Developmental Biology, Pulmonology Department, Maastricht University Medical Centre (MUMC) in Maastricht, Netherlands, conducted this retrospective study to evaluate whether recently identified genomic LCNEC subtypes are clinically relevant for a chemotherapy outcome.

Identified genomic subtypes of LCNEC are mutually exclusive

Two mutually exclusive genomic LCNEC subtypes have been identified by molecular studies: one shows co-mutated TP53 and RB1, which is similar to SCLC, and the STK11/KEAP1 subtype with predominantly RB1 wild-type that is similar to NSCLC.

Reviewing clinical data and tumour specimens of 232 patients in the Netherlands Cancer Registry and Pathology Registry from 2003 to 2012, this retrospective analysis identified 148 patients who had been diagnosed with LCNEC. Of these, samples from all patients receiving first-line chemotherapy for panel-consensus diagnosed LCNEC were included for next-generation sequencing (NGS) for the TP53, RB1, STK11, and KEAP1 genes. RB1 (pRB1, 13A10) was analysed by immunohistochemistry, with samples having an H-score of ≥50 considered positive.

The results of the NGS and pRB1 were correlated with overall survival (OS) and progression-free survival (PFS) by Kaplan Meier plots and Log-rank test.

Quality control of the samples determined that 79 samples were sufficient for NGS and that pRB1 could be analysed in 109 samples. RB1 mutations were identified in 37 (47%) samples and loss of pRB1 expression was found in 78 (72%) of samples. It was further determined that mutations in RB1 were mutually exclusive with mutations in the STK11 gene that were identified in 8 samples (p = 0.006).

Survival was longer with NSCLC (platinum-gemcitabine/taxane) than SCLC (platinum-etoposide) chemotherapy in the overall cohort, in patients with wild-type RB, and in patients with tumours expressing pRB1; however, no differences were observed in RB1 mutated LCNEC

The researchers assessed the utility of chemotherapy used to treat both NSCLC and SCLC, but omitted pemetrexed (which is used in some NSCLC cases) from the analysis due to reports of resistance in neuroendocrine carcinomas.

The OS was significantly longer overall in patients treated with chemotherapy for NSCLC than OS in patients receiving SCLC-chemotherapy. In the 15 patients with LCNEC treated with NSCLC-chemotherapy, median OS was 9.6 (range 7.7 to 11.6) months compared to OS of 5.8 (range 5.5 to 6.1) months demonstrated by the 13 patients receiving a SCLC-type chemotherapy regimen (p = 0.026).

Improved Survival Demonstrated with NSCLC-chemotherapy in Pulmonary Large Cell Neuroendocrine Carcinoma with RB1 wild-type.

© Jules Derks.

Patients with LCNEC tumours that expressed pRB1 also had longer OS when treated with a NSCLC chemotherapy regimen; of the patients treated with NSCLC-type chemotherapy, 14 patients with tumour expression of pRB1 had median OS of 9.6 (range 7.4, 11.8) months compared to 1.9 (range 1.7 to 2.1) months in 9 patients with tumours not expressing pRB1 (p = 0.001).

Similarly, PFS was significantly longer in RB1 wild-type patients with NSCLC chemotherapy treatment than with SCLC chemotherapy (p = 0.018). Prolonged PFS also favoured NSCLC over SCLC treatment in patients with pRB1 (p = 0.023).

However, OS and PFS did not significantly differ between NSCLC and SCLC chemotherapy in patients with tumours having RB1 mutation

Conclusions

The authors called for the initiation of prospective studies to support the findings of this retrospective analysis. These findings demonstrated that survival was improved in patients with LCNEC and RB1 wild-type, who were treated with NSCLC- type of chemotherapy as opposed to chemotherapy designed to treat SCLC. However, no differences in clinical outcomes were observed with either chemotherapy regimen in patients with RB1 mutated LCNEC.

These data also suggest that genomic profiling may aid in informing treatment decisions for patients with LCNEC.

Mauro Cives who discussed the study results said that hypothesis has not been met as there is no correlation between SCLC-like genomics and benefit from SCLC treatment and outcomes in patients with NSCLC-like tumours mirror those in unselected population.