Phase II, noncomparative, open label, multicenter, study of osimertinib, in patients with locally advanced or metastatic EGFR mutated, T790M undetectable or unknown non-small cell lung cancer (Stage IIIB-IV) after no immediate prior EGFR TKI (OSIRIS study).

e21116 Background: Osimertinib (OSI) is a potent irreversible EGFR TKI approved for 1st line therapy advanced EGFR+ NSCLC and for 2nd line T790M+ pts. The AURA trial showed promising results even in pts with T790M- NSCLC after no immediate prior OSI in locally advanced or metastatic EGFR+ NSCLC, T790M undetectable or unknown, after 1st line EGFR TKI and subsequent chemotherapy. Methods: This phase II trial was performed to investigate the role OSI in locally advanced or metastatic EGFR+ NSCLC, T790M undetectable or unknown, after 1st line EGFR TKI (1 or 2 generation) and subsequent chemotherapy. Eligible pts (M or F, > 18 years, ECOG 0-2) received OSI (80 mg/day) until disease progression or unacceptable toxicity. Objective response rate (ORR) was the primary endopoint. Assuming a 10% attrition rate, 90 pts were planned to be enrolled according to the Optimal Simon’s 2 stage design. In the first stage, 32 pts were planned to be accrued, and if ≤ 3 responses were observed the study would be stopped. Otherwise, 49 additional pts were planned to be accrued. The null hypothesis would have been rejected if ≥ 12 responses were observed. This design yields a type I error rate of 0.05 and 80% power. Results: From May 2017 to October 2020, a total of 54 pts were enrolled (17 M and 37 F, mean age 66 years). The study was stopped early due to an extremely slow enrolment rate. However, the ORR of 31.5% (95% CI 19.5% - 45.5%) was significantly higher than the null hypothesis of 9% (p<0.0001). 17 pts obtained a partial response and 20 a stable disease with an overall disease control rate of 68.5%. Median PFS was 9 mos and median OS was 15 mos. Forty-one pts experienced at least 1 adverse event (51.8% treatment related), more than 90% Grade 1 or 2, the most common being diarrhea. Conclusions: Despite early termination and incomplete recruitment, the treatment with OSI in pts with undetectable or unknown T790M showed a significant ORR and a PFS in line with results in T790M+. Currently, OSI represents the preferred option in naïve pts with EGFR+ NSCLC, regardless of T790M status and for pts who progress the recommended subsequent therapies include local therapy, continuing OSI or chemotherapy; in this new scenario, our results confirm that OSI rechallenge in subsequent line after chemotherapy should be explored. This research was conducted with support from AstraZeneca. Clinical trial information: 2016-002555-17.