Osimertinib Less Effective for Treatment of Complex EGFR Mutation in NSCLC

July 10, 2020

Osimertinib may not be as effective for the treatment of with non–small cell lung cancer who harbor complex EGFR mutations as it is for those who harbor traditional EGFR mutations, according to a recent study.

Progression-free survival (PFS) and overall survival (OS) may be negatively impacted with the administration of osimertinib (Tagrisso) in patients with non–small cell lung cancer (NSCLC) harboring complex EGFR mutations, and patients whose prior treatment duration with EGFR tyrosine kinase inhibitors (TKI) was shorter than usual, according to a study that assessed sequenced cancer tissues by RNA reverse transcription polymerase chain reaction ,published in Lung Cancer.

“Osimertinib is currently the standard of care for [patients with] NSCLC with secondary EGFR T790M mutation after frontline EGFR TKI treatment failure. However, a subset of patients’ cancer still responds poorly to osimertinib,” Jin-Yuan Shih, MD, told Targeted Oncology. “In this study, we found advanced [patients with] NSCLC with complex EGFR mutations and secondary T790M had lower osimertinib response rate, shorter osimertinib PFS, and shorter OS than patients with a single EGFR mutation and secondary T790M. Whether osimertinib is still the drug of choice for these patients becomes questionable.”

Prior to this analysis, no study had evaluated the efficacy of osimertinib in patients with complex EGFR mutations with secondary T790M. But following the success of the phase II AURA2 trial (NCT02094261), the serial AURA trials utilized the Roche cobas EGFR Mutation Test V2 (Cobas V2) assay, which was suspected to have missed complex EGFR mutations, rationalizing the need for this assessment.

“Complex EGFR mutations cannot be detected by hotspot mutation analyses for EGFR, such as the authority-approved Cobas V2. Complex mutations would be misclassified as its partner common mutation solely. For instance, L858R-positive, Q791H-positive, [and] T790M in this study found by direct sequencing would be reported as “L858R-positive” T790M by Cobas V2,” Shih added.

“Since most complex EGFR mutations co-occur with L858R rather than exon 19 deletion, it might be one of the reasons that deletion 19 with T790M seemed to respond to osimertinib better than L858R with T790M in the AURA extension and AURA2 [trials], while the 2 studies used Cobas V2 to detect EGFR mutations. To identify complex EGFR mutations with T790M, a hotspot mutation analysis is not enough. Next-generation sequencing or direct sequencing of the EGFR should be used.”

The study included 165 patients with advanced NSCLC and secondary T790M mutation who were treated with osimertinib. Patient characteristics recorded at baseline showed that 51.0% of subjects received a minimum of 1 EGFR TKI before osimertinib, 62% of whom received prior gefitinib (Iressa), another 67% received erlotinib (Tarceva), and 29% received afatinib (Gilotrif). The remaining 10% of patients were previously treated with a novel EGFR TKI, including rociletinib (CO-1686) for 1, nazartinib (EGF816) for 7, olmutinib (HM61713) for 2, HS-10296 for 4, and AZD3759 for 3. Prior EGFR TKIs were administered for a median duration of 18.8 months at a median follow-up time of 15.8 months. Between the prior treatments, no differences in median duration or follow-up time was observed, even across mutational groups.

The population of patients in this study were predominantly male (36%) with a median age of 62 years (range, 53-71). Eighteen percent had an ECOG performance status greater than 2 prior to receiving osimertinib in the study.

A total of 162 patients were included in the final analysis, 11 of whom were classified as having an EGFR mutation and secondary T790M. Of the patients with rare EGFR mutations, 1 had an exon 19 deletion with A755G and the remaining 10 patients had a complex EGFR mutation, including an L859R mutation. The response rate for the 11 patients was 27% and the disease control rate (DCR) was 54%.

In the overall study population, 73% of patients experienced disease progression or death; the median PFS observed with osimertinib was 9.1 months (95% CI, 7.6-10.6). Among the subgroup of patients with complex EGFR mutations, the median PFS was notably shorter at 2.9 months (95% CI, 0.3-5.6) compared with those who had a single EGFR mutation at 9.7 months (95% CI, 7.6-11.8), which resulted in a P value of less than .001.

Looking deeper into mutational subgroups, subjects who had an EGFR exon 19 deletion had a median PFS of 9.3 months (95% CI, 6.6-12.0), those with an L858R mutation had a median PFS of 10.1 months (95% CI, 6.1-14.1), and patients with complex EGFR mutations had a median PFS of 2.9 months (95% CI, 0.3-5.6). The difference between these 3 subgroups was P <.001. According to a log rank test, PFS was also shorter among patients who received prior EGFR TKI therapy for shorter durations. Median PFS for patients with shorter prior EGFR TKI therapy was 7.3 months (95% CI, 5.3-9.4) compared with 13.8 months (95% CI, 8.9-18.6) for patients who had longer periods of EGFR TKI treatment (P <.001).

The hazard ratio (HR) for shorter PFS in patients with complex EGFR mutations was calculated as 3.06 (95% CI, 1.60-5.84; P = .001), and for shorter PFS among individuals with shorted prior EGFR TKI treatment was 1.71 (95% CI, 1.18-2.49; P = .005), according to a Multivariate Cox analysis. For another subgroup of patients who had brain metastasis prior to receiving osimertinib, the HR was 1.48 (95% CI, 1.02-2.17; P = .042). These were all factors determined to be independently significant for PFS in patient treated with osimertinib. In addition, a shorter PFS was observed with patients who received a novel EGFR TKI prior to osimertinib (HR, 1.70; 95% CI, 0.98-2.94; P = .060).

Treatment failure occurred in 116 patients who stopped osimertinib during the study. The median time to treatment failure (TTF) was 11.4 months (95% CI, 7.6-15.2) overall. In the complex EGFR mutation subgroup, the median TTF was 6.3 months (95% CI, 3.2-9.4) compared with patients who had a single mutation at 13.1 months (95% CI, 7.7-18.5), with a difference of P <.001. For patients with an EGFR exon 19 mutation, the median TTF was 11.3 months (95% CI, 7.6-15.0), compared with 17.6 months (95% CI, 9.9-25.5) among patients with an L8558R mutation, and 6.3 months (95% CI, 3.2-9.4) for patients with a complex EGFR mutation (P = .001). Similar to the results observed with PFS, TTF was shorter among patients with shorter duration of prior EGFR TKI therapy of 8.5 months (95% CI, 6.8-10.2) compared with 22.4 months (95% CI, 14.5-30.3) among patients who had longer periods of prior EGFR TKI treatment (P <.001).

Following treatment with osimertinib in the study, the median OS in the overall population was 29.1 months (95% CI, 21.3-36.8). Forty-six percent of patients died. Among those with complex EGFR mutations, the median OS was shorter at 17.8 months (95% CI, 7.4-28.2), versus the single EGFR mutation group 31.0 months (95% CI, 26.0-36.0). The difference between the groups was significant (P = .01).

The median OS observed in patients with EGFR exon 19 mutation, specifically, was 36.5 months (95% CI, 22.9-50.1), versus 27.7 months (95% CI, 20.3-35.0) in the L858R mutational subgroup, and 17.8 months (95% CI, 7.4-28.2) in the complex EGFR mutational subgroup. Patients who had shorter duration of prior EGFR TKIs had a median OS of 21.5 months (95% CI, 15.3-27.8) compared with those who had longer EGFR TKIs therapy for whom the median OS was 36.7 months (95% CI, 25.5-47.9; P = .003).

Overall, complex EGFR mutations led to shorter OS than that seen in the overall population (HR, 2.17; 95% CI, 25.5-47.9; P = .003), according to the Multivariate Cox regression analysis. Also, shorter EGFR TKI treatment before osimertinib showed shorter OS (HR, 1.87; 95% CI, 1.16-3.02; P = .01). These factors were independent determinants of OS, just as observed with PFS.

“Rare (or uncommon) EGFR mutations account for around 10% to 15% of all EGFR mutations. Their responsiveness to EGFR TKIs is heterogeneous. Except for EGFR exon 20 insertions, in general, EGFR TKIs are still active against most of the rare EGFR mutations. However, their effectiveness against rare mutations may not be as good as the effectiveness against common EGFR mutations,” said Shih. “For first-line treatment for [patients with] advanced NSCLC with rare EGFR mutations, afatinib may be superior to first-generation TKIs gefitinib and erlotinib. For rare EGFR mutations with secondary T790M, it is important to remember that osimertinib has not been tested in a randomized controlled trial. Considering its few adverse effects, we might still consider osimertinib for these patients. However, it is prudent to monitor the tumor response and disease activity closely.”

With low response to osimertinib observed with complex EGFR mutation in this study, as well as shorted PFS and OS, there may be a need to alter treatment patterns in everyday practice. Shih stated: “In general, osimertinib is an ideal TKI, with few adverse effects and a very potent anticancer activity against NSCLC with activating EGFR mutations. If a cancer’s response to osimertinib is unexpectedly poor, consider checking the method for EGFR mutation analysis. Hot-spot analyses are sensitive and cost-effective methods to detect EGFR mutations for advanced NSCLC. However, they are insensitive for rare EGFR mutations. Osimertinib has been reported to be less effective against several rare mutations, in both the first-line setting (in the KCSG-LU15-09 trial) and second-line treatment for secondary T790M (in this study). In such [circumstances], a next-generation sequencing study or sequencing of EGFR for the cancer tissue may be necessary.”


Lin YT, Tsai TH, Wu SG, et al. Complex EGFR mutations with secondary T790M mutation confer shorter osimertinib progression-free survival and overall survival in advanced non-small cell lung cancer. Lung Cancer. 2020;145:1-9. doi:10.1016/j.lungcan.2020.04.022