Up-front treatment with osimertinib reduced the risk of brain progression-free survival but provided a comparable overall survival benefit compared with sequential treatment with gefitinib followed by osimertinib in patients with advanced non–small cell lung cancer harboring EGFR mutations.
Up-front treatment with osimertinib (Tagrisso) reduced the risk of brain progression-free survival (PFS) but provided a comparable overall survival (OS) benefit compared with sequential treatment with gefitinib (Iressa) followed by osimertinib in patients with advanced non–small cell lung cancer (NSCLC) harboring EGFR mutations, according to data from the final OS analysis of the phase 2 APPLE trial (NCT02856893).1
Findings presented at the 2023 European Lung Cancer Congress showed that at a median follow-up of 50 months, evaluable patients treated with up-front osimertinib (arm A; n = 53) experienced a median OS that was not yet reached (NR; 95% CI, NR-NR) compared with 42.8 months (95% CI, 28.58-NR) in those treated with gefitinib followed by osimertinib (arms B and C; n = 103; HR, 1.01; 90% CI, 0.61-1.68). The 18-month OS rate was 84.4% (95% CI, 70.1%-92.3%) for patients in arm A vs 82.3% (95% CI, 72.7%-88.7%) for those in arms B/C.
In patients from arms B/C whose tumors harbored EGFR T790M mutations detected by circulating tumor DNA (ctDNA) and switched to osimertinib due to molecular progression (n = 8), the median OS was NR compared with 42.84 months (95% CI, 28.58-NR) for those in arms B/C who switched treatment after experiencing radiological progressive disease per RECIST v1.1 criteria (n = 58).
Moreover, patients in arm A experienced a median brain PFS of 34.3 months (95% CI, 26.94-NR) vs 22.3 months (95% CI, 18.56-28.8) for those in arms B/C (HR, 0.54; 90% CI, 0.34-0.86). The 18-month brain PFS rates were 82.2% (95% CI, 67.6%-90.7%) for arm A and 63.5% (95% CI, 52.4%-72.6%) for arms B/C.
“The APPLE trial demonstrated that serial monitoring of T790M status by ctDNA is feasible in patients with EGFR-mutant advanced NSCLC for making treatment decisions,” lead study author Jordi Remon, MD, of Paris-Saclay University and the Institut Gustave Roussy in Villejuif, France, explained in a presentation of the data.
Previously, patients with EGFR-mutant advanced NSCLC who progressed following frontline treatment with an EGFR TKI had limited treatment options; however, acquired EGFR exon 20 T790M mutations were identified as a mechanism of resistance following treatment with an EGFR TKI, one that occurs in up to 60% of tumors. This discovery shifted the treatment landscape for this population, according to Remon.
“It is not always feasible to perform a tissue biopsy at the time of progressive disease with the aim to search for this mechanism of resistance,” Remon said. “It has been reported that the liquid biopsy could be an alternative for genomic profiling at baseline and also at the time of progression.”
One of the aims of the APPLE trial was to evaluate whether serial monitoring of EGFR T790M status through ctDNA was feasible to identify patients who could benefit from an earlier switch to osimertinib prior to radiologic progression per RECIST v1.1 criteria.
Data previously published in the Annals of Oncology showed that ctDNA identified 17% of patients in arm B who experienced molecular progression in the form of an EGFR T790M mutation.2 The 18-month PFS rate with osimertinib was 67.2% (84% CI, 56.4%-75.9%) for arm B vs 53.5% (84% CI, 42.3%-63.5%) for arm C, meeting the primary end point of the trial. The median PFS was 22.0 months in arm B vs. 20.2 months in arm C.
The APPLE trial enrolled patients with advanced NSCLC who harbored EGFR exon 19 deletions or exon 21 L858R mutations and who had not received prior treatment with an EGFR TKI.1 Patients were also required to have an ECOG performance status of 0 to 2. Those with stable brain metastases were permitted to enroll if they were not receiving steroids.
Patients were randomly assigned in a 1:1:1 fashion to the 3 treatment arms. Those in arm A were administered 80 mg of osimertinib per day until progressive disease per RECIST v1.1 criteria. Patients in arm B received 250 mg of gefitinib per day until central ctDNA testing revealed an acquired EGFR T790M mutation or they experienced progressive disease; they were then treated with 80 mg of osimertinib per day until progressive disease. Those in arm C were given 250 mg of gefitinib per day until progressive disease; subsequently, they received 80 mg of osimertinib per day until disease progression.
Stratification factors included EGFR mutation subtype (exon 19 deletion vs exon 21 L858R), brain metastases (present vs absent), and EGFR T790M mutational status at baseline (positive vs negative).
Although liquid biopsies were performed once every 4 weeks for all patients enrolled on the trial, results were only applied as a predictive for making therapeutic decisions in arm B.
The trial’s primary end point was investigator-assessed 18-month PFS rate on osimertinib per RECIST v1.1 criteria. Overall response rate and OS served as key secondary end points.
“For the analysis presented [at the meeting], we decided to pool arm B and arm C of the trial because these two arms explored the sequential treatment approach, and the results of these arms were compared with arm A of the trial [featuring] up-front osimertinib,” Remon explained.
The median age for all patients in arm A was 68 years (range, 46-87) compared with 65 years (range, 31-89) for those in arms B/C. The majority of patients in arm A and arms B/C were female (56.6% and 69.9%, respectively), Caucasian (75% and 78%), harbored EGFR exon 19 deletions (66% and 64%), and were never smokers (66% and 65%). Forty percent of patients in arm A had a World Health Organization (WHO) performance status of 0 and 57% had a WHO performance status of 1; those rates in arms B/C were 48% and 49%, respectively.
The median tumor volume based on target lesions was 52 mm and 51 mm in arm A and arms B/C, respectively. Nineteen percent and 29% of patients in arm A and arms B/C, respectively, had brain metastases.
Subsequent treatment at progression for patients in arm A (n = 20) included chemotherapy (50%), radiotherapy (15%), switch to osimertinib (15%), and targeted therapy (15%). Among the 3 patients to receive subsequent targeted therapy, the regimens consisted of gefitinib, erlotinib (Tarceva), and osimertinib plus savolitinib. Subsequent treatments for patients who progressed in arms B/C (n = 81) included switch to osimertinib (82%), targeted therapy (10%), and osimertinib outside the study (6%). Targeted therapy regimens in arms B/C included gefitinib (n = 2), erlotinib (n = 5), and afatinib (Gilotrif; n = 1).
Additional data showed that the median PFS was 19.5 months (95% CI, 14.42-26.97) for patients treated in arm A compared with 21.39 months (95% CI, 18.43-30.82) for those treated in arms B/C (HR, 0.87; 90% CI, 0.60-1.26). The 18-month PFS rates were 51% (95% CI, 35.8%-64.5%) and 61% (95% CI, 49.6%-69.8%) for arm A and arms B/C, respectively.
In an exploratory analysis of ctDNA clearance of EGFR mutations and PFS, patients in arm A who had no mutation at 4 weeks (n = 25) achieved a median PFS of 16.6 months (95% CI, 12.3-26.1) vs 9.1 months (95% CI, 1.2-14.6) for those in arm A with an EGFR mutation detected in ctDNA at this timepoint (n = 9). In arms B/C, the median PFS was 10.6 months (95% CI, 9.0-12.9) for those with no mutation detected in ctDNA at week 4 (n = 44) vs 5.0 months (95% CI, 3.4-9.3) for those who had a mutation detected (n = 17).
At week 8, those in arm A with no mutation detected in ctDNA (n = 27) had a median PFS of 19.8 months (95% CI, 10.7-not estimable [NE]) compared with 13.0 months (95% CI, 4.6-NE) for those in arm A with a detectable mutation (n = 5). In arms B/C, the median PFS was 10.1 months (95% CI, 9.0-12.6) and 3.6 months (95% CI, 2.3-18.4) for those with no mutation (n = 44) and those with a mutation (n = 9) detected at week 8, respectively.
Regarding safety for those in arm A (n = 53) and arms B/C (n = 71) who received treatment with osimertinib, 98.1% and 90.1% of patients, respectively, experienced at least 1 any-grade adverse effect (AE). Rates of grade 3 or higher AEs were 39.6% and 28.2% for those in arm A and arms B/C, respectively. Additionally, any-grade treatment-related AEs (TRAEs) occurred in 86.8% of patients in arm A and 60.6% of those in arms B/C; the rates of grade 3 or higher TRAEs were 7.5% and 7.0%, respectively.
The most common any-grade TRAEs in arm A and arm B/C, respectively, were diarrhea (39.6% vs 23.9%), dry skin (37.7% vs 12.7%), and acneiform rash (20.8% vs 9.9%). Dose reductions were required in 7.5% of patients in arm A and 7.0% of those in arms B/C.
Editor’s note: Dr Remon reported serving in a consulting/advisory role with Merck Sharp & Dohme, Bristol Myers Squibb, Pfizer, Ose-Immunotherapeutics, Boehringer Ingelheim, Sanofi, Janssen, and Takeda; speaking at a company’s organized public event for Merck Sharp & Dohme, Roche, and Janssen; receiving grants/research support from Merck Sharp & Dohme and AstraZeneca, and receiving sponsored research from Merck.