PROFILE 1014 was the first study to define the role of the ALK inhibitor crizotinib in the first-line treatment of patients with ALK-positive lung cancer. It compared crizotinib 250 mg twice daily (n = 172) with pemetrexed plus cisplatin (n = 171) in patients with ALK-positive, locally advanced, recurrent or metastatic non-squamous NSCLC in the first-line setting. The primary efficacy endpoint (i.e., superiority of crizotinib vs. chemotherapy in terms of PFS) was met, with an HR of 0.454 (median PFS, 10.9 vs. 7.0 months for crizotinib and chemotherapy, respectively; p < 0.0001) . ORR was significantly higher with crizotinib than with chemotherapy (74 % vs. 45 %; p < 0.001). At that time, median OS had not been reached in either group at data cut-off.
Long-term OS advantage in PROFILE 1014
After a median follow-up of approximately 46 months in both arms, Mok et al. presented the updated OS and safety analysis . According to these data, crizotinib gave rise to a 24 % reduction in mortality risk compared to chemotherapy (HR, 0.76), although this difference was not statistically significant (p = 0.0978). Median OS had still not been reached for crizotinib, with a lower margin of 45.8 months, which resembled the median OS of 47.5 months obtained for chemotherapy. The four-year OS rates were 56.6 % vs. 49.1 %. This is one of the highest 4-year survival rates for any TKI therapy in patients with stage IV NSCLC to date.
As crossover had been permitted in PROFILE 1014, the proportion of patients randomised into the chemotherapy arm who received subsequent TKI therapy with crizotinib was substantial. Using a rank-preserving structural failure time model adjusted for crossover, it was estimated that the HR for OS would be 0.346 if no crossover had occurred (median OS, 59.8 vs. 19.2 months). With regard to the impact of subsequent therapies, it was shown that patients who received crizotinib followed by another ALK TKI had the longest OS, whereas those randomised to chemotherapy followed by no ALK TKI or other treatment fared worst (Figure 1). No unexpected toxicities were revealed with long-term crizotinib treatment.
Figure 1: Impact of various treatment sequences on overall survival
ALEX: head-to head comparison
However, as progression is inevitable in patients treated with the first-generation ALK inhibitor crizotinib, further targeted options have become available. The second-generation ALK inhibitor alectinib has shown systemic and CNS efficacy in patients previously treated with crizotinib in two pivotal phase II trials [3, 4]. Based on these studies, alectinib was approved for the treatment of patients with ALK-positive NSCLC who have progressed on crizotinib or are intolerant to it.
In the first-line setting, alectinib was compared to crizotinib in the ALEX trial that investigated both the systemic and CNS efficacy of these two drugs in patients with ALK-positive, stage IIIB/IV NSCLC. Overall, 303 patients participated in ALEX, with 152 and 151 receiving alectinib 600 mg twice daily and crizotinib 250 mg twice daily, respectively. The primary endpoint of the ALEX study was met: alectinib significantly improved PFS compared to crizotinib (not reached vs. 11.1 months; HR, 0.47; p < 0.001) .
The CNS is a common site of metastasis and disease progression in ALK-positive NSCLC patients. As many as 30 % of patients already have CNS lesions at initial diagnosis , and the CNS is the first site of progression in up to 50 % of patients receiving crizotinib [7, 8]. Patients with asymptomatic brain metastases were permitted in the ALEX trial, irrespective of whether treatment for them had been administered or not. All of the patients underwent brain imaging prior to study entry and every 8 weeks thereafter. At the ESMO 2017 Congress, Gadgeel et al. presented the CNS efficacy results from the ALEX trial after a median follow-up of approximately 18 months .
Activity across multiple CNS endpoints
Among the total study population, 122 individuals had CNS disease at baseline. Here, 64 were randomised to alectinib and 58 to crizotinib. Approximately 60 % in each arm had not received any treatment for their brain metastases prior to study entry. Compared with crizotinib, alectinib significantly improved PFS both in patients with baseline CNS metastasis (not reached vs. 7.4 months; HR, 0.40; p < 0.0001) and those without (not reached vs. 14.8 months; HR, 0.51; p = 0.0024). PFS was also assessed by prior radiotherapy in patients with baseline CNS metastasis. Alectinib gave rise to significant PFS prolongation regardless of whether radiotherapy had been administered or not (HRs, 0.34 and 0.44, respectively).
Progression in the CNS at the time of first progression was less frequent with alectinib than with crizotinib in both patients with and without CNS metastases at baseline. This also applied to patients with baseline lesions independent of prior radiotherapy. A key secondary endpoint of the ALEX trial was time to CNS progression. Based on a competing risk analysis, it was shown that alectinib significantly delayed CNS progression both in patients with CNS metastases at baseline (cumulative incidence rates at 12 months, 16.0 % vs. 58.3 % with alectinib and crizotinib, respectively; cause-specific HR, 0.18; p < 0.0001) and those without (cumulative incidence rates at 12 months, 4.6 % vs. 31.5 %, respectively; cause-specific HR, 0.14; p < 0.0001). This suggests that alectinib has protective effects against the development of CNS progression. Again, alectinib treatment benefited both patients with and without prior radiotherapy concerning the cumulative incidence rate of CNS progression (HRs, 0.11 and 0.22, respectively).
Superiority with respect to intracranial responses
CNS responses according to RECIST were assessed separately in patients with and without prior radiation who had measurable CNS disease at baseline. In those who had received radiotherapy, CNS ORR for alectinib was 85.7 %, and complete remissions in the CNS occurred in 28.6 % (Figure 2). For crizotinib, these rates were 71.4 % and 0 %, respectively. Patients without prior radiation showed CNS overall and complete response rates of 78.6 % and 42.9 %, respectively, for alectinib, and 40.4 % and 6.7 %, respectively, for crizotinib. Duration of response obtained with alectinib also exceeded the corresponding results observed with crizotinib in patients with and without prior radiation.
Similar outcomes resulted for CNS response in patients with both measureable and non-measureable CNS disease at baseline. The group without prior radiotherapy fared best; here, CNS overall response and complete remission rates were 74.4 % and 61.5 %, respectively. For crizotinib, these percentages were 24.3 % and 10.8 %, respectively. Again, alectinib performed better with regard to duration of response in patients with and without prior radiotherapy.
In the ALEX trial, efficacy was also assessed by use of the RANO criteria. The analysis showed that data generated by RECIST and RANO criteria were consistent. According to the RANO criteria, the cumulative incidence rates of CNS progression at 12 months were significantly lower with alectinib than with crizotinib (8.0 % vs. 32.2 %; cause-specific HR, 0.18; p < 0.0001). Along with the systemic results, these findings consolidate alectinib as the new standard of care for patients with previously untreated, advanced, ALK-positive NSCLC.
Figure 2: CNS responses in patients with measurable CNS disease in the ALEX trial: patients with (left) and without prior radiation (right)
ALUR: alectinib versus chemotherapy
Until recently, no studies have directly compared alectinib with standard chemotherapy in patients with ALK-positive NSCLC after crizotinib failure. This gap was closed by the randomised phase III ALUR trial. Patients enrolled in this study had already received crizotinib and one line of platinum-based chemotherapy. They were randomised to either alectinib 600 mg twice daily (n = 72) or chemotherapy with pemetrexed or docetaxel as per investigator’s choice (n = 35).
The primary endpoint, which was PFS in the ITT population according to investigator assessment, was met, with an HR of 0.15 . Median PFS was 9.6 vs. 1.4 months (p < 0.001). All of the subgroups experienced markedly greater PFS benefit from alectinib treatment than from chemotherapy. Likewise, the analysis according to the independent review committee (IRC) showed a clear advantage for alectinib, with median PFS being 7.1 vs. 1.6 months (HR, 0.32; p < 0.001). A similar magnitude of effect was observed for the differences in overall response rates; these were 37.5 % vs. 2.9 % by investigator, and 36.1 % vs. 11.4 % by IRC. Disease control was obtained in 80.6 % vs. 28.6 % according to investigator, and duration of response was 9.3 vs. 2.7 months.
CNS responses confined to the alectinib arm
In ALUR, approximately 70 % of patients in each arm had CNS metastases at the time of study entry. CNS overall response rate by IRC in patients with measurable CNS lesions at baseline was defined as a key secondary endpoint of the trial. Alectinib conferred significant benefit with regard to this outcome, as 54.2 % of patients in the experimental arm responded to the treatment (Table). One patient developed CR, and 12 showed PR. In contrast, none of the patients included in the control arm experienced any CNS remissions (p < 0.001).
The median time on treatment was more than three times longer with alectinib than with chemotherapy (20 vs. 6 weeks, respectively). Despite this greater exposure to treatment, AEs of all grades occurred less frequently with alectinib compared to chemotherapy (77.1 % vs. 85.3 %). This also applied to grade 3–5 AEs (27.1 % vs. 41.2 %). Furthermore, alectinib therapy showed advantages with respect to AEs leading to treatment discontinuation (5.7 % vs. 8.8 %) and AEs leading to dose reductions (4.3 % vs. 11.8 %). Overall, the results of the ALUR trial further confirmed the previously proven benefit of alectinib for ALK-positive patients with advanced or metastatic NSCLC.