Thank you for your interest in CCO content. As a guest, please complete the following information fields. These data help ensure our continued delivery of impactful education.
Become a member (or login)? Member benefits include accreditation certificates, downloadable slides, and decision support tools.
EGFR was the first targetable mutation identified in NSCLC, and it now has 5 approved TKIs. Gefitinib and erlotinib were the first-generation EGFR TKIs to be approved in advanced NSCLC. We now have the second-generation agents afatinib and dacomitinib, and most recently, the third-generation agent osimertinib. Gefitinib and erlotinib have reversible inhibition, whereas afatinib, dacomitinib, and osimertinib are irreversible EGFR inhibitors.
One of the initial insights in developing targeted therapy for advanced NSCLC was that appropriately targeted TKIs outperformed chemotherapy. This table shows 6 clinical trials, all well-designed, randomized, controlled studies, that conclusively showed improved PFS with use of a targeted TKI over standard platinum-based chemotherapy as first-line therapy in EGFR mutation–positive disease; the OS was not consistently improved, presumably because patients would often receive a targeted therapy in the second line of treatment.
The phase III FLAURA trial was the first attempt to see if the third-generation EGFR TKI osimertinib could improve outcomes over the first-generation agents erlotinib or gefitinib as first-line treatment for advanced NSCLC.[9,10] It was a double-blind study of 556 treatment-naive patients with EGFR mutations: either an exon 19 deletion or an exon 21 L858R point mutation. Stable brain metastases were permitted.
Patients were randomized to osimertinib 80 mg a day or investigator’s choice of erlotinib 150 mg or gefitinib 250 mg daily. Patients continued treatment until progressive disease or toxicity, and crossover was allowed if a patient’s tumor was positive for T790M at the time of progression. The primary endpoint was investigator-assessed PFS, with secondary endpoints including OS as well as duration of response, disease control, patient-reported outcomes, and safety.
The results of FLAURA clearly demonstrated that osimertinib improved median PFS compared with erlotinib or gefitinib, nearly doubling it from 10.2 to 18.9 months (HR: 0.46; P < .001).[9,10] This led to the FDA expanding the approval of osimertinib in 2019 for first-line treatment of patients with advanced NSCLC and EGFR exon 19 deletions or an exon 21 L858R mutation. The OS results also showed an improvement with osimertinib, although it was not statistically significant.
An important difference between osimertinib and the first-generation EGFR TKIs is that osimertinib was designed to penetrate the blood–brain barrier more effectively. Results from FLAURA showed that patients who had central nervous system (CNS) metastases at baseline clearly did better on osimertinib than on erlotinib or gefitinib, with a median PFS of 15.2 vs 9.6 months, respectively (HR: 0.47; P < .001).[9,10] Also, CNS progression events were more common in patients receiving 1 of the first-generation TKIs, at 15% compared with only 6% in the osimertinib arm.
Osimertinib was generally well tolerated. Even though a substantial minority of patients had the typical acneiform rash associated with EGFR TKIs, the severity varied substantially: only 4% of rashes in the osimertinib arm were grade 2 compared with 18% with the first-generation agents, and no grade 3/4 rash was seen. Aspartate aminotransferase (AST) and alanine aminotransferase (ALT) were less likely to be increased with osimertinib, with grade 3 elevations in approximately 1% of patients, vs 4% to 8% with first-generation TKIs.[9,10]
Cardiac issues were also a concern with osimertinib, including corrected QT interval prolongation and heart failure, which has been seen in a small number of patients. A baseline echocardiogram is recommended at the initiation of treatment.
Unfortunately, tumors treated with targeted therapy inevitably develop resistance to these agents. Osimertinib was an initial attempt to combat this, because it was designed to target the EGFR T790M gatekeeper mutation. This mutation develops in approximately 50% of patients treated with first-generation and second-generation EGFR TKIs.
Now that osimertinib is approved for use as a first-line treatment, we are learning more about how tumors develop resistance to it. Although repeat biopsy after osimertinib treatment is not the standard of care at the present time, we will eventually learn how to better treat a tumor at progression with targeted therapies through repeat biopsies. In a small study of 32 patients with EGFR-mutant NSCLC who developed resistance to single-agent osimertinib, the mechanism of acquired resistance was MET amplification in 22% of patients, T790M/C797S in 19%, loss of T790M in 34%, and no T790M in 9%.
With regard to MET amplification as a resistance mechanism to osimertinib, results from the TATTON trial and other early studies have suggested that adding a MET inhibitor at the time of development of MET amplification can lead to a good response rate and duration of benefit. New mechanisms of resistance are being seen as well. For example, EGFR C797S is a resistance mutation only seen after osimertinib; patients with this mutation may be able to be treated with first-generation TKIs.
Finally, as with first-generation and second-generation agents, a subset of patients treated with osimertinib will develop histologic small-cell lung cancer transformation; for these patients, chemotherapy—especially with platinum and etoposide—may be the best next option. Certainly, multiple clinical trials are in development to address resistance after osimertinib and thereby lengthen the duration of treatment with targeted agents.
Many osimertinib combinations are currently being investigated in early-phase studies to combat resistance from first-line agents or to address resistance in the second line. Key approaches include combining with antiangiogenic agents such as bevacizumab in both the first and second line, and ramucirumab as second-line therapy for T790M-positive NSCLC. There is increasing interest in using EGFR antibodies, such as necitumumab, to combat resistance to osimertinib, especially in the setting of EGFR amplification. Other mechanisms, such as the inhibition of MEK, Axl, Bcl2, Met, and TORC1/2, are also promising and under clinical investigation.
Understanding the importance of EGFR in advanced NSCLC and the development of EGFR inhibitors set the stage for targeted agents in this setting. Subsequently, ALK fusion proteins were identified as a therapeutic target in NSCLC. Crizotinib, which was originally developed as a MET inhibitor, was also found to be an ALK inhibitor.
The first phase III study comparing crizotinib vs chemotherapy in ALK-rearranged NSCLC (PROFILE 1014) showed a clear advantage (both ORR and PFS) over chemotherapy as first-line therapy. Subsequently, the second-generation ALK TKIs ceritinib, alectinib, and brigatinib also showed improvement over chemotherapy or crizotinib in this setting.[16-19] Today, alectinib, crizotinib, brigatinib, and ceritinib are all approved for newly diagnosed ALK-positive NSCLC.
The randomized, open-label phase III ALEX trial was designed to determine if the second-generation ALK TKI alectinib could improve outcomes compared with the first-generation TKI crizotinib as first-line therapy for ALK-positive advanced NSCLC (N = 303). Because of the confidence in alectinib’s ability to cross the blood–brain barrier, asymptomatic brain or leptomeningeal metastases were permitted.
Patients were randomized to alectinib 600 mg BID or crizotinib 250 mg BID and treated until progressive disease, unacceptable toxicity, or withdrawal; crossover was not permitted. The primary endpoint was investigator-assessed PFS, and secondary endpoints included OS, ORR, CNS ORR, time to CNS progression, and safety.
Results from ALEX showed a significant benefit of alectinib over crizotinib, with a very large median PFS improvement from 10.9 months to 34.8 months (HR: 0.43), which led to alectinib’s approval for ALK-positive NSCLC. OS data were not yet mature and no benefit for alectinib has been seen yet.
As mentioned, alectinib penetrates the blood–brain barrier better than crizotinib, and this is reflected in superior outcomes for patients with CNS metastases at baseline. In these patients, there was a striking difference in CNS progression reflecting the cumulative incidence rates of 16.0% for alectinib vs 58.3% for crizotinib at 12 months (HR 0.18; P < .0001). Likewise, in the subset of patients without CNS metastases at baseline, the incidence of CNS progression was also significantly decreased with alectinib vs crizotinib, with 12-month cumulative incidence rates of 4.6% vs 31.5%, respectively (HR: 0.14; P < .0001).
Brigatinib, another second-generation ALK inhibitor, was compared with crizotinib as first-line therapy for advanced ALK-positive NSCLC in the randomized, open-label, phase III ALTA-1L study (N = 275). The study design was very similar to the ALEX study, using brigatinib 90 mg/day for a 7-day lead-in followed by 180 mg/day vs the standard 250-mg BID dose of crizotinib. Asymptomatic or treated CNS metastases were allowed, and crossover from crizotinib to brigatinib was allowed upon disease progression. The primary endpoint was PFS, with secondary endpoints including ORR, intracranial PFS, OS, and safety.
Results from ALTA-1L showed a clear PFS improvement with brigatinib vs crizotinib. The median PFS was not reached in the brigatinib arm compared with 9.8 months for crizotinib (HR 0.49; P < .001). At 1 year, 67% of patients in the brigatinib arm had not progressed vs 43% in the crizotinib arm.
The ORR was higher with brigatinib (71%) than with crizotinib (60%), with an even larger difference in the intracranial response rate, at 78% vs 29%, respectively.
Like alectinib, brigatinib is effective at crossing the blood–brain barrier. Analysis of PFS by baseline CNS metastases clearly showed superior outcomes for brigatinib in patients with and without brain metastases. In patients with brain metastases, the median PFS was not reached with brigatinib (whole body and intracranial) vs 5.6 months with crizotinib. In those without baseline brain metastases, again median PFS was not reached with brigatinib; in this group, the median PFS with crizotinib in a whole-body assessment was 11.1 months but was not reached by an intracranial assessment, where the benefit appeared very similar to brigatinib.
Brigatinib is now approved by the FDA for first‑line use based on these results becoming an effective alternative to alectinib in the first-line setting.
As discussed, resistance to TKIs in patients with NSCLC inevitably develops. Even though some patients can achieve impressive PFS lasting 30 months or longer, we still need strategies for when resistance develops. Lorlatinib is the first of the third-generation ALK inhibitors to be developed.
This table shows IC50 results of an assay in which the 5 available ALK inhibitors were tested on ceritinib-resistant, patient-derived cell lines engineered to express a variety of ALK fusion mutations. The presence of ALK resistance mutations was associated with sensitivity to lorlatinib. Some mutations, such as G1202R, were detected at resistance, which lorlatinib appears effective against. Crizotinib, ceritinib, alectinib, and brigatinib all seem to be fairly inactive against G1202R.
These promising preclinical data led to clinical trials of lorlatinib. Solomon and colleagues conduced an open-label, phase II trial in 276 patients with ALK-positive or ROS1-positive advanced NSCLC and previous treatment with an ALK/ROS1 TKI. The study had an interesting design in which cohorts were defined by the number of previous ALK TKIs, including patients with 3 or more lines. Patients were enrolled into 6 expansion cohorts on the basis of ALK and ROS1 status and previous therapy. All patients were treated with lorlatinib 100 mg once daily in 21-day cycles. The primary endpoint was ORR and intracranial response in ALK-positive patients.
Of note, lorlatinib has also been tested in advanced, ROS1-positive NSCLC, where it has been shown to have some benefit. In a phase I/II trial of 364 patients, including those with CNS metastases and previous crizotinib treatment, lorlatinib showed clinical activity.
Expansion cohorts 4 and 5 consisted of heavily pretreated ALK-positive patients who had received 2 or 3 previous ALK TKIs, respectively, with or without chemotherapy. Even with 75% of patients having baseline brain metastases, there was a robust response rate of 39%, including a good intracranial response rate of 53%.[24,26]
The median PFS was 6.9 months, suggesting good activity. Results from this study led the FDA to approve lorlatinib for ALK-positive patients whose tumors have progressed on earlier ALK agents.
Lorlatinib does have distinct toxicities compared with other ALK inhibitors, although it is generally well tolerated. One key toxicity is hyperlipidemia, which commonly manifests as hypercholesterolemia and hypertriglyceridemia and often requires medical management. In an analysis of pooled phase I data (N = 295), hypercholesterolemia was seen in more than 80% of patients, including 14% that were grade 3 cases and 5 that were grade 4 (1.7%). Hypertriglyceridemia was seen in more than 60% of patients, including 13% grade 3 cases and 7 grade 4 cases (2.4%).
Also notable are mood effects, cognitive effects, and speech effects in a significant minority of patients. These are generally only a nuisance, but can be distressing, particularly to younger patients with NSCLC and an otherwise good performance status. This is important to be aware of, and patients should be counseled to expect it.
In a single-arm study of crizotinib in ROS1-positive NSCLC (N = 53), the median PFS was 19.3 months and the median OS was 51.4 months, suggesting excellent benefit for this first-generation agent. At 4 years, more than one half of the patients remained alive.
In a similar study of entrectinib in ROS1-positive NSCLC (N = 53), the median PFS was also at 19 months and the median OS was not reached, suggesting very good survival outcomes in these patients. Of note, entrectinib appears to be active in patients with CNS metastases with a median PFS of 13.6 months. Therefore, it is often my first choice when a patient with ROS1-positive NSCLC presents with CNS metastases at baseline.
Doebele and colleagues conducted a pooled analysis of entrectinib-related AEs in 355 patients from 3 clinical trials. The results showed that entrectinib is generally well tolerated, with most AEs being grade 1/2. The most common all-grade AEs were dysgeusia (41%; 1 grade 3/4), fatigue (28%; 10 grade 3/4), dizziness (25%; 2 grade 3/4), constipation (24%; 1 grade 3/4), nausea (21%; 0 grade 3/4), and diarrhea (23%; 5 grade 3/4).
That said, serious AEs were seen in 8.5% of patients, including grade 3/4 anemia in 4.5% of patients (n = 16) and grade 3/4 weight gain in 5% of patients. More than one quarter of patients had to either dose reduce or interrupt entrectinib due to treatment-related AEs.
BRAF V600E Mutation–Positive NSCLC
Although BRAF V600E mutation is much more common in melanoma, it also represents a subset of patients with NSCLC. After the approval of the combination of dabrafenib and trametinib in melanoma, it was also studied in NSCLC.
Planchard and colleagues[34,35] conducted an open-label, multicohort phase II study of dabrafenib and trametinib in patients with metastatic BRAF V600E–positive NSCLC. In this study, the ORR was 64% in previously untreated patients (n = 36) and 63% in previously treated patients (n = 57). The median PFS was 10.9 months in the previously untreated group vs 9.7 months in those who received previous treatment. In both cohorts, these results were significantly better than we have seen with dabrafenib monotherapy.
As with melanoma, combining treatment with dabrafenib and trametinib carries the risk of certain side effects that need monitoring. Pyrexia is a strikingly different toxicity compared with many other targeted therapies in NSCLC, and was seen in 64% of previously untreated patients (4 cases grade 3 or worse) and in 46% of previously treated patients (1 case grade 3 or worse).[34,35]
Other toxicities were as expected, including nausea/vomiting and diarrhea in more than one third of patients. Of note, ALT elevation was seen in 17% of previously untreated patients, including grade 3 or worse in 4 patients. Grade 3 or worse hypertension was noted in 4 previously untreated patients and 2 previously treated patients.
To aid in treatment planning, Clinical Care Options has developed a very useful interactive decision support tool for NSCLC treatment, which can be found at clinicaloptions.com/Lungtool. A user selects a combination of patient and disease characteristics through a series of questions, and the tool then provides treatment recommendations based on the entered characteristics from myself and 4 other experts in NSCLC care.