CME
Physicians: Maximum of 1.25 AMA PRA Category 1 Credits™
Released: September 01, 2022
Expiration: August 31, 2023
Zofia Piotrowska, MD, MHS:
There were a lot of exciting new data on the use of targeted therapy in NSCLC at ASCO 2022, including updated reports on the KRAS G12C inhibitor adagrasib.
Another KRAS G12C inhibitor, sotorasib, became the first of this new class of therapies to receive FDA approval for treatment of adult patients with KRAS G12C–mutated locally advanced or metastatic NSCLC in May 2021 based on the data from the CodeBreaK 100 study.27 CodeBreaK 100 was a phase I trial evaluating sotorasib in patients with advanced solid tumors harboring the KRAS G12C mutation.28 Of 129 recruited patients, 59 had NSCLC, and sotorasib achieved a confirmed objective response of 32.2% in this subset. There were no dose-limiting toxicities or treatment-related deaths.
Zofia Piotrowska, MD, MHS:
The KRYSTAL‑1 registrational cohort is part of an ongoing, multicenter, open-label, nonrandomized phase I/II study.29,30 In this presentation, 116 patients with unresectable/metastatic NSCLC harboring KRAS G12C mutations identified in tumor tissue who had previously been treated with PD‑1/PD‑L1 inhibitors, in combination or in sequence with platinum‑based chemotherapy, were treated with adagrasib. Adagrasib was given at 600 mg twice daily in a fasted state. Recall that sotorasib is dosed once daily without the need for fasting, so there are some differences in administration between these 2 agents. KRYSTAL-1 was open to patients with stable CNS metastases, and the primary endpoint was ORR by BICR. Secondary endpoints were disease control rate (DCR), DoR, PFS, OS, and 1-year survival rate.
Zofia Piotrowska, MD, MHS:
Most patients (84%) were White, with smaller subsets of other races represented in this cohort.29,30 Most (96%) were current or former smokers, 43% had received 1 previous line of therapy, and 12% had received ≥4 previous lines of therapy. Almost all (98%) patients had received previous platinum‑based chemotherapy and ICI therapy. Twenty-one percent of patients had brain metastases at entry.
Zofia Piotrowska, MD, MHS:
The ORR with adagrasib in patients with advanced KRAS G12C–positive NSCLC was 42.9%.29,30 Of these, 42.0% were PR, with 1 (0.9%) CR, and a DCR of 79.5%.
Zofia Piotrowska, MD, MHS:
Tumor shrinkage occurred in 89 (79.5%) patients, highlighting that many patients with stable disease (SD) did gain some tumor regression.29,30
Zofia Piotrowska, MD, MHS:
The median DoR among the 48 responders was 8.5 months (95% CI: 6.2-13.8) and median time to treatment response was 1.4 months.29,30 Responses were seen across a range of doses.
Zofia Piotrowska, MD, MHS:
The median PFS was 6.5 months (95% CI: 4.7-8.4) and the median OS was 12.6 months (95% CI: 9.2-19.2).29,30
Zofia Piotrowska, MD, MHS:
There were no major differences in ORR across different subgroups analyzed, including age, number of previous systemic regimens, and performance status.29,30 Adagrasib was active in patients with and without brain metastases at baseline. In fact, numerically, ORR appeared to be slightly higher in patients with baseline brain metastases.
Zofia Piotrowska, MD, MHS:
The KRYSTAL-1 investigators conducted a post hoc analysis of the intracranial activity of adagrasib in a subset of 33 patients with radiographically evaluable CNS metastases at baseline.29,30 It is important to recall that the study design permitted stable brain metastases only and to be aware that most radiographically evaluable patients in this post hoc analysis (81.8%) had received radiation therapy prior to treatment with adagrasib.
The intracranial ORR among this subset was 33.3% (95% CI: 18.0%-51.8%), with a median duration of intracranial response of 11.2 months (95% CI: 2.99 to not evaluable). Median intracranial PFS was 5.4 months (95% CI: 3.3-11.6). Of 13 patients with evaluable brain metastases, most gained some tumor shrinkage, with 6 reaching PR, 4 with SD, and 1 CR.
Zofia Piotrowska, MD, MHS:
The impact of comutations on response to KRAS G12C inhibitors has been of interest and was the subject of an exploratory analysis by the KRYSTAL-1 investigators.29,30 Overall, response rates appeared broadly similar across the different comutation profiles evaluated. Numerically, patients with STK11 and KEAP1 wild‑type tumors had a higher ORR than those with mutated STK11 and KEAP1 tumors, whereas patients with mutated TP53 or CDKN2A had higher ORR than those with wild-type TP53 or CDKN2A, respectively. Again, patient numbers were small and I would urge caution when interpreting these data. As expected, the response rates across PD‑L1 subgroups appeared similar. This is still an area where we need more data before we can use these comutations to guide our practice.
Zofia Piotrowska, MD, MHS:
The rate of any-grade TRAEs with adagrasib was 97%, with 40.5% and 2.6% of patients experiencing a grade 3 or 4 TRAE, respectively. Sixty-three percent of patients receiving adagrasib experienced diarrhea, with nausea seen in 62.1%, and vomiting in 47.4%, although these were largely grade 1/2 toxicities.29,30 Fatigue occurred in 40.5%, and approximately one quarter of patients had an increase in alanine aminotransferase (ALT)/aspartate aminotransferase (AST). Two patients (1.7%) in this trial had a grade 5 TRAE, and these were cardiac failure in a patient with a history of pericardial effusion and pulmonary hemorrhage.
Dose reduction was needed in 51.7% of patients, dose interruption in 61.2%, and treatment discontinuation in 6.9%. The most common TRAEs leading to dose modifications were gastrointestinal events, hepatic events (increased ALT, AST), and fatigue.
Zofia Piotrowska, MD, MHS:
How best to use these drugs in patients with brain metastases is an important clinical question, given that this is a common population we see in clinic. The activity and safety of adagrasib in patients with KRAS G12C–mutated NSCLC and untreated CNS metastases was evaluated in separate cohort of KRYSTAL-1. This CNS cohort of 25 patients had asymptomatic, active, stable, untreated CNS metastases at baseline, and all received adagrasib 600 mg twice daily.Sabari 2022 The objectives of this study were to evaluate the safety and intracranial and systemic activity of adagrasib, and to measure adagrasib concentration in the cerebrospinal fluid (CSF).
Zofia Piotrowska, MD, MHS:
At baseline, 20% of patients in this cohort had no target CNS lesions, 48% had 1, and 28% had 2-5.31 Twenty-four percent had no nontarget CNS lesions, 28% had 1, 40% had 2-5, and 1 (4%) patient had more than 5.
Zofia Piotrowska, MD, MHS:
Overall, 6 of 19 (32%) evaluable patients had an intracranial response to treatment with adagrasib.31 The intracranial DCR was 84% compared with a systemic disease control rate of 70%; and concordance between systemic and intracranial disease control was 88%.
In 2 patients with CSF pharmacokinetic data available, concentrations of adagrasib were indicative of CNS penetration and antitumor activity, which is encouraging and correlates with the response data.
Zofia Piotrowska, MD, MHS:
Intracranial tumor shrinkage with adagrasib was observed in the majority of patients in the CNS cohort, and 32% of patients had an intracranial response.31 Median PFS among this group was 4.2 months. Median DoR and OS in the intracranial responders was not reached. The median follow-up in this study was only 6.6 months, so we will have to wait for more information on the durability of intracranial response. Nevertheless, this is encouraging information about potential intracranial activity. The median DoR in patients with a systemic response was 9.6 months, and the median PFS was 5.6 months.
A post hoc analysis of CodeBreak 100 was presented at the 2021 World Conference on Lung Cancer that provided preliminary information on the intracranial activity of sotorasib.32 Patients were eligible for this study if they had stable brain metastases previously treated with radiation or surgery. Intracranial DCR was 88% in a small cohort of 16 patients with evaluable brain metastases.
We will need more data on both KRAS G12 inhibitors to be able to evaluate this CNS activity more completely, but I was excited to see these early signs of CNS activity with adagrasib.
Zofia Piotrowska, MD, MHS:
In the KRYSTAL-1 CNS cohort, durable responses were seen in patients treated with 400 mg and 200 mg twice daily dosing, suggesting that the maximal dose of 600 mg twice daily may not be needed to achieve an intracranial response.31 Reduced dosing may enable patients to stay on treatment longer, and these data suggest it may be feasible to take that approach, although again, this is a small number of patients.
Zofia Piotrowska, MD, MHS:
Looking across the 2 trials of sotorasib and adagrasib, it looks like the efficacy of these agents is broadly comparable in terms of PFS, response rate, and durability. However, safety outcomes do look a little different. In the CodeBreaK 100 study of sotorasib, diarrhea was reported in 29.5% of patients and AST/ALT elevation in approximately 12% to 13%.28 Numerically, this looks lower than rates seen with adagrasib, which were 70.7% and approximately 27%, respectively,33 although caution should be taken when comparing across trials. The rate of dose reduction for AEs was 22% in the sotorasib study and 52% in the adagrasib study.
We expect that adagrasib will likely receive FDA approval in the coming months. We would then have 2 good options for treating patients with KRAS G12C mutations. Stephen, how will you choose between them?
Stephen V. Liu, MD:
I agree that these therapies appear similar. Although the toxicity rates were a little different for adagrasib, we are comparing different studies and it is worth noting that the rate of discontinuation for toxicity was 7% in both studies.
Patients with KRAS G12C–mutated lung cancer can respond well to IO and so the current standard is to use IO first. Most patients, therefore, have received IO before these drugs. One wonders if there would be less toxicity if we were to give these therapies before IO? I suspect there may be. I also expect we would see more toxicity if we combined them with IO. There are higher frequencies of AEs in the adagrasib study, but there was also a somewhat higher proportion of patients in that study who received previous treatment with both chemotherapy and IO than in the sotorasib study (98% vs 90%, respectively). Thus, I would not read too much into the toxicity data; I think both therapies have similar profiles. However, we do know that there are differences in the resistance profiles of adagrasib and sotorasib; mutations at H95D/Q/R result in resistance to adagrasib but not to sotorasib.34 Although there is more to learn here as well, this suggests we may be able to sequence one drug after the other.
Zofia Piotrowska, MD, MHS:
I am encouraged by the data from the CNS cohort, as we do need CNS‑active drugs for our patients with KRAS‑mutant cancers. The patient numbers are small, however, and we must take care not to “over‑extrapolate” the CNS activity of this drug based on this small cohort.
Stephen V. Liu, MD:
I agree; there is huge unmet need in this setting. We know that the rate of brain metastases in KRAS G12C–mutated NSCLC is significant, and so having drugs that are active in the CNS is very important.
Data from the other cohort of KRYSTAL‑1, in which patients had pretreated brain metastases, are much less informative because control of disease can be attributed to the radiation, and so it is hard to know what effect, if any, the drug is having there. Those data are not nearly as meaningful as the CNS cohort data that shows efficacy in patients with untreated brain metastases, who are often excluded from trials. This setting needs further study in future trials.
Zofia Piotrowska, MD, MHS:
In other settings, brain imaging is often done at the discretion of the investigator and may happen only at times when patients present with symptoms or have systemic progression, and so it is hard to extrapolate the CNS activity in that context. This type of CNS‑specific cohort, which usually include standard serial CNS imaging and surveillance, is a much better way to assess the efficacy of these drugs in the CNS.
Stephen V. Liu, MD:
If both sotorasib and adagrasib are available in the future, we are going to need to work out how to sequence them. If adagrasib is showing clear evidence of efficacy in the brain, that could be a distinguishing factor. We now need to see if there are similar outcomes with sotorasib in that setting.
Zofia Piotrowska, MD, MHS:
EGFR exon 20 insertion mutations are seen in up to 4% of NSCLC cases,35 but are associated with worse survival than more common EGFR mutations.36 Two drugs—mobocertinib and amivantamab—are now approved by the FDA for use in patients with advanced or metastatic NSCLC with EGFR exon 20 insertions whose disease has progressed on or after platinum-based chemotherapy. Mobocertinib is an oral EGFR kinase inhibitor and achieved an ORR of 28% in this setting.37,38 Amivantamab is a bispecific antibody targeting EGFR and MET that is given by IV infusion and achieved an ORR of 40%.39,40 Both of these therapies are associated with significant toxicity that healthcare professionals should be aware of. Mobocertinib carries a black box warning for risk of life-threatening QT prolongation and is linked to increased risk of interstitial lung disease (ILD)/pneumonitis, cardiac toxicity, and diarrhea. Amivantamab is also associated with risk of ILD/pneumonitis, along with infusion-related reactions (IRR), ocular toxicity, and dermatologic AEs.
There is certainly room for improvement, both in terms of efficacy and the tolerability of drugs targeting EGFR exon 20 insertions, and the next study we will review is an attempt to meet that need.
Zofia Piotrowska, MD, MHS:
CLN‑081 is a novel EGFR TKI with specific activity against EGFR exon 20 insertions and better selectivity for these mutations as compared with wild‑type EGFR.41,42 Our hope is that this will translate into improved safety over other therapies that target these mutations.
At ASCO 2022, we presented updated data from an ongoing multicenter, open-label phase I/II study of CLN‑081 in patients with EGFR exon 20 insertion–mutated advanced NSCLC, for which I was a coauthor.43 This is a dose-escalation and dose-expansion study with multiple cohorts, some of which are moving into dose expansion. Patients had received previous treatment including platinum-based chemotherapy. The primary endpoints were TRAEs, dose-limiting toxicity, laboratory abnormalities, and ORR in the phase IIa dose-expansion cohorts. Key secondary endpoints were ORR, DoR, DCR, PFS, and OS in the phase I dose-escalation cohorts.
Zofia Piotrowska, MD, MHS:
Thirty-six percent of patients had received previous EGFR TKI therapy with either afatinib, gefitinib, or osimertinib, which are generally less active against EGFR exon 20 insertions.43 Three (4%) of the 73 patients enrolled on this study had received previous poziotinib or mobocertinib, which target EGFR exon 20 insertions. Thirty-eight percent of patients had a history of CNS involvement, and 55% had received prior IO.
Zofia Piotrowska, MD, MHS:
The most common AEs were rash (seen in 80% of patients overall), paronychia (32%), and diarrhea (30%).43 However, most TRAEs were grade 1/2 in severity. The most common grade 3 AE was anemia, seen in 10% overall. The rate of grade ≥3 AEs was somewhat higher in the 150-mg twice-daily dosing cohort, whereas most AEs seen in the larger 100-mg twice-daily cohort were low grade. This is encouraging, given that the rate of diarrhea with mobocertinib, as reported from a pooled safety population, was 93%, including 20% of patients having grade 3/4 diarrhea.37 Of course, if patients stay on treatment for longer, even low‑grade AEs can be quite bothersome for their quality of life, but nevertheless, these are encouraging preliminary data on the safety of CLN-081.
Zofia Piotrowska, MD, MHS:
At the time of this analysis, 33% of patients were still receiving CLN-081.43 Overall, 14% of patients required dose reduction and 8% of patients had to discontinue treatment with CLN-081 due to AEs.
Zofia Piotrowska, MD, MHS:
Among the 73 patients treated across dose levels, the confirmed PR rate was 38.4%. The median DoR among these patients was 10 months and the median PFS was 10 months.43 In the cohort of 39 patients treated with 100 mg twice daily, the confirmed PR rate was 41% and the median DoR was more than 21 months. For those receiving ≤65 mg of CLN-081 twice daily, the median DoR was more than 19 months.
I think these data are encouraging, although we would like to see even more improvement in efficacy, given that we are now used to seeing higher response rates with many of our other approved targeted therapies.
Of 3 patients with CNS target lesions at baseline, 1 had an intracranial and systemic response and remains in PR, 1 had SD, and the other had progression of disease in the brain. These are very small numbers, but this may be a hint that CLN-081 has some CNS activity here. Clearly, more data will be needed before we can draw conclusions.
Zofia Piotrowska, MD, MHS:
Other drugs are also in development in this space, including sunvozertinib (DZD9008) a selective EGFR inhibitor that received breakthrough therapy designation from the FDA in January 2022.44 Updated data from a phase I/II trial of sunvozertinib in patients with advanced NSCLC and an EGFR exon 20 insertion mutation previously treated with platinum-based chemotherapy and anti–PD-1/PD-L1 therapy were presented at ASCO 2022.33 In this report, ORR ranged from 22% to 56% across dosing arms, and the safety profile was encouraging with fewer grade 3 AEs than with some of the currently approved agents in this setting.
Stephen V. Liu, MD:
I was encouraged by the safety data with CLN-081, and wonder if it may be possible to escalate the dose further, with a view to perhaps increasing CNS efficacy? That small but important data set on the CNS efficacy of this agent was encouraging and, as we have noted, remains an unmet need. But I agree that there is room for improvement with the efficacy rate. We know that EGFR exon 20 insertions are a heterogeneous group of mutations, and perhaps there is a subset of patients where we might see greater efficacy. I would also like to see more detail on safety outcomes when the data set is larger, because although we are certainly interested in grade 3 AEs vs all-grade AEs, there is also an important difference between grade 1 and grade 2 AEs.
Zofia Piotrowska, MD, MHS:
As mentioned above, amivantamab is a bispecific antibody targeting EGFR and MET that is approved by the FDA for treatment of adult patients with locally advanced or metastatic NSCLC with EGFR exon 20 insertion mutations whose disease has progressed on or after platinum-based chemotherapy.39 This approval was based on the results of a cohort of patients with EGFR exon 20 insertion–positive NSCLC enrolled on the CHRYSALIS trial, which is an open-label phase I multicohort dose-escalation and dose-expansion study (NCT02609776).
An additional cohort of the CHRYSALIS trial is investigating the efficacy of amivantamab in MET exon 14 mutation–positive NSCLC.45,46 Patients must have failed or be ineligible for standard-of-care treatment. Patients received amivantamab at the recommended phase II dose of 1400 mg in people with body weight ≥80 kg, or 1050 mg if <80 kg given intravenously, weekly in cycle 1 and every 2 weeks from cycle 2 onwards. The key study objectives were to assess safety and efficacy.
Initial results in 19 patients suggested that amivantamab monotherapy is active in this cohort, whether patients were treated or untreated.45 The updated analysis presented at ASCO 2022 included 55 patients.46
Zofia Piotrowska, MD, MHS:
Treatment history varied across the cohort. Nine patients were treatment naive, 18 had not received prior MET inhibitor therapy, and 28 patients had received previous MET inhibitor therapy. Between 11% and 25% had a history of brain metastases, and 47% were current smokers compared with the remaining 53% who had never smoked. This is a good reminder that MET exon 14 skipping mutations are commonly seen in patients with a smoking history, unlike some of the other targetable alterations we see in lung cancer, such as EGFR and ALK.47
Zofia Piotrowska, MD, MHS:
The ORR across the entire cohort was 33%, although there were differences in response rate according to treatment history.46 Keeping in mind that these were small patient groups, the highest response rate was seen in the treatment‑naive subset (57%) followed by 47% in patients who had not received a previous MET inhibitor. The response rate was more modest among patients who had progressed on a previous MET inhibitor, at just 17%, or 4 PRs out of 24 patients.
The clinical benefit rate (defined as CR or SD for ≥11 weeks) was 71% overall. Again, the rate was highest in the treatment naive setting (71%), although rates were similar among patients with and without previous MET inhibition (58% and 53%, respectively).
The median PFS in the overall cohort was 6.7 months, 8.3 months in the patients with no previous MET inhibitor therapy, and 4.2 months in those with previous MET inhibitor use. Median PFS was not yet reached in the treatment‑naive subset.
Zofia Piotrowska, MD, MHS:
Safety outcomes in the cohort of patients with MET exon 14 skipping mutations was consistent with observations in the larger CHRYSALIS safety cohort.46 The most common treatment-emergent AEs were IRR (67%), rash (36%), dermatitis (36%), and paronychia (45%), and these were predominantly grade 1/2 in severity. Amivantamab does cause hypoalbuminemia and peripheral edema (27% and 20%, respectively, in the MET exon 14 cohort). These are MET‑related effects that also occur with other MET TKIs.48 Pneumonitis/ILD occurred in 4% of patients. Twelve percent of patients required dose reduction and 5% discontinued treatment due to AEs.
Zofia Piotrowska, MD, MHS:
Two selective MET inhibitors, capmatinib and tepotinib, are FDA approved for patients with metastatic NSCLC and MET exon 14 skipping alterations.49 Of interest, in the GEOMETRY mono-1 study of capmatinib, ORR was highest in the treatment‑naive population at 68% vs 41% in previously treated patients. The VISION study of tepotinib reported similar ORR in treatment-naive and treatment-experienced patients (43% in both groups). The ORR seen with amivantamab in CHRYSALIS appears comparable to some MET‑targeted therapies, although I was disappointed to see that efficacy was more limited in patients who had had received previous MET inhibitors (ORR: 17%), where I think this drug is likely to be the most appealing.
It will be interesting to see more data about the types of resistance patients develop to MET inhibitor therapy because that may help identify patient populations who might gain particular benefit from the sequence of a MET inhibitor followed by amivantamab. Overall, these are intriguing data, but we will need to learn more about how to select patients with MET exon 14 skipping–positive advanced NSCLC who would benefit from amivantamab and the optimal sequence with current standard-of-care options.
Stephen V. Liu, MD:
I think the data presented at ASCO 2022 are more informative than the initial report from WCLC 2021.45,46 However, we have to be careful not to overinterpret data when patient numbers are small, and this updated data set is still relatively small. I look forward to seeing more data on this therapy in larger numbers of patients.
I think the lower rate of edema with amivantamab compared with the available MET inhibitors is likely to be appealing.50,51 Dosing of current MET inhibitors tends to be limited by peripheral edema, for which there are no good solutions, apart from dose reduction. Perhaps the reason we are seeing less edema with amivantamab is a reflection of efficacy. I do not know if that is the case, but if we can effectively and durably treat a MET exon 14–altered lung cancer without significant edema, there is value in that. We may also start to understand why this edema occurs in this setting and how we can prevent it.
Zofia Piotrowska, MD, MHS:
CHRYSALIS‑2 is a multicohort dose-escalation and dose-expansion phase I study assessing the combination of amivantamab with the third-generation EGFR inhibitor lazertinib in patients with advanced EGFR-mutated NSCLC with disease progression on osimertinib and chemotherapy.52
The analysis presented at ASCO 2022 is an update on the cohort of 162 patients with the common EGFR mutations, exon 19 deletion or L858R, who had previously received osimertinib and platinum‑based chemotherapy. Patients received amivantamab IV at the recommended phase II dose (1050 mg if <80 kg and 1400 mg if ≥80 kg) weekly in cycle 1 and every 2 weeks for cycle 2 and beyond with lazertinib 240 mg orally once daily. The primary endpoint was ORR.
Zofia Piotrowska, MD, MHS:
After a median follow-up of 10 months, the ORR with amivantamab plus lazertinib was 33% (95% CI: 26%-41%).52 The median DoR was 9.6 months, the clinical benefit rate was 57%, and the median PFS was 5.1 months (95% CI: 4.2-6.9).
Of interest, among 27 patients with untreated brain metastases who completed ≥1 postbaseline brain scan, 26% had complete CNS lesion clearance, and none had on-target progression of CNS disease. These are early data, but there is a suggestion of CNS activity with this EGFR inhibitor in combination with amivantamab.
Zofia Piotrowska, MD, MHS:
Interestingly, response rates appeared slightly higher among the more heavily pretreated patients. ORR by BICR was 21% in those who received lazertinib followed by chemotherapy; 36% in patients who had had a first-generation or second-generation TKI, then lazertinib followed by chemotherapy; and 39% the most heavily pretreated patients.52 Some of this may relate to patient selection and the type of heavily pretreated patients who meet eligibility criteria for a phase I study like this. But the ORR of 39% in the heavily pretreated population is noteworthy nonetheless.
At ASCO 2021, we saw some information about biomarker selection and encouraging signs that patients with EGFR-driven or MET‑driven resistance mechanisms to osimertinib seemed to reap the most benefit from the combination of amivantamab with lazertinib.53 This makes sense, given the mechanism of action of these 2 drugs. We did not see any updated data regarding the biomarker selection in the ASCO 2022 presentation, but that could be important in selecting patients for this combination.
Zofia Piotrowska, MD, MHS:
The safety of amivantamab plus lazertinib was consistent with earlier data.52 The frequency of rash and paronychia appeared slightly higher with the combination of amivantamab with lazertinib than we saw in CHRYSALIS with amivantamab alone.46,52 In CHRYSALIS-2, hypoalbuminemia was reported in 43% of patients, and peripheral edema in 27%, but rates of grade ≥3 toxicities were low (7% and 1%, respectively).52 IRRs, which we know are a common AE of amivantamab, were seen in 67% of patients. Seven percent of patients had pneumonitis or ILD, including 4% at grade ≥3. Cleary, that is important to keep in mind when treating patients.
Zofia Piotrowska, MD, MHS:
Overall, these data show encouraging activity with amivantamab plus lazertinib in the postosimertinib, postchemotherapy setting, where we do not have approved targeted therapy options. Other drugs are also in development in this setting, including antibody–drug conjugates such as patritumab deruxtecan,54 so this is an active space with lots of trials ongoing.
Stephen V. Liu, MD:
I agree that the activity observed in this study is very encouraging, but you are right—we need to define the population where this drug combination is going to have the greatest impact. For example, I do not think that this drug makes sense for a patient with RET fusion‑mediated resistance. The biomarker data from ASCO 2021 showed that patients with resistance mechanisms other than EGFR/MET-based resistance had a lower ORR.53 Defining where to best use this combination is the way to maximize its impact.
Zofia Piotrowska, MD, MHS:
OPAL was a multicenter, open‑label phase II study of the combination of osimertinib with either carboplatin or cisplatin (physician’s choice) and pemetrexed in previously untreated, EGFR‑mutated advanced nonsquamous NSCLC.55 Essentially, this was a single‑arm study of osimertinib plus chemotherapy as frontline therapy for these patients. After 4 cycles of osimertinib plus carboplatin or cisplatin and pemetrexed, patients went on to received osimertinib plus pemetrexed as maintenance therapy. The coprimary endpoints were safety and ORR, and the secondary endpoints were CR rate, DCR, and PFS.
Zofia Piotrowska, MD, MHS:
Overall, 46% of patients had an EGFR exon 19 deletion and 52% had an L858R mutation.55 Just 1 patient (1.5%) had both mutations. Most patients had stage IV disease (stage IVA: 40%; stage IVB: 45%) and 15% had recurrent disease.
Zofia Piotrowska, MD, MHS:
Osimertinib monotherapy can cause mild myelosuppression,56 as, of course, does chemotherapy. In OPAL, the combination resulted in a decreased neutrophil count, anemia, and decreased platelet count in the majority of patients.55 Among patients who received osimertinib with carboplatin and pemetrexed, 61% had grade ≥3 neutrophil count decrease, 42% had grade ≥3 platelet count decrease, and 27% had grade ≥3 anemia. Comparable rates for those receiving cisplatin were lower, but this is clearly something healthcare professionals should be aware of. Generally, the observed myelosuppression appears to be manageable. Rates of febrile neutropenia were considerably lower (6% in the carboplatin arm), and 12% to 15% of patients discontinued treatment for TRAEs across arms.
Zofia Piotrowska, MD, MHS:
The objective response rates were impressive: 91% in both treatment arms with an overall DCR of 97%.55 At a median follow-up of 21.4 months, the overall 24-month PFS was 70% and the 24-month OS was 92%.
Zofia Piotrowska, MD, MHS:
Of course, we will need randomized data to see if this combination is truly better than frontline osimertinib alone in the long term, but it does make me excited to see the much anticipated data from the ongoing phase III FLAURA‑2 trial, which is comparing osimertinib alone with osimertinib plus either cisplatin or carboplatin and pemetrexed in previously untreated EGFR-mutated advanced NSCLC.57
Stephen V. Liu, MD:
I agree, the response rates are high but, as you say, OPAL is a teaser for FLAURA‑2, where we hope we will ultimately see a survival impact.
We will also see results from the randomized, double-blind phase III COMPEL study in due course, which we hope will validate this approach in previously treated patients.58 COMPEL is open to adults with nonsquamous EGFR-mutated (exon 19 deletion or L858R) locally advanced, metastatic, or recurrent NSCLC, and non-CNS progression after an initial response to first-line osimertinib.
Osimertinib is the standard of care in this setting, and our ability to combine it with other therapies will be important to maximizing patient benefit. So, I look forward to seeing the results of these trials.
Zofia Piotrowska, MD, MHS:
Patritumab deruxtecan is an antibody–drug conjugate targeting HER3, which has shown promising activity (ORR: 39%) in the setting of EGFR‑mutant NSCLC.59
At ASCO 2022, we saw data from an ongoing phase I dose-escalation and dose-expansion study of patritumab deruxtecan in a cohort of patients with locally advanced or metastatic NSCLC without common EGFR mutations following failure of platinum-based chemotherapy with or without IO.54 This study enrolled patients with and without acquired resistance to EGFR TKIs, defined as genomic alterations other than EGFR exon 19 deletion, L858R, L861Q, or G719X mutations.
The expansion cohort of 47 patients was treated with patritumab deruxtecan at the recommended phase II dose of 5.6 mg/kg every 3 weeks. The primary endpoint was confirmed ORR by BICR, and secondary endpoints were DCR, DoR, PFS, time to response, and safety.
Zofia Piotrowska, MD, MHS:
At baseline, 45% of participants had driver genomic alterations present, and 55% had no driver genomic alterations.54 The median number of previous lines of therapy was 3 (range: 1-8), and 19% had received genomic-directed treatment. Forty-five percent of patients had a driver genetic alteration.
Zofia Piotrowska, MD, MHS:
Use of patritumab deruxtecan resulted in an ORR of 26.9% (95% CI: 11.6%-47.8%) in 26 patients without identified driver genomic alterations, with a median DoR of 9.6 months.54 Among 21 patients with identified driver genomic alterations, ORR was similar at 28.6% (95% CI: 11.3%-52.2%), and the median DoR was 9.4 months. Median PFS was slightly longer in the group with driver mutations vs those without (10.8 vs 4.2 months, respectively).
Zofia Piotrowska, MD, MHS:
Responses were seen among patients with a diverse array of genomic alterations, including KRAS G12C/F mutations, RET fusions, NRAS Q61L mutations, and EML4-ALK rearrangements.54 Similarly, there was variation in response among patients without identified genomic alterations, which highlights the need for a better biomarker of response to this particular agent than genomic sequencing gives us.
Zofia Piotrowska, MD, MHS:
These data are intriguing, and I look forward to seeing more follow-up from this study. Of most importance, I think we need better biomarker selection to identify which patients will benefit from patritumab deruxtecan, as the message from this study appears to be that driver genomic alterations are not adequate for that task. Earlier data suggest that HER3 expression by immunohistochemistry also does not seem to be the right biomarker to select patients for this therapy,59,60 but I wonder if there will be others that prove more informative.
Stephen V. Liu, MD:
I think this therapy could fill a huge unmet need, and its efficacy across different resistant settings is important. First, it provides a good option for patients who do not have a targetable mechanism to their disease. In my experience of combining multiple targeted agents, I find that initial responses can be very impressive, but they are often transient. I think that speaks to some of the heterogeneity we see here in relation to resistance. If we had an approach that was active across different mechanisms of resistance, that could potentially overcome that heterogeneity. I look forward to seeing more data with this drug in the post‑TKI space.
Back
