CME
Physicians: Maximum of 1.25 AMA PRA Category 1 Credits™
Released: November 18, 2021
Expiration: November 17, 2022
Thomas E. Stinchcombe, MD:
Moving on to the setting of advanced NSCLC, we had an update on the phase III EMPOWER‑Lung 1 study, which enrolled patients with locally advanced and metastatic NSCLC. Patients were required to have a PD‑L1 ≥50% and no alterations in EGFR, ALK, or ROS1.15,16 They were randomized to receive either cemiplimab 350 mg every 3 weeks for 108 weeks or until progression, or investigator’s choice of chemotherapy for 4-6 cycles. Upon disease progression, patients in the chemotherapy arm were allowed to cross over and receive cemiplimab; patients in the cemiplimab arm could continue cemiplimab along with 4 cycles of chemotherapy.
The primary endpoints were OS and PFS, which have been reported previously.16 Based on those results, cemiplimab was approved earlier this year for first-line treatment of advanced NSCLC with high PD-L1 expression (TPS ≥50%).17 The current analysis focused on the patient subset with locally advanced disease, which was defined as stage IIIB/IIIC disease that was ineligible for definitive concurrent chemoradiation or that recurred after initial treatment with concurrent chemoradiation.15
Thomas E. Stinchcombe, MD:
In the United States, many patients with locally advanced NSCLC get chemotherapy and radiation therapy, and a small subset who aren’t candidates for concurrent chemoradiation therapy often get treated as if they have metastatic disease. Approximately 15% of the patients enrolled on EMPOWER-Lung 1 had locally advanced disease ineligible for chemoradiation, 45 patients in the cemiplimab arm and 42 in the chemotherapy arm. In both arms, approximately 80% of these patients had stage IIIB disease. Recall that PD‑L1 expression ≥50% was an inclusion criterion for the trial, with a breakdown of 40% to 43% with PD‑L1 ≥90%, approximately 25% with PD-L1 of 60% to 90%, and approximately 35% of patients with PD-L1 50% to 60%.
Thomas E. Stinchcombe, MD:
The outcomes in this locally advanced disease subgroup are very similar to those in the ITT patient population. For median OS, the HR was 0.48 (95% CI: 0.20-1.14) in the locally advanced subgroup and 0.57 (95% CI: 0.42-0.77) in the ITT population.15,16 For PFS, the HR was 0.49 (95% CI: 0.27-0.88) in the locally advanced subgroup and 0.54 (95% CI: 0.43-0.68) in the ITT population.
Thomas E. Stinchcombe, MD:
The ORR was 44% with cemiplimab vs 31% with chemotherapy in the locally advanced patient subgroup, with an OR of 1.8 (95% CI: 0.7-4.2; nominal P = .20).15
Thomas E. Stinchcombe, MD:
Overall, first-line cemiplimab improved clinical outcomes vs chemotherapy in patients with locally advanced NSCLC and PD-L1 ≥50%. Although the sample size was small, I think what this tells us is that if you have a patient who’s not a candidate for or who doesn’t want concurrent chemoradiation therapy but whose disease has PD‑L1 ≥50%, cemiplimab may offer activity similar for that seen for patients in the setting of metastatic disease. Single-agent pembrolizumab is also approved by the FDA for patients with locally advanced, unresectable NSCLC who are not candidates for definitive chemoradiation but with a requirement of TPS ≥1%. These occurrences are rare in my clinic, but it’s helpful to know there are options when that situation arises.
Anne Chiang, MD, PhD:
Given the demonstrated benefit of immune checkpoint inhibitors targeting the PD-1/PD-L1 pathway in patients with metastatic NSCLC, trials have been evaluating whether combining anti–CTLA-4 antibodies with anti–PD-L1 antibodies may provide additional benefit. The randomized phase III POSEIDON trial evaluated durvalumab with or without tremelimumab plus chemotherapy vs chemotherapy alone in patients with treatment‑naive, stage IV NSCLC with no EGFR or ALK alterations and with an ECOG PS of 0/1.18 In total, 1013 patients were randomly assigned to durvalumab plus 4 cycles of chemotherapy followed by durvalumab maintenance; durvalumab and tremelimumab plus 4 cycles of chemotherapy followed by durvalumab/tremelimumab maintenance; or the control regimen of chemotherapy for up to 6 cycles. Pemetrexed maintenance was also given to patients with nonsquamous NSCLC who received first-line pemetrexed as part of their chemotherapy regimen. The coprimary endpoints were PFS and OS (in the durvalumab/chemotherapy vs chemotherapy arms). Positivity for either primary endpoint triggered analysis of key secondary endpoints.
Anne Chiang, MD, PhD:
Looking at the baseline characteristics, approximately 36% to 37% of patients had squamous cell histology, with the majority being nonsquamous. CNS metastases were present in 8% to 10% of patients in the immunotherapy arms and were slightly more frequent in the chemotherapy arm at 13%. Rates of liver metastases were also slightly higher in the chemotherapy arm vs the immunotherapy arms, at 24% vs 18% to 20%. However, overall patient characteristics between the arms were similar.
Anne Chiang, MD, PhD:
PFS was significantly improved with durvalumab plus chemotherapy vs chemotherapy alone, with an HR of 0.74 (P = .00093). There was also a positive trend for OS with a median follow-up of approximately 35 months, but this was not statistically significant (HR: 0.86; P = .076). We do see separation at the tail end of both curves, which is expected with immunotherapy benefit.
Anne Chiang, MD, PhD:
Looking at the subgroup analysis of OS in the CheckMate 9LA trial, the addition of nivolumab and ipilimumab to first-line chemotherapy was associated with a significant benefit including in patients with squamous cell histology and PD-L1–negative tumors.19 In contrast, the benefit from the addition of durvalumab (with or without tremelimumab) to chemotherapy in the POSEIDON trial was not as evident in patients with squamous cell histology, PD-L1 negativity, or stage IVB disease.
Differences in duration of treatment with the anti–CTLA-4 antibody may be a contributing factor to these differences. In POSEIDON, only 5 doses of tremelimumab were administered, whereas in the 9LA trial, ipilimumab was continued for up to 2 years. Another possible contributing factor is the use of gemcitabine chemotherapy in POSEIDON but not CheckMate 9LA—many clinicians now use taxanes instead of gemcitabine in the first line setting. In my practice, I prefer to combine platinum chemotherapy with a taxane for patients with squamous histology.
Anne Chiang, MD, PhD:
The response rate was 41.5% with durvalumab plus chemotherapy, 38.8% with durvalumab/tremelimumab plus chemotherapy, and 24.4% with chemotherapy alone. It is important to remember that this trial was not designed to compare the durvalumab plus chemotherapy arm with the durvalumab/tremelimumab plus chemotherapy arm. The statistical analysis is designed to compare each of those immunotherapy-based arms with chemotherapy alone.
Anne Chiang, MD, PhD:
In patients with nonsquamous histology, durvalumab plus chemotherapy and durvalumab/tremelimumab plus chemotherapy both performed better than chemotherapy alone. Looking at the PFS curves, there was clear separation of the curves for both immunotherapy-containing arms vs chemotherapy alone with HRs of 0.77 and 0.66, respectively.
There was a trend toward an OS benefit. Median OS was longest with durvalumab/tremelimumab plus chemotherapy, at 17.2 months, compared with 14.8 months with durvalumab plus chemotherapy and 13.1 months with chemotherapy alone. A longer OS is what we’re looking for with the anti–CTLA-4 plus the anti–PD‑L1 combination.
Anne Chiang, MD, PhD:
By contrast, in patients with squamous histology, neither durvalumab plus chemotherapy nor durvalumab/tremelimumab plus chemotherapy were significantly better than chemotherapy alone for PFS or OS. As I mentioned above, this is different from the CheckMate 9LA study where benefit was seen with combination immunotherapy plus chemotherapy in patients with squamous disease, possibly due to differences in type of chemotherapy given and the duration that immunotherapy was received.19 In POSEIDON, almost 90% of patients with squamous histology who received chemotherapy were treated with gemcitabine plus platinum, whereas gemcitabine was not used in CheckMate 9LA. Furthermore, the anti–CTLA-4 antibody was given for a longer duration in CheckMate 9LA as compared with POSEIDON.
Anne Chiang, MD, PhD:
Looking at responses by histology, response rates were higher with the addition of immunotherapy to chemotherapy in patients with nonsquamous histology, but we did not see that extent of benefit in patients with squamous histology.
Anne Chiang, MD, PhD:
Now looking at safety outcomes, one may wonder if the addition of the anti–CTLA-4 antibody to combination chemoimmunotherapy would add significant toxicity. However, what we saw was that safety outcomes were similar across all 3 arms. The rate of serious treatment-related AEs was slightly higher with both durvalumab and tremelimumab—28% vs 20% with durvalumab plus chemotherapy and 18% with chemotherapy alone—suggesting the combination is slightly more toxic, but not as toxic as might be expected.
Anne Chiang, MD, PhD:
Immune-mediated AEs of any grade occurred at a higher rate in the durvalumab/tremelimumab arm at 33% vs 19% in the durvalumab arm. However, focusing on just grade 3/4 events, the rates were quite similar at 10% and 7%, respectively. In addition, rates of individual AEs were similar between arms, suggesting that the addition of tremelimumab did not substantially increase the incidence of immune-related AEs.
Anne Chiang, MD, PhD:
POSEIDON was a positive study. The primary endpoint of PFS was reached and there was a trend toward an OS benefit for durvalumab plus chemotherapy vs chemotherapy alone. If, with maturation of the data, we see that OS is significantly improved, then certainly durvalumab plus chemotherapy may be another regimen we could consider for our patients with nonsquamous histology. Results with durvalumab/tremelimumab plus chemotherapy also look promising, but we need more information to see if that continues to bear out over time. We will need to keep an eye on the tail of the curves, and if the benefit is prolonged, then it may make sense to use all 3 drugs. It was certainly notable that the addition of the anti–CTLA-4 therapy didn’t result in much more toxicity for patients as compared with the combination of anti–PD-L1 therapy plus chemotherapy. However, even if approved, it is not clear to me that these approaches have clear advantages over currently available therapy options.
Thomas E. Stinchcombe, MD:
I anticipate that the POSEIDON trial results will lead to the approval of the combination of durvalumab/tremelimumab plus chemotherapy. And I agree: It’s always good to have options, although I am also not sure where this treatment will fit among the other choices that I currently have available and how often I may use it in my own clinical practice.
Anne Chiang, MD, PhD:
The outcomes from POSEIDON also make me wonder whether some PD-L1–positive patients would do fine with single-agent immune checkpoint inhibition upfront, rather than using all our tools, including chemotherapy and/or an anti–CTLA-4 antibody at diagnosis. Could we instead sequence these therapies, using immune checkpoint inhibitor monotherapy in the frontline then adding chemotherapy later, thereby sparing our patients chemotherapy-related toxicities if not necessary? There are ongoing clinical trials directly evaluating the timing of immunotherapy and chemotherapy for PD-L1–positive disease, such as the phase III INSIGNA trial comparing first-line pembrolizumab alone vs combination carboplatin/pemetrexed/pembrolizumab up front. This trial also has an arm that tests whether adding chemotherapy post-progression to pembrolizumab is superior to changing from pembrolizumab to chemotherapy upon progression in patients with stage IV nonsquamous NSCLC.20
Anne Chiang, MD, PhD:
The randomized phase III CheckMate 9LA trial demonstrated a significant improvement in OS, PFS, and ORR with nivolumab, ipilimumab, and chemotherapy vs chemotherapy alone in 719 patients with previously untreated stage IV or recurrent NSCLC.19 Based on those results, nivolumab/ipilimumab plus 2 cycles of chemotherapy received FDA approval for the first-line treatment of adults with metastatic or recurrent NSCLC with no EGFR/ALK alterations, regardless of PD-L1 expression level.
At WCLC 2021, Carbone and colleagues21 presented results of a post hoc analysis evaluating outcomes in CheckMate 9LA based on the presence of baseline brain metastases. Patients with brain metastases were allowed on study if they were treated and asymptomatic for at least 2 weeks before the first study dose. In the post hoc analysis, the endpoints were systemic efficacy and safety in patients with and without baseline brain metastases and intracranial efficacy in patients with baseline brain metastases.
Outcomes in this subset are important to assess, given that brain metastases occur in approximately 10% of patients with NSCLC and are associated with a poor prognosis.22,23
Anne Chiang, MD, PhD:
Baseline patient characteristics were well matched between arms. One point to highlight is that patients with baseline brain metastases had higher rates of being a never smoker in the nivolumab/ipilimumab arm compared with those without baseline brain metastases (22% vs 11%).
Anne Chiang, MD, PhD:
In the efficacy analysis, the addition of nivolumab/ipilimumab to chemotherapy was associated with a significant improvement in OS and PFS in patients with treated, asymptomatic brain metastases; 2-year OS rates were 35% with nivolumab/ipilimumab plus chemotherapy vs 12% with chemotherapy alone. These outcomes were similar to those observed in patients without brain metastasis, where 2-year OS rates were 39% and 29%, respectively. Outcomes with chemotherapy were worse in patients with brain metastases than in patients without them.
Anne Chiang, MD, PhD:
There was also good intracranial response with the addition of nivolumab/ipilimumab to chemotherapy in patients with treated, asymptomatic brain metastases, with a median intracranial PFS of 13.5 months vs 4.6 months with chemotherapy.
Anne Chiang, MD, PhD:
Treatment with nivolumab/ipilimumab plus chemotherapy was also associated with a lower rate of developing new brain lesions (16% vs 30% with chemotherapy alone) and a longer median time to developing new brain lesions (9.0 vs 4.6 months with chemotherapy alone) in patients with brain metastases. The ability to control and prevent new brain lesions is key to patients’ quality of life and outcomes.
Anne Chiang, MD, PhD:
In this post hoc analysis of CheckMate 9LA, patients with treated, asymptomatic brain metastases did well on nivolumab/ipilimumab plus chemotherapy vs chemotherapy alone, with outcomes similar to patients without brain metastases. Patients with brain metastases who received the chemoimmunotherapy regimen developed fewer new brain metastases and over a longer period of time. However, when interpreting these data, we need to keep in mind that this post hoc analysis was not prespecified, and among patients with brain metastases, baseline characteristics were somewhat unbalanced.
I think looking at real-world data will be important to teasing out the utility of using immunotherapy in patients with brain metastases. In a retrospective study of 570 patients with metastatic NSCLC receiving pembrolizumab-based therapy, survival outcomes were similar between the 22% of patients who had brain metastases (76% treated) and the 78% of patients who did not have brain metastases.24
Thomas E. Stinchcombe, MD:
I’ll add that seeing in this analysis that patients with baseline brain metastases had very similar benefit to the intent-to-treat patient population tells us that brain metastases are not a poor prognostic factor for use of nivolumab/ipilimumab plus chemotherapy.
Thomas E. Stinchcombe, MD:
One of the common clinical situations we face is that patients with newly diagnosed advanced NSCLC may have small, asymptomatic brain metastases along with extensive disease or symptomatic disease outside the brain. As we discussed, there’s some evidence that immune checkpoint inhibitors may have intracranial efficacy alone or in combination with chemotherapy. Patients with untreated brain metastases have historically been underrepresented in prospective clinical trials. The ATEZO-BRAIN trial aimed to help fill that gap.
ATEZO-BRAIN is a single‑arm phase II trial that evaluated atezolizumab plus chemotherapy in treatment-naive patients with stage IV nonsquamous NSCLC who had untreated brain metastases.25 The patients were required to be EGFR and ALK negative and could have any PD‑L1 expression level. Patients received atezolizumab plus carboplatin/pemetrexed for four to six 3-week cycles, then were able to transition to atezolizumab and pemetrexed maintenance therapy given every 3 weeks for 2 years or until progression or unacceptable toxicity. The coprimary endpoints were safety and investigator‑assessed PFS by Response Evaluation Criteria in Solid Tumors (RECIST) v1.1 and Response Assessment in Neuro-Oncology Brain Metastases (RANO-BM) criteria, which is a standard for assessing response to brain metastases.
Thomas E. Stinchcombe, MD:
This was a small trial that enrolled 40 patients who represented the standard characteristics that we often see in patients with advanced NSCLC: Approximately 80% were current or former smokers, 15% were never smokers, 65% had an ECOG PS of 1, and approximately 40% were on some form of baseline glucocorticoid.
Thomas E. Stinchcombe, MD:
The observed safety profile with atezolizumab and carboplatin and pemetrexed was similar to previous chemoimmunotherapy combinations. The toxicity was acceptable, with no new safety signals in this patient population.
Thomas E. Stinchcombe, MD:
There are 2 components of efficacy to consider: systemic efficacy and intracranial efficacy. The median systemic PFS was 8.9 months and 18‑month systemic PFS was approximately 25%. The median intracranial PFS by RANO-BM criteria was 6.9 months, with an 18-month intracranial PFS of 10.4%.
When we look at the best responses, we see that the systemic and intracranial responses were pretty concordant. Four patients had a discordance, 2 patients who had progressive disease in the body and stable disease in the brain, and 2 patients who had progressive disease in the brain and a partial response in the body.
Thomas E. Stinchcombe, MD:
I think this trial reassures me about what I was already doing. For patients with small, asymptomatic brain metastases, I think many of us have become comfortable starting systemic therapy and doing regular MRI surveillance on the brain metastases to watch for growth or following for worsening of symptoms, then calling our radiation oncologist for potential radiation therapy to the brain.
Anne Chiang, MD, PhD:
Other data also show a benefit with upfront immune checkpoint inhibitors in patients with asymptomatic brain metastases. In an analysis of 37 patients with stage IV PD-L1–positive NSCLC with untreated brain metastases from a phase II trial, pembrolizumab was associated with a brain metastasis response rate of 29.7%.26
Going back to the ATEZO-BRAIN study, I would like to know about other clinical factors that may have affected the response rate, such as size and number of the brain metastases as well as the tumor PD‑L1 status. Given that 43 patients were on steroids at baseline, what effect does steroid use have on intracranial response? I think such factors may play a role in understanding which patients will respond to immunotherapy.
Thomas E. Stinchcombe, MD:
A common clinical question we face is, what is the preferred therapy for a patient with an EGFR mutation or ALK rearrangement who has progressed on SoC TKI therapy? Currently, platinum‑based chemotherapy is the SoC, but it only achieves a response rate of approximately 30% and a PFS of approximately 5‑6 months. Clearly, we are in need of new therapies in this setting. Furthermore, single-agent immunotherapy, for the most part, has not shown significant activity in these patients either, with very low response rates in the second‑line setting. However, investigators have wondered whether the combination of pembrolizumab plus carboplatin/pemetrexed, which is frequently used as a first‑line therapy, could be effective in these patients with disease progression on TKIs.
At WCLC 2021, Gadgeel and colleagues27 presented results from a small number of patients with NSCLC with EGFR mutations, including the common mutations (exon 19 deletions or exon 21 L858R mutations) and the so‑called uncommon EGFR mutations, or ALK rearrangements who were treated with pembrolizumab plus carboplatin/pemetrexed. Of the patients enrolled, 26 had EGFR mutations and 7 had ALK rearrangements. The primary endpoint was ORR by RECIST criteria, and key secondary endpoints included PFS and OS.
The response rate was 42.3% in the EGFR mutation–positive group and 28.6% in patients with ALK rearrangements. Median PFS and OS were 8.3 months and 22.2 months, respectively, for patients with EGFR mutations, and both were 2.9 months for those with ALK rearrangements. Although there have been some concerns that giving immunotherapy after a TKI could lead to a higher rate of immune‑related toxicities, no unexpected AEs were seen in this trial.
Overall, I see this as an interesting hypothesis‑generating study with some preliminary safety data. The ALK‑positive cohort in particular, with only 7 patients, was very exploratory. I think the effect on clinical practice is going to be relatively limited. In my practice for patients who progress on a TKI, I generally offer carboplatin/pemetrexed as an option, or I offer the 4‑drug combination of atezolizumab, carboplatin/paclitaxel, and bevacizumab from Impower 150. Other studies investigating pembrolizumab plus carboplatin/pemetrexed are ongoing or have completed accrual, and I think these may provide better information about the activity of this combination in patients with NSCLC and receptor tyrosine kinase mutations. For example, KEYNOTE-789 is evaluating pemetrexed and carboplatin or cisplatin with or without pembrolizumab in patients with EGFR-mutated metastatic NSCLC.28 I expect we likely will have to extrapolate outcomes from EGFR-based trials to clinical decisions around the use of immunotherapy in patients with NSCLC and other targetable mutations.
Anne Chiang, MD, PhD:
I think this is an important study because it does show that you can have a good response rate with pembrolizumab and chemotherapy in patients with EGFR-mutated NSCLC who have progressed on TKI therapy. This is a clinical scenario where we are still trying to figure out which regimens are best to use. Many other chemoimmunotherapy trials, including the front-line trials that led to our current FDA approved chemoimmunotherapy options along with POSEIDON that we discussed above, did not include patients with EGFR mutations. This study also demonstrated the feasibility of analyzing circulating tumor cells, which was done prior to the first and third cycles, in a multicenter trial.
Back
