CME
Physicians: Maximum of 1.00 AMA PRA Category 1 Credit™
Released: August 29, 2022
Expiration: August 28, 2023
Allison Betof Warner, MD, PhD:
Individuals with stage IIB/IIC melanoma face a high risk of relapse after 24 months post surgery.1,2 Given the success of adjuvant therapy for stage III melanoma, the randomized phase III KEYNOTE-716 trial was conducted to compare adjuvant pembrolizumab with placebo in patients with high-risk stage IIB/IIC melanoma.3 As a reminder, all patients included in KEYNOTE-716 had a negative sentinel lymph node biopsy, and the primary endpoint was investigator-assessed RFS. The patient characteristics were as expected for the trial population.
Previously reported data from KEYNOTE-716 showed that pembrolizumab significantly improved RFS vs placebo at the first (HR: 0.65; 95% CI: 0.46-0.92) and second (HR: 0.61; 95% CI: 0.45-0.82) interim analyses.3
At ASCO 2022, the investigators reported new data for DMFS, as well as updated RFS data.4
Allison Betof Warner, MD, PhD:
The DMFS data represent an important update. Median DMFS has not yet been reached in either arm despite a median follow-up of 27.4 months.4 Although we think of patients with stage IIB/IIC disease as a high-risk population, it is important to remember that not all of these patients are going to have disease progression or progress as quickly as sometimes we think they might. Nevertheless, DMFS was improved in the pembrolizumab arm vs the placebo arm. Of interest, the lung was the most common site of distant metastasis, and lung metastases were much more common in patients who received placebo vs those who received pembrolizumab.
Hussein Tawbi, MD, PhD:
I think these data are pretty interesting. When we saw the RFS data from the 2 interim analyses, there was considerable discussion about whether these recurrence events are actually important. If you have a patient with stage II disease who has locoregional recurrence, you can perform surgery again, which may cure those patients. Distant metastasis is the closest surrogate available for melanoma mortality. I think that is why this analysis is so relevant for demonstrating the potential benefit of adjuvant pembrolizumab in this setting.
Allison Betof Warner, MD, PhD:
I do think these important data swayed some people who were still sitting on the fence regarding use of adjuvant therapy in patients with stage II disease.
Allison Betof Warner, MD, PhD:
The third interim analysis of RFS was consistent with prior reports.3,4 We continue to see the pembrolizumab arm performing better than the placebo arm.
Allison Betof Warner, MD, PhD:
In addition, there were no new safety signals from what we had seen previously.4
However, I do think it is important to recognize that there are still important safety concerns related to administration of a PD-1 inhibitor, particularly in patients with stage II melanoma who are not guaranteed to recur and who may not ever need anti–PD-1 therapy. Of note, 16% of patients did discontinue pembrolizumab due to adverse events (AEs), and there was a reasonably high rate (38%) of immune‑related AEs in the pembrolizumab arm.
Allison Betof Warner, MD, PhD:
Based on the KEYNOTE-716 findings, I still think it is important to think long and hard about which patients may be appropriate for adjuvant immunotherapy in the stage II setting. I know of many healthcare professionals who saw the approval of pembrolizumab for stage IIB/IIC melanoma and who thought every patient with stage IIB or IIC disease should now be receiving adjuvant pembrolizumab. I would argue that the safety data do not support that approach for all patients.
Hussein Tawbi, MD, PhD:
I agree. The risk–benefit analysis becomes harder every time the overall benefit becomes smaller. KEYNOTE-716 was definitely positive in terms of the HRs. We saw a 36% reduction in the risk of recurrence and a 36% reduction in the risk of distant metastasis. Thus, treatment did decrease the endpoints of relevance in this disease. However, those data reflect the relative risk reduction. When you look at the absolute risk reduction, it is still a pretty small benefit. So, absolutely, the AEs gain a lot more significance and suddenly become more relevant among patients who have an 80% probability of being cured with surgery alone. In stage III disease where patients are cured 60% of the time with surgery alone, we already deliberate long and hard about adjuvant therapy. Now it is even a more delicate and difficult risk–benefit analysis in patients with stage IIB/IIC disease.
Allison Betof Warner, MD, PhD:
Yes. One of the things we always teach our medical students is to think about the number needed to treat vs the number needed to harm. This is a really important calculation to apply here. When you do that analysis, even in stage IIC melanoma, it is not clear that the risk–benefit ratio favors treatment, as you said.
Hussein Tawbi, MD, PhD:
I must say, however, that I do have patients at very high risk for recurrence for whom I do not think twice about administering adjuvant pembrolizumab now that I have a regimen that I can give them—for example, a patient with a 12-mm melanoma who somehow has a negative sentinel lymph node biopsy and who, against the odds, does not have metastatic disease.
Hussein Tawbi, MD, PhD:
Desmoplastic melanoma is a unique type of cutaneous melanoma that usually occurs on sun-exposed areas.5 It is unique pathologically, as well, in that it spreads very locally but with deep infiltration and neurotropism and typically arises in areas where preserving function and aesthetics is very important. Surgical resection is helpful, but because of local infiltration, this particular disease tends to recur locally, sometimes with a vengeance.6 Thus, it is important for patients to be treated beyond just surgical resection. We know that in the metastatic setting, this particular histologic subtype responds really nicely to single-agent PD-1/PD-L1 inhibitors, and there have been reports of response rates of up to 70%.7
SWOG 1512 was a phase II study of neoadjuvant immunotherapy with single-agent pembrolizumab in patients with resectable desmoplastic melanoma.8 This study was designed to take advantage of the activity of single-agent anti–PD-1 therapy to try to prevent recurrence and cure as many patients as possible. Study participants received 3 cycles of single-agent pembrolizumab given at 200 mg every 3 weeks. That dosing regimen is important to note because pembrolizumab can be administered at other doses and schedules. In the case of response, patients went on to surgery. In the case of progression, patients received 1 additional cycle of pembrolizumab to confirm progression of disease and to exclude possible pseudoprogression.
The primary endpoint was pathologic complete response (pCR). This was a well-designed study statistically, with a plan to detect a pCR rate of 25% with 90% power.
Hussein Tawbi, MD, PhD:
The patient characteristics are relevant in that approximately 70% of patients had a head and neck primary.8 In addition, looking at the American Joint Committee on Cancer categories, a sizeable proportion of patients had N2c disease (10%).
The mutational analysis highlights how different desmoplastic melanoma is from our usual cutaneous melanomas. There is a high prevalence of NF1 mutations, no NRAS mutations, and a very small proportion of BRAF mutations. We also see that this is a disease with a high tumor mutational burden.
Hussein Tawbi, MD, PhD:
In total, 30 patients were enrolled on the study, but 1 patient refused treatment and withdrew consent.8 Of the 29 patients who received pembrolizumab, 28 eventually underwent resection. I think the patient who refused surgery had a complete response (CR), which was one of the reasons they did not proceed to resection.
Hussein Tawbi, MD, PhD:
The pCR rate, defined using the International Neoadjuvant Melanoma Consortium criteria as the absence of any viable melanoma in the resected tissue, was 55%, which is an impressive response rate.8 Among the 45% of patients who did not achieve pCR, 1 patient had a major pathologic response (MPR) according to the International Neoadjuvant Melanoma Consortium criteria based on a very small amount of residual disease. None of the patients had inoperable disease. That is a very important point, because sometimes when you intervene with systemic therapy you worry about the disease progressing and then not being operable. That was not an issue.
In subgroup analyses, the pCR rate was generally consistent across subgroups defined by age, sex, Eastern Cooperative Oncology Group performance status, disease status, and baseline LDH.
Hussein Tawbi, MD, PhD:
The objective response here is very important, because we use Response Evaluation Criteria in Solid Tumors standards before we get patients to surgery to gain insight into the pathologic response. The ORR was again high at 46%, and 15 patients had a CR. 8 Remember, this is with single-agent PD-1 inhibition.
Hussein Tawbi, MD, PhD:
In SWOG 1512, pCR was associated with durable disease control.8
Allison Betof Warner, MD, PhD:
These data are consistent with what we have seen in cutaneous melanoma and emphasize the importance of pCR.
Hussein Tawbi, MD, PhD:
Yes. For the desmoplastic melanoma population, RFS and OS were generally pretty impressive and probably highlight the importance of treating these patients with neoadjuvant immunotherapy.
Hussein Tawbi, MD, PhD:
The safety data were a bit surprising in that the rate of grade 3/4 toxicity was very low—only 7%—suggesting that single-agent pembrolizumab is quite safe for this population.8 I expected a slightly higher incidence of severe toxicity, but these patients did pretty well.
Hussein Tawbi, MD, PhD:
In cutaneous melanoma that is nondesmoplastic, getting a pCR rate in the 50% range usually requires combination immunotherapy with a CTLA-4 inhibitor and a PD-1 inhibitor. A pCR rate of 55% with single-agent PD-1 blockade is quite impressive in a disease that is otherwise difficult to resect, has a high rate of local recurrence, and that features neurotropism. This population clearly benefited from neoadjuvant pembrolizumab, and there was not a lot of toxicity.
My conclusion based on these data is that the next time I see a patient with desmoplastic melanoma, I will treat them neoadjuvantly with pembrolizumab.
Allison Betof Warner, MD, PhD:
These are really exciting data. I agree that when we next see a patient with desmoplastic melanoma, neoadjuvant pembrolizumab should be the approach. A key question that arises from this—knowing how morbid these surgeries are—is whether we really need to do surgery in patients who have an MPR, particularly in the head and neck, which often involve a flap or scalp reconstruction, resection of the sentinel node, and even potentially a neck dissection. I think it is critical that we do not jump to the conclusion that resection is not necessary. SWOG 1512 sets us up to study that issue in a separate trial.
Hussein Tawbi, MD, PhD:
Yes. From my perspective, another consideration here is whether we need to use combination immunotherapy in a population that already has such a high rate of response to single-agent immunotherapy. We achieve higher responses with combination immunotherapy, but there is more toxicity.
Allison Betof Warner, MD, PhD:
That might be a place where nivolumab plus relatlimab may be helpful.
Hussein Tawbi, MD, PhD:
Excellent point.
Allison Betof Warner, MD, PhD:
PRADO is a phase II trial of response-directed surgery and adjuvant therapy following neoadjuvant nivolumab plus ipilimumab for stage III melanoma. This study builds on a growing body of evidence supporting use of neoadjuvant therapy in the stage III setting.9-11 There has been great enthusiasm for this treatment approach. PRADO12 builds on the phase II OpACIN-neo trial,13 which really defined the optimal dosing of neoadjuvant therapy with combined immunotherapy. These studies utilize the so-called “flip dosing” schedule of ipilimumab 1 mg/kg and nivolumab 3 mg/kg that many of us have gotten used to using in the neoadjuvant setting.
PRADO included patients with stage IIIB/IIIC melanoma with measurable disease.12 They had an index lymph node that was marked so that it could be found if they had an excellent response.
Patients received 2 cycles of the flip-dosed ipilimumab plus nivolumab and then underwent index node resection. Patients who had a pCR or near pCR—which means <10% viable tumor—did not undergo a therapeutic lymph node dissection. These individuals were followed by CT and ultrasound and underwent surgery if they had a recurrence. Patients with a pathologic partial response—that is, with 10% to 50% viable tumor—did undergo a therapeutic lymph node dissection and then were followed by CT. Pathologic nonresponders underwent therapeutic lymph node dissection and then were treated with adjuvant nivolumab or, if they had a BRAF mutation, adjuvant dabrafenib and trametinib. Of note, 51% of the patient population had a BRAF mutation at baseline.
The primary endpoints were pathologic response rate in the index node and RFS at 2 years.
Allison Betof Warner, MD, PhD:
The PRADO results were very confirmatory of what we had seen in the OpACIN-neo trial, which is reassuring.12 Of the 99 patients included, 49% had a pCR and 12% had a near pCR, yielding an MPR rate of 61%.
Allison Betof Warner, MD, PhD:
The results showed a very typical immune-related AE profile with ipilimumab plus nivolumab.12 The AEs were decreased compared with the standard dosing of ipilimumab plus nivolumab but were still higher in incidence than what is usually observed with a single-agent PD-1 inhibitor.
Allison Betof Warner, MD, PhD:
This trial also included an analysis of surgery-related AEs, which is an important consideration for patients who undergo therapeutic lymph node dissection.12 Approximately two thirds of patients did not undergo therapeutic lymph node dissection, whereas the remaining one third did. As you would imagine, there was a lower incidence of surgical-related AEs for the patients who did not undergo a therapeutic lymph node dissection, which reached statistical significance when compared with individuals who did undergo lymph node dissection.
Allison Betof Warner, MD, PhD:
Quality of life also was improved for patients who did not undergo therapeutic lymph node dissection compared with those who did. 12 Although this is expected, it is a very important endpoint of this study.
Allison Betof Warner, MD, PhD:
RFS and DMFS looked excellent among the all-comers population.12 The 2-year RFS rate was 85% in this high-risk patient population, and the 2-year DMFS rate was 89%—so really quite promising data with this therapeutic approach.
Allison Betof Warner, MD, PhD:
When we break the data down further, patients who had an MPR—that is, those with a pCR or near pCR—did exceptionally well in terms of both RFS and DMFS.12 Patients with a pathologic near response also did reassuringly well, with a 71% RFS rate at 2 years and a 76% DMFS rate at 2 years.
It is important to note the difference between RFS and DMFS, particularly in patients with pathologic partial response. Patients who have a locoregional recurrence may undergo surgery and then be cured. Really, the key numbers here are those for DMFS, and I think they look quite promising, including among patients who did not undergo therapeutic lymph node dissection, which is a key endpoint here.
Allison Betof Warner, MD, PhD:
My conclusion from PRADO is that the data confirm what many of us have been doing with neoadjuvant therapy in terms of administering ipilimumab 1 mg/kg and nivolumab 3 mg/kg in this patient population. Although the results did not meet the intended endpoint of only 1 distant recurrence, I think that we can still argue for skipping a therapeutic lymph node dissection for patients with an MPR.
Hussein Tawbi, MD, PhD:
Yes, I agree. This was a very intriguing study that takes the concept of neoadjuvant therapy one step further by examining whether we need to do extensive surgery in patients with a good response. This is the first neoadjuvant study where we are altering the surgical approach based on the response, which I think is really elegant. The results show that those patients who did not have a therapeutic lymph node dissection did just as well as those who did. Remember, those patients with a pCR/near pCR who did not undergo node dissection were not treated with anything after the index node resection—not even with adjuvant therapy. This really highlights the power of neoadjuvant combination immunotherapy.
The second aspect that I find interesting is that the patients without a pathologic response, who were switched to BRAF/MEK inhibitors if they had BRAF-mutated disease or to additional nivolumab if they did not, actually seem to do pretty well. These results may start to inform not only what to do adaptively with the surgical approach, but also what to do adaptively for those patients in the adjuvant setting.
This is a great study in terms of answering the neoadjuvant question definitively and starting to ask the question about the need for surgery and preparing us for asking the adaptive adjuvant question. All are really important issues.
Allison Betof Warner, MD, PhD:
Yes, and I think you highlighted a key point. In addition to confirming the role for adjuvant therapy for patients without a pathologic response, it is also important to understand whether we need to give adjuvant therapy to patients with a response. I think many of us in our clinical practice have been administering 1 year of adjuvant therapy out of fear. But late toxicity is certainly real, as is financial toxicity and the inconvenience for patients. PRADO raises the important question of whether we need to continue adjuvant immunotherapy for patients who have an excellent response.
Hussein Tawbi, MD, PhD:
Absolutely.
Hussein Tawbi, MD, PhD:
NeoTrio was a neoadjuvant study that combined checkpoint inhibitors with BRAF and MEK inhibitors. The idea is based on the fact that BRAF inhibitors have been shown to favorably modulate the tumor microenvironment by inducing CD8+ T-cell tumor infiltration,14,15 along with the fact that immunotherapy in the neoadjuvant setting is relevant and potentially can improve outcomes.16,17 These findings raise the question for patients with resectable high-risk stage III disease with a BRAF mutation of whether it is better to administer all 3 agents—BRAF and MEK inhibitors plus immunotherapy—with the goal of trying to maximize the cure rate or to rely on immunotherapy alone.
NeoTrio was designed to answer that question and had 3 arms.18 One arm received single-agent pembrolizumab at 200 mg every 3 weeks for 2 doses. There were 2 triplet regimens. In the first, which used a sequential approach, patients received 1 week of BRAF/MEK inhibitors and then single-agent pembrolizumab. In the second triplet regimen, all 3 drugs were given for the entirety of the 6 weeks of neoadjuvant therapy. All arms went to surgery at 6 weeks.
The primary endpoint was the pathologic response rate at 6 weeks and the pCR rate.
Patients’ baseline characteristics were very consistent with other trials in this setting.
Hussein Tawbi, MD, PhD:
Pembrolizumab behaved as expected, with a 30% pCR rate.18 Of interest, 1 week of dabrafenib and trametinib followed by pembrolizumab did not have a high pCR rate; it was only 15%. The triplet regimen where all agents were administered concurrently had a 50% pCR rate.
Hussein Tawbi, MD, PhD:
Of importance, when we look at event-free survival and RFS, all 3 arms performed similarly, with a 1-year event-free survival rate of 80% in all 3 arms and 1-year RFS rates ranging from 80% to 89%.18 Looking at the differences between patients who achieved CRs vs those who did not, again, patients who had a pCR/near pCR had an excellent RFS rate at 1 year of 91% to 100%, whereas patients who did not respond had higher rates of relapse.
Hussein Tawbi, MD, PhD:
The toxicity profile is particularly relevant for the arm that combined the 3 drugs at the same time.
Consistent with prior reports, the toxicity with 2 doses of pembrolizumab monotherapy was relatively low, with only 5% of patients reporting grade 3/4 toxicities.18 The arm that received 1 week of dabrafenib/trametinib followed by pembrolizumab had a higher rate of grade 3/4 toxicity at 25%. The arm that received all 3 agents concurrently had the highest incidence of grade 3/4 AEs at 55%.
Of importance, the discontinuation rate during the neoadjuvant therapy phase was highest for the concurrent triplet regimen, as well, with 40% of patients discontinuing ≥1 of the 3 drugs. Also, 95% of patients who received that regimen required treatment interruption, which is again not very surprising. We have seen in other neoadjuvant studies that this population is at an increased risk of toxicity, perhaps because patients with localized disease are not as immune suppressed as patients with metastatic melanoma.
Hussein Tawbi, MD, PhD:
We know that dabrafenib and trametinib elicit a higher incidence of cytokine-release types of events. When combined with immunotherapy, there seems to be a high incidence of these events in the arm that received targeted therapy and immunotherapy concurrently.18
Hussein Tawbi, MD, PhD:
When it comes to surgical considerations, the triplet arm appeared to have fewer complications from surgery, potentially due to higher objective and pathologic response rates.18
The surgeons filled out a questionnaire indicating whether operability at Week 6 was easier, harder, or no different than it was at Week 0. They found it easier to perform surgery in most patients who received the triplet. So, perhaps as we start considering various aspects of surgery in this situation, the triplet regimen may have a role.
Hussein Tawbi, MD, PhD:
In conclusion, treatment with 1 week of dabrafenib/trametinib followed by pembrolizumab did not seem to improve outcomes in any way. It increased toxicity and did not improve the response rates, so that is not a recommended approach. The highest response rate was seen in the arm treated concurrently with the triplet for 6 weeks. However, the higher rate of pCR again comes at the risk of higher toxicity.
The triplet is an interesting regimen. There are unique situations in which I would choose the triplet over other neoadjuvant regimens. For example, for patients with BRAF-mutated disease and bulky tumors, obtaining an 80% response rate and potentially helping the surgeon operate are 2 considerations that I will take into account. I think we have all gotten pretty good at managing toxicity, so I would not necessarily be afraid of that.
Allison Betof Warner, MD, PhD:
My interpretation of NeoTrio is that most patients should not receive triplet therapy. The toxicity rates are quite concerning. I do not know that the degree of improvement in the pCR rate makes the increase in toxicity worth it for most patients. Although I think we can manage the toxicity, I am concerned that this can lead to surgical issues, particularly fevers, liver function test abnormalities, and things that may lead to surgical delays or complications. I do agree that if a patient has particularly bulky disease, concurrent dabrafenib/trametinib and pembrolizumab may be an appropriate approach, but I do not think this changes the standard of care for most patients.
Back
