CME
Physicians: Maximum of 1.00 AMA PRA Category 1 Credit™
Released: August 07, 2020
Expiration: August 06, 2021
Axel Grothey, MD:
We begin our discussion with a pivotal study presented at the ASCO 2020 Virtual Meeting, the KEYNOTE-177 trial (Capsule Summary).1 This was an open-label, randomized phase III trial that enrolled 307 treatment‑naive patients with MSI-H or dMMR metastatic CRC. Patients were treated with pembrolizumab or investigator’s choice of chemotherapy (mFOLFOX6 or FOLFIRI, with or without bevacizumab or cetuximab). The study had coprimary endpoints of PFS and OS. Crossover was permitted from the chemotherapy arm to the pembrolizumab arm upon disease progression.
Of importance, this study was the first randomized comparison of an anti–PD-1 antibody with a chemotherapy‑based standard of care in the first-line setting for MSI‑H/dMMR CRC.
Axel Grothey, MD:
The initial primary endpoint of the study was PFS.1 Median PFS was twice as long for patients treated with pembrolizumab vs chemotherapy (16.5 vs 8.2 months; HR: 0.60; 95% CI: 0.45-0.80; P = .0002). An interesting phenomenon was observed in the PFS curves, wherein an initial dip was seen in the pembrolizumab curve that later crossed the chemotherapy curve, with longer-term benefit observed for patients receiving pembrolizumab. This phenomenon has also been observed with other malignancies, including gastric cancer.2 One potential explanation for this phenomenon can be found in the response rate finding from this study, as described next.
Axel Grothey, MD:
The ORR was significantly higher with pembrolizumab compared with chemotherapy (44% vs 33%; P = .0275).1 Of interest, a greater proportion of patients receiving pembrolizumab had progression of disease as the best response (29.4% vs 12.3% with chemotherapy). This observation is likely driving the early dip and cross in the PFS curves and suggests that there is a population of MSI-H patients who do not benefit from the use of pembrolizumab/experience disease progression early in the course of treatment. In the KEYNOTE-177 trial, this population represented approximately 30% of patients. Identification of these early-progressing patients is an important clinical issue that needs to be addressed.
Axel Grothey, MD:
Virtually all patient subgroups appeared to benefit from first-line pembrolizumab vs chemotherapy.1 There was an unexpected finding wherein patients with KRAS-mutant or NRAS‑mutant tumors did not seem to derive a PFS benefit with pembrolizumab. This is not consistent with data from other studies, and of note, the analysis did not adjust for multiple variables, so this finding should be further evaluated before being applied to clinical practice.
Axel Grothey, MD:
The KEYNOTE-177 study also found that the duration of response with pembrolizumab was significantly superior compared with standard-of-care chemotherapy.1 After a median follow up of 28.4 months, the median duration of response for pembrolizumab was not reached. In comparison, the median duration of response was 10.6 months for patients receiving chemotherapy (median follow-up: 27.2 months).
Axel Grothey, MD:
As expected, pembrolizumab was well tolerated.1 Fewer grade ≥ 3 treatment-related adverse events were observed with pembrolizumab vs chemotherapy (22% vs 66%, respectively). More immune-mediated adverse events were reported with pembrolizumab (grade ≥ 3: 9% vs 2% with chemotherapy).
Axel Grothey, MD:
The identification of early progressors after first-line pembrolizumab in the KEYNOTE-177 trial is an interesting phenomenon that needs to be addressed in future studies.1
Manish A. Shah, MD:
I agree. At present, we do not know what molecular features would help us to identify those patients who would benefit from first-line pembrolizumab or those patients who might be early progressors. Perhaps the best thing to do at present is to perform an early scan and monitor closely if treating with pembrolizumab. If a patient does not respond and has an aggressive tumor, then it might be possible to salvage him or her with chemotherapy.
It is also worth considering data from patients with CRC who received dual checkpoint inhibitor therapy with nivolumab plus ipilimumab (which is approved for patients with metastatic CRC who have progressed after chemotherapy). These data showed a slightly higher response rate in dMMR CRC.3 It would be interesting to see the response rate for nivolumab plus ipilimumab in the first-line setting and whether the 30% progression as best response rate would be diminished.
Axel Grothey, MD:
The data we have seen for nivolumab plus ipilimumab in first-line setting, albeit in only 45 patients, did show a lower percentage of early progressors, with 13% of patients progressing.4 It is possible that there is an early synergism between anti–CTLA-4 and PD-1 antibody activity.
Manish A. Shah, MD:
When thinking about the crossover in the PFS curves, it is also worth considering whether the patients who progressed early could have benefited from chemotherapy. If so, an appropriate strategy might be to start with 2 months of chemotherapy and then continue with immunotherapy. This would obviously require further study.
Axel Grothey, MD:
Combination or sequential use of chemotherapy and immunotherapy is an interesting strategy to consider. There are some data in lung cancer suggesting that the combination of an anti–PD-1 antibody with chemotherapy in tumors that perhaps are not as immunogenic may be effective.5
There are also data showing that the likelihood of response to checkpoint inhibitors in MSI‑H tumors actually correlates with tumor mutational burden (TMB).6 It would be interesting to check if some patients in the early progressor group of KEYNOTE-177 actually had a low TMB score. Unfortunately, the KEYNOTE-177 study did not require comprehensive next-generation sequencing profiling, so complete data on TMB is unlikely to be available. This would be valuable to explore in future studies.
Manish A. Shah, MD:
One other important point to highlight about the KEYNOTE-177 study is that more than one half of the patients in the chemotherapy arm (59%) crossed over to receive pembrolizumab during the study.1 This made it difficult to assess the OS benefit and further underscores the need for additional study to determine if the early progressors would have benefited from chemotherapy upfront before starting immunotherapy.
Axel Grothey, MD:
Yes, it was unclear as to whether the endpoint of OS was achieved or is going to be achieved.
Manish A. Shah, MD:
Looking to the future, there is much promise for MSI‑H/dMMR tumors regarding immune checkpoint inhibition.
Let’s return to one of the questions we asked at the beginning of the activity.
Axel Grothey, MD:
The IDEA study was an international collaboration of 6 consortia in which 12,834 patients received either 3 months or 6 months of adjuvant leucovorin/fluorouracil/oxaliplatin (FOLFOX) or capecitabin/oxaliplatin (CAPOX) (per investigator choice) for stage III colon cancer.7
The initial data published from the IDEA study evaluated the primary endpoint, which was a noninferiority analysis of the 3‑year disease‑free survival (DFS) of 3 months vs 6 months of adjuvant therapy. Overall, the initial analysis could not confirm the statistical noninferiority of 3 months of adjuvant therapy vs 6 months, with an absolute delta of 3‑year DFS of -0.9% (HR: 1.0; 95% CI: 1.00-1.15).
Unexpectedly, the analysis discovered that there was a differential regimen effect. Patients who were treated with CAPOX could be effectively treated with 3 months of adjuvant therapy compared with 6 months of CAPOX therapy, whereas in almost all studied subgroups 3 months of FOLFOX was inferior to 6 months of FOLFOX.
There are various theories as to why CAPOX performed better than FOLFOX in terms of noninferiority between 3 and 6 months of adjuvant therapy. I think the most likely explanation is that the administration of the fluoropyrimidine is more continuous with the CAPOX regimen compared with the more intermittent administration of 5-FU with the FOLFOX regimen.
At the ASCO 2020 Virtual Meeting, an updated 5‑year survival analysis of 12,835 patients was presented, with a new statistical analysis plan that looked at CAPOX vs FOLFOX and low-risk vs high-risk subgroups after a median follow-up of 6 years.8
Axel Grothey, MD:
In terms of OS, the Kaplan Meier curves for the 3- and 6-month therapy groups were almost completely superimposable.8 The updated analysis showed again that statistical noninferiority was not reached for the overall patient population, with a delta for OS of -0.4% at 5 years.
The 5-year analysis confirmed the differential regimen effect seen in the 3-year analysis. Patients given CAPOX experienced similar 5-year OS and DFS rates with 3 or 6 months of adjuvant CAPOX; however, better outcomes were seen with a longer duration (6 months) of adjuvant FOLFOX therapy, particularly in patients with high-risk (T4/N2) colon cancer.
These data must also be considered in the context of the dramatic toxicity reduction—particularly neurotoxicity—with 3 months of adjuvant therapy. Overall grade ≥ 3 adverse events occurred in 38% and 24% of patients who received 3 months of adjuvant FOLFOX and CAPOX, respectively. In comparison, the rates were 57% and 37% in patients who received 6 months of adjuvant therapy. Grade ≥ 3 neurotoxicity rates were 3% with 3 months of adjuvant FOLFOX or CAPOX compared with 16% and 9% with 6 months of adjuvant therapy.
Axel Grothey, MD:
The IDEA investigators concluded that 3 months of CAPOX adjuvant therapy is appropriate for low-risk and high-risk groups with stage III colon cancer, with minimal to no sacrifice of OS vs 6 months of adjuvant therapy.8 However, if FOLFOX is given, especially for high‑risk patients with T4 and/or N2 disease, 6 months of adjuvant FOLFOX therapy is recommended.
In the future, the hope is that we will identify additional risk factors beyond T and N stage classification, perhaps molecular markers such as Immunoscore or circulating tumor DNA as indicators of measurable residual disease. This will further assist in how we identify the appropriate intensity of adjuvant therapy for individual patients.
Manish A. Shah, MD:
In the United States, we are rarely able to give the full doses of capecitabine that are used in clinical trials. Across the IDEA study centers, I believe there was some treatment selection heterogeneity regarding the CAPOX regimen. Is that right?
Axel Grothey, MD:
That is correct, and you bring up a good point. Different studies within the IDEA consortia selected CAPOX over FOLFOX with different frequencies (range: 0% to 75%).8 The ACHIEVE, SCOT, and TOSCA study investigators selected CAPOX more commonly, whereas the CALGB SWOG study did not use CAPOX at all. There were some differences in regional use of CAPOX vs FOLFOX that translated to differences in the outcomes and deltas observed in the IDEA study. Of note, the SCOT study was the largest contributor of CAPOX patients and also had the best noninferiority results for adjuvant CAPOX at 3 months compared with 6 months.
Manish A. Shah, MD:
This heterogeneity in CAPOX use between sites is slightly concerning, although one might argue that in such a large patient population, you should expect some heterogeneity, even by random chance.
As far as my own plans for clinical practice, for high‑risk patients (ie, the N2 group), I prefer 6 months of FOLFOX, partly because it is not easy to give full doses of the capecitabine. For low‑risk patients, I will likely use CAPOX and FOLFOX interchangeably for 3 months. I am interested to know how you will apply these data to your own practice.
Axel Grothey, MD:
In our clinic, the capecitabine dose we use is 1000 mg/m2 twice daily, 2 weeks on, 1 week off. This is tolerable for 3 months of CAPOX treatment (4 cycles), but it is not usually tolerable for 6 months. If I ever choose 6 months of adjuvant treatment, particularly for patients with T4 tumors who are performing very poorly, then I will recommend 6 months of FOLFOX. However, if the plan is for 3 months of adjuvant therapy in these patients, I will recommend the standard dose of CAPOX. Based on discussions with our chemotherapy nurses, I also try to avoid placing a port-a-cath, since there are just 4 IV infusions of oxaliplatin, to avoid the associated arm pain.
The IDEA study collaboration has given us a valuable framework for educated discussions with our patients on this topic. We now have very good data that show what we can achieve with varying durations of adjuvant therapy with different regimens.
Manish A. Shah, MD:
This is very true. I certainly appreciate the incredible amount of work and thousands of patients that participated in this pooled analysis. It is truly a once-in-a-lifetime study. We now have the data and options to know for certain that 6 months of treatment is not needed for many patients.
Axel Grothey, MD:
There were numerous interesting rectal cancer studies at ASCO 2020. Three studies that were presented will affect, if not inform, the standard of care for patients with resectable, locally advanced rectal cancer: the OPRA study,9 the RAPIDO study,10 and the PRODIGE 23 study (Capsule Summary).11 All 3 studies showed that total neoadjuvant therapy strategies that include more treatment prior to surgery are the future of rectal cancer management.
The RAPIDO trial was an international phase III trial that randomized 920 patients with stage II and stage III colon cancer with 1 or more high‑risk features (T4a/b status, extramural vascular invasion, N2 disease, mesorectal fascia involvement, or enlarged lateral lymph nodes) to receive either standard preoperative chemoradiation followed by surgery and adjuvant chemotherapy (control arm) or short‑course radiation therapy (5 x 5 Gy), followed by chemotherapy (6 cycles of CAPOX or 9 cycles of FOLFOX for 18 weeks), and then surgery 2-4 weeks later with no further postoperative adjuvant therapy (experimental total neoadjuvant therapy arm).10 It is worthwhile to highlight that in the experimental arm, surgery occurred relatively soon after the end of chemotherapy but was somewhat delayed after the end of radiation therapy.
The primary endpoint of the RAPIDO study was disease‑related treatment failure defined as either distant metastases, locoregional failure, new primary CRC, or treatment‑related death.
Axel Grothey, MD:
Overall, lower rates of disease‑related treatment failure occurred in the total neoadjuvant therapy arm (23.7% at 3 years vs 30.4% in the control arm; HR: 0.75; P = 0.019).10 Significantly fewer patients receiving total neoadjuvant therapy experienced distant metastases (20.0% vs 26.8% in the control arm at 3 years; HR 0.69; P = .005), and there was no significant difference in locoregional failure rates between arms (HR: 1.45; 95% CI: 0.93-2.26; P = .09).
Furthermore, twice as many patients receiving total neoadjuvant therapy experienced a pathologic CR, as determined by evaluation of the resected rectal tumor (28.4% vs 14.3% in the control arm; P < .001), suggesting the possibility of effective nonoperative management even in patients with a very poor prognosis. Analysis of OS was premature, but high rates were observed at 3 years (approximately 89% in both arms).
The investigators also did a very interesting quality-of-life analysis that showed that moving all the treatment upfront did not affect the quality of life of patients compared with the old standard of care.
Axel Grothey, MD:
The investigators of the RAPIDO trial suggested that short-course radiation plus chemotherapy as total neoadjuvant therapy might establish a new standard of care, and I agree. However, it is important to note that the experimental arm of this study had 2 unique factors vs the control arm: 1) a difference in radiation therapy (prolonged vs short course) and 2) a difference in the timing of chemotherapy (postoperative vs preoperative). It is difficult to determine which factor is the driver of the improved outcomes.
The results of the RAPIDO trial combined with the rectal cancer data from the OPRA and PRODIGE trials give promise to individualization of treatment choices for patients with rectal cancer based on various parameters and the potential to have a nonoperative management option for patients who have a CR or clinical CR to preoperative therapy. It was very encouraging to see the total neoadjuvant therapy strategy resulting in the highest rate of responders and the best responses for patients with rectal cancer.
Manish A. Shah, MD:
As you observed, it is important to note that the RAPIDO study introduced 2 variables—the short‑course radiation and the complete neoadjuvant setting of chemotherapy—that potentially contributed to a high pathologic CR rate. I think both components are very important, and we do not yet know the relative value of short‑course vs long‑course radiation.
For low-risk rectal tumors, locoregional treatment and the avoidance of a permanent colostomy are our goals. Our group would, therefore, likely prefer long‑course radiation followed by 4 months of chemotherapy. That being said, I do think that the paradigm that the RAPIDO study highlights is likely to be practice changing.
Axel Grothey, MD:
I think there are several factors that could make short‑course radiation therapy attractive. For instance, consider radiation options in this time of the COVID-19 pandemic. Patients have less exposure to healthcare facilities with a shorter course of radiation (5 days compared with 5.5 weeks with a standard course), along with lower resource utilization and costs.
Treatment for every patient with rectal cancer should be discussed by an interdisciplinary team—including surgeons, radiation oncologists, and medical oncologists—to determine goals of treatment and available treatment options. The team should decide whether patients are going to have a low anterior resection with sphincter preservation or whether they will require an abdominoperineal resection with permanent colostomy. Different treatment decisions would be made for different circumstances. It is fascinating to see how individualized rectal cancer therapy has become.
Manish A. Shah, MD:
Of the pancreatic cancer data presented at ASCO, I was most interested in the data from SWOG S1505, a randomized, open-label phase II study in which 102 patients with previously untreated, resectable pancreatic adenocarcinoma were randomized to receive either modified leucovorin/fluorouracil/irinotecan/oxaliplatin (FOLFIRINOX) or gemcitabine plus nab-paclitaxel before and after surgery (Capsule Summary).12
Studies in the metastatic setting have demonstrated significant improvements in OS for FOLFIRINOX vs gemcitabine (11.1 vs 6.8 months)13 and also for gemcitabine/nab‑paclitaxel vs gemcitabine (8.5 vs 6.7 months).14 FOLFIRINOX and gemcitabine/nab‑paclitaxel have not been directly compared in the metastatic setting.
In the adjuvant setting, modified FOLFIRINOX was shown to be superior to gemcitabine in a phase III study and is now considered a standard option.15 However, gemcitabine/nab‑paclitaxel was not proven to be superior to gemcitabine in a phase III trial.16
The primary endpoint for SWOG S1505 was a “pick the winner” design wherein the 2-year OS rate of each arm was first compared with the historical rate of 40% and then the arms were to be compared if the 2-year OS rates were ≥ 58%. The goal of the study was to determine which regimen would proceed in a subsequent study comparing a neoadjuvant approach vs an adjuvant approach.
Manish A. Shah, MD:
There was no difference observed in the 2‑year OS rates with mFOLFIRINOX and gemcitabine/nab‑paclitaxel (43.1% vs 46.9%).12 The median OS rates were 22.4 and 23.6 months, respectively.
In terms of resection outcomes, 85% of patients in both groups achieved an R0 resection. Patients in the gemcitabine/nab‑paclitaxel group had a slightly higher major or pathologic CR rate (42% vs 25%).
Since neither arm achieved a 2‑year OS rate > 58%, the study did not meet its primary endpoint and the investigators decided not to move forward with a subsequent study comparing these neoadjuvant approaches with adjuvant therapy.
Manish A. Shah, MD:
For me, the take-home message of the SWOG S1505 study is that neoadjuvant therapy is feasible and also that gemcitabine/nab‑paclitaxel is an active regimen and should not be discounted in patients with resectable pancreatic cancer.
This study should be considered along with the ESPAC‑5F study (also presented at ASCO 2020), which was a randomized phase II trial that evaluated neoadjuvant FOLFIRINOX, gemcitabine plus capecitabine, chemoradiation, or upfront surgery in patients with borderline resectable pancreatic cancer.17 This study demonstrated that patients who received neoadjuvant therapy, whether it was chemotherapy or chemoradiation, had better overall outcomes than patients who underwent upfront surgery followed by adjuvant therapy. Specifically, neoadjuvant therapy was associated with a 12-month OS rate of 77% compared with 42% with upfront surgery (HR: 0.28; P < .001). Taking these 2 studies together, I feel more comfortable recommending a neoadjuvant approach for locally advanced pancreatic cancer.
Axel Grothey, MD:
As you noted, SWOG S1505 is the first study to directly compare upfront FOLFIRINOX with gemcitabine/nab‑paclitaxel. Previously, the general clinical opinion had been that FOLFIRINOX is more active in the adjuvant setting than gemcitabine plus nab‑paclitaxel. I was surprised to see how well gemcitabine/nab‑paclitaxel performed in the SWOG S1505 trial. Even though the primary endpoint was not achieved in this study, the results made me more comfortable in recommending upfront gemcitabine plus nab‑paclitaxel, and I no longer consider FOLFIRINOX our only option in this setting.
Manish A. Shah, MD:
I agree. I think that is a very important point for pancreatic cancer treatment: Our patients now have viable options other than FOLFIRINOX, which can be a rather toxic regimen. Recent results from a randomized phase II study of gemcitabine plus cisplatin demonstrated a very high response rate with this regimen in patients with a BRCA1 or BRCA2 mutation, suggesting another group of patients that may not need FOLFIRINOX.18
Axel Grothey, MD:
The point you make about BRCA1/2 patients is very well taken. I, myself, have used gemcitabine and cisplatin in my patients with BRCA mutations and have experienced remarkable success in some cases.
Manish A. Shah, MD:
In the area of upper GI cancers, there was an important phase III study presented at ASCO 2020 that evaluated a novel strategy with trastuzumab plus trimodality treatment (chemotherapy, radiation, and surgery) for patients with resectable HER2-positive esophageal adenocarcinoma. The current standard of care for patients with locally advanced esophageal cancer is based on the practice‑changing CROSS study, which demonstrated that radiation therapy sensitized with weekly paclitaxel and carboplatin prior to surgery results in improved outcomes for both squamous cell and adenocarcinoma.19 The ToGA trial demonstrated that patients with HER2‑positive advanced gastroesophageal cancer experienced a significant improvement in survival rates when HER2-targeting trastuzumab was added to chemotherapy compared with chemotherapy alone.20 Whether trastuzumab is also active in the adjuvant setting in upper GI cancer has been an area of interest.
Studies of adjuvant treatment for HER2-positive breast cancer have shown that trastuzumab is associated with a significant benefit in DFS and OS compared with standard therapy.21 However, gastric and esophageal cancers have been shown to have a more heterogenous expression of HER2 than breast cancers; correspondingly, the activity of HER2-targeted therapy has not been demonstrated to be as strong in upper GI cancers as in breast cancer, meaning many of the drugs that are effective in breast cancer (ie, lapatinib, pertuzumab or TDM1) have not been shown to be effective in gastroesophageal cancers.20,22
RTOG 1010 was a randomized phase III study that evaluated the use of trastuzumab plus trimodality therapy consisting of chemoradiation plus surgery vs trimodality therapy alone in 203 patients with locally advanced esophageal adenocarcinoma, defined as node‑positive or T2 or 3 tumors of the esophagus. Tumors had to be HER2 positive, which was defined similarly to the ToGA study as IHC 3+ or FISH positive. Eligible patients were candidates for potential curative resection (Capsule Summary).23 Trimodality therapy consisted of paclitaxel plus carboplatin for 6 weeks plus concurrent radiation over 5.5 weeks, followed by surgery 5-8 weeks after radiation. Trastuzumab was given as 4 mg/kg in Week 1, then 2 mg/kg weekly in Weeks 2-6 during chemoradiotherapy, and 6 mg/kg prior to surgery. Maintenance trastuzumab therapy was also given at 6 mg/kg every 3 weeks for 13 treatments after surgery.
Manish A. Shah, MD:
Overall, there was no difference in DFS (primary endpoint) or OS with the use of trastuzumab as part of chemoradiation therapy for patients with HER2-positve, resectable gastroesophageal adenocarcinoma.23 Median DFS was 14.2 months without trastuzumab and 19.6 months with trastuzumab, with the survival curves extensively overlapping and an HR of 0.97 (95% CI: 0.69-1.36).
This study provides another example of how HER2 overexpression in upper GI cancers is not the same as HER2 overexpression in breast cancer.
The investigators also performed a DFS multivariate analysis and found that the most important prognostic features of worse DFS in the study were older than the age of 65 years (HR 1.90 vs younger than 65 years; P = .0003) and T3 tumor (HR 1.92 vs T1 and T2 tumors; P = .0058). Typically, we think of nodal status as the most important feature after resection (ie, if the patient is node positive after resection/chemoradiation, they have an 80% chance of recurrence in the first year). However, nodal status was not a significant predictor of DFS in this multivariate analysis.
Manish A. Shah, MD:
Overall, the RTOG 1010 trial is a very important study because a lot of physicians (and patients) were strongly considering the addition of trastuzumab to a neoadjuvant or adjuvant approach (based on the breast cancer data), and I think these results should give everyone pause.
Axel Grothey, MD:
I agree. This is one of those studies that everyone was excited about when it was initiated and many thought that this experimental regimen might become a new standard of care. There was a clear hypothesis and underlying data that suggested the trastuzumab-containing regimen should work. And this is exactly why we need clinical trials.
The RTOG 1010 study also underscored the relevance of HER2 status and differences in tumor biology between breast and gastroesophageal cancers. In contrast to breast cancer, we see that HER2 overexpression is not a real negative prognostic factor in gastroesophageal cancer. It is evident that what we can achieve with trastuzumab and other HER2‑targeted agents in breast cancer is stronger than what we can achieve in HER2-positive gastric and esophageal cancers. I am glad we have these data; this was a very important question to answer and now we must move on.
Manish A. Shah, MD:
Exactly. It also brings up a question about the level of HER2 overexpression in patients in the RTOG 1010 study. We know from the ToGA trial that patients with tumors that highly overexpressed HER2 (IHC 3+) did far better than those with lower levels of HER2 overexpression.20 The RTOG 1010 study enrolled patients who were IHC 3+ or FISH positive, and we know that up to 20% of patients who are FISH positive are likely to be IHC 0 or 1+. I hope that the investigators will do a subsequent analysis looking at the subset of patients who are IHC 3+ to see if there was any benefit in this patient group.
Let’s return to one of the questions we asked at the beginning of the activity.
Manish A. Shah, MD:
That discussion provides an excellent lead-in to our next topic. Two studies were presented at ASCO 2020 related to a novel HER2-targeted antibody–drug conjugate called trastuzumab deruxtecan (also known as T-DXd or DS-8201): DESTINY‑Gastric01, which enrolled patients with HER2-positive, locally advanced or metastatic gastric or GEJ cancer, and DESTINY‑CRC01, which enrolled patients with HER2-expressing, unresectable/metastatic CRC. In both studies, patients had received 2 or more previous regimens. Let’s take a closer look at these studies now.
Manish A. Shah, MD:
The DESTINY-Gastric01 study was a randomized phase II study of T‑DXd vs physician’s choice of chemotherapy (either paclitaxel or irinotecan) for patients with HER2-positive, locally advanced or metastatic gastric or GEJ cancer who had progressed on 2 or more previous regimens (Capsule Summary).24,25 HER2 overexpression was defined as IHC 3+ or IHC 2+ and FISH positive. The primary endpoint of the study was ORR and secondary endpoints included OS, duration of response, and PFS. The study almost exclusively enrolled patients in Japan and Korea.
The study arms of DESTINY-Gastric-01 were well balanced in terms of median age (65-66 years). Most patients had gastric cancer (86% to 89%) vs GEJ cancer, which is typical of what is seen in Asia. Approximately 75% of patients were IHC 3+, and all patients had previously received trastuzumab.
Manish A. Shah, MD:
The ORR with T-DXd was remarkable, with 51.3% of patients receiving this agent experiencing a CR or PR compared with 14.3% of patients receiving chemotherapy (P < .001).24,25 Furthermore, 8% of patients experienced a CR on T-DXd (vs 0% with chemotherapy), so clearly this drug had significant and activity vs standard-of-care chemotherapy.
Manish A. Shah, MD:
The Kaplan-Meier survival curves in the DESTINY-Gastric01 trial were also quite remarkable. Median OS with T-DXd was 12.5 months vs 8.4 months with chemotherapy (HR: 0.59, P = .0097).24,25 This is a clear demonstration of activity. Based on this study, T-DXd was granted an accelerated pathway through FDA review.
Axel Grothey, MD:
It is important to understand the mechanism of action of T-DXd in the context of this study. T-DXd is an antibody–drug conjugate that is essentially a monoclonal antibody with the same amino acid sequence as trastuzumab linked to a topoisomerase I inhibitor by a tumor-selective cleavable tetrapeptide linker. The antibody is internalized and then releases the cytotoxic agent, so this is effectively a targeted chemotherapy. Given this, it is interesting that patients in the DESTINY-Gastric01 trial—who all had pretreatment with trastuzumab—still benefited from this approach. Data, thus far, for second-line and third-line HER2‑targeted agents in gastroesophageal cancer have not been positive. T-DXd is the first HER2-targeted agent to show such impressive responses after pretreatment with trastuzumab in patients with gastric cancer.
Axel Grothey, MD:
The DESTINY-Gastric01 trial noted several cases of interstitial lung disease (ILD) in patients receiving T-DXd (9.6% vs 0% in patients receiving chemotherapy).24 ILD is an adverse event of interest with T-DXd treatment and has been reported in patients with breast cancer in the DESTINY-Breast01 trial who received T-DXd (9% total, with 2.6% developing grade ≥ 3 ILD, all of which resulted in fatal outcomes).26
Manish A. Shah, MD:
The investigators recommended close monitoring for symptoms of ILD and felt that if you identify shortness of breath early, then this adverse event can be managed. Clinicians must be trained to identify these early signs as part of learning how to use this new drug. The current prescribing information for this agent (which is approved for specific patients with metastatic HER2-positive breast cancer as of July 2020) contains a boxed warning for ILD and pneumonitis and recommends holding T-DXd and considering steroids if asymptomatic grade 1 ILD/pneumonitis and discontinuing this agents of symptomatic grade ≥ 2 events.27
Axel Grothey, MD:
Our next study is DESTINY‑CRC01, which was a phase II multicohort trial evaluating T-DXd for patients with HER2‑expressing metastatic CRC who had progressed on 2 or more previous regimens (Capsule Summary).28
The hypothesis of the study was that patients with HER2 expression would benefit from a HER2-targeted agent, such as T-DXd. Based on this rationale, the DESTINY-CRC01 study investigated 3 different cohorts:
All patients were administered 6.4 mg/kg T-DXd every 3 weeks as third‑line or later treatment.
Axel Grothey, MD:
Perhaps the most important finding of the DESTINY-CRC01 trial was that there were no responses in patients with mCRC who were HER2 IHC 2+/FISH negative (cohort B) and those who were HER2 IHC 1+ (cohort C).28 In patients with HER2 overexpression (IHC 3+ or IHC 2+/FISH positive cohort A), the response rate was quite remarkable, with 45.3% of patients experiencing a CR or PR. Approximately 30% of patients in cohort A had received previous anti‑HER2 treatment (with trastuzumab, pertuzumab, T-DM1, or tucatinib) and several showed a response to T-DXd, which is reassuring and mirrors the results we discussed in gastric cancer.
Axel Grothey, MD:
Responses to T-DXd in patients with metastatic CTC with HER2 overexpression (IHC 3+ or IHC 2+/FISH positive) were quite durable and were associated with a median PFS of 6.9 months, which is remarkable for a later‑line treatment setting.28 Median OS was not reached at the time of the data analysis.
Axel Grothey, MD:
Some safety concerns were again raised regarding T-DXd and ILD. In the DESTINY-CRC01 trial, there were 5 patients who developed ILD, with 2 having a fatal event related to ILD.28
Questions remain as to the best management course for these patients. Early identification of this adverse event is critical, and ILD most commonly presents as a dry cough with some shortness of breath. For my patients receiving anti-HER2 agents, I have them walk with a pulse oximeter to determine whether they show early signs of desaturation. Of interest, the symptoms overlap somewhat with what we are seeing with COVID-19, so everyone is currently more aware of these. The CT findings are also very similar, with bilateral opacity in the chest CT scan documenting an interstitial injury in the lungs.
ILD is the most important adverse event to watch for with T-DXd (and other anti-HER2 agents), particularly since it can lead to a fatal outcome. Once the early symptoms are recognized, the agents should be held or discontinued (depending on severity) and corticosteroids should be administered along with close patient follow up.
Axel Grothey, MD:
Overall, the results of the DESTINY-CRC01 trial show great promise and indicate that T-DXd has the potential to become part of our armamentarium of drugs that can be effectively used in biomarker‑selected subgroups. This has currently been examined in a later‑line setting, but it may be possible to use this agent in earlier treatment lines once we gain a better understanding of the ILD complication.
Manish A. Shah, MD:
I agree. Just as we saw in the HERACLES study, which showed good responses with the combination of trastuzumab and lapatinib,29 the DESTINY-CRC01 trial gives us another example that HER2-targeted agents can have efficacy in CRC. The unique mechanism of action of T-DXd, wherein it not only inhibits HER2 signaling but also delivers a toxic payload to the targeted cancer cells, is also promising as it provides an even more targeted approach to treatment. This agent has shown efficacy even for patients who have progressed on previous anti-HER2 therapy.
The issue you highlighted with ILD may make it challenging to combine T-DXd with other agents, but the fact that there is such remarkable activity as a single agent makes this a very compelling future treatment option.
Let’s return to one of the questions we asked at the beginning of the activity.
Back
