Atrayee Basu-Mallick, MD, discusses the emerging role of ctDNA as a predictive biomarker in early-stage CRC and the growing importance of molecular testing in the advanced-stage setting.
Atrayee Basu-Mallick, MD
Circulating tumor DNA (ctDNA) has demonstrated utility as a biomarker in patients with early-stage colorectal cancer (CRC), said Atrayee Basu-Mallick, MD. However, prospective data are needed to solidify ctDNA as a prognostic and predictive marker for intensified or de-escalated treatment in this patient population.
“[Making] decision changes based on ctDNA will help us to better [determine] the predictive value [of ctDNA testing],” said Basu-Mallick. “In the future, we will hopefully be able to escalate or de-escalate care based on ctDNA because it has been shown to be such an important and reliable prognostic marker. It also might help us in trying new drugs to get quicker answers from studies.”
Although ctDNA is still under study, broad molecular testing for KRAS, BRAF, and HER2 mutations, microsatellite instability–high (MSI-H) status, and even NTRK fusions, has been established as a critical component of treatment selection for patients with advanced-stage CRC, Basu-Mallick explained. As the armamentarium for HER2-directed therapies, for example, becomes more crowded, early efficacy data and adverse effect (AE) profiles should also be taken into account when selecting a frontline regimen without comparative data between regimens.
In an interview with OncLive® during an Institutional Perspectives in Cancer webinar on gastrointestinal (GI) malignancies, Basu-Mallick, a clinical assistant professor of medical oncology at Thomas Jefferson University Hospital, discussed the emerging role of ctDNA as a predictive biomarker in early-stage CRC and the growing importance of molecular testing in the advanced-stage setting.
Basu-Mallick: [ctDNA] could be a very good prognostic marker. [Data from a study presented by Tenna V. Henriksen, a PhD candidate at Aarhus University, at the 2020 ASCO Virtual Scientific Program, showed] that patients who had ctDNA-positive [disease] after surgery had a hazard ratio of 11 in terms of risk of recurrence. These data showed that chemotherapy worked. However, if we go by the trial, the reduction with chemotherapy was about 20% as opposed to what we traditionally [see], at about 30% to 40%.
We need a study to be done prospectively. Although the samples [from the Danish trial] were collected prospectively, they were analyzed retrospectively. [Additionally,] no treatment decisions were made based on ctDNA results. Future trials have to be done prospectively with [treatment] decision changes; some of these are already ongoing.
One is the COBRA trial [NCT04068103], which is looking into an escalation approach for patients with stage IIA colon cancer who are not recommended otherwise for adjuvant therapy.
The other trial is the NRG-GI008 trial [CIRCULATE], which will enroll patients with lower-risk stage III [disease]. It will evaluate therapy escalation for ctDNA-positive [patients] and therapy de-escalation for those who are ctDNA negative.
We’ve had quite a few years without any significant development in early-stage CRC. It’s very important for us in the GI oncology community to encourage patient enrollment on the NRG trials because [it’s too early] to use ctDNA in a prognostic or predictive manner. It would be best to have more patients enrolled in these trials so that we have definitive answers to convert these important ctDNA data into a prognostic marker.
Right now, [it is being evaluated] purely in the early-stage setting.
Any patient with advanced CRC should definitely have molecular testing done because 40% of [patients are] KRAS wild type. [KRAS testing] is always done, but we have to go beyond that. We want to [identify] the 8% of patients who have BRAF mutations, the 3% of patients who have HER2-amplified tumors, and the 8% of [patients with] MSI-H [disease].
Additionally, although it is less than 1% [of patients, we want to identify] NTRK [fusions] because the responses patients with NTRK fusions have are remarkable.
In CRC, about 40% of patients have KRAS wild-type [disease], for which we have cetuximab [Erbitux] and panitumumab [Vectibix].
About 8% of patients have BRAF mutations. We have seen that these patients do poorly and have a median survival of 13 months compared with 30 months in the [BRAF] wild-type population. This population is relatively chemotherapy refractory, or even if [chemotherapy] works, it only works for a short time. The combination of encorafenib [Braftovi] and cetuximab [is available], as well as the triplet drug combination [with binimetinib (Mektovi)]. However, we are mostly not using [the triplet] given that the data are almost equal with the doublet and triplet.
For BRAF-mutant patients, everybody in the community is hopeful that [the BEACON CRC trial regimen of encorafenib, binimetinib and cetuximab will] move to the first-line setting. We have to wait for the ANCHOR-CRC data to mature, but we are all looking forward to [seeing that happen].
In terms of MSI-H patients, the PFS2, which is time from randomization to second-line treatment, also showed that it favored the immunotherapy arm. We can [interpret] that in 2 ways. Almost 59% of patients in the chemotherapy arm had crossed over to another immunotherapy. That tells us that possibly the overall survival [OS] will not reach statistical significance because of the crossover. When we get the OS data, it will tell us whether using [pembrolizumab] in the first-line setting is more important than using it in the second-line setting. That is something that we can look forward to because sometimes even the sequencing of immunotherapy is important; sometimes tumor biology changes if a tumor is exposed to more chemotherapy.
The trastuzumab deruxtecan study was unique in the sense that it allowed patients to have prior HER2-directed therapies. This was the first time that patients who had had prior HER2-directed therapy could get further HER2-directed treatment. Whether patients were treated with HER2-directed therapy before or not, the ORR [objective response rate] was similar at 43% vs 45%, [respectively].
In general, HER2-directed therapies are showing response rates of anywhere from 30% to 50%. We know that trastuzumab [Herceptin] plus pertuzumab [Perjeta], as well as trastuzumab plus lapatinib [Tykerb], showed good ORRs of 30% to 40% in patients with HER2-positive [CRC].
The first important thing to note is that we consider HER2-directed therapy only in patients who are KRAS wild-type because we don’t have any data on responses in KRAS-mutated patients. Number two is that if a patient is HER2 amplified or HER2 positive, the recommendation, based on response data, is that we do not go to third- or fourth-line standards of care, such as regorafenib [Stivarga] or TAS-102 [trifluridine/tipiracil; Lonsurf] because those [drugs induce] response rates of less than 2% [in that population]. HER2-directed therapy in a KRAS wild-type patient who is HER2 positive has [shown] response rates of 30% to 50%.
Based on data so far, the first-line HER2-directed therapy would be [physician’s] choice of trastuzumab plus pertuzumab, trastuzumab plus lapatinib, or even trastuzumab plus tucatinib [Tukysa], which could be an option too once more data come out. As a provider, I would choose [therapy] based on the AE profile. Almost all [the regimens are associated with] diarrhea, but [the regimens] are a little bit different in terms of other AEs.
I would keep trastuzumab deruxtecan for later after initial HER2-directed therapy because it was tested in that setting. We want to use the few weapons we have in our hand judiciously. I would use [trastuzumab deruxtecan] as a second-line HER2-directed therapy also because of the concern for grade 2 to grade 5 interstitial lung disease [(ILD) associated with that agent]. Up to 2.6% of patients in the DESTINY-CRC01 trial had grade 5 ILD, which is concerning. Based on the AE profile, any of those combinations [could be considered] in the first-line setting and trastuzumab deruxtecan [in the] second-line setting.