Circulating tumor DNA is working to transform care in colorectal cancer and beyond and positivity is actionable at this time.
Axel Grothey, MD
Circulating tumor DNA (ctDNA) is working to transform care in colorectal cancer (CRC) and beyond and positivity is actionable at this time, according to Axel Grothey, MD, who added that persistent ctDNA following surgery is the strongest poor prognostic factor available, and this marker also possesses positive predictive value for disease recurrence. However, several questions need to be addressed regarding its optimal use in oncology.
“ctDNA is going to change the way we monitor and treat patients,” Grothey, a medical oncologist and director of Gastrointestinal Cancer Research at West Cancer Center and Research Institute, said during a presentation at the 6th Annual School of Gastrointestinal Oncology. “ctDNA is not perfect, but it lowers the threshold of where we can detect tumor burden in a patient better than with imaging scans; this is an important point, because we see limitations in tumor detection by imaging, but we lower the threshold of tumor detection via ctDNA. This will change how we deal with cancer—not just in the treatment of established cancer by monitoring minimal residual disease [MRD], but also in diagnosis and population screening.”
During his presentation, Grothey focused on how ctDNA be utilized to monitor MRD, predict disease recurrence, and guide treatment strategies.
To assess for ctDNA as a marker for MRD, two types of tests can be utilized: tumor-agnostic and tumor-informed assays. Tumor-agnostic approaches use a next-generation sequencing (NGS) or polymerase chain reaction (PCR) panel to identify common mutations in CRC, through the use of REVEAL, or methylation markers, through the use of COLVERA. Although this approach may be easier to utilize logistically, it can have a lower sensitivity. Meanwhile, tumor-informed approaches can detect mutations in the primary tumor via NGS or PCR; this can be done with tests like like Signatera Blood Test. Although these tests tend to be more complicated in terms of logistics, they have a high sensitivity and specificity, according to Grothey.
In one study, investigators utilized a tumor-informed assay to examine whether the detection of MRD correlated with disease stage, said Grothey.2 In the study, patients with stage II, T3N0 CRC had a post-surgery MRD rate of 5.6% compared with 28.6% in patients with stage II T4N0 disease. Moreover, patients with stage III, low-risk T1-3N1 CRC had a post-surgery MRD rate of 9.3% vs 39.4% in patients with stage III, high-risk T4, N1-2, T any, N2 disease following surgery.
“This is important because we know it has consequences,” Grothey said. “The presence of ctDNA after surgery is associated with an almost guaranteed recurrence.”
In another study, investigators performed a ctDNA analysis to detect MRD and predict recurrence in patients with stage II colon cancer. Data indicated that among patients who did not receive treatment with adjuvant chemotherapy, those who were ctDNA negative following surgical resection (n = 164) had a 3-year recurrence-free survival (RFS) of 90%.3 All patients who tested positive for ctDNA following surgery (n = 14) experienced disease recurrence within 3 years (HR, 18; 95% CI, 7.9-40; P <.001). “This hazard ratio was highly statistically significant,” noted Grothey.
Moreover, among patients who were clinically low risk, defined as mismatch repair proficient (pMMR) or deficient (dMMR) with no poor prognostic features, those who tested negative for post-operative ctDNA (n = 122) had a higher RFS than those who tested positive for ctDNA (n = 7; HR, 28; 95% CI, 8.3-93; P <.001). Similarly, among patients who were clinically high risk, defined as pMMR plus at least 1 poor prognostic feature, those who were ctDNA negative also had a higher RFS compared with those who were ctDNA positive (HR, 7.5; 95% CI, 2.6-22; P <.001).
“Even if you adjust for high-risk and low-risk factors, you can still see that the presence of ctDNA is highly predictive of what’s going to happen with a patient,” Grothey explained. “It’s clearly a very strong prognostic marker.”
When considering ctDNA and 3-year recurrence prediction, there is 100% positive predictive value with the test, noted Grothey. Of the 8% of patients who were ctDNA positive, none of them did not recur within 3 years. On the other side, the negative predictive value is also strong, according to Grothey, at 91%.
“There is hardly a test in medicine where you have such strong negative and positive predictive values as with using ctDNA,” Grothey said. “We’ve seen the same biologic principle in breast cancer, melanoma, lung cancer, bladder cancer, etc. We are generating a lot of data in many malignancies…One of the critical issues—and this is a very important point—is, does it mean anything and can we reverse the poor prognosis?”
One study examined the predictive value of ctDNA with regard to outcomes in 96 patients with stage III colon cancer who had received adjuvant treatment.4 Similar to patients with earlier-stage disease, those who were negative for post-operative ctDNA had a higher RFS vs those who were ctDNA positive (HR, 3.8; 95% CI, 2.4-21.0; P <.001).
“Patients with [negative] post-operative ctDNA, of course, do better than [those with] positive ctDNA,” Grothey explained. “These are only 96 patients, 20 [of whom] were positive. It’s not a strong database; [this is] not prospective, randomized, or retrospective data…Positive ctDNA was not 100% predictive of tumor recurrence within the first 2 to 3 years…It means we can salvage some patients; we do not always see tumor recurrence in the context of adjuvant therapy.”
Patients who tested negative for ctDNA following treatment with chemotherapy similarly had better outcomes vs ctDNA-positive patients (HR, 6.8; 95% CI, 11.0-15.0; P <.001).
“The predictive value of ctDNA becomes ever stronger when you look at patients after chemotherapy. If patients stayed positive with ctDNA, they have a poorer prognosis than patients who are negative,” Grothey said. “The conversation rate of patients who transition from being ctDNA positive to negative have better outcomes than if they stayed positive over time [HR, 3.7; 95% CI, 1.1-17.0; P = .04].”
Additionally, patients who went from being ctDNA negative to positive experienced poorer outcomes (HR, 6.5; 95% CI, 7.2-642.0; P <.001). “As such, there is a treatment effect on ctDNA with consequence,” noted Grothey.
In a more recent analysis presented during the 2021 Gastrointestinal Cancers Symposium, investigators attempted to classify patients with CRC as being at a higher or lower risk of recurrence based on the presence of MRD.5 Investigators also set out to assess risk of relapse following treatment in patients who were ctDNA positive and determine the lead time of ctDNA detection vs computerized tomography (CT) recurrence. Tumor tissue samples were collected after surgery, and patients were followed for 3 years. Plasma samples were subsequently collected every 3 months, and DNA exome sequencing and ctDNA assays were performed on both tumor tissue and plasma samples. CT scans were done at 12 months and 36 months.
In total, 260 patients were analyzed, 165 of whom had received adjuvant treatment, at a median follow-up of 29.9 months in non-relapsed patients (n = 212). Patients had stage I (n = 4), II (n = 90), and III (n = 166) disease.
Results indicated that ctDNA-negative patients had a higher RFS than those who were ctDNA positive (HR, 11; 95% CI, 5.9-21; P <.0001). Eighty-seven percent of patients who were negative for ctDNA did not experience relapse compared with 20% of those who were ctDNA positive.
After adjuvant chemotherapy, ctDNA-negative patients had a higher RFS and 87.5% did not relapse compared with 16.7% of those who were ctDNA positive (HR, 12; 95% CI, 4.9-27; P <.0001). Longitudinal monitoring yielded similar results, with 96.6% of ctDNA-negative patients not experiencing relapse vs 10.7% of ctDNA-positive patients (HR, 51; 95% CI, 20-125; P <.001).
According to Grothey, the following questions regarding ctDNA and adjuvant therapy still need to be answered:
Several clinical trials that seek to answer these questions are ongoing. Grothey elaborated on 1 particular concept regarding ctDNA-based adjuvant therapy.
“[What do you do with] a patient who has already had surgery, radiation therapy, or adjuvant chemotherapy and how do you utilize ctDNA?” Grothey questioned.8 “You could use the test as a screening tool and randomize patients to have a test or not have a test to then treat those patients who are ctDNA positive with an intensification of treatment, whether it’s chemotherapy or targeted agents.
However, in some cases, such as those with low-risk, stage II cancers, adjuvant treatment may not be administered, according to Grothey. Here, ctDNA may be potentially be utilized to examine its predictive and prospective value by randomizing patients to have a test or not, and then act on that test if it’s positive by using adjuvant treatments such as 5-fluorouracil, leucovorin, and oxaliplatin (FOLFOX) and capecitabine and oxaliplatin (CAPOX).
For example, in the ongoing phase 2/3 COBRA trial (NRG-GI005; NCT04068103), patients with resected microsatellite stable (MSS), stage II colon cancer (n = 1408) will be randomized to either be tested for ctDNA to determine whether adjuvant chemotherapy is needed or to undergo standard-of-care observation.8 Patients who test positive will receive adjuvant FOLFOX, while ctDNA-negative patients will be placed under observation.
“This is a test that will tell us prospectively whether patients will benefit from ctDNA-based adjuvant treatment in the MRD setting,” Grothey said.
For patients with stage III CRC, adjuvant therapy is typically used, according to Grothey. Here, there may be opportunities for the escalation of standard treatments in those who are ctDNA positive and opportunities for de-escalation of such treatments in those who are ctDNA negative. To this end, various concepts are under consideration.
“We actually have just recently approved the ctDNA-informed study CIRCULATE-US or NRG-GI 008, a collaboration between NRG Oncology and SWOG. It will address these different components in stage III colon cancer,” Grothey said.
In the study, patients with T1-3, N1 stage III colon adenocarcinoma will be stratified based on their ctDNA status, with ctDNA-negative patients being randomized to a CAPOX or FOLFOX regimen per investigator decision or serial monitoring of ctDNA over the course of 2 years to determine whether patients remain negative.9 If patients become ctDNA positive, they will go on to receive a FOLFOX or CAPOX regimen or treatment with folinic acid, 5-FU, oxaliplatin, and irinotecan for 6 months in the adjuvant setting.
“That’s not a large study, but it really tries to establish the proof of principle using ctDNA as a decision marker for when and how to treat patients. I’m really excited about the de-escalation [portion of the research],” Grothey said. “…There’s the idea of eventually correlating this with disease-free survival because that is currently the primary end point of adjuvant studies.”
A similar concept is also under investigation in the Australian phase 2/3 DYNAMIC-III study.10 Patients with stage III disease who have undergone surgery will be randomized based on clinical risk factors to receive either standard of care (n = 500), which will be clinician’s choice of no chemotherapy, 6 months of 5-FU and capecitabine, or 3 to 6 months of FOLFOX or CAPOX, or a ctDNA-informed approach (n = 500).
“For [patients who are] ctDNA negative, you de-escalate. For instance, you can go from just using fluoropyrimidine, you go to no treatment or a shorter treatment; you go from FOLFOX to just 5-FU or capecitabine; and you go from 6 months to 3 months of de-escalation in intensity or duration,” Grothey explained. “You can also escalate in the ctDNA-positive group, where you go from no chemotherapy planned to just fluoropyrimidine. You go just 1 step up, which I believe to be a brilliant design. I hope they have enough power in these 1000 patients to really answer these questions that are being asked.”
Additionally, in the phase 3 SU2C ACT3 study (NCT03803553),11 patients with stage III colon cancer are being tested for ctDNA following adjuvant therapy. Investigators will stratify patients based on whether they are ctDNA positive or negative. Patients will be screened for microsatellite instability (MSI), MSS, and BRAF mutations and appropriately treated based on those results. Patients who do not have actionable mutations will be given 5-FU, leucovorin, and irinotecan, while those whose tumors are BRAF-mutated will receive encorafenib (Braftovi), binimetinib (Mektovi), and cetuximab (Erbitux). Patients who are MSI-high will receive nivolumab (Opdivo).
“This is a trial that will try to move immunotherapy and targeted therapy to a select patient population that is ctDNA positive and has actionable tumors,” Grothey noted.
“We know that the consistent presence of ctDNA after surgery is the strongest poor prognostic factors that we have; this is more important than T stage or N stage. We also know that when you use adjuvant therapy, you decrease the likelihood of cancer recurrence in the ctDNA-positive [population]. It’s not a point of no return,” Grothey concluded. “ctDNA positivity is actionable at this time. Some patients [who] have a positive test [can] turn negative and stay negative. The easiest situation that we can potentially act on right now is a situation where you do not recommend adjuvant therapy based on conventional risk factors, but if ctDNA is present, it could lead to the use of adjuvant therapy.”