Summary
Researchers have identified histologic biomarkers that could guide treatment decisions about duration of adjuvant therapy in stage III colorectal cancer (CRC). Results were presented at the ASCO 2020 Annual Meeting (Abstract 4065) and published in the Annals of Oncology (2020;31:487-494).
Specifically, researchers found that low tissue-infiltrating lymphocytes (TILs) density and high tumor budding independently predicted shorter disease-free survival in patients with stage III CRC treated with FOLFOX for 6 months. In addition, combining these two variables further increased prognostic precision in both high- and low-risk individuals.
To explore additional resources offered by Oncology Times and to subscribe, visit https://journals.lww.com/oncology-times/pages/default.aspx
Original Article
Biomarkers Identified to Guide Treatment Duration in Stage III Colorectal Cancer
Oncology Times
By Veronica Hackethal
Researchers have identified histologic biomarkers that could guide treatment decisions about duration of adjuvant therapy in stage III colorectal cancer (CRC). Results were presented at the ASCO 2020 Annual Meeting (Abstract 4065) and published in the Annals of Oncology (2020;31:487-494).
Specifically, researchers found that low tissue-infiltrating lymphocytes (TILs) density and high tumor budding independently predicted shorter disease-free survival in patients with stage III CRC treated with FOLFOX for 6 months. In addition, combining these two variables further increased prognostic precision in both high- and low-risk individuals.
“Our data underscore limitations of T and N staging and demonstrate the ability of budding/TILs to enhance prognostication beyond anatomic tumor staging,” noted senior author Frank A. Sinicrope, MD, FACP, Professor of Medicine and Oncology at the Mayo Clinic & Mayo Comprehensive Cancer Center.
TILs are part of the ant-tumor response, while tumor budding refers to four or fewer tumor cells at the tumor’s invasive margin. Low TIL density and high tumor budding have been linked to metastasis. Both TILs and tumor budding can be evaluated using routine histopathologic techniques, making it a cost-effective approach that could potentially be applied to routine clinical practice.
“Since all patients were treated with FOLFOX for 6 months, our data do not provide predictive information for these biomarkers,” Sinicrope said. “Our data suggest that budding/TILs warrant evaluation in patients who received 3 versus 6 months of adjuvant chemotherapy. Such data have the potential to inform recommendations for the duration of adjuvant chemotherapy in stage III disease.”
The research team is currently using digital pathology and image analysis software to enable comparison and ensure reproducibility of their findings.
Current management of CRC is based primarily on the Tumor, Node, Metastasis (TNM) staging system. Yet patients often show substantial variability within stages, which may affect outcomes. For example, the landmark IDEA study found that treatment for 3 months with CAPOX may be as effective as 6 months for patients with stage III CRC, especially for lower-risk individuals. However, disease-free survival was improved among patients treated with FOLFOX for 6 months, especially for patients with high-risk disease. Overall, shorter duration of therapy was associated with fewer adverse events. The results provided support for a risk-based approach for determining duration of adjuvant therapy, pointing to the need for biomarkers that could better inform treatment decisions (N Engl J Med 2018;378:1177-1188).
To identify such factors, researchers analyzed data from 1,532 patients with stage III surgically resected CRC who had received 6 months of FOLFOX alone, or combined with cetuximab as part of the Alliance phase III clinical trial (JAMA 2012;307:1383-1393). Because the trial had found no statistically significant difference in disease-free survival with the addition of cetixumab, researchers pooled data from both groups for the purposes of this analysis. However, they divided participants into low-risk (T1-3, N1; n=804) and high-risk (T4 or N2; n=729) groups, according to the risk classification strategy proposed by the IDEA trial.
Researchers used microscopy on routine biopsies to quantify TIL densities, tumor budding, and micropapillary features (small clusters of five or more malignant epithelial cells that have been associated with tumor progression).They determine optimal cutoffs based on 5-year disease-free survival (DFS) in randomly selected training (n=766) and validation (n-766) sets. They also analyzed biopsies for mismatch repair deficiency (MMR) status, as well as KRAS and BRAF mutations. Results were adjusted for age, T/N stage, primary tumor sidedness, histologic grade, KRAS/BRAF, MMR, performance status (PS), and treatment. Median follow-up was 83 months.
Overall, tumors with high budding/low TIL density had the poorest DFS (60%), while tumors with low budding/high TIL density had the best DFS (78%).
Furthermore, low versus high TIL density was significantly associated with worse DFS in the training group (HR=2.01; 95% CI, 1.41% to 2.85%; P<0.0001), and confirmed in the validation group (HR=1.53; 95% CI, 1.06% to 2.21%; P=0.0196).
Likewise high versus low tumor budding was significantly associated with worse DFS in training and validation groups (HR=1.56; 95% CI, 1.19% to 2.05%;P=0.0002; HR=1.24; 95% CI, 0.94 to 1.64; P=0.0307, respectively).
A combined high budding/low TIL variable predicted DFS better than either variable alone (HR + 2.07, 95% CI 1.50-2.88, p<0.0001). The high budding/low TIL variable was significantly linked to DFS in both low risk (HR 1.59; 95% CI, 1.02-2.48; p=.0273) and high risk (HR 2.82; 95% CI, 1.72-4.63; p<.0001) individuals.
Overall, the relative contribution of the budding/TIL variable to predict DFS was second only to N stage. In the high risk group, the TIL/budding variable was the most important contributor. In the low-risk group, the TIL/budding variable ranked second to KRAS (see tables).
The presence of micropapillary features was significantly associated with DFS in the training set (HR=1.43; 95% CI, 1.08% to1.91%; P=0.0158), but not in the validation set (HR=0.94; 95% CI, 0.68% to 1.31%; P=0.7232).
| Variable/Predictor/Prognosticator | Percent (%) |
|---|---|
| KRAS | 45.5 |
| Budding/TILS | 24.4 |
| Treatment | 7.2 |
| MMR | 6.1 |
| Performance Status | 5.0 |
| BRAF | 5.0 |
| Age | 3.8 |
| Sidedness | 2.3 |
| Histologic Grade | 0.8 |
| Variable/Predictor/Prognosticator | Percent (%) |
|---|---|
| KRAS | 45.4 |
| Budding/TILS | 13.0 |
| Treatment | 12.0 |
| MMR | 10.4 |
| Performance Status | 9.1 |
| BRAF | 5.1 |
| Age | 2.8 |
| Sidedness | 2.2 |
Veronica Hackethal is a contributing writer.