SAN FRANCISCO – Advances in molecular prognosis may soon make it possible to predict the risk of recurrence of early colorectal cancer after curative resection – and thus the need for adjuvant therapy in patients who might not otherwise receive further treatment.
Five-Gene Tumor Signature
A new five-gene tumor prognostic signature has performed well in external validation testing among patients with stage I or II disease, according to Dr. Peter F. Lenehan, chief medical officer of Everist Genomics Inc., in Ann Arbor, Mich.
Dr. Lenehan and his coinvestigators validated the gene signature among 115 patients who had undergone curative resection for stage I or II colon cancer or stage I rectal cancer and did not receive adjuvant therapy. Some 40% experienced a recurrence within 3 years.
For predicting recurrence, the signature had a sensitivity of 70%, a specificity of 55%, a positive predictive value of 51%, and a negative predictive value of 73%. The overall accuracy was 61%.
Other prognostic tests can differentiate patients who will not recur, Dr. Lenehan said, but "we alone are best able to predict who is going to recur. For this population, we believe this is more clinically significant because the default position is not to treat stage I and II.
"So if we are able to detect those that actually need more aggressive follow-up or high consideration for adjuvant chemotherapy, then our test is better suited for this patient population."
Patients with a signature-predicted high risk of recurrence were more than twice as likely to have a recurrence as were their counterparts with a signature-predicted low risk (hazard ratio, 2.06; P = .02).
In contrast, the combination of criteria suggested by the National Comprehensive Cancer Network (NCCN) (T4 stage, fewer than 12 lymph nodes examined, grade of 3 or 4, lymphovascular invasion, bowel obstruction, localized perforation, or close, indeterminate, or positive margins) did not significantly predict risk.
"We are the first prognostic test in colorectal cancer to target stage I. The others target stage II and III," Dr. Lenehan added.
"The importance for us is that 10%-15% of stage I colorectal cancer patients actually recur and succumb to their disease," he said. "The standard of care for this stage is no treatment. And therefore, you have 3,000 or 4,000 patients per year which are potentially missing treatment that could potentially help them."
The signature could also be used to select only high-risk stage I patients for a clinical trial of adjuvant therapy, which would dramatically reduce the number of patients needed to determine benefit, he added.
GCC Expression in Lymph Nodes
Expression in lymph nodes of mRNA of the enzyme guanylyl cyclase C (GCC) predicts recurrence in patients with stage II colon cancer, according to Daniel J. Sargent, Ph.D., a professor of biostatistics and of oncology at the Mayo Clinic in Rochester, Minn.
"GCC is a marker that is only present in the colon, and if you find the GCC marker in the lymph nodes, it indicates that the tumor has metastasized from the colon to the lymph nodes," he explained in an interview. "So the presence of GCC in the lymph nodes should be a negative prognostic marker, and that’s indeed what we found."
The investigators retrospectively studied 241 patients who underwent resection for stage II colon cancer, did not receive adjuvant therapy, and had a minimum follow-up of 36 months. With a median follow-up of 60 months, 12% had a recurrence or cancer-related death. Some 10-41 lymph nodes (median, 15) were evaluated per patient.
The GCC assay results were expressed as a lymph node ratio (number of positive nodes divided by total number of informative nodes). In all, 35% of patients had a lymph node ratio of greater than 0.1.
These patients had more than double the risk of recurrence, compared with their counterparts who had a lower ratio (HR, 2.38; P = .02). The estimated 5-year rate of recurrence was 27% for the former and 10% for the latter.
The findings were even stronger in the subgroup of 181 patients with highly favorable clinical and pathological factors – namely, a T3 tumor, at least 12 lymph nodes examined, and negative margins – a third of whom had a lymph node ratio of greater than 0.1 (HR, 5.06; P = .003). Here, the estimated 5-year rate of recurrence was 27% with a ratio greater than 0.1 and just 4% with a lower ratio.
In a multivariate analysis among all patients that included potential confounders (age, T stage, number of lymph nodes assessed, and mismatch repair status), a GCC lymph node ratio greater than 0.1 was associated with a 2.61-fold higher risk of recurrence (P = .02).