For patients with stage III colon cancer, prognoses based on risk category have been described in the IDEA trial (EudraCT 2009-010384-16). However, the stratification standards are based on selected risk factors that do not cover all high-risk characteristics that patients may present.

By analyzing the Surveillance, Epidemiology and End Results (SEER) database, a group of researchers sought to assess how risk factors such as pathological T4 disease, perineural invasion, lymphovascular invasion, inadequate lymphadenectomy consisting fewer than 12 lymph nodes removed and poorly differentiated tumors may impact overall survival (OS) as well as cancer-specific survival (CS) in stage III colon cancer.

The study demonstrated that individuals with high-risk characteristics have lower OS and CS than those who do not harbor high-risk characteristics. The results also indicated that patients with high-risk characteristics should be considered at high risk for mortality.

In an interview with Targeted Oncology™, Luv Hajirawala, MD, resident in the Colon and Rectal Surgery Section at LSU Health New Orleans School of Medicine, discussed the relevance of high-risk characteristics in patients with a stage III colon cancer and how recent real-world analysis might change risk stratification in the future.

TARGETED ONCOLOGY: Can you describe the high-risk features of stage II colon cancer and how these features impact outcomes?

Hajirawala: There are many high-risk features, including perineural invasion, which is an invasion of neurons into nervous system structures adjacent to the colon; lymphovascular invasions; poor differentiation at presentation; advanced stage T, which corresponds to the growth of the tumor through the wall; having fewer than 12 lymph nodes; and insufficient margins.

In stage II colon cancer, these have been associated with a poor prognosis for patients. So some of the patients who have stage II disease without these, we tend not to give them adjuvant therapy, because surgery is curative, but when patients have these high-risk features, they are at risk higher rate of recurrence or unfavorable survival. Therefore the [depending on the] arrange [for] patients with these risk factors, we will choose to administer chemotherapy. It is very important to understand how these factors affected patients with stage III disease.

How have previous studies explored this topic?

First, oncologists need to understand that our incidence rates are going down. We are doing something with advanced screening practices, and new cancers are being discovered and we are treating them. Our incidence of colon cancer has also decreased and our survival has been improved. But what we’re seeing is that an equal number of patients have localized disease, which is stage I and II, and locally advanced disease, which is stage III at the time of presentation. So when you look at all the patients who are diagnosed with colon cancer in a population, when we diagnose them, an equal proportion, about 35%, is localized. Surgery is therefore curative, and an equal number of them have locally advanced, stage III disease, and chemotherapy is the mainstay of treatment after surgery for these patients.

Thus, our understanding of risk in this population is evolving. More recently, if you look at the NCCN [National Comprehensive Cancer Network], our current risk stratification in these patients depends on TNM staging. The IDEA trial established this high-risk group of patients with stage III colon cancer. What they found was that patients with T4 disease or N2 disease had a poor prognosis and benefited from longer adjuvant treatment with FOLFOX. And that’s kind of where we are in the NCCN guidelines for risk stratification. We have no data on stage III disease and how these characteristics affect outcomes, in addition to our normal high risk stratification, which is T4 or N2.

How did you conduct the SEER database analysis of mortality risk in patients with stage III colon cancer?

We reviewed our SEER database between 2010 and 2017 and identified patients with stage III colon cancer. We first compared patients by stratification based on high-risk characteristics. We therefore divided the group into 3 categories, which included stage III colon cancer with no risk factors, stage III colon cancer with 1 risk factor, and stage III colon cancer with 2 or more. And those risk factors included inadequate lymphadenectomy, meaning fewer than 12 nodes removed during surgery; high-grade disease; poor differentiation; perineural invasion; and T for disease.

What we wanted to examine was how does the presence of high-risk features affect overall patient survival and cancer-specific survival?

What were the main conclusions of the analysis?

Patients who had 1 high-risk feature had lower overall and cancer-specific survival than patients who did not. And when we looked at more than 2 high-risk characteristics, this overall and cancer-specific survival was even lower. What we wanted to know then was, is it because of a single risk factor, or do they all have a combined effect on this one?

We then examined the impact of individual risk factors on overall and cancer-specific survival in patients with stage III disease. We found that fewer than 12 lymph nodes, inadequate lymphadenectomy, high-grade disease, and perineural disease all have a similar impact on overall and cancer-specific survival. These factors bring survival down to 65% in this range, and T4 disease had a more pronounced effect on this. T4 disease brings your survival down to 57%.

We also found that patients who had more than 2 risk factors, their survival was nearly 44% for overall survival and 51% for cancer-specific survival, which is lower than the lowest conferred survival. by any high-risk individual characteristic. So what can be summarized is that if a patient has more than 2 risk factors, the effect on survival is cumulative.

Were there any notable subgroup results?

In the stage III group, we also stratified the patient into a low-risk and high-risk group. This was done based on the IDEA trial, which is currently the gold standard for risk stratification in stage III disease. We therefore examined the impact of high-risk features in low-risk stage III cancer and found that the presence of one or more high-risk features had both an impact on overall and cancer-specific survival. .

Then we repeated this analysis for a patient who had a high-risk disease, and we found the same thing. We found that the presence of at least one high-risk feature had a negative effect on overall and cancer-specific survival, which is cumulative. This means that the highest stats have an impact on survivability.

How do you interpret these results?

These results imply that our current risk stratification based on TNM status is probably not sufficient to provide adequate prognostic information. Our current risk stratification is based on T4N2 disease, but what we find is that despite your overall tumor nodes and metastatic disease status, high-risk features have unfavorable survival. Thus, high-risk features or the presence of more than 2 high-risk features should be considered a high-risk population in Stage III disease.

How should future research assess risk stratification for stage III colon cancer?

Further studies need to focus on these patients to see which chemotherapeutic agents or adjuvant treatments would be beneficial. We wanted to watch [in] especially in the SEER database, so we were only able to look at some high-risk features. There are other high risk features that are not captured in this database. So we would like to examine the impact of these. I think in the future we want to characterize all the high-risk characteristics and risks that may predict survival for these patients because ultimately cancer treatment is going to be personalized. And I think each patient, based on their high-risk characteristics, should receive appropriate treatment that may not be applicable to someone else. So we want to find the best therapy and the best individual therapy that we can offer for stage III colon cancer. I think it’s a bit far, but we’re making progress.


Hajirawala, LN, Yi Y, Bergeron MA, et al. 6315 – Are a subset of patients at higher risk of mortality in stage III colon cancer: an analysis of the SEER database. Presented at: 2022 AACR Annual Meeting; April 8-13, 2022; New Orleans, LA. Abstract 6315.