MRI has ‘big upside’ predicting which patients with rectal cancer could avoid surgery
Click Here to Manage Email Alerts
Key takeaways:
- MRI can help determine which individuals with rectal cancer may not need surgery following neoadjuvant therapy.
- Those who achieved complete response likely can forego total mesorectal excision.
MRI can help identify which patients with rectal cancer could avoid surgery based on response to neoadjuvant therapy, according to study results.
Researchers reported 5-year DFS of more than 80% among individuals who achieved complete response as determined by MRI, indicating these patients may benefit from a watch-and-wait approach.
“Many patients who show complete response on MRI [after] chemoradiation can likely forego total mesorectal excision. We can hopefully take that immediate option off the table,” Arun Krishnaraj, MD, MPH, professor of radiology and medical imaging, and chief of the abdominal imaging section at University of Virginia, told Healio. “There’s a big upside about MRI’s use in rectal cancer.”
Background
In the 1990s, colorectal cancer had been the fourth-leading cause of cancer deaths for men and women younger than age 50 years. Now it is the leading cause of cancer death among men in this age group and the second leading cause among women behind only breast cancer, an American Cancer Society report showed.
Total mesorectal excision has improved outcomes for many people with rectal cancer, but patients then require a lifelong colostomy bag and could have sexual dysfunction, too, Krishnaraj said.
Over the last several years, more patients have received neoadjuvant therapy such as induction chemotherapy and consolidated chemoradiation.
Digital rectal examination, flexible sigmoidoscopy and MRI all have been used to evaluate tumor response after treatment. However, 25% of those deemed to have achieved complete response have local regrowth and 8% develop distant metastases, according to study background.
An MRI-based tumor regression grade system has not incorporated advances in practice, such as diffusion-weighed imaging and lymph node evaluation.
“Eighteen leading cancer centers from across the United States decided to pool their data on MRI and asked, ‘Are any of the findings on the MRI helpful in predicting those who have complete response, near-complete response and an incomplete response?’” Krishnaraj said. “Can we preselect those patients who would likely benefit from just a watch-and-wait approach and avoid surgery? Given all the incumbent risks and associated complications that can occur, and quality-of-life issues with surgery, we felt like this would be a big boon to patients to be able to know.”
Methods
Krishnaraj and colleagues investigated these questions using data from the Organ Preservation in Rectal Adenocarcinoma (OPRA) trial.
The OPRA trial — a prospective study in which investigators analyzed oncologic outcomes among individuals undergoing watch-and-wait surveillance — included 324 patients with stage II or III rectal adenocarcinoma. Researchers randomly assigned participants to one of two neoadjuvant treatment regimens.
Krishnaraj and colleagues included 277 adults (median age, 58 years; interquartile range, 17 years; 64.6% men) who underwent restaging MRI following therapy into their analysis.
They stratified those patients into three categories based on whether they had achieved complete response (41.9%), near-complete response (45.1%) or incomplete response (13%).
Mean time between treatment completion and MRI was 8 weeks (standard deviation, ±4).
Outcomes at 5 years such as DFS, OS and local regrowth served as primary endpoints.
Results
Results showed 5-year total mesorectal excision survival of 45.9% (95% CI, 40.2-52.5) in the entire cohort.
Researchers reported higher 5-year total mesorectal excision survival (65.3% vs. 41.6%; P < .001) and lower rates of local regrowth (24.4% vs. 36.6%; P = .02) in the complete response group than the near-complete response group.
Results showed 5-year DFS of 71.3% (95% CI, 65.7-77.2) in the entire cohort.
Researchers observed a higher 5-year DFS rate among those who had achieved complete response (81.8%) than those with near-complete response (67.6%)or incomplete response (49.6%; P < .001).
Results revealed similar trends for 5-year OS (92.9% vs. 79.7% vs. 67.8%; P < .001) and 5-year distant RFS (86.4% vs. 76.8% vs. 61.8%).
About half (48.5%) of the 266 participants who had at least 2 years of follow-up had residual disease. MRI features independently associated with residual disease included restricted diffusion (OR = 2.5; 95% CI, 1.22-5.24) and abnormal nodal morphologic features (OR = 5.04; 95% CI, 1.43-23.9).
Researchers acknowledged study limitations, including the small number of patients with incomplete response. They also cited lack of data about extramural vascular invasion and tumor deposits, both of which may predict outcomes and deserve more research, Krishnaraj said.
‘Future is bright’
Krishnaraj called the results “encouraging” but emphasized MRI’s ability to stratify patients could be enhanced with endoscopy and physical exam.
“This study was designed to look at the impact of MRI’s predictive ability in isolation,” he said. “If you couple that with endoscopy and physical exam, it would likely be more powerful in being able to predict 5-year DFS.”
AI also could improve evaluation. MRIs can take 30 to 45 minutes to “obtain all the sequences,” Krishnaraj said. “Many people are working on techniques and AI that will improve the efficiency and speed at which MRI images can be acquired. That’s going to benefit patients because they’re not going to have to sit in the scanner as long.
“The other area that I am excited about is there is likely findings that are either below the threshold of the human eye to see, or very subtle findings that even the experienced reader overlooks, that the machine-learning algorithm may detect,” he added.
Krishnaraj described himself as “bullish on the technological innovations in radiology in general.”
“I think the future is bright with the advances in processing power, in the actual MRI machines themselves, and with the software tools that are going to improve the field across the board,” he said.
For more information:
Arun Krishnaraj, MD, MPH, can be reached at ak8jj@uvahealth.org.
Collapse
Click Here to Manage Email Alerts