By Kate O’Rourke
Washington—At the 2013 annual meeting of the Society of Surgical Oncology, two surgeons discussed the following controversial topic: In patients with rectal cancer who receive neoadjuvant chemoradiotherapy, when is local excision or nonoperative management most appropriate?
Philip Paty, MD, attending surgeon in the Colorectal Surgery Service at Memorial Sloan-Kettering Cancer Center in New York City, discussed nonoperative management of patients with T3/T4 rectal lesions after a clinical complete response (cCR) from neoadjuvant chemoradiotherapy.
Studies have demonstrated that after neoadjuvant chemoradiotherapy, 10% to 45% of patients with rectal cancer have a pathologic complete response (pCR), and when this occurs, 95% of these patients are cured (Ann Surg Oncol 2012;19:384-391; Lancet Oncol 2010;11:835-844).
A 1995 study demonstrated that in rectal adenocarcinoma, nearly all patients who achieve a cCR after primary external beam radiotherapy, do so by eight months after initiation of radiation treatment and 60% by four months (Int J Radiat Oncol Biol Phys 1995;312:255). Disappearance of the tumor is necessary for cure and evolves over time. “If you are going to practice nonoperative management, you have to wait,” said Dr. Paty. “You have to allow time for the tumor to regress fully after radiotherapy.”
So, what are the criteria for determining a cCR? “If you can reach the tumor with your finger, it should be a completely flat mucosa, no nodularity, no mass,” said Dr. Paty. “It’s okay if there is a scar with some smooth induration, but it has to have a very benign feel to it.” Visually, he said, there should be normal, flat mucosa, with or without a pale scar, and there can be telangioalatsia. “What is not clear is whether a small ulcer is an exclusion criteria,” said Dr. Paty. “For me, it is.”
Dr. Paty highlighted data from four published series of patients who underwent nonoperative treatment of T2/T4 rectal lesions after achieving a cCR from neoadjuvant chemotherapy. Patient characteristics were “all over the map” with variable T stage, nodal stage and size (Table). “Smaller tumors are going to have a higher CR rate, but even larger tumors can melt away,” said Dr. Paty.
Dr. Paty pointed out that when managed nonoperatively, the vast majority of cCR patients will avoid rectal resection within the first five years. “Local failure rates range from 10% to 25% and most occur early, within the first 18 months after radiotherapy,” he said. Nearly all of the local failures can be salvaged with R0 [curative] resections.
In a study from the Netherlands, only one of the 21 nonoperatively treated patients had a local recurrence and had surgery as salvage (J Clin Onc 2011;29:4633-4640). In a consecutive series of 49 patients receiving neoadjuvant chemoradiotherapy in the United Kingdom, 12% demonstrated a cCR and had been managed without surgery (Colorectal Dis 2012;14:567-571).
In a Brazilian study, patients with stage 0 distal rectal cancer, who were nonoperatively managed after achieving a cCR, were matched to patients who received surgery and achieved a pCR. five-year overall survival (OS) rates (100% vs. 88%) and disease-free survival (DFS) rates (92% vs. 83%) favored the nonoperatively managed patients (Ann Surg 2004;240:711-717).
An update from this study highlighted the need for long-term follow-up. “The local recurrences seen in the rectum at 5% had a mean time interval of over four years, so there is a long tail to the recurrence curve. This is the most concerning aspect,” said Dr. Paty. “Do we really know what the late failure rates are? There [aren’t] enough data in the literature to assess this accurately.”
In a Memorial Sloan-Kettering Cancer Center study, 32 patients with stage I to III rectal cancer were treated nonoperatively after achieving a cCR after chemoradiation; these patients were compared with a control group of 57 patients who underwent rectal resection and had a pCR (Ann Surg 2012;256:965-972). The two-year distant DFS (88% vs. 98%; P=0.27) and OS (96% vs. 100%; P=0.56) were similar for the nonoperative group and the control group. Rectal resection was successfully avoided in 81% of the patients who were treated nonoperatively.
Assessment of cCR in the four studies was variable, including time to assessment, which ranged from two to 12 months. In the Brazilian study, clinicians waited a full year before declaring sustained clinical responses (Ann Surg 2004;240:711-717).
Other issues in assessing a cCR include the role of biopsy; the role of imaging; assessment of lymph-node metastases; trigger to operate (any mass, viable cancer on biopsy, lack of regression and local tumor progression); and the role of local (suspicious scar), small residual cancer. Clinical assessment with proctoscopy is essential, but other practice patterns vary. “There is really no standardization in imaging, but MRI is emerging as the test of choice,” said Dr. Paty. “There are too many false-positives on PET scans, and CT scans are not sensitive enough to detect small residual tumors.” Biopsies are not always reliable and can miss disease. “Most viable tumors are actually deep in the surface rather than at the surface,” he said.
One critical question is whether clinicians should be looking for the complete disappearance of a tumor or simply for its regression. For anal cancers, deferring surgery as long as patients aren’t progressing is reasonable, he said, but clinicians should look for continued regression for an adenocarcinoma.
While nonoperative management is controversial in itself, Dr. Paty said, there are other reasons surgeons aren’t likely to jump on board soon. “You have to follow patients, you are at medical-legal risk, and it is a lot more work. It is much easier to operate,” he said. He examines his nonoperative patients every three months in the first year, every four months the second year, and then every six months up to five years.
Heidi Nelson, MD, professor of surgery, Mayo Clinic, Rochester, Minn., discussed local excision with T1/T2 rectal lesions. She highlighted the case of a a 37-year-old man with no family history of rectal cancer who presented with an 6.5-cm rectal lesion that was soft, flat, 1.5 cm from the top of the anal canal, and negative for microsatellite instability. The colonoscopy was negative for other lesions, and a biopsy revealed an invasive grade 2 carcinoma with mucinous features. Metastatic workup was negative; both MRI and ultrasound suggested a T1N0 anterior lesion.
“Based on the size of the lesion, his young age and the mucinous features and histopathology, we chose to go ahead with an ultra lower anterior resection. It turned out to be T1N1, stage III disease,” said Dr. Nelson. “This shows the challenge of these cases.”
Candidates suitable for local excision are elderly patients who have a limited life expectancy or a serious medical condition. Tumors should be located below the peritoneal reflection, less than 2 or 3 cm in size, not amenable to lower anterior resection or coloanal, amenable to full-thickness excision and have confirmed negative margins. If a tumor has features of lymphovascular invasion, mucinous features or signet ring features, clinicians should be cognizant that local excision is not the best choice in most cases.
Most patients prefer local excision, but at what price? A study in the journal Cancer (1989;637:1421-1429) examined the relationship between depth of tumor and influence of lymph nodes and showed that local excision will most likely miss lymphatic disease.
In 2000, researchers at the University of Minnesota, Minneapolis, published data on 108 patients with early rectal cancer. The study analyzed local failures with local excision compared with standard resection (Dis Colon Rectum 2000;43:1064-1071). Local recurrence rates were higher in the local excision group overall (28% vs. 4%) and for T1 lesions (18% vs. 0). OS also was worse for those receiving local excision (69% vs. 82%), but slightly less so with T1 lesions (72% vs. 80%).
This helped spur a National Cancer Database (NCDB) analysis of 35,179 rectal cancer patients between 1989 and 2003. In a subanalysis of 2,000 patients, researchers identified a difference in disease-specific survival (DSS), favoring standard resection (97% vs. 93%; P=0.004), but no difference in OS.
Patients should be educated about expected results for T1 lesions treated with local excision and then individualized treatment. “If [patients] are more concerned with cancer recurrence risks, they will likely prefer standard resection, but if they are concerned with having a permanent stoma, then they may opt for local excision,” said Dr. Nelson.
Selection criteria, however, need to be much more restrictive for T2 lesions. An NCDB analysis identified a survival advantage for standard resection (77% vs. 68%; P=0.01) and although DSS was not statistically different between the two groups, one-fourth of patients who had local excision had a local failure. “The survival data and local failure data should really bring caution to us,” Dr. Nelson said.