CHICAGO — Neoadjuvant chemotherapy for operable colon cancer failed to reduce the risk of disease recurrence, but was associated with significantly fewer surgical complications and improved surgical outcomes versus adjuvant therapy, a randomized trial showed.
Patients who received chemotherapy before surgery had a 2-year recurrence rate of 13.6% as compared with 17.2% for patients who had adjuvant chemotherapy. The difference met the prespecified reduction in the hazard ratio (0.75), but did not achieve statistical significance because of a lower-than-expected event rate in the control arm, reported Matthew T. Seymour, MD, of the University of Leeds in England, at the American Society of Clinical Oncology (ASCO) annual meeting.
“The trial did not reach target significance for the primary endpoint; in that sense, it was a negative trial,” said Seymour. “However, we have shown with some confidence that moving 6 weeks of chemotherapy ahead of surgery … was safe. We saw less surgical morbidity, significantly downstaged tumors, and a reduced rate of incomplete resection of tumors. We saw a strong trend toward improved 2-year cancer control.”
“On the basis of this, we feel that this approach can be considered a new therapeutic option for locally advanced, operable colon cancer,” he added.
Invited ASCO discussant Jordan Berlin, MD, of Vanderbilt University Medical Center in Nashville, acknowledged what the study showed (feasibility, safety, and less surgical morbidity) but pointed out some unresolved issues. For example, almost half of the control arm (49%) had stage I or II disease, a group of patients for which chemotherapy is not indicated except for high-risk cases. Neoadjuvant chemotherapy could pose a risk of overtreatment for that group.
“The study is not definitive in terms of efficacy, but it is promising and certainly opens up this neoadjuvant approach as a treatment option,” said Berlin. “This is not a practice-changing study in terms of standard of care. A neoadjuvant approach can be considered for colon cancer without causing detriment. It is safe, effective, and appears to increase the percentage of patients who receive at least one dose of chemotherapy.”
In his introduction, Seymour pointed out that no advances in postoperative therapy for colon cancer have occurred over the past 15 years, since the approval of oxaliplatin. Neoadjuvant chemotherapy has been used successfully in several types of solid tumors and offers several potential benefits: a reduced risk of incomplete surgical resection; earlier treatment of micrometastases; and an opportunity to assess tumor response and developed personalized treatment strategies.
No large-scale clinical trials had compared neoadjuvant and adjuvant chemotherapy for operable, locally advanced colon cancer. To address the lack of data, investigators in the National Institute for Health Research Clinical Research Network performed a multicenter randomized FOxTROT trial. The study compared outcomes with a 6-month course of postoperative oxaliplatin-based chemotherapy versus 6 weeks of neoadjuvant therapy, followed by surgery and an additional 18 weeks of adjuvant chemotherapy.
The primary endpoint was the frequency of relapse or persistent disease during the first 2 years after surgery. Investigators assumed a 2-year treatment failure rate of 25%-32% for the control arm and hypothesized that neoadjuvant therapy would reduce the rate by 25%. On that basis, they enrolled and randomized 1,052 patients 2:1 to pre- and postoperative therapy or to adjuvant therapy only.
Within the neoadjuvant arm, investigators optionally enrolled patients in a phase II randomized placebo-controlled trial of panitumumab (Vectibix) prior to initiation of preoperative chemotherapy. Limited to patients with KRAS-wild type tumors, the study-within-a-study occurred before the availability of extended KRAS testing, said Seymour.
The study population had a median age of 65, and men accounted for two-thirds of the patients. More than 98% of the patients had WHO Performance Status 0-1, primary tumor location was evenly distributed between the right and left side, and three-fourths of the patients had CT scan-predicted T4/T3 tumors (≥5 mm extramural extension).
The 3.6% absolute difference in the primary endpoint translated into an HR of 0.75, but the confidence intervals crossed 1.0 (95% CI 0.55-1.04, P=0.08). The 5-year recurrence rates were 27% for the control arm and 21% for the neoadjuvant arm.
Attempted curative surgery was successful in 98% of patients in both treatment groups. Significantly, more patients in the control arm did not receive planned chemotherapy (27% vs 4%, P<0.0001), either because the patient was too sick or refused treatment (11%) or the tumor was considered low risk (16%). Seymour did not dwell on chemotherapy-associated toxicity, noting that "there were no surprises."
Investigators had some concerns that neoadjuvant chemotherapy might lead to more surgical morbidity, but that proved not to be the case. Most complications occurred with similar frequency in both arms, but several types of events occurred less often in the neoadjuvant arm:
- Anastomotic leak/intra-abdominal abscess: 4.7% vs 7.4%
- Complications requiring additional surgery: 4.3% vs 7.1%
- Complications leading to prolonged hospitalization: 11.6% vs 14.3%
Fewer patients receiving preoperative chemotherapy went to surgery but had no resection (0.3% vs 1.1%), had macroscopically incomplete surgery (0.3% vs 1.1%), or had microscopically incomplete surgery (4.2% vs 8.8%), leading to a higher rate of microscopically complete surgical resection (93.1% vs 88.4%). Collectively, the differences resulted in a significant advantage for the neoadjuvant group (P=0.001).
Neoadjuvant therapy led to substantial tumor downstaging, including more patients with pT0 (4.1% vs 0%) and pT1/pT2 (11.7% vs 5.8%) at surgery and fewer patients with pT4 disease (20.5% vs 29.8%). The differences represented another significant advantage for neoadjuvant therapy (P<0.0001). Nodal stage at surgery also favored neoadjuvant chemotherapy (P<0.0001), including half as many patients with positive apical nodes (3.8% vs 7.5%, P=0.013).
Results with Panitumumab
A sensitivity analysis showed that use of panitumumab did not explain the effects observed in the primary analysis, said Seymour. Subgroup analysis showed that patients whose tumors had mismatch repair (MMR) deficiency had no tumor regression overall, even though a few patients had pathologic complete responses. In contrast, patients with MMR-proficient tumors had a trend toward a lower recurrence rate at 2 years (P=0.082).
During a panel discussion that followed the presentation, Seymour noted that 4% of patients in the control group had stage I disease at surgery and 20% had stage II disease without any high-risk characteristics. Chemotherapy would not be indicated for those patients.
“I think we have to ask ourselves whether 6 weeks of a relatively low-risk chemotherapy regimen … where we see good surgical morbidity data … is a reasonable thing for that small percentage of patients who might otherwise not get chemotherapy,” he said. “We have to balance that against what we do in the clinic every week, which is to expose patients to postoperative chemotherapy knowing that over 80% of them will not benefit because either they are already cured or because they will relapse anyway.”
“Personally, I think treating that 24% of patients with 6 weeks of chemotherapy that we might not otherwise have used seems quite a reasonable thing to do,” he said.
The study was supported by the University of Birmingham.
Seymour disclosed a relevant relationship with Amgen.