A 65-year-old woman diagnosed with a T3N2 colon cancer treated with resection and 10 cycles of adjuvant FOLFOX, switched to 5-FU/LV alone for the final two cycles secondary to Grade II neuropathy. Six months postsurgery, a rise in CEA levels was linked to a positive CT finding of new bilobar hepatic metastases, including a 2-cm lesion in segment VI and a 3-cm lesion in segment III. A PET scan confirmed the findings, but ruled out extrahepatic disease and she immediately underwent metastectomy of both the right and left small hepatic lesions without complication. She received six months of postoperative capecitabine.
Case Discussion:
DR HALLER: This 65-year-old woman was diagnosed with T3N2 colon cancer four years ago and received adjuvant FOLFOX. After 10 cycles, the oxaliplatin was discontinued because of Grade II sensory peripheral neuropathy and she received an additional two cycles of LV/5-FU-2.
We then put her on our surveillance schedule — which means performing a CEA every three months and CT scans taken of the abdomen and pelvis every six months, typically for five years. By that time, at least 85 to 90 percent of all recurrences will have occurred, so continuing the same schedule after that point becomes less productive.
Three months ago, the patient’s CEA level seemed to be elevated at 6.5, but CT was negative. Certainly, CEA levels can reach 6.5 during treatment, because of chemotherapy-induced fatty liver and the effect on the excretion into the bile duct of CEA. However, this was a little far out to be elevated, especially since she had been negative until then.
We repeated the CEA three months later and it rose to 10.3, which is pretty close to the level at which it’s never a false-positive. CT then showed a 2-cm lesion in segment VI and a 3-cm lesion in segment III, so this was bilobar hepatic disease. A PET scan showed these lesions to be FDG-avid, but there were no other abnormalities seen in the liver, abdomen or chest.
We also performed a colonoscopy, which showed no new lesions. In a case like this, if the patient has not had a colonoscopy within a year and we are considering a liver resection, we conduct the study to ensure we are not missing a new primary colon tumor, which is pretty common in people who’ve had prior colon cancer.
The patient then went to surgery for a hepatic resection. Whether one should use neoadjuvant therapy in these patients is debatable. This patient’s disease was clearly resectable, so conversion therapy or so-called downstaging was not needed.
If preoperative therapy is not necessary for downstaging, then what are you accomplishing? If you look critically at the data, number one, you’re exposing patients to chemotherapy that increases the morbidity of surgery. Number two, although people make the argument that nonresponders maybe should not be taken to the operating room, that data concern patients who are treated with conversion therapy. It’s clear that patients with unresectable disease that progresses won’t fare well no matter what surgery you perform.
Another argument is that patients might experience disease progression during chemotherapy, making their previously resectable disease nonresectable. However, in the EORTC trial, that hardly ever happened, because you’re only treating with six cycles of therapy. At the worst, the patients end up undergoing surgery no matter whether their disease progressed or responded. So, it doesn’t help you as a decision tool with regards to taking patients to surgery or not.
Whether patients with resectable disease should have surgery up front and then receive chemotherapy remains controversial, and I’ve seen a fair amount about it in the literature with commentary by Nick Petrelli and others. The general belief is, if you will, that the EORTC study was a proof-of-principle trial showing that some systemic therapy is better than surgery alone, but it didn’t establish neoadjuvant therapy as a treatment standard.
In the Journal of Clinical Oncology, a combined analysis was published of two trials evaluating postoperative 5-FU and leucovorin. Neither reached final accrual, but the data were virtually superimposable with the neoadjuvant study data. Therefore, many of us take patients with resectable disease to the operating room first and then administer postoperative chemotherapy.
Now the interesting issue that remains is what to treat the patient postoperatively. If this patient had not already received chemotherapy, we probably would have administered up to 12 cycles of postoperative FOLFOX. However, she had residual neuropathy secondary to this regimen and we didn’t want to worsen her quality of life. So, in this case one might ask, why not administer FOLFIRI? The problem is that FOLFIRI was used in the PETTAC-3 trial for Stage III disease and it was a negative study. It’s never been published, but hopefully it will be soon.
The ACCORD trial in high-risk Stage III colon cancer was a negative trial for FOLFIRI as well. Also, a study presented at ASCO in 2008, but not published, compared LV/5-FU-2 versus FOLFIRI after liver resection and showed no benefit. They observed a slight delay up to one year, at which point the curves came together.
We would have treated with LV/5-FU-2, but the port had been removed and she simply didn’t want to return to the pump. So we treated with capecitabine in a manner similar to the X-ACT trial, the Phase III study comparing capecitabine with standard intravenous 5-FU/LV in the adjuvant treatment of early stage colon cancer.
