Colorectal Cancer Update GI Think Tank 2007

DR NEAL J MEROPOL: The management of Stage II colon cancer is a difficult decision-making scenario, insofar as the prognosis is good with surgery alone and the potential benefit of adjuvant therapy for the population as a whole is marginal. If you were to treat all cases of Stage II colon cancer with adjuvant therapy, the absolute benefit would probably be in the range of five percent or less for the most active chemotherapy regimen.

The challenges are in defining which patients are at the greatest risk of relapse from the group of patients with Stage II disease and in selecting those patients for adjuvant therapy because their potential for benefit is greater.

Patients have different values with regard to the tradeoffs of the potential benefits and side effects. This requires a discussion about the option of adjuvant therapy — the potential hazards, which are well defined, and the potential benefits, which are less well defined for any individual.

The doctor and patient have to come to an agreement and understanding about what is best for that individual. With regard to the question of whether all patients with Stage II colon cancer should be referred to a medical oncologist, my answer is that all of those patients should engage in a discussion about adjuvant therapy and whether it’s right for them.

Colorectal Cancer Update GI Think Tank 2007

DR HOWARD S HOCHSTER: I believe that every patient with Stage II cancer would benefit from seeing a medical oncologist. Even if they decide against adjuvant therapy, they reap other benefits, such as discussions about the risk of colorectal cancer for relatives and how they can be screened.

Patients also need to know how their health-related issues for the next 25 years will be different as a result of the cancer, particularly in terms of future screening, so that if he or she is the one patient in five who develops a recurrence, we can capture it when it’s likely to be curative.

Click on the image to enlarge

Colorectal Cancer Update 2006 (5)

DR CHARLES S FUCHS: In deciding about adjuvant chemotherapy off protocol, particularly for patients with Stage II disease, I use the clinical features we are familiar with, such as perforation and obstruction and the number of lymph nodes sampled.

I try to pool together patients I don’t believe would benefit from adjuvant therapy, those with whom I’m comfortable using a fluoropyrimidine alone and those for whom I would use FOLFOX.

In patients with higher-risk disease — those with few lymph nodes analyzed or those with adherence to or invasion of adjacent structures who might have obstruction or perforation — I’m comfortable using FOLFOX. I have to admit, however, that I’m still willing to use fluoropyrimidine monotherapy for patients with more standard-risk disease, although I know some of my colleagues routinely use FOLFOX in all circumstances.

Although the proportional benefit of FOLFOX is fairly consistent across patients with Stage II or Stage III disease, I also want to consider the absolute benefits, in a particularly low-risk setting. Is the addition of oxaliplatin, with its inherent neuropathy risk, necessary in patients for whom the absolute benefit is not so great?

Colorectal Cancer Update 2007 (4)

DR AXEL GROTHEY: The key issue surrounding lymph node sampling is that we need to make sure that we identify Stage II patients as Stage II patients with the adequate number of sample lymph nodes — that’s the first step. So we talk about appropriate diagnosis and appropriate staging. The more lymph nodes we evaluate, the more likely we are to correctly classify cases as Stage II or Stage III.

By definition or by convention, we normally assume 12 lymph nodes are adequate in terms of a sample. However, this is not a clear-cut threshold. This is more or less a gray zone over numbers. I believe the more lymph nodes we have, the more certain we are that our staging is appropriate, and therefore, the better we identify prognosis.

Beyond the number of lymph nodes that are found in the specimens, we’ve also learned that the better the ratio is between affected lymph nodes and recognized lymph nodes, the better the prognosis is.

A big difference is evident in the patient who has two positive nodes out of 12 lymph nodes or the patient who has two positive nodes out of 45 lymph nodes for disease. The lymph node ratio is actually one of the most important prognostic factors for a patient with Stage II or Stage III disease.

Colorectal Cancer Update 2007 (5)

DR CATHY ENG: The fewer lymph nodes dissected, the more likely you’ve had an inadequate surgical evaluation or pathological evaluation. Ideally, there should be 12 nodes evaluated. At MD Anderson, we prefer to evaluate a minimum of 14 lymph nodes.

If a patient has an inadequate number of lymph nodes sampled, I consider that to be a high-risk factor. If a patient has four lymph nodes dissected and one of those is positive — so 25 percent of the dissected nodes are positive — I will follow that patient more closely than a patient who had one of 23 lymph nodes positive.

I might perform a CAT scan on that patient twice a year because the chance of a recurrence may be much higher based on the small number of lymph nodes dissected.

Cancer Conference Update 2007 (Post-ASCO Edition)

DR LEONARD B SALTZ: If a patient with colorectal cancer has fewer than 12 lymph nodes evaluated, I consider that a poor-risk factor. If the number of lymph nodes sampled is not adequate to give me confidence that the disease is node-negative, I err toward aggressive treatment.

I’m not certain that is the right thing to do in all cases, but that’s a discussion I have with patients. This issue has raised my anxiety level enough to make me believe that oxaliplatin is worth considering for these patients.

Colorectal Cancer Update GI Think Tank 2007

DR MEROPOL: When oncologists were asked nationally in a survey, “What are you likely to recommend to a woman who has a 10 or 20 percent risk of relapse from breast cancer?” the vast majority said they would likely recommend chemotherapy, but in colon cancer, far fewer doctors said they would treat a patient with colon cancer and the same risk of relapse.

I believe part of this is cultural and part of it is data driven. With regard to the data in breast cancer, prospective randomized studies involving thousands of patients have conclusively shown and defined the small benefit from the addition of adjuvant therapy for an individual who’s at low risk of relapse.

In colon cancer, we don’t have those kinds of data. We have extrapolations from a higher-risk situation and pooled analyses that either show no benefit or marginal benefit.

Click on the image to enlarge

Colorectal Cancer Update 2007 (4)

DR DANIEL J SARGENT: From the FDA’s perspective, no trial has ever demonstrated that chemotherapy is better than observation for patients with Stage II disease. Although the QUASAR trial for patients with Stage II disease has been presented in abstract form at ASCO, it still has not been published.

Therefore, when the FDA was presented with the MOSAIC trial of FOLFOX versus 5-FU/leucovorin that showed a nonsignificant benefit in the subgroup of patients with Stage II disease, the FDA also considered that no other data exist showing that treatment benefits patients with Stage II disease. Based on this, the FDA did not consider an interpolation justified.

That was the FDA’s perspective, but Dr Grothey and I wrote an editorial in the Journal of Clinical Oncology criticizing that decision. Our feelings were based on a fundamental paradigm with clinical trials, which is that, in the absence of compelling data, the best result is the overall result.

The overall results should be based on all patients in the trial — that result is consistent with the prospectively designed, planned analysis of the clinical trial.

Based on that paradigm and the fact that in MOSAIC the relative risks of relapse for patients with Stage III disease (HR = 0.76) and Stage II disease (HR = 0.80) were similar and the formal test result for interaction was highly nonsignificant — suggesting that the benefit of treatment was the same for Stage III disease as it was for Stage II — we did not see any compelling reason to go against the fundamental principle of clinical trials, which is to use the data from the entire trial.

Click on the image to enlarge

Colorectal Cancer Update GI Think Tank 2007

DR NORMAN WOLMARK: I believe that from a biologic standpoint, we have no reason to think that Stage II patients form a unique subset relative to their responsiveness to adjuvant therapy. It’s just that they’re at lower risk for recurrence. The heterogeneity between breast and colon cancer does not lie in the tumor or in the patient. Patients want to be treated for the same low risk whether they have breast or colon cancer. The heterogeneity lies in the fact that, traditionally, the medical oncologist who specializes in colorectal cancer is less enthusiastic about adjuvant therapy.

The irksome part from my perspective is not that all patients with Stage II disease should be treated — it’s that all patients with Stage II disease should be apprised of the benefits of adjuvant chemotherapy. I believe what we need to resolve this issue is a tool that allows us to evaluate patients beyond the traditional factors we’ve used to decide which patients with early-stage colorectal cancer to treat. We need an assay equivalent to the Oncotype DX™, and I believe we’re making significant progress relative to that.

Colorectal Cancer Update 2007 (3)

DR ALAN P VENOOK: We enroll patients on ECOG-E5202 (Figure 7), which I believe is an incredibly important study. Patients with Stage II disease are risk stratified based on the molecular features of their cancer.

Patients at low risk — who are expected to make up approximately 60 percent of the patients — are observed. Patients at high risk — deletion on 18q and/or microsatellite stability — receive FOLFOX or FOLFOX/bevacizumab.

I believe that’s such an absolutely important study to distinguish who doesn’t need chemotherapy and to see whether the data hold up.

Colorectal Cancer Update 2007 (4)

DR GROTHEY: The key recent data set discussing adjuvant therapy is the MOSAIC trial, which compared 5-FU/LV to FOLFOX. It was first presented at ASCO in 2003. This was a large trial with approximately 2,300 patients, 40 percent of whom had Stage II disease, and 60 percent had Stage III disease.

At the initial presentation, the trial had already met its primary endpoint of improvement in disease-free survival overall for the patients who received FOLFOX. This was also seen clearly in a subgroup analysis for Stage III disease, but not necessarily for Stage II disease.

In 2003 we didn’t know whether a benefit in disease-free survival would translate to an overall survival benefit. However, at ASCO in 2007 with a median follow-up of six years, the primary endpoint — disease-free survival — was still solid, and we saw that overall survival was improved for the patients who received FOLFOX.

We now need to closely examine the subgroups of Stage II disease and Stage III disease. A significant effect is clear for patients with Stage III disease, but interestingly, unselected patients with Stage II disease did not benefit in terms of overall survival, and only marginally in terms of disease-free survival with the use of FOLFOX over 5-FU/leucovorin.

My takeaway from the MOSAIC trial is that it clearly validates the use of FOLFOX as adjuvant therapy for patients with Stage III disease but not for all patients with Stage II disease. A group of patients with high-risk Stage II disease may benefit from oxaliplatin, surely in terms of disease-free survival and perhaps even overall survival, but it’s probably a little more complicated. This illustrates that we still need to research the idea of the optimal therapy for Stage II colon cancer.

Colorectal Cancer Update 2007 (4)

DR AIMERY DE GRAMONT: In the update of the MOSAIC trial, the disease-free survival advantage that was observed for FOLFOX4 at three years’ follow-up was confirmed at five years. At ASCO 2007, I presented the overall survival data with a minimum of six years’ follow-up in all patients, and the benefit was observed only in patients with Stage III colorectal cancer.

Click on the image to enlarge

Among patients with Stage II disease, no benefit in survival is evident. The population was a mix of patients with low-risk and high-risk disease. We could define an advantage in the high-risk population in terms of disease-free survival, but even in this population we could not observe a survival advantage. However, the numbers were small and these were exploratory analyses.

It’s good news that in the MOSAIC trial, patients with Stage II colorectal cancer — high-risk disease and low-risk disease — have a six-year probability of survival of 87 percent. The overall relapse rate was nearly 10 percent, and a 3.8 percent difference in disease-free survival between the two treatment arms was evident.

Colorectal Cancer Update 2006 (5)

DR GROTHEY: Recently, we have moved from 5-FU to FOLFOX as the standard in the adjuvant setting for colon cancer. So the new question is how to integrate the biologic agents that we know work in the palliative setting — cetuximab and bevacizumab — into the adjuvant setting.

Both of these biologics have demonstrated efficacy in established colorectal cancer tumors, but it has not been determined whether they play any role in the adjuvant setting.

We can’t immediately transfer all our knowledge and our experience from the palliative setting into the adjuvant setting. This was demonstrated by the relative failure of irinotecan to show benefit in the adjuvant setting, whereas in the palliative setting FOLFOX and FOLFIRI are equivalent.

Cetuximab is an EGFR antibody, and bevacizumab is a VEGF inhibitor. Both are currently being tested in the United States in cooperative group trials. NSABP-C-08 evaluated FOLFOX with or without bevacizumab as adjuvant therapy for Stage II and III colon cancer. NCCTG-N0147 is evaluating FOLFOX with or without cetuximab in the adjuvant setting.

ECOG has a Stage II trial (E5202, Figure 7) for patients at molecular high risk as defined by microsatellite stability status and loss of heterozygosity at the 18q chromosome. Those patients will be randomly assigned to FOLFOX with or without bevacizumab — analogous to the NSABP-C-08 trial.

Colorectal Cancer Update GI Think Tank 2007

DR VENOOK: I believe one of the important endpoints of the NSABP-C-08 trial will be determining whether there are any long-term consequences from using bevacizumab for a year in patients with Stage II or Stage III colon cancer. Meanwhile, we should be cautious about assuming that long-term adjuvant use of bevacizumab is beneficial for patients.

Colorectal Cancer Update for Surgeons 2007 (1)

DR WOLMARK: If you evaluate the NSABP-C-07 trial, which compared adjuvant FLOX versus 5-FU/LV in patients with Stage II and Stage III colon cancer, you see that Grade III toxicities are 50 percent versus 41 percent, and they’re largely gastrointestinal.

The most troubling toxicity, of course, is the neurotoxicity, especially Grade III, which causes persistent pain or interference with daily life. At one year, the rate of Grade III neurotoxicity in patients who received the FLOX regimen was approximately 0.5 percent in this study.

On the other hand, if you ask patients to self-report toxicities, you see lingering toxicities even 12 months down the line, particularly in the lower extremities. While this is not trivial, adding oxaliplatin makes a statistically significant difference and is likely to save lives. It’s a tradeoff, but quite frankly, we anticipated greater toxicity with oxaliplatin than what we observed in NSABP-C-07.

Colorectal Cancer Update 2007 (5)

DR RICHARD M GOLDBERG: One of the questions I’m asked by other physicians who treat colorectal cancer is, “What about the use of oxaliplatin for the patient who is older and has diabetes?” My feeling is that in most patients, the neurologic toxicity associated with oxaliplatin is not an on-off switch — it’s something that is “dialed up” over time.

Dan Sargent and I conducted a meta-analysis of four different trials comparing patients older than 70 to those younger than 70. We examined the MOSAIC trial, two first-line advanced-disease trials — 9741 and the de Gramont original study — and the Rothenberg trial in the second-line setting.

What we found across the board in all three settings was that patients benefited equally from the addition of oxaliplatin, regardless of age. Toxicity was minimally elevated, and it was white blood cell counts and thrombocytopenia that were affected. These were laboratory parameters, not clinical parameters. Except for the second-line setting, there was an advantage of equal value to younger and older patients in terms of FOLFOX therapy.

The people enrolled in these trials were the best of the older patients. Taking the data from them and applying them to the patient who walks in your office every day is a little more complicated. My hope is that this study will free oncologists to think more liberally about the use of FOLFOX in older patients.

In my own practice, if I’m worried about someone, I’ll start with LV/5-FU. If they tolerate that, I’ll either add full-dose oxaliplatin or, if I’m being cautious, administer 60 mg/m² of oxaliplatin initially. If they pass that test, I’ll go up to the full dose. I don’t believe that when you first see the patient you have to make a decision, “I’m going to administer FOLFOX or I’m going to administer just 5-FU.”

Click on the image to enlarge

Click on the image to enlarge

Colorectal Cancer Update GI Think Tank 2007

DR VENOOK: When you use Adjuvant Online!, as the age of the patient increases, the absolute benefit of adjuvant therapy decreases because of competing causes of mortality. In my opinion, age is a factor in your decision-making and in how aggressively you follow patients.

However, age by itself should not be a reason not to offer patients treatment. One problem with the clinical trials is that we get our trial data on 80-year-old “Olympic athletes,” and whether that’s relevant to the patient in your office is a different question.

Colorectal Cancer Update 2007 (5)

DR ENG: I have colleagues who use capecitabine and oxaliplatin in the adjuvant setting. We now know with the NO16966 study that FOLFOX and CAPOX were equivalent in the metastatic setting. That may also apply in the adjuvant setting. As with any oral chemotherapy, you must be cautious and make sure they’re adherent to your recommendations and learn how to treat toxicities.

In terms of tolerability, capecitabine is good for select patients. If patients have diarrhea and they don’t take the supportive medications, they can end up dehydrated and in the hospital. You also have to be concerned about hand-foot syndrome, so I try to see these patients every other cycle, to follow them closely.

Colorectal Cancer Update 2006 (5)

DR FUCHS: In the CALGB-C89803 adjuvant study of 5-FU/LV versus IFL, which was a negative trial in terms of the primary endpoint, we provided patients with a 16-page questionnaire on diet and lifestyle. Patients completed the questionnaire, which was validated through various studies such as the Nurses’ Health Study. In the CALGB trial, the surveys were administered midway through adjuvant therapy and about six months after its completion.

Compliance with the questionnaire was excellent — approximately 95 percent of the patients completed it. Physical activity was highly protective and associated with a significant improvement in disease-free survival. The more physically active the patient, the better the disease-free and overall survival rates.

The physical activity took place during and after chemotherapy. If the patients walked an average of approximately six hours per week, they had a 47 percent improvement in disease-free survival. One might argue that the physically active patients were healthier and the patients who were inactive had occult cancer. We considered that and repeated the analysis by excluding all of the events in the first six months after completing the questionnaire, and we found the same results.

We also repeated the analysis for the patients with colon cancer in the Nurses’ Health Study who had completed the same questionnaire, and the findings were identical.

Select Publications