Colorectal Cancer Update 2007 (2)

DR SALTZ: The XELOX-1/NO16966 study confirms prior studies that showed bevacizumab increases progression-free survival. I believe it justifies my continued feeling that bevacizumab is an appropriate component of first-line chemotherapy, except for patients who have a significant contraindication, such as a history of significant arterial thrombotic events, serious wound-healing issues and so on.

As with other trials, it indicates nothing about whether bevacizumab should be continued in multiple lines of therapy. The revised package insert for bevacizumab says it is approved for first- or second-line therapy — it does not say first- and second-line therapy — so it is my practice to use bevacizumab in one line of therapy. I use it in first-line therapy unless there is a contraindication. If that contraindication is resolved so that bevacizumab becomes appropriate to use in second- or third-line therapy, then I might consider it.

Click on the image to enlarge

Colorectal Cancer Update 2007 (5)

DR GOLDBERG: In determining which therapy to administer first line in the metastatic setting, I first try to enroll patients on the CALGB-80405 trial, which is a study of dealer’s choice, FOLFOX or FOLFIRI, with either cetuximab, bevacizumab or both.

If a patient is unwilling to participate in the trial, I’ll tell him or her that FOLFOX and FOLFIRI provide basically equivalent outcomes and that, in general, we’re adding bevacizumab to first-line therapy. I don’t believe the data from the XELOX/NO16966 trial presented at ASCO will change my first-line approach off study. I will still offer patients FOLFOX with bevacizumab as first-line therapy.

The XELOX/NO16966 trial evaluated XELOX versus FOLFOX with or without bevacizumab. The most surprising factor was that no difference in response rate appeared when they added bevacizumab. Modest differences in survival of about a month were evident. Does that mean that bevacizumab is not worth adding? Not to me. I still believe it’s worth adding bevacizumab to FOLFOX, but I’m watching for additional information to either reinforce or change my opinion.

Click on the image to enlarge

Colorectal Cancer Update 2006 (5)

DR GROTHEY: From an efficacy point of view, FOLFOX and FOLFIRI are not different in the palliative setting. I believe whether you use FOLFOX or FOLFIRI is “a wash.”

Outside of a clinical trial, I talk to my patients about which toxicity they would prefer. It’s either the neurotoxicity or the higher risk of developing early-onset diarrhea. We don’t have direct comparisons of FOLFOX/bevacizumab versus FOLFIRI/bevacizumab, but we can make some indirect conclusions from the BICC-C trial data and the TREE trial data. The TREE trials and the BICCC trial were similar. The TREE-1 and BICC-C trials both started the year before bevacizumab was approved, so both trials did not include bevacizumab in their first phase, and both trials evaluated what is the best fluoropyrimidine, in combination with either oxaliplatin in the TREE trials or irinotecan in the BICC-C trial.

After bevacizumab was approved in the United States, the TREE trial, which evaluated oxaliplatin combinations, was amended to include bevacizumab in all three treatment arms. The capecitabine dose in arm three was reduced to account for toxicities observed in TREE-1, and the BICC-C trial discontinued the capecitabine and irinotecan arm and added bevacizumab to FOLFIRI and IFL.

What we have now is a cross-trial comparison. When you compare FOLFOX and FOLFIRI with bevacizumab in TREE-2 and in the BICC-C trial, it’s interesting to see that the progression-free survival in both trials — a cross-trial comparison for FOLFOX with bevacizumab and FOLFIRI with bevacizumab — was 9.9 months, exactly identical. The response rates were 54 to 55 percent, almost identical.

We have data on FOLFOX with bevacizumab in terms of overall survival, which was 26 months in TREE-2. For FOLFIRI with bevacizumab, the endpoint for overall survival in the BICC-C Phase II trial had not yet been reached, but the survival curve suggested that it will be beyond two years. So we have similar data on FOLFIRI and FOLFOX using almost all efficacy parameters.

Click on the image to enlarge

Colorectal Cancer Update 2007 (5)

DR HERBERT I HURWITZ: In addressing the question of whether adding bevacizumab to an oxaliplatin-containing regimen is as beneficial as adding it to an irinotecan-based regimen, I don’t think it’s an issue of oxaliplatin versus irinotecan.

If you consider the second-line FOLFOX study from Dr Bruce Giantonio in ECOG — FOLFOX/ bevacizumab versus FOLFOX alone versus bevacizumab monotherapy — the benefits in terms of response rates and survival with the addition of bevacizumab to second-line FOLFOX were significant. The improvements using bevacizumab were as large as those in any other study, and the second-line setting probably includes a harder-to-treat population.

A comparison of other data to those of the first-line study of bevacizumab with oxaliplatin is confounded by several clinical study variables, which is why we need large randomized studies and why data from cross-study comparisons need to be interpreted with great caution.

Click on the image to enlarge

Colorectal Cancer Update GI Think Tank 2007

DR VENOOK: In deciding whether or not it is reasonable to use bevacizumab — with informed patient consent — for patients with metastatic disease who have experienced a recent arterial event, I don’t believe a global “no” or “yes” is the right answer. With every treatment decision, we weigh the benefits and risks of therapy.

Clearly the evidence suggests that patients who’ve had prior arterial thromboembolic events (ATEs) are at greater risk for developing subsequent events when treated with bevacizumab, and one should remember that the contemporary bevacizumab studies excluded patients who had experienced a variety of ATEs within a specific time frame, usually a year.

Click on the image to enlarge

Therefore, I believe it’s prudent to think twice before treating a patient who’s had a myocardial infarction (MI) or ATE in the past year. Still, it’s a relative contraindication because the average patient will benefit from bevacizumab, and if you examine the data, you’ll see that the incidence of an ATE is only about four percent.

When I examine this issue, I consider a couple of angles. One, what is the goal of therapy? If you believe that an excellent response to treatment might render a patient curable and free of disease, then you want to give that patient the best shot, and that might be a reason to use bevacizumab even with a relative contraindication.

On the other hand, if a patient has massive metastatic disease and the goal is clearly palliative, you might think twice about using bevacizumab if that patient is marginal in terms of risk factors for these events.

In terms of whether a patient on anticoagulation should receive bevacizumab, a bit of a myth suggests that patients who are anticoagulated can’t safely receive bevacizumab, and a fair amount of data indicate that’s not the case. I see patients who do not receive bevacizumab because of a perceived contraindication that I believe to be incorrect.

A prime example is the patient who’s been anticoagulated for a coronary stent placed five years ago. I see no contraindication to bevacizumab when treating a patient like that.

I believe that the root of this misconception may revolve around the presumption at the beginning of the studies — there was global concern regarding clotting and bleeding. The existing data set is a retrospective collation of results on patients who were anticoagulated, so it’s flawed by the questions of patient selection and why they were anticoagulated. However, Julie Hambleton has compiled data that I believe compellingly demonstrate that if bevacizumab is otherwise indicated, the patient who’s been anticoagulated can safely receive bevacizumab.

Colorectal Cancer Update 2007 (1)

DR JORDAN D BERLIN: Patients who were 65 years of age and older with a prior arterial event had more than a 17 percent risk of a second event while on bevacizumab — quite a substantial risk. However, we have a number of 65-year-old patients who have had a previous MI and are receiving bevacizumab. We have warned them about the potential for arterial events, but it’s hard not to recommend a drug with a survival benefit this good.

For treatment of hypertension associated with bevacizumab in patients with metastatic disease, we tend to use the beta blockers or the ACE inhibitors. We treat patients on bevacizumab more aggressively for hypertension because of the potential for reversible posterior leukoencephalopathy syndrome (RPLS), which can be mistaken for a stroke or a TIA.

The syndrome can include confusion, symptoms of a stroke, seizures or even coma or death. RPLS happens rarely but almost always in conjunction with at least some level of hypertension, and treating the hypertension usually leads to reversibility.

Colorectal Cancer Update GI Think Tank 2007

DR MEROPOL: For me, the scenario that sometimes plays out is with the patient for whom cure is not possible and we’re discussing prolongation of survival. Patients may not be willing to take on the added risk of a thrombotic complication with bevacizumab if they’re in a high-risk group in the front-line setting.

However, in the second- or third-line setting, when we know that bevacizumab can also improve survival after failure of front-line therapy, the tradeoff may be different. It may be the last chance to receive bevacizumab, and they might be more willing to take the risk as they get further along.

DR GOLDBERG: Regarding the issue of tolerability of bevacizumab in the elderly, we currently don’t have data to suggest that bevacizumab is tolerated less well by older people, unless they have a history of an ATE and are over 65, so I recommend it routinely without consideration for age, but I do consider the patient’s arterial thrombotic history.

Colorectal Cancer Update 2007 (5)

DR HURWITZ: The most important issue in the management of potentially curable metastatic disease is probably multidisciplinary care. There has to be a lot of discussion with the surgeon who is experienced in that kind of surgery.

In general, I think there are two potential approaches. One approach is to resect the tumor as soon as the disease is considered potentially resectable for cure. If that’s at the time of de novo presentation, one could proceed to surgery without any prior chemotherapy. The issue then is what additional chemotherapy to administer afterward. I make the extrapolation that the patient should be treated with the best first-line therapy, which is usually 5-FU combined with either oxaliplatin or irinotecan with bevacizumab. Therapy should not be started until patients have healed up from their surgery.

I approach the duration of therapy by noting how the patient tolerates treatment. As in current adjuvant studies — AVANT and the NSABP study — one may need to adjust the chemotherapy around the sixth-month mark, particularly with oxaliplatin where treatment with that agent beyond six months becomes problematic. Continuing 5-FU or 5-FU/bevacizumab to the one-year mark is a hybrid approach, consistent with the adjuvant studies and common practice in the first-line setting.

Colorectal Cancer Update GI Think Tank 2007

DR HOCHSTER: In the EORTC 40983 (EPOC) Phase III trial, patients with potentially resectable liver metastases were randomly assigned to receive three months of FOLFOX before and three months of FOLFOX after surgery or to undergo surgery without chemotherapy.

The trial demonstrated a progression-free survival benefit for chemotherapy. If a survival benefit emerges, then we will know to treat these patients with chemotherapy up front. However, you have to bear in mind that the systemic therapy in the trial is FOLFOX alone, not FOLFOX with bevacizumab.

For patients with either unresectable liver disease or one or two metastases in the lung in addition to the liver metastases, I prefer FOLFOX with bevacizumab because that combination has the highest response rate. The first-line data from the somewhat abbreviated SWOG study show that cetuximab also improves the response rate when combined with chemotherapy, so we have two antibodies that increase the response rate over chemotherapy alone.

The CALGB-C80203 study, presented by Alan Venook at ASCO 2006, was supposed to accrue 2,200 patients but was stopped early after enrolling approximately 280. The design was a two-by-two randomization of FOLFOX versus FOLFIRI with or without cetuximab. The addition of cetuximab showed approximately a 10 to 20 percent increase in the response rate for both arms, so this study also suggests that the addition of cetuximab to first-line chemotherapy improves the response rate.

Colorectal Cancer Update GI Think Tank 2007

DR GROTHEY: In terms of treating patients with potentially resectable metastatic disease, I believe what we are seeking in the neoadjuvant setting is response rate, more than delaying of tumor progression, but we don’t want to get to the point where we can no longer see the metastases. I have heard the statement, “The medical oncologist’s dream is a surgeon’s nightmare.” We know that a complete response in a liver metastasis is not a complete response by the pathology criteria.

I can easily see the rationale for using EGF receptor antibodies based on Alan Venook’s data and other data. A consistent response rate benefit occurs when we add cetuximab.

Whether we should also add bevacizumab is a different question. For now I believe that in the first-line neoadjuvant therapy, response rate is our surrogate marker for resectability, but how we get there is an area of discussion. Off study, I’ve used FOLFOX/bevacizumab and I’ve used FOLFOX/cetuximab.

Colorectal Cancer Update 2006 (5)

DR FUCHS: I consider the possibility of not sending patients who present with synchronous primary and metastatic colon cancer to up-front surgery.

If the tumor is on the right side, where the risk of obstruction is reasonably low, if they don’t demonstrate any obstructive symptoms and if there isn’t any obvious bleeding and I’m not concerned about perforation, sending them for a resection would just delay the start of systemic therapy. I would start them on a regimen of chemotherapy with bevacizumab without sending them for a resection. Some may be concerned about the possibility that perforation might occur if the primary is in place, but data have not borne that out.

Colorectal Cancer Update GI Think Tank 2007

DR WOLMARK: As surgeons, we have all been taught that the appropriate treatment when patients present with a simultaneous primary colonic tumor and metastatic disease is to resect the primary lesion.

I have actively participated in this practice and taught medical students and residents that if you leave the primary tumor, then you will end up with inordinate rates of obstruction, perforation and bleeding, and you will pay a greater price later than you would by performing the operation sooner.

This was a carryover from an era when therapy for metastatic disease didn’t have the same benefits as it does currently. We wanted to know the magnitude of the issue and whether it really is a problem today to leave the primary intact, and we hope to answer those questions with the NSABP-C-10 trial.

This is a Phase II trial evaluating FOLFOX6 with bevacizumab in patients who present with untreated primary colon cancer and concomitant metastatic disease not considered surgically resectable for cure. Patients do not undergo surgery unless down the road an obstruction or perforation makes it necessary for their safety. Patients with liver metastases amenable to hepatic resection or resection and ablation to render them “disease free” are ineligible for this study.

The endpoints consist of monitoring the rate of operations for complications of the primary tumor, such as bleeding, perforation, fistula or obstruction, or death related to the primary tumor.

In addition, these are patients who are deemed inoperable, so a tertiary endpoint is to determine how many patients can be converted from inoperable to operable, in terms of metastatic disease and the primary, with a modern regimen such as FOLFOX6 with bevacizumab.

Colorectal Cancer Update GI Think Tank 2007

DR GROTHEY: The BRiTE registry followed prospectively approximately 2,000 patients treated with chemotherapy and bevacizumab for metastatic colorectal cancer, and about 16 percent of those patients had the primary intact.

The registry revealed that 3.4 percent of these patients experienced gastrointestinal perforations, which is about twice as high as in the overall registry population, but of course they eliminated the risks of primary surgery, so it’s a tradeoff.

Currently, I believe that if the metastases are the dominating life-threatening factor, I start with chemotherapy right away, with the primary intact. Also, I do use bevacizumab in that situation.

DR DANIEL G HALLER: I believe that leaving the primary tumor intact makes sense. However, this approach is highly individualized, so it’s between the patient and the surgeon to make the final decision. If a patient has a nonobstructing, nonbleeding primary tumor, more often than not, we leave it in place.

DR GOLDBERG: My default position is not to operate on patients who present with widely metastatic disease unless they’re obstructed or bleeding, and I have been willing to administer bevacizumab.

Colorectal Cancer Update 2006 (5)

DR MEROPOL: One of the key clinical research questions being asked in current trials is whether one should continue bevacizumab after the failure of a frontline regimen containing bevacizumab.

That is, perhaps, the most important clinical question we have in the treatment of metastatic disease.

Studies are in development that we hope will answer this question. One study will randomly assign patients who experience disease progression on a frontline bevacizumab-containing regimen to continue or not continue bevacizumab with their next line of therapy. At this point, I am not continuing bevacizumab with second-line therapy.

Colorectal Cancer Update 2007 (4)

DR GROTHEY: In terms of continuing bevacizumab beyond progression, I’ve used both approaches in my clinical practice. I’ve treated patients with bevacizumab beyond progression, and I’ve stopped bevacizumab upon progression.

Outside of a clinical trial, I individualize therapy based on the effect of therapy on patients in the first-line setting. For instance, if a patient receives FOLFOX/bevacizumab as front-line therapy — which many of our patients receive right now in clinical practice — I routinely stop oxaliplatin after about four months and try to maintain the response with 5-FU/bevacizumab.

For some patients it works well, and we see the tumor stabilize for more than a year or a year and a half — as long as we’ve had experience with bevacizumab. Then you see this slow creeping up of metastases, though it’s never a rapid progression that would suggest it is a completely useless therapy. I don’t believe that bevacizumab plays a role as second-line therapy for patients whose disease progresses right after you start a bevacizumab-containing regimen.

Colorectal Cancer Update 2007 (5)

DR HURWITZ: If the disease clearly progresses on therapy, whatever that therapy is, it should be stopped and patients should receive whatever other options seem reasonable. The difficulty — particularly if slow progression is occurring — is knowing whether or not the progression is slow due to bevacizumab or due to indolent disease. That’s why we have to use our best judgment until we obtain better data.

A study known as iBET is being run by the cooperative groups. iBET (Figure 31) will address the question of whether or not patients do better with bevacizumab continued into that so-called second-line setting and whether they do better on the 5-versus the 10-mg dose.

Colorectal Cancer Update 2007 (4)

DR GROTHEY: The BRiTE registry was founded in February 2004 when bevacizumab was approved as a component of treatment for first-line colorectal cancer. Later, the approval was extended to second-line colorectal cancer, but it was clear at the time bevacizumab became approved that we had limited clinical experience with the drug.

We had data from one pivotal trial. Certain toxicities like GI perforations were reported, and later, atherothrombotic events were recognized, though only in a small number of patients.

The idea of the BRiTE registry was to obtain information on a larger number of patients — eventually it was 1,953 patients — enrolled in a “real-life” setting. Oncologists who use bevacizumab in combination with whatever chemotherapy regimen they deem appropriate document the clinical course of their patients over a long period of time. We are developing a nice database on these patients.

When patients experienced their first progression on therapy, some physicians continued bevacizumab and some did not continue bevacizumab in combination with a different treatment regimen. Outcomes, progressive disease and overall survival data were documented in the BRiTE registry.

This is a nonrandomized setting, but we tried to compare therapies within the BRiTE registry — the outcomes for patients who continued bevacizumab and those who did not. The effects were quite profound because patients who continued bevacizumab had a remarkably longer overall survival than the patients who did not receive bevacizumab.

We tried to account for all of these factors in a multivariate analysis by considering age, performance status, the number of metastatic sites, some laboratory analyses, duration of first-line therapy, et cetera. Still, the continuation of bevacizumab beyond progression turned out to be a significant factor in this analysis. Of course, if there is a profound effect, such as this one-year difference in overall survival (31 months versus 19 months) using bevacizumab beyond progression, we need to validate this finding in a prospective clinical trial.

Colorectal Cancer Update 2007 (3)

DR VENOOK: A study that has recently opened but will be important to clinical practice patterns is the iBET trial (SWOG/NCCTG/NCIC iBET S0600, Figure 31). It will evaluate whether to continue bevacizumab for patients experiencing progression on first-line chemotherapy and bevacizumab.

Patients experiencing progression on FOLFOX/bevacizumab will be randomly assigned to either irinotecan/cetuximab or irinotecan/cetuximab and bevacizumab. This trial will approach the biggest question I’m asked regularly: “Do you continue bevacizumab or don’t you continue bevacizumab at the time of progression?” Generally, outside of a protocol, our instinct is to not continue bevacizumab in that situation, although I’ll admit we often reintroduce it later on.

For a 30-year-old whose risk of stroke or myocardial infarction I believe to be sufficiently low, I might keep it going. For a 70-year-old with whom I suspect I’ve already tempted fate, I might stop. I don’t have a one-size-fits-all answer.

We factor in how the patient’s tumor is responding to the therapy, although it can cut either way. For a patient who shows a dramatic response but whose treatment is only palliative, you could argue to stop the bevacizumab because you’ve already obtained considerable value out of it.

The other way to consider it is, don’t stop the bevacizumab because it may be contributing to the response. So it’s a gray area. There’s little black and white about the decision.

Click on the image to enlarge

Colorectal Cancer Update 2007 (2)

DR JOHN L MARSHALL: My general approach to managing metastatic colorectal cancer in clinical practice is to follow an OPTIMOX-type strategy. Whether I’m administering OPTIMIRI or OPTIMOX, I back off from irinotecan or oxaliplatin after I see the optimum response, which is usually around four months of therapy. Generally, I continue with 5-FU and bevacizumab.

We are beginning to design more trials to test new agents in the chemotherapy-free window to see if they are able to stabilize or prolong progression.

When patients reprogress on one regimen, I change to the other base chemotherapy. However, many physicians like to resume the old chemotherapy — if the patient was on oxaliplatin, they bring was on irinotecan, they bring back the irinotecan.

When I switch regimens, I frequently bring in an EGFR blocker, even though that has not yet been established by the data. More recent trials support the practice of not waiting until the disease becomes irinotecan refractory before bringing in an EGFR inhibitor. In fact, data for the EGFR inhibitors now indicate that in almost every setting they’ve been tested — last line, second line, and now we have some first-line data — they’ve shown a positive impact.

For the patient whose disease is progressing, if they progressed rapidly or didn’t respond well to the treatments, I’m less enthusiastic about keeping a drug on board, so I’ll switch it. But if they’ve received a nice benefit from a drug, I don’t usually give that up.

With bevacizumab, you can recognize a change in the biology of these tumors — they slow down. It’s not necessarily that the tumors respond further, it’s just that you have “turned them off,” so I hesitate to pull patients off bevacizumab. Clinically, that’s what we’re seeing — this quieting of colon cancer and long-term survivors with metastatic disease.

Colorectal Cancer Update 2007 (3)

DR VENOOK: I’ve always been a proponent of using drug holidays in metastatic disease, without any basis for it — I just rely on common sense and try to do the best thing possible for my patients.

One of my patients is approximately four years into his metastatic disease. He initially responded to front-line FOLFOX/bevacizumab therapy and then developed neuropathy. We allowed him six months off the treatment. During that time, his disease progressed a bit and the neuropathy was persistent. We switched to FOLFIRI and bevacizumab at the time of progression.

This was some years ago, and we stopped everything for that patient. My instinct now is to continue the components of therapy without the oxaliplatin.

This patient was switched to FOLFIRI/bevacizumab and had another response, which then peaked. His neuropathy resolved, and his disease progressed. We went back to FOLFOX and bevacizumab, and he responded again. He’s four years out now on cetuximab and responding.

Click on the image to enlarge

Colorectal Cancer Update GI Think Tank 2007

DR HOCHSTER: The BOND-2 trial evaluated patients whose disease had progressed on an irinotecan regimen but who had not received bevacizumab previously. Patients who received both bevacizumab and cetuximab had approximately a 20 percent response rate, as compared to 11 percent among patients who received cetuximab only.

The toxicities were as expected from either antibody, with no unexpected or synergistic toxicities. They saw vascular side effects and perforations from the bevacizumab and skin toxicities from the cetuximab.

One study continues to examine this double antibody strategy in the front-line setting. The CALGB Intergroup study, C80405, is a three-arm study powered for survival. The physician selects either FOLFOX or FOLFIRI, and then patients are randomly assigned to bevacizumab alone, cetuximab alone or the combination of the two.

That’s an excellent study, which, if completed, will demonstrate the merits of using an anti-VEGF antibody, an anti-EGF antibody or a combination of the two. That should provide some clear data.

Colorectal Cancer Update 2007 (1)

DR ROBERT A WOLFF: I believe that in select cases the use of chemotherapy with bevacizumab and cetuximab is rational off protocol, particularly in preop situations in which you are aiming for cure. For a subset of patients, the biologic doublet, regardless of the chemotherapy backbone, will provide more bang for your buck.

However, I would not be in favor of using the combination for all patients. I usually have a sequential way of going through drugs. If patients start with FOLFOX/bevacizumab, they usually receive either FOLFIRI/bevacizumab or irinotecan as a single agent and then irinotecan and cetuximab.

I tell many of my patients that what they’re trying to accomplish is not a race — it’s a marathon. You want to stretch out the clock.

If you plow through your cytotoxics and molecular therapies and put them all into “the soup” at once, I don’t know what you’re going to have left.

Select Publications