ALAN P. VENOOK, MD, is the Madden Family Distinguished Professor in Medical Oncology and Translational Research at Helen Diller Family Comprehensive Cancer Center of the University of California, San Francisco.
The incidence of colorectal cancer is decreasing thanks to some success with screening, yet there will still be more than 140,000 patients diagnosed with the disease this year. And although more patients with colon cancer are cured than are not, about 50,000 people a year will be diagnosed with metastatic disease. As opposed to the majority of stage IV solid tumor patients who are being treated for palliation, however, patients with limited metastatic colorectal cancer may be treated with curative intent. That makes every decision an important one.
Sometimes, the first decision has been made before the oncologist is involved.
As many as one-third of patients will present with metastatic disease and the primary cancer in place, often reflecting the ability of PET scan and CT imaging to detect low-volume metastases. Depending on who sees the patient first, therefore, the patient with metastatic disease may be recovering from surgery, which may necessitate a delay before treatment can be initiated. If the patient has not yet undergone surgery, then that is where the debate begins.
The pendulum swings back and forth as to whether the primary should or should not be left in place. The balance between immediate chemotherapy or surgery is a nuanced one and is even more complicated if the tumor is in the rectum (the distal 12 cm of the large bowel), where radiation to maximize local control also needs consideration.
In general, patients with markedly symptomatic primaries should be strongly considered for definitive upfront management (either surgery or chemotherapy and radiation) unless the difference of six or eight weeks in delaying systemic treatment while the patient recovers from surgery is a concern in the patient with bulky metastases.
Regardless of the events leading up to the visit to the oncologist, there is always a push to start chemotherapy as soon as possible. However, it is important for oncologists to make sure they know enough details of the tumor biology to proceed. Someday that may include routine multi-gene platform assays, but as of today, that would mean assessment of KRAS status; and if the patient is KRAS wild-type, a check for BRAF mutations.
With that in mind, it is important to check with the pathologist to be sure there is adequate material for these assays. (The expectation, though, is that KRAS testing will soon give way to “expanded” RAS testing, which may require more tissue and more time.)
Which Combination of Chemotherapy?
With that information either in hand or pending, the complex decision of which combination of chemotherapy to begin with can be addressed. Actually, based on market information, this is not really a complex decision for most practitioners: the majority of patients with metastatic colorectal cancer start treatment with FOLFOX and bevacizumab. This is certainly a reasonable choice, albeit not based on any comparative data. And because the first-line treatment will impact subsequent lines of therapy by the process of elimination, if not by complications or toxicities, the initial steps may be determinant of whether the patient can be cured or not and/or will be saddled with a toxicity or complication that will persist throughout the remainder of their life.
FOLFOX and FOLFIRI are for all intents and purposes equivalent in outcome but carry different risks and toxicities.
FOLFIRI causes alopecia and can be poorly tolerated, particularly in patients with Gilbert's syndrome. FOLFOX is rarely problematic for patients—until six or seven doses have been given, when nearly every patient develops peripheral neuropathy and when patient and clinician must decide whether the neuropathy is enough to warrant a change in approach. Even then it may be too late, since common wisdom aside, neuropathy may get worse rather than better even when the oxaliplatin has been discontinued.
For that reason (and not only to be contrarians), we typically start with FOLFIRI as the first-line backbone (after a check of prior serum bilirubin levels makes the likelihood of Gilbert's nil—about 10 percent of the population actually has a UGT1A1 polymorphism that makes irinotecan a bad idea.)
We continue FOLFIRI as long as it is well tolerated and in the absence of progression, with a check at two months if the plan is to go to curative metastectomy lest the patient develop hepatic complications from chemotherapy. Capecitabine can be substituted for the 48-hour infusional component, but we do not mix it with irinotecan regimens.
The chemotherapy actually may work remarkably well by itself, so bevacizumab or an EGF-R antibody (cetuximab or panitumumab) are not mandatory at first or even later on. The antibodies—with the recently approved ziv-aflibercept—add to overall survival on average when used in multiple lines, but there are occasional severe complications and the sequence of their use, the possibility that discontinuing bevacizumab may lead to a sort of rebound effect, and the acneform rash of the EGF-R antibodies are all factors to be considered.
A biomarker for bevacizumab or ziv-aflibercept would be a great advance just as the identification of KRAS and other RAS mutations enabled us to target patients likeliest to benefit from EGF-R blockade.
The decision to change treatments is obvious when there is overt progression. We follow patients with good old-fashioned contrast CT scans to inform the less obvious cases and look for clear-cut progression before doing so. We do not let changes in PET imaging dictate therapeutic decisions, by and large, because these are often misleading. In fact, I believe that chemotherapy decision-making in colorectal cancer should be a PET-free zone.
However the decisions are made, when the above combinations have been spent, we turn either to regorafenib or to an experimental therapy we deem promising, with an emphasis on combinations with the capacity to reverse KRAS mutation and resistance. Many patients will have had their tumors analyzed for a plethora of mutations by this time, but most of the “findings” do not beget a treatment. The exception is the BRAF mutation: we determine that status at the beginning, and such patients are targeted early for experimental approaches, either at our center or nationally.
The algorithm described above is just one person's opinion. We await the results of CALGB 80405 to determine if there is one best approach for all patients or one best approach for an individual patient.
A your comments in the blog version of this article—http://bit.ly/OT-TREAT-Venook—or email OT@LWWNY.com