The First Step: Identifying the Goal of Therapy
When I first see a patient with metastatic colorectal cancer, one of my key issues is to establish the overall goal of therapy. It is well known that some patients with Stage IV colorectal cancer can be cured from their disease or at least, rendered free of their disease for some time to kind of reset the “tumor clock.”
Which Patient is a Candidate for Potentially Curative Therapy?
Patients who present with limited metastatic spread—for instance, four to five metastases to the liver without other organ involvement—should always be considered for a potential curative approach.
Now, every liver surgeon will tell us that it is not the actual number of metastases, but the amount of functional liver parenchyma with adequate inflow and outflow after resection that determines resectability. Therefore, I routinely get an experienced liver surgeon involved right upfront in the decision-making process when I establish the goal of therapy with my patient.
How Do I Use PET/CT Scans in This Setting?
Before I get a surgeon involved, I obtain a CT scan of abdomen and pelvis (and a CT chest for patients with rectal cancer) as well as a PET/CT scan.
This is one of the very few indications I see for a PET/CT scan in the management of patients with advanced colorectal cancer: The Stage IV preoperative setting before the start of potentially curative therapy.
It is important to obtain this PET/CT scan before initiation of systemic chemotherapy, since even after just one single cycle of chvemotherapy, in particular, when administered with bevacizumab, PET/CT scans can be falsely negative.
The goal of this scan is not to establish response to therapy, but to identify metastatic lesions that could have been missed on regular CT scans, the most common location here being retroperitoneal lymph node metastases or peritoneal deposits.
Can a Patient with Liver and Lung Metastases Be Cured?
Oligometastatic disease, which includes limited lung metastases, even if bilateral, can in select cases also be targeted for curative resection.
It is obvious that this requires a very individualized approach, and no firm guidelines can be given regarding what amount of disease is still considered resectable—along the lines of the famous quote: “You know it when you see it.”
Conversion Therapy versus Neoadjuvant Treatment
Some patients present with metastatic disease that is easily resectable—for instance, a solitary liver lesion peripherally in the right lobe. Other patients have “borderline resectable” disease. This is when a surgeon tells you that he could take it out but would rather have you shrink the lesion by administering preoperative therapy.
Other patients present with metastatic lesions, which require a significant tumor response before anyone would even think of surgical resection.
Whenever chemotherapy is needed to obtain a potentially resectable state, we are talking about “conversion therapy.” In these cases, a treatment regimen with the highest response rate is likely the best to achieve this goal.
What Is the Best “Conversion Therapy”?
Most data in this setting are available for a FOLFOX-based treatment approach, although recent data suggest that FOLFIRI could also serve as backbone.
A few studies have also demonstrated that combining all active chemotherapy agents—5-FU, irinotecan, and oxaliplatin (FOLFOXIRI)—can achieve high response rates and can lead to a high rate of liver resections in previously unresectable tumors.
I personally routinely use FOLFOX in this setting and would add an EGFR antibody to the treatment if the tumor has been characterized as KRAS wild-type.
EGFR antibodies have routinely been associated with an increase in response rate when added to chemotherapy (FOLFOX or FOLFIRI), which is exactly what is required in a conversion therapy setting.
I add bevacizumab to FOLFOX in KRAS-mutated tumors, although I could envision that using FOLFOXIRI without any biologic agent could also be an attractive alternative in KRAS-mutated colorectal cancer.
Unfortunately, I do not have a lot of experience yet with this aggressive chemotherapy combination, although this might change with the emergence of FOLFIRINOX as one of the standard therapies for advanced pancreatic cancer, which uses even higher dose of irinotecan in the combination than FOLFOXIRI.
When Do I Consider Neoadjuvant Therapy in Resectable Stage IV Disease?
In patients with resectable metastatic disease, one could wonder if neoadjuvant therapy was indicated, or if surgical resection would be the preferred first step, followed by postoperative adjuvant therapy.
In my eyes, the optimal sequence of treatment steps depends of the clinical situation we are facing. Two different patient cases might outline the different scenarios I have in mind here.
The first case is a patient who presents with colon cancer and synchronous, albeit resectable liver metastases. The clinical information we have on this patient at the time of diagnosis is a one-time snapshot of his disease; we have no real understanding of the particular biology of this cancer.
We could be facing yet undetected metastases, which could declare themselves if treatment was withheld for some time—but I don't think that any patient or oncologist would be willing to wait long before initiation of therapy.
In patients with synchronous metastases, I routinely use neoadjuvant therapy to administer hopefully effective cytotoxic treatment early on to target occult micrometastatic disease and—most importantly—to obtain information on the chemosensitivity of the cancer.
This approach provides an in vivo test of chemo-responsiveness of this particular disease, which can help guide postoperative adjuvant therapy, too.
It is important to note that an asymptomatic primary tumor can be safely left in place for resection at a later point. In addition, the preoperative treatment duration does not need to be long. After two to three cycles of FOLFOX, a response should be visible on CT scans.
With regard to biologics, I do not think that bevacizumab is needed in this setting, since it would have to be discontinued six weeks before planned surgery anyway. Similarly, the use of EGFR antibodies in KRAS wild-type cancers is not mandatory here.
Now, the decision with regard to potential neoadjuvant therapy in resectable Stage IV disease can be different in the following scenario: A patient who underwent colon resection for a Stage II cancer four years ago and now presents with a solitary liver metastasis.
In this case, the tumor took four years to declare itself as metastatic disease—and then came up with a solitary, easily resectable metastasis. This scenario provides us with a lot of information on this patient's tumor biology; no neoadjuvant therapy is indicated. In such cases, I get a liver surgeon involved right away and offer postoperative adjuvant therapy, mainly FOLFOX (or XELOX).
Of course, a lot of patients will fall somewhere in between these two scenarios and individual decisions on the most appropriate approach have to be made in agreement with my surgical colleagues.
Should Biologics Be Used in the Postoperative Setting after Curative Resection of Stage IV Disease?
Since the goal of postoperative, adjuvant therapy is to prevent the recurrence of macrometastatic disease, I consider patients with R0-resected Stage IV disease to be in the same clinical situation as patients with resected high-risk Stage III colon cancer.
Unfortunately, all recent Phase II trials on the use of cetuximab and bevacizumab as adjuvant therapy in Stage III colon cancer have been completely negative, with no identifiable subgroup who might have benefitted from the antibodies. Therefore, I do not add bevacizumab or EGFR antibodies to the postoperative therapy, even if I had used them successfully before as part of a preoperative therapy as outlined above.
What Molecular Markers Do I Like to Have to Make Treatment Decisions in the Palliative Therapy for Colorectal Cancer?
Before any treatment for metastatic disease is considered, I obtain a KRAS test on available tumor tissue (primary tumor or biopsy of metastases) to get a better understanding of the treatment options available for my therapeutic approach.
Patients with KRAS-mutated tumors have no chance of benefit from EGFR antibodies, cetuximab, and panitumumab.
I also like to get to know if BRAF is mutated, since patients with BRAF mutations, which make up around 7% of all cases, have a very poor prognosis, with a median overall survival of about 12 months.
These patients need to be watched closely on therapy and are not necessarily candidates for “stop-and-go” strategies and “chemo-light” maintenance approaches.
In my practice, I have seen several patients with BRAF-mutated cancers who initially responded well to therapy, but then showed rapid, unstoppable and eventually lethal progression of their disease. Unfortunately, no prospective studies in these patients are available yet to help outline an optimal therapeutic strategy.
What Chemotherapy Do I Use as First-Line Treatment?
I generally prefer combination regimens such as FOLFOX and FOLFIRI as first-line therapy. In select cases, I also use XELOX, although I find that a lot of patients experience this regimen as more toxic than FOLFOX.
FOLFOX and FOLFIRI are largely equivalent as palliative therapy in this setting, and it is up to one's individual preference and personal experience to choose between one or the other.
I personally use FOLFOX a lot more commonly than FOLFIRI for several reasons: FOLFOX normally does not lead to alopecia—meaning that patients are not being stigmatized as cancer patients with regard to their outer appearance.
Secondly, oxaliplatin can be safely administered in patients with liver and renal dysfunction—a key difference to irinotecan.
Which of the Various “FOLFOXes” Is the Best?
In a palliative setting, I routinely use modified FOLFOX7, which is the widely known modified FOLFOX6 without bolus 5-FU. I very much doubt that bolus 5-FU contributes significantly to the activity of this regimen, but it is associated with neutropenia, mucositis, and diarrhea.
In a palliative setting, where the treatment goal is to end the duration of a patient's life, but also maintain the quality of his life as long as possible, bolus 5-FU should not be administered.
Should a Biologic Agent Be Added to First-Line Chemotherapy?
In most patients, I add bevacizumab to the chemotherapy backbone, regardless if I choose FOLFIX, FOLFIRI, XELOX, or—perhaps in older patients—5-FU/leucovorin or capecitabine single agent.
I do not think that EGFR antibodies provide a significant advantage over bevacizumab in this setting, not even in KRAS wild-type cancers. In fact, data suggest that the activity of EGFR antibodies is maintained in later lines of therapy.
One could even say that the “bang-for-the-buck” (as measured by hazard ratio) is higher, when cetuximab or panitumumab are used as salvage therapy in a third-line setting, with or without chemotherapy. I would only consider EGFR antibodies as part of first-line therapy if the patient has contraindications against bevacizumab. But then, not every patient needs biologics upfront.
How Long Should First-Line Therapy Be Continued?
Since I mainly use FOLFOX + bevacizumab as first-line chemotherapy, the issue of cumulative neurotoxicity plays an important role in my choice of an appropriate treatment strategy and decision on duration of therapy.
It is quite obvious that patients cannot tolerate more than five to six months of oxaliplatin-containing therapy, which is shorter than the median progression-free survival achievable nowadays.
In fact, the severity of sensory neurotoxicity seems to jump after a cumulative threshold dose of about 750 mg/m2 oxaliplatin has been administered.
Therefore, in a palliative setting I always—and I repeat: always—stop oxaliplatin after eight cycles of FOLFOX (or six cycles of XELOX with the higher 3-weekly oxaliplatin dose of 130 mg/m2), and continue treatment with 5-FU/LV (or capecitabine) plus bevacizumab until disease progression.
This greatly limits the (neuro-)toxicity of oxaliplatin and allows the use of an oxaliplatin-based regimen at a later point in the treatment sequence.
I actually also believe that irinotecan could be stopped in a similar fashion if FOLFIRI has been chosen as front-line therapy—data to support this notion actually exist.
However, FOLFIRI is much less plagued by cumulative toxicities than FOLFOX, although over time, myelosuppression can become an issue.
What Do I Do If All “Normal” Treatment Options Are Exhausted, but My Patient Is Still in Excellent Performance Status?
More and more I see patients who have gone through two to three years of therapy with various combinations of the available active agents and eventually present with progressive disease, but are still in excellent shape—and willing to undergo further active therapy.
The good news here is that this indicates that the right treatment strategy can actually keep patients alive and in good performance status without compromising their quality of live and functional status too much.
The bad news is that we are facing a shortage of novel promising agents and treatment approaches.
In fact, it has been almost seven years since the last new classes of agents, bevacizumab and cetuximab, were approved in February 2004.
Salvage therapy options that I consider are liver-directed therapy—e.g., in the form of radioembolization in patients with liver-dominant disease who are still candidates for such an approach.
One can also consider going back to the treatment regimen used longest ago, with the idea that because of tumor dynamic genetic heterogeneity, tumor cells might be sensitive to previously used agents.
Participation in clinical trials, such as Phase I studies, are an option, if available and if the patient is considered eligible. Of course, best supportive care needs to be brought up, and I do believe that this is a much better and more honest option for a lot of patients than utilizing a potentially toxic, most likely ineffective chemotherapy as a costly placebo.
We Need Participation in Clinical Trials!
I could not finish this treatment outline without appealing to all oncologists to enroll patients in clinical trials. We owe it to our patients to move this field forward, and there are so many gaps in our knowledge that haven't been filled yet by solid data.
In the US, it seems to be difficult to accrue to trials run in first- and second-line therapy in advanced colorectal cancer—likely because most oncologists think that an appropriate standard of care is available for most patients.
I would vehemently dispute this notion, since we are still not curing most of our patients with Stage IV colorectal cancer, and a median overall survival somewhere between two and three years at this point might look like a major improvement compared with 15 to 18 months 10 years ago, but it is still not close to what I would understand as “chronic disease.”
Advances cannot be made by preserving the status quo, but only by moving forward in innovative ways. This is where we all need the help of the oncologic community.