Selection of biologics for patients with metastatic colorectal cancer: the role of predictive markers

Abstract

With the advent of biologics, the survival and quality of life of patients diagnosed with metastatic colorectal cancer (mCRC) has improved. These agents include anti-VEGF agents (bevacizumab/aflibercept), anti-EGFR antibodies (panitumumab/cetuximab) and regorafenib (an oral multi-kinase inhibitor). With the use of the novel therapeutics, there was a notable increase in the median survival of patients with mCRC; however, there are still several unmet needs with the use of novel therapeutics in mCRC. The focus of this editorial is to highlight some of these issues alongside results of some of the important clinical trials and other studies presented at the American Society of Clinical Oncology meeting last year.

Keywords:

• aflibercept
• bevacizumab
• biologics
• cetuximab
• EGFR
• metastatic colorectal cancer
• panitumumab
• RAS
• regorafenib
• VEGF

With the advent of biologics, the survival and quality of life of patients diagnosed with metastatic colorectal cancer (mCRC) has improved. These agents include anti-VEGF antibodies (bevacizumab/aflibercept), anti-epidermal growth factor (EGFR) antibodies (panitumumab/cetuximab) and regorafenib (an oral multi-kinase inhibitor).

Review of institutional databases show that with the usage of these novel therapeutics that became available around 2004 in the USA, there was a notable increase in the median survival of patients with mCRC (from around 18 months earlier [95% CI: 15.8–20.2 months] to 29.2 months later [95% CI: 24.3–34.2 months]) [1].

In addition, their usage has allowed selection of better toxicity profile and attractive maintenance treatment options for not only in patients with mCRC but also as viable treatment options in treatment refractory cases [2–5]. However, there are still several unmet needs with the usage of novel therapeutics in mCRC.

First is the selection of the right biologic agent for the right patient. Currently, there are few considerations in the algorithm of selecting a particular biologic agent for patients with mCRC. The main determinant as is known is the Kirsten rat sarcoma (KRAS) mutation status of the tumor; with patients who lack the mutation (wild-type; KRAS wild-type [wt] tumors) being the right candidates for anti-EGFR-based therapies like Cetuximab and Panitumumab.

Recently, we realize that only testing for mutations in KRAS – exon 2 at codons 12 and 13 to identify patients who could benefit from anti-EGFR-based therapies is not sufficient because there are still a significant proportion of patients who harbor other activating RAS mutations who also do not benefit from anti-EGFR-based therapies [6]. The usage of these agents in this patient population actually is detrimental as noted by Douillard et al.; whereby the progression-free survival (PFS) was actually shorter; 7.3 versus 8.0 months (p = 0.26) in groups receiving FOLFOX with and without Panitumumab, respectively [7].

These findings have been corroborated by other studies presented at the American Society of Clinical Oncology meeting this year. This included the analysis of the OPUS study on the addition of cetuximab to FOLFOX and the CRYSTAL study on the addition of cetuximab to FOLFIRI [8,9]. This gives further emphasis to the importance of personalizing treatment based on predictive markers and avoiding the simplistic approach taken in stratifying these patients [10].

As we continue to identify potentially more mutations that may be clinically useful, agreement on the panel being tested and standardization of the methods/procedures used across different hospitals and centers is going to be challenging but essential [11].

On the other hand, there has been little advancement to help identify predictive markers for bevacizumab, which targets the VEGF-A [12]. There is some interesting work done on the relative expression of VEGF receptors in patients with metastatic colon versus rectal cancers; patients with rectal cancers benefiting more from bevacizumab than patients with colon cancers in a large randomized controlled trial from Denmark (median overall survival [OS] of 24.1 vs 16.6 months; HR: 0.73; 95% CI: 0.61–0.87) [12]. It would be helpful as suggested by the authors to look back at the prior completed studies to assess the actual impact of the primary tumor location in patients with mCRC.

In addition, work on miRNAs and other plasma-based ‘angiogenic signatures’ shows some promise but is not yet ready for primetime to help guide treatment decisions [13,14]. Studies on single nucleotide polymorphisms in the VEGF pathway also show they may have predictive potential, but the data would have to be validated by larger studies [15].

One key difference to realize between the anti-EGFR monoclonal antibodies and anti-VEGF agents are the systems they are targeting. Anti-EGFR directly target tumor cells. They are, therefore, subject to the genetic instability of intracellular pathways which can lead to primary (e.g., RAS mutation) and secondary resistance (e.g., emergence of RAS mutations or activation of alternate pathways). VEGF inhibitors target a genetically stable system: the host-provided endothelial cells and the tumor–host microenvironment. This has significant implications for likelihood to find predictive biomarkers and accounts for the work accomplished thus far.

After the initial development and addition of the aforementioned anti-VEGF and anti-EGFR-based therapies to the standard chemotherapies, further addition of other targeted agents has been unyielding thus far [16]. It would be helpful to look more into the pharmacogenomics and pharmacokinetic aspects of these biologics, which were later noted to be important for some of the other chemotherapies (e.g., irinotecan) and biologic agents (e.g., rituximab) [17,18].

At the same time, the concomitant advancement in the surgical techniques with specific attention to operable liver metastases in mCRC has called into question the usage of biologics in this unique clinical setting. The results of the New EPOC randomized controlled trial are worth mentioning where addition of cetuximab to the oxaliplatin (or irinotecan) and fluoropyrimidine chemotherapy backbone in patients with operable liver metastases showed a detrimental outcome, at least for patients treated with an oxaliplatin-based chemotherapy [19]. The PFS was significantly shorter, 14·1 months versus 20·5 months (hazard ratio: 1·48; p = 0·030).

Thus, there is also the question of which combination chemotherapies to combine with some of these biologics with postulated negative interactions repeatedly being reported with combining cetuximab and oxaliplatin, which, however, have not borne out in other studies [20]. The NORDIC-VII study cited, which is a Phase III trial of cetuximab with continuous or intermittent fluorouracil, leucovorin, and oxaliplatin (Nordic FLOX) versus FLOX alone in first-line treatment of mCRC did not show any benefit of the addition of the biologic, even in patients who were deemed KRAS wild type [20]. As rightly put by the authors, ‘a so-called one size fits all strategy might have seemed rational when a fluoropyrimidine was the only systemic agent available, but fails us when we use targeted agents’ [5,19].

As first line agents, a head-to-head comparison of cetuximab and bevacizumab in combination with different chemotherapies was much awaited but did show conflicting results in the form of the FIRE-3 and the CALGB/SWOG 80405 clinical trials [21,22]. The median PFS was the reported to be the same (median 10.3 vs 10.0 months) for cetuximab versus bevacizumab respectively; however, the OS was noted to be significantly different for the two arms (28.7 vs 25.0 months; p = 0.017) in the FIRE-3 study [21]. The reasons for this are not entirely clear. Some considerations raised by the group from the FIRE-3 were the following: FIRE-3 patients received FOLFIRI as the chemotherapy regimen whereas for CALGB/SWOG 80405 most patients received FOLFOX; detailed analysis of CALGB/SWOG 80405 based on the extended RAS testing; and FIRE-3 study was restricted to mCRC only whereas the other study did allow locally advanced but unresectable patients to be included in their study [21]. The other question is if a particular first-line therapy affects clonal selection and sensitivity of cancer cells to subsequent treatments; thus raising the question of optimal sequence of the different treatment options for mCRC.

Updated results from the CALGB/SWOG 80405 and FIRE-3 studies were presented at the European Society for Medical Oncology 2014 meeting and addressed some of the issues raised. One question from the CALGB/SWOG trial was the issue of results changing if an expanded RAS testing was performed. There was still no significant difference reported between cetuximab and bevacizumab groups (OS: 32 months vs 31 months; although there was a higher response rate reported 68.6 vs 53.6% [p < 0.01], respectively). Results of independent radiological review of the FIRE-3 study at the same time showed discordant results, which may help explain the discrepancy in the OS noted in the two groups.

Finally, an important aspect for biologics in mCRC, which is true for all the newer biologics, is the cost of the novel therapeutics. Although a lot of the indications and clinical situations are approved in the USA, the recent guidance statement from National Institute for Health and Care Excellence published in January 2012 questioned the cost–effectiveness of using some of these biologic agents for mCRC that has progressed after first-line therapy [23].

Moving forward, these are important public health considerations. Although not undermining the efforts and the costs that go into drug development, the final cost of a particular biologic agent would have to be justified with the actual benefits seen and the relative costs of competitor drugs in the same class. ‘Refining the patient population’ and identifying the subgroup that would benefit the most from a particular biologic agent through more work on predictive markers may help at least improve the cost–benefit ratio [3,7,24].

Methodology

The selection of articles presented in this review was based on expert opinion with focus on prospective randomized controlled trials that are landmark and have had practice changing implications for patients with mCRC in the opinion of the authors. Data from unpublished clinical trials that were presented in the form of abstracts from recent meetings at the American Society of Clinical Oncology was also considered.

Financial & competing interests disclosure

JM Hubbard has received research support from Genentech, Boston Biomedical Inc. and PRISM Pharma Co. A Grothey has received research support from Genentech, Bayer, Eisai, Pfizer and Eli-Lilly. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.