Steven M Sorscher
Assistant Professor of Medicine
Washington University School of Medicine
For most of the last 40 years patients with metastatic colon or rectal cancer had few treatment options. The mainstay of treatment, 5-fluorouracil (5FU) combined with leucovorin, imparted a small survival advantage compared with best supportive care. The agents irinotecan, oxaliplatin, capecitabine, bevacizumab, cetuximab, vatalanib and panitumumab are now having a substantial impact on the outlook for patients with metastatic colon or rectal cancer. Patients who receive these drugs may expect to live an average of 25 months, more than twice the overall survival compared to that expected with 5FU/leucovorin alone.(1) The prospects for further improving the outlook for these patients involves understanding how the host genotype affects the efficacy and toxicity of available therapies, developing new therapies and using what has been learned from recent key findings.
In many ways, the diagnosis and treatment of early-stage colon cancer is evolving and improving. Identification of families harbouring germline polymorphisms of mismatch repair genes (familial non-polyposis colon cancer syndrome) which place patients at high risk of developing colorectal, uterine and other malignancies has led to more aggressive screening and prophylaxis that should translate into fewer cancer deaths in these families. There is evidence that cancers in patients from these families show biologies different from their sporadic cancer counterparts. Understanding these differences will help optimise both adjuvant therapy and treatment of metastatic disease. Other intrinsic tumour molecular abnormalities in cancers appear to predict greater or less aggressive behaviour. For example, patients with tumours lacking loss of chromosome 18q might carry a more favourable prognosis, such that these patients could safely elect to forgo adjuvant chemotherapy.(2) These are areas of active investigation.
This brief review focuses on treatment of metastatic colon or rectal cancer. The promise of individualising therapy based on host polymorphisms (which affect the safety and effectiveness of a drug) will be discussed. The key studies demonstrating benefit for newer agents will be reviewed as well.
Understanding the host
As William Osler wrote: “It is more important to know what kind of patient has the disease than what kind of disease the patient has.” Through painstaking efforts, laboratory research has delineated many of the molecular changes characterising different cancers. Can genetic differences between patients predict the best therapy as well?
Polymorphisms have been identified which alter the efficacy and toxicity of some common therapies. Germline polymorphisms in the thymidylate synthase (TS) gene and the dihydropyrimidine hydrodenase (DPD) gene affect 5FU metabolism and thereby safety and efficacy. Glutathoine S-transferase P (GSTP1) polymorphisms alter the effect of oxaliplatin.(3) Both 5FU and oxaliplatin are commonly employed drugs in treating colorectal cancer.
Irinotecan is another drug commonly used in treating colon and rectal cancer. A sizeable percentage of the general population harbours a germline poly‑morphism in the gene encoding uridine disphosphate glucuranosyl transferase (UGT) 1A1. Patients with this polymorphism have less ability to metabolise SN-38, the active metabolite of irinotecan, and therefore are at greater risk of toxicity from irinotecan. In fact, the package insert for irinotecan now notes: “Individuals who are homozygous for the UGT1A1*28 allele are at increased risk for neutropenia following initiation of camptosar treatment. A reduced dose should be considered for patients known to be homozygous for the UGT1A1*28 allele.”(4,5) Oncologists will increasingly test patients for these polymorphisms in order that patients can more adequately be advised of how both toxicity and efficacy might be altered by specific polymorphisms. As our knowledge of the effect of polymorphisms on the metabolism of other agents grows, the genetic background of an individual patient will become increasingly critical in planning treatment.
The fluoropyrimidine 5FU has been the basis for the treatment of metastatic colon and rectal cancer for the last several decades. However, 5FU (which has been given in bolus or infusional forms, with or without leucovorin) only produces a 20–30% response rate, and only results in a small survival advantage compared to best supportive care. More recently, bolus 5FU/leucovorin has been used in combination with infusional 5FU and either oxaliplatin (FOLFOX) or irinotecan (FOLFIRI) with an approximate doubling of response rates. Now FOLFOX and FOLFIRI are being used in combination with the so-called targeted agents bevacizumab and cetuximab. (The later two drugs might more appropriately be referred to as “targeting”, since they are designed to target proteins produced by the malignancy).
Bevacizumab is a recombinant humanised monoclonal antibody that targets vascular endothelial growth factor (VEGF).(6) Colon and rectal cancers typically express VEGF and expression inversely correlates with prognosis, making an antibody thought to interfere with the angiogenic function of VEGF a promising strategy.(7–9)
In the first-line treatment of metastatic colon or rectal cancer the first key trial to demonstrate bevacizumab efficacy was a randomised double-blind placebo-controlled trial. Patients receiving bolus 5FU/irinotecan/leucovorin (IFL) and bevacizumab (n = 397) showed a median survival advantage (20.3 vs 15.6 months, p = 0.0001) and greater response rate (44.8 vs 35%, p = 0.024) compared with the same chemotherapy plus placebo.(10) Hypertension, gastrointestinal perforation, proteinuric and thrombo‑embolic events have been associated with bevacizumab. All three cases of grade 4 bleeding were seen among patients who received bevacizumab.
The ECOG (Eastern Cooperative Oncology Group) 3200 study evaluated patients previously treated with a fluoropyrmidine and an irinotecan-based regimen after randomly being assigned to receive FOLFOX4, either alone or in combination with bevacizumab (10mg/kg every two weeks).(11,12) Overall survival favoured the patients who received bevacizumab compared with those assigned to the non-bevacizumab-containing treatment group (12.5 vs 10.7 months). There were increased risks of bleeding, hypertension and neurotoxicity for those patients who received bevacizumab.(13)
Other studies have shown a benefit to the use of bevacizumab in combination with chemo‑therapeutic regimens combining commonly used drugs. A recent report of the Bevacizumab Expanded Access Trial (BEAT) analysed 1,915 patients treated with 5FU- or capecitabine-based chemotherapy and bevacizumab. At a median follow-up of 6.7 months, this study showed uncommon but serious adverse effects, including thromboembolic disease (1.7%), pulmonary embolus (1.1%), bleeding (1%), gastro‑intestinal perforation (0.9%) and arterial thrombo‑embolic disease (0.8%).(14)
The oral agent vatalanib was designed to interfere with VEGF tyrosine kinase activity. A randomised phase III trial, CONFIRM-1 (Colorectal Oral Novel Therapy for the Inhibition of Angiogenesis and Retarding Metastases in the Second Line) has enrolled 1,090 patients and showed that compared with FOLFOX alone, vatalinib with FOLFOX improved progression-free survival (PFS).(15,16)
Cetuximab is the other commonly used targeted agent. This chimaeric IgG1 monoclonal antibody has high affinity for the extracellular binding domain of the epidermal growth factor receptor (EGFR), and 60–80% of colorectal cancers express or overexpress EGFR, and EGFR appears critical to the malignant phenotype. While the exact mechanism resulting in cetuximab efficacy is incompletely understood, interference with downstream signalling from EGFR results from cetuximab binding.(17–21)
The BOND-1 trial enrolled patients previously treated with oxaliplatin and irinotecan. Patients received irinotecan and cetuximab or cetuximab alone followed by cetuximab 250 mg/m2 every week. Irinotecan was given as it had been previously for each irinotecan-treated patient. Time to progression (TTP) favoured the combination therapy (4.1 vs 1.5 months). High-grade skin reaction correlated with high response rates and TTP. The response rates favoured the combination therapy (22.9 vs 10.8%, p = 0.0071). Severe anaphylaxis occurred in 2% of patients and acneform rash in 80%.(22)
In CALGB (Cancer and Leukemia Group B) 80203, patients received FOLFIRI or FOLFOX with or without cetuximab. Response rates favoured those patients receiving cetuximab (52 vs 38%), with similar grade 3 neutropenia in the cetuximab- and non-cetuximab-containing arms.(23)
Panitumumab is a fully humanised monoclonal antibody to the EGFR extracellular domain. In heavily pretreated patients, panitumumab produced an improved PFS compared with best supportive care (HR 0.54, 95% CI 0.44, 0.66). Three percent of patients experienced grade 3 rash, and no grade 3, 4 infusion reactions were seen.(24)
Future directions and conclusions
The described studies have resulted in significant progress in the treatment of metastatic colon and rectal cancer. The key to future success naturally lies partly in the development of drugs active in this disease. In the same way that years of laboratory studies have demonstrated profound difference in the effects of drugs in laboratory animals compared to patients, our improved understanding of pharmaco‑genetics will show that polymorphic differences between individual patients will have a major impact on the utility of drugs used alone or in combination. In addition, as we have come to realise that there are subgroups of breast cancer (eg, HER-2 overexpression or lack of overexpression, oestrogen/progesterone receptor negative or positive) such that therapy must be tailored appropriately (eg, trastuzumab, antioestrogen therapies), the same will be true concerning truly novel targeted therapies for colorectal cancer. Future studies (metastatic or adjuvant) should carefully evaluate the difference in benefit between molecular subgroups; otherwise, a statistical benefit of a subgroup could be lost in analysing the patients as a single group. For example, a recent report suggests that the downstaging of metastatic colorectal cancer in the liver with bevacizumab-based therapy in men is twice as effective compared with women. The authors suggested this result was due to expression of VEGF by endogenous oestrogens, resulting in less activity of bevacizumab in the female patients.25 This is one of numerous emerging examples of the need to understand the role of the host more fully.
In summary, understanding the genotype of the patient with colon or rectal cancer, the particular molecular subgroup of an individual’s cancer and the development of new drugs will undoubtedly result in still greater improvements in survival for patients with metastatic colon or rectal cancer.
1. Hobday TJ, Goldberg RM. Clin Colorect Canc 2002;2(3):161-9.
2. Lanza G, Matteuzzi M, Gafá R, et al. Int J Cancer 1998;79(4):390-5.
3. Lenz H-J. Pharmacogenomics and colorectal cancer. New York: Springer; 2006.
4. Center for Drug Evaluation and Research. Camptosar®: irinotecan hydrochloride injection. Rockville MD: FDA; 2005. Available at: www.fda.gov/cder/foi/label/2005/020571s024,027,028lbl.pdf
5. Center for Devices and Radiological Health. Invader UGT1A1 molecular assay 510 (k) summary. Rockville MD: FDA; 2005. Available at: www.fda.gov/cdrh/pdf5/K051824.pdf
6. Ferrara N, Hillan KJ, Gerber HP, et al. Nat Rev Drug Discov 2004:3(5):391-400.
7. Lee J, Chow N, Wang S, et al. Eur J Cancer 2000:4(6):748-53.
8. Choi HJ, Hyun MS, Jung GJ, et al. Oncology 1998;55(6):575-81.
9. Takahashi Y, Kitadai Y, Bucana C, et al. Cancer Res 1995;55(18):3964-8.
10. Hurwitz H, Fehrenbacher L, Novotny W, et al. N Engl J Med 2004;350(23):2335-42.
11. Giantonio BJ, Catalano PJ, Meropol NJ. Paper presented at American Society of Clinical Oncology Gastrointestinal Cancers Symposium, USA, 2004; Abstract 241.
12. Reddy GK. Clin Colorect Canc 2005;4(5):300-1.
13. Giantonio BJ, Catalano PJ, Meropol NJ, et al. Paper presented at American Society of Clinical Oncology Gastrointestinal Cancers Symposium, USA, 2005; Abstract 169.
14. Van Cutsem E, Michael M, Berry S, et al. Paper presented at: 2006 Gastrointestinal Cancers Symposium: Multidisciplinary Approaches to the Prevention, Diagnosis, and Therapy of GI Cancers; San Francisco, CA, USA; 26–28 January 2006; Abstract 250.
15. Hecht JR, Trarbach T, Jaeger E, et al. Paper presented at American Society of Clinical Oncology Annual Meeting, USA, 2005; Abstract LBA3.
16. Tyagi P. Clin Colorect Canc 2005;5(1):24-6.
17. Sato JD, Kawamoto T, Le AD, et al. Mol Biol Med 1983;1:511-29.
18. Mendelsohn J, Baselga J. J Clin Oncol 2003;21(14):2787-99.
19. Baselga J. Oncologist 2002;7:2-8.
20. Baselga J, Albanell J. Hematol Oncol Clin N Am 2002:10(5):1041-63.
21. Salomon DS, Brandt R, Ciardiello F, et al. Crit Rev Oncol Hematol 1995:7(3):183-232.
22. Cunningham D, Humblet Y, Siena S, et al. N Engl J Med 2004;351(4):337-45.
23. Venook A, Niedzwiecki D, Hollis D, et al. Paper presented at American Society of Clinical Oncology Annual Meeting, USA, 2006; Abstract 3509.
24. Peeters M, Van Cutsem E, Salvatore S, et al. Paper presented at Annual Meeting of the American Association for Cancer Research, USA, 2006; Abstract CP-1.
25. Aderka D, Wolf I, Shmueli E, et al. Paper presented at Gastrointestinal Cancers Symposium 2007; Abstract 360.