‘Sidedness’ in Colon Cancer: Using the Data in the Clinic

Get Permission

The evidence from clinical trials has established that “side matters” when it comes to colorectal cancer outcomes. How do clinicians use this information in their practices? Christina Wu, MD, of Emory University, shared her thoughts with attendees at the 2017 Debates and Didactics in Hematology and Oncology Conference in Sea Island, Georgia.

“Patients have heard about this, and they have been calling and asking if their cancer was in the right or left side. It’s now a hot topic,” she noted.

Biologic and Clinical Differences by Side 

Scientists are searching for reasons why metastatic right-sided (proximal) cancers have worse outcomes than left-sided (distal) ones, with several hypotheses emerging. One pertains to the tissue of origin. The right side of the colon originates from the midgut, and the left originates from the hindgut. These different origins may explain what we see biologically, she said.

Different signaling pathways are also involved in the development of cancer—conventional pathways, serrated pathways, and familial pathways. The serrated pathway is more prevalent in the right side and is where BRAF mutations develop and CpG island hypermethylation occurs. The conventional pathway, where mutations in KRAS, TP53, and APC occur most, is more often a left-sided phenomenon. 

Clinical characteristics also differ between right- and left-sided cancers. Patients with right-sided tumors tend to be older and female; to have tumors with BRAF, PIK3CA, and KRAS mutations; and to have tumors classified as microsatellite instability–high (MSI-H). Prognosis is worse for these patients with metastatic disease.

Patients with left-sided tumors are usually younger, have KRAS and NRAS wild-type disease, and demonstrate amplification in epidermal growth factor receptor (EGFR) and HER2 as well as high EGFR ligand expression. Prognosis for these patients is better than for those with right-sided tumors.

Does Side Matter in Early-Stage Disease?

The impact of side is clear in metastatic disease, but not so much in early-stage disease, and this has implications in the clinic, Dr. Wu said. 

In a Canadian population-based retrospective cohort study of 6,265 patients, those with right-sided tumors tended to be older and female, and the tumors were more likely to be staged as T4 and poorly differentiated but less likely to be node-positive.1 No difference was observed for overall survival or cancer-specific survival between right- and left-sided tumors—overall or at any stage of the disease (stages I–III). 


  • Right-sided tumors: Offer chemotherapy plus bevacizumab. 
  • Left-sided tumors: Offer chemotherapy and bevacizumab or an anti-EGFR antibody (if KRAS and NRAS wild-type tumors). 
  • MSI-H tumors: Enroll the patient on a clinical trial of immunotherapy. 
  • BRAF-mutant tumors: Treat with FOLFOXIRI and bevacizumab.

In a Surveillance, Epidemiology, and End Results (SEER) analysis of 58,801 patients with early-stage cancer, right-sided tumors more often occurred in older women, were diagnosed at a more advanced age, and were poorly differentiated.2 Again, no significant differences in mortality were observed according to side.

“I think we still have to gather more information for early-stage colon cancer, to know for sure whether right vs left matters,” Dr. Wu said. Meanwhile, sidedness should not be a factor in treatment selection for early-stage disease.

Side Does Matter in Metastatic Disease

In contrast, several trials have found outcomes to be worse when tumors of the right colon metastasize. This was shown in two studies of patients not molecularly selected for KRAS: the E2290 trial, in which 1,120 patients received fluorouracil (5-FU)–based chemotherapy,3 and another study, where 2,053 patients received one of several chemotherapy regimens with bevacizu-mab (Avastin).4 In addition, in an analysis of patients with KRAS wild-type disease in the CO.17 trial who received cetuximab (Erbitux) plus best supportive care, right-sided cancers progressed sooner.5 

“The key question is the first-line choice of biologic. Do we offer a chemotherapy backbone with bevacizumab or an anti-EGFR antibody?” she asked. Three major trials have been informative for this question. 

FIRE-3 evaluated FOLFIRI (5-FU, irinotecan, leucovorin) plus cetuximab or bevacizumab in the first-line metastatic setting and showed more benefit for cetuximab in left-sided tumors.6 Bevacizumab seemed to benefit patients with tumors in either location, although among patients receiving bevacizumab, those with left-sided cancers had better survival. Most strikingly, among patients receiving FOLFIRI plus cetuximab, median overall survival was 38.7 months for left-sided cancers but 16.1 months for right-sided cancers (hazard ratio [HR] = 0.26; P < .0001). Comparing the drugs in left-sided cancers, cetuximab greatly improved survival over bevacizumab (HR = 0.63; P = .002) but had no significant impact in right-sided cancers.7 

I think we still have to gather more information for early-stage colon cancer, to know for sure whether right vs left matters.
— Christina Wu, MD

Tweet this quote

Similar findings emerged from an analysis of CALGB/SWOG 80405, which also evaluated bevacizumab and cetuximab plus FOLFOX (5-FU, leucovorin, oxaliplatin) or FOLFIRI as first-line treatment. The regimens produced similar results in terms of overall survival and progression-free survival, but left-sided tumors were associated with better survival than right-sided tumors.8 In an analysis restricted to the 1,137 patients with wild-type KRAS (for whom either agent could be effective), patients with tumors originating in the right side had a much shorter median overall survival (19.4 months) than those with left-sided tumors (33.3 months; HR = 1.60; P < .001).9 

Importantly, the study also showed the relative efficacy of cetuximab and bevacizumab, depending on the tumor location. Similar to the FIRE-3 data, median survival with cetuximab was 36 months for left-sided tumors but 16.7 months for right-sided tumors. With bevacizumab, the difference was less—31.4 vs 24.2 months, respectively. 

Finally, the CRYSTAL study evaluated cetuximab given with FOLFIRI in the first-line setting and also showed that for left-sided tumors, cetuximab improved progression-free and overall survival, whereas it had little impact on right-sided tumors.7 

Treatment Algorithm Taking Side Into Account

Dr. Wu put these factors together to make treatment recommendations. She said that right-sided cancers are more likely to be MSI-H, which should prompt consideration of pembrolizumab -(Keytruda, which was recently approved for chemotherapy--refractory MSI-H or mismatch repair–deficient tumors). Right-sided tumors may also harbor KRAS and BRAF mutations, indicating a poor prognosis; in patients with BRAF mutations, triplet chemotherapy with FOLFOXIRI (5-FU/leucovorin, -oxaliplatin, irinotecan) plus bevacizumab is warranted. 

For left-sided tumors, clinicians should consider the possibility of HER2 amplification, which would make trastuzumab (Herceptin) and lapatinib (Tykerb) treatment options, per the phase II -HERACLES trial. With more KRAS and NRAS wild-type tumors seen in the left colon, anti--EGFR antibodies should be strongly considered here. 

Bevacizumab can be useful for both right- and left-sided cancers, but anti-EGFR antibodies are useful only in left--sided tumors in the first-line setting. This has been incorporated into the NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines®), which suggest giving cetuximab or panitumumab (Vectibix) only to patients with KRAS/NRAS wild-type disease and left-sided tumors as initial therapy, she said.

Dr. Wu described her own approach as follows:

  • Right-sided tumors: Offer chemotherapy plus bevacizumab.
  • Left-sided tumors: Offer chemotherapy and bevacizumab or an anti-EGFR antibody (if KRAS and NRAS wild-type tumors). 
  • MSI-H tumors: Enroll the patient on a clinical trial of immunotherapy. 
  • BRAF-mutant tumors: Treat with FOLFOXIRI and bevacizumab.

As to which regimen—FOLFOX or FOLFIRI—is preferred in the first-line setting, “The jury is still out,” she said. “We know they are equal in the first-line setting, so we often tailor the chemotherapy to the patient’s tolerability of its side effects.” She added that she would use cetuximab with either regimen.

In the second line, patients can receive alternative chemotherapy plus a drug targeting vascular endothelial growth factor pathway (ie, bevacizumab, aflibercept [Zaltrap], or ramucirumab [Cyramza]). In the third line, patients with RAS wild-type disease benefit from the addition of an anti--EGFR agent to irinotecan if they have not already received an anti-EGFR antibody. Patients with tumors that harbor BRAF mutations should be considered for a clinical trial (ie, phase III BEACON CRC study with encorafenib, binimetinib, and cetuximab). Patients’ tumors should be tested for HER2 so that overexpressers can be offered HER2-directed therapy or enrolled in a trial (ie, phase II MOUTAINEER study with tucatinib and trastuzumab). Finally, patients with refractory disease could be offered trifluridine/tipracil (Lonsurf) or regorafenib (Stivarga), she suggested.

Dr. Wu added that other interesting approaches are being studied for late-line therapy, such as the combination of cobimetinib (Cotellic; a MEK inhibitor) plus atezolizumab (Tecentriq; a checkpoint inhibitor). Many phase I trials are underway for patients with actionable mutations. ■

DISCLOSURE: Dr. Wu reported no conflicts of interest.


1. Karim S, et al: JAMA Oncol. June 8, 2017 (early release online).

2. Weiss JM, et al: J Clin Oncol 29:4401-4409, 2011

3. O’Dwyer PJ, et al: J Clin Oncol 19:2413-2421, 2001.

4. Loupakis F, et al: J Natl Cancer Inst. 107:dju427, 2015.

5. Brulé SY, et al: Eur J Cancer 51:1405-1414, 2015

6. Heinemann V, et al: 2013 ASCO Annual Meeting. Abstract LBA3506. Presented June 1, 2013.

7. Tejpar S, et al: JAMA Oncol. October 10, 2016 (early release online).

8. Venook AP, et al: 2014 ASCO Annual Meeting. Abstract LBA3

9. Venook AP, et al: 2016 ASCO Annual Meeting. Abstract 3504. Presented June 5, 2016.