Genomics in Practice at the 2013 European Cancer Congress

Maurie Markman, MD


October 31, 2013

Many highly clinically relevant abstracts related to the revolution in our understanding of the molecular basis of malignant disease were presented at the 2013 European Cancer Congress. This report provides highlights of several of these provocative presentations.

70-Gene Panel Predicts Breast Cancer Recurrence

Nijenhuis and colleagues[1] sought to determine whether it is possible to define a patient population at high risk for both local and distant recurrence of breast cancer and, conversely, a population with a relatively low risk for recurrence. For this analysis, they used a set of 1053 breast cancer patients whose tumors were examined with a previously defined 70-gene panel.[2]

Their report at this conference[1] focused on locoregional recurrence -- and their answer seems to be "yes." They found that the 492 patients with a "high-risk" signature had a 9.5% 5-year risk for recurrence and 13.0% 10-year risk for experiencing such an event, whereas the 561 patients with a "low-risk" signature experienced a considerably reduced incidence of locoregional recurrence, showing a 5-year risk of only 2.7% and a 10-year risk of only 6.1% (P < .001). They postulate that the high-risk patients may benefit from more extensive adjuvant treatment to lower their risk, but precisely how these data will inform clinical practice remains to be seen.

Mutations in PI3K Pathway Predict Response to HER2-Targeting Therapy

A more direct application for genetic testing information can be seen in the study by Baselga and colleagues,[3] who reported on the negative impact of mutations in the PI3K signaling pathway and response of HER2-positive cancer to therapy delivered in the neoadjuvant setting.

For the approximately one quarter of tumors demonstrating a mutation in PIK3CA, KRAS, BRAF, or AKT1, there was a significant reduction in the percentage of patients who showed pathologic complete response at surgery, regardless of whether they were treated with trastuzumab, lapatinib, or both agents (odds ratio, 0.45; P = .015 vs wild-type). The authors of this provocative report suggested that such data may help clinicians identify a patient population unlikely to achieve clinical benefit if treated with a HER2-targeted antineoplastic agent.

KRAS Mutations Predict Colon Cancer Survival

Yoon and colleagues[4] analyzed the effect of specific KRAS mutations in codons 12 and 13 on outcomes among 2165 colon cancer patients with documented BRAF wild-type malignancies who were enrolled in a phase 3 trial of adjuvant therapy with FOLFOX alone or plus cetuximab. Whereas other reports have noted the negative association between outcomes in KRAS mutation-positive cancers if patients are managed with cetuximab,[5] the strength of the current study was the large sample size and the homogenous nature of the treated population with regard to disease stage and chemotherapy regimens used. (The treatment arms were combined for this analysis, as there was no evidence that KRAS mutation status affected therapeutic outcome.[6,7])

In this analysis, the investigators clearly demonstrated the negative impact on disease-free survival associated with the presence of codon 12 and 13 KRAS mutations vs KRAS wild-type. The 3-year disease-free survival rate was 68% in those with codon 12 KRAS mutations (P < .001) and 65% in those with a codon 13 KRAS mutation (P < .0032), but 77% in those with KRAS wild-type.

PIK3CA Mutations Predict Response to Aspirin in Colon Cancer Patients

Genetic testing in colon cancer patients can also be used to predict whether strategies to reduce the risk for recurrence might prove useful. Previously reported highly provocative retrospective data had suggested that patients with colon cancer possessing a mutation in PIK3CA who take aspirin regularly after their cancer diagnosis demonstrate a lower risk for cancer recurrence, but that aspirin use has no impact on tumors with wild-type PIK3.[8]

Following on these data, Church and colleagues[9] looked at the impact of aspirin use among 896 colon cancer patients who participated in a trial comparing rofecoxib with placebo after surgical removal of the cancer.[10] They found that the presence of a PIK3CA mutation had no effect on outcome for patients who received rofecoxib. However, regular use of aspirin after the diagnosis of cancer was shown to substantially decrease the risk for relapse in PIK3CA-mutant cancer (hazard ratio, 0.11; P = .027) but not in the absence of such a mutation (hazard ratio, 0.92). These confirmatory findings, which the investigators further elaborated on in the full study results published in the Journal of Clinical Oncology,[11] provide a strong rationale for examining colon cancers for the presence of PIK3CA mutations and considering long-term aspirin use if a mutation is discovered.

The Road Ahead

Although the reports discussed here must be considered preliminary until peer-reviewed publications are available, it is clear that these and other research programs in precision cancer medicine are rapidly altering our concepts of the optimal management of malignant disease. Such changes have the potential to rather substantially favorably affect the risk/benefit ratio associated with antineoplastic therapy.