At the 2013 European Cancer Congress, two investigative teams attempted to explain how aspirin may protect against colorectal cancer recurrences, with one study showing PIK3CA mutations associated with protection from aspirin, but not a COX-2 inhibitor, and the other study implicating HLA class I antigens but finding no PIK3CA association.
Aspirin and other nonsteroidal anti-inflammatory drugs (NSAIDs) are already believed to protect against colorectal cancer, having been associated with reduced disease recurrence following primary treatment. Since these drugs carry toxicity, research has focused on finding a subgroup that might benefit most from the drugs’ potential effects.
Biomarkers of benefit have been proposed, including cyclooxygenase-2 (COX-2) overexpression and germline UGT1A6 polymorphisms. Recent data suggest aspirin’s benefit may be limited to patients with PIK3CA-mutant cancers, as reported recently in The New England Journal of Medicine,1 where aspirin use was associated with a hazard ratio (HR) of 0.18 (P < .001) for colorectal-cancer survival in patients with PIK3CA-mutant tumors.
VICTOR Affirms Role of PIK3CA Mutations
The VICTOR study sought to determine the benefit of COX-2 inhibition in 2,424 patients who were randomly assigned to receive rofecoxib (Vioxx) at 25 mg/d or placebo following primary resection.2 In a biomarker substudy of 896 patients,3 of whom 11.6% had a PIK3CA mutation, relapse-free survival and overall survival were compared between the rofecoxib and placebo arms, according to mutation status, and between low-dose (< 100 mg/d) aspirin use (14%) and nonuse (86%), as noted at randomization. The study was discontinued after a median duration of rofecoxib therapy of 7.4 months, due to adverse cardiovascular events.
“We found no evidence of a greater benefit from rofecoxib treatment compared to placebo in patients whose tumor had PIK3CA mutations, compared to PIK3CA wild-type cancers,” said David Church, MD, of the Oxford Cancer Centre at the University of Oxford in the United Kingdom. Among patients with PIK3CA-mutant tumors, recurrence-free and overall survival were not significantly different, whether they received rofecoxib or not.
In contrast to the findings for rofecoxib, regular aspirin use after diagnosis was associated with a reduced rate of recurrence in patients with PIK3CA-mutant cancers (HR = 0.11; P = .036) but not in cases lacking PIK3CA mutations (HR = 0.94; P = .71). The test for interaction between the mutation and recurrence was statistically significant (P for interaction = .024).
Among the 14 PIK3CA-mutated patients taking low-dose aspirin, there were no relapses, whereas there were 23 relapses among the 90 patients with mutations but not taking aspirin. The hazard ratio for overall survival was 0.29, but due to the small number of deaths (including only one death in the aspirin group), this was not statistically significant, he added.
The discordance between the benefits of rofecoxib and aspirin in PIK3CA-mutant tumors suggests that aspirin’s effects may be, at least partly, COX-2–independent, Dr. Church suggested.
“PIK3CA mutation merits further evaluation as a predictive biomarker for aspirin therapy. Our findings are concordant with recent data and support the prospective investigation of adjuvant aspirin in PIK3CA-mutant colorectal cancer,” he said.
Aspirin May Affect Immune System
Dutch investigators have another theory for how aspirin improves outcomes after a colorectal cancer diagnosis, showing its benefits in patients whose tumor cells express human leukocyte antigen (HLA) class I, a cell-surface protein produced by genes involved in immune function.
Marlies S. Reimers, MD, of Leiden University Medical Center in The Netherlands, and colleagues examined the expression of HLA class I proteins and COX-2 in 999 tumor samples of colorectal cancer from patients whose aspirin use was known. They also examined PIK3CA gene mutations in 663 tumors. Tumor markers were related to survival.4
Dr. Reimers reported that low-dose (80 mg/d) aspirin use after diagnosis was associated with a 47% better overall survival in patients with tumors expressing HLA class I, at a median time of 4 years postdiagnosis.
“Aspirin benefit was not associated with a PIK3CA mutation or COX-2 expression of the tumor. Low-dose aspirin use after diagnosis improved survival only in patients with tumors expressing HLA class I,” Dr. Reimers reported.
“We think that platelets are involved in metastasis by shielding tumor cells in the bloodstream, so that they cannot be recognized by the immune system. Aspirin could help to ‘unmask’ those tumor cells by attacking platelet formation, so that the immune cells can detect and eliminate them,” she suggested.
Interestingly, their investigation did not find a preferential benefit of aspirin in patients with PIK3CA mutations, nor was it related to COX-2 expression. It is possible that aspirin is acting on two different pathways in colon cancer: one related to the evolution of polyps to cancer and to cancer recurrence (PI3KCA, COX-2), and the other (HLA class I) related to metastasis by influencing platelets in the bloodstream, Dr. Reimers commented.
“Although speculative, it may be that the interaction of platelets with HLA-positive tumor cells circulating in the blood promotes the metastatic potential of these cells,” she said. Dr. Reimers added that it is easier to measure HLA class I on tumor cells than to determine PIK3CA mutation status.
“It is important to acknowledge that the data presented by Dr. Reimers are preliminary and await publication,” said senior author Gerrit-Jan Liefers, MD. “Also, our data need to be validated, preferably in published cohorts. However, we feel that designing a randomized aspirin trial on the basis of PIK3CA mutations is premature. It would be better to stratify patients for various molecular markers (eg, PIK3CA, HLA1). ■
Disclosure: Drs. Church, Reimers, and Liefers reported no potential conflicts of interest.
1. Liao X, Lochhead P, Nishihara R, et al: Aspirin use, tumor PIK3CA mutation, and colorectal-cancer survival. N Engl J Med 367:1596-606, 2012.
2. Church D, Domingo E, Sieber O, et al: Evaluation of PIK3CA mutation as a predictor of benefit from NSAID therapy in colorectal cancer. 2013 European Cancer Congress. Abstract 2161. Presented September 29, 2013.
3. Domingo E, Church DN, Sieber O, et al: Evaluation of PIK3CA mutation as a predictor of benefit from nonsteroidal anti-inflammatory drug therapy in colorectal cancer. J Clin Oncol. September 23, 2013 (early release online).
4. Reimers MS, Bastiaannet E, Langley RE, et al: Low dose aspirin use after a colon cancer diagnosis and survival in HLA class I expressing tumors. 2013 European Cancer Congress. Abstract 2183. Presented September 30, 2013.