SAN FRANCISCO ― Most cases of colorectal cancer are preceded by adenomas. Data show that chemoprevention with selective cyclooxygenase-2 (COX-2) inhibitors, particularly celecoxib (Celebrex, GD Searle), can reduce the incidence of these lesions.
However, there has been some hesitation over the use of celecoxib and related agents for chemoprevention because of concern over cardiovascular risks, as well as other possible side effects. But now a potential biomarker that identifies individuals who would derive most benefit has been found, which would personalize use of such chemoprevention.
The new findings were presented here at the Gastrointestinal Cancers Symposium (GICS) 2017.
Using archived adenoma tissue, the authors sought to identify predictors of response to celecoxib. They focused on several biomarkers within the COX metabolic pathway.
The findings revealed that neither COX-2 nor 15-prostaglandin dehydrogenase (15-PGDH) was prognostic for adenoma recurrence, based on pretreatment drug levels, explained lead author Jiping Wang, MD, PhD, of the Dana-Farber Cancer Institute, Boston, Massachusetts.
Both markers did, however, show predictive strength.
Celecoxib significantly reduced adenoma recurrence in both high-expression and low-expression COX-2 groups, with a trend toward better results in the high COX-2 group.
For 15-PGDH, it was its absence, not its presence, that was associated with a benefit from celecoxib, noted Dr Wang.
Celecoxib inhibits expression of prostaglandin E2 (PGE2), which is an inflammatory mediator produced by fatty acid metabolism via cyclooxygenases.
COX-2 promotes PGE2 expression, and 15-PGDH is a tumor suppressor that degrades PGE2. The knockout of 15-PGDH increases the susceptibility to carcinogens.
"We hypothesized that the expression of COX-2 and 15-PDGH might relate to the tumor-preventive efficacy of the selective COX-2 inhibitor celecoxib," Dr Wang said when he presented the study findings at the meeting.
Reduction of Adenomas in APC
The Adenoma Prevention With Celecoxib (APC) trial found at the 3-year end point that patients at high risk of developing an adenoma who received 400 mg of celecoxib had 33% fewer lesions and 57% fewer advanced lesions.
The study was halted early when the researchers became aware of associated cardiac toxicity. At that point, nearly all of the enrolled patients had received treatment for 3 years. It was decided to continue monitoring patients for an additional 2 years to assess safety and efficacy.
After 5 years, the rate of adenomas was reduced by 41% in patients who took the lower dose (400 mg) of celecoxib and by 25% in patients who took the higher dose (800 mg).
Analyzing the Archived Tissue
In the current study, Dr Wang and colleagues used tissue samples from the APC trial to validate their hypothesis.
Of the original cohort of 2035 participants, biomarker data were available for a subset of 1295 (71%) patients. These data were used for the current analysis.
The adenomas that were collected from APC trial participants were assessed for expression of COX-2 (high vs low) and for 15-PGDH (present vs absent) using immunohistochemistry.
A combined variable to estimate tumor PGE2 levels identified cases as PGE2 low (COX-2 low/15-PGDH present) or PGE2 high (COX-2 high/15-PGDH present; COX-2 high/15-PGDH absent; or COX-2 low/15-PGDH absent).
"We found that the majority of participant had a loss of PGDH as well as PGE2 high," said Dr. Wang.
In the placebo group (n = 440) 15-PGDH was absent in 88.2%; 6.4% had high COX-2, and 90% had high PGE2.
In the celecoxib group, 15-PGDH was absent in 88.9%; 9.1% had high COX-2, and 91.8% had high PGE2.
This is consistent with what has been seen in the literature, he explained.
In samples obtained from patients who had received placebo rather than celecoxib, there was no difference in adenoma recurrence between the high and low COX-2 groups, which indicated that pretreatment levels of COX-2 were not prognostic for adenoma recurrence, Dr Want noted.
What they did find was that celecoxib was able to reduce the risk for recurrence in both the high and low COX-2 group. Those with high COX-2 expression at baseline had a 52% reduction in the risk for adenoma recurrence, as compared with placebo (relative risk [RR], 0.37; P = .0001). For those with low COX-2, the risk was also reduced with celecoxib, but at a lower rate (30.5%; RR, 0.64; P < .0001).
"These findings suggest that higher COX-2 level predicted a greater response to celecoxib," explained Dr Wang.
The presence of 15-PGDH was associated with a small risk reduction (RR, 0.73; P = .15; risk reduction, 17.7%), but a loss or absence was associated with a significant risk reduction of 33.8% (RR, 0.60; P < .0001).
"This tells us that 15-PDGH is not prognostic for adenoma recurrence but is predictive for response to celecoxib," said Dr Wang. "Those who retain 15-PGDH expression may not benefit from celecoxib."
They found that the estimated adenoma PGE2 levels yielded results consistent with the 15-PGDH analysis.
There was no strong evidence to show that celecoxib would benefit the subgroup with low PGE2 (RR, .95; P = .82; risk reduction, 0%). For high PGE2, the risk reduction was highly significant, at 34.9% (RR, 0.59; P < .0001).
"These results suggest that PGE2 is a significant driver of tumorigenesis in the colorectum for many, but not all, patients with adenoma," Dr Wang said.
Discussing the paper, Frank A. Sinicrope, MD, FACP, of the Mayo Clinic, noted that the "future is precision prevention, where biomarkers can be used to select patients most likely to benefit," and that he would like to "congratulate the authors on their study and their attempt to try to make precision prevention a reality."
However, he pointed out that the biomarker analyses were generally underpowered, "given the small number of patients in both the COX-2 high and 15-PGDH present groups.
"The absence of 15-PGHD expression predicted benefit from celecoxib, but the mechanism is unclear and awaits trial," he added.
No funding for the study was disclosed. Dr Wang has relationships with Cisco Systems and GenomiCare; the other authors have disclosed no relevant financial relationships. Dr Sinicrope has relationships with Gilead Sciences, illumine, EMD Serono, and Ventana Medical Systems.
Gastrointestinal Cancers Symposium (GICS) 2017. Abstract 524, presented January 21, 2017.