Kate Johnson

December 12, 2011

December 12, 2011 (Dubai, United Arab Emirates) — In an impressive public relations feat stretching from Dubai to San Antonio, Texas, representatives from sanofi-aventis, the makers of Lantus (insulin glargine), scrambled this week to deal with new concerns about the drug's cancer risk.

Just as a meta-analysis by the company showing a decreased risk for cancer with the use of glargine, compared with other insulins, was about to be presented here at the International Diabetes Federation (IDF) World Diabetes Congress 2011, a contradictory study was being presented at the 34th Annual San Antonio Breast Cancer Symposium.

In that study, Hakan Olsson, MD, from the Departments of Oncology and Cancer Epidemiology at Lund University in Sweden, analyzed overall cancer risk in 1499 glargine users and 8075 metformin users, and found that glargine was associated with an odds ratio (OR) of 2.88 (95% confidence interval [CI], 1.15 to 6.64), whereas metformin was associated with a nonsignificant trend toward lower risk (OR, 0.92; 95% CI, 0.82 to 1.09).

But Peter Boyle, MD, principal investigator and presenter of the meta-analysis in Dubai, and president of the International Prevention Research Institute in Lyon, France, said once this Swedish study is added to the meta-analysis, it loses significance.

"With all the fuss that was in the press, once you absorb that into all the available data, it adds nothing at all," he said.

To demonstrate this, he presented the results of the meta-analysis, first without and then with the results from the Swedish study.

The original meta-analysis looked at 7 database studies examining the association of glargine and other insulins with all forms of cancer, 10 studies looking specifically at breast cancer, and 5 studies each looking at colorectal and prostate cancer. Overall, the analysis involved 80,000 patients enrolled in clinical trials with 38 million patient-years of treatment exposure, said Dr. Boyle.

Dr. Boyle's team calculated the summary relative risks (SRR) and found a decreased risk for overall cancer among glargine users, compared with users of other insulins, and no increased risk for breast, colorectal, or prostate cancer.

"We're not saying that glargine protects against all types of cancer, but if we're conservative, there's no indication it increases the risk," Dr. Boyle emphasized.

Specifically, the SRR for overall cancer risk was 0.87 (95% CI, 0.78 to 0.96), for breast cancer it was 1.08 (95% CI, 0.92 to 1.27), for colorectal cancer it was 0.80 (95% CI, 0.61 to 1.05), and for prostate cancer it was 1.16 (95% CI, 0.90 to 1.49).

When the Swedish study was incorporated into the analysis, the SRR for breast cancer rose marginally, from 1.08 (95% CI, 0.92 to 1.27) to 1.10 (95% CI, 0.94 to 1.30), said Dr. Boyle.

In a press release from sanofi-aventis, Dr. Boyle said the meta-analysis "highlights the need to go beyond single-study reports...and place the findings from any single study in the context of all available information."

"We've got to make the best we can of the available data and be very cautious of any interpretations, particularly of any one study," he said.

However, he bemoaned the lack of good data on the subject.

"I think it's extremely unfortunate that this has grown into such a big topic. I think it's a business, it's an industry. But to me, my background is in cancer epidemiology, and these database studies have not been designed for the use that everyone is putting them to these days.... I think we may be torturing them a little too much."

"It's another meta-analysis, and meta-analyses cannot, in my mind, replace clinical trials," said Steven Kahn, MD, head of the IDF's scientific program committee, and professor of medicine at the University of Washington in Seattle. He cochaired the session in which the data were presented.

"You can't rely on one study...because even though it had some effects that suggest increased risk, when you put it together with everything, it gets washed out," he told Medscape Medical News.

Dr. Boyle is the principal investigator of another large-scale study sponsored by sanofi-aventis, requested by the European Medicines Agency. The results of that study are expected soon, he said.

In addition, 2 retrospective cohort studies and 1 case–control study are currently being conducted by independent investigators, according to the company.

"When studies produce conflicting results, it can be confusing for the clinician," said Anne Peters, MD, professor of clinical medicine at the Keck School of Medicine at the University of Southern California (USC), and director of the USC clinical diabetes program in Los Angeles, California, when asked to comment on the findings.

In some cases, a well-done large, single RCT [randomized controlled trial] can answer a specific question regarding drug safety and/or efficacy. A good RCT can change how we treat patients. An epidemiologic study [such as the Swedish study] generally suggests trends, but is limited in terms of proving causality. A meta-analysis [such as that conducted by Dr. Boyle and colleagues] is limited by the quality of the pooled data and the robustness of the mathematical approach to account for differences in datasets and covariates," she told Medscape Medical News.

"It is known that type 2 diabetes increases the risk of certain types of cancer. It is also generally accepted that using insulin tends to occur later in the progression of the disease, when many complications, perhaps including cancer, have already started to develop. Unless the Olsson study is carefully adjusted for duration of diabetes between those using glargine and those using metformin, as well as obesity and other variables that might make one subset more likely to develop breast cancer than other, then the data aren't particularly valid," Dr. Peters said.

"Since I have not read the study methods and results, I can only speculate."

The results of the meta-analysis fit "with what I would like to believe — that glargine is not associated with an increased risk of cancer. But I await long-term RCTs to help inform my final opinion."

As a clinician, Dr. Peters said she is guided by the following principles: "Metformin is a great drug for treating type 2 diabetes and should be used alone or in combination with other agents, as well as insulin. Obesity is a significant risk factor for cancer and a host of other issues, so I work hard on lifestyle modification with my patients. Following appropriate screening guidelines for cancer is imperative. Finally, I use glargine because it is an effective long-acting insulin that is associated with less hypoglycemia (which is a serious problem in patients with diabetes). I follow as carefully as I can all reports about the safety of all the drugs I use in my patients, adjusting my treatment when evidence renders the risk greater than the benefit."

Dr. Boyle reports receiving honoraria from and being a consultant for sanofi-aventis, the manufacturer of glargine. Dr. Peters reports serving as an advisor or consultant for Amylin Pharmaceuticals, AstraZeneca Pharmaceuticals, Eli Lilly and Company, Takeda Pharmaceuticals North America, Bristol-Myers Squibb, Novo Nordisk, Roche, Dexcom, Medtronic MiniMed, Merck & Co., and Abbott Diabetes Care.

International Diabetes Federation (IDF) World Diabetes Congress 2011. Presented December 8, 2011.