Confusing evidence emerges on insulin glargine
Some wonder whether the data reflect the real cancer risk factors
Keywords: Diabetes quality improvement initiatives - Confusing evidence links insulin glargine to cancer
by Phyllis Maguire
Published in the September 2009 issue of Today's Hospitalist
A MAINSTAY OF INPATIENT GLYCEMIC CONTROL, insulin glargine (Lantus) made headlines this summer when a series of studies conducted in Europe and published in the journal Diabetologia published suggested a link between that therapy and cancer risk.
Those publications brought a swift disclaimer from the American Diabetes Association (ADA), which urged patients using glargine to continue the therapy, calling the research “conflicting and inconclusive.” That assessment is echoed by Jeffrey L. Schnipper, MD, MPH, director of clinical research for Brigham and Women’s Hospital’s academic hospitalist service and a
member of the Society of Hospital Medicine’s glycemic control task force.
|“The predictor of cancer isn’t the Lantus, but the reasons why a physician would choose to put a patient on that therapy by itself.”|
–Jeffrey L. Schnipper, MD, MPH
Brigham and Women’s Hospital
“The state of things right now is as follows,” says Dr. Schnipper. “Should we be treating these patients’ hyperglycemia? Yes. Should we change what we’re doing as far as prescribing Lantus? No. Do we need more studies? Yes.”
In the first of four retrospective studies, German researchers found a correlation between all insulins and cancer, as well as a dose-dependent association between cancer risk and Lantus, which is manufactured by Sanofi Aventis. (For patients taking a dose of 50U, for instance, the increased cancer risk was 31% compared to human insulin. The magnitude of possible risk was one additional cancer for every 100 patients taking Lantus for a year.)
The second study, based on a Swedish database, found no increased cancer risk in patients taking Lantus in combination with other forms of insulin, but it did find a two-fold increased risk of breast cancer in patients taking Lantus alone. A third study, using a Scottish database, likewise found a higher cancer risk among patients taking Lantus alone, including a higher risk of breast cancer.
But a fourth study done in the United Kingdom did not find higher risk associated with Lantus, either when Some wonder whether the data reflect the real cancer risk factors used alone or in combination with other therapies. It did, however, find a lower cancer risk associated with metformin. Most of the patients across all four studies who took Lantus alone were older, type 2 diabetics.
Today’s Hospitalist spoke with Dr. Schnipper about those results.
Have patients and colleagues been asking about the findings?
I’ve heard from colleagues, and I mentor other hospitals that are developing diabetes quality improvement initiatives, so the questions come up in that context. And I’ve heard a lot from Sanofi.
What do you hear from Sanofi?
They’re just trying to reassure people that the track record is solid and that these data are conflicting. It’s easy for the ADA to say, “Go talk to your doctor, if you’re a patient.” Sanofi wants to know what you the doctor are going to say when the patient comes in to talk.
You continue to use Lantus, correct?
Absolutely. In the hospital, the decision to put patients on Lantus or NPH (in addition to other insulins) has to do with patients’ individual circumstances. We’ll often use NPH for medical patients in the hospital, mainly for cost reasons, but for surgical patients, Lantus makes more sense because you don’t have to adjust the dose for whether or not they’re eating.
And when I need to send patients home on insulin for the first time, that’s probably going to be Lantus because it’s once a day and the easiest regimen for patients to take. You know that they will need close follow-up and they may need a more complicated regimen down the road, but when you’re discharging patients on insulin for the first time, you can’t make it that complicated.
The studies had conflicting results, but singled out patients using Lantus alone. What does that suggest to you?
Those kinds of conflicting results from observational data suggest confounding by indication. In other words, the predictor of cancer isn’t the Lantus, but the reasons why a physician would choose to put a patient on that therapy by itself.
In the outpatient setting, some patients may do wonderfully on Lantus alone. But often, the person who’s on Lantus alone is someone a physician has decided can’t handle a more complicated regimen. That’s the case even if a lot of those people may need a different insulin as meal-time coverage.
So their cancer risk may have nothing to do with Lantus, but everything to do with how often those patients get cancer screening or their diet—or any one of a thousand things having to do with lifestyle—that influences what medications a patient may be prescribed.
Theoretically, we know that insulin stimulates insulin-like growth factor, which is one of its major receptors. Is it possible that IGF stimulation predisposes you to cancers? There are certainly good pathophysiological reasons it could.
And with Lantus being the longest-acting insulin and causing probably the most sustained hyperinsulinemia, it could do that better than others. That’s part of why it works, and why it causes good glucose control.
Do any of these studies suggest safer alternatives?
They don’t, so questions that would have to be considered would include: How effective is Lantus with metformin or Lantus with pioglitazone, so you could get away with less Lantus? Or low-dose Lantus vs. pioglitazone and metformin without any insulin at all?
If you give people medications so they’ll need to take less insulin, is that good for their heart and longevity? And how does Lantus once a day compare to NPH twice a day, which is always an alternative? Some people like Lantus more because it causes less nighttime hypoglycemia.
Those are the questions you’d like answered with a randomized controlled trial, not with an observational study, and I don’t think those studies have been done with a long enough follow-up period. But no one is saying we should under-treat hyperglycemia because of the potential for increased cancer risk, especially with the magnitude of risk they’re talking about in these studies.
What’s the lesson that you take away from these studies?
That we need more follow-up data, particularly from randomized controlled trials, and I’m sure the FDA is asking Sanofi for those follow-up data. The FDA also needs more regulatory “teeth” and more funding for rigorous post-marketing studies to look at long-term endpoints for drugs it has already approved.
Now that the Agency for Healthcare Research and Quality is getting $300 million to look at comparative effectiveness research, I think the first study we need in the inpatient setting is to look at whether standardized glucose control with sub-Q insulin in non-critical care units is better than usual care for hard outcomes.
We don’t have randomized controlled trial data showing that tighter control matters in the non-critical care setting, and comparing Lantus to NPH as basal insulin could be a part of that trial. But we’re never going to find the cancer risk of any type of insulin on the inpatient side because we’re talking about needing years of follow-up. Right now, there just aren’t enough data to change what we do.
Phyllis Maguire is Executive Editor of Today’s Hospitalist.