Sunday, June 28, 2009

Lantus and Cancer

Everyone seems to be talking about the recent report that insulin glargine (Lantus) might increase cancer rates. Needless to say, this is very upsetting news to a lot of people, although Lantus has been linked with cancer in the past.

But like most news stories, the facts aren't as simple as the summaries would indicate. If you like reading all the details, they're available here in full text, free.

The method of publishing these papers was interesting. Apparently the first study was submitted to the journal Diabetolgia last year. But the editors decided not to publish it without further evidence. So they commissioned three additional studies to see if the results could be reproduced. Those plus the original article are the four articles available at the journal’s Web site.

Was this irresponsible or responsible? Not publishing results suggesting some bad side effect of a drug might mean that people were harmed because they didn’t know about it. On the other hand, publishing preliminary results suggesting some bad side effect of a drug might mean that people were harmed because the huge publicity in the popular press that usually follows such news would scare patients who could be helped by the drug so they’d stop taking it.

This happened with the drug rosiglitazone (Avandia) last year. One analysis suggested that it caused heart disease, there was a huge glut of articles and blogposts saying Avandia could kill you, and a lot of people stopped taking it. Their A1cs then increased, and high A1cs also increase the risk of heart disease.

So it’s sort of a damned if you do and damned if you don’t situation.

Most patients don’t have the statistical skills to analyze research articles, especially meta-analyses, which try to combine results of numerous trials to see if they can come to overall conclusions. Because the patient populations, time frames, and endpoints being studied in different studies differ, this is difficult to do, and it’s easy to disagree with the results of a meta-analysis.

But regardless of whether or not we approve of the journal’s method of dealing with the Lantus-cancer articles, there are some things patients should understand.

In the editorial accompanying the articles, the authors make a few interesting points:

1. Type 2 diabetes is associated with increased rates of cancer of the colon, breast, and pancreas. People with type 2 diabetes have insulin resistance and hence need more insulin, either produced by their own beta cells or injected.

2. Cancers of the colon, breast, and pancreas have been associated with increased circulating levels of insulin in nondiabetic people. Obviously they wouldn’t be injecting insulin. They’re probably insulin resistant.

3. Metformin reduces the rates of cancer of the colon and pancreas but not cancer of the breast and prostate. (Metformin also seems to reduce the pancreatitis that can result from sitagliptin [Januvia].)

4. Insulin is a growth factor for both healthy cells and cancer cells in cell culture.

5. Evidence suggests that insulin doesn’t cause cancer, but it may increase the rate of growth of cancer cells that have been caused by something else.

6. Most elderly people have some early cancer cells. For example, 90% of men older than 90 years have prostate cancer cells. (The body often keeps these cells in check, or even destroys them.)

7. One early insulin analogue, B10Asp, was found to increase cancer growth in rodents and was never marketed. But "B10Asp would have passed the carcinogenicty testing to which insulin glargine was subjected and would now be in clinical use."

8. Lantus increases mitogenic potency (mitogenic means it causes cell division or transformation into another cell type, for example, a malignant one) six to eight times. The short-acting insulins (e.g., Novolog and Humalog) have little effect. Detemir (Levemir) seems to reduce the mitogenic potency in vitro, but the authors say that because of technical difficulties, this was difficult to measure.

There were all kinds of confusing factors in the four studies published by Diabetologia. For instance, they classified everyone diagnosed when older than 30 as type 2. We know that a lot of type 1s and LADAs aren’t diagnosed until later in life.

In the German study, the patients receiving regular insulin were taking larger doses than the patients taking insulin analogues. In another study, those taking only Lantus were older than those taking Lantus plus a bolus insulin. The ones taking only a basal insulin were most likely type 2s, who would need large doses of insulin, and we know that cancer rates increase with age among all groups.

In their editorial, the authors conclude that

1. “There is no evidence that insulin, however formulated, causes cancer.”

2. But “the growth of some tumor cells lines is clearly enhanced by insulin.”

3. “Circulating levels of endogenous insulin appear to be associated with cancer risk in obesity and other insulin-resistant conditions, including type 2 diabetes.”

4. “There is no evidence of harm in type 1 diabetes, or in males, or in premenopausal breast cancer.”

5. “On current evidence, the short-acting analogues do not appear to present a potential problem.”

In other words, there is evidence that very high levels of insulin, no matter what the source, may increase the growth of pre-existing cancers. People with insulin resistance (metabolic syndrome or type 2 diabetes) are at increased risk whether or not they inject insulin of any kind.

Hence, for us, the best approach would be to do whatever we can to reduce the amount of insulin that we need. Reducing our insulin resistance through exercise and, if possible, weight loss should help.

Another approach, available to everyone regardless of ability to exercise or lose weight, is to reduce the amount of carbohydrate we eat.

The less carbohydrate you eat, the less insulin you need. The less insulin you need (either your own or injected), the lower your risks of encouraging the growth of cancer cells. It seems me this is pretty easy to understand.

But some people don't seem to be able to grasp this. How long will it be before the American Diabetes Association stops telling people with type 2 diabetes to eat more carbohydrate and “Make starch the star”?

Saturday, June 27, 2009

Body Build and Diabetes

You can't tell if someone is diabetic simply by looking at them.

Yet a lot of people still think you can. They buy into the idea that eating too much and exercising too little causes obesity and obesity causes type 2 diabetes. Hence, if shown a bunch of photos of strangers, many would predict that the fat ones had diabetes and the thin ones didn't.

It is true that being overweight is
associated with type 2 diabetes, and many people with the disease are, in fact, overweight. But not everyone who is overweight, or even obese, has diabetes, and not everyone who has type 2 diabetes is overweight.

While filing sometime recently, I came across a popular press article that I had found fascinating.

A British newspaper had asked 10 people between the ages of 35 and 50 who had never been diagnosed with diabetes to take "a blood test." The test was simply a
fingerstick test in a home-type meter. If the participants said they hadn't eaten (most likely not rigorously controlled, so it wasn't a real fasting test but closer to a premeal test), any result over 5.9 mmol/L (106 mg/dL) was considered suggestive of diabetes. If they said they'd eaten recently, the cutoff was 8.9 mmol/dL (160 mg.mL).

If the results were over these limits, the people were told to see their physician for more rigorous tests. They don't say which readings were
premeal and which were after eating, so one person with a reading of 8.9 mmol/L (160 mg/dL) was labeled diabetic and another with a reading of 9.1 (164 mg/dL) was labeled "needs investigation." Probably the former reading was premeal and the latter was after eating.

Here's the interesting part. They then photographed the participants wearing gym togs, so you can compare their body builds with high blood glucose (
BG) readings. It's also interesting to compare the body builds with the BMIs.

In this tiny sample, there were a lot of results that go against "common knowledge."

One person labeled obese (
BMI 34) had normal BG levels. A person with a BMI of 22 had high BG levels. A woman with a pear shape, which is supposed to be healthy (it's the apple shape with most of the weight in the stomach that is thought to be dangerous), had high BG levels. A woman with a lot of risk factors had normal BG levels. A vegetarian was obese. And the woman with the lowest BG reading said she got no exercise at all.

Scroll down through the initial text, which is the usual popular press stuff about diabetes, and try looking at the photos before you read the descriptions underneath them to see how your predictions compare with the facts.

Many of these participants had relatives with diabetes, which is probably one reason they volunteered for the tests. And they were relatively young. Some of those who tested normal with this fairly uncontrolled test may develop diabetes when they get older.

But it's still interesting to see how different people are and how misleading body build can be as a predictor of diabetes.

The sad thing is that because of the constant barrage of news stories saying that obesity is the cause of type 2 diabetes, many people (probably most people) believe it. And that includes a lot of physicians, who might not bother to do
BG tests on someone who was thin, said they "ate healthy" (whatever that means), and got regular exercise.

With the advent of home meters, it's easy to test friends and relatives if you think they might be at risk. Just make certain you use fresh lancets when you do, as well as cleaning off the tip of the finger pricker, to avoid passing on any blood-borne diseases. Even better, ask them to get their own finger-pricking devices.

Because type 2 diabetes usually starts with elevated
postmeal BG levels, measuring after a large meal, especially a carby meal, would be the best place to start. People can have normal fasting BG levels and elevated postmeal levels for years before they get a diagnosis. The earlier they learn they're at risk, the easier it will be to take corrective action, like limiting the carbohydrate content of their meals.

And if you are testing relatives, don't ignore the thin ones who exercise. Diabetes is a complex disease, and it happens to apparently healthy people too.

Thursday, June 11, 2009

Monday, June 1, 2009

Right on!

Sometimes a cartoon says it all. This one does it for me.

It's funny, but it's also sad. So many small discoveries are magnified by PR departments and then news media that the public has become cynical about every new discovery.

Some day we'll find cures for cancer, AIDS, and diabetes, and people will read the stories in the newspaper or see them on TV and think, "Yeah, right. Next week they'll announce that the cure causes something even worse."

Sadly, this sort of thing has been going on for a long time, and I doubt that it will change in the near future.