September 29, 2015

The Newly Approved Basal Insulin Tresiba's Label is Disappointing

We have been hearing for quite a while about the benefits of Novo Nordisk's new, longer acting basal insulin, Tresiba, whose approval was delayed for a year while the FDA waited for cardiovascular safety data .

Well, the FDA finally approved it last week. And lo and behold, none of the main claims that have been made for Tresiba until now appear in the official, FDA-sanctioned Prescribing Information (PI), A.K.A. "label."

You can read the whole PI/label here:

The hype was that Tresiba caused fewer hypos than existing basal insulins. If this is so, it is not demonstrated anywhere on the label. There were hefty hypo rates reported in 6 of the 7 studies summarized in the label. The percent of of people in these studies who had hypos where blood sugar dropped below 56 mg/dl via a blood sugar meter test were: 46.5%,  28.5%,  50%,  43.8%,  50.9%,  80.9% and 42.5%.  (Page 9 of the PI)

There is no way to directly compare these studies with studies of other basal insulins, as it is clear that there is really no way to compare any of these studies of Tresiba with any other study of Tresiba. When you have a range of outcomes from 28.5% to 80.9%, means and medians are worthless. There is just too much variation from study to study. So comparisons with studies of other insulins that would demonstrate superiority would have to be cherry picked. I am certain, of course, that they will be. Drug salespeople are extremely good at coming up with creative ways to promote drugs in ways that are not actually substantiated by existing research.

(It's worth noting that Toujeo's claims of causing fewer hypos were also removed from its U.S. when it was recently approved.  Toujeo is the latest incarnation of Lantus, which is being pitched as a "new" insulin though it is actually the same insulin molecule as is found in Lantus, just dressed up with a new pen and a new, higher price tag.)

The implication of the rest of the hype for Tresiba was that its longer action  would make it superior to Lantus, which currently dominates the niche for basal insulins. But oddly--and one has to assume the FDA insisted on this--the charts displayed from page 20 to page 24 of the PI clearly demonstrate that study after study found that patients on Lantus, be they Type 1s or Type 2s, saw better improvements in their A1c on Lantus than those on Tresiba. Lantus consistently achieved an improvement of  .1 to .2% in A1c.--i.e. if Tesiba lowered A1c on average to 7.1%, Lantus (called by its generic name, glargine, on the charts) brought its group down on average to 6.9%.

The reason for this may have to do with the way that Tresiba works. The chart showing the activity curve of Tresiba is worrisome. The insulin is active for 42 hours but the curve is far from flat. Since patients are told to inject it once a day, and furthermore told that they can inject it different times from day to day, the overlap between the previous day's insulin and the next day's insulin may be unpredictable.

We see from the comparison charts where Tresiba was matched with Lantus that despite the lower impact on A1c patients on Tresiba had to reduce their fast-acting insulin at meal times. This is undoubtedbly because of that very long tail of the previous day's insulin working in unpredictable ways.

The only real advantage with Tresiba is that it is available in a U200 concentrated form that will allow a single dose of up to 160 units. This is helpful for people with extreme insulin resistance who need huge doses and who, until now, had to take two or more shots of basal each day. But the population of those needing 160 units is small. For most people the main impact on them of switching to Tresiba will be that they or their insurer will be paying more for basal insulin while they will be likely to see less improvement in their control and a continuing possibility of experiencing serious hypos.

Another supposed benefit of Tresiba is that it can be mixed in the pen with fast acting Novolog, allowing it to be sold as a 70/30 insulin.  This 70/30 Tresiba/Novolog mix will be sold under the name Ryzodeg.

70/30 insulin is probably the worst form of insulin available, as putting basal and fast acting insulin into a single shot makes it impossible to match the dose of fast-acting insulin being given to the person's insulin resistance or to the the amount of carbs they eat at the meal that comes after their two daily shots. So this generic dosing of  meal-time insulin which occurs with 70/30 insulins guarantees the worst of both worlds. The choice is either more hypos or poorer control than is possible with dosing each meal separately for the rest. Insurers like it because they pay for only one prescription, not two. Some patients like it because it requires fewer shots. But no doctor who knows anything about insulin--or cares about his patients' long term outcome would prescribe this kind of insulin.

Nevertheless, you can expect to see plenty of advertising promoting this mixture of two insulins in one shot as a "convenience."  The main reason that Novo Nordisk is going to be selling Ryzodeg, of course, is not that it is good for patients. They introduced it because they know that that Lantus can not be mixed with a fast acting insulin--and Novo Nordisk is locked in a battle to the death for diabetes dollars with Sanofi, the maker of Lantus and Toujeo.

Keep in mind, both Tresiba and Toujeo come to the market under the shadow of the approaching launch of biosimilar glargine--Lantus equivalents which can be sold now since the Lantus patent has expired. For now, "can be sold" doesn't mean "will be sold." Sanofi is doing all it can to delay the day that cheaper but still effective basal insulins are available to American consumers. They just signed an agreement with Lilly, that will delay the launch of Lilly's biosimilar Lantus until 2017. This is what would be called "restraint of trade" anywhere that had real regulation of drug pricing and marketing (like, say, the EU, where biosimilar Lantus is already available). But here in the U.S. "free market" economics appear mean that companies are "free" to keep helpful drugs off the "market" as long as they pay off their big pharma peers--peers who will do the same for them in the future, to keep the profits high for both.

All of this is bad news for people with diabetes, of course, who are paying more than twice what they were paying for basal insulin just 8 years ago, while still contending with unexpected hypos. But their interests are overlooked since the business world appears to believe that the only important question about any new insulin is how much money can be extracted from those who are obliged to use it.