Long-Term Phentermine for Weight Loss Appears Safe, Effective

Miriam E. Tucker

November 15, 2018

NASHVILLE, Tenn — Long-term, off-label use of the weight-loss drug phentermine appears to be effective and safe, new data suggest.

Findings from an electronic record analysis involving nearly 14,000 adult phentermine users were presented November 14 in a late-breaking session here at Obesity Week 2018 by Kristina H. Lewis, MD, of Wake Forest University Health Sciences, Winston-Salem, North Carolina.

Phentermine, a generic appetite suppressant approved by the US Food and Drug Administration in 1959, is indicated as a "short-term adjunct (a few weeks) to lifestyle-based programs," for weight loss. Concerns about long-term use include the potential for cardiovascular disease and addiction, Lewis explained.

"Phentermine is the most widely used weight loss medication in the United States but currently it's labeling is for short-term use. That doesn't align with the idea of obesity as a chronic condition...but it's generic, and there's not a lot of incentive to do a big trial," said Lewis.

The 2012 approval of the combination drug phentermine/topiramate (Qsymia, Vivus) based on data demonstrating safety and efficacy up to 2 years prompted her group to attempt to collect similar data for the solo generic drug.

"Now that phentermine/topiramate is approved, data suggest that it's safe and effective for longer-term use, but we haven't really looked at it on its own, so that's what we tried to do," she told Medscape Medical News.  

There is a need, since many payers will only cover phentermine and not the more expensive branded combination. "If you look at the epidemiology of obesity and who's more affected, you want to be able to prescribe affordable drugs for lower-income patients," Lewis commented.

Although her data show that longer-term phentermine use is associated with clinically significant greater weight loss at 2 years — compared with short-term use — and without increased incidence of cardiovascular disease or death, she cautioned that the analysis is not definitive. "I really feel strongly that clinical trials are needed," she stressed.

Asked to comment, Stephen R. Smith, MD, chief scientific officer of Florida Hospital Research Services, Orlando, told Medscape Medical News, "I think it's a nice preliminary study...I think it will give physicians some confidence since there are now more data in the public domain."

But Smith also pointed out that although this phentermine study was large, it was underpowered to detect a safety signal, noting that "they had a big enough cohort that we can preliminarily see there's nothing here that makes us worry."

On the other hand, "Your population needs to be sick enough. This was generally healthy, predominantly female, and middle-aged — representative of those who take phentermine. I think it's a nice preliminary study. It isn't conclusive, but it's a great start."

Longer Phentermine Use Associated With Greater Weight Loss

Lewis and colleagues used electronic health records data from eight systems participating in the Patient Outcomes Research to Advance Learning cohort, which includes Kaiser sites in seven US regions and Denver Health and Health Partners. The study period was January 2010 to September 2015.

Participants included 13,972 adults with a body mass index ≥ 27kg/m2 (average 37.2 kg/m2) and a first phentermine prescription fill during the study period. Overall, 84% were women, 45% were white, and were a mean 43.5 years of age. Rates of comorbidities were relatively low and included hypertension (21%) and diabetes (12%), and 34% were ever or current smokers.

The investigators divided participants into five groups based on phentermine exposure:

  • Per label short-term users (referent), who had just one episode of use lasting ≤ 3 months (n = 6764).

  • Short-term intermittent users, with multiple episodes of use all lasting ≤ 3 months (n = 2938).

  • Medium-term continuous users, with one episode > 3 months but no more than 12 months (n = 1703).

  • Medium-term intermittent users, with multiple episodes of use including at least one lasting > 3 months to 12 months (n = 2423).

  • Long-term continuous users, with one episode > 12 months (n = 144). 

In multivariable analysis adjusting for sex, age group, race/ethnicity, BMI category, comorbidities, and other covariates, the estimated percentage weight loss at 6 months was already significantly greater among those who would continue with greater phentermine exposure, at –2.7% for the referent on-label group, –7.8% for the medium-term continuous users, and –6.9% for the long-term continuous users.

At 12 months, the referent group had regained some of their weight (–1.4%), the medium-term groups regained less, and the long-term continuous group maintained a –7.0% weight reduction.

By 24 months, the short-term use group had regained most of the weight they had lost (–0.2%), while the other four groups all had maintained significantly greater weight loss by comparison: –0.4%, –1.9%, –3.6%, and –7.5% for the short-term intermittent, medium-term continuous, medium-term intermittent, and long-term continuous users, respectively.

And to account for the fact that those who continued to use phentermine were likely to be those in whom it worked the best from the start, the investigators conducted a sensitivity analysis looking just at those who responded to the drug initially.

Those results were similar but stronger, ranging from –6.3% for the referent to –8.5% for long-term continuous use at 6 months.

By 24 months, the referent group had regained about half the weight they had lost (–3.0%) but the long-term continuous users had lost even more, averaging –10.7%.

No Cardiovascular Risk Found at 2 Years

Overall, the composite safety outcome of incident myocardial infarction, stroke, cardiovascular intervention or death was rare, occurring in just 0.3% (n = 41) of the 13,972 participants.

Compared with the short-term use group, hazard ratios were 0.72 for both the short- and medium-term intermittent groups and 1.54 for the combined medium- and long-term continuous groups; investigators combined the two groups because there were no events in the long-term continuous group. None of the differences were significant.

Smith, who was an investigator on the recent CAMELLIA-TIMI 61 cardiovascular outcomes study of another obesity drug, lorcaserin, noted that the 0.3% event rate in this phentermine study is lower than the 1% to 1.5% or greater target for most big cardiovascular outcomes studies. "You try to have a bigger disease burden so you have more confidence," he said.

Moreover, he said that the sample size in this phentermine study was "probably a little underpowered for small increases in risk, but it's pointing in the right direction."

He agreed that there's little incentive for industry to conduct an adequately powered trial on an old generic drug, but he believes a government entity such as the Centers for Medicare and Medicaid Services or National Institutes of Health should fund such a study.

"No pharma company is going to write a check for it, so we're left with a big gap in knowledge...and we need to know because there are so many prescriptions for it...I believe our citizens deserve that answer."   

The study was funded by the Patient-Centered Outcomes Research Institute. Lewis has reported receiving unrelated research funding from Nestle.

Obesity Week 2018. November 14, 2018; Nashville, TN. Abstract OR-LB-2074.

For more diabetes and endocrinology news, follow us on Twitter and on Facebook.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.