Upholding Standards: Should the FDA approve a new lupus drug with unclear benefits?

If you are one of the estimated 160,000 300,000 to 4 million patients in the US who suffer from lupus, you may feel excited about a drug called Benlysta (generically, belimumab).  There is a need for better drugs to treat symptoms, reduce the use of other medications that can have imperiling side effects, and reverse disease progression.  The US Food and Drug Administration (FDA) plans to decide whether to approve belimumab by December 9th.

For unknown reasons, in lupus the immune system attacks the body, leading to a chronic inflammatory disease.  The illness can vary greatly for each patient and between patients.  Flares can appear unexpectedly and with an unpredictable severity, several times a year.  Symptoms often are “general” (fever, weight loss and fatigue) and organ-specific, including joint pain, skin lesions, kidney injury, and heart injury among others.  In severe cases, a patient may die from kidney failure, severe infection, stroke or other causes.

A lupus diagnosis is no longer a death sentence.  Sixty years ago, the 5-years survival rate was 40%, whereas since the 1980s it has been over 90%.  This may have occurred as physicians learned to identify and treat the disease sooner or use medications “off label” to help patients.  While the only FDA-approved medications for lupus are aspirin, prednisone and the anti-malarial drug hydroxychloroquine, some patients also find benefit from immune suppression drugs like cyclophosphamide and other glucocorticoids related to prednisone.  Yet taking these medicines brings many side effects, often severe.  They can include increased risk of severe infection, body weight increase and redistribution, rashes, eye cataracts, and increased risk of major cardiovascular illness including heart attack and stroke.

Enter the hope for belimumab.  Its makers, Human Genome Sciences and GlaxoSmithKline, conducted two Phase 3 trials that each enrolled 800 to 900 patients, all of whom received standard of care medicines like prednisone.  In addition, the patients either received placebo or belimumab.  Trial researchers tracked patients’ symptoms to determine whether they responded to the drug.

Did the drug reduce lupus symptoms?  In one trial with patients who were from Asia and Latin America, 58% of patients receiving the target dose of belimumab responded compared to 44% in the placebo arm.  This difference reached statistical significance, meaning it was highly unlikely for these groups to have differed solely by chance.  But in the second trial, with patients in the US, Canada and Europe, the improvement was smaller at 12 months (43% response compared to 34% with placebo) and disappeared by 18 months.  In addition, patients of African-American or African heritage worsened with belimumab.  In the first trial, 46% of these patients responded compared to 64% on placebo.  In the second trial, it was 33% for belimumab versus 39% for placebo.

Was the drug safe?  Of 2,133 patients treated, death was 1.8 times more likely for patients on belimumab (0.9% of patients) than on placebo (0.4% of patients), and this was a statistically significant result.  71% of patients treated with the drug got an infection (of any severity) compared to 67% of those receiving placebo.  Serious infections were slightly more frequent with belimumab than placebo (6% of all patients in the Phase 2 and 3 trials, versus 5% for placebo).

Considering these findings, will the FDA approve the drug?  On the one hand, some industry analysts have written, approval would encourage other lupus drug makers.  Yet lowering the bar would only benefit short-term industry interests at the expense of patients.  Difficult problems require innovation and perseverance.  Perhaps suggesting it would not lower its standards, in a briefing to the FDA’s recent advisory panel, the agency raised several major criticisms of the belimumab results.

Yet surprisingly, the panel voted 13 to 2 in support of belimumab.  Meeting minutes have not yet been released.  Reuters quoted one panel member as saying, “The efficacy is mild, but I think there is a need for drug even with mild efficacy.”   The FDA is not required to follow the panel’s recommendation, though several journalists have indicated it usually does.  One additional variable is that the drug may cost $20,000 to 30,000 per year.  While this is expensive, it is not nearly as expensive as some drugs for cancer or other rare diseases (see list here), though it would rack up costs as patients took it chronically.

One compromise is the FDA could approve the drug with restrictions.  For instance, the drug might be permitted only for patients the drug helped, such as those of Caucasian heritage or with skin or musculoskeletal forms of the illness.  Or it may be restricted to 12 months or less of continuous therapy.  In a few weeks we will learn of the decision.  Hopefully it will be made in the interest of patients first.

Further reading:

  1. Mayo Clinic overview of lupus, medically known as systemic lupus erythematosus (SLE): http://www.mayoclinic.com/health/lupus/DS00115
  2. Mayo Clinic overview on the pros and cons of glucocorticoid drugs: http://www.mayoclinic.com/health/steroids/HQ01431
  3. UpToDate Patient information on lupus: http://www.uptodate.com/patients/content/topic.do?topicKey=~44azQ566S.O5Xa&selectedTitle=3~150&source=search_result
  4. Article about epidemiology of lupus (for instance, how common the disease is by country): http://lup.sagepub.com/content/15/5/308
  5. Wikipedia article about lupus and with prevalence estimates: http://en.wikipedia.org/wiki/Lupus#Epidemiology
  6. Articles on FDA panel backing Benlysta:
    1. FierceBiotech: http://www.fiercebiotech.com/story/breaking-news-hgs-wins-crucial-panel-backing-benlysta/2010-11-16
    2. LA Times: http://articles.latimes.com/2010/nov/17/news/la-heb-lupus-drug-20101117
    3. New York Times: http://www.nytimes.com/2010/11/17/health/17drug.html
  7. FDA Arthritis Advisory Commitee 2010 meeting materials: http://www.fda.gov/AdvisoryCommittees/CommitteesMeetingMaterials/Drugs/ArthritisDrugsAdvisoryCommittee/ucm203434.htm
  8. Overview of FDA and the clinical trial process for drugs seeking approval in the USA:  http://www.fda.gov/drugs/resourcesforyou/consumers/ucm143534.htm
  9. How can a drug have a perceived benefit compared to placebo, but not actually have one due to randomness?  Read an overview of statistical significance here:  http://en.wikipedia.org/wiki/Statistical_significance
  10. Interesting summary of an investigation into the high percentage of patients enrolled outside the US for drugs garnering approval in the US: http://www.nytimes.com/2010/06/22/health/research/22trial.html
  11. Note the clinical trials discussed are BLISS-52 (HGS press release) and BLISS-76 (press release) as the “first” and “second” trials, respectively.


  1. As a lupus patient who has experienced major organ involvement (heart, lungs, bone marrow, kidneys, and brain) and a person who works daily with lupus patients at the Lupus Foundation of Florida, I hope that Benlysta WILL be approved. Cytoxan is a particularly nasty drug. Cytoxan, at higher doses, is a breast cancer chemotherapy with wretched side effects which frequenlty include sterility, a major concern since 9 out of 10 lupus patients are women.

    I am fortunate that Cellcept (mycophenolate mofetil) in conjunction with prednisone and hydroxychloroquine AND a proactive life style have brought my lupus to remission. Cellcept is ‘off lable.’

    We desperately need new treatments. Black box warnings are no stranger to lupus patients. We know the risks and make informed choices. Yes, Benlysta is pricey. So are other treatments. The average lupus patient already spends $6,000-$10,000 for treatment and that is for mild to moderate lupus.

    The Lupus Foundation of America estimates the number of patients with lupus at 1.5 million, which is 10 times the number you cited.

    We lupus patients are people. We struggle with an intractable and often debilitating disease. I was on disability for four years, but am now working more than full time. I am one of the fortunate ones. I am 58 years old. I also know a number of lupus patients who were part of this study and experienced significant improvement.

    ALL of the medications that we are routinely given have horrid side effects. The only three drugs that are FDA approved specifically for lupus are aspirin, prednisone and hydroxychloroquine which was approved in 1958! The question is whether the risk vs. benefit ration of belimumab is better than that of other treatments (which happen to be off label) like cytoxan, rituxan, azathioprine, methotrexate, etc.

    I can understand your desire to present ojective clinical evidence, but we are not creatures in a petri dish. We are real people who suffer greatly. Fortunately, my medical team treats me with great human compassion.


  2. […] Upholding Standards: Should the FDA approve a new lupus drug with unclear benefits? (mckee.wordpress.com) […]


  3. […] by Andrew One day after my post on belimumab, the FDA announced it would be taking 3 more months to announce its decision on whether to approve […]


  4. Dear Linda,

    Thank you for your thoughtful reply. It is great to hear that you have been persistent in trying to find a way of life that helps you cope with the disease and I felt inspired by your progress. I also apologize if the tone of my essay was too clinical. My intention was that the article could be educational about the potential of the new drug, and that I could convey the topic in a clear, compassionate way without relying on clinical and scientific jargon. Do you have any suggestions on what specifically I could have done differently in the article? I had thought of interviewing one of my family members who suffers from lupus. But then as a father of an infant, I decided to prioritize my immediate family over writing a more in-depth article, especially since I’m doing this in my free time. I realize that could have made for a more compassionate and warming tone to the article. Any other feedback you have would help me improve my writing.

    Your response also raises a few issues that are difficult in medicine. I empathize with the desire for new therapies. My mother died of metastatic cancer over ten years ago and I remember the pain, the sense of helplessness, when we wished one of her therapies would work, yet none did. The doctors tried three different chemo agents and ultimately there was nothing else we could do. I know of some cancer clinics that help patients take experimental therapies, before they have been approved by the FDA, because the patients may only have months to live.

    Also the trials with Benlysta did not actually demonstrate that the drug has less side effects than current therapies, as all patients were ethically required to take the best drugs available to them, in addition to placebo or Benlysta. In other words, patients in all arms of the trials were getting the drugs with nasty side effects such as glucocorticoids and immune suppressants. However, I do agree with your implied point: if Benlysta were to be approved, doctors may try to prescribe it before prescribing more toxic drugs like the immune suppressants, even though there may not be evidence that Benlysta would be any better. Also it is worth noting that Benlysta suppresses B cells in the immune system, so there is no guarantee (nor any data that I’m aware of) that Benlysta would be better than immune suppressants like Cytoxan. It could be worse.

    Another question is, at least under the current system with the FDA we have in the US, what if a drug does not work better on average than placebo? Still, some patients may do really well on a new drug. Is it fair that they may never get the medicine? And sometimes trials that show a drug works are not repeatable, as my new post summarizes. With Benlysta, as discussed, the drug doesn’t even work that well, in fact it hardly works better than placebo.

    That doesn’t mean I’m opposed to patients trying the drug. With any drug, patients will experience a range of responses. Some people may really benefit. Is it fair to not let those patients try the drug? The challenge can be that we don’t know beforehand who will benefit. So, if a drug is on average no better than placebo, then you’re rolling the dice with a high chance of nothing different happening, and a low chance of injury or benefit. Some patients will fare worse than if they hadn’t taken the drug, others might not notice any improvement, and some may do better. If Benlysta had some test that helped you predict whether you would feel better on the drug, or whether awful side effects would occur, that could help patients know whether it might help them.

    It suggests a broader question. Should we change how the drug regulatory system works? Should we lower the bar, so that more drugs are released to the population, letting patients harness more of the risk (risk of positive events, as well as negative events), and then over time, figure out which drugs really work? If so, how would we change the system? I don’t have any answers to this, just thinking out loud.

    Finally, thank you for your point about how common lupus is in the US. You’re right it looks like my figure, pulled from wikipedia, is a bit low for the US. It also depends on how you measure lupus. The CDC below summarizes that there could be between 300,000 and 4 million Americans with lupus, depending on how it is categorized. I have also corrected the post.


    In any event, I wish you well. I hope that as I continue to work on my journalistic writing skills that readers like you may find my thoughts more helpful in the future.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: