Stopping statins in serious illness doesn’t change survival…or does it? Sorting out the message

by: Alex Smith, @AlexSmithMD

We’ve talked previously on GeriPal about Dumb Medicine: continuing preventative treatments near the end of life.

• In one study of patients diagnosed with advanced incurable cancer and an average life expectancy of less than 2 years, up to 15% were being screened for another cancer that had no chance of harming them in their lifetime.
• In another studyof women with dementia and an average life expectancy of less than 3 years, 18% were being screened with mammograms for a breast cancer that if detected, would not harm them in their lifetime.

Key features of both of these studies:

• Preventive treatments have a lag time to benefit.  Cancer screening is designed to detect slow growing cancers that will not cause symptoms or harms for about 10 years.
• The harms of testing are real.  False positives are no joke.  Detecting a clinically insignificant cancer can lead to tests and treatments (think biopsy, surgery, chemotherapy, radiation) for a cancer that is unlikely to harm the patient in their lifetime.
• The focus on prevention displaces time that could be spent on supportive and palliative interventions.

This week, we have a new study by the fabulous palliative care researcher Jean Kutner and colleagues to add to this list.  I think the new study pushes the envelope of Dumb Medicine.  It’s self-evident that we should not continue cancer screening for patients with life expectancies of 2-3 years.  The new study examines statin usefor patients with a life expectancy of less than a year.

Now statins are a slightly different beast from cancer screening.  They are preventative, certainly.  They have a lag time to benefit on the order of years.  They have harms, ranging from pill burden to cost, and risks, ranging from myopathies to rhabdomyolosis.

But in the back of our heads, maybe we’re wondering, might this medication be preventing a heart attack, or a stroke?  Maybe it’s not so dumb to continue it after all.

Now we have new evidence that demonstrates that stopping statins (probably) doesn’t change survival and likely improves quality of life.

Some of the aspects of the study are really important, yet because they’re buried or in “researchees, i.e. the language of researchers” they are inscrutable to the average reader.

• 381 subjects with a life expectancy between a month and a year, about 1/3 in hospice, were enrolled from 15 sites in the Palliative Care Research Co-operative (PCRC).  The PCRC alone is BIG news.  You know how cancer trials are conducted in these research networks?  Now we have one for palliative care research! (disclosure – I’m writing this on a plane to the PCRC meeting)
• 22% were unwilling to participate.  That’s not a sky high number for enrollment in a study near the end of life.  However, it does suggest that some patients may be unwilling to even consider stopping statins.
• Participants were randomized to continue statins or not.
• This was a pragmatic trial.   This means that instead of trying to figure out why exactly statins might or might not impact survival under optimal research conditions, the authors were interested in
• The study was not blinded.  People knew if they were still getting a statin or not. They didn’t get sham pills that were possibly a statin or a placebo.  Their doctors knew they were either still getting a statin or not.  This is important, because it’s possible that stopping the statin may have clued the doctor in to stop other medications. Indeed the group that stopped statins also stopped on average almost one additional medication.
• Quality of life was modestly higher in the group that stopped statins over time.
• Mean cost savings given the average online drug price were $716 per patient.
• On the one hand, there was no difference in the primary endpoint of death at 60 days following randomization.  24% of those who stopped statins died within 60 days, and 20% of those who continued statins died within 60 days.  The p-value for this difference was 0.36, and did not reach the threshold our society accepts as reasonable for saying a difference exists, or 0.05.  Theoretically, if we repeated the study 100 times, 36 times we would find no difference between stopping and continuing, and 64 times we would find a difference.  If the p had been .04, we would have said 4 only times we would find no difference and 96 times we would find a difference.
• On the other hand, the study did not reach it’s non-inferiority endpoint.  What does this double negative mean in plain english?  If we untwist the double negative, does it mean that stopping statins is superior?  Not really.  Not non-inferior doesn’t mean superior.  Here’s what it means.  It means that the range of possible outcomes given the sample size of the study included the possibility that stopping statins could result in as much as a 10% increase in mortality at 60 days compared to continuing.  At the other end of the range, stopping might decrease survival by 3%, but it’s the possibility that stopping might increase mortality that is the focus of the “non inferiority” analysis.

What are we to make of these last two points?  Is this a “mixed message”?  Possibly.  The authors conclude that for those patients whose goals are quality of life, the physician could counsel the patient to stop statins.  For those patients focused on maximizing chance of survival, we can’t really say there isn’t a slim chance they might prolong life.

So now I’m going to take a stand that is a more definitive stance than the authors could take in an academic journal:

Stop. The. Statins.

Continuing statins for patients with a life expectancy of less than 1 year is nonsense.  Just stop it.  It’s dumb medicine.  Continuing has no demonstrable impact on survival, and stopping leads to a higher quality of life at a reduced cost to society.

The biggest barrier to stopping isn’t that niggling voice in the physician’s head saying, “maybe there’s a chance the statin is helping.”  The biggest barrier is the lack of support, lack of training, and fear of talking with the patient about the reality of their limited prognosis, and the implications of that prognosis for treatment.