A Three-arm N-of-1 Trial With Statin, Placebo and Tablet Free Periods, to Verify Side Effects and Identify Their Cause: The SAMSON Trial
Author Block: James Philip Howard, Frances Wood, Judith Finegold, Imperial Coll London, London, United Kingdom; Alexandra Nowbar, London; David Thompson, Univ Coll London, London; Ahran D Arnold, Christopher Rajkumar, Imperial Coll London, London, United Kingdom; Stride B Christopher, Univ of Sheffield, Sheffield, United Kingdom; susan connolly, Jaimini Cegla, Peter Sedgwick Sever, Imperial Coll London, London, United Kingdom; Christine Norton, Kings Coll London, London, United Kingdom; Simon Thom, Matthew Shun-Shin, Imperial Coll London, London; Darrel Francis, Imperial Coll
Disclosure Block: J.P.Howard: None. J.Cegla: n/a. P.S.Sever: Research Grant; Significant; Amgen, Pfizer, Speaker/Speaker's Bureau; Modest; Amgen. C.Norton: None. S.Thom: None. M.Shun-shin: None. D.Francis: None. F.Wood: n/a. J.Finegold: n/a. A.Nowbar: None. D.Thompson: None. A.D.Arnold: Honoraria; Modest; Medtronic, Other; Modest; Bayer. C.Rajkumar: None. S.B.Christopher: n/a. S.Connolly: Honoraria; Modest; Astra Zeneca.
Side effects are a major reason for statin discontinuation, despite blinded trials showing no difference in symptoms between statin and placebo. We used a three-arm double-blinded n-of-1 design to investigate whether symptoms were re-induced by statin tablets and by placebo tablets.
1) Over 50% of symptom burden is nocebo rather than pharmacological
2) The majority of participants, at 6 months after completion, will either be taking statins or have declined statins for reasons other than perceived side effects.
We recruited participants who had abandoned statins due to side effects. Participants received 12 one-month medication bottles in a random sequence, 4 containing atorvastatin 20mg, 4 placebo, and 4 empty. Participants reported the intensity of their symptoms daily via a smartphone app on a scale from 0 (“no symptoms”) to 100 (“worst imaginable”). If symptoms became intolerable, they could stop tablets for that month. The primary endpoint was the nocebo ratio: the ratio of symptom intensity induced by taking a statin tablet that was also induced by taking a placebo tablet.
We recruited 60 participants between June 2016 and March 2019. 49 participants completed the full 12 month protocol.
Figure 1 shows the symptom score for every month for every patient. The right side shows patients who did not complete the full protocol.
The nocebo ratio was 0.90. The mean symptom intensity was 8.0 during no-tablet months (95% confidence interval 4.7 to 11.3), 15.4 during placebo months (12.1 to 18.7; p<0.0005 versus no-tablet months) and 16.3 during statin months (13.0 to 19.6; p<0.0005 versus no-tablet months; p=0.388 versus placebo months).
6 months after completing the trial, 30 out of 60 entrants (50%) had successfully restarted statins, 4 planned to do so, and 1 was uncontactable. The remaining 25 were off statins and not planning to restart, giving the following reasons: side effects in 18, cholesterol spontaneously improved in 4, recollection that their cholesterol had not been reduced by a statins in 1, a new diagnosis of a progressive neurodegenerative disorder in 1, and feeling “too old” in 1.
Including no-tablet periods in N-of-1 trials permits measurement of not only the effect of the active ingredient, but also the nocebo effect. In participants who have abandoned statins because of side effects, these side effects are real and predominantly due to the act of taking tablets, not what is in them. This study design allowed half of eligible patients to successfully restart a statin.