The Lancet has retracted a study of Novartis’ blood pressure drug valsartan (Diovan) that has been subject to an investigation following the retraction of a related study earlier this year.
Here’s the detailed notice for “Valsartan in a Japanese population with hypertension and other cardiovascular disease (Jikei Heart Study): a randomised, open-label, blinded endpoint morbidity-mortality study”:
The Lancet published the Jikei Heart Study in April, 2007.1 On July 31, 2013, we were informed by Professor Kazuhiro Hashimoto (Jikei University) that there had been a press conference on July 30 reporting an interim conclusion from an internal investigation into this research. The report considered that “The data on blood pressure are not reliable…”. Given this finding, we now wish to retract the Jikei Heart Study on the grounds that we no longer have confidence in the published results.
When the Jikei Heart Study was first published, Staessen and Richart2 congratulated the investigators on their report. But we soon became aware that there was concern in Japan about the trial. Indeed, there was not only concern about the Jikei Heart Study, but also the Kyoto Heart Study, which was published in the European Heart Journal in 2009.3 Both reports were about valsartan and had several authors in common. In October, 2011, we received correspondence expressing concern about the similarity of the standard deviations of the mean blood pressure measurements in the two studies, and we invited the authors to comment. Although we were initially satisfied with their response, in 2012 we published a letter from Dr Yui reporting further concerns.4
The retraction of the Kyoto Heart Study5 in February, 2013 led to an investigation into the conduct of the Jikei Heart Study. An investigating committee headed by Professor Hashimoto from Jikei University was established. We became aware of this development on April 29, 2013, and on May 2 we wrote to Jikei University asking for details of the investigation and requesting that we be kept informed. We wrote again on June 4 and June 19 asking when the investigation might be completed. We wrote again on July 31 after we were made aware that a press conference had been held. By return we received a synopsis of the investigation’s findings. The report identifies concern over the reliability of the blood pressure data: “We believe, therefore, that the data were intentionally altered.” The report goes on to say: “We suspect that the data were altered during their statistical analysis.”
The report also comments on the affiliation of the person entrusted with the statistical analysis, Nobuo Shirahashi. Mr Shirahashi’s affiliation, as provided to The Lancet, was: “Clinical epidemiology, Osaka City University Graduate School”. In April of this year, The Lancet received a letter from a professor at Osaka City University Graduate School of Medicine informing us that there has never been a Department of Clinical Epidemiology in the university, that Osaka City University “has never had the statistical analysis group for Jikei Heart Study”, and that Mr Shirahashi “has never been staff of our university” but was employed by Novartis Pharma Japan.
Our attempts to obtain a response from Mr Shirahashi were hindered by his recent retirement from Novartis, but the Global Head of Medical Affairs, Dr Usman Azam, did share with The Lancet that Mr Shirahashi had been employed by Novartis throughout the period of the trial and its publication. Dr Azam also told us: “In addition, it appears Mr Shirahashi obtained an unpaid position in about 2001 with the Osaka City University Graduate School’s Medical Research Faculty as a part-time lecturer in the Department of Medicine. He held the position until 2011.” Dr Azam further acknowledges “…that Mr Shirahashi should have requested the studies’ authors to identify him as a Novartis Pharma K.K. employee”.
Taken together, these findings indicate that there is now sufficient doubt as to the integrity of the Jikei Heart Study and the obfuscation over affiliation of the study statistician for The Lancet formally to retract the paper from the scientific record.
1 Mochizuki S, Dahlof B, Shimizu M, et al. Valsartan in a Japanese population with hypertension and other cardiovascular disease (Jikei Heart Study): a randomised, open-label, blinded endpoint morbidity-mortality study. Lancet 2007; 369: 1431-1439.
2 Staessen JA, Richart T. Sum and substance in the Jikei Heart Study. Lancet 2007; 369: 1407-1408.
3 Sawada T, Yamada H, Dahlof B, Matsubara H. Effects of valsartan on morbidity and mortality in uncontrolled hypertensive patients with high cardiovascular risks: Kyoto Heart Study. Eur Heart J 2009; 30: 2461-2469.
4 Yui Y. Concerns about the Jikei Heart Study. Lancet 2012; 379: e48.
5 Retraction of: Effects of valsartan on morbidity and mortality in uncontrolled hypertensive patients with high cardiovascular risks: KYOTO HEART Study [Eur Heart J (2009) 30: 2461—69, DOI:10.1093/eurheartj/ehp363]. Eur Heart J 2013; 34: 1023.
The study has been cited 178 times, according to Thomson Scientific’s Web of Knowledge.
Larry Husten, at Cardiobrief, has been following the story in detail, and notes that the saga began with questions about papers by Hiroaki Matsubara, who now has eight retractions. Matsubara is not an author on the Jikei study.