Informative Neutral Studies Matter—Why the Targeting Inflammation With Salsalate in Cardiovascular Disease (TINSAL-CVD) Trial Deserves Our Attention

Clinical research requires a fundamental state of equipoise allowing the physician responsible for care to state with assurance to a potential trial participant that the best treatment alternative is unknown, that an important but untested hypothesis has been raised in the scientific community, and that the only way to address validity of the hypothesis is to perform a randomized, placebo-controlled clinical trial in which the therapy of interest is determined by random allocation. This covenant between physician and patient is a delicate one because, if true scientific equipoise exists, roughly half of all trials will be neutral. In a survey of 104 contemporary trials funded by not-for-profit agencies between 2000 and 2005, 49% reported evidence favoring new treatments over standard of care, whereas 51% did not.¹ Yet, when a well-conducted trial is based on solid pathophysiologic principles, neutral outcomes can be highly informative.

In this issue of JAMA Cardiology, Hauser et al² present primary results from the National Heart, Lung, and Blood Institute–funded Targeting Inflammation Using Salsalate in Cardiovascular Disease (TINSAL-CVD) trial. Salsalate is a nonacetylated prodrug of salicylate that has the capacity to inhibit the nuclear factor–κB signaling pathway and is used in the treatment of inflammatory disorders, such as rheumatoid arthritis. Because inflammation is a major contributor to both atherothrombosis and diabetes mellitus, there has been considerable interest in salsalate as a potential intervention for these disorders. In fact, much of this work has been accomplished by the TINSAL investigators themselves. In 2010, the TINSAL group reported in an initial dose-escalation trial among patients with diabetes mellitus that salsalate modestly reduced glycated hemoglobin levels and led to improvement of other markers of glycemic control.³ The same group then confirmed this finding for glycated hemoglobin in 2013 in a 12-month trial of 286 patients (TINSAL-T2D), which additionally demonstrated lower levels of circulating leukocyte, neutrophil, and lymphocyte counts.⁴ In related work, the TINSAL investigators have also demonstrated that salsalate increases adiponectin levels, while reducing adipose tissue nuclear factor–κB activity in a manner not related to peripheral insulin sensitivity.⁵ On the other hand, in these same patients, salsalate had no significant effect on late advanced glycation end products⁶ or on flow-mediated endothelial dependent dilation or nitroglycerin-mediated dilation (TINSAL-FMD).⁷ These latter data are less consistent with direct anti-inflammatory effects within arteries; therefore, equipoise is clearly present with regard to potential vascular benefits of salsalate.

In this context, the TINSAL-CVD trial² sought to address whether salsalate can reduce progression of noncalcified coronary plaque assessed by multidetector computed tomographic angiography (MDCTA), a highly relevant question given interest in inflammatory mechanisms of atherosclerosis. In brief, 190 overweight to obese patients with coronary artery disease taking statin therapy were allocated to salsalate (3.5 g) daily or to matching placebo and followed up over 30 months. As in the earlier TINSAL trials (but now in a predominantly nondiabetic population), salsalate again decreased the total white blood cell, lymphocyte, monocyte, and neutrophil counts and increased adiponectin levels. However, no difference was observed between the salsalate and placebo groups with regard to change in noncalcified coronary plaque volume, the trial’s primary end point.² Of considerable interest and perhaps telling about the ability of salsalate to affect the central interleukin 1 to interleukin 6 pathway,⁸ no reduction in C-reactive protein level or fibrinogen level was observed in TINSAL-CVD² (as was also the case in the 2013 TINSAL-T2D⁴), nor were significant reductions observed for other downstream mediators of inflammation, such as myeloperoxidase or serum amyloid A.

What lessons should we take from these neutral data by Hauser et al²? First, because salsalate is an inexpensive anti-inflammatory agent, reliably understanding its effects on the vasculature remains a highly important topic, and this well-conducted, well-powered trial takes a major step in that direction. Second, and of particular interest moving forward, the TINSAL group also found no progression of noncalcified plaque as assessed by MDCTA in the placebo group. While it is possible that this effect reflects a limitation of MDCTA, it is also possible that this neutral finding represents an important reality of current cardiovascular care. All participants in TINSAL-CVD were appropriately treated with statin therapy,² and these agents have been shown not only to slow atherosclerotic progression overall but also to reverse progression in some patients.⁹ Moreover, we now understand that statins increase coronary atheroma calcification, while reducing event rates.¹⁰ Third, in TINSAL-CVD, salsalate reduced neutrophil and lymphocyte counts but did not reduce C-reactive protein level, a pattern of anti-inflammatory effect virtually opposite to that of statins.² As nicely shown, the cell count effects are unlikely to be due to bone marrow suppression because hemoglobin levels and hematocrits increased. Additional lessons from these data are that the underlying mechanisms for specific anti-inflammatory agents matter and that we cannot easily judge one by another. Fourth, it is possible that a different result might be obtained if salsalate therapy was evaluated at an early rather than a late stage of the disease process.

All of these issues have considerable relevance for ongoing and planned investigations of the inflammatory hypothesis of atherothrombosis. Therefore, TINSAL-CVD² is an informative neutral trial that deserves our attention.

Corresponding Author: Paul M Ridker, MD, MPH, Center for Cardiovascular Disease Prevention, Brigham and Women’s Hospital, Harvard Medical School, 900 Commonwealth Ave, Boston, MA 02215 (pridker@partners.org).

Published Online: May 25, 2016. doi:10.1001/jamacardio.2016.0604.

Conflict of Interest Disclosures: The author has completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Ridker reported being listed as a coinventor on patents held by Brigham and Women’s Hospital that relate to the use of inflammatory biomarkers in cardiovascular disease and diabetes mellitus that have been licensed to AstraZeneca and Seimens and reported currently serving as principal investigator and trial chairman of the National Heart, Lung, and Blood Institute–funded Cardiovascular Inflammation Reduction Trial and the Novartis-funded Canakinumab Anti-inflammatory Thrombosis Outcomes Trial. No other disclosures were reported.