Scrutinizing science to save lives: uncovering flaws in the data linking L-type calcium channels blockers to CRAC channels and heart failure.

ABSTRACT

Hypertension is estimated to affect almost 1 billion people globally and significantly increases risk of myocardial infarction, heart failure, stroke, retinopathy and kidney disease. One major front line therapy that has been used for over 50 years involves L-type Ca 2+ channel blockers (LCCBs). One class of LCCBs is the dihydropyridine family, with amlodipine being widely prescribed regardless of gender, race, ethnicity or age. In 2020, Johnson et al. 7 reported that all LCCBs significantly increased the risk of heart failure, and attributed this effect to non-canonical activation of store-operated Ca 2+ entry. A major approach on which they based many of their arguments was to measure cytosolic Ca 2+ using the fluorescent Ca 2+ indicator dye fura-2. We recently demonstrated that amlodipine is highly fluorescent within cells and overwhelms the fura-2 signal, precluding the use of the indicator dye with amlodipine 24 . Our meta-analyses and prospective real world study showed that dihydropyridines were not associated with an increase in heart failure, likely explained by the lack of consideration by Johnson et al. 7 of well-known confounding factors such as age, race, obesity, prior anti-hypertensive treatment or diabetes 24 . Trebak and colleagues have responded to our paper with a forthright and unwavering defence of their work 27 . In this paper, we carry out a forensic dissection of Johnson et al., 7 and conduct new experiments that address directly points raised by Trebak et al. 27 . We show that there are major flaws in the design and interpretation of their key experiments, that fura-2 cannot be used with amlodipine, that there are fundamental mathematical misunderstandings and mistakes throughout their study leading to critical calculations on heart failure that are demonstrably wrong, and several of their own results are inconsistent with their interpretation. We therefore believe the study by Johnson et al. 7 is flawed at many levels and we stand by our conclusions.