Key Points
- The HARP trial, the largest prospective multi-modality imaging study in MINOCA to date, identified an underlying cause in 79% of patients using combined optical coherence tomography (OCT) and cardiac magnetic resonance (CMR), compared with 44% with OCT alone and 69% with CMR alone.
- Among 284 patients with both studies performed, 59% had confirmed myocardial infarction, 20% had a MINOCA-mimicking condition, and 21% remained idiopathic.
- No significant differences in ischemic or non-ischemic CMR findings were detected between women and men, suggesting that the well-established female predominance of MINOCA reflects differences in coronary anatomy rather than underlying pathophysiology.
Myocardial infarction with no obstructive coronary arteries accounts for approximately 6% of all MI cases and up to 15% of women presenting with NSTEMI; a threefold higher prevalence than in men. Current guidelines recommend systematic imaging to identify the underlying etiology and direct therapy, which varies substantially depending on whether the cause is atherothrombotic, a dissection or spasm, or a mimicking condition such as myocarditis or Takotsubo syndrome. Whether mechanisms differ by sex has remained poorly defined. The HARP trial NCT02905357 (Heart Attack Research Program), presented as Featured research at ACC.26 was designed to address these gaps using blinded, core laboratory adjudicated OCT and CMR in the largest prospective MINOCA cohort to date.
HARP is an AHA Go Red for Women-funded prospective multicenter observational study enrolling MI patients without known obstructive CAD and without an alternate explanation for troponin elevation. The study initially enrolled women only before expanding to include men to examine sex differences. Eligible patients with confirmed MINOCA (≤50% stenosis in all major vessels) underwent multi-vessel OCT at diagnostic catheterization and CMR within one week. Core laboratories at the Cardiovascular Research Foundation (OCT) and Brigham and Women’s Hospital (CMR) interpreted all studies blinded to clinical data, sex, and results of the other modality. The final cohort included 336 patients enrolled across 28 sites in the US, Canada, and the United Kingdom; 284 (85%) completed both studies.
The cohort had a median age of 58 years, 48% hypertension, 18% diabetes, and a median peak troponin of 32 times the upper reference limit. Only 5% presented as STEMI, and 46% had a normal coronary angiogram by site read. Multi-vessel OCT was performed in 93%, with three-vessel imaging in 64%, a median of two days post-MI.
On OCT, a culprit lesion was identified in 45% of patients with no significant difference by sex (44% women vs. 53% men, p=0.18). The most common findings were intraplaque hemorrhage (17%), healed plaque (12%), plaque rupture (7%), and plaque erosion (5%). Multivariable predictors of an OCT culprit lesion included older age, angiographic abnormality, and three-vessel OCT imaging.
CMR was abnormal in 63%: ischemic findings in 40% (infarction on LGE in 28%, regional injury in 12%) and non-ischemic findings in 23%, including myocarditis (10%), non-ischemic cardiomyopathy (6%), and Takotsubo syndrome (3%). Importantly, 44% of patients with ischemic CMR findings had no identifiable OCT culprit, implicating coronary spasm or thromboembolism in a substantial proportion. Even among patients in the lowest troponin decile, less than four times the upper reference limit, 38% had an abnormal CMR, underscoring that troponin level alone cannot reliably identify who warrants advanced imaging.
Synthesizing both modalities, a causal diagnosis was reached in 79% of patients. Sex-stratified analyses found no significant differences in ischemic or non-ischemic CMR findings between women and men, challenging the assumption that female predominance of MINOCA reflects distinct pathophysiology.
According to lead investigator Harmony Reynolds, MD, of NYU Grossman School of Medicine, the clinical implication is straightforward: “If you have OCT available in the cath lab, go ahead and do that at the same sitting. And if you have cardiac MRI easily available, get it as soon as possible.” She noted the importance of identifying mimicking conditions, which “completely changes the therapy; from multiple drugs after an MI to you didn’t have a heart attack and you might not need anything.”
These findings provide the strongest prospective evidence to date that combined OCT and CMR achieves a causal diagnosis in the large majority of MINOCA presentations, that neither modality alone is sufficient, and that the underlying mechanisms do not differ meaningfully by sex, reinforcing guideline recommendations for systematic imaging-guided evaluation after a MINOCA presentation.