Missing incremental diagnostic value analysis: does NH-IMRangio add anything to readily available clinical and cardiac magnetic resonance metrics?

We read with interest the work by Sedoud et al. (1), which examined whether angiography-derived non-hyperemic index of microcirculatory resistance (NH-IMRangio), a quantitative angiographic index, could help differentiate myocardial disease from electrical heart disease in survivors of sudden cardiac death. The authors reported a mean NH-IMRangio of 46.5±13.1 in the myocardial disease group vs. 34.1±10.8 in the electrical heart disease group, with an area under the receiver operating characteristic (ROC) curve of 0.75 and an optimal cutoff of 41.5. While we appreciate the effort to derive a novel angiographic marker in this challenging population, we wish to bring attention to a key question that the study did not address: does NH-IMRangio offer any diagnostic information beyond what is already available from simpler, routinely collected clinical and cardiac magnetic resonance (CMR) metrics?

The authors’ central claim is that NH-IMRangio “could help orient the etiological diagnosis” and “might help clinicians decide which diagnostic pathway to prioritize”. However, the study did not evaluate whether this angiographic index offers incremental diagnostic information over measures already embedded in routine post-arrest care. Left ventricular ejection fraction (LVEF) represents one such readily available metric. It can be derived immediately from echocardiography or from the same angiographic procedure without incurring additional expense. Within this cohort, LVEF measured 44.4%±13.6% in the myocardial disease group compared with 55.5%±4.19% in the electrical heart disease group (P=0.03), a distinction that separates the two groups without requiring dedicated software. The authors did not conduct a multivariate logistic regression to assess whether NH-IMRangio maintains an independent association with the diagnosis after LVEF is taken into account. In the absence of such adjustment, it remains unclear whether the observed NH-IMRangio difference represents a genuine microvascular phenomenon or merely reflects the underlying disparity in systolic function between groups.

An even more striking observation emerges from the late gadolinium enhancement (LGE) findings. CMR with LGE is widely considered a valuable tool for identifying structural myocardial involvement in patients presenting with ventricular arrhythmias or after cardiac arrest, including conditions such as takotsubo syndrome and other cardiomyopathies (2). The authors reported that 19 of 20 patients (95%) in the myocardial disease group had LGE, whereas none of the 10 patients (0%) in the electrical heart disease group showed LGE. This near-perfect separation (P<0.0001) suggests that the presence or absence of LGE alone already accomplishes the diagnostic task that NH-IMRangio was proposed to perform. When a binary test achieves such performance, the clinical necessity of a more complex, software-dependent angiographic index becomes questionable. The authors did not evaluate whether NH-IMRangio adds any meaningful information in the subgroup of patients without LGE (where diagnostic uncertainty truly exists) or whether it improves classification beyond what LGE already provides.

Beyond LVEF and LGE, the study also omitted analysis of other readily accessible clinical variables that could influence or confound the observed association. For instance, the authors noted in their discussion that patients received vasopressor support during angiography, yet they did not assess whether vasopressor dose or duration differed between groups or correlated with NH-IMRangio values. Previous investigations have shown that vasopressor agents can directly affect coronary microvascular resistance through α-adrenergic and other vasoconstrictor pathways (3), and such agents are frequently administered in post-resuscitation care (4). Similarly, the time interval between angiography and CMR ranged from 2 to 27 days, yet the authors did not explore whether this delay influenced the alignment between the index test and the reference standard. In acute presentations such as takotsubo syndrome, microvascular and tissue changes can evolve substantially over days, which may affect the apparent diagnostic performance of a test measured early but compared with a standard performed later (5,6).

We fully recognize that the authors worked within the constraints of a retrospective, single-center cohort and that their sample size (n=30) limited the number of variables that could be included in formal models. However, the question of incremental value does not necessarily require large-sample multivariate modeling. With the data already collected, the authors could take three straightforward steps to determine whether NH-IMRangio adds diagnostic value beyond conventional measures. The first step would be to fit a multivariate logistic regression model using myocardial disease as the dependent variable, incorporating LVEF (or LGE status) together with NH-IMRangio as independent variables. Even with a sample of 30 patients, such an approach could establish whether NH-IMRangio maintains an independent relationship with the diagnosis after controlling for LVEF. The second step would involve computing the net reclassification improvement (NRI) and integrated discrimination improvement (IDI) to estimate the number of patients correctly reclassified when NH-IMRangio is added to a baseline model that already includes LVEF or LGE. These measures are specifically intended for assessing incremental diagnostic value and are routinely employed in diagnostic accuracy research (7). The third step would be to present the positive and negative predictive values (PPVs and NPVs) associated with the 41.5 threshold for the entire cohort, and, if the subgroup size allows, also for those without LGE. PPV and NPV are the metrics that most directly inform clinical decision-making (8), yet they were not provided in the original report. ROC curve-derived measures alone can be misleading when evaluating the practical utility of a test (9).

Such analyses do not demand fresh data gathering, extra patient recruitment, or additional imaging. They draw exclusively from information already obtained and can be carried out with commonly available statistical software. Furthermore, these assessments are consistent with the Standards for Reporting Diagnostic Accuracy Studies (STARD) 2015 reporting recommendations, which specifically advise that diagnostic accuracy investigations examine whether a new test contributes information beyond what is already available from existing clinical measures (10). The authors stated that their study followed STARD; adding these analyses would further strengthen their adherence to this important reporting standard.

Our intent is not to challenge the practicality of obtaining NH-IMRangio measurements, nor to overlook the genuine interest in evaluating microvascular status among patients who survive sudden cardiac death. Rather, we propose that the study’s claims regarding clinical applicability would become considerably more convincing, and their interpretation considerably clearer, if the authors were able to show that NH-IMRangio conveys diagnostic insights beyond those already obtainable through simpler, well-established metrics. At a time when health systems increasingly require evidence that new diagnostic tools offer tangible value, establishing incremental benefit over existing approaches goes beyond scholarly discussion; it stands as an essential condition for sound clinical implementation. We share these thoughts with appreciation for the authors’ efforts and trust that the suggested analyses may be taken into account to better define the place of NH-IMRangio in this challenging clinical setting.

Appendix 1: Response to “Missing incremental diagnostic value analysis: does NH-IMRangio add anything to readily available clinical and cardiac magnetic resonance metrics?”

Acknowledgments

None.

Footnote

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2026-0769/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Sedoud B, Caullery B, Aouati M, Marliere S, Vautrin E, Piliero N, Ormezzano O, Bouvaist H, Lantuejoul LR, Vanzetto G, Riou L, Barone-Rochette G. Diagnostic relevance of angiography-derived coronary microcirculatory resistance in sudden cardiac death. Quant Imaging Med Surg 2026;16:234. [Crossref] [PubMed]
2. Eitel I, von Knobelsdorff-Brenkenhoff F, Bernhardt P, Carbone I, Muellerleile K, Aldrovandi A, Francone M, Desch S, Gutberlet M, Strohm O, Schuler G, Schulz-Menger J, Thiele H, Friedrich MG. Clinical characteristics and cardiovascular magnetic resonance findings in stress (takotsubo) cardiomyopathy. JAMA 2011;306:277-86. [Crossref] [PubMed]
3. Overgaard CB, Dzavík V. Inotropes and vasopressors: review of physiology and clinical use in cardiovascular disease. Circulation 2008;118:1047-56. [Crossref] [PubMed]
4. Jozwiak M, Bougouin W, Geri G, Grimaldi D, Cariou A. Post-resuscitation shock: recent advances in pathophysiology and treatment. Ann Intensive Care 2020;10:170. [Crossref] [PubMed]
5. Singh T, Khan H, Gamble DT, Scally C, Newby DE, Dawson D. Takotsubo Syndrome: Pathophysiology, Emerging Concepts, and Clinical Implications. Circulation 2022;145:1002-19. [Crossref] [PubMed]
6. Kadoya Y, Chong AY, Small GR, Chow BJW, deKemp RA, Ruddy TD, Beanlands RS, Crean AM. Global Microvascular Dysfunction and Incomplete Recovery in Takotsubo Syndrome: Insights From Serial PET Imaging. Circ Cardiovasc Imaging 2025;18:e018266. [Crossref] [PubMed]
7. Pencina MJ, D'Agostino RB Sr, Steyerberg EW. Extensions of net reclassification improvement calculations to measure usefulness of new biomarkers. Stat Med 2011;30:11-21. [Crossref] [PubMed]
8. Trevethan R. Sensitivity, Specificity, and Predictive Values: Foundations, Pliabilities, and Pitfalls in Research and Practice. Front Public Health 2017;5:307. [Crossref] [PubMed]
9. Cook NR. Use and misuse of the receiver operating characteristic curve in risk prediction. Circulation 2007;115:928-35. [Crossref] [PubMed]
10. Bossuyt PM, Reitsma JB, Bruns DE, Gatsonis CA, Glasziou PP, Irwig L, Lijmer JG, Moher D, Rennie D, de Vet HC, Kressel HY, Rifai N, Golub RM, Altman DG, Hooft L, Korevaar DA, Cohen JF. STARD 2015: An Updated List of Essential Items for Reporting Diagnostic Accuracy Studies. Clin Chem 2015;61:1446-52. [Crossref] [PubMed]