Beyond Classification

Corresponding Author

In this issue of JACC: Asia, Ling et al¹ report a meticulously designed multicenter retrospective study comparing 8 imaging-based classification systems for predicting conservative management failure in isolated superior mesenteric artery dissection (ISMAD). Drawing from a cohort of 156 patients from 3 participating centers, their analysis represents the most comprehensive head-to-head evaluation of existing morphological frameworks to date. Their key finding—that Tan's classification achieved the highest discriminative ability (C-statistic: 0.750) and net clinical utility—deserves particular attention. By incorporating long-segment dissection exceeding 5 cm (subtype L) and severe true-lumen stenosis surpassing 80% (subtype S) into the existing framework, Tan's system succeeded where its predecessors had only partially delivered: it identified, with quantifiable precision, the subset of patients most likely to fail conservative management and thus require early endovascular intervention.^1,2

Yet, as we welcome this refinement, we should also pause to ask a more fundamental question: Are we approaching the ceiling of what anatomical imaging classification alone can tell us about a disease whose defining feature is not the vessel itself, but the viability of the bowel it supplies?

The current treatment paradigm for ISMAD rests upon an assumption that has remained the silent cornerstone of every major classification system—Sakamoto et al,³ Yun et al,⁴ Li et al,⁵ and now Tan et al²—without ever having been rigorously validated: that the degree of true-lumen stenosis seen on CTA reliably predicts intestinal ischemia severity. Sakamoto introduced false-lumen morphology; Yun added hemodynamic stratification; Li incorporated true-lumen stenosis severity; and Tan has now layered in dissection length as an independent predictor. Yet, all 4 systems share a common limitation: They classify the vessel, not the organ it supplies. This assumption seems intuitively sound. But intuition, in medicine, is a treacherous guide.

Accumulating evidence exposes this fragility. Kim et al⁶ (2014) first demonstrated that clinical symptoms in symptomatic ISMAD resolved rapidly despite progressive angiographic stenosis on follow-up imaging—a finding that quietly undermined the assumption that anatomical severity predicts clinical course. Yu et al⁷ (2022) confirmed that symptom profiles overlapped across all Yun subtypes, including type III (complete SMA occlusion). Even more instructive is Anzai et al⁸ (2020): A patient with CTA-demonstrated proximal SMA “complete occlusion” and elevated lactate who showed no radiographic signs of intestinal ischemia whatsoever—no bowel wall thickening, no mesenteric edema, no pneumatosis. The dissection had compressed the true lumen to near-invisibility, yet the bowel, perfused through collaterals from the celiac trunk and inferior mesenteric artery, remained fully viable. Anatomical catastrophe; functional normality.

This anatomical-functional disconnect highlights a physiological principle that ISMAD classifications have largely overlooked: the extraordinary resilience of the intestinal collateral circulation. Unlike the heart, which depends almost entirely on the coronary arteries, the intestine is protected by a rich collateral network connecting the celiac trunk, superior mesenteric artery, and inferior mesenteric artery through the gastroduodenal, pancreaticoduodenal, and marginal arteries of Drummond. Even severe SMA stenosis may be hemodynamically tolerated if collateral recruitment is adequate. A classification system that assesses only the primary vessel and ignores collateral status is, by design, incomplete.

The implications for treatment are considerable. Under the Li classification, type IIb lesions are categorized as candidates for intervention.⁵ Yet Heo et al⁹ (2017) showed that among 83 symptomatic patients with ISMAD managed conservatively, 96% achieved pain resolution without intervention—regardless of radiographic features that would classify many as Li type IIb or III. Similarly, Zhu et al¹⁰ (2018) reported that initial conservative treatment was used in 87.0% of symptomatic cases, with an overall success rate of 84.7% and a crossover-to-intervention rate of only 14.3%. These figures strongly suggest that “intervention-indicated” classifications may lead to unnecessary stent placement in patients whose bowel was never truly ischemic.

This is precisely where cardiology's experience with fractional flow reserve (FFR) offers a compelling parallel. Two decades ago, coronary revascularization was guided by angiographic stenosis severity—much as ISMAD management is guided by CTA today. The landmark FAME trial demonstrated that physiology-guided PCI, treating lesions that caused ischemia (FFR ≤0.80) rather than those that merely looked severe, reduced the composite endpoint of death, myocardial infarction, and repeat revascularization by 28%.¹¹ FAME 2 further established that deferring PCI in lesions with FFR >0.80 was safe and reduced unnecessary procedures.¹² The field of ISMAD is, in many respects, where coronary intervention stood before FFR: relying on anatomical images to make functional decisions, and paying the price in overtreatment.

Recently, Peng et al¹³ (2024) reported the first application of mesenteric FFR in an ISMAD patient—a 56-year-old man with severe true-lumen compression on CTA but a measured FFR of 0.72. Rather than proceeding to stent placement based on morphological criteria (which would have classified the patient as requiring intervention under both Li IIb and Tan S subtypes), the clinical team elected conservative management with bowel rest and anticoagulation. Symptoms resolved within 14 days, and he remained recurrence-free at 5-year follow-up. This single case offers preliminary, hypothesis-generating proof-of-concept that functional assessment may safely reclassify patients who would otherwise be triaged to intervention.

But we must be candid about where this technology stands. Mesenteric FFR is at the very beginning of its conceptual life. The optimal ischemia threshold for the SMA territory is entirely unknown; whether the coronary cutoff of 0.80 applies to visceral arteries is speculative at best. Whether papaverine or adenosine can reliably induce maximal mesenteric hyperemia remains unanswered, and the postprandial physiology of splanchnic circulation may introduce confounders with no coronary equivalent.¹⁴ The entire published experience rests on a single case report. What mesenteric FFR offers now is not a tool, but a direction.

To translate these concepts into practice, a multidimensional approach integrating morphology, perfusion, and biomarkers will ultimately be needed. Alongside invasive physiological assessment, noninvasive functional imaging may provide a more immediately accessible intermediate step. Dual-energy CT (DECT) iodine maps have emerged as a promising technique for quantifying bowel wall perfusion directly. D'Angelo et al¹⁵ (2024) demonstrated in 142 patients with nonocclusive mesenteric ischemia that late arterial phase iodine concentration measurements discriminated ischemic from nonischemic bowel segments, with an AUROC of 0.966, markedly outperforming conventional CT attenuation values (AUROC: 0.828). For ISMAD, integrating DECT perfusion assessment could answer the critical question that no morphological classification can: Does the dissected SMA still deliver adequate oxygen to the intestinal wall?

Complementary to perfusion imaging, intestinal ischemia-specific biomarkers offer a molecular dimension. Intestinal fatty acid-binding protein (I-FABP), a 15-kDa cytosolic protein specific to small intestinal enterocytes, demonstrated pooled diagnostic accuracy for acute intestinal ischemia in a meta-analysis, with serum I-FABP showing a sensitivity of 0.74 and specificity of 0.91, and urinary I-FABP showing a sensitivity of 0.88 and specificity of 0.79.¹⁶ This performance stands in sharp contrast to conventional markers such as D-dimer (sensitivity 0.94 but specificity only 0.50) and serum lactate (which lacks intestinal specificity and shows wide performance variability across studies), reflecting systemic responses rather than intestinal cell injury per se.^17,18 Unlike these nonspecific indicators, I-FABP is released directly from damaged enterocytes, offering a more precise signal of intestinal mucosal integrity. While requiring assay standardization and ISMAD-specific validation, molecular markers of tissue damage may ultimately prove more informative than anatomical images of the supplying vessel.

Ling et al¹ should be commended for bringing sorely needed methodological rigor to a field long suffering from fragmented classification systems. Their work provides evidence-based rationale for adopting Tan's classification. But as the field matures, we should move beyond exclusive reliance on anatomical staging. The future of ISMAD management likely lies in a multidimensional risk assessment integrating 3 complementary domains: 1) refined morphological classification (Tan's system), 2) functional confirmation of intestinal perfusion (mesenteric FFR or DECT iodine mapping), and 3) biomarker evidence of enterocyte injury (I-FABP) (Figure 1). Only when these 3 domains converge—when the vessel, the perfusion, and the cell are simultaneously assessed—can we confidently distinguish patients who genuinely need urgent intervention from those who can be safely observed.

Figure 1.

Multidimensional Risk Stratification Framework for Isolated Superior Mesenteric Artery Dissection

The proposed paradigm integrates 3 complementary domains to guide clinical decision-making: 1) morphological classification using Tan's system (type L: dissection length >5 cm; type S: true-lumen stenosis >80%; type N: neither), which is vessel-centric and identifies structural risk; 2) intestinal perfusion assessment, including invasive mesenteric fractional flow reserve (FFR) and noninvasive dual-energy computed tomography (DECT) iodine mapping, which is organ-centric and confirms adequate oxygen delivery; and 3) enterocyte injury biomarkers, notably intestinal fatty acid-binding protein (I-FABP), which is cell-centric and detects mucosal integrity loss. These 3 dimensions converge at a central multidimensional integration node. Conservative management is appropriate when intestinal perfusion is adequate and biomarkers are negative, regardless of morphological severity. Endovascular intervention is favored when severe morphology is accompanied by compromised perfusion and elevated biomarkers, indicating enterocyte injury. Open surgery with bowel resection is reserved as a rescue pathway for patients with established bowel necrosis or peritonitis, representing a complication-driven rather than risk-stratified decision. The timeline at the bottom illustrates the conceptual evolution from the anatomy-only era (Sakamoto, Yun, Li, and Tan classifications), through the functional assessment era (FFR and DECT), toward the current integrated multidimensional era, shifting the paradigm from “classifying the vessel” to “assessing the organ.” AUROC = area under the receiver operating characteristic curve; Se = sensitivity; SMA = superior mesenteric artery; Sp = specificity.

As we refine our ability to classify the dissected artery, let us not forget the organ at stake. In ISMAD, the vessel is merely the conduit; the bowel is the patient. The next decade of research should focus less on how to categorize the dissection and more on how to determine whether the dissection is causing harm. From the perspective of the at-risk intestine, anatomy is only the beginning of the story.

Funding Support and Author Disclosures

This work was supported by the Zhejiang Provincial Natural Science Foundation (Grant No. KLY26H020025). Dr Liu reports collaboration with Dr Peng, the first author of the mesenteric FFR case report cited in this manuscript (reference 13). Dr He has reported that he has no relationships relevant to the contents of this paper to disclose.