M1, M2: Time to Bid Adieu!

Abstract

Progress in the field of transcatheter thrombectomy for acute stroke has come up against the hurdle of scarce data about occlusions in the terminal divisions of the middle cerebral artery and their branches in the lateral sulcus. This unhappy state of affairs can be traced to the practice of grouping these occlusions as lesions of the M2 segment of middle cerebral artery. An alternative way of characterizing them is proposed: it is based only on anatomic features that are likely to be of clinical relevance and represents information that is already available to the treating physician. Adopting it for presenting results of endovascular treatment of acute stroke could go at least some way toward filling the knowledge gap facing researchers in the field.

Nonstandard Abbreviations and Acronyms

ET

endovascular thrombectomy

MCA

middle cerebral artery

Clinical Perspective

What Is New?

• The conventional method for grouping intracranial arterial occlusions implicated in acute stroke is riddled with ambiguity.

• A simple and practical alternative, the arborization‐diameter‐territory system, is provided. It is more objective and incorporates information of relevance to endovascular thrombectomy in the context of acute stroke.

What Are the Clinical Implications?

• Information necessary for analyzing published data on endovascular thrombectomy will no longer be obfuscated, making meaningful comparison of the results of different studies possible.

• By allowing a more granular stratification of occlusions, much needed light would be thrown on the utility of endovascular thrombectomy in the management of acute stroke caused by relatively peripheral lesions.

The M2 segment of middle cerebral artery (MCA) has become one of the frontiers in endovascular treatment of acute stroke. ¹ Reviews of the topic almost uniformly support transcatheter extraction of a symptomatic occlusive embolus in the segment. ² , ³ , ⁴ , ⁵

This growing consensus on the therapy of these lesions is associated with an inexplicable lack of agreement about which vessels are the focus of interest. It is just not that definitions of the M2 segment are inconsistent with respect to the use of gross or angiographic anatomy, as the frame of reference, but also to transition to the M1 and M3 segments. ⁶ , ⁷ , ⁸ , ⁹ , ¹⁰ If read closely, some are incomprehensible, ¹¹ , ¹² whereas others deviate from anatomic reality. ¹³ , ¹⁴ , ¹⁵ , ¹⁶ , ¹⁷ This is perhaps not surprising when the landmarks for identifying the segment, such as the “apex of the circular sulcus,” ⁹ , ¹⁰ are not mentioned in anatomy textbooks. The uncertainty is compounded by research that questions the existence of a circular sulcus and posits instead that the insula is actually limited by 4 distinct peri‐insular sulci, or by 2 sulci and an anterior fissure. ¹⁸ , ¹⁹ Likewise one would be hard put to identify the “mesial margin” or the “top” of the Sylvian fissure, ⁷ , ⁸ , ¹¹ , ¹² given that the latter is actually a deep sulcus, the lateral sulcus of Terminologia Anatomica, that separates the frontal lobe from the temporal. ²⁰

The confusion about what the term M2 segment, and by extension M1 and M3 segments, represent begs the question: why do we not simply ignore them and instead provide data on the variables that are actually of importance in the context of endovascular thrombectomy (ET): location of embolus, diameter of occluded artery, part of the brain that was at risk. As a result of our unfortunate reluctance to do so, it is well‐nigh impossible to determine whether the impressive results of ET in M2 segment really apply to the smaller arteries coursing in the lateral sulcus or simply reflect patient samples biased toward occlusion of the relatively large and easily accessible terminal divisions of MCA. It has also led to the absurd situation where both the American Stroke Association and the European Stroke Organisation strongly recommend ET for occlusion of MCA immediately medial to limen insulae, but not if the same artery is occluded a few millimeters more peripheral in the lateral sulcus. ²¹ , ²² Fresh attempts to arrive at a consensus on the boundaries of the segments of MCA are very likely to founder, as a perusal of a recent review of the anatomy of MCA reveals. ²³ Therefore, it is high time we step away from umbrella terms of little merit such as “M2 occlusions” and instead unambiguously characterize each lesion treated in terms of features that are of actual relevance.

Arborization Level of Occlusion

The more peripheral the occlusion, the longer the time to reperfusion is likely to be and the lower the magnitude of neurologic response and lower the chance of success, and possibly higher the risk of complication. ² , ²⁴ , ²⁵ , ²⁶ Whether this is indeed the case is unknown. Reports purportedly to the contrary are difficult to take at face value, as long as an “M2 occlusion” can represent anything from an occluded MCA to occlusion of a third‐order branch at the level of the superior peri‐insular sulcus. Consequently, we can only speculate about how far one can venture into the territory of MCA, before the number needed to treat, or the number needed to harm, begins to become a cause for concern.

Diameter of Occluded Artery

If providing the diameter of the occluded artery had been the standard, we could perhaps have known by now whether there exists a threshold beyond which ET could to advantage give way to intraarterial thrombolysis. Likewise, the data could perhaps have clarified whether different thrombectomy techniques do differ in efficacy, or whether the differences reported simply reflect the confounding effect of arterial diameter on the results. ¹² , ²⁷

Territory Supplied by Occluded Artery

The part of the brain at risk for infarction plays little, if any role today in the selection of patients for thrombectomy. But it could very well be that the clinical response to the procedure is a function of the territory that is reperfused. ²⁸ , ²⁹ As a corollary, aggressive attempts at reestablishing flow could perhaps be reserved for patients who would most likely to benefit, whereas others may be better served by no endovascular intervention at all. After reviewing the anatomy of the MCA and its branches, Shapiro et al recommended that “endovascular classifications should be based on supply territory rather than non‐vascular landmarks or sites of vessel origin”. ²³ ,  

Table  .

ADT System for Characterizing Occlusions in MCA and Its First‐, Second‐, and Third‐Order Branches

Variable	Value (encoding)
Arborization level of occlusion	MCA (Z), first order (F), second order (S), third order (T)
Diameter of occluded artery	<1 mm (0)*, 1 mm (1), 2 mm (2), 3 mm (3), …
Territory supplied by occluded artery 30	Orbitofrontal (of), prefrontal (pf), precentral (pc), central (c), anterior parietal (ap), posterior parietal (pp), angular (a), temporo‐occipital (to), posterior temporal (pt), middle temporal (mt) anterior temporal (at), temporopolar (tp)

ADT indicates arborization‐diameter‐territory; and MCA, middle cerebral artery.

^* Given the limits of in‐plane spatial resolution of computed tomography angiograms and magnetic resonance angiograms, diameters less than 1 mm are encoded as 0, to reduce intra‐ and interobserver errors.

The particulars about an occlusion can be encoded as shown in Table and presented in logical sequence in a format that is easy to understand: branch order of occluded artery, diameter of occluded artery, and number of subdivisions of the territory of MCA that are supplied by the occluded artery. The identity of the subdivisions can be appended in parentheses. ³⁰ Occlusion of a second order branch of MCA beyond the limen insulae would thus be classified as S2/2(pc, c), if it, for example, was 2 mm in diameter and supplied the precentral and central regions of the cerebrum (Figure 1). The same finding in a second‐order branch, 1 mm in diameter and coursing toward the prefrontal region, would represent an S1/1(pf) occlusion (Figure 2). Under current practice, both lesions would be lumped under M2 occlusions, misleadingly implying that they are of similar significance with respect to treatment and prognosis and thereby comparable. By the addition of suitable prefixes, the use of the arborization‐diameter‐territory system can be extended to the anterior cerebral and posterior cerebral circulations. Likewise, other characteristics of an occlusion that prove to be germane in the future can be incorporated as suffixes.

Figure 1.

S2/2(pc, c) occlusion.

A, Computed tomography angiogram (maximum intensity projection) occluded second‐order branch of left middle cerebral artery, 2 mm in diameter. B and C, Perfusion computed tomography (time‐to‐peak map) upper and lower boundaries of the territories primarily supplied by the occluded artery (precentral, and central).

Figure 2.

S2/1(pf) occlusion.

A, Computed tomography angiogram (maximum intensity projection) occluded second‐order branch of right middle cerebral artery, 1 mm in diameter. B and C, Perfusion computed tomography (time‐to‐peak map) upper and lower boundaries of the territory primarily supplied by the occluded artery (prefrontal).

Characterizing occlusions as described here is of course not without its limitations. In clinical practice, mapping the territory served by an occluded artery to regions defined by Gibo et al is likely to pose a challenge in some cases. ³⁰ Critics will also fault that no allowance is made for partial involvement of adjacent regions. But as the literature on the Alberta Stroke Program Early Computed Tomography Score reveals, an acceptable level of interobserver agreement can be achieved even with subjective assessments of cross‐sectional images of the brain. ³¹ With the progressive improvement in the quality of software tools available for segmentation, fusion, and analysis of images of the brain, it is hoped that these drawbacks will be overcome in the near future. Besides, as envisaged, the more important 3 features of an occlusion are all objective in nature.

The arborization‐diameter‐territory system for characterizing intracranial vascular occlusions represents the first iteration in what, it is hoped, will become a work in progress. Transition to it, and thereafter extension of its use to the territories of anterior and posterior cerebral arteries, would go a long way toward reducing the guesswork involved in interpreting results of ET. As stratification of occlusions based on clinically important variables would become a reality, comparisons between “apples and apples” would finally become possible. As a result, making decisions about therapy would become more informed by the patient's condition and less by the personal preferences of the treating physician.

Sources of Funding

None.

Disclosures

None.