Ravinder-Jeet Singh, MBBS, DM
Infarct growth among patients with large vessel occlusion (LVO) is highly variable. In some patients, infarct progresses very quickly (rapid progressor) and they have no or small penumbra even during early hours after their stroke onset, while others progress more slowly (slow progressor) and have large penumbral tissue at later time windows. Therefore, size of pre-treatment penumbra and response to reperfusion therapies, especially endovascular thrombectomy, would vary depending on time from symptom onset and rate of infarct growth, resulting in patient-specific time-windows to intervene. While rapid progressors could benefit from reperfusion therapy during very early time-window, the slow progressors can potentially benefit from treatment in either early- or late-windows This concept has been tested in the recent early- and late-window thrombolysis and thrombectomy trials. Therefore, early distinction between rapid vs slow progressor might prove particularly useful in making time-sensitive decisions, especially interfacility transfer decisions, typically between more peripheral primary stroke centers to larger endovascular therapy capable centers.
The variability in infarct growth is determined by multiple demographic, clinical, and imaging factors, such as age, blood pressure, blood glucose, stroke severity, initial infarct size, and time from ictus; these factors can influence “final” infarct volume and determine functional outcomes. Collateral blood flow status plays an especially major role in providing residual flow, and infarct size. Whether these same factors also underlie “early” infarct growth is less well studied. In the present study, the authors investigated clinical and imaging factors associated with early (hyperacute) infarct growth.
A total of 178 patients with known stroke onset, anterior circulation LVO (terminal ICA, M1 occlusion, or both), and baseline CT perfusion were analyzed. The collateral score, as assessed by the regional leptomeningeal score (rLMC), was correlated with the rate of infarct growth. When rLMC, a 20-point score, was trichotomized into good (rLMC >15), moderate (11-15), and poor (<11) collaterals, the rate of infarct growth varied considerably among the three groups. It was 0.17 mL/min, 0.26 mL/min, and 0.41 mL/min in patients with good, moderate, and poor collateral, respectively. In multivariable linear regression, only the rLMC score was independently associated with hyperacute infarction growth (adjusted β=−0.35; P<0.001).
The authors concluded that “hyperacute infarction growth is highly associated with leptomeningeal collateral status.” Of note, the association was evident irrespective of infarct growth model, linear vs logarithmic. Further, association between collateral and infarct growth was lost if less granular collateral score, for example, five-point Maas score, was used.
Few limitations are of note. First, this was a single-center cohort and included less than 25% of all registry patients, thereby introducing selection bias. Second, infarct size was based on “predicted core” derived using cerebral blood volume threshold (<1.2 ml/100 ml) with its inherent limitations. Finally, generalizations cannot be made to patients with stroke in posterior circulation. However, the rLMC can be easily calculated without sophisticated software, but requires some expertise in reading the score. rLMC is also a bit more cumbersome than simpler three- or five-point collateral score but is more robust.
Despite limitations, the study demonstrates the value of collateral assessment as a useful tool to help predict infarct growth rate and distinguish patients in whom infarct progresses more slowly from those who progress rapidly with limited window for intervention. The collateral assessment, therefore, might allow physicians to determine suitability of an individual patient for transfer based on his or her infarct growth curve. Only patients with infarct growth rate slow enough to keep their penumbral tissue volume stable during “transport time” will eventually have enough salvageable brain by the time they arrive to EVT centers, to make them eligible to endovascular therapy. This subset of patients, thus, would benefit from transfers. This will also help to reduce futile transfers and harmonize resource use.
