To model infarct growth rate as a logistic growth function and determine its association to hypoperfusion intensity ratio (HIR)

Understanding IGR in patients with Acute Ischemic stroke (AIS) due to Large Vessel Occlusion (LVO) may allow for individualized treatment to improve neurological recovery. Other IGR models, like slow vs fast progressor classification, may be too broad. Additionally, animal studies have successfully modelled IGR as a logistic growth function and found its parameters are dependent on collateral circulation. We present the first logistic model for IGR in human patients.

This a secondary analysis of a retrospective cohort of AIS patients due to anterior circulation LVO that underwent successful endovascular therapy in two comprehensive stroke centers. Two time measures were used: from stroke onset to CT perfusion and from this to reperfusion.  Final infarct volume was measured with DWI-MRI within 24 hours after EVT. IGR was modelled as a logistic growth function with a maximum infarct volume parameter (Vmax) and infarct growth rate parameter (r). These were estimated for each patient. To assess correlation to HIR, we used Pearson’s correlation coefficient.

We included 34 patients; 24 classified as having good collaterals (HIR<0.4) and 10 as having poor collaterals (HIR>0.4). Median age was 72.5 (IQR 56.25 - 83.0) and 50% were male. Median baseline NIHSS was 16 (IQR 12.5 – 18). None were different between groups. Vmax was 51.5 (37.6 in good HIR vs 59.6 in poor HIR, p=0.322), while r was 0.024 (0.031 vs 0.024, respectively, p=0.724). HIR’s Pearson’s coefficient was 0.455 (0.13 – 0.68, p=0.006) for Vmax, while it was 0.11 (-0.22 – 0.43, p=0.5076) for r.

IGR can be modelled as a logistic growth function with parameters dependent on multiple factors. When modelling IGR, we found that HIR was associated with infarct maximum growth, but not growth rate.