Futile complete recanalization: patients characteristics and its time course

As the goal of mechanical thrombectomy is shifting toward mTICI-3 rather than mTICI-2b, we sought to clarify the limitation of the effect of mTICI-3. A post-hoc analysis of a registry of large-vessel occlusion stroke from 46 centers was conducted. Among 2,420 registered patients, 725 patients with anterior circulation occlusion who achieved successful reperfusion were analyzed. We compared outcomes between patients with mTICI-3 and mTICI-2b, and investigated how the effect of mTICI-3 changed according to baseline characteristics and time course. The proportion of patients with favorable outcomes (mRS 0–2 at day 90) was higher among patients with mTICI-3 compared to those with mTICI-2b (adjusted OR, 2.10; 95% CI, 1.49–2.97). There was no heterogeneity in the effect of mTICI-3 with respect to age, neurological deficit, alteplase use, occluded vessels, or infarct size. mTICI-3 was associated with favorable outcomes when the puncture-to-reperfusion time was <80 minutes (adjusted OR, 2.28; 95% CI, 1.52–3.41), but not when the puncture-to-reperfusion time was ≥80 minutes. A significant heterogeneity was found in the effect of mTICI-3 reperfusion across the puncture-to-reperfusion time subgroups (P for interaction = 0.025). Until when operators should continue the procedure after mTICI-2b has been achieved, needs to be studied.

In this study, we sought to identify the subgroup of patients in whom the effect of mTICI-3 reperfusion was absent. For this purpose, we examined how the effect of mTICI-3 reperfusion changes according to baseline characteristics and the time to reperfusion using a nationwide large data set of patients with acute large vessel occlusion stroke.
Imaging and endovascular therapy. Diagnostic modalities before EVT were not unified in the RESCUE-Japan Registry 2. The Alberta Stroke Program Early Computerized Tomography Score (ASPECTS) was derived from computed tomography (CT) or magnetic resonance imaging (MRI) data. If both CT and MRI were performed before EVT, the lower ASPECTS was used. The final mTICI scores were assessed by the attending interventionalists. A successful reperfusion was defined as a score of mTICI-2b or 3 2 . The puncture to reperfusion time was defined as the interval between time of groin puncture and time when reperfusion status reached the final state, and was recorded by 5 minute intervals.
Outcomes. We set the modified Rankin Scale (mRS) 18 of 0-1 and 0-2 at 90 days after stroke onset as excellent and favorable outcomes, respectively. The primary outcome of this study was the favorable outcome, and the secondary outcomes were the excellent outcome and mortality at 90 days after stroke onset. As safety outcomes, any intracranial hemorrhage and symptomatic intracranial hemorrhage within 72 h of EVT according to SITS-MOST criteria 19 were considered. Statistical analysis. Patient characteristics and outcomes were compared between patients who achieved mTICI-3 and mTICI-2b reperfusion. These values are presented as median with interquartile range for continuous variables, and as numbers and percentages for the categorical variables. Continuous variables were compared using Wilcoxon rank sum test. Fisher's exact test or chi-squared test was used for categorical variables, when appropriate. We constructed univariate and multivariate logistic regression models. Adjusted variables in the multivariate logistic regression model are as follows: age, sex, National Institutes of Health Stroke Scale (NIHSS) score, ASPECTS, target occlusion location (internal carotid artery, the horizontal segment of the middle cerebral artery, or other vessels), intravenous alteplase administration, and time from onset to reperfusion.
In order to assess the heterogeneity of the effect of complete reperfusion by baseline characteristics, subgroups were defined as follows: (1) age ≤ 75 or> 75 years 17 ; (2) NIHSS score ≥10 or <10; (3) use or non-use of alteplase; (4) proximal (internal carotid artery or the horizontal segment of the middle cerebral artery) or distal (the insular segment or more peripheral segment of the middle cerebral artery, or anterior cerebral artery) vessel occlusion 20 ; and (5) ASPECTS ≥ 6 or <6 20,21 . In each subgroup, we investigated the association between mTICI and clinical outcomes. The adjusted variables in multivariate logistic regression models were same as above.
To assess how the delay of reperfusion modifies the effect of complete reperfusion, we divided the onset-to-reperfusion time, the onset-to-puncture time, and the puncture-to-reperfusion time by their upper quartile. Then, we investigated the association between mTICI and clinical outcomes in each subgroup. The adjusted variables in multivariate logistic regression models were the same as above except these: puncture-to-reperfusion time was adjusted for in analyzing the subgroups of onset-to-puncture time; onset-to-puncture time was adjusted for in analyzing the subgroups of puncture-to-reperfusion time. Furthermore, we analyzed how the mTICI3 effect changes according to time to reperfusion in continuous valuables. We report graphically plots of predicted probabilities of functional outcomes stratified by mTICI.
The interaction was tested using a multiplicative interaction term (mTICI*variable) included in the models. The time course was used as both dichotomized and continuous variable. All reported P values were 2-tailed, and statistical significance was established at P < 0.05. We did all analyses using SAS University Edition (SAS Institute Inc, Cary, NC) and drew figures with JMP Pro 14.3 (SAS Institute Inc, Cary, NC).

Results
Population. Among the 2,420 enrolled patients, 21 patients were excluded because 12 patients did not meet the eligibility criteria and 9 patients were lost to follow-up. Patients with posterior circulation occlusion were also excluded (n = 335). In the remaining 2,064 patients, 876 patients underwent EVT using stent retriever or aspiration catheter within 12 hours of onset. Among them, patients who achieved successful recanalization (mTICI-2b and 3) were included (n = 752). Of those patients, 27 were excluded due to incomplete records. As a result, a total of 725 patients were analyzed in the present study (Fig. 1).

Subgroup analyses based on baseline characteristics. The result of our subgroup analysis by baseline
characteristics is summarized in a forest plot (Fig. 2). Favorable outcome was observed significantly more often in patients with mTICI-3 than in those with mTICI-2b in the subgroups of age ≤ 75, age > 75, alteplase use, no alteplase, NIHSS score ≥ 10, ASPECTS ≥ 6, and proximal vessel occlusion. There was no heterogeneity in the effect of mTICI-3 reperfusion on favorable outcomes with respect to age, use of alteplase, NIHSS score, ASPECTS, or target occlusion location.
The ORs for excellent outcome are shown in Supplementary Fig. S1. Excellent outcome was observed significantly more often in patients with mTICI-3 than in those with mTICI-2b in all the subgroups, except subgroups of NIHSS < 10 and ASPECTS < 6. There was no heterogeneity in the effect of mTICI-3 reperfusion on excellent outcomes with respect to the baseline characteristics. www.nature.com/scientificreports www.nature.com/scientificreports/ in a forest plot (Fig. 3). In the subgroup of onset-to-reperfusion time <365, onset-to-puncture time <305, puncture-to-reperfusion time <80, favorable outcome was observed significantly more often in patients with mTICI-3 than those with mTICI-2b (adjusted OR, 2.32; 95% CI, 1.56-3.45; adjusted OR, 2.20; 95% CI, 1.47-3.31; and adjusted OR, 2.28; 95% CI, 1.52-3.41, respectively). Contrarily, there was no significant association between mTICI score and favorable outcome in the subgroup of puncture-to-reperfusion time ≥80 (adjusted OR, 1.15; 95% CI, 0.53-2.50), and there was significant heterogeneity in the effect of mTICI-3 reperfusion across the puncture-to-reperfusion time subgroups (P for interaction = 0.025), suggesting that the effect of mTICI-3 reperfusion diminishes within the highest quartile of puncture-to-reperfusion time.  Supplementary Fig. S2. In the subgroup of onset-to-reperfusion time <365, onset-to-puncture time <305, and puncture-to-reperfusion time <80, excellent outcome was observed significantly more often in patients with mTICI-3 than in those with mTICI-2b (adjusted OR, 2.35; 95% CI, 1.55-3.55; adjusted OR, 2.13; 95% CI, 1.39-3.26; and adjusted OR, 2.35; 95% CI, 1.53-3.61, respectively). On the other hand, there was no significant association between mTICI score and excellent outcome in the subgroup of puncture-to-reperfusion time ≥80 (adjusted OR, 1.17; 95% CI, 0.46-3.00), and there was a marginal heterogeneity in the effect of mTICI-3 reperfusion across the puncture-to-reperfusion time subgroups (P for interaction = 0.073).
Changes in the probability of favorable outcome with time course in continuous variables stratified by mTICI score are shown in Fig. 4. As a continuous variable, none of the onset-to-reperfusion time, onset-to-puncture time, or puncture-to-reperfusion time significantly modified the effect of complete reperfusion (P for interaction = 0.77, 0.70, 0.25, respectively). Changes in the probability of excellent outcome with time course in continuous variables stratified by mTICI score are shown in Supplementary Fig. S3. As a continuous variable, none of the time course variables significantly modified the effect of complete reperfusion.

Discussion
Our study demonstrated that the effect of mTICI-3 reperfusion was not significantly modified by patients' baseline characteristics such as age, sex, NIHSS score, ASPECTS, and target occlusion location among patients  www.nature.com/scientificreports www.nature.com/scientificreports/ with anterior circulation occlusion. Previous studies have reported the effect of complete reperfusion on outcomes [4][5][6][7][8][9][10][11][12][22][23][24] , but the heterogeneity of this effect according to patients' characteristics has not been fully investigated. Although the superiority of mTICI-3 reperfusion did not reach statistical significance in some subgroups, presumably due to insufficient power, our findings provide further support for the importance of achieving mTICI-3 reperfusion regardless of the baseline characteristics.
On the other hand, the association between mTICI-3 reperfusion and clinical outcome was not significant in patients who achieved mTICI-3 later than 80 minutes after the puncture, and the effect of complete reperfusion was modified by the puncture-to-reperfusion time ≥80 minutes. This is not surprising because the ischemic core grows into the penumbral region as time passes 25 , and the brain tissue that can be rescued becomes less. This study could not determine until when an operator should continue the procedure aiming for mTICI-3 reperfusion after mTICI-2b has been achieved; however, it is worth noting that long procedure time would limit the advantage of mTICI-3 reperfusion.
While a significant heterogeneity in the effect of mTICI-3 reperfusion was observed according to puncture-to-reperfusion time, whether the effect of mTICI-3 reperfusion changes by onset-to-puncture time remained unclear. Few benefits of mTICI-3 reperfusion were suggested by point estimates among patients with the highest onset-to-puncture time quartile, but the heterogeneity was not statistically significant. A recent study reported that the effect of complete reperfusion was retained among patients who underwent mechanical thrombectomy after 6 hours of stroke onset 26 . Onset-to-puncture time may be less critical than puncture-to-reperfusion time with respect to the effect of complete reperfusion, because onset-to-puncture time is sometimes imprecisely determined or documented 27 , while puncture-to-reperfusion time is generally accurately documented.
It is known that discrepancy exists between the assessment of the mTICI score by local operators and by independent core lab, although the degree of discrepancy varies 28,29 . As operators tend to overestimate the degree of reperfusion 28 , the successful (mTICI-2b/3) reperfusion rate in the RESCUE-Japan Registry 2 may be overestimated. In addition, more detailed grading scores including -2c (near-complete reperfusion) have been proposed to predict clinical outcomes precisely 30,31 . Those grading scores, which our cohort did not employ, may help to identify patients who can benefit from adding endovascular procedures. Further studies with core lab assessment and more detailed grading score are needed to clarify the limitation of the effect of complete reperfusion.
There are several limitations in this study. First, in this observational study, there might be unmeasured factors that inhibited complete reperfusion, and those could be confounded with the outcome. Second, assessments of mRS scores could be biased because mTICI scores were not blinded to physicians who rated mRS scores. Third, ischemic penumbra was not assessed. Finally, the heterogeneity in the effect of mTICI-3 reperfusion according to puncture-to-reperfusion time was not shown by the analysis where puncture-to-reperfusion time was included as a continuous variable. This may be because the relationship between clinical outcomes and puncture-to-reperfusion time is not linear, that is, the rate of favorable outcome decreases rapidly as time elapses in the early period of puncture-to-reperfusion time, and it reaches a plateau later 32 . The threshold of puncture-to-reperfusion time in this study is derived from our own registry, and its generalizability has not been confirmed. Therefore, it needs to be validated against other registries. conclusions mTICI-3 reperfusion, especially when achieved in less than 80 minutes, was superior to mTICI-2b reperfusion, regardless of baseline characteristics. Achieving early and complete reperfusion is important to improve outcomes. Until when operators should continue the procedure after mTICI-2b has been achieved, needs to be studied.

Data availability
The datasets generated during and/or analyzed during the current study are not publicly available but are available from the corresponding author on reasonable request.