Local Anesthesia Without Sedation During Thrombectomy for Anterior Circulation Stroke Is Associated With Worse Outcome

Patients undergoing MT for LVO stroke of the anterior circulation at 4 comprehensive stroke centers between January 1, 2018, and December 31, 2018, were included. Nine centers participate to date in the prospective multicenter ETIS registry in France, but data from 4 centers were used due to the participation of the other 5 centers after December 31, 2018. Data of the ETIS registry were collected and stored in electronic Case Report Forms (URL: https://www.clinicaltrials.gov. Unique identifier: NCT03776877) and informed consent was obtained. Previously published work described ETIS methodology in details.¹⁰ Briefly, the ETIS registry included consecutive patients with acute ischemic stroke with LVO treated with MT in high-volume (annual MT >250/y) comprehensive stroke centers in France. All patients 18 years of age or older were included in this study. No patients were excluded based on National Institutes of Health Stroke Scale (NIHSS) and Alberta Stroke Program Early CT Score scores on presentation, administration of intravenous thrombolysis, or time from onset to puncture.

CS as first-line sedation management before MT was defined as CS group, while the patients enrolled at one center, which used no sedation management before MT, was defined as no sedation group. No protocol was present in the no sedation group and protocol exists in every CS center. No anesthesiologist was present in the angiosuite during MT in the setting of LA, whereas anesthesiologist was systematically present for CS. Subcutaneous injection of ≈10 mL of Xylocaine was done for LA and was also applied in the setting of CS. CS was performed with intravenous injection of Remifentanil. Blood pressure was monitored noninvasively in both groups without specific objectives, except that mean blood pressure should be around >100 mm Hg in patients treated under CS. In both groups, the systolic blood pressure was maintained <185 mm Hg in case of prior intravenous thrombolysis in compliance with the American Heart Association/American Stroke Association guidelines.¹¹ Hypotension was treated with intravenous injection of Ephedrine during MT only in the CS group. The institutional review board at each site approved this study.

Our primary outcome was the favorable outcome, defined as a modified Rankin Scale (mRS) score 0–2 at 90 days or a mRS equal to prestroke situation. Ninety days after the acute event, functional outcome was assessed by board-certified vascular neurologists during a routinely scheduled clinical visit or by a study nurse certified in administering the mRS during a standardized telephone interview if the patient was unable to attend. Secondary outcomes included excellent outcome (90-day mRS score 0–1 or equal to prestroke mRS), early neurological improvement (24-hour NIHSS decrease ≥4 or NIHSS at 24-hour 0–1), 24-hour NIHSS shift (defined as baseline NIHSS- 24-hour NIHSS), final reperfusion status which was assessed by the modified Thrombolysis in Cerebral Ischemia grade, and the workflow metrics. Successful and complete reperfusion was defined as an modified Thrombolysis in Cerebral Ischemia grade 2b–3 and modified Thrombolysis in Cerebral Ischemia grade 3 at the end of MT, respectively. Safety outcomes included the all-causes mortality at 90 days, the occurrence of any procedural complication (dissection, perforation, and embolus in a new territory) and the occurrence of cerebral hemorrhage according to the ECASS II (European Collaborative Acute Stroke Study) classification.¹² Imaging criteria were evaluated by a local investigator of each center.

Quantitative variables are expressed as means (SD) in the case of normal distribution or medians (interquartile range) otherwise. Categorical variables are expressed as numbers (percentage). Normality of distributions was assessed using histograms and the Shapiro-Wilk test. Patients were divided in 2 groups according to center’s anesthetic protocol (LA versus CS). Baseline characteristics were described according to the 2 study groups, and the magnitude of the between-group differences was assessed by calculating the absolute standardized difference; an absolute standardized difference >10% was interpreted as a meaningful difference.¹³ We compared the outcomes between the 2 study groups after taking into account the potential confounding factors by using prespecified propensity score methods.¹⁴

The effects of the anesthetic approach were estimated by using propensity score matching method (propensity score was used to assemble well-balanced groups) as primary analysis and by using inverse probability of treatment weighting (IPTW) propensity score method (using stabilized inverse propensity score as weighty in regression models). The propensity score was estimated using a nonparsimonious multivariable logistic regression model, with anesthetic protocol as dependent variable and all the characteristics listed in the Table as covariates. Patients included in LA group were matched 1:1 to patients in CS group according to propensity score using the greedy nearest neighbor matching algorithm with a caliper width of 0.2 SD of logit for propensity score.^15,16 To evaluate bias reduction after using the propensity score matching method, absolute standardized differences were calculated after propensity score matching. Because of missing baseline data (range from 0% to 12.2%), we estimated the effect sizes in matched- and IPTW-propensity score cohorts after handling missing covariate values by multiple imputation,¹⁷ using a regression switching approach (chained equations with m=10 imputations obtained).¹⁸ Imputation procedure was performed under the missing at random assumption using all variables listed in the Table (including anesthesia groups) with a predictive mean matching method for continuous variables and multinomial or binary logistic regression model for categorical variables. In each imputed data set, we calculated the propensity score and assembled a matched cohort to provide both matched and IPTW-propensity score-adjusted effect sizes, which were later combined by using Rubin’s rules.¹⁹

In propensity score matched cohort, between-group comparisons were done using a generalized estimating equations model (binomial distribution, log function with a compound symmetry working correlation structure) for binary outcomes, or using a linear mixed model with a matched blocks as random effect for continuous outcomes (admission NIHSS was included as covariate fixed effect for comparison in NIHSS change from baseline to 24 hours). In IPTW-propensity score-adjusted cohort, comparisons were done using weighted log-binomial regression models for binary outcomes or linear regression models for continuous outcomes (admission NIHSS was including as covariate fixed effect for comparison in NIHSS change from baseline to 24 hours). Using patients in CS group as reference, we derived from these regression models, relative risks and mean difference, as treatment effect size measures, with their 95% CIs.

Our first analyses covered the whole study group according to an intention-to-treat principle. All analyses were repeated after excluding patients who did receive sedation (CS or GA) from the center with LA as first-line anesthetic approach and patients who received LA or GA from the centers with CS as first-line anesthetic approach (defined as a sensitivity per-protocol analysis). Statistical testing was conducted at the 2-tailed α-level of 0.05. Data were analyzed using the SAS software version 9.3 (SAS Institute, Cary, NC).

A total of 1034 acute ischemic stroke patients with LVO were admitted for MT at the participating centers during the study period. Of these, 762 were enrolled at 3 centers using a CS as first-line anesthetic approach, and the remaining 272 patients were enrolled at one center using LA. Of these, 126 patients from CS and 34 from the LA group were excluded from analysis due to absence of LVO or missing information on site of occlusion on noninvasive imaging (Figure 1). Of the patients from CS group, 7 (1.1%) received no sedation and 41 (6.6%) received GA. Of the patients from LA group, 37 (16.3%) received CS and 5 (2.2%) received GA. In patients who did not received MT, sedation management (either CS or GA) was used in 100% of cases (n=78/78) from CS group and 2.4% of cases (n=1/41) from LA group.

Two hundred twenty-two matched pairs were found in main intention-to-treat analysis, and 167 matched pairs were found in the sensitivity per-protocol analysis. The Table shows the baseline characteristics after handling missing values by multiple imputations according to the 2 study groups before and after propensity score matching (see Table I in the Data Supplement for baseline characteristics before matching and handling missing values, and Table II in the Data Supplement for baseline characteristics before and after matching in per-protocol analysis). Before matching, several meaningful differences were found; the stronger differences (absolute standardized difference >30%) were observed for cardioembolism cause, site of occlusion in noninvasive imaging, use of intravenous thrombolysis prior MT and time from onset to angiosuite admission. These differences were reduced after propensity score matching with a maximum absolute standardized difference of 13.9% for time from onset to angiosuite admission (Table; Figure I in the Data Supplement).

In the propensity score matched cohort, favorable outcome (primary outcome) was achieved significantly less often in the LA group (40.0%) than in the CS group (52.0%, matched relative risk=0.76 [95% CI, 0.60–0.97], Figure 2). Similarly, the rate of successful reperfusion was significantly lower in the no sedation group than in the CS group (76.6% versus 87.1%; matched relative risk=0.88 [95% CI, 0.79–0.98]). In IPTW-propensity score-adjusted cohort, similar significant differences were found both for favorable outcome and successful reperfusion. The difference in favorable outcome in favor of CS is mainly observed for carotid T and tandem occlusion, although the between-group difference did reach the significance level only in IPTW-adjusted analysis for Carotid T occlusion (Figure 3). Regarding other study efficacy outcomes (excellent outcome, early neurological improvement, 24-hour NIHSS shift, or complete reperfusion), the effectiveness criteria of the no sedation group were worse than those of the CS group, with only a significant difference for complete reperfusion in IPTW-propensity score-adjusted cohort (relative risk, 0.73 [95% CI, 0.57–0.93]). The mean NIHSS improvement from baseline to 24 hours was not different between the 2 groups in both matched (mean log difference, −0.80 [95% CI, −2.76 to 1.16]; P=0.42) and IPTW (−1.4 [95% CI, −2.9 to 0.06]; P=0.059) propensity score cohorts. In patients with successful reperfusion, there was no significant difference in the time from angiosuite arrival to reperfusion in both matched (mean loge difference, −0.04 [95% CI, −13.2 to 0.06]; P=0.44) and IPTW (−0.02 [95% CI, −10.1 to 0.06]; P=0.055) propensity score cohorts.

With respect to the safety outcomes, we found no significant differences in the matched-propensity score cohort (Figure 2). However, in IPTW, a higher mortality rate (odds ratio, 1.66 [95% CI, 1.27–2.15]) and lower parenchymal hematoma rate (OR, 0.59 [95% CI, 0.35–0.97]) were observed in the LA group. The sensitivity analysis restricted to our per-protocol sample provided similar results across all studied outcomes in the matched-as well as in the IPTW-propensity cohorts (Table III in the Data Supplement).

Fondation Adolphe de Rothschild: Michel Piotin, Raphael Blanc, Hocine Redjem, Simon Escalard, Jean-Philippe Desilles, Gabriele Ciccio, Stanislas Smajda, Mikael Mazighi, Mikael Obadia, Candice Sabben, Ovide Corabianu, Thomas de Broucker, Didier Smadja, Sonia Alamowitch, Olivier Ille, Eric Manchon, Pierre-Yves Garcia, Guillaume Taylor, Malek Ben Maacha.

Hôpital Foch: Adrien Wang, Serge Evrard, Maya Tchikviladze, Nadia Ajili, Bertrand Lapergue, David Weisenburger, Lucas Gorza, Oguzhan Coskun, Arturo Consoli, Federico Di Maria, Georges Rodesh, Morgan Leguen, Julie Gratieux, Fernando Pico, Haja Rakotoharinandrasana, Philippe Tassan, Roxanna Poll, Sylvie Marinier.

CHU Bordeaux: Gaultier Marnat, Florent Gariel, Xavier Barreau, Jérôme Berge, Louis Veunac, Patrice Menegon, Igor Sibon, Ludovic Lucas, Stéphane Olindo, Pauline Renou, Sharmila Sagnier, Mathilde Poli, Sabrina Debruxelles, Thomas Tourdias, Jean-Sebastien Liegey.

CHU Nantes: Romain Bourcier, Lili Detraz, Benjamin Daumas-Duport, Pierre-Louis Alexandre, Monica Roy, Cédric Lenoble, Vincent L’allinec, Jean-Baptiste Girot, Hubert Desal.

CHU Nancy: Benjamin Gory, Isabelle Costa, Serge Bracard, René Anxionnat, Marc Braun, Anne-Laure Derelle, Romain Tonnelet, Liang Liao, François Zhu, Emmanuelle Schmitt, Sophie Planel, Sébastien Richard, Lisa Humbertjean, Gioia Mione, Jean-Christophe Lacour, Nolwenn Riou-Comte, Gabriela Hossu, Marine Beaumont, Mitchelle Bailang, Gérard Audibert, Marie Reitter, Agnès Masson, Lionel Alb, Adriana Tabarna, Marcela Voicu, Iona Podar, Madalina Brezeanu.

CHU Montpellier: Vincent Costalat, Caroline Arquizan, Cyril Dargazanli, Grégory Gascou, Pierre-Henri Lefèvre, Imad Derraz, Carlos Riquelme, Nicolas Gaillard, Isabelle Mourand, Lucas Corti.