Ischemic stroke Original research Health economic impact of first-­pass success among patients with acute ischemic stroke treated with mechanical thrombectomy: a United States and European perspective Osama O Zaidat ‍ ‍,1 Marc Ribo ‍ ‍,2,3 Heinrich Paul Mattle,4 Jeffrey L Saver,5 Hormozd Bozorgchami,6 Albert J Yoo,7 Alexandra Ehm,8 Emilie Kottenmeier,9 Heather L Cameron,10 Rana A Qadeer,10 Tommy Andersson11,12 ►► Additional material is ABSTRACT economic burden was estimated to be €45 billion published online only. To view, Background First-­pass effect (FPE), restoring complete in the European Union (EU) alone in 2015.3 please visit the journal online (http://​dx.​doi.​org/​10.​1136/​ or near complete reperfusion (modified Thrombolysis in Approximately 87% of all strokes are acute isch- neurintsurg-­​2020-­​016930). Cerebral Infarction (mTICI) 2c-­3) in a single pass, is an emic strokes4 and 24% to 38% of acute ischemic independent predictor for good functional outcomes in strokes are caused by large vessel occlusions, which For numbered affiliations see the endovascular treatment of acute ischemic stroke. The are associated with additional costs.5 A study from end of article. economic implications of achieving FPE have not been the USA demonstrated that total costs per patient assessed. were higher for patients with large vessel occlu- Correspondence to Dr Osama O Zaidat, Department Objective To assess the economic impact of achieving sions than for those without large vessel occlusions of Neuroscience, Mercy Saint complete or near complete reperfusion after the first (US$18 815 vs US$15 174).6 Vincent Medical Center, Toledo, pass. Complete reperfusion of brain tissue is the OH 43608, USA; ​OOZaidat@​ Methods Post hoc analyses were conducted using primary goal in treatment of acute ischemic stroke. Mercy.​com ARISE II study data. The target population consisted of Rapid revascularization is vital as reduced time Received 4 October 2020 patients in whom mTICI 2c–3 was achieved, stratified from symptom onset to revascularization is strongly Revised 23 November 2020 into two groups: (1) mTICI 2c–3 achieved after the first correlated with improved clinical and functional Accepted 24 November 2020 pass (FPE group) or (2) after multiple passes (non-­FPE outcomes.7 8 Complete revascularization with Published Online First group). Baseline characteristics, clinical outcomes, mechanical thrombectomy, a cost-­ effective treat- 21 December 2020 and healthcare resource use were compared between ment for patients with acute stroke, might require groups. Costs from peer-­reviewed literature were applied multiple passes, which prolongs procedure time and to assess cost consequences from the perspectives of increases the risk of arterial endothelial injury.9 10 the United States (USA), France, Germany, Italy, Spain, Research suggests that first-­pass effect (FPE)—that Sweden, and United Kingdom (UK). is, restoring complete or near complete reperfu- Results Among patients who achieved mTICI 2c–3 sion (modified Thrombolysis in Cerebral Infarction (n=172), FPE was achieved in 53% (n=91). A higher (mTICI) 2c–3) in a single pass, is an independent proportion of patients in the FPE group reached good predictor of good functional outcome (modified functional outcomes (90-­day modified Rankin Scale Rankin Scale (mRS) score 0–2).9 Although the clin- score 0–2 80.46% vs 61.04%, p<0.01). The patients ical benefits of achieving complete or near complete in the FPE group had a shorter mean length of stay reperfusion after the first pass have been investi- (6.10 vs 9.48 days, p<0.01) and required only a single gated, the potential economic benefit has not been stent retriever, whereas 35% of patients in the non-­FPE assessed. group required at least one additional device. Driven The objective of this analysis was to assess the by improvement in clinical outcomes, the FPE group economic impact of achieving complete or near had lower procedural/hospitalization-­related (24– complete reperfusion after the first pass, FPE, 33% reduction) and annual care (11–27% reduction) as compared with the need for multiple passes costs across all countries. to achieve the same level of reperfusion. Clinical Conclusions FPE resulted in improved clinical outcomes were derived from patient-­level data of outcomes, translating into lower healthcare resource use the ARISE II (Analysis of Revascularization in Isch- and lower estimated costs. emic Stroke with EmboTrap) study, and the analysis © Author(s) (or their employer(s)) 2021. Re-­use was done from the perspective of various health- permitted under CC BY-­NC. No care systems (USA, France, Germany, Italy, Spain, commercial re-­use. See rights Sweden, and United Kingdom (UK)). and permissions. Published INTRODUCTION by BMJ. Worldwide, stroke is the leading cause of mortality To cite: Zaidat OO, Ribo M, and disability with a substantial economic impact.1 METHODS Mattle HP, et al. The direct and indirect costs of stroke were esti- Data source J NeuroIntervent Surg mated at US$33.9 billion in the United States (USA) Post hoc analyses were conducted using patient-­ 2021;13:1117–1123. in 2012–2013.2 Similarly, in Europe, the annual level data from the ARISE II study, a prospective Zaidat OO, et al. J NeuroIntervent Surg 2021;13:1117–1123. doi:10.1136/neurintsurg-2020-016930    1117 Ischemic stroke single-­arm international multicenter clinical trial, investigating passes (non-­FPE group). The two groups were defined based on the efficacy and safety of the EmboTrap device (Cerenovus, the core laboratory adjudicated mTICI scores measured after the Irvine, California, USA) in 227 treated patients. Although the first pass (ie, total number of passes did not affect categorization). EmboTrap device was used in all procedures as a first-­line treat- ment, the technique used in the ARISE II study was based on Primary analyses investigator’s choice (ie, the physicians established an appro- Primary analyses were conducted in two steps. First, baseline priate treatment plan based on the subject’s medical condition characteristics, clinical outcomes, and healthcare resource use and available diagnostic screening procedures); specific recom- were evaluated for each group (ie, FPE and non-­FPE) using data mendations were not provided for the use of intermediate cath- from the ARISE II study. Second, cost data from peer-­reviewed eters (data as to whether the stent retriever was 'engulfed' or literature were applied to assess cost consequences using a deter- 'corked' are not available); and pump suction was not used. ministic approach from two different perspectives, provider and Details of the ARISE II study are provided in Zaidat et al 2018.11 payer. Target population and comparators Aligned with the ARISE II publication,11 FPE was defined as Baseline characteristics complete or near complete reperfusion (mTICI 2c–3) after the Baseline characteristics were compared between the FPE and first pass with the EmboTrap device. As such, the target popu- non-­ FPE groups to examine the risk of systematic bias and lation consisted of patients in whom mTICI 2c–3 was achieved confounders (table 1). (n=172) in the ARISE II population (patients in whom mTICI 2c–3 was not achieved (n=55) were excluded from the analyses Clinical outcomes to avoid potential selection bias). The target population was then The following clinical outcomes were assessed: core laboratory stratified into two groups: (1) patients in whom mTICI 2c–3 was adjudicated mTICI score after each pass and at the end of the achieved after the first pass (FPE group) or (2) after multiple procedure, 90-­ day mRS score (categorized into good (mRS Table 1 Baseline characteristics for the FPE and non-­FPE groups Variable FPE (n=91) Non-­FPE (n=81) P Value* Demographics  Age, mean (SD) 68.4 (11.9) 67.7 (14.3) 0.70  Sex, male, n (%) 43 (47.3%) 33 (40.7%) 0.39 Vascular risk factors, n (%)  Hypertension 66 (72.5%) 57 (70.4%) 0.75  Atrial fibrillation 32 (35.2%) 38 (46.9%) 0.12  Diabetes mellitus 16 (17.6%) 17 (21.0%) 0.57  Dyslipidemia 38 (41.8%) 37 (45.7%) 0.61  Smoking 19 (20.9%) 21 (25.9%) 0.43  Previous MI/CAD 23 (25.3%) 11 (13.6%) 0.06  Previous stroke 16 (17.6%) 13 (16.0%) 0.79 Clinical presentation, median (IQR)  Baseline NIHSS score 16 (11–19) 16 (13–20) 0.40  Baseline systolic BP, mm Hg† 144 (130–158) 148 (134–158) 0.31  Baseline diastolic BP, mm Hg† 80 (70–91) 81 (72–93) 0.29 Occlusion location, n (%)  Internal carotid artery 14 (15.4%) 13 (16.0%) 0.90  M1 middle cerebral artery 48 (52.7%) 46 (56.8%) 0.60  M2 middle cerebral artery 22 (24.2%) 20 (24.7%) 0.94  Posterior 7 (7.7%) 2 (2.5%) 0.17 Procedural factors  Time from onset to puncture (min), median (IQR) 214 (161–263) 220 (153–270) 0.54  General anesthesia, n (%) 33 (36.3%) 25 (30.9%) 0.46  IV tPA use, n (%) 59 (64.8%) 56 (69.1%) 0.55  Balloon guide catheter use, n (%) 71 (78.0%) 60 (74.1%) 0.54  Intermediate catheter use, n (%) 20 (22.0%) 41 (50.6%) <0.01 *P values presented for t-­test beside means, Wilcoxon rank sum (Mann-­Whitney) test beside medians, and chi-­square or Fisher’s exact tests beside proportions. †Four patients missing data for systolic and diastolic BP. BP, blood pressure; CAD, coronary artery disease; FPE, first-­pass effect; IQR, interquartile range; IV, intravenous; MI, myocardial infarction; NIHSS, National Institutes of Health Stroke Scale; SD, standard deviation; tPA, tissue plasminogen activator. 1118 Zaidat OO, et al. J NeuroIntervent Surg 2021;13:1117–1123. doi:10.1136/neurintsurg-2020-016930 Ischemic stroke score 0–2) and excellent (mRS score 0–1) functional outcomes), resource costs were increased/decreased by 20% as confidence 90-­day mortality, occurrence of symptomatic intracranial hemor- intervals for these costs were not reported in the literature. A rhage within 24 hours postprocedure based on the Heidelberg hypothetical cost-­efficiency analysis was also conducted using a Bleeding Classification, embolization into new territory, and cohort of 200 cases, where the procedural/hospitalization-­related discharge status (home, acute care facility, or extended care) impact of using two devices with variable rates of achieving FPE assessed at 7 days post-­procedure. was compared. In this analysis, it was assumed that one device achieved FPE in an additional 5% of the cases as compared with Healthcare resource use the other device. Procedural/hospitalization-­related resources, associated with the initial stroke, were compared between the FPE and non-­FPE Statistical analyses groups (total hospital length of stay (LOS), days in the inten- Continuous variables were described using the mean or median sive care unit, standard bed days, and devices used during the and standard deviation (SD) or IQR, while categorical variables procedure). were described using the number of observations and propor- tions. Statistical significance testing was based on a p value of <0.05 for the following tests: t-­test and Wilcoxon rank-­sum Economic outcomes (Mann-­Whitney) test for continuous variables and chi-­square or Cost consequences were compared between the FPE and non-­ Fisher’s exact test for categorical variables. All statistical analyses FPE groups from the perspective of healthcare systems in the of the ARISE II data were performed using Stata 15 (StataCorp, USA, France, Germany, Italy, Spain, Sweden, and UK with two College Station, Texas, USA). separate time horizons (procedural/hospitalization-­related and annual care in the first year after stroke, details provided below). Clinical and healthcare resource use was obtained from the RESULTS ARISE II study, and country-­specific healthcare resource costs In the ARISE II study, complete or near complete reperfusion were obtained from peer-­reviewed literature and public market (ie, mTICI 2c–3) was achieved in 76% of the patients (n=172); research reports, which were validated by clinical experts in among these patients, FPE was observed in 53% (n=91) (online interviews. All costs were reported as 2020 currencies and 2020 supplemental figure I). Baseline characteristics were mostly United States dollars (US$), and were inflated using country-­ balanced between the FPE and non-­FPE groups. Notably, inter- specific inflation indices if required. The Consolidated Health mediate catheter use was lower in the FPE group than in the Economic Evaluation Reporting Standards (CHEERS) statement non-­FPE group (22.0% vs 50.6%, p<0.01; table 1). was used for the reporting of the economic analyses.12 A significantly higher proportion of patients in the FPE The procedural/hospitalization-­ related economic impact group achieved good (mRS 0–2, 80.46% vs 61.04%, p<0.01) was calculated in the following steps. First, the procedural/ or excellent (mRS 0–1, 63.22% vs 46.75%, p=0.03) func- hospitalization-­related healthcare resource use was obtained for tional outcomes as compared with patients in the non-­ FPE patients in the FPE and non-­FPE groups from the ARISE II study. group (table 2 and online supplemental figure II). Although the Second, country-­specific costs valuing those resources were taken proportion of patients who died or had symptomatic intracra- from peer-­reviewed literature (LOS costs) and public market nial hemorrhage were lower in the FPE group, the differences research reports (device costs).13–21 The resources captured in were not statistically significant. The proportion of patients the LOS costs were similar across the literature (online supple- who had an embolization into a new territory was significantly mental table I). Third, total costs per patient from a provider lower in the FPE group (2.20% vs 11.11%, p=0.03; table 2). perspective in the acute care setting were calculated. The proportions of patients discharged to their home or to The costs per patient arising during the first year after the acute/extended care facilities (skilled nursing home, rehabilita- stroke from a payer perspective were also calculated in three tion), respectively, were similar between the FPE and non-­FPE steps. First, proportions of patients achieving each mRS score (ie, groups (43.53% and 44.74% discharged to home and 56.47% mRS 0 to 6) at 90 days were obtained for each group. Second, and 55.26% discharged to acute/extended care facilities, respec- country-­specific annual care costs, based on 90-­day mRS score, tively, p=0.88). were obtained from the literature but were not available for Patients in the FPE group required only a single EmboTrap Germany and Spain.16 22–25 The resources captured in the annual device, whereas 35% of patients in the non-­FPE group required care costs mainly focused on direct healthcare resources (online at least one additional device. The patients in the FPE group supplemental table I). Third, total costs per patient arising from were also discharged significantly earlier with a shorter mean a payer perspective were calculated. LOS (6.10 (IQR=3.00–8.00) vs 9.48 (IQR=3.00–11.00) days, p<0.01). Patients in the FPE group had significantly fewer mean number of days spent in a standard bed (3.05 (IQR=0.00–5.00) Sensitivity analyses vs 6.13 (IQR=1.00–8.00), p<0.01), whereas the mean number To test the robustness of the results, univariate sensitivity of days spent in the intensive care unit was similar between the analyses varying key input parameters were conducted. These two groups (3.39 (IQR=2.00–4.00) vs 3.58 (IQR=2.00–4.00), included analyses with two alternative definitions of FPE11 26: (1) p=0.70; table 2). 'FPE (mTICI 3)', defined as achieving mTICI 3 after the first pass Subsequent to the improvements in clinical outcomes and and (2) 'FPE (mTICI 2b–3)', defined as achieving mTICI 2b–3 lower healthcare resource use during the acute care phase, the after the first pass. Accordingly, based on the chosen definition economic cost assessment accordingly estimated lower costs of FPE, the target population was also adjusted (eg, patients in for the FPE group. When the acute care costs arising within whom mTICI 3 was not achieved were excluded from the anal- the hospital (ie, procedural/hospitalization-­related costs) were yses when the definition of FPE was changed to FPE (mTICI 3)). assessed, achieving FPE led to potential per-­patient cost savings in Two additional parameters were also varied: (1) total LOS was every country studied (US$6575 for the USA, €1560 (US$1833) varied within its interquartile range (IQR) since this parameter for France, €2202 (US$2587) for Germany, €2901 (US$3409) was not normally distributed and (2) country-­specific healthcare for Italy, €4548 (US$5343) for Spain, Kr 29 468 (US$3364) Zaidat OO, et al. J NeuroIntervent Surg 2021;13:1117–1123. doi:10.1136/neurintsurg-2020-016930 1119 Ischemic stroke Table 2 Clinical and healthcare resource use outcomes for the FPE and non-­FPE groups Outcome FPE (n=91) Non-­FPE (n=81) P value* Clinical outcomes 90-­Day mRS score (categorized into good and excellent functional outcomes)†  Good outcomes (mRS score 0–2), n (%) 70 (80.46%) 47 (61.04%) <0.01  Excellent outcomes (mRS score 0–1), n (%) 55 (63.22%) 36 (46.75%) 0.03 90-­Day mortality,‡ n (%) 5 (5.68%) 11 (13.75%) 0.08 sICH within 24 hours, n (%) 2 (2.20%) 4 (4.94%) 0.42 ENT, n (%) 2 (2.20%) 9 (11.11%) 0.03 Healthcare resource use outcomes LOS  Total LOS, mean (IQR)§ 6.10 (3.00–8.00) 9.48 (3.00–11.00) <0.01  Standard bed days, mean (IQR)¶ 3.05 (0.00–5.00) 6.13 (1.00–8.00) <0.01  ICU days, mean (IQR)** 3.39 (2.00–4.00) 3.58 (2.00–4.00) 0.70 *P values presented for the following tests: t-­test or Wilcoxon rank-­sum (Mann-­Whitney) test for continuous variables and chi-­square or Fisher’s exact test for categorical variables. †Eight patients missing data for 90-­day mRS score. ‡Four patients missing data for 90-­day mortality. §Represents the date of discharge minus date of admission to study hospital; seven patients missing data for LOS. ¶Represents the normal ward days; 10 patients missing data for standard bed days. **Eight patients missing data for ICU days. ENT, embolization into new territory; FPE, first-­pass effect; ICU, intensive care unit; IQR, interquartile range; LOS, length of stay; mRS, modified Rankin Scale; NA, not applicable; sICH, symptomatic intracerebral hemorrhage. for Sweden, and £1751 (US$2285) for the UK; table 3, online the main analyses, limiting the FPE group to FPE (mTICI 3) led supplemental figure III). Similarly, achieving FPE led to poten- to potential per-­patient procedural/hospitalization-­related and tial per-­patient annual care cost savings in the first year after annual care cost savings (online supplemental table III). When the stroke in every country studied (US$4116 for the USA, €2131 criteria for FPE were broadened (ie, defined as mTICI 2b–3 after (US$2503) for France, €701 (US$823) for Italy, Kr 13 333 for the first pass), FPE was seen in 56% (n=117) of the patients who Sweden (US$1522), and £2132 (US$2783) for the UK; table 4, finally achieved at least mTICI 2b. Notably, achieving mTICI online supplemental figure IV). 2b–3 after the first pass was also associated with improved clin- When the criteria for FPE were defined to be more limited ical outcomes and reduced healthcare resource use and costs (ie, mTICI 3 after the first pass), this FPE was observed in 57% (n=67) of the patients who finally achieved mTICI 3 (n=118). (online supplemental tables III and IV). Variations of the total The FPE (mTICI 3) group was associated with improved clinical LOS within the IQR and increasing or decreasing the costs by outcomes and reduced healthcare resource use as compared with 20% did not alter the direction of the results but did change the non-­FPE group, including patients with mTICI 2c reperfu- the magnitude of the estimated cost savings (online supplemental sion after first-­pass (online supplemental table II). Similar to table III). Table 3 Average per-­patient procedural and hospitalization-­related healthcare resource use costs for the FPE and non-­FPE groups US France Germany Italy Spain Sweden UK (2020 US$) (2020 €) (2020 €) (2020 €) (2020 €) (2020 Kr) (2020 £) FPE (n=91)  LOS* $9832 €1381 €3096 €3124 €6542 Kr 68 942 £3019  Devices/methods used $7886 €2325 €1415 €3458 €2709 Kr 26 424 £2416  Total $17 718 €3706 €4511 €6581 €9252 Kr 95 365 £5435 Non−FPE (n=81)  LOS* $13 616 €2146 €4811 €4854 €10 167 Kr 89 376 £3955  Devices/methods used $10 678 €3120 €1902 €4628 €3632 Kr 35 457 £3231  Total $24 293 €5266 €6713 €9483 €13 799 Kr 124 833 £7186 Difference (2020 currencies) −$6575 −€1560 −€2202 −€2901 −€4548 −Kr 29 468 −£1751 Difference (2020 US$)† −$6575 −$1833 −$2587 −$3409 −$5343 −$3364 −$2285 Numbers may not sum due to rounding. *Costs for LOS for US, UK, and Sweden were based on total LOS and proportion of standard bed vs intensive care unit days (standard bed and intensive care unit days costed separately). Costs for LOS for France, Italy, Spain, and Germany were based on total LOS (separate costs for intensive care unit stay and standard bed days not found in the literature; as such, a general cost for LOS was used). †Exchange rates reported for August 03, 2020, 16:00 UTC were used. The exchange rates were as follows: €1.00=US$1.17, Kr1.00=US$0.11, and £1.00=US$1.31. FPE, first-­pass effect; LOS, length of stay; UTC, coordinated universal time. 1120 Zaidat OO, et al. J NeuroIntervent Surg 2021;13:1117–1123. doi:10.1136/neurintsurg-2020-016930 Ischemic stroke Table 4 Average per-­patient annual care costs, based on 90-­day mRS score, for the FPE and non-­FPE groups % Achieving mRS score Average costs for annual care/mRS score by country mRS score FPE (n=91) Non-­FPE (n=81) US (2020 US$) France (2020 €) Italy (2020 €) Sweden (2020 Kr) UK (2020 £) 0 41.38% 19.48% $12 176 €12 817 €2072 Kr 74 717 £3255 1 21.84% 27.27% $12 538 €12 817 €2437 Kr 98 093 £3829 2 17.24% 14.29% $14 504 €12 817 €2803 Kr 153 609 £4403 3 9.20% 7.79% $24 905 €22 244 €5524 Kr 186 481 £15 666 4 2.30% 11.69% $50 408 €40 144 €7474 Kr 244 084 £21 196 5 2.30% 5.19% $74 108 €40 144 €16 599 Kr 297 096 £34 418 6 5.75% 14.29% – – – – – Results by country FPE (n=91) $15 430 €14 203 €2934 Kr 108 413 £5661 Non-­FPE (n=81) $19 546 €16 334 €3635 Kr 121 746 £7794 Difference (2020 currencies) −$4116 −€2131 −€701 −Kr 13 333 −£2132 Difference (2020 US$)* −$4116 −$2503 −$823 −$1522 −$2783 Numbers may not sum due to rounding. *Exchange rates reported for August 03, 2020, 16:00 UTC were used. The exchange rates were as follows: €1.00=US$1.17, Kr1.00=US$0.11, and £1.00=US$1.31. FPE, first-­pass effect; mRS, modified Rankin Scale; UTC, coordinated universal time. In the hypothetical analysis that applied procedural/ use. Patients in the FPE group required fewer thrombectomy hospitalization-­related costs for the FPE versus non-­FPE groups devices; the number of devices used to achieve mTICI 2c–3 is (eg, for the US analyses, per-­patient costs for the FPE and non-­ inevitably linked to the number of passes even though up to three FPE group were $17 718 and $24 293, respectively) and assessed attempts with the same device were allowed in the EmboTrap the impact of two devices with variable rates of achieving FPE instructions for use. Patient recovery was faster for patients with in a cohort of 200 a per year, the device that achieved a 5% FPE as indicated by the significantly shorter total hospital LOS. higher FPE rate led to cost savings of US$65 753 in the USA, As such, achieving mTICI 2c–3 after the first pass led to poten- US$18 328 in France, US$25 868 in Germany, US$34 087 in tial per-­ patient procedural/hospitalization-­ related cost savings Italy, US$53 433 in Spain, US$33 637 in Sweden, and US$22 847 in every country studied. Notably, the reductions in healthcare in the UK. resource use and cost savings were driven by the improvements in clinical outcomes. An analysis of the GAIN international trial DISCUSSION demonstrated that LOS in hospital during the first 90 days after We assessed the economic impact of achieving FPE in patients a stroke was positively associated with 90-­ day mRS score.28 with acute ischemic stroke from the perspective of healthcare Long-­term direct healthcare costs of stroke are mainly driven systems in the USA, France, Germany, Italy, Spain, Sweden, and by patients who do not achieve functional recovery and their UK using patient-­level data from the ARISE II study and cost degree of disability.29 30 Thus, due to the higher rate of patients data from peer-­reviewed literature. The results demonstrated achieving functional independence in the FPE group, our anal- that patients achieving FPE had improved functional outcomes, ysis showed that patients in the FPE group had lower costs in lower risk for embolization into a new territory, and a faster the first year after stroke than those in the non-­FPE group. recovery time (shorter LOS). These clinical improvements led to These findings are also aligned with a recent Markov model reduced procedural/hospitalization-­related and annual care costs that showed that achieving expanded TICI (eTICI) 3 resulted in in every country studied. The findings proved to be robust after healthcare (US$10 327) and societal (US$20 224) cost savings conducting a range of sensitivity analyses that varied key inputs, as compared with achieving eTICI 2b among patients with large including the definition of FPE. vessel occlusions.31 The clinical findings in this analysis of the ARISE II data are Rapid revascularization is widely accepted to be strongly supported by previous literature.9 26 Zaidat et al demonstrated correlated with improved clinical outcomes.7 8 It might be that patients in the North American Solitaire Stent Retriever argued that the improvements in clinical outcomes seen in this Acute Stroke (NASA) registry (n=354) with FPE had improved study were not due to FPE per se but due to the prolonged functional outcomes (90-­day good functional outcome: 55.1% time between groin puncture and reperfusion associated with vs 31.3%, p<0.01) and reduced mortality (16.3% vs 36.5%, multiple passes.26 However, a matched case–control analysis that p<0.01) as compared with patients without FPE.9 Similarly, a adjusted for time from symptom onset to complete reperfusion study by Nikoubashman with 164 consecutive patients in whom showed that the occurrence of good functional outcomes was mTICI 3 was achieved showed that complete reperfusion after almost twice as common among those who achieved complete the first pass was associated with improved functional outcomes reperfusion after the first pass as compared with those who (90-­day good functional outcome: 62% vs 40%, p=0.013).26 achieved complete reperfusion after multiple passes.26 Addition- Additionally, the association between FPE and favorable clinical ally, Garcia-­Tornel and colleagues reported a linear association outcome (mRS score at 90 days of 0–2 or equal to pre-­stroke between the number of passes and good functional outcomes mRS score) has been shown to be independent of the technique among patients who underwent mechanical thrombectomy used (stent retriever vs aspiration, p=0.29).27 procedures for large vessel occlusion in the anterior circulation This analysis demonstrated that achieving mTICI 2c–3 after and achieved recanalization as compared with patients who did the first pass was associated with reduced healthcare resource not achieve recanalization.32 Jindal et al also reported that the Zaidat OO, et al. J NeuroIntervent Surg 2021;13:1117–1123. doi:10.1136/neurintsurg-2020-016930 1121 Ischemic stroke proportion of patients who achieved good functional outcome costs for death (ie, mRS score 6), which had a lower incidence was inversely related to the number of passes among anterior in the FPE group. Fourthly, the economic impact of FPE was cerebral circulation stroke thrombectomy cases.33 Similarly, assessed in a deterministic analysis (ie, a probabilistic model multivariable analyses from Zaidat et al demonstrated that was not created as we used point estimates from one clinical achieving FPE was an independent predictor of good functional study and costs from the literature, which often did not report outcome.9 the associated measures of variability). However, we used Although there are factors that might affect the likelihood of the best available estimates for all key inputs and conducted achieving FPE (eg, factors related to the treating physician such numerous sensitivity analyses, including those that altered the as training, settings such as available equipment, and patients definition of FPE. such as clot composition), recent research has suggested that certain techniques involving the combined use of stent retrievers and intermediate aspiration catheters could improve recanaliza- CONCLUSIONS tion rates and the rate of FPE.34 35 Thus it might be beneficial to According to our knowledge, this is the first study which assesses 'throw everything in' immediately to improve the likelihood of the health economic impact of FPE. The analysis showed that achieving FPE. In this context it might be surprising that the use patients in the FPE group had improved clinical outcomes as of intermediate catheters was less frequent in the FPE group than compared with patients in the non-­FPE group, which resulted in the non-­FPE group (22.0% vs 50.6%, p<0.01), as several in lower healthcare resource use and improved economic published techniques (eg, SAVE, PROTECTPLUS) that incorpo- outcomes. Overall, the findings indicate that the performance rate intermediate catheters report a high percentage of FPE.34 36 within the thrombectomy procedure affects both the clinical Potential explanations for the higher use of intermediate cath- outcome and healthcare spending in the short and long term. eters in the non-­FPE versus FPE group include electing to use Thus FPE represents a relevant procedural goal for endovas- the intermediate catheter after first-­pass failure (perhaps, due to cular treatment of acute ischemic stroke. Moreover, the first-­line thrombus composition and/or lack of standardized practices). treatment should ideally involve a thrombectomy technique that Research examining the reasons for improved clinical outcomes provides the best chance of succeeding in the first pass. Future with FPE is ongoing; current literature suggests that repeated economic studies that obtain clinical and economic data from thrombectomy passes may be associated with an increased risk the same data source are required to validate the findings from of vessel injury and distal embolization, which could negatively these analyses. Additionally, studies are required to determine impact clinical outcomes.9 26 Additionally, more passes increase the factors and techniques/methods that maximize the likelihood the risk of losing embolic fragments into previously unaffected of achieving FPE. territories. This hypothesis is supported by the results of this Author affiliations study, which showed that the proportion of patients who had 1 Department of Neuroscience, Mercy Saint Vincent Medical Center, Toledo, Ohio, USA embolization into a new territory was significantly lower in 2 Stroke Unit, Neurology, Hospital Vall d’Hebron, Barcelona, Spain the FPE group than in the non-­FPE group (2.20% vs 11.11%, 3 Departament de Medicina, Universitat Autònoma de Barcelona, Bellaterra, Spain 4 p=0.03). Both distal embolization and embolization into new 5 Department of Neurology, Inselspital, University of Bern, Bern, Switzerland vascular territories damage brain tissue directly and can also Department of Neurology, UCLA, Los Angeles, California, USA 6 Oregon Health and Science University, Portland, Oregon, USA cut off leptomeningeal collaterals and thus accelerate dying of 7 Department of Neurointervention, Texas Stroke Institute, Plano, Texas, USA penumbral tissue. 8 Johnson & Johnson Medical GmbH, Norderstedt, Germany This study has several strengths, including the use of patient-­ 9 Cerenovus, Johnson & Johnson, Irvine, California, USA 10 level data from a multi-­country study to inform the health EVERSANA, Burlington, Ontario, Canada 11 economic analyses, inclusion of several countries in the health Departments of Radiology and Neurology, AZ Groeninge, Kortrijk, Belgium 12 Departments of Neuroradiology and Clinical Neuroscience, Karolinska University economic analyses for increased generalizability, and differ- Hospital, Karolinska Institutet, Stockholm, Sweden entiation between acute and annual care costs. Furthermore, as compared with previous literature,9 the target population Twitter Marc Ribo @marcriboj for this study excluded patients for whom mTICI 2c–3 was Contributors All authors made substantial contributions to the conception or not achieved to obtain a homogeneous patient cohort and to design of the work; or the acquisition, analysis, or interpretation of data for the conduct an unbiased assessment of FPE. However, the find- work. Additionally, all authors were involved in drafting the work or revising it ings of this study should be interpreted in the context of the critically for important intellectual content and final approval of the version to be following limitations. First, there were limitations in relation published. All authors agree to be accountable for all aspects of the work. to the procedural/hospitalization-­ related and annual care Funding This study was funded by Cerenovus, a subsidiary of Johnson and Johnson. cost data used in this analysis, which were obtained from Competing interests OOZ serves as a consultant for Neuravi, Stryker, Penumbra, the literature. There were differences in resources captured and Medtronic. MR is a shareholder in Anaconda Biomed; consultant for Neuravi/ in costs across countries and in the patient populations that Cerenovus, Medtronic, Stryker, Apta Targets, and Vesalio. HPM reports personal fees from Covidien/Medtronic, Neuravi/Cerenovus, Servier, and Bayer outside the were assessed in the cost literature and in the ARISE II study, submitted work; served on the steering committees of the SWIFT PRIME and ARISE respectively.16 23 25 However, this study used the best available studies. JLS is an employee of the University of California, which holds a patent on cost estimates, which were validated by clinical experts in retriever devices for stroke. The University of California, Regents receives funding interviews. Second, it is plausible that resource use may vary for the services of JLS as a scientific consultant regarding trial design and conduct to Covidien/Medtronic and Stryker; serves as a consultant for Modest, Abbott, across countries owing to variations in clinical practices. Sensi- Medtronic, Stryker, and Neuravi/Cerenovus; has contracted stock options for Modest tivity analyses assessing differences by region were not feasible and Rapid Medical. HB serves as a modest consultant for Neuravi/Cerenovus, due to small sample sizes. Third, procedural/hospitalization-­ and Stryker. AJY is a consultant for Cerenovus, Penumbra, Genentech, and Zoll related and annual care costs were not comprehensive. Anal- Circulation; receives research grant support from Cerenovus, Penumbra, Medtronic, yses assessing the procedural/hospitalization-­related economic Stryker, and Genentech; and has equity interest in Insera Therapeutics. AE is an employee of Johnson and Johnson. EK is an employee of Cerenovus, a subsidiary impact did not include costs for procedure time and inter- of Johnson and Johnson. HLC and RAQ are paid consultants for Cerenovus. TA mediate catheters (lower use in the FPE group), and analyses is a consultant for Neuravi/Cerenovus, Anaconda, Amnis Therapeutics, and Rapid assessing the annual care economic impact did not include Medical; served on the steering committees of the ARISE studies. 1122 Zaidat OO, et al. J NeuroIntervent Surg 2021;13:1117–1123. doi:10.1136/neurintsurg-2020-016930 Ischemic stroke Patient consent for publication Not required. 13 Gershengorn HB, Garland A, Gong MN. Patterns of daily costs differ for medical and surgical intensive care unit patients. Ann Am Thorac Soc 2015;12:1831–6. Ethics approval Research approval was not obtained as this study is a post-­hoc 14 Chen BP, Clymer JW, Turner AP, et al. Global hospital and operative costs associated analysis of an existing database/study (ARISE II). with various ventral cavity procedures: a comprehensive literature review and analysis Provenance and peer review Not commissioned; externally peer reviewed. across regions. J Med Econ 2019;22:1210–20. Data availability statement No data are available. 15 Dodel RC, Haacke C, Zamzow K, et al. Resource utilization and costs of stroke unit care in Germany. Value Health 2004;7:144–52. Supplemental material This content has been supplied by the author(s). It 16 Fattore G, Torbica A, Susi A, et al. The social and economic burden of stroke survivors has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have in Italy: a prospective, incidence-­based, multi-­centre cost of illness study. BMC Neurol been peer-­reviewed. Any opinions or recommendations discussed are solely those 2012;12:137. of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and 17 Vega Valdés P, Murias Quintana E, Morales Santos E, et al. Análisis de los costes responsibility arising from any reliance placed on the content. Where the content directos asociados a la trombectomía mecánica ya la fibrinólisis intravenosa en el includes any translated material, BMJ does not warrant the accuracy and reliability Hospital Universitario central de asturias. Revista de Neurología 2018;66:7–14. of the translations (including but not limited to local regulations, clinical guidelines, 18 Steen Carlsson K, Andsberg G, Petersson J, et al. Long-­term cost-­effectiveness of terminology, drug names and drug dosages), and is not responsible for any error thrombectomy for acute ischaemic stroke in real life: an analysis based on data from and/or omissions arising from translation and adaptation or otherwise. the Swedish stroke register (Riksstroke). Int J Stroke 2017;12:802–14. Open access This is an open access article distributed in accordance with the 19 Xu X-­M, Vestesson E, Paley L, et al. The economic burden of stroke care in England, Creative Commons Attribution Non Commercial (CC BY-­NC 4.0) license, which Wales and Northern Ireland: using a national stroke register to estimate and report permits others to distribute, remix, adapt, build upon this work non-­commercially, patient-­level health economic outcomes in stroke. Eur Stroke J 2018;3:82–91. and license their derivative works on different terms, provided the original work is 20 Decision Resource Group (DRG). Clot management devices: market insights, 2018. properly cited, appropriate credit is given, any changes made indicated, and the use 21 Vidal GA, Milburn JM. The penumbra 5MAX ACE catheter is safe, efficient, and cost is non-­commercial. See: http://​creativecommons.​org/​licenses/​by-​nc/​4.​0/. saving as a primary mechanical thrombectomy device for large vessel occlusions in acute ischemic stroke. Ochsner J 2016;16:486–91. ORCID iDs 22 Shireman TI, Wang K, Saver JL, et al. Cost-­effectiveness of Solitaire stent retriever Osama O Zaidat http://​orcid.​org/​0000-​0003-​4881-​4698 thrombectomy for acute ischemic stroke: results from the SWIFT-­PRIME trial (solitaire Marc Ribo http://​orcid.​org/​0000-​0001-​9242-​043X with the intention for thrombectomy as primary endovascular treatment for acute ischemic stroke). Stroke 2017;48:379–87. 23 Spieler J-­F, Lanoë J-­L, Amarenco P. Costs of stroke care according to handicap levels REFERENCES and stroke subtypes. Cerebrovasc Dis 2004;17:134–42. 1 GBD 2016 Stroke Collaborators. Global, regional, and national burden of stroke, 24 Dental and Pharmaceutical Benefits Agency (TLV). Cost-­effectiveness analysis of 1990-­2016: a systematic analysis for the global burden of disease study 2016. Lancet thrombectomy for treatment of acute severe ischemic stroke; 2015. Neurol 2019;18:439–58. 25 Lobotesis K, Veltkamp R, Carpenter IH, et al. Cost-­effectiveness of stent-­retriever 2 Benjamin EJ, Blaha MJ, Chiuve SE, et al. Heart disease and stroke statistics-­2017 thrombectomy in combination with IV t-­PA compared with IV t-­PA alone for acute update: a report from the American Heart Association. Circulation ischemic stroke in the UK. J Med Econ 2016;19:785–94. 2017;135:e146–603. 26 Nikoubashman O, Dekeyzer S, Riabikin A, et al. True first-­pass effect. Stroke 3 Wilkins E, Wilson L, Wickramasinghe K, et al. European cardiovascular disease 2019;50:2140–6. statistics 2017, 2017. 27 Ducroux C, Piotin M, Gory B, et al. First pass effect with contact aspiration and stent 4 Centers for Disease Control and Prevention (CDC). Stroke facts, 2020. retrievers in the aspiration versus stent retriever (ASTER) trial. J Neurointerv Surg 5 Rennert RC, Wali AR, Steinberg JA, et al. Epidemiology, natural history, and clinical 2020;12:386–91. presentation of large vessel ischemic stroke. Neurosurgery 2019;85:S4–8. 28 Dawson J, Lees JS, Chang T-­P, et al. Association between disability measures and 6 Rai AT, Evans K, Riggs JE, et al. Intravenous thrombolysis of large vessel occlusions healthcare costs after initial treatment for acute stroke. Stroke 2007;38:1893–8. is associated with higher hospital costs than small vessel strokes: a rationale for 29 Kavanagh S, Knapp M, Patel A. Costs and disability among stroke patients. J Public developing stroke severity-­based financial models. J Neurointerv Surg 2016;8:423–8. Health Med 1999;21:385–94. 7 Goyal M, Jadhav AP, Bonafe A, et al. Analysis of workflow and time to treatment and 30 Luengo-­Fernandez R, Yiin GSC, Gray AM, et al. Population-­based study of acute- and the effects on outcome in endovascular treatment of acute ischemic stroke: results long-­term care costs after stroke in patients with AF. Int J Stroke 2013;8:308–14. from the SWIFT PRIME randomized controlled trial. Radiology 2016;279:888–97. 31 Kunz WG, Almekhlafi MA, Menon BK, et al. Public health and cost benefits of 8 Shi Z-­S, Liebeskind DS, Xiang B, et al. Predictors of functional dependence successful reperfusion after thrombectomy for stroke. Stroke 2020;51:899–907. despite successful revascularization in large-­vessel occlusion strokes. Stroke 32 García-­Tornel Álvaro, Requena M, Rubiera M, et al. When to stop. Stroke 2014;45:1977–84. 2019;50:1781–8. 9 Zaidat OO, Castonguay AC, Linfante I, et al. First pass effect: a new measure for stroke 33 Jindal G, Carvalho HDP, Wessell A, et al. Beyond the first pass: revascularization thrombectomy devices. Stroke 2018;49:660–6. remains critical in stroke thrombectomy. J Neurointerv Surg 2019;11:1095–9. 10 Khunte M, Wu X, Payabvash S, et al. Cost-­effectiveness of endovascular thrombectomy 34 Maegerlein C, Berndt MT, Mönch S, et al. Further development of combined in patients with acute stroke and M2 occlusion. J Neurointerv Surg 2020. techniques using stent retrievers, aspiration catheters and BGC : The PROTECTPLUS doi:10.1136/neurintsurg-2020-016765. [Epub ahead of print: 19 Oct 2020]. Technique. Clin Neuroradiol 2020;30:59–65. 11 Zaidat OO, Bozorgchami H, Ribó M, et al. Primary results of the multicenter ARISE 35 Delgado Almandoz JE, Kayan Y, Young ML, et al. Comparison of clinical outcomes in II study (Analysis of Revascularization in Ischemic Stroke with EmboTrap). Stroke patients with acute ischemic strokes treated with mechanical thrombectomy using 2018;49:1107–15. either Solumbra or ADAPT techniques. J Neurointerv Surg 2016;8:1123–8. 12 Husereau D, Drummond M, Petrou S, et al. Consolidated Health Economic Evaluation 36 Maus V, Behme D, Kabbasch C, et al. Maximizing first-­pass complete reperfusion with Reporting Standards (CHEERS) statement. Cost Eff Resour Alloc 2013;11:6. save. Clin Neuroradiol 2018;28:327–38. Zaidat OO, et al. J NeuroIntervent Surg 2021;13:1117–1123. doi:10.1136/neurintsurg-2020-016930 1123