Thrombolysis in Patients With Mild Ischemic Stroke: A Comparative Effectiveness Study From the Norwegian Stroke Registry

ABSTRACT

Background

Intravenous thrombolytic therapy (IVT) is recommended for patients with acute ischemic stroke (AIS) and disabling deficits, regardless of National Institutes of Health Stroke Scale (NIHSS) score. The benefit of IVT in patients with mild AIS was, however, not documented in recent trials. The aim of this study was to evaluate the effectiveness of IVT in mild AIS in a nationwide registry‐based study.

Methods

AIS patients from the Norwegian Stroke Registry (2016–2022) with a National Institutes of Health Stroke Scale (NIHSS) score $\le$ 5 at admission were included. Using 1:1 propensity score matching, 1736 IVT‐treated patients were matched with 1736 controls. Primary outcome was excellent functional outcome (modified Rankin Scale [mRS] score 0–1) at 90 days. Secondary outcomes included mRS 0–2, a favorable mRS shift (range 0–6), and death within 28 and 90 days. Data on symptomatic intracerebral hemorrhage (sICH) were available only for the IVT group.

Results

At 90 days, 68.4% of IVT‐treated patients and 61.2% of controls achieved mRS 0–1 (OR 1.39; 95% CI 1.20–1.60). IVT was associated with a higher probability of achieving excellent functional outcome in patients with NIHSS 4–5 (OR 1.53; 95% CI 1.06–2.21) and NIHSS 2–3 (OR 1.64; 95% CI 1.32–2.03), but not NIHSS 0–1 (OR 1.14; 95% CI 0.89–1.44). No significant differences in overall mortality were observed. SICH occurred in 3.7% of IVT‐treated patients.

Conclusion

IVT was associated with favorable functional outcomes for mild AIS, but not in patients with NIHSS 0–1.

In this longitudinal study from the Norwegian Stroke Registry (2016–2022), the effectiveness of intravenous thrombolysis (IVT) was compared in 1736 IVT‐treated patients and 1736 not IVT‐treated patients with mild acute ischemic stroke by propensity score matching. Mild stroke was defined as National Institutes of Health Stroke Scale (NIHSS) score 0–5, and functional outcome at 3 months was assessed by the modified Rankin Scale (mRS). IVT was associated with a favorable functional outcome in patients with NIHSS 2–5.

1. Introduction

Mild or minor acute ischemic strokes (AIS), often defined as a score of 0–5 points on the National Institutes of Health Stroke Scale (NIHSS) [1], are common and generally account for most AIS cases [2, 3, 4]. Mild AIS can be further categorized as disabling or non‐disabling, based on the severity of symptoms at presentation [5, 6]. However, a standardized definition of the term has yet to be established [6, 7]. Despite a high incidence, the effect of intravenous thrombolytic therapy (IVT) for mild AIS has been less extensively studied than for AIS in general [4, 8, 9, 10]. Early randomized trials of IVT for AIS mostly excluded patients presenting with mild AIS, and in trials that allowed enrolment, the number of participants with mild AIS was relatively low [11, 12, 13, 14, 15, 16]. Clinical equipoise regarding the proper course of action has resulted in different treatment practices over time [17, 18, 19]. Recently updated guidelines advise against the use of IVT in patients with mild non‐disabling deficits, while IVT is recommended in AIS patients with disabling deficits, regardless of NIHSS score [10]. Results from recent trials specifically targeting patients with minor stroke do not support the use of IVT in these patients with or without disabling deficits [5, 20, 21, 22, 23].

While randomized trials remain the reference standard for assessing efficacy, real‐world observational data can be a valuable complement. This comparative effectiveness study aims to assess the effectiveness and safety of IVT for patients with mild AIS in a real‐world clinical setting, employing propensity score‐matched data from the Norwegian Stroke Registry.

2. Patients and Methods

Data were obtained from the Norwegian Stroke Registry, a nationwide quality registry that covers all hospitalizations for acute stroke in Norway [24]. Inclusion in the Norwegian Stroke Registry is mandatory, thus not based on individual patient consent. The registry's data is deemed to be adequately complete and correct for research purposes [25, 26].

This study was approved by the Regional Committee for Medical and Health Research Ethics in North Norway (REK Nord 504,501) and by the Data Protection Officer at UiT The Arctic University of Norway. It is reported in accordance with the Strengthening the Reporting of Observational studies in Epidemiology (STROBE) statement [27].

2.1. Study Subjects

Patients included in the Norwegian Stroke Registry with a diagnosis of AIS (ICD‐10 code I63) and an NIHSS score $\le$ 5 at admission, and hospitalized between January 1, 2016 and December 31, 2022 (N = 28,677) were eligible for study inclusion. We excluded patients with onset‐to‐admission time > 4.5 h (n = 11,950), patients who used anticoagulant medication at stroke onset (n = 1068), and patients with wake‐up strokes or otherwise unknown time of onset (n = 6574) as they were not eligible for thrombolysis according to Norwegian guidelines in effect during the observation period. We also excluded patients with missing information on IVT (n = 5), and patients who received thrombectomy (n = 199) or had unknown thrombectomy status (n = 17). Furthermore, we excluded patients with missing information on mRS score prior to stroke onset (n = 654), a pre‐stroke mRS score $\ge$ 3 (n = 732), missing mRS score at three‐month follow‐up (n = 1278), or with late follow‐up > 6 months after hospital admission (n = 294) (Figure 1).

FIGURE 1.

Flow chart of study participation. IVT, Intravenous thrombolytic therapy; mRS, Modified Rankin Scale.

Propensity score matching was used to define a cohort of 1736 pairs (in total, 3472 patients), matching individuals who received IVT with individuals who did not receive IVT. Details on the matching procedure are described in the statistical section below.

2.2. Data Collection

Information on demographics, known cardiovascular risk factors prior to stroke onset, contraindications to IVT, NIHSS score and mRS score at admission, mRS score at three‐month follow‐up, onset‐to‐admission time (minutes), door‐to‐needle time (minutes from admission to IVT initiation), onset‐to‐treatment time (minutes from onset to IVT initiation), and hospital stroke volume were obtained from the registry. Hospital stroke volume was defined by the average number of stroke cases admitted per year and grouped as high (> 300 cases/year), medium (100–300 cases/year), and low (< 100 cases/year) volume centers.

Continuous surveillance of data quality is an integrated and routine part of the Norwegian Stroke Registry's work. Several validation rules have been implemented in the reporting system to reduce the risk of logical errors. In addition, the registry secretariat regularly performs quality controls of reported data using comprehensive scripts. Records containing errors or inconsistencies are returned to the reporting hospital for correction. The secretariat also conducts audits comparing reported data with patient medical records [25, 26, 28].

2.3. Outcome Measurements

The Norwegian Stroke registry collects information on functional outcome after 90 days of follow‐up. The information is obtained by telephone interview of the patient (or carers when appropriate), letter, or in‐person interview in outpatient clinical settings. Vital status information is obtained through linkage to the national Norwegian Causes of Death Registry.

The primary outcome was an excellent functional outcome at three‐month follow‐up, defined as a score of 0–1 point on the modified Rankin Scale (mRS). Secondary outcomes were good functional outcome at three‐month follow‐up, defined as an mRS score of 0–2, and a favorable shift across the full mRS range (scores 0–6). Safety outcomes were death within 28 days and 90 days after stroke onset. Information on symptomatic intracerebral hemorrhage (sICH) within 36 h after treatment was only available in patients receiving IVT. An ICH was considered symptomatic if occurring in conjunction with neurological deterioration (operationalized by a 4‐point increase in the NIHSS score from the patient's baseline value or previous lowest measurement) or causing death and documented by CT imaging or autopsy.

2.4. Statistical Analysis

Participants were divided into one treatment group (receiving IVT) and one control group (not receiving IVT). We used a four‐step approach to propensity score matching (PSM) [29]. Individual probability of receiving IVT in all eligible participants was assessed with logistic regression. Propensity scores were estimated using a generalized boosted model with Average Treatment Effect on the Treated (ATT) as the estimand [30]. Balance metrics from the “twang” package were employed as a stopping criterion for matching, utilizing both mean and maximum values of the Effect Size (ES), and Kolmogorov–Smirnov (KS) statistics [31]. The number of trees was set to 8000, interaction depth = 3, and shrinkage = 0.01. Matching was done by applying the MAny‐Group‐MAtching‐algorithm (MAGMA) in R, using 1:1 nearest neighbor matching [32]. We selected an optimal matched cohort size according to comprehensive calculations of Pillai's Trace value, mean effect, Cohen's d (< 0.20), and the adjusted d‐ratio across all potential cohort sizes. A thorough propensity score analysis indicated a cohort size of 1736 individuals per group as the optimal choice, based on the best adjusted d‐ratio. This resulted in a Pillai's trace value of 0.00, a mean g of 0.02, a d‐ratio of 1.00, and an adjusted d‐ratio of 1.00.

The following variables were significantly associated with whether an individual received IVT (p < 0.05) and were included in the propensity score estimation: Age, NIHSS score at admission, baseline mRS score, onset‐to‐admission time, hospital stroke volume, prior transient ischemic attack (TIA), living situation at stroke onset, and marital status at stroke onset. Sex was also included, despite not having a statistically significant influence on the propensity score estimation.

Baseline characteristics of matched and unmatched cases were presented as medians (interquartile ranges) for continuous variables, and numbers (percentages) for categorical variables, with standardized mean differences for comparison of groups before and after matching.

Logistic regression analyses were used to assess the effect of the IVT on the primary outcome after 90‐day follow‐up (excellent functional outcome; mRS scores 0–1), the secondary outcome (good functional outcome; mRS 0–2), and mortality within 28 and 90 days. Ordinal logistic regression was applied to assess shifts across the full mRS range (0–6), utilizing the Brant test to validate the proportional odds assumption. All regression analyses were adjusted for age and sex. Additional sensitivity analyses were done with adjustment for atrial fibrillation and for diabetes, as these showed baseline imbalances.

Interaction by NIHSS on the primary outcome was tested by adding an interaction term (product of IVT‐status and NIHSS) to the model. Subgroup analyses of functional outcome and safety outcomes were performed by stratifying patients according to baseline NIHSS scores 0–1, 2–3, and 4–5.

p‐values less than 0.05 were considered statistically significant. No formal correction for multiple testing was applied, as all analyses beyond the primary analysis were considered exploratory.

Calculations were performed in R (version 4.3.2) through RStudio (version 2023.06.2), with packages “MAGMA.R” (version 1.0.1) and “twang” (version 2.6.0) for propensity score matching. ChatGPT (GPT‐4 and GPT‐4o, OpenAI), an AI‐based language model, was used to troubleshoot parts of code for statistical programming when issues occurred and to assist in generating preliminary code for Figures 2 and 3. ChatGPT was not provided data access, and the code fed to the model was censored with generic names.

FIGURE 2.

Modified Rankin Scale scores at three‐month follow‐up in the propensity score matched cohorts of patients treated with or without IVT. IVT, Intravenous thrombolytic therapy; mRS, Modified Rankin Scale.

FIGURE 3.

Subgroup analysis of effectiveness (A) and safety outcomes (B), stratified by baseline NIHSS scores, adjusted for age and sex. IVT, Intravenous thrombolytic therapy; OR, Odds ratio; CI, Confidence interval; NIHSS, National Institutes of Health Stroke Scale.

3. Results

Within the unmatched cohorts, patients receiving IVT were younger, more likely to be male or married, and less likely to live alone or require assisted living compared to controls. Patients who received IVT had lower proportions of atrial fibrillation, diabetes mellitus, prior TIA, and were less likely to use antihypertensive or lipid‐modifying medications. Furthermore, they had lower median pre‐stroke mRS scores and higher median NIHSS scores at admission (Table 1).

TABLE 1.

Baseline characteristics in matched and unmatched cohorts.

	Unmatched Cohorts			Matched Cohorts
	IVT, n = 2961	No IVT, n = 2945	SMD	IVT, n = 1736	No IVT, n = 1736	SMD
Age in years, median (IQR)	71 (62–78)	73 (63–80)	0.13	73 (63–79)	73 (63–80)	0.02
Female sex, n (%)	1159 (39.1)	1236 (42.0)	0.06	704 (40.6)	697 (40.2)	0.01
Marital status
Married, n (%)	2077 (70.1)	1969 (66.9)	0.07	1182 (68.1)	1180 (68.0)	0.00
Widowed, n (%)	386 (13.0)	420 (14.3)	0.04	254 (14.6)	237 (13.7)	0.03
Unmarried, n (%)	448 (15.1)	499 (16.9)	0.05	268 (15.4)	281 (16.2)	0.02
Living situation
Alone, n (%)	869 (29.3)	957 (32.5)	0.07	546 (31.5)	528 (30.4)	0.02
Not alone, n (%)	2059 (69.5)	1958 (66.5)	0.07	1168 (67.3)	1183 (68.1)	0.02
Institution, n (%)	5 (0.2)	6 (0.2)	0.01	4 (0.2)	5 (0.3)	0.01
Living conditions
Independent, no assistance, n (%)	2779 (93.9)	2690 (91.3)	0.09	1614 (93.0)	1587 (91.4)	0.06
Independent, assisted, n (%)	163 (5.5)	215 (7.3)	0.07	109 (6.3)	121 (7.0)	0.03
Assisted living facility, n (%)	13 (0.4)	25 (0.8)	0.05	8 (0.5)	18 (1.0)	0.07
Nursing home, n (%)	5 (0.2)	12 (0.4)	0.05	4 (0.2)	8 (0.5)	0.04
Smoking
No, n (%)	1203 (40.6)	1219 (41.4)	0.02	728 (41.9)	693 (39.9)	0.07
Yes, n (%)	510 (17.2)	530 (18.0)	0.00	282 (16.2)	306 (17.6)	0.03
Former (> 1 month), n (%)	904 (30.5)	950 (32.3)	0.01	526 (30.3)	575 (33.1)	0.05
Atrial fibrillation, n (%)	348 (11.8)	396 (13.4)	0.05	206 (11.9)	226 (13.0)	0.04
Prior myocardial infarction, n (%)	336 (11.3)	329 (11.2)	0.01	195 (11.2)	211 (12.2)	0.03
Prior TIA, n (%)	251 (8.5)	336 (11.4)	0.10	160 (9.2)	172 (9.9)	0.02
Diabetes mellitus, n (%)	407 (13.7)	470 (16.0)	0.06	262 (15.1)	293 (16.9)	0.05
Use of medication
Antihypertensive drugs, n (%)	1418 (47.9)	1526 (51.8)	0.08	857 (49.4)	906 (52.2)	0.05
Antiplatelet drugs, n (%)	988 (33.4)	1001 (34.0)	0.01	600 (34.6)	597 (34.4)	0.00
Lipid‐modifying drugs, n (%)	1011 (34.1)	1048 (35.6)	0.03	600 (34.6)	635 (36.6)	0.04
Pre‐stroke mRS score, median (IQR)	0 (0–0)	0 (0–1)	0.17	0 (0–1)	0 (0–1)	0.04
0, n (%)	2286 (77.2)	2075 (70.5)	0.15	1256 (72.4)	1225 (70.6)	0.04
1, n (%)	469 (15.8)	554 (18.8)	0.08	315 (18.2)	333 (19.2)	0.03
2, n (%)	206 (7.0)	316 (10.7)	0.13	165 (9.5)	178 (10.3)	0.03
NIHSS score at admission, median (IQR)	3 (2–4)	1 (0–2)	0.92	2 (1–3)	2 (1–3)	0.01
0, n (%)	215 (7.3)	1077 (26.2)	0.76	215 (12.4)	201 (11.6)	0.02
1, n (%)	504 (17.0)	771 (26.3)	0.22	501 (28.7)	557 (32.0)	0.08
2, n (%)	712 (24.1)	520 (17.7)	0.16	481 (27.7)	448 (25.8)	0.05
3, n (%)	598 (20.2)	294 (10.0)	0.29	300 (17.3)	270 (15.6)	0.05
4, n (%)	530 (17.9)	173 (5.9)	0.38	141 (8.1)	162 (9.3)	0.04
5, n (%)	402 (13.6)	110 (3.7)	0.36	98 (5.7)	98 (5.7)	0.00
Onset‐to‐admission time in minutes, median (IQR)	78 (50–125)	105 (60–170)	0.02	88 (55–148)	90 (50–150)	0.02
Door‐to‐needle time in minutes, median (IQR)	29 (19–45)	—	—	30 (20–48)	—	—
Hospital stroke volume
> 300 cases/year, n (%)	554 (18.7)	619 (21.0)	0.06	332 (19.1)	360 (20.7)	0.04
100–300 cases/year, n (%)	1527 (51.6)	1185 (40.2)	0.23	803 (46.3)	761 (43.8)	0.05
< 100 cases/year, n (%)	880 (29.7)	1141 (38.7)	0.19	601 (34.6)	615 (35.4)	0.02

Note: Age, pre‐stroke mRS score, NIHSS score at admission, onset‐to‐admission time, and door‐to‐needle time are presented as medians (interquartile ranges).

Abbreviations: IVT, Intravenous thrombolytic therapy; SMD, Standardized mean difference; TIA, Transient ischemic attack; mRS, Modified Rankin Scale; NIHSS, National Institutes of Health Stroke Scale.

In the overall matched study cohort (n = 3472), median age was 72 years (interquartile range [IQR] 63–79), and females comprised 40.4% (n = 1401). There were no substantial differences in baseline characteristics between the two matched groups (Table 1).

Excellent functional outcome (mRS scores 0–1) was achieved by 68.4% (n = 1188) of patients in the matched IVT group at three‐month follow‐up, compared to 61.2% (n = 1062) in the control group, with an odds ratio (OR) of 1.39 (95% confidence interval [CI] 1.20–1.60; p < 0.001). The effect estimate was similar after additional adjustment for atrial fibrillation and diabetes (OR 1.38; 95% CI 1.19–1.59). Similarly, IVT was associated with a higher likelihood of achieving a good (mRS scores 0–2) functional outcome (OR 1.57; 95% CI 1.29–1.91; p < 0.001). The ordinal regression analysis demonstrated an overall favorable shift towards lower mRS scores in the IVT group (OR 1.41; 95% CI 1.25–1.59; p < 0.001) (Table 2, Figure 2).

TABLE 2.

Effectiveness and safety of IVT vs. no IVT in propensity score matched cohorts of patients with mild ischemic stroke, adjusted for age and sex.

	IVT, n = 1736 n (%)	No IVT, n = 1736 n (%)	Odds ratio (95% CI)	p‐value
Effectiveness outcomes
Excellent functional outcome	1188 (68.4)	1062 (61.2)	1.39 (1.20–1.60)	< 0.001
Good functional outcome	1543 (88.9)	1451 (83.6)	1.57 (1.29–1.91)	< 0.001
Ordinal analysis	1736 (100.0)	1736 (100.0)	1.41 (1.25–1.59)	< 0.001
Safety outcomes
Death within 28 days	32 (1.8)	34 (2.0)	0.96 (0.59–1.57)	0.87
Death within 90 days	61 (3.5)	75 (4.3)	0.82 (0.58–1.17)	0.28
Symptomatic ICH	64 (3.7)	—	—	—

Abbreviations: CI, Confidence interval; ICH, Intracerebral hemorrhage; IVT, Intravenous thrombolytic therapy.

NIHSS score at admission significantly modified the association between IVT and functional outcome (p for interaction = 0.03). Subgroup analyses showed no significant association between IVT and achieving an excellent functional outcome (OR 1.14; 95% CI 0.89–1.44; p = 0.30) or a good functional outcome (OR 1.12; 95% CI 0.74–1.68; p = 0.60) among patients with NIHSS scores 0–1 at admission. In patients with baseline NIHSS scores 2–3, IVT was associated with increased odds of achieving both excellent (OR 1.64; 95% CI 1.32–2.03; p < 0.001) and good (OR 1.63; 95% CI 1.21–2.20; p = 0.001) functional outcomes. Similarly, among patients with baseline NIHSS 4–5, IVT was associated with an OR of 1.53 (95% CI 1.06–2.21; p = 0.02) for excellent and 2.40 (95% CI 1.58–3.68; p < 0.001) for good functional outcomes (Figure 3).

Overall, there were no significant differences in case fatality between the two groups, neither at 28 days (IVT group: 1.8%, n = 32; controls: 2.0%, n = 34; OR 0.96; 95% CI 0.59–1.57; p = 0.87), nor at 90 days (IVT group: 3.5%, n = 61; controls: 4.3%, n = 75; OR 0.82; 95% CI 0.58–1.17; p = 0.28) (Table 2). Subgroup analyses showed a significant association between IVT and decreased mortality at 28 days (OR 0.36; 95% CI 0.12–0.90; p < 0.05) and 90 days (OR 0.50; 95% CI 0.25–0.94; p < 0.05) in patients with NIHSS 4–5 at admission. For patients with baseline NIHSS 0–1 and 2–3, there were no significant associations between IVT and mortality at either time point (Figure 3).

Symptomatic intracerebral hemorrhage (sICH) was observed in 3.7% (n = 64) of patients within 36 h of initiating IVT (Table 2). Subgroup analyses showed that sICH occurred in 3.0% (n = 22/716) of patients with baseline NIHSS 0–1, 4.1% (n = 32/781) with NIHSS 2–3, and 4.2% (n = 10/239) with NIHSS 4–5.

IVT‐treated patients who were excluded due to loss to follow‐up or delayed follow‐up beyond six months did not differ substantially from the matched IVT‐treated patients, except for having a higher median NIHSS score at admission. Median NIHSS score in IVT‐treated excluded patients was 3 (IQR 2–4), compared with 2 (IQR 1–3) in the matched IVT group (Supplemental Table S1).

4. Discussion

We observed a significant association between IVT and favorable functional outcomes at three‐month follow‐up (mRS scores 0–1, 0–2, and across the full mRS range) in patients with mild AIS. When stratifying patients by baseline NIHSS scores, we found no significant association between IVT and favorable functional outcomes in patients with NIHSS 0–1 at admission. Additionally, there were no significant associations between IVT and mortality at 28 or 90 days in the total study cohort, nor in those with NIHSS 0–3. However, a significantly lower 28‐day and 90‐day mortality was observed in IVT‐treated patients with baseline NIHSS 4–5.

Several studies have reported a positive association between IVT and favorable functional outcomes in patients with mild AIS. Our findings align with a previous meta‐analysis that included data from 6756 individual patients across nine randomized controlled trials, of whom 666 had baseline NIHSS 0–4. Subgroup analysis showed a significant association between IVT and an excellent functional outcome at three to six months post‐stroke in patients with NIHSS 0–4. The benefit was most pronounced when IVT was administered within 3 h of symptom onset (OR 1.75; 95% CI 1.35–2.27; p < 0.001), but remained significant when administered between 3 and 4.5 h (OR 1.26; 95% CI 1.05–1.51; p = 0.01) [33]. A subsequent meta‐analysis of ten studies (including one RCT, two RCT post hoc analysis studies, three prospective studies, and four retrospective studies) showed a positive association between IVT and a good functional outcome in 4333 patients with AIS and baseline NIHSS scores $\le$ 5 [34].

Previous registry‐based studies have also supported these findings. A retrospective study from Korea of 368 propensity score‐matched AIS patients with baseline NIHSS score $\le$ 5 and treated within 4.5 h of symptom onset showed that a significantly higher proportion of patients in the IVT group achieved excellent functional outcome at three months compared to controls [35]. A prospective study from the Austrian Stroke Unit Registry that included 445 matched pairs with NIHSS $\le$ 5 showed a significant reduction in mRS scores at three months in IVT‐receiving patients compared to controls [36]. Further stratification by baseline NIHSS scores showed a positive association between IVT and a favorable functional outcome in patients with baseline NIHSS 4–5, while no significant benefit was demonstrated for patients with NIHSS scores $\le$ 3 [36]. This was also the case in a registry‐based study from China that included 1163 propensity score‐matched patients [20]. Furthermore, studies from the Austrian Stroke Unit Registry found no benefit and potential harm associated with IVT for patients with NIHSS scores 0–1 at baseline [36, 37]. Similarly, we observed no significant associations between IVT and a favorable functional outcome in patients with baseline NIHSS 0–1. Overall, these findings suggest that the commonly used NIHSS score range of 0–5 for mild AIS may be too broad to be defined as a homogeneous group.

Results from more recent randomized controlled trials have failed to demonstrate a clear benefit of IVT in patients with mild AIS. In the ARAMIS trial (2018–2022), dual antiplatelet therapy (DAPT) was non‐inferior to IVT in patients with non‐disabling AIS and baseline NIHSS scores 0–5. The study included 350 patients in the IVT group and 369 patients in the DAPT group [20]. Similarly, the PRISMS trial (2014–2017) failed to demonstrate a significant benefit of IVT compared to aspirin in 313 patients with mild AIS (baseline NIHSS $\le$ 5) and whose deficits, if they were to persist, were deemed not to interfere with activities of daily living or prevent return to work, as judged by local clinicians in consultation with patients and available family. The trial was terminated early due to slow recruitment, thereby precluding any definitive conclusions [5]. The TEMPO‐2 trial assessed the effect of tenecteplase in 884 patients enrolled within 12 h of symptom onset with NIHSS 0–5 and intracranial arterial occlusion or focal perfusion abnormality [21]. IVT showed no significant benefit, but possible harm with increased risk of death and sICH in the IVT group compared to controls. An exploratory secondary analysis of the TEMPO‐2 trial found no significant treatment effect of tenecteplase in either the 100 patients (11.3%) with disabling deficits or the 784 with non‐disabling deficits, and disabling status did not modify the neutral treatment effect [38]. The definition of “disabling” was based on evaluation of NIHSS items. Similarly, no significant benefit of IVT was shown in trials with urokinase [22] or prourokinase [23] in patients with NIHSS $\le$ 5 given within 6 h of symptom onset. Furthermore, a recent meta‐analysis of three RCTs and seventeen observational studies found no significant association between IVT and an excellent functional outcome at three months in patients with mild AIS (baseline NIHSS scores $\le$ 5) [39].

Following the negative results from recent RCTs, the recently updated American Heart Association/American Stroke Association guideline for early management of acute ischemic stroke advises against use of IVT for patients with mild non‐disabling stroke, while IVT is recommended in patients with disabling stroke, regardless of NIHSS score [10]. The recommendations for disabling AIS are based on findings from a meta‐analysis of earlier trials [33]. The guideline further points to the significant knowledge gap with regard to the lack of a standardized definition for minor non‐disabling vs. disabling stroke deficits. Available data from the Norwegian Stroke registry were insufficient for applying the criteria used in PRISMS or TEMPO‐2 in the present study.

The case fatality analysis revealed no significant difference in mortality at 28 days or 90 days between the IVT and control groups, consistent with a recent meta‐analysis of data from 56 studies [4]. The subgroup analysis, however, identified a significant association between IVT and lower 28‐day and 90‐day mortality in patients with baseline NIHSS 4–5. This finding further suggests that patients with mild AIS and NIHSS scores at the higher end of the spectrum may benefit from IVT. Symptomatic intracerebral hemorrhage (sICH) occurred in 3.7% of patients within 36 h after initiating IVT. This aligns with previously reported rates ranging from 0.7% to 4.1% [2, 3, 5, 20, 21, 35, 36, 37, 39, 40].

Key strengths of our study include a large, matched sample size of 1736 individuals in each group, utilizing data from a nationwide high‐quality registry [25, 26]. Data were collected in a real‐world clinical setting, reflecting contemporary mild AIS treatment practices in Norway. To minimize bias, a comprehensive propensity score analysis was performed to ensure thorough balance estimation before matching. To reduce the risk of overfitting, we employed a gradient boosting machine with a relatively high number of trees, a shallow interaction depth, and a low shrinkage rate.

This study is limited by the variables available in the Norwegian Stroke Registry, increasing the risk of residual confounding due to unmeasured covariates and of confounding by indication. Our results may not be directly comparable to studies that define mild AIS differently. Additionally, the absence of sICH data in the control group precluded a direct comparison of sICH risk between groups. Another limitation is the lack of information on the use of early treatment with DAPT. Although no noticeable differences in baseline characteristics were found between patients lost to follow‐up and patients included in the matched study cohorts, a loss to follow‐up rate of 21.0% (n = 1572/7478) could adversely affect the results (Supplemental Table S2).

The results of this study suggest that IVT administered within 4.5 h of symptom onset may improve functional outcomes in patients with mild AIS, defined as an NIHSS score $\le$ 5 at admission. This is particularly true for those scoring at the higher end of the mild AIS spectrum. Future studies should incorporate stratification by narrower baseline NIHSS score intervals to enable a more precise evaluation of treatment effectiveness and safety. Moreover, the modified Rankin Scale serves as a broad tool for assessing the effects of IVT in cases of mild stroke. There may be a need for more granular assessment tools specifically designed to evaluate treatment effects in patients with the lowest NIHSS scores at admission. Additionally, patient‐reported outcome measures may complement clinician‐rated scales by capturing aspects of daily functioning and health‐related quality of life that are otherwise missed [41].

5. Conclusion

IVT was associated with favorable functional outcomes in patients with AIS and baseline NIHSS scores of 0–5. Subgroup analyses indicated that the treatment effect could be restricted to patients in the upper end of NIHSS scores 0–5. However, the proportion of symptomatic intracerebral hemorrhage of 3.7% highlights the importance of conducting a thorough risk–benefit assessment prior to IVT.

Author Contributions

Bent Indredavik: writing – review and editing. Torunn Varmdal: writing – review and editing. Agnethe Eltoft: supervision, conceptualization, writing – review and editing, methodology, investigation, validation, funding acquisition, resources. Ellisiv B. Mathiesen: supervision, project administration, conceptualization, writing – review and editing, methodology, investigation, validation, funding acquisition, resources. Elise K. Kristensen: data curation, software, formal analysis, writing – original draft, writing – review and editing, investigation, funding acquisition, visualization, validation.

Funding

UiT The Arctic University of Norway, Tromsø, Norway.

Ethics Statement

This study was approved by the Regional Committee for Medical and Health Research Ethics in North Norway (ID: REK Nord 504,501) and by the Data Protection Officer (ID No: 619,612) at UiT The Arctic University of Norway.

Consent

Informed consent was not sought for the present study because inclusion in the Norwegian Stroke Registry is mandatory, thus not based on individual patient consent.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Table S1: Baseline characteristics in excluded participants due to loss to follow‐up or late follow‐up > 6 months, compared to matched cohorts.

Table S2: Missing information on baseline characteristics in matched and unmatched cohorts, presented as n (%).

Data Availability Statement

The data are not publicly available due to ethical and legal restrictions. However, access may be granted upon reasonable request via www.helsedata.no, subject to approval by the relevant authorities.