Use of Intravenous TPA in Patients Outside the Defined Criteria

Abstract and Introduction

Abstract

Currently, intravenous thrombolysis is by far the most effective treatment of acute ischemic stroke, and its use can independently strongly increase the proportion of stroke patients surviving. While the use of recombinant tissue plasminogen activator in accordance to licensed criteria has continuously risen, off-label use is also frequent. In this review the most important reasons to transcend current license criteria are discussed and evidence is summarized from new studies, such as IST-3, contributing to the balance of increased benefit as opposed to possible harm in situations of off-label use of recombinant tissue plasminogen activator in stroke. In addition, several scores to predict risk and outcome in patients undergoing thrombolysis are compared.

Introduction

Intravenous thrombolysis is the single most effective treatment of acute ischemic stroke. Depending on time of treatment after stroke onset, between 50 and 280 additional patients out of 1000 will be alive and independent after 12 months compared with placebo treatment.^[1]

Recombinant tissue plasminogen activator (rtPA) or alteplase has been approved in the USA since 1996 for treatment of acute stroke, and in the EU since 2002, for use within a 3-h time window.^[2] In the EU in 2011, the license was extended to a time window of 4.5 h, 3 years after the publication of the ECASS-III trial.^[3] Since then, the percentage of patients treated with rtPA is steadily increasing, especially in countries with good coverage of stroke units, such as Germany or Austria.^[4] Furthermore, the knowledge resulting from randomized controlled trials (RCTs) has increased since the approval of alteplase, especially with the largest thrombolysis trial, IST-3.^[5] This trial included a large number of patients, especially the elderly, that would not have fallen under the current license criteria.

A considerable number of patients treated to date do not fulfill all excluding or including criteria formulated by the license and would therefore be precluded from treatment. In such situations, it is not uncommon to treat patients outside the limits of the approvals, in a so-called off-label use. Most off-label use concerns the age limit, the time window, mild strokes, concomitant oral anticoagulation (OAC) or blood pressure (BP), outside the limits used for the main studies.

It is noteworthy that there are important differences between the guidelines of the USA and the EU. In the USA, there is no age limit of 80 years, and the permitted minimum interval between antecedent surgery and the stroke is 14 days in the USA, in comparison with 3 months in the EU countries.^[6] The differences in the approval criteria probably result from different expert opinions.

Exclusion criteria can be classified into the following four categories: patients groups that have not been tested (or not sufficiently studied) by RCTs, that is, subjects aged over 80 years before publication of the IST-3 study; exclusion criteria used in trials due to safety concerns (i.e., surgery, history of intracranial bleeding, high BP); factors that have been highlighted as of concern from trials, cohort studies or meta-analysis of randomized trials – that is, high blood glucose – and likelihood of low efficacy in specified patient groups – that is, in patients with mild stroke symptoms or with rapidly dissolving symptoms. Thus, restrictions for use of rtPA are based on quite different types of evidence balancing favorability and risks.

Table 1. Blood glucose values and risk of symptomatic intracranial hemorrhage in study cohorts of intravenous recombinant tissue plasminogen activator-treated patients.
Study (year)	Patients (n)	Cohort	Mean age (years)	Diabetes (%)	Median baseline NIHSS	Cutoff criterion	OR	95% CI	Rate SICH	Ref.
NINDS (1995)	312	RCT rtPA treatment group	68	22	14	Admission glucose per 5.6 mmol/l increase	1.75 multivariate	1.11–2.78	6.4	[25]
ECASS-II (1998)	409	RCT rtPA treatment group	68	21	11	Postadmission blood glucose >7.8 mmol/l	5.69 multivariate	2.05–15.8	8.1	[21]
Houston/Calgary (1999)	138	Stroke registry rtPA-treated patients	66	23	16	Admission glucose per 5.5 mmol/l increase	2.26 multivariate	1.07–4.29	9	[28]
SITS-MOST (2008)	6483	Stroke registry rtPA-treated patients	68	16	12	1 SD above mean admission glucose value	1.32 univariate	1.22–1.43	8.5	[35]
CASES (2009)	1098	Stroke registry rtPA-treated patients	73	10	14	Admission blood glucose >8 mmol/l	1.65 multivariate	0.95–3.00	4.5	[27]

Table 2. Characteristics of different scores for prediction of outcome and symptomatic intracranial hemorrhage in recombinant tissue plasminogen activator-treated stroke patients.
Score	Derivation sample (n)	Observed SICH rate (%)	Outcome predicted	Predictive variables	Included off-label characteristics	Statistic	External validation	Ref.
Stroke-TPI	2131	Data not shown	3-month outcome	Age, sex, NIHSS, OTT, DM, SBP, prior stroke	OTT 3–6 h	ROC: 0.788	X	[57]
HAT	302	6	SICH (NINDS definition)	DM, BG, NIHSS, early hypodensity on CT		C-statistic: 0.88	X	[58]
SITS	31,627	1.8	SICH (SITS definition)	Age, OTT, NIHSS, antithrombotic therapy, BG, SBP, hypertension, bodyweight		C-statistic: 0.7		[59]
GRASP	10,242	4.8	SICH (NINDS definition)	Asian race, male sex, NIHSS, SBP, BG	Age >80 years; SBP ≥180; OAC with INR <1.7; NIHSS >20	C-statistic: 0.7	X	[60]
iScore	1696	6.9	SICH (any symptomatic hemorrhage), outcome at discharge, 30-day mortality	Age, sex, CSS, stroke subtype, Afib, CHF, comorbidity, disability, BG	Age >80 years; severe strokes (CSS)	ROC: 0.85	X	[61]
DRAGON	1319	Data not shown	3-month outcome	Age, prestroke mRS, NIHSS, OTT, hyperdense MCA sign, BG	Age >80 years	ROC: 0.84	X	[62]

Afib: Atrial fibrillation; BG: Blood glucose; CHF: Congestive heart failure; CSS: Canadian Stroke Scale; MCA: Median cerebral artery; mRS: Modified Rankin scale; NIHSS: National Institute of Health Stroke Scale; OTT: Onset to treatment time; ROC: Receiver operator characteristic; SBP: Systolic blood pressure.