Intravenous t-PA followed by endovascular therapy combines the advantages of a rapid start of intravenous t-PA with a greater likelihood of early recanalization with the use of endovascular therapy in patients with persistent occlusion after treatment with intravenous t-PA. This combined approach was first tested in 1994 with the Emergency Management of Stroke (EMS) Trial (Table 8.1), a randomized controlled study of 35 patients, that was designed to test feasibility and safety of combined IV rt-PA (0.6 mg/kg over 30 min) and IA rt-PA therapy (up to 20 mg delivered over 2 h at the clot), compared to IV placebo plus IA t-PA therapy [21]. Intravenous treatment in the EMS Study was to be started within 3 h of stroke onset. This study was not powered to examine efficacy differences between the two treatment groups, but rather to demonstrate feasibility and safety of the combined strategy. Recanalization was better (p = 0.03) in the IV/IA group with TIMI 3 flow in 6 of 11 IV/IA patients versus 1 of 10 placebo/IA patients and correlated to the total dose of r-TPA (p = 0.05). The proportion of improved outcomes (prospectively defined as a seven point improvement of the NIHSSS at 7–10 days) did not differ between groups (both groups = 24 %).

The prospectively determined primary safety measure for the EMS Study was life-threatening bleeding during the first 24 h after therapy. There were no symptomatic or asymptomatic intracerebral hematomas in the 35 EMS patients. One symptomatic hemorrhagic infarction occurred during the first 24 h among 18 patients treated with placebo/IA rt-PA therapy. Two symptomatic hemorrhagic infarctions occurred in the combined IV/IA group 24–72 h after treatment. Asymptomatic intracerebral hemorrhagic infarction during the first 72 h was more frequent in the IV/IA group (n = 5) than in the IA only group (n = 1).

Results for only participants with an M-1 or M-2 occlusion in the EMS Study were compared to the results of the PROACT II Study, which only included MCA occlusions. The number of patients is small but there was a trend toward better outcomes (mRS of 0–2 = 57 % EMS, 40 % PROACT II, same median NIHSSS at baseline for EMS and PROACT II patients) and better rates of TIMI Grade 3 flow (complete recanalization) in the patients treated with combined IV/IA rt-PA. Likely these differences are related to the quicker time to initiation of therapy in the EMS Trial.

The Interventional Management of Stroke (IMS) Study was undertaken to test the feasibility and safety of a combined IV/IA approach to recanalization in a larger group of ischemic stroke patients [22]. It was designed as a multi-center, open-label; single arm study comparing eligible patients treated with an IV/IA rt-PA approach to historical subjects of similar age and baseline NIHSS score from the NINDS rt-PA Stroke Trial. Patients enrolled in the IMS Study were treated with low dose IV rt-PA (0.6 mg/kg, 15 % as a bolus over 1 min followed by the remainder over 30 min) and up to 22 mg of IA rt-PA if clot was identified at angiography. The IV and IA doses of t-PA were chosen so that the patients would rarely receive a greater dose than standard IV t-PA dose of 0.9 mg/kg (90 mg maximum total dose). Intravenous and intra-arterial rt-PA was to begin within 3 and 5 h of stroke onset, respectively. A 2,000 U bolus of heparin was administered before IA rt-PA therapy and a low dose infusion (40 U/h) was continued until the catheter was removed. The primary safety endpoint of the trial was life-threatening bleeding within 36 h of rt-PA infusion (symptomatic intracerebral hematoma/hemorrhagic infarction or systemic bleeding requiring transfusion of ≥3 units of blood or major surgical intervention). Although the study was not powered for efficacy, the primary outcome endpoint was a mRS of 0–1 at 3 months. Secondary outcome measures included rates of TIMI grade II or III flow at the completion of angiography, a mRS of 0–2 at 3 months, and other 3-month-endpoints from the NINDS rt-PA Stroke Trial. Additional secondary endpoints were also identified based upon a prospective analysis of the most sensitive endpoints from the NINDS rt-PA Stroke Trial [22, 23].

Of 1,477 subjects screened over 9 months, 80 patients were enrolled. Only 1 patient was lost to follow-up. Median time from symptom onset to IV rt-PA administration was 140 min, while median time to start of IA rt-PA was 212 min, compared to a median 90 min for IV rt-PA in the NINDS rt-PA trial [1]. Mean and median NIHSS scores for IMS patients were both 18.

Combined IV/IA therapy proved comparatively safe in the selected patients. Symptomatic ICH was more common in IMS and NINDS rt-PA patients than NINDS placebo patients (IMS patients 6.3 %, NINDS rt-PA patients 6.6 %, NINDS placebo patients 1.0 %). While the mortality and symptomatic intracerebral hemorrhage rates were similar between IMS subjects and rt-PA treated subjects in the NINDS rt-PA Stroke Trial, the rate of asymptomatic intracerebral hemorrhage during the first 36 h was higher in the IMS Study. However, only 6 (7.5 %) of the participants had PH2 type hematomas that were associated with neurological deterioration and poor outcome [24]. In addition, the overall rate of intracerebral hemorrhage in the IMS Study was similar to that in the PROACT II Study. IMS subjects with occlusions of the internal carotid artery, with or without distal occlusions, were particularly likely to have some hemorrhagic change (20 of 28–71 %). By comparison, only 15 of the 36 (42 %) subjects with occlusions in the middle cerebral artery or vertebral-basilar artery system had any hemorrhagic change.

Outcome was better in IMS subjects than matched NINDS placebo-treated subjects, with 30 % and 18 % of respective patients achieving a mRS of 0–1 at 3 months (OR 2.26, 95 % CI 1.15–4.47). IMS subjects had similar 3-month primary and secondary outcomes as rt-PA treated NINDS patients, although odds ratios trended in favor of IMS subjects.

For subjects receiving IA rt-PA in addition to IV rt-PA, rates of recanalization to TIMI 3 versus TIMI 2 or 3 were 13 % and 67 % respectively. Among these patients, achievement of TIMI 2 or 3 flow was associated with a better outcome (mRS 0-1 at 3 months) than TIMI 0 or 1 flow (p = 0.013). Initiation of IA rt-PA within 3 h of stroke onset was associated with a trend toward better outcome compared to initiation after 3 h (p = 0.095).

The IMS II Trial immediately followed completion of the IMS I Trial and was very similar in design except that intra-arterial t-PA was administered via the EKOS micro-infusion catheter or a standard microcatheter at the site of the thrombus up to a total dose of 22 mg over 2 h of infusion or until thrombolysis [9]. The EKOS micro-infusion catheter delivered ultrasound energy at the catheter tip while it was also delivering IA t-PA. The 81 subjects in the trial had a median baseline NIHSS score of 19 and the median time to initiation of IV t-PA was 142 min as compared with 108 min for placebo and 90 min for rt-PA-treated subjects in the NINDS rt-PA Stroke Trial (P = 0.0001). The 3-month mortality in IMS II subjects was 16 % as compared with the mortality of placebo (24 %) and rt-PA–treated subjects (21 %) in the NINDS rt-PA Stroke Trial who had the same range of baseline NIHSS scores and ages as patients enrolled in IMS II. The rate of symptomatic intracerebral hemorrhage in IMS II subjects (9.9 %) was not significantly different than that for rt-PA treated subjects in the NINDS t-PA Stroke Trial (6.6 %). IMS II subjects had significantly better outcomes at 3 months than NINDS placebo-treated subjects for all end points (OR—2.7) and better outcomes than NINDS rt-PA–treated subjects as measured by the Barthel Index and Global Test Statistic.

Based upon the pilot IMS I and IMS II Trials as well as other reported experience with the combined IV t-PA/endovascular approach, the NINDS-funded Interventional Management of Stroke (IMS) III trial was initiated. IMS III was the first international, multi-center (US, Europe, Canada, and Australia), phase 3, randomized, open-label clinical trial with blinded outcome designed to directly compare the approach of intravenous thrombolysis followed by endovascular treatment (intra-arterial t-PA and FDA-approved embolectomy devices) with intravenous thrombolysis alone [11]. This trial included only patients with moderate to severe strokes as measured by a NIHSS greater than 10; or greater than 8 if large vessel occlusion was documented on vascular imaging. At the time of the design and initiation of IMS III, a poll of participating centers indicated approximately 20 % performed CT angiography as part of local standard of practice, a low figure, giving no assurance CT angiography should/could be used as an inclusion factor in a trial with time-dependent therapies. However, pre-treatment CT angiographic data was collected prospectively as the trial proceeded and the use of CT angiography at IMS III sites increased substantially during the trial. At completion of the IMS III Trial, 306 (47 %) IMS III subjects had a baseline CT angiography prior to initiation of treatment [11].

Intravenous thrombolysis had to be started within 3 h in both groups to be included in the trial. Initially the combination arm received only 2/3’s the standard dose of IV thrombolytic which had been part of the design of the IMS I and IMS II Trials, but based on safety data that accumulated in the literature during the study, full dose IV t-PA therapy was used as of June 2011 (the trial was conducted from 2006 to 2012).

Patients in the endovascular arm underwent angiography as soon as possible after infusion of IV thrombolysis, but the procedure was intended to begin no later than 5 h after symptom onset and completed by no later than 7 h. Method of endovascular therapy was chosen by the site interventionalist and included IA thrombolysis with or without intravascular ultrasound as well as embolectomy devices that were approved for use in the trial. The IMS III trial was designed to evaluate and approve new embolectomy technology and add new devices to the available options included in the endovascular approach; this design allowed the trial to keep up with advances in technology after review and approval of the devices by the FDA and the IMS III Executive Committee. Given the trial spanned from 2006 to 2012, this trial incorporated both the Penumbra aspiration system and the Solitaire Stent Retriever as they became approved by the FDA and approved for use in the Trial (Table 8.2). The trial was planned to include 900 patients, but was halted after enrollment of 656 patients. The Data Safety Monitoring Committee recommended stoppage because a predefined futility boundary related to the primary efficacy outcome had been crossed.