Background Older patients undergoing thrombectomy for emergent large vessel occlusion have worse outcomes. However, complete or near-complete reperfusion (modified Thrombolysis in Cerebral Ischemia (mTICI) score of 2 c/3) is associated with improved outcomes compared with partial recanalisation (mTICI 2b).
Objective To examine the relationship between outcomes and age separately for the mTICI 2c/3, 2b and 0-2a groups in patients undergoing thrombectomy for anterior circulation emergent large vessel occlusion.
Methods Retrospective review of 157 consecutive patients undergoing thrombectomy at a single centre with an occlusion of the internal carotid artery (ICA), M1 or proximal M2 segments of the middle cerebral artery (MCA). Angiograms were graded in a blinded fashion. Patients were divided into three groups: mTICI 0-2a, mTICI 2b, and mTICI 2c/3. Demographics and workflow parameters were compared. Outcomes at 90 days were compared as a function of age, using both the conventional modified Rankin scale (mRs) and utility weighted mRs (UWmRs).
Results There were 72, 61 and 24 patients in the mTICI 2c/3, 2b and 0-2a groups, respectively. Outcomes were significantly worse with increasing age for the mTICI 2b group, but not for the mTICI 0-2a and 2c/3 groups (P=0.0002). With increasing age, outcomes of the mTICI 2b group approached those of the mTICI 0-2a group. However, outcomes of the mTICI 2c/3 groups were similar for all ages. This association was present for both the original mRs and UWmRs.
Conclusion Increasing age was associated with worse outcomes for those with partial (mTICI 2b) recanalisation, not in patients with complete (mTICI 2c/3) recanalisation.
Statistics from Altmetric.com
Mechanical thrombectomy in addition to best medical care (including intravenous tissue plasminogen activator (tPA)) dramatically improves outcomes in patients with emergent large vessel occlusion in the anterior circulation, even up to 24 hours from stroke onset.1–7 In the HERMES patient-level meta-analysis, a consistent benefit was seen from thrombectomy across all subgroups.8 When patient age was specifically evaluated, there was a benefit of thrombectomy across the entire age spectrum. However, there does appear to be an age-dependent response, with older patients much less likely to achieve independence with larger infarcts.9 Singer examined the outcomes of 362 patients in the multicentre ENDOSTROKE registry, and showed increasing rates of poor outcomes despite recanalisation as patients’ age increased.10 Kurre and colleagues examined 109 patients aged 80 or older who underwent thrombectomy, and suggested that despite recanalisation only 17.4% achieved independence at 90 days.
Historically, ‘successful’ recanalisation following thrombectomy has been defined as achieving modified Thrombolysis in Cerebral Ischemia (mTICI) 2b or 3 recanalisation. However, recent series have shown a greater benefit to achieving mTICI 3 versus 2b.11 Additionally, a new category, mTICI 2c has also been proposed, and outcomes with mTICI 2c appear to be similar to mTICI 3.12 13 The purpose of our study was to examine the relationship between degree of recanalisation (mTICI 2b versus 2c/3) and outcomes as a function of patient age. It is hypothesized that age will have a deleterious association with outcomes, but this association will be different for different levels of recanalisation.
Institutional Review Board approval was obtained for a retrospective review of our prospectively collected quality assurance (QA) stroke centre database. All patients with acute ischaemic stroke are entered into our QA database. We queried all patients who were treated with thrombectomy over a 21-month period with occlusions of the internal carotid artery (ICA), and M1 or proximal M2 segments of the middle cerebral artery (MCA). Patients who were not independent prior to treatment (ie, pretreatment modified Rankin score (mRs)>2), and those with posterior circulation occlusions were excluded from this series. Patient demographics including age, sex, admission National Institute of Health Stroke Scale (NIHSS), premorbid mRs, site of intra-cranial occlusion, and treatment with systemic tPA were collected. In addition, time to recanalisation from symptom onset, non-contrast computed tomography (NCCT) and CT angiography (CTA) were also prospectively collected.
Pretreatment imaging and treatment criteria
Patients with anterior circulation stroke were considered for thrombectomy at our centre if they had an acute occlusion of the ICA, M1 or proximal M2 segment of the MCA, with a corresponding disabling clinical deficit, typically defined as an admission NIHSS of 6 or higher. All patients underwent NCCT and CTA prior to treatment, at the first presenting hospital. During the initial portion of the study, CTA was performed as single phase, and later, as multi-phase CTA, using a previously described protocol.14 We did not use CT perfusion in any cases. Patients who had an admission Alberta Stroke Programme Early CT score (ASPECTS) less than 6 on NCCT, or those with absent collateral filling of the affected territory on single-phase or multi-phase CTA were typically not treated. For patients who presented beyond 6 hours from symptom onset, MRI was performed pretreatment, using criteria similar to those used in the DEFUSE-3 trial.15 We typically treated patients with a core infarct volume (defined as ADC value of less than 620) smaller than 70 mL, with a concomitant perfusion mismatch (defined as tMax volume of greater than 15 mL and mismatch ratio greater than 1.8).
The endovascular procedures were performed under moderate sedation, using midazolam and fentanyl. We did not electively intubate any patients, although a small number had been previously intubated for airway concerns. The details of our endovascular technique have been previously described, and our primary technique uses a combination of a stent retriever and local aspiration catheter in all cases.16 In the latter portion of the study, a small number of patients were treated with a cervical balloon guide catheter (BGC) in conjunction with the local aspiration catheter and stent retriever. A standardised setup for the procedure was used by all three neuro-interventionalists who performed the procedures. Adjunctive systemic heparin (3000 international units) was administered as a one-time dose following arterial puncture in patients who had not received intravenous tPA. Our usual protocol is to perform mechanical thrombectomy to recanalise the proximal occlusions (ICA, M1 and proximal M2). We did not perform mechanical thrombectomy in more distal occlusions (such as distal M2, M3 or A2 segment occlusions), nor did we administer adjunctive intra-arterial tPA in any cases.
All post-procedure angiograms were interpreted in an independent, blinded fashion by two experienced interventional neuroradiologists (MVJ and RAM). At the time of image review, the readers were blinded to the age and clinical outcome of the patient. Angiograms were scored according to the mTICI scale incorporating the mTICI 2c category. Cases of disagreement between the two expert readers were handled with a consensus adjudication session. For the purposes of this paper, we define mTICI 2c/3 as ‘near complete’ recanalisation, TICI 2b as ‘partial’ and TICI 0-2a as ‘incomplete’.
Study treatment and intervention
Based on the aforementioned angiographic review, patients were divided into the mTICI 2c/3, mTICI 2b and mTICI 0-2a groups.
The 90-day mRs was obtained at the time of follow-up in the vascular neurology clinic. For patients who were not able to return for in-person evaluation, a telephone follow-up was obtained. Patients for whom we were not able to obtain 90-day mRs were excluded from the analysis. When obtaining the 90-day mRs, the evaluator had access to the reported mTICI score in the medical record, but did not have access to the adjudicated, blinded angiographic interpretation. In addition, we calculated the utility weighted mRS (UWmRS), both pretreatment and at 90 days, using the previously described weights which were used in the DAWN trial.17 Symptomatic intracranial haemorrhage was defined as a four-point or more increase in the NIHSS from baseline, with imaging evidence of new parenchymal haematoma, intra-ventricular or subarachnoid haemorrhage.
Absolute 90-day mRS and UWmRS were modelled by age and TICI score. Because 90-day mRS may be influenced by pre-stroke mRS, change between baseline to 90-day mRS and UWmRS (Δ) was also examined. All modelling was accomplished using generalised linear modelling, assuming a binomial distribution (0–6 for mRS and 0–1 for UWmRS) and a normal distribution for Δ, where the interaction of age by TICI score (discrete 0-2a, 2b, 2c-3) was examined. TICI score was modelled by age using a generalised linear model assuming a binomial distribution (0–2). The mRS was also examined by dichotomising age at 80; however estimates from these analyses should be interpreted with caution.18–20 All modelling was accomplished using SAS software 9.4 (SAS Inc., Cary, North Carolina, USA) with the GLIMMIX procedure. Baseline differences by TICI were evaluated using a Kruskal–Wallis test and Fisher exact test. α was established, a priori, at the 0.05 level, with interval estimates calculated for 95% confidence. Post-hoc comparisons were adjusted using Bonferroni correction.
During the study period, we identified a total of 165 patients treated with thrombectomy who met the inclusion criteria (ICA, M1 or proximal M2 occlusion; independent at baseline). Ninety-day follow-up was missing in eight patients, leaving 157 available with complete data for analysis. The demographics for all patients, subdivided based on mTICI score (0-2a, 2b or 2c/3) are summarised in table 1. There were no differences observed in baseline characteristics across the groups, with the exception of more patients in the mTICI 0-2a group who received intravenous tPA and a higher proportion of ICA occlusions in that group. However, the mTICI 2b and 2c/3 groups were not statistically significantly different in their proportion of ICA, M1 and M2 occlusions, or pretreatment with tPA.
In figure 1A, we examined 90-day mRS as a function of age for each TICI group (mTICI 0-2a, 2b and 2 /3). A significant interaction was observed between age and the TICI groups (P=0.0002), in that there was a significant relationship between 90-day mRS and age for those in the TICI 2b group (OR 1.06, 95% CI 1.04 to 1.08, P<0.0001), while the relationship between mRS and age for the TICI 0-2a group only approached significance (OR 1.02, 95% CI 0.99 to 1.05, P=0.0815) and no statistically significant relationship was observed between age and 90-day mRS for those in the TICI 2c/3 group (OR 1.01, 95% CI 0.99 to 1.02, P=0.1360). That is, the likelihood of functional independence did not significantly increase with age for patients who achieved TICI 2c/3 recanalisation, but did so for those with TICI 2b recanalisation. This relationship was confirmed when dichotomising age at 80 years: no difference was observed between those over and those under 80 years of age for TICI 2c/3 and 0-2a (P=0.1672 and P=0.999, respectively); a difference was observed for 2b (OR 4.7, 95% CI 4.0 to 5.1 vs. OR 2.3, 95% CI 1.9 to 2.6, P<0.0001).
Because 90-day mRS may be influenced by a function of pre-stroke mRS, baseline functioning was controlled by examining the change between pre-stroke and 90-day mRS (Δ). As seen in figure 1B, a significant interaction was observed in mRS change between age and TICI *P=0.0307). Specifically, for every 1 year increase in age, the change in mRS increased 0.06 points (95% CI 0.03 to 0.10) in those in the TICI 2b group (P=0.0007). No significant relationship was observed between age and mRS change in those in the TICI 0-2a or 2c/3 groups (P=0.5321 and P=0.9873, respectively). As indicated in figure 1B, the change between pre-stroke and 90-day mRS is largely unchanged across ages for TICI 0-2a and 2c/3. Thus, the mean change in mRS between admission and 90 days for those in the TICI 0-2a group was 3.8 points (95% CI 2.9 to 4.6) and 1.6 points (95% CI 1.2 to 2.0) in the TICI 2c/3 group, a statistically significant difference (P<0.0001). Again, this relationship was confirmed when dichotomising age at 80 years: no difference was observed between those over and those under 80 years of age for TICI 2c/3 and 0-2a (P=0.999 and P=0.999, respectively); a difference was observed for 2b (3.8, 95% CI 2.8 to 4.8 vs. 1.9, 95% CI 1.4 to 2.5, P<0.0158).
Thus, as illustrated in figure 1B, in the TICI 2b group, age seemed to have a deleterious influence, such that with each 1 year advance in age, the increase in mRS change worsened and approached the value of mRS change in the TICI 0-2a group, while the TICI 2c/3 group seems to be resilient to increasing age (ie, relationship between age and mRS was flat).
We also examined 90-day UWmRS as a function of age for each TICI group (mTICI 0-2a, 2b and 2c/3). As seen in figure 2A, there was a significant relationship in UWmRS score and age for the TICI 2b group (OR 0.94, 95% CI 0.89 to 0. 98, P<0.0061), while no statistically significant relationship was observed between mRS and age for the TICI 0-2a and 2c/3 groups (OR 0.98, 95% CI 0.92 to 1.04, P=0.4904, and OR 0.99, 95% 0.95 to 1.02, P=0.5124). Again, this relationship was confirmed when dichotomising age at 80 years: no difference was observed between those over and those under 80 years of age for TICI 2c/3 and 0-2a (P=0.999 and P=0.999, respectively); a difference approached statistical significance for 2b (0.26, 95% CI 0.11 to 0.5 vs. 0.69, 95% CI 0.54 to 0.80, P<0.0557).
Additionally, we examined the change between pre-stroke and 90-day UWmRS (Δ) by age and TICI score. As seen in figure 2B, a significant interaction was observed in UWmRS change between age and TICI (P=0.0228). Specifically, for every 1 year increase in age, the change in UWmRS decreased 0.011 points (95% CI −0.01 to −0.02) in those in the TICI 2b group (P=0.0002). No significant relationship was observed between age and UWmRS change for the TICI 0-2a or 2c/3 groups (P=0.4725 and P=0.7691, respectively). As indicated in figure 2B, the change between pre-stroke and UWmRS is largely unchanging across ages for TICI 0-2a and 2c/3. Thus, the mean change in UWmRS for the TICI 0-2a group was −0.59 points (95% CI −0.43 to −0.76) and −0.24 points (95% CI −0.17 to −0.31) in the TICI 2c/3 group, a statistically significant difference (P<0.0001). This relationship was also confirmed when dichotomising age at 80 years: no difference was observed between those over and those under 80 years of age for TICI 2c/3 and 0-2a (P=0.999 and P=0.999, respectively); a difference was observed for 2b (−0.64, 95% CI −0.81 to −0.47 vs. −0.28, 95% CI −0.37 to −0.18, P<0.0045).
As with mRS change, UWmRS change, in the TICI2b group, age seemed to have a deleterious influence, such that with each 1 year advance in age, the increase in mRS change worsened and approached the value of mRS change in those in the TICI 0-2a group, while those in the TICI 2c/3 group seemed to be resilient to increasing age (ie, relationship between age and UWmRS is flat) (see figure 2B).
In order to show that the likelihood of achieving complete (mTICI 2c/3) recanalisation was not a function of age, we examined this relationship as well. As shown in figure 3, there was no relationship observed between TICI score and age (P=0.8118). Older patients were as likely to achieve complete recanalisation as younger patients in our series.
In this series, we show that patients with mTICI 2c/3 levels of recanalisation have similar outcomes irrespective of age, whereas patients with mTICI 2b recanalisation have worse outcomes with increasing patient age. This held true for both the standard categorical mRs as well as the utility weighted scale, and also for the absolute score at 90 days as well as the change from baseline. Importantly, there was not an age-dependent relationship on the likelihood of achieving mTICI 2c/3 recanalisation. That is to say, the likelihood of procedural success was not a function of age. Indeed, there appears to be a greater benefit of complete recanalisation in older patients than younger ones.
We believe these results are relevant to the modern thrombectomy era for a variety of reasons. First of all, this underscores the importance of separately reporting mTICI 2b versus mTICI 2c/3 recanalisation in all future series of thrombectomy. As others have suggested, we believe that mTICI 2c should be the goal of thrombectomy.21 Additionally, further maturation of thrombectomy techniques may allow greater rates of mTICI 2c/3 reperfusion, and evaluation of the relative efficacy of novel techniques should account for the age of the patient. As our results highlight, achieving mTICI 2b reperfusion in an 80-year-old patient is not the same as in a 50 year old.
As previously discussed, studies by Singer10 and Kurre22 have shown worse outcomes despite recanalisation in older patients but both of those grouped mTICI 2b/3 together, rather than examining them separately. The relationship between pretreatment infarct size and outcomes has also been examined as a function of age. Ribo and colleagues suggested the concept of a Maximal Admission Core Lesion Compatible with Favorable Outcome (MALCOM), using CT perfusion derived measurements of infarct core.23 They showed that in octogenarians, the MALCOM was 15 mL compared with 40 mL in younger patients. Similarly, Hwang showed a lower likelihood of achieving independence in those older than 75 years old, with a more restrictive baseline DWI ASPECTS and core lesion size than in the youngest group.24
Among the current generation of randomised thrombectomy trials, prespecified analysis based on an age threshold showed significant benefit in older subgroups in the ESCAPE and MR CLEAN trials (categorised as 80 years or older vs. younger) and in the SWIFT-PRIME (categorised as 70 years or older vs. younger) trial, with no heterogeneity of benefit. In the HERMES patient-level meta-analysis, while overall outcomes were worse in older patients than younger ones, there was no difference in the benefit of thrombectomy across the entire age spectrum using a mixed methods linear regression. Additionally, when divided into subgroups, the adjusted common odds ratio (aCOR) was not significantly different among the 50–59, 60–69, 70–79 and 80+ age groups, with 218, 333, 371 and 198 patients in those groups, respectively. Similarly, even in the 6–24 hour time window, DAWN showed similar benefit to those less than 80 versus 80 years and older. However, the differences in outcomes between mTICI 2b and 3 recanalisation were not specifically examined in any of those studies. An analysis of 269 patients from the SWIFT and STAR trials showed that successful recanalisation (mTICI 2b or higher) was strongly associated with favorable outcome, and that the benefit of recanalisation increased with age.25 However, differences between mTICI 2b and 3 were not explored.
A previous study examined the benefit of complete recanalisation (TICI 3) versus incomplete recanalisation (TICI 0-2b) in a series of 80 patients, dichotomised by age ≥80 years versus <80 years26. They found that independence at 90 days (mRs 0–2) was 65% in the older cohort and 68% in the younger cohort, when TICI 3 recanalisation was achieved, and 21% versus 45% for TICI 0-2b in the older and younger cohorts, respectively. However, they did not consider TICI 2b separately from TICI 0-2a, although 89% of all their patients achieved at least TICI 2b recanalisation. Our study expands on this work with a larger sample size, the use of the mTICI 2c category, and also in examining the relationship as a function of age.
The workflow times in this series were faster than those published in the randomised trials. For instance, our median and fastest quartile for symptom onset to recanalisation times were 225 and 154 min in the TICI 2c/3, and 243 and 181 min in the TICI 2b groups, respectively. In contrast, the median and fastest quartile times in the HERMES dataset were 285 and 210 min, respectively. While we feel this may allow us to demonstrate the importance of achieving near complete recanalisation, it may also be the case that the difference in recanalisation time contributes tosome extent to the improved outcomes.
Our study has several limitations. This is a single-centre, retrospective series, and as such, the generalisability of these results across multiple centres is unknown. However, this also ensures that workflow processes and procedural techniques were held constant in our patient sample. It is important to acknowledge that we did not aggressively pursue more distal occlusions (such as distal M2/M3 of the MCA or A2 segment of the anterior cerebral artery) which remained after primary thrombectomy of a proximal lesion. It is unknown whether there is a benefit to performing additional endovascular manoeuvres (whether mechanical or pharmacological) to recanalise distal occlusions remaining after thrombectomy of the more proximal primary lesion. Additionally, at our centre, we do not perform endovascular therapy on primary distal occlusions, and it is unknown based on these results if there would be a benefit to such a treatment or not.
For patients undergoing thrombectomy for anterior circulation emergent large vessel occlusion, there is an association between increasing age and worse outcomes in patients with mTICI 2b recanalisation, but not in those with mTICI 2c/3 recanalisation. With complete recanalisation, there is a similar likelihood of independence at 90 days irrespective of age. Further investigation of this relationship between degree of recanalisation (separating mTICI 2b and 2c/3) is warranted. As endovascular technology and techniques improve and the likelihood of achieving complete recanalisation increases, the overall outcomes in older patients may approach those of younger patients.
Contributors All authors: made substantial contributions to the design of the analysis, interpretation of data, drafting and critically revising the manuscript as well as final approval for the version published. They agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Patient consent Not required.
Ethics approval Lifespan Institutional Review Board.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.