Mechanical Thrombectomy for Pediatric Arterial Ischemic Stroke from Acute M2 Occlusion

DISCUSSION

In this case series, we demonstrated successful use of mechanical thrombectomy in 6 pediatric patients with M2 occlusion. All 6 patients had excellent clinical outcomes (ped-mRS = 0–1) at 3 months poststroke and final follow-up. While the sample sizes are too small for meaningful comparative statistical testing, the 5 pediatric patients with M2 occlusion managed with best medical treatment alone from our previously published case-control study all had mild-to-moderately severe disability at 3 months poststroke (ped-mRS = 2–4).³

No patients developed symptomatic intracranial hemorrhage or required decompressive craniectomy. The procedural complications (detailed above) were all minor with no long-term sequelae. In our view, postretrieval vasospasm in young children should be anticipated, and antispasmodic medications should be drawn up and ready for intra-arterial infusion. Overall, MeVO thrombectomy in this small pediatric sample appears relatively safe.

Close collaboration between pediatric neurology and interventional neuroradiology colleagues is essential for appropriate patient selection. During multidisciplinary team discussions for pediatric M2 occlusions, one should consider the following points:

• 1) Patient age and premorbid clinical status

• 2) Ped-NIHSS score

• 3) Probable stroke etiology

• 4) Nondominant-versus-dominant hemisphere involvement

• 5) Opercular-versus-perirolandic cortical involvement⁵

• 6) Presence of speech or dominant hand impairment

• 7) Overall risk profile of thrombectomy tailored to the individual patient

• 8)Availability of pediatric stroke and neurointerventional expertise.

Patients with suspected focal cerebral arteriopathy or possible Moyamoya syndrome should be considered carefully, with a bias toward medical therapy alone.⁷ The informed consent process undertaken with the patients’ caregivers must be tailored on the basis of the above information, and the balance of risks versus benefits should be clearly communicated.

We would argue against using the Ped-NIHSS score in isolation for treatment decisions. Delayed deterioration of patients with LVO presenting with lower NIHSS scores is described in both pediatric¹¹ and adult settings.¹² Collateral blood flow within regions of the penumbra can fluctuate with time, and deterioration of such flow results in delayed infarct growth.¹³ Therefore, the condition of the child with a fluctuating Ped-NIHSS score may deteriorate during periods of relative hypotension (such as sleep), with a potential loss of treatment options. Interrater reliability of the Ped-NIHSS is lower when testing sensory/visual domains and is not validated in children younger than 2 years of age.¹⁴ Therefore, the Ped-NIHSS score should be considered in the wider context of clinical presentation, occlusion location, and potential long-term disability during a longer life span.

The role of perfusion imaging for the treatment decision-making in pediatric LVO stroke remains uncertain. Existing pediatric stroke guidelines make no recommendations regarding the use of CT perfusion or MR perfusion imaging for patient selection.^7,15 Quantitative parameters on CT perfusion imaging are known to vary with age during childhood, and this aspect must be considered during interpretation.¹⁶ Recent evidence suggests that performing perfusion imaging does not delay the time to recanalization in children treated with mechanical thrombectomy, that target mismatch criteria do not correlate with functional outcome, but a favorable hypoperfusion intensity ratio does correlate with clinical outcomes.¹⁷ In our experience, perfusion imaging can be helpful in borderline cases but is not sufficient to exclude a patient from being offered thrombectomy.

Our study has several limitations. First, the sample size is small but does form the largest published series of pediatric patients with M2 occlusion treated with thrombectomy. Second, there is probable underlying selection bias when selecting pediatric patients for thrombectomy in this context. This can be seen in the significantly greater delays in acquiring imaging in the control group from our previously published case-control study.³ Third, the role of thrombectomy for M2 occlusions in adults is still being assessed by at least 3 randomized trials. Fourth, 1 patient having undergone thrombectomy (case 4) was 11 months of age. The role of thrombectomy in infants remains controversial. Fifth, the main stroke etiologies in our study were cardioembolism and dissection. Pediatric stroke is heterogeneous, precluding generalization to the broader pediatric stroke population. Finally, the ped-mRS is potentially biased toward motor outcomes.