Near-Complete Reversal of Large Diffusion-Weighted Imaging Lesion after Thrombectomy: A Case Report and Literature Review

Stroke is the leading cause of death and disability, and its incidence increases rapidly in low- and middle-income countries [1]. It is estimated that in Vietnam, over 150,000 new cases of stroke occurred in 2021 [2]. Endovascular therapy (EVT) has become a cornerstone of treatment for selected patients who have acute ischemic stroke (AIS) and large vessel occlusion (LVO) [3], as soon as possible in order to rescue the penumbra tissue and decrease the extent of the infraction core [4]. More recently, results published from randomized clinical trials demonstrated that thrombectomy in patients with AIS due to anterior LVO and a large ischemic core (defined as an ASPECTS value lower than 6) led to significantly better functional outcomes than best medical treatment alone [5‒9]. However, the data in these clinical trials mainly focused on patients with moderate ischemic core with ASPECTS of 3 and 5 but was very limited to the patient group with an extremely large core (ASPECTS value of 0 to 2) [10].

The determination of the irreversible infraction areas is of crucial importance in the assessment of the potential pros and cons associated with making clinical decisions [3]. Regarding mechanism, diffusion-weighted imaging (DWI) can detect restricted water diffusion in ischemic brain tissues rapidly, thus showing evidence of the cytotoxic edema phenomenon that occurs in AIS [11, 12]. Therefore, a hypertense lesion on DWI with decreased apparent diffusion coefficient (ADC) values was considered the ischemic region that has been consumed to be irreversibly injured [13]. Besides, the ischemic region is evaluated as a core when this area has DWI hypertense with an ADC value less than 620 × 10⁻⁶ mm^2/s on b0/b1000 [14].

This conception, however, has been a controversial topic at present because some observational studies have demonstrated that some DWI hypersensitive regions could be reversible if the occlusive vessel is recanalized promptly [13, 15, 16]. The reversibility of DWI lesions accounts for approximately 25% of AIS [13]. This presents a challenge in accurately defining the infarct core, particularly during the thrombectomy era. Overestimation of the ischemic core volume by more than 10 mL compared to the final infarct volume on DWI, also known as the “ghost infarct core” [17]. The reversibility of the DWI lesion was linked to early neurological improvement as well as a favorable functional outcome. Herein, we reported a case of AIS with anterior LVO and a large ischemic region on DWI with near-complete reversal after thrombectomy and achieved a dramatic recovery.

A 58-year-old male patient arrived at our emergency department experiencing a sudden generalized tonic-clonic seizure and confusion, accompanied by left hemiparesis following the seizure. He was admitted to the emergency room 3 h after the onset of symptoms and was sedated by intravenous midazolam. His past medical history revealed a diagnosis of hypertension, chronic coronary syndrome, and heart failure. Prescription medications used concurrently were aspirin 81 mg (1 tablet), atorvastatin 20 mg (1 tablet), lisinopril 10 mg (1 tablet), bisoprolol 2.5 mg (1 tablet), felodipine 5 mg (1 tablet), and furosemide 40 mg (1 tablet). On clinical examination, the patient had confusion with a Glasgow Coma Scale (GCS) score of 9; he exhibited left-sided hemiplegia and forced deviation of the eyes and head to the right side, with an estimated NIHSS score of 26 (with sedation). A non-contrast computed tomography (NCCT) scan was accomplished within 45 min after hospitalization. It showed no evidence of hemorrhage, but some suspected early low-density signs in the right lentiform nucleus, insular cortex, and M1-M2-M3 cortical regions (shown in Fig. 1a, b), with ASPECTS of 5. Because of the atypical onset of stroke symptoms, the on-call stroke neurologist suggested brain magnetic resonance imaging (MRI) for diagnosis. The subsequent MRI, performed 1 h and 43 min after hospital admission, revealed a new acute ischemic lesion in the right hemisphere with occlusion of the terminal right internal carotid artery on time-of-flight MR angiography. The ischemic region was tremendously large on DWI, with an ASPECTS value of 2 (shown in Fig. 2a). Using RAPID software, the ischemic core volume on baseline imaging was retrospectively calculated to be 141 mL (shown in Fig. 3a). In particular, the lesion on fluid-attenuated inversion recovery (FLAIR) imaging did not appear clearly.

After informed consent was signed by his relatives, the patient was immediately transferred to our digital subtraction angiography unit. The procedure lasted 30 min, during which time the patient was locally anesthetized with lidocaine and fentanyl. In the first angiogram, the patient had very poor collateral circulation, with a grade of 1/4 according to the American Society of Interventional and Therapeutic Neuroradiology/Society of Interventional Radiology (ASITN/SIR) collateral grading scale. During the procedure, multiple small thrombi were removed after one attempt using the SolitaireX stent retriever, achieving TICI-3 recanalization (shown in Fig. 4a, b). After the procedure, neurological symptoms improved significantly within 24 h of admission, with a GCS score of 15, a slight deviation in gaze to the left, and mild left hemiparesis. The NCCT was repeated the next day and showed no evidence of intracranial hemorrhage (ICH), with an ASPECT score of 6 (shown in Fig. 1c, d). On the 8th day after hospitalization, a brain MRI was performed to estimate the ischemic lesion, revealing an ASPECT score of 5 (on FLAIR). Notably, on DWI, the ischemic regions showed dramatic reversal, with the lesion area predominantly in the cortical insular and temporal lobes on FLAIR imaging (shown in Fig. 2b). RAPID software showed a remarkably decreased infarct volume of 61 mL (shown in Fig. 3b).

On evaluating stroke etiology, atrial fibrillation was detected on the surface electrocardiography. Transthoracic echocardiography showed evidence of hypokinesis in the apical segment of the left ventricle, elevated pulmonary artery pressure of 40 mm Hg, and a decreased left ventricular ejection fraction of approximately 40%, without moderate-to-severe mitral valve stenosis or any thrombosis in cardiac chambers. Furthermore, there was no significant evidence of severe intracranial atherosclerosis. Finally, stroke etiology was classified as cardioembolism.

The patient was discharged on day 9 with a fully awake state, no recurrent seizures, and an NIHSS score of 3 points. Considering factors such as the final infarction core, comorbidities, and balancing between stroke recurrence and the risk of hemorrhagic transformation, the prescribed medications are as follows: Direct oral anticoagulant (Dabigatran) 110 mg b.i.d., atorvastatin 20 mg (1 tablet), isosorbide mononitrate 60 mg twice a day, furosemide 40 mg (2 tablets), valsartan 80 mg (2 tablets), and spironolactone 25 mg (2 tablets). He was compliant with the treatment. At 90-day follow-up, he did not experience any seizures or recurrent strokes, with a mRS score of 1.

In this case, we report a case of AIS with an initial generalized seizure at onset, also known as stroke chameleon. An epileptic seizure can be an early sign of acute stroke as well as one of the differential diagnoses, making it a diagnostic and therapeutic challenge for stroke neurologists in clinical practice [18]. On neuroimaging, hyperperfusion patterns and not conforming to vascular regions are important signs in order to suspect brain injury abnormalities due to an underlying cause other than stroke [19, 20]. In this patient, brain MRI was a priority choice for several reasons. First and foremost, we quickly and accurately excluded stroke mimics. Additionally, this imaging assessed the severity of AIS and LVO to aid clinical decision-making.

The efficacy and safety of EVT for AIS with anterior LVO and a large ischemic core have been proven in recent clinical trials [5‒9]. However, data on the treatment effectiveness of EVT in patient extremely low ASPECT group are still very limited. Most recently, the Trial of Thrombectomy for Stroke with a Large Infarct of Unrestricted Size (LASTE) trial has provided valuable evidence on the efficacy and safety of thrombectomy in patients with very low ASPECTS (with a rate of more than half in enrolled patients). Despite increasing the risk of ICH but not symptomatic ICH, this trial show that EVT is associated with better functional outcomes and a reduced mortality rate when compared to medical treatment alone [9]. LASTE trial also demonstrated a statistically significant improvement in the mRS ordinal shift analysis, with a number needed to treat of 4.2, comparable to that observed in trials involving patients with small-to-moderate core infarctions [21]. In a subsequent meta-analysis based on these previous clinical trials that enrolled AIS patients with very low ASPECTS (0–2), the pooled analysis showed a significant shift toward better 90-day mRS scores in favor of EVT compared with medical management alone (generalized OR 1.62, 95% confidence interval 1.29–2.04), suggesting the notable efficacy of EVT [22].

DWI is now widely used as one of the standard diagnostic imaging techniques to define the brain infarct core in the selection criteria of numerous pivotal clinical trials, including for large ischemic lesion [5‒9]. In other aspects, this perspective has been challenged because the DWI reversal phenomenon has been observed in human and animal models [13, 16]. This phenomenon is associated with faster successful reperfusion of the ischemic regions, with or without tissue plasminogen activator or EVT, a shorter time to thrombectomy, and the regions with milder reduced diffusion (characterized by a higher ADC value that is greater than 620 × 10⁻⁶ mm^2/s) [13, 15‒17, 23]. In clinical settings, early neurological improvement was seen within 2 h up to 7 days of stroke when DWI reversal occurred and was significantly linked with good long-term functional outcomes. In patient groups with LVO and large ischemic regions, reversibility of DWI lesions was also observed, but the data were scarce. One study indicated the prevalence of this phenomenon is reported to be 18.9% (based on the DWI-ASPECTS criteria of 0–5). It was significantly related to early neurological improvement and long-term favorable outcomes, with only successful recanalization (mTICI 2c-3) being the relevant prognostic factor [24]. Notably, complete reversal is seldom seen (about 0.8%), and the DWI reversal volume greater than 10% (defining “ghost infarct core”) is uncommon [13, 17, 24, 25]. One retrospective study showed that the prevalence of “ghost infraction core” was about 16% (using the ischemic core definition of cerebral blood flow less than 30% as the reference parameter) [25]. Another retrospective study demonstrated similar findings that ghost infarct core was observed in approximately 4% (using the ischemic core criteria as regions had DWI positive with ADC ≤620 × 10⁻⁶ mm^2/s) [17]. Overestimated infarct core volume was also associated with a shorter time to reperfusion, fast, and successful recanalization [25].

Although the estimated ischemic lesion on DWI is extremely large, this patient presented within a 5-h time window from symptom onset with a significant mismatch between DWI and FLAIR profile. Our treatment decision, similarly to the inclusion criteria in the Recovery by Endovascular Salvage for Cerebral Ultra-Acute Embolism-Japan Large Ischemic Core (RESCUE-Japan LIMIT) and LASTE trials, selected patients whose ischemic lesions on DWI had no signal change in the initial image on FLAIR [8, 9]. This conception suggests very severe hypoperfusion and an uncertain extent of necrotic tissue [16, 26]. Additionally, as mentioned above, results from high-quality clinical trials have demonstrated that EVT may be beneficial and not harmful for patients with very low ASPECTS below 3, as determined by NCCT or DWI [22]. Moreover, despite its greater accuracy, the DWI-ASPECT score is likely to overestimate the estimated ischemic core when compared with the CT-ASPECT score in the hyperacute phase [27]. A symptomatic seizure may aggravate a brain lesion assessed on DWI. Finally, decision-making in the acute stroke phase is not solely based on the infarction core. In this case, the occlusion site in the terminal intracranial carotid artery may represent a larger ischemic penumbra than the ASPECTS would suggest, and EVT might provide substantial benefit. Younger age (<60 years) is also a significant factor associated with better outcomes and fewer complications following thrombectomy. Based on these promising benefits, we decided to perform EVT for the patient. One limitation of this case report is that we lacked the data from neuroimaging on day 30 from stroke onset, so we cannot evaluate the final ischemic core.

In conclusion, we report a case involving the reversibility of an extremely large DWI ischemic lesion following EVT, which was associated with good short- and long-term clinical outcomes. This finding aligns with previous literature. The clinical implications of this phenomenon indicate that using DWI to evaluate the infarct core should be interpreted with caution. A large lesion on DWI may not be the sole factor in excluding patients from EVT. Given the growing evidence of the potential benefit of EVT in AIS with a large core, further studies are needed to evaluate prognostic factors related to a good functional outcome and identify patients who can truly benefit from endovascular treatment.

The authors express their gratitude to Dr. NQP for treating the patient and providing the data.

Written informed consent was obtained from the patient for publication of this case report and any accompanying images. Due to the retrospective nature of the study, ethics approval was not required in this case report. The study was conducted according to the Declaration of Helsinki.

The authors have no conflicts of interest to declare.

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article. T.Q.N. was funded by the PhD Scholarship Programme of Vingroup Innovation Foundation (VINIF), code VINIF.2023.TS.139. The funder had no role in the design, data collection, data analysis, and reporting of this study.

T.Q.N., D.N.C., T.V.S.L., R.A.-Q.N., and H.N: conceptualization, methodology, and literature review supervision, H.Q.D: manuscript preparation, literature review preparation, and project administration, T.H.N: project supervision, literature review edition, manuscript review, and final approval, H.Q.D: case description, manuscript review, and final approval. All the authors have read and agreed for publication of the current version of the manuscript.

Huy Quang Dang and Trung Quoc Nguyen contributed equally to this work.

All data generated or analyzed during this study are included in this article. Further inquiries can be directed to the corresponding author.