Introduction

Lacunar infarction (LI) accounts for up to 25% of acute ischemic strokes[1] and is defined as a small subcortical ischemic lesion with a maximal diameter <15 mm located in the basal ganglia, thalamus, internal capsule, corona radiata or brainstem caused by occlusion of a single penetrating artery due to lipohyalinosis or microatheroma[2]. The clinical presentation comprises five typical so-called lacunar syndromes (pure motor stroke, pure sensory stroke, sensorimotor stroke, dysarthria-clumsy hand syndrome, and ataxic hemiparesis)[3]. For unknown reasons, a deterioration in the clinical course can be observed in many patients with LI[4]. While LI has been assumed to be associated with a rather benign prognosis for a while[1], more recent studies demonstrated a higher mortality and risk of recurrent stroke in patients with LI in the longer perspective[5].

While multimodal MRI including magnetic resonance angiography (MRA), and perfusion-weighted imaging (PWI) can be used to characterize the underlying vessel pathology and degree of perfusion changes in large territorial ischemic stroke in detail, its ability to demonstrate corresponding changes in acute LI are limited due to the relatively small diameter of the penetrating arteries and the small lesion size. Only a few studies investigated PWI in LI and their results are inconsistent[6-14]. Furthermore, the question whether acute LI may benefit from acute treatment with recombinant tissue plasminogen activator (rtPA) has been a matter of debate as it was hypothesized that intravenous thrombolysis would only be effective in patients with large vessel occlusion[15,16].

In this case series of patients with acute LI admitted to our hospital we aimed to describe infarction patterns on DWI and perfusion patterns on PWI in the acute stage and at follow-up in order to provide a better understanding of its pathophysiology and to indentify ischemic lesions suited for intravenous thrombolysis.

Material and Methods

Results

MRI analysis

Diffusion weighted imaging demonstrated an acute ischemic infarction in the thalamus in 8 (50%), in the internal capsule in 4 (25%), in the corona radiata in 3 (18.8%) patients, and in the mesencephalon in 1 (6.3%) patient. The right side was affected in 6 (37.5%), and the left side in 10 (62.5%) patients. On initial DWI, the ischemic lesions had a mean volume of 0.46±0.29 cm³. On follow-up DWI, the ischemic lesions had a mean volume of 0.67±0.47 cm³. Growth of DWI lesions was observed in 7 (43.8%) patients, reversal of DWI lesions in 3 (18.8%) patients (for details see Table 1 and Figure 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 1. Three examples of perfusion patterns in acute lacunar infarction.

Mismatch (1), inverse mismatch (2), and match (3) between diffusion-weighted and perfusion-weighted images (DWI, PWI). DWI (a) shows the initial ischemic lesion (white arrow). PWI derived maps demonstrate the perfusion deficit: time to peak (b), cerebral blood flow (c), and cerebral blood volume (d). Follow-up DWI (e) and FLAIR (f) show the ischemic lesion (white arrow). Note that case 2 is also an example of lesion reversal after intravenous thrombolysis.

https://doi.org/10.1371/journal.pone.0077428.g001

Perfusion weighted imaging showed a circumscribed hypoperfusion in 10 patients (62.5%), whereas the other perfusion studies were normal. The hypoperfused areas had a mean volume of 0.61±0.48 cm³. In comparison to the corresponding DWI lesion size (0.47±0.30 cm³) there was no significant difference (p=0.50). The mean CBF and CBV ratios were 0.65±0.28 and 0.84±0.41 respectively. With regard to the predefined perfusion patterns, 3 patients demonstrated a match, 4 an inverse mismatch, and 3 a mismatch. Figure 1 shows exemplary cases of each perfusion pattern.

Reversal of the DWI lesion was observed only in patients with hypoperfusion on PWI irrespective of the perfusion pattern (3/10 vs. 0/6, p=0.25) and treated with intravenous thrombolysis (3/7 vs. 0/9, p=0.63). In detail there were 2 patients with match and 1 patient with inverse mismatch. Lesion growth was shown in 6 patients with hypoperfusion but only in one patient with unremarkable perfusion study (6/10 vs. 1/6, p=0.15) and was more common in patients who did not receive intravenous thrombolysis (5/9 vs. 2/7, p=0.36). In detail there were 3 patients with mismatch, 2 patients with inverse mismatch and 1 patient with match.

No significant differences in CBF (0.79±0.22 vs. 0.59±0.29, p=0.33) and CBV ratios (1.04±0.28 vs. 0.76±0.45, p=0.37) as well as hypoperfused brain tissue volumes (0.60±0.24 cm³ vs. 0.61±0.57, p=0.97) in patients with DWI lesion reversal were observed compared to those patients with stable or growing ischemic lesion size.

Discussion

The most striking result of the present study of DWI-PWI patterns in LI is that all patterns that are well-known in large territorial infarction (match, inverse mismatch, mismatch) can also be observed in patients with LI. In particular the finding of a DWI-PWI mismatch in three LI patients was unexpected as comparable cases are not described in the medical literature. Due to a minor clinical deficit all three patients were not treated with intravenous thrombolysis and two of them showed an early clinical deterioration which was paralleled by a marked infarction growth on follow-up MRI. Whether the different DWI-PWI patterns represent different stages in stroke evolution or various underlying etiologies, however, remains open and needs to be addressed in further studies. Another important finding is the lesion reversal in three patients with LI after intravenous thrombolysis. Thus, acute therapy with rtPA may be regarded as an effective treatment option in this stroke subtype.

Although the existence of a penumbra has been demonstrated in an animal model of LI[21], until today no reports on its widely used MRI surrogate, the DWI-PWI mismatch have been published. In general, the medical literature concerning combined DWI and PWI in LI is sparse and the few reported study results are controversial. In one of the first studies with 6 patients, the authors observed no perfusion deficit at all in LI[6]. In contrast to this, Chalela and colleagues reported on a single patient with LI on DWI and matching hypoperfusion on PWI successfully treated with intravenous thrombolysis. Treatment was followed by a clinical improvement and reversal of the initially observed DWI abnormalities on follow-up MRI[7]. Similarly, we observed a partial or complete reversal of lesions in 3 patients. In the same year, another small case series (n=6 patients) demonstrated perfusion abnormalities to be present in all patients, but in all cases the perfusion deficits were found to be smaller than the initial ADC lesion. In the course, an increase in size of ADC as well as PWI abnormalities were found. Perfusion deficits always remained smaller than the corresponding ADC lesion. In the present series, however, only 4 of 10 patients with perfusion deficit demonstrated an inverse mismatch. The authors hypothesized that these lesions spread beyond the hypoperfused area possibly induced by cytotoxic mechanisms at the border of the tissue damage[8]. The results were confirmed in a larger case series featuring 19 patients more recently by the same study group[13]. In thalamic infarction, a DWI-PWI mismatch was only found in stroke due to large vessel disease or cardioembolism but not in LI in 5 patients[9]. In another study on DWI and PWI in acute ischemic infarction, all patients with LI (n=4) had a perfusion deficit matching the DWI abnormalities[10]. In the only larger case series, hypoperfusion on PWI was not only observed in two thirds of LI (n=22) but could also be related to an early clinical deterioration. Hypoperfusion was not associated with infarction growth, nor with a less favorable clinical outcome at 3-month follow-up. This stands in contrast with our own observation of infarction growth in 7 patients with LI and perfusion deficit. However, the authors did not attempt to determine the presence of a DWI-PWI mismatch[11]. In the brainstem, a matching hypoperfusion is far less common in lacunar pontine infarction(n=5) than in pontine infarction due to large vessel disease[12].

Similar to these inconsistent and partly contradictory results, the question whether intravenous thrombolysis is beneficial in patients with LI still has not been answered convincingly. Several authors are of the opinion that intravenous thrombolysis might be effective only in large vessel occlusion[15] and more recently proposed an observational study in order to define more precisely the effectiveness of rtPA in acute ischemic infarction without large arterial occlusion[16]. This hypothesis may be supported by the finding that three fourths of ischemic stroke patients without angiographically detectable vessel occlusion in the acute phase finally had a favorable clinical outcome at 3-month follow-up[22]. Pathophysiological considerations may also challenge the clinical practice to treat LI patients with intravenous thrombolysis. In the 1960ies and 70ies Fisher demonstrated in his meticulous pathoanatomical studies that LI are caused by occlusion of small perforating arteries due to lipohyalinosis or microatheroma[2]. Thus, rtPA might be ineffective to recanalize the occluded vessel in LI since it is an enzyme activating fibrinolysis. Furthermore, as pointed out earlier some authors also hypothesize a different pathophysiological mechanism in LI compared to large vessel occlusion: Infarction growth may not be the consequence of hypoperfusion but instead of cytotoxic mechanisms at the border of the initial ischemic lesion[8,13].

However, given the low mortality of LI, most of autopsies took place months to years after the index event and as a consequence, the observed changes may represent an already organized vascular lesion[23]. In the end, the pathoanatomical appearance of the vascular lesion and the underlying pathophysiological processes in acute LI remain unknown and the efficacy of rtPA cannot be excluded a priori.

The only available data from a large clinical trial derives from the National Institute of Neurological Disorders and Stroke study (NINDS) in which 81 patients with LI were included. Patients who received rtPA were more likely to have a favorable clinical outcome in comparison to patients who received the placebo[24]. Furthermore, several case series and larger observational studies addressed the important question of rtPA safety and effectiveness in acute LI. In a case series, Lee and colleagues found that hemorrhagic complications are unusual in patients with LI (n=13) treated with rtPA[25]. In a cohort study, patients with LI (n=101) treated with thrombolysis had the best clinical outcome of all stroke subtypes and did not suffer from symptomatic intracerebral hemorrhage[26]. Fuentes and colleagues did not observe a difference in 3-month clinical outcome between LI (n=60) and other stroke subtypes[27]. Correspondingly, in the largest cohort study published 2012 no difference in treatment response and a low rate of symptomatic intracerebral hemorrhage were observed in LI (n=195) compared to other stroke subtypes[28]. Compared to the reported frequency of LI (up to 25%), the rates of thrombolysed LI patients are rather low, ranging from 4 to 11%. There are several possible explanations for the reluctance to treat patients with LI. First, LI is still thought be associated with a benign prognosis in general[29]. Second, clinical presentation can be rather mild and deterioration may occur only later in the clinical course[4,30]. Finally, pathophysiological considerations might also restrain stroke neurologists from treating acute LI with intravenous thrombolysis. With regard to this ongoing debate, the findings of our study may support the argumentation in favor of an acute treatment in patients with LI[16].

The present study has some limitations. First, this is a retrospective study of moderate size since only patients who underwent initial MRI within 6 hours after symptom onset and follow-up MRI within one week were included. Follow-up MRI was primarily performed in patients who underwent intravenous thrombolysis or demonstrated clinical deterioration and this may explain the high thrombolysis rate of 43.1% in this study. Second, although all included patients underwent follow-up MRI, only 4 patients had follow-up PWI. Thus, we cannot draw any conclusions about the evolution of perfusion abnormalities with time. Third, characterization of penumbra in LI with multimodal MRI is difficult due to the small lesions, the limited spatial resolution of PWI and its susceptibility to motion artifacts. As a consequence punctuate hypoperfusion might be missed in small lacunar lesions. While the manual definition of the ROIs for measurement of the perfusion deficits may be considered to be a drawback of this study, a similar approach was used in other earlier studies [6,12]. Although we cannot exclude an under- or overestimation of PWI lesion size by this approach, there are no generally accepted or recommended PWI quantitative analysis approaches in LI. Even automated delineation of PWI lesions may result in exaggerated volumes as recently reported by Galinovic and colleagues [31]. Fourth, the diagnosis of LI is based on the clinical syndrome, characteristic MRI findings, and the exclusion of competing etiologies like cardioembolism or large vessel disease. Since the underlying vascular lesion could only be determined by neuropathological examination, some uncertainty remains regarding the final diagnosis. Finally, a detailed comparison of patients treated with intravenous thrombolysis and those without acute treatment was not performed due to the moderate study size. Consequently, the results have to be interpreted cautiously and should not be regarded as proof of efficacy of thrombolysis in acute LI.

In conclusion, multimodal MRI including DWI and PWI is feasible to demonstrate a mismatch in acute LI as well as growth or reversal of an ischemic lesion and consequently may serve for a better characterization of this stroke subtype and support treatment decisions in daily clinical practice.