6 Left P2 Posterior Cerebral Artery Stenosis

Left P2 Posterior Cerebral Artery Stenosis

Clinical Presentation

A 25-year-old woman was admitted to a district general hospital with a disturbance affecting the right visual fields of both eyes and a right-sided hemihypesthesia. She had no history of migraine and had no other vascular risk factors other than using an estrogen-containing contraceptive pill. No headaches were reported. Ischemic strokes of the left occipital lobe and the left thalamus were diagnosed on MRI. No magnetic resonance angiography (MRA) was performed. Echocardiography, electrocardiogram (ECG), and transcranial Doppler (TCD) revealed normal findings. Laboratory workup demonstrated mildly raised levels of lipoprotein(a) and slight hyperhomocysteinemia. Antiplatelet therapy with aspirin was started. Four weeks later, the patient was admitted to our Emergency Department because of a subjective deterioration in her right-sided visual fields.

Initial Neuroradiologic Findings

MRI on the day of admission showed the known posterior cerebral artery (PCA) infarct in the left occipital region in addition to a small area of ischemia of the left thalamus, identical to the initial finding 4 weeks previously. T1-weighted images revealed a mild hyperintense signal in the region of the cortical PCA infarction, indicating a slight hemorrhagic transformation. Time-of-flight (TOF) MRA was suggestive of an occlusion of the left distal P2-PCA segment (Fig. B6.1 and Fig. B6.2).

Suspected Diagnosis

Hemorrhagic transformation of the known left-sided PCA infarction.

Questions to Answer by Ultrasound Techniques

Was there an occlusion or stenosis within the left PCA?
Was there evidence of vascular changes in the extracranial brain-supplying arteries, in particular within the vertebrobasilar system?

Initial Neurosonologic Findings (Day 2)

Extracranial Duplex Sonography

Assessment of the carotid and vertebral arteries (VAs) revealed normal findings. There was no evidence of atherosclerosis.

Transcranial Duplex Sonography

Normal and symmetric flow signals were seen in both anterior cerebral arteries (ACAs) and middle cerebral arteries (MCAs) (not shown). Flow velocities in both P1-PCA segments and the right P2-PCA segment were within the normal range. A distinct turbulent flow was evident in the left distal P2-PCA segment. Doppler spectrum analysis in this area revealed an increased flow velocity (156/75 cm/s) (Fig. B6.3, Fig. B6.4, Fig. B6.5, Fig. B6.6; see also Video B6.1).

Conclusion

Distal left P2-PCA stenosis of unknown origin.

Clinical Course

On MRI there was no evidence of subsequent ischemic events. The mild hemorrhagic transformation in the PCA infarct was considered to be the cause of the clinical deterioration. Neurosonologic examination demonstrated a stenosis in the distal left P2-PCA segment, which was probably overlooked during the initial TCD study 4 weeks previously. In light of the ultrasound findings, the small residual MRA vessel signal in the projection of the left distal P2-PCA segment was thought to result from the weak blood flow distal to a high-grade stenosis or to belong to the superior cerebellar artery. As the only known potential vascular risk factors were a mildly raised level of lipoprotein(a), a mild hyperhomocysteinemia, and the use of an estrogen-containing contraceptive pill, an in-situ thrombus was suspected. Because of the hyperhomocysteinemia the patient was prescribed folic acid. Furthermore, we recommended that she stopped taking the combined contraceptive pill. Aspirin therapy for secondary stroke prevention was continued as no new ischemic event had occurred. Repeated clinical and ultra-sound follow-up over a 3-year period demonstrated a stable neurologic status and unchanged ultrasound findings.

Fig. B6.1 Left: MR FLAIR image, axial plane. Hyperintense ischemic lesions in the left occipital lobe as well as in the left thalamus (arrows). Right: T1-weighted image, axial plane. Mild hyperintense signals in the area of infarction, suggestive of hemorrhagic transformation (arrows).

Fig. B6.2 3D TOF-MRA, axial maximal intensity projection (MIP). Absent signal of the left distal P2-PCA main stem, suggesting high-grade stenosis or occlusion (large arrow). Note the weak vessel signal more distally (small arrows), probably corresponding to a PCA branch or to the superior cerebellar artery.

Fig. B6.3 TCCS (transtemporal approach), right-sided insonation, midbrain plane. Normal flow in the right P1-PCA (flow velocity 79/37 cm/s).

Fig. B6.4 TCCS (transtemporal approach), right-sided insonation, midbrain plane. Normal flow in the right distal P2-PCA (flow velocity 78/43 cm/s).

Fig. B6.5 TCCS (transtemporal approach), left-sided insonation, midbrain plane. Left P1-PCA shows a normal flow signal (flow velocity 54/29 cm/s).

Fig. B6.6 TCCS (transtemporal approach), left-sided insonation, midbrain plane. Intrastenotic flow signal in the distal left P2-PCA (flow velocity 156/75 cm/s).

Final Diagnosis

Left occipital PCA territory infarction with concomitant thalamic involvement, probably caused by an in-situ thrombus with residual left distal P2-PCA stenosis.

Discussion

Clinical Aspects

This patient is a 25-year-old woman with a left PCA infarction probably caused by a distal P2-PCA stenosis. In view of the absence of the classic vascular risk factors, the etiology of the stenosis remained unclear.

In the United States and Europe, 5–10% of stroke patients are less than 45 years of age (Jacobs et al 2002, Marini et al 2001). The incidence ranges from 11.3/100,000 per year in primarily white populations to 22.8/100,000 per year in black people (Kittner et al 1993, Kristensen et al 1997). Young stroke patients more frequently have cardiac embolism associated with a patent foramen ovale (PFO). Also, a hypercoagulable state, illicit drug use, and inherited blood clotting disorders are more frequently found in young stroke patients than in the older population (Pezzini et al 2003). One of the largest population-based studies from 15 European stroke centers (the 15 Cities Young Stroke Study) reported ~3,331 patients aged 15 to 49 years with first-ever ischemic stroke. According to the TOAST criteria the cause of stroke remained undetermined in 39.6% of cases, 17.3% had a cardioembolism, 12.2% small-vessel disease, 9.3% large-vessel disease, and 21.6% had another determined etiology. PFO was the most frequent cardioembolic cause with a proportion of 6.6% of all strokes in this series followed by cardiomyopathy (2.0%), PFO and atrial septal aneurysm (1.8%), and ventricular wall hypo- or akinesia (1.2%). In strokes of other determined etiologies, nonatherosclerotic noninflammatory and inflammatory arteriopathies, hematologic disorders, coagulopathies (genetic, acquired, or related to systemic disorders), and miscellaneous rare causes have to be mentioned. Cervical artery dissection was the most common cause with 12.8% of all strokes in the series, distantly followed by antiphospholipid syndromes (1.2%), systemic vasculitis (0.8%), hematologic diseases (0.6%), systemic lupus erythematosus (0.5%), primary angiitis of the CNS (0.5%), migrainous infarction (0.4%), illicit drug use (0.4%), moyamoya (0.4%), pregnancy or puerperium-related (0.3%), reversible cerebral vasoconstriction syndrome (0.2%), fibromuscular dysplasia (0.2%), hyperhomocysteinemia or homocysteinuria (0.2%), CADASIL (0.2%), mitochondrial disease (0.2%), and other rare causes like HIV-related vasculopathy in 0.1% (Putaala et al 2009, Yesilot Barlas et al 2013). For further reading on moyamoya disease, see Case 9); for cervical artery dissection, see Case 11 and Case 19; for fibromuscular dysplasia, see Case 13; for migrainous infarction, see Case 22; for pregnancy-related and reversible cerebral vasoconstriction syndrome, see Case 36; and for HIV-related stroke, see Case 17.

CADASIL (cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy) is a hereditary small-vessel disease without pronounced ultrasound findings but with a prolonged cerebral circulation time (Chabriat et al 2009, Liebetrau et al 2002). Fabry’s disease, a lysosomal storage disease also rarely related to stroke, was not specified in the 15 Cities Young Stroke Study. A separate large European multicenter trial including 5,023 patients aged 18–55 years revealed definite and probable Fabry’s disease as main cause of stroke in 0.5% and 0.4%, respectively (Rolfs et al 2013). Sickle cell disease, also not specified in the two European studies, is discussed in Case 43.

In our patient none of the classic vascular risk factors like arterial hypertension, diabetes mellitus, or hyper-lipidemia were present and she did not smoke. However, her homocysteine and lipoprotein(a) levels were mildly raised and she was taking an estrogen-containing contraceptive pill. Metabolic disorders like hyperhomocysteinemia and homocystinuria are associated with ischemic stroke in the young population (Mineyko and Kirton 2013, Sébire et al 2004). The prevalence of mild or moderately raised homocysteine levels ranges between 10% and 20%, depending on the nutritional status of the studied population. Hyperhomocysteinemia is considered to be a risk factor for the development of atherosclerotic vessel wall changes in large arteries. A raised serum homocysteine is further associated with a two- to threefold increased risk of stroke (Bostom et al 1999, Giles et al 1998). Several consecutive studies have demonstrated that homocysteine levels can be lowered by 10–15% if vitamin B and folic acid supplements are taken. No benefit of vitamin B and folic acid treatment in stroke patients with raised levels of homocysteine was reported, e.g., in the VISP (Vitamin Intervention for Stroke Prevention) study (Toole et al 2004). Also, a recent Cochrane review affirmed that lowering hyperhomocysteinemia will not decrease a patient’s cerebrovascular or cardiovascular risks for primary or secondary prevention (Martí-Carvajal et al 2015).

Lipoprotein(a) as an independent risk factor for stroke is also controversial. Some authors have reported higher mean lipoprotein(a) levels in stroke patients compared with controls (Pedro-Botet et al 1992). A prospective study demonstrated that the lipoprotein(a) level is an independent predictor of stroke and vascular death (Ariyo et al 2003). A more recent study found that raised lipoprotein(a) levels were associated with a higher incidence of ischemic stroke. However, this was true only for white women and nonwhites of both sexes (Ohira et al 2006). Others saw increased values related with an increased hereditary risk for suffering a vascular event. Its role as an independent stroke risk factor remains unclear (Kamstrup et al 2009). Most investigations of vascular change in relation to lipoprotein(a) have focused on the extracranial arteries. Data concerning intracranial atherosclerosis are scarce. One study reported an association with the extent of intracranial atherosclerotic vessel wall changes (Arenillas and Alvarez-Sabín 2005).

Female sex hormones used for contraception or for postmenopausal hormone replacement therapy increase the risk for vascular events including stroke. Since the first reports of an association between oral contraceptives and ischemic strokes (Vessey and Doll 1969), a large number of studies have addressed this issue. In these studies, oral contraceptives were confirmed to be an independent risk factor for stroke in young women (Chan et al 2004, Gillum et al 2000, Petitti et al 1996, WHO 1996). However, this risk is low and is probably further lowered by the current use of low-dose estrogens and third-generation progestogens. A review of progestogen-only oral contraceptives showed no increased risk of stroke (Chakhtoura et al 2009). But the risk can increase if other risk factors such as thrombophilia, age >30 years, smoking, hypertension, diabetes, migraine, and obesity are additionally present.

The combination of the three mentioned and discussed risk factors in our patient might have contributed to the development of an intracranial atherosclerotic lesion. However, as only one lesion was found, and because of the young age of our patient, atherosclerosis seemed unlikely and a local nonatherosclerotic vessel disease more probable.

The frequency of PCA infarctions reported by most stroke databases lies between 5% and 10% and is therefore lower than the incidence of MCA infarctions (Brandt et al 2000). This may be one reason why there are only a few large studies on PCA infarction and its etiology, and modes of clinical presentation are less well analyzed. Few clinical studies with large number of patients and detailed neurologic and cardiologic evaluation have been published (Brandt et al 2000, Steinke et al 1997, Yamamoto et al 1999). In these studies, embolic PCA infarctions occurred in 53–79% of cases. The majority of embolic events were of cardiac origin (28–41%), and to a lesser extent of artery-to-artery origin (22–32%). In-situ thrombi were seen in 8–16% of cases. Rare causes, such as coagulopathies, were found in 3–15% of cases while the cause remained unclear in up to 24% of cases. A study in a Korean population reported atherosclerotic macro-angiopathy in 42.4% of 205 patients as the most frequent cause of PCA stroke. Within this group, arterio-arterial embolism was postulated in 47.5%, perforator artery occlusion in 25.3%, an in-situ thrombosis in 12.6%, and a combination of arterio-arterial embolism and perforator occlusion in 7.8%, respectively. Of the patients with a macroangiopathic cause, a pathology limited to the PCA alone was seen in 18.5%. The ventrolateral thalamus was the most frequent infarct location, followed by occipital lobe infarction in patients with isolated macroangiopathic PCA lesions (E. Lee et al 2009). The high rate of involvement of the thalamus emphasizes that the occlusion of perforators at the level of the P1-PCA or proximal P2-PCA segment is a relevant stroke mechanism. In cases of artery-to-artery emboli, atherosclerotic vessel wall changes and dissections of the extracranial VA, followed by the intracranial VA and the basilar artery (BA), have been identified as the most common embolic sources (Yamamoto et al 1999). Occasionally, atherosclerotic lesions of the internal carotid artery (ICA) can also cause PCA infarction in those individuals with a fetal-type PCA variant (Steinke et al 1997).

The high percentage of observed macroangiopathy might have been caused by the higher prevalence of intracranial atherosclerosis in the Asian population as well as by a better and more detailed vascular analysis of the studied cases; however, the prevalence also seemed remarkable in a white population. In a Danish TIA population a PCA stenosis ≥50% according to the Baumgartner criteria was found in 11% of all intracranial stenoses (von Weitzel-Mudersbach et al 2012). An even higher prevalence of 30% was observed in an Italian study with TIA and stroke patients (Viaro et al 2012).

Besides the rarer conditions such as hypercoagulopathies or Sneddon’s syndrome, migraine has repeatedly been discussed as a potential cause of PCA ischemia. A migraine-associated vasospasm with secondary development of thrombi has been discussed, and its proportion was estimated to be as high as 10% of all PCA infarctions. One reason for this hypothesis is that PCA infarctions are accompanied by headaches in up to 50% of cases, unlike ischemic events in the anterior circulation (Brandt et al 2000, Pessin et al 1987). However, the most favored hypothesis at present is that PCA infarctions may trigger a migraine in those who are currently experiencing migraine attacks (Olesen et al 1993). Migraine as a basic underlying pathomechanism seems unlikely and is not supported by the available pathoan-atomic studies (Caplan 1991) (for further discussion on migraine and stroke, see Case 22).

Angiologic and Anatomic Aspects

The PCA can be subdivided into four different segments from P1 to P4 (for further information, see Chapter 2, “Posterior Cerebral Artery” under “Special Arterial Anatomy and Ultrasound Anatomy”). The pattern of PCA infarctions follows the anatomic paths of blood supply. The P1-and proximal P2-PCA segments mainly supply the paramedian midbrain and the medial and posterolateral thalamus via small perforating arteries. Relevant cortical PCA branches start in the midpart of the P2-PCA segment with a highly variable anatomy. Usually the anterior temporal artery is the first cortical branch followed by the occipitotemporal artery mainly supplying the middle and posterior parts of the basal temporal lobe. The following P3-PCA segment starts in the quadrigeminal cistern and quickly separates into the two main final branches, the parietooccipital and calcarine arteries, which supply the mesial parietal and occipital cortex, respectively. Again, variations of vessel courses and branching are more the rule than the exception. Depending on the location of occlusion or stenosis and the capacity to develop collateral pathways, typical infarcts and corresponding clinical pictures appear. Embolic PCA occlusions may therefore range from total PCA infarction to a circumscribed partial cortical/subcortical ischemia. The latter frequently occurs in the calcarine artery territory as emboli generally follow the most direct vessel pathways. Subsequently, visual disturbance is the most common symptom, occurring in up to 90% of cases. Involvement of the thalamus indicates involvement of perforating arteries. In our case an involvement of the P1-PCA and/or proximal P2-PCA segment was assumed. However, the stenosis detected was clearly distal to the origin of the thalamogeniculate or thalamoperforating arteries which can best be explained by a dynamic development of the vessel pathology. The first evaluation in our clinic was performed 4 weeks after the initial ischemic event. Therefore, an initial proximal occlusion, caused, for example, by an in-situ thrombus involving the perforator arteries, seems possible, followed by secondary partial recanalization. The observed secondary hemorrhagic transformation is another positive indicator of recanalization (Molina et al 2001) possibly coinciding with the reported secondary deterioration of the patient’s visual field.

Ultrasound diagnostics of the posterior intracranial circulation have considerably improved with the introduction of transcranial color-coded sonography (TCCS) in the early 1990s. Unlike the TCD approach, this enables the PCA, and in particular the P1-, P2-, and P3-PCA segments, to be reliably identified. In our case, an initial TCD examination in the first admitting hospital was normal. We suspect that the superior cerebellar artery (SCA) signal could have been mistaken for the PCA as the two vessels are closely related and flow velocities and flow profiles are comparable (Pade et al 2010). However, even TCCS carries the risk of such confusion because of the close vicinity of both vessels (Baumgartner et al 1999). For further details on insonation of the SCA, see Chapter 2, “Superior Cerebellar Artery” under “Special Arterial Anatomy and Ultrasound Anatomy.”

Despite the described advantages of TCCS, not much data exists on evaluation and quantification of P2-PCA stenoses. In analogy to their evaluation of MCA stenoses, Baumgartner and coworkers (1999) described flow velocity cut-off values for determination of ≥50% and <50% P1- and P2-PCA stenoses. Compared with DSA results, a systolic flow velocity ≥145 cm/s yielded a sensitivity, specificity, and positive and negative predictive values of 100%, 100%, 100%, and 91% for the detection of a ≥50% PCA stenosis, respectively, and a systolic flow velocity ≥100 cm/s yielded values of 100%, 100%, 100%, and 100% for the detection of a <50% PCA stenosis, respectively. Another study reported cut-off values of >200 cm/s systolic flow velocity for the detection of a P2-PCA stenosis (Kimura et al 2000). However, the authors used angle-corrected values in all patients, in contrast with the former group, which might explain the apparent difference between the two studies. An exact angle correction in intracranial vessels is often difficult to obtain because of the vessel elongations. Angle correction should therefore preferentially used in the midpart of a vessel segment within a straight vessel course of at least 10 mm. As this is seldom the case within the proximal course of the PCA, angle correction should not be attempted.

Scarce data are available comparing ultrasound techniques and MRA or CT angiography (CTA) for evaluation of proximal PCA occlusions and stenoses. In an Italian study of 292 symptomatic patients, a ≥50% PCA stenosis (according to the Baumgartner criteria) was confirmed by a second modality in approximately 30% of patients with an intracranial stenosis at any site (Viaro et al 2012). Similar high rates of ≥30% PCA stenosis (26% and 25%, respectively) in symptomatic patients were identified by CTA (Homburg et al 2011, Ovesen et al 2013). 3D TOF-MRA, however, is particularly prone to misinterpret low flow for occlusion, as could be seen in our patient, who was initially wrongly diagnosed as having a P2-PCA occlus ion.

Only gold members can continue reading. Log In or Register to continue