Article Text


Impact of posterior communicating artery on basilar artery steno-occlusive disease
  1. J M Hong1,
  2. J Y Choi1,
  3. J H Lee1,
  4. S W Yong1,
  5. O Y Bang2,
  6. I S Joo1,
  7. K Huh1
  1. 1
    Department of Neurology, Ajou University School of Medicine, Suwon, South Korea
  2. 2
    Department of Neurology, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, South Korea
  1. Correspondence to Dr J M Hong, Department of Neurology, School of Medicine, Ajou University, 5 San, Woncheon-dong, Yongtong-gu, Suwon-si, Kyunggi-do, 442-749, South Korea; dacda{at}


Background: Acute brainstem infarction with basilar artery (BA) occlusive disease is the most fatal type of all ischaemic strokes. This report investigates the prognostic impact of the posterior communicating artery (PcoA) and whether its anatomy is a safeguard or not.

Methods: Consecutive patients who had acute brainstem infarction with at least 50% stenosis of BA upon CT angiography (CTA) were studied. The configuration of PcoA was divided into two groups upon CTA: “textbook” group (invisible PcoA with good P1 and P2 segment) and “fetal-variant of PcoA” group (only visible PcoA with absent P1 segment). Baseline demographics, radiological findings and stroke mechanisms were analysed. A multiple regression analysis was performed to predict clinical outcome at 30 days (modified Rankin disability Scale (mRS⩽2)).

Results: Among all 95 patients, 58% (n = 55) had good prognoses (mRS⩽2). Interestingly, 44 patients (46.3%) had at least one fetal-variant PcoA (26 bilateral, 18 unilateral). By multiple logistic regression analysis, the atherosclerotic mechanism (OR 18.0; 95% CI 3.0 to 107.0) and presence of fetal-variant PcoA (OR 5.1; 95% CI 1.4 to 18.8) were independent predictors for good prognosis and initial NIH stroke scale score (OR 1.24 per one-point increase; 95% CI 1.1 to 1.4) for poor prognosis.

Conclusions: Fetal-variant PcoA appears to act as a safeguard against ischaemic insult in acute stroke victims involving the brainstem with BA occlusive disease. This result can be explained by the fact that patients with fetal-variant PcoA have a smaller area of posterior circulation and a possibility of retrograde filling into the upper brainstem through the fetal-variant PcoA.

Statistics from

When an artery in the brain is blocked by acute ischaemic insult, collateral vessels can be used as a “detour” around the blockage. Sufficient collateral flow is a “protector” of ischaemic penumbra against ischaemic insult.1 This arterial network has two compartments: the primary collaterals (ie, circle of Willis), which are used in acute ischaemia, and the secondary collaterals (ie, ophthalmic, leptomeningeal and intracerebellar vessels), which develop when the primary collaterals are insufficient.2 3 Therefore, the posterior communicating artery (PcoA) can play an important role as a primary collateral pathway.

Basilar artery (BA) occlusive disease is often associated with a high mortality and a poor prognosis.4 5 The clinical importance of PcoA has been addressed in relation to anterior circulation stroke.6 Therefore, we explored whether the anatomical status of PcoA can predict short-term clinical outcomes in patients with acute brainstem infarction due to severe BA occlusive disease.



A retrospective study was done with patients who were enrolled in the Ajou Stroke Registry between May 2002 and Jun 2008. The study was confined to patients with a symptom onset within 7 days. We included the following eligible patients if they had (1) brainstem or cerebellar signs on admission, (2) a final diagnosis of acute ischaemic stroke with a brainstem DWI confirmation, and (3) complete occlusion or at least 50% stenosis of the BA on CT-based angiography (CTA). All included patients underwent diagnostic studies that included cerebral vascular studies (ie, CTA), routine blood tests and cardiological work-up (electrocardiogram, echocardiogram). We excluded patients whose BA occlusion could be recanalised by thrombolytic therapy. This study was approved by the Institutional Review Board at Ajou University Hospital.

Imaging analysis

Brain MR scanning (1.5 T, GE Medical, Milwaukee, Wisconsin) with conventional T2 and DWI was conducted in the axial plane, with 5 mm thick sections. CTA images were obtained using with a 16 multidetector row spiral computed tomography scanner (Sensation 16, Siemens, Germany). Following the reconstructive CTA technique, the anatomy of PcoA was divided into (1) “textbook type” (no visible PcoA with good P1 and P2 segment) and (2) “fetal-variant PcoA type” (visible PcoA with an absent P1 segment).7 8 9 We calculated infarct volume by multiplying the lesion area per slice by section thickness. The degree of stenosis was defined as the narrowest vessel diameter divided by a normal diameter of the vessel.10 We designated ⩾50% stenosis of the BA as significant BA occlusive disease. To compare the PcoA configuration of the patients with those of the age-matched populations, we evaluated brain CTA of patients >30 years without stroke history who visited outpatient clinic from January 2007 to June 2007.

Short-term prognosis and stroke mechanisms

All patients were evaluated according to the National Institute of Health Stroke Scale score (NIHSS) and modified Rankin Disability Score (mRS) on admission, on days 1, 3 and 7 after admission, before hospital discharge, after 30 days via face-to-face or telephone interview. First, the clinical outcome was dichotomised into good (mRS 0–2) and poor (mRS 3–6) using a 30-day mRS after admission. Second, the outcome of patients was redistributed as excellent (mRS 0–1), minor deficit (mRS 2–3), major deficit (mRS 4–5) and death (mRS 6). Based on the results of clinical, vascular and cardiological studies, we divided the patients’ stroke mechanisms into two groups by the New England Medical Center Posterior Circulation Registry classification: (1) atherosclerosis or (2) embolism.5

Statistical analysis

Differences between groups were analysed using the Student t test and χ2 test, for continuous and categorical variables. Potential predictors for outcome were entered into a univariate logistic regression model: (1) age, sex and stroke mechanisms; (2) stroke risk factors: history of hypertension, diabetes, currently smoking and previous stroke; (3) potential radiological variables: brainstem only (infarct confined within brainstem) or brainstem plus (infarct in brainstem and elsewhere) in the involving lesion, and stroke volume; and (4) angiographic factors: degree of BA stenosis or occlusion, presence or absence of fetal-variant PcoA. Potential predictors that were not significant (p>0.2) in the univariate analysis were deleted from the full multivariate model. A stepwise backward conditional method was performed in the final multivariate model. p Values <0.05 were deemed statistically significant.


Among all 2517 consecutive patients admitted to our stroke centre with acute ischaemic stroke or TIA, there were 130 patients with BA occlusion or severe stenosis (⩾50% stenosis). Thirty-five patients were excluded due to the following: no lesion on DWI, n = 19; no CT angiography, n = 7; success of BA recanalisation, n = 9, leaving a total of 95 patients who met our criteria. Cerebral vascular studies comprised CT angiography (n = 95), additional MR angiography (n = 17) and distal subtraction angiography (n = 26). The mean age of the patients was 65.9 (SD 10.6) years, and 36 (37.9%) were female. We compared the configuration of PcoA between patients (n = 95) and control group (n = 75) who were matched in age (64.3 (9.4) years, p = 0.269) and sex (female 40%, p = 0.780). Compared with the control, a significant number of patients with BA steno-occlusive infarct had at least one fetal-variant PcoA (46.3 vs 28.0%, p = 0.015), along with bilateral fetal-variant PcoA (27.4 vs 2.7%, p<0.001).

Table 1 shows details of the baseline features of the patients with the significant BA occlusive disease according to the presence of fetal-variant PcoA. Age, gender and stroke risk factors were not different according to the configuration of the Willisian circle. However, baseline NIHSS and infarct volume were significantly higher (p = 0.001) and larger (p = 0.001) in the textbook group. The proportions of the atherosclerotic mechanism (70.2 vs 93.2%, p = 0.005) and brainstem only lesion (31.4 vs 54.5%, p = 0.023) were also significantly higher in the fetal-variant group. The short-term outcome according to mRS favours the fetal-variant PcoA group. There was a greater proportion of excellent or minor disability deficits in the fetal-variant PcoA group, as compared with the textbook group (p<0.001).

Table 1

Clinicoradiological findings according to Willisian circulation configuration and univariate association predictive of good prognosis (mRS 0–2)

To predict the good outcome after 30 days, a final multivariate model was done after univariate analysis (table 1). The atherosclerotic stroke mechanism (OR 18.01; 95% CI 3.03 to 106.95, p = 0.001) and the “fetal-variant PcoA type” configuration of the Willisian circle (OR 5.11; 95% CI 1.39 to 18.83, p = 0.014) were independently associated with good outcome. However, initial NIHSS (OR 1.24 per one-point increase; 95% CI 1.12 to 1.38, p<0.001) was independently associated with poor outcome.


This study illustrates the significance of the fetal-variant PcoA (fetal-type variant of the Willisian circle), which is a common and independent determinant for predicting short-term prognosis of brainstem infarct patients with significant BA occlusive disease

We have shown that the fetal-variant PcoA was detected in 46% of all patients. There was a significantly higher proportion in the patient group when compared with the control. In particular, the bilateral fetal-variant PcoA showed a significant prevalence. Current studies on the prevalence of fetal-variant PcoA showed that 11–29% of the study participants had a unilateral fetal-variant PcoA, and 1–9% of normal healthy subjects or patients with stroke had a bilateral hyperplastic PcoA.3 Moreover, 18.5% of Asian patients with stroke had a fetal-variant PcoA.11 As a result, the fetal-variant PcoA appears to be a common configuration of the Willisian circle in patients due to brainstem infarction with significant BA occlusive disease.

We have found a new radiological factor, the presence of fetal-variant PcoA, to be an independent short-term predictor for all patients with brainstem stroke. Fetal-variant patients can be less vulnerable to ischaemic insult in the brainstem or cerebellum. The safeguard role of the fetal variant PCoa can be explained both anatomically and physiologically (fig 1). Anatomically, fetal-variant PcoA inevitably has no bilateral or unilateral posterior cerebral arterial flow from the BA and has a smaller posterior circulation territory, as compared with the textbook group. Physiologically, there is a possible development of a secondary collateral flow in the upper brainstem via the primary collateral flow of the fetal-variant PcoA. The fetal-variant PcoA can contribute to the posterior circulation by branching off short interpenduncular or long mensencephalic penetrating arteries to the midbrain or thalamus.12 13 14

Figure 1

Plausible defence mechanisms in the subject with a fetal-variant PcoA against the BA occlusive stroke. (A) Subject with the normal configuration of the circle of Willis. (B) Subject with the configuration of the fetal-variant of the circle of Willis, who can be less vulnerable to an ischaemic insult because of (i) a smaller territory supplied from the BA and (ii) a possibility to develop secondary collateral flows (arrows) into the upper brainstem via fetal-variant PcoA.

It is still unclear whether the fetal-variant of PcoA is congenital or acquired. Liebeskind et al asserted that such a primary collateral flow is a dynamic feature rather than a static feature, illustrating a case with a great change in PcoA configuration during a short-term period after BA occlusion.15 Because time is a critical variable in the development of collateral circulation, progressive or chronic obstruction of BA flow resulting from atherosclerosis can allow for more extensive development of collateral flow via PcoA.2 This study documented that the atherosclerotic stroke mechanism was more common than the embolic mechanism in the BA steno-occlusive patients with fetal-variant PcoA. Consequently, fetal-variant PcoA may be an acquired form as well as a congenital form, particularly in patients with ischaemic stroke with BA occlusive disease.

In conclusion, the presence of fetal-variant PcoA on CTA appears to be a safeguard mechanism to patients with brainstem stroke due to acute BA occlusive disease, whereas fetal-variant PcoA was more common in patients with BA occlusive disease. Therefore, PcoA may act as a double-edged sword: with the first prediction that subjects with fetal-variant PcoA might be prone to the BA occlusion because of reduced flow in the posterior circulation and with the second prediction that it might be beneficial in a stroke victim with acute brainstem infarction.


View Abstract


  • Competing interests None.

  • Ethics approval Ethics approval was provided by Ajou Institutional Review Board.

  • Provenance and Peer review Not commissioned; externally peer reviewed.

Request permissions

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.