T2-Weighted Hyperintense MRI Lesions in the Pons in Patients With Atherosclerosis
Background and Purpose Pontine hyperintense lesions (PHL) on T2-weighted MRI have been recognized recently. Histopathological findings resemble periventricular leukoaraiosis, and a vascular etiology has been suggested. We studied the frequency and the associated factors of PHL in patients with symptomatic atherosclerosis.
Methods Two independent observers assessed brain MRIs in a prospective cohort of patients with symptomatic atherosclerosis. Only patients in whom both observers scored PHL on T2- and proton density–weighted images, but not on T1-weighted images, were considered to have the lesion.
Results We studied 229 patients: 31% presenting with ischemic stroke, 31% with myocardial infarction, and 38% with peripheral artery disease. Both observers scored PHL in 23% of all patients. Patients with PHL were significantly older and had more lacunar infarcts and periventricular leukoaraiosis than patients without PHL. There were more women, more hypercholesterolemic and diabetic patients, and more cortical infarcts on MRI (P=NS). After logistic regression the presence of leukoaraiosis (odds ratio, 2.4; 95% confidence interval, 1.6 to 3.4) and lacunar infarcts (odds ratio, 2.2, 95% confidence interval, 1.5 to 3.1) remained independently associated with PHL. PHL was more common in patients with ischemic strokes (39%) than in patients with myocardial infarctions (11%) or peripheral artery disease (19%) (P<.001).
Conclusions We found that PHL on T2- and proton density–weighted MR images are often found in patients with symptomatic atherosclerosis. The association with periventricular leukoaraiosis and lacunar infarcts suggests that PHL is a variant of leukoaraiosis, with possibly the same pathophysiology. The clinical symptoms and consequences of PHL, however, are not yet clear.
In 1995 Pullicino et al1 reported 16 cases with “ischemic rarefaction” of the pons in a series of 85 patients examined with an MRI. They examined two cases histopathologically. The pontine areas that were hyperintense on T2W MRI showed white matter pallor with reactive astrocytosis, primarily in the central parts of the pons, with arteriosclerotic changes in the small arteries. The authors also saw lacunar infarcts in and around these areas. The histopathological findings were similar to those found in LA.2 3 Patients with pontine rarefaction more often suffered from hypertension and diabetes mellitus than patients without these lesions, but the differences were not statistically significant. The clinical relevance of pontine rarefaction is not clear.
The findings in the two cases of Pullicino et al1 suggest a vascular etiology. We therefore expected to find PHL on T2W MRI more frequently in a population of atherosclerotic patients. We examined the frequency of PHL in a cohort of atherosclerotic patients and studied its association with vascular risk factors and other manifestations of vascular brain disease.
Subjects and Methods
Patients with symptoms of atherosclerosis were recruited prospectively from January 1992 to February 1996 in two teaching hospitals with both regional and secondary referral functions. We recruited patients with recent ischemic stroke, myocardial infarction, or peripheral artery disease because these diseases are the most important clinical manifestations of atherosclerosis. Recent ischemic stroke was defined as an acute focal neurological deficit persisting at least 1 week. Hemorrhage was ruled out by a CT scan. The type of stroke was determined according to the criteria of Bamford et al.4 Severity was graded on a clinical scale, as follows: mild stroke was defined as symptoms or signs that did not cause impairment in ADL; moderately severe stroke was defined as paresis of arm or leg that caused moderate impairment in ADL or dysphasia with residual verbal communication; and severe stroke was defined as paresis of arm or leg or both that caused severe impairment of ADL or a severe communication disorder. Myocardial infarction was defined as characteristic angina pectoris for more than 20 minutes and laboratory or electrocardiographic findings characteristic of myocardial infarction. Peripheral artery disease was defined as intermittent claudication with characteristic leg pain on walking that disappeared with rest or a history of intermittent claudication treated with vascular surgery. Patients with dementia were excluded.
The study was approved by the medical ethics committees in both hospitals. All patients gave their informed consent. We recorded smoking status and medical history. Hypertension was defined as a diastolic blood pressure of ≥95 mm Hg on repeated prior measurements or current treatment for hypertension. Hypercholesterolemia was defined as a total plasma cholesterol level of ≥7 mmol/L or current treatment for hypercholesterolemia.
MRI was performed with a Siemens Magnetom 63 SP/4000 (1.5 T). Transversal T1W, T2W, and PDW images were made. Two investigators (V.I.H.K., B.V.) independently assessed the MRI scans without knowledge of clinical or laboratory data.
We defined PHL as confluent hyperintense areas on T2W, PDW, or both images, without distinct borders, but with normal or slightly hypointense corresponding signals on T1W images. Lacunar infarcts were defined as well-demarcated hyperintense lesions on T2W images with corresponding demarcated hypointense lesions of <2-cm diameter on T1W images. With these definitions the agreement for the presence or absence of PHL between the two observers was good (κ=0.63). We considered patients to have PHL only if both observers agreed on its presence. In all other cases PHL was judged absent.
We assessed hemispheric white matter lesions (LA) on the T2W and PDW images. LA was graded according to an adapted version of the scoring system of Fazekas et al.5 Periventricular hyperintensities limited to small caps and pencil-thin linings were considered normal.
The presence or absence of PHL was related to other clinical and MRI characteristics. Frequency differences were analyzed with the χ2 test with Yates’ correction for small sample sizes. Continuous data were analyzed with the Student’s t test or the Mann-Whitney U test, when appropriate. Since we made multiple comparisons, a value of P≤.005 (0.05/10) was considered significant. This analysis was also performed in the different vascular subgroups. Last, we compared patients with PHL only to patients with neither LA nor PHL.
We examined the association between PHL and a number of variables with multivariate analysis using a logistic regression model. We assigned the presence or absence of PHL as dependent variable and the other data of interest as independent variables using a stepwise forward selection strategy. For this analysis age was dichotomized according to the median (65 years). Patients with missing values were excluded from the analysis. A second logistic regression analysis was performed in patients without supratentorial LA to compare patients with PHL only to patients with neither LA nor PHL.
We analyzed 229 patients in this study (Table 1⇓). Thirty-one percent had an ischemic stroke, 31% a myocardial infarction, and 38% peripheral artery disease. The mean age was 61.8 years (SD, 11.7). Most patients were male (68%). Forty-one percent had hypertension, and a majority were smokers (86%). Approximately half had hypercholesterolemia, and only a few patients were diabetic. On the MRI many patients showed LA (39%) and lacunar infarcts (46%), while only a minority had cortical infarcts (12%).
Of the patients with ischemic stroke, 60% had symptoms of a (sub)cortical infarct and 40% of a lacunar infarct. Most of them (76%) had a moderately severe stroke, and a minority had a mild (17%) or a severe (7%) stroke. All stroke patients had hemispheric infarcts, except for 3 with brain stem signs. Among the stroke patients 15% had a history of transient ischemic attack or minor stroke, while among the myocardial infarction and peripheral artery disease patients 6% and 5% had such a history, respectively. Conversely, 9% and 11% of the stroke patients had a history of myocardial infarction or peripheral artery disease, respectively.
In 52 cases (23%), both observers independently agreed on the presence of PHL (Figs 1⇓ and 2⇓). In 16 patients there was disagreement; these patients were added to the control group. The 3 patients with brain stem signs all had PHL and also had a lacunar infarct in a different location: 2 approximately 10 mm caudally and 1 only unilaterally, while the PHL was seen bilaterally. Thirty-two patients (14%) had both PHL and supratentorial LA. Thus, 20 patients (9%) had PHL without evidence of LA, 119 patients (52%) had neither LA nor PHL, and 58 patients (25%) had LA only.
Patients with PHL were significantly older and more often had lacunar infarcts and LA than patients without PHL (Table 1⇑). There were more women and more hypercholesterolemic and diabetic patients, but fewer smokers (not statistically significant). The frequencies of hypertension and cardiac disease were similar in patients with and without PHL. Cortical infarcts were seen more frequently in patients with PHL, but this was not statistically significant. The presence of PHL was significantly correlated with the severity of LA (P<.001, Mann-Whitney U test).
After logistic regression, only the presence of LA (OR, 2.4; 95% CI, 1.6 to 3.4) and lacunar infarcts (OR, 2.2; 95% CI, 1.5 to 3.1) remained independently associated with PHL.
To clarify the differences between the symptomatic groups, we analyzed the frequencies of the vascular risk factors and the MRI findings within the subgroups (Table 2⇓). PHL was more frequent in patients with ischemic strokes (39%) than in patients with myocardial infarctions (11%) or peripheral artery disease (19%) (P<.001, χ2 test). In all subgroups patients with PHL were older, but this was statistically significant only in those with peripheral artery disease. The only other factor significantly associated with PHL was the presence of LA in patients with ischemic stroke. The other factors did not differ significantly within the subgroups, but there were more lacunar infarcts in patients with PHL within the stroke and myocardial infarction groups.
We analyzed the patients with PHL only (without supratentorial LA) and contrasted them with patients with neither LA nor PHL. In a univariate analysis, only age was significantly correlated with PHL (P<.001, Student’s t test). The presence of lacunar infarcts was weakly correlated with PHL (P=.02, χ2 test). In a logistic regression analysis, age was the only factor independently associated with isolated PHL (OR, 1.09; 95% CI, 1.03 to 1.15).
In a population of patients with different manifestations of atherosclerosis, we found a substantial proportion of patients with PHL (23%). We used a strict definition of PHL to be as specific as possible. Good interobserver agreement was achieved with our definition of PHL. The observed proportion of PHL was comparable with an earlier study.1 To our knowledge the first report on MRI PHL was published in 1993 by Watanabe et al (in Japanese).6
PHL on T2W MRI may be caused by multiple sclerosis, neoplasms, central pontine myelinolysis, infarcts, and wallerian degeneration. It is highly unlikely that one of these conditions caused PHL in our patients. They did not have a history of multiple sclerosis, cancer, severe alcoholism, or electrolyte disturbances. Pontine infarcts would leave corresponding sharply demarcated, hypointense signals on the T1W images. This was only seen (in a different location than the PHL) in three patients with clinical signs of brain stem infarct. We found no evidence for wallerian degeneration, such as lesions that are more rostrally located in the pyramidal tract or other long tracts.
Pullicino et al1 suggest that PHL is a variant of LA and that PHL is caused by hypoperfusion. In two of their patients the histopathological features of PHL resembled those of LA. The central part of the pons may be particularly vulnerable for effects of hypoperfusion. As shown in a recent study, this area corresponds to a vascular border zone, supplied by the anteromedial, lateral, and posterior groups of small penetrating arteries arising from the basilar and superior cerebellar arteries.7 A similar border zone exists in the periventricular areas between the vascular territories of the long penetrating pial arteries and the lenticulostriate arteries.8 LA is probably caused by atherosclerosis of small arteries and arterioles of the brain.9 10 11 12 13 Hyaline deposits and intimal atheroma formation may narrow the orifices of the penetrating arteries and the lumen of small arteries. This so-called small-vessel disease is likely to impair the microcirculation. When a substantial part of the small arteries becomes affected, the tissue in the border zones of these arteries may become compromised, leading to demyelination.8 Arteriolosclerosis, expressed by increased wall thickness of the arterioles, is strongly associated with LA.3 Occlusion of a single small artery results in a lacunar infarct.8 This type of infarct occurs frequently in patients with LA.14
We postulate that atherosclerosis in the small arterioles of the pons leads to PHL in the central pontine areas by the same mechanism as in LA. Two findings support this hypothesis. First, in our patients LA was associated with PHL, and the severity of LA correlated with the presence of PHL. This may point to a common pathophysiological mechanism, namely small-vessel disease. Second, lacunar infarcts are also associated with PHL. As mentioned above, this type of infarct is probably also caused by a small-vessel disease. In addition, Pullicino et al1 provided pathological evidence for this hypothesis when they noted “typical” arteriosclerotic hypertensive changes of the small arteries in the pons, although they did not describe these changes in detail.1
Several case-control studies show that the strongest risk factor for LA is age.15 In our patients with PHL (without supratentorial LA), age was the only factor independently associated with PHL. Hypertension was also associated with LA and was likely to be important in its pathogenesis.15 16 17 A prior history of stroke, myocardial infarction, and diabetes mellitus has been variably reported in patients with LA.15 16 18 19 20 21 We found more hypercholesterolemia and diabetes mellitus in patients with PHL, but this was not statistically significant. Pullicino et al1 found higher frequencies of hypertension and diabetes mellitus in patients with PHL, but this was not statistically significant.
Within the symptomatic subgroups, the only factors associated with PHL were age (in patients with peripheral artery disease) and LA (in patients with stroke) (Table 2⇑). The higher frequencies of LA and lacunar infarcts (although the latter was not statistically significant) within the stroke group could explain the higher prevalence of PHL in our stroke patients. In fact, we found a larger proportion of patients with lacunar infarcts than in other series of stroke patients,4 probably because a number of patients with large infarcts could not give consent, were not able to undergo an MRI, or died before MRI could be performed.
To summarize, we found that patients with different clinical manifestations of atherosclerosis often have PHL on MRI T2W and PDW images. The association with periventricular leukoaraiosis and lacunar infarcts suggests that PHL are a variant of leukoaraiosis with possibly the same pathophysiology. This is supported by histopathological findings in a previous study on this subject.1 We believe that the term “pontine leukoaraiosis” is appropriate for these lesions. The clinical significance of PHL, if any, is not clear. Some authors speculate that these patients may have neurological symptoms,1 6 but this remains to be demonstrated in clinical studies.
Selected Abbreviations and Acronyms
|ADL||=||activities of daily living|
|PHL||=||pontine hyperintense lesions (on T2- and proton density–weighted images)|
This study was supported in part by Sanofi Winthrop. The Amsterdam Vascular Medicine Group consists of the following: Academic Medical Center: H.R. Büller, MD, PhD, Center for Hemostasis, Thrombosis, Atherosclerosis and Inflammation Research; R.J.G. Peters, MD, PhD, Department of Cardiology. Slotervaart Hospital: J.J. van der Sande, MD, PhD, Department of Neurology; D.P.M. Brandjes, MD, PhD, Department of Internal Medicine; R.H. Bakker, MD, Department of Cardiology. Amstelveen Hospital: J.A. Lawson, MD, PhD, Department of Surgery.
Drs Blok and Stam are involved in research that is partially funded by Sanofi Winthrop. Drs Stam and Kwa collaborate in an international trial that is funded by Boehringer Ingelheim.
- Received February 4, 1997.
- Revision received May 5, 1997.
- Accepted May 6, 1997.
- Copyright © 1997 by American Heart Association
van Swieten JC, van den Hout JH, van Ketel BA, Hijdra A, Wokke JH. Periventricular lesions in the white matter on magnetic resonance imaging in the elderly: a morphometric correlation with arteriolosclerosis and dilated perivascular spaces. Brain. 1991;114:761-774.
Fazekas F, Chawluk JB, Alavi A, Hurtig HI, Zimmerman RA. MR signal abnormalities at 1.5 T in Alzheimerís dementia and normal aging. AJNR Am J Neuroradiol. 1987;8:421-426.
Watanabe M, Takahashi A, Arahata Y, Motegi Y, Furuse M. Clinical significance of pontine high signals identified on magnetic resonance imaging [in Japanese]. Rinsko Shinkeigaku. 1993;33:721-725.
Tatu L, Moulin T, Bogousslavsky J, Duvernoy H. Arterial territories of human brain: brainstem and cerebellum. Neurology. 1996;47:1125-1135.
Caplan LR. Binswanger’s disease: revisited. Neurology. 1995;45:626-633.
Kawamura J, Meyer JS, Terayama Y, Weathers S. Leukoaraiosis correlates with cerebral hypoperfusion in vascular dementia. Stroke. 1991;22:609-614.
Kobari M, Meyer JS, Ichijo M, Oravez WT. Leukoaraiosis: correlation of MR and CT findings with blood flow, atrophy, and cognition. AJNR Am J Neuroradiol. 1990;11:273-281.
Terayama Y, Meyer JS, Kawamura J, Weathers S, Mortel KF. Patterns of cerebral hypoperfusion compared among demented and nondemented patients with stroke. Stroke. 1992;23:686-692.
Hijdra A, Verbeeten B Jr, Verhulst JA. Relation of leukoaraiosis to lesion type in stroke patients. Stroke. 1990;21:890-894.
Pantoni L, Garcia JH. The significance of cerebral white matter abnormalities 100 years after Binswanger’s report: a review. Stroke. 1995;26:1293-1301.
Matsushita K, Kuriyama Y, Nagatsuka K, Nakamura M, Sawada T, Omae T. Periventricular white matter lucency and cerebral blood flow autoregulation in hypertensive patients. Hypertension. 1994;23:565-568.
Breteler MMB, van Swieten JC, Bots ML, Grobbee DE, Claus JJ, van den Hout JHW, van Harskamp F, Tanghe HLJ, de Jong PTVM, van Gijn J, Hofman A. Cerebral white matter lesions, vascular risk factors, and cognitive function in a population-based study: the Rotterdam Study. Neurology. 1994;44:1246-1252.
Lindgren A, Roijer A, Rudling O, Norrving B, Larsson E, Eskilsson J, Wallin L, Olsson B, Johansson BB. Cerebral lesions on magnetic resonance imaging, heart disease, and vascular risk factors in subjects without stroke: a population-based study. Stroke. 1994;25:929-934.
Ylikoski A, Erkinjuntti T, Raininko R, Sarna S, Sulkava R, Tilvis R. White matter hyperintensities on MRI in the neurologically nondiseased elderly: analysis of cohorts of consecutive subjects aged 55 to 85 years living at home. Stroke. 1995;26:1171-1177.
Jorgensen HS, Nakayama H, Raaschou HO, Olsen TS. Leukoaraiosis in stroke patients: the Copenhagen Stroke Study. Stroke. 1995;26:588-592.