The present case illustrates that pure sensory lacunar infarcts, most often located in the ventral posterior tier nuclei which is the main sensory relay nuclei to the cerebrum [1, 8], should not be considered benign. Sensory symptoms that may appear as harmless may easily be ignored, but acute sensory symptoms defined with a remarkable midline split, especially when both head and trunk are involved, is often considered unique to thalamic or thalamocortical lesions , and augur a minor stroke. Bilateral thalamic lesions are in general uncommon . However, bilateral paramedian thalamic infarcts, due to occlusion of the artery of Percheron, are well known , but lateral ‘mirror lesions’ as seen in the present patient must be considered very rare. Combined with a Déjerine–Roussy syndrome , the case is probably unique. The only case we have found in the literature was that of Cordery and Rossor (1999) who described a patient with bilateral thalamic pain secondary to bilateral thalamic infarcts, but the lesions were not well-located and not demonstrated radiologically .
Establishing exact stroke mechanisms is very difficult. Despite thorough investigations, only plausible mechanisms may be launched in the majority of cases. The thalamus is supplied by very small arteries susceptible to slowly evolving lipohyalinosis and microthrombosis . The principal inferolateral arteries, arising from the P2 branch of the posterior cerebral artery, supply the major part of the thalamic sensory nuclei . In the present case, no focal atherosclerotic posterior artery disease was seen on CT-angiography and, as expected, we were not able to identify the principal inferolateral arteries. The size of the thalamoperforating branches of the posterior cerebral arteries varies from 100 to 400μm in diameter , and the artery of Percheron has only a few times been successfully demonstrated on conventional angiography .
On the basis of the clinical history, risk profile, the absence of findings of large artery disease and potential cardiac embolic source, small vessel disease was the most plausible cause of the lacunes in the present case. One might only speculate if an occlusion due to thrombosis in situ of either an unpaired (anomalous) inferolateral artery or two separate arteries occurred. Diabetes seems to be a strong predictor for relatively early recurrence of a lacunar infarction in the wake of a first ever lacunar stroke . One might also argue that there is increased risk of a new sensory stroke due to the fact that thalamoperforating branches of the posterior cerebral arteries in general are of smaller caliber than the lenticulostriate vessels , but this hypothesis has to be confirmed in future studies. However, in a follow-up study of 695 patients who had a first ever lacunar infarction , the mean time for recurrence was 58 months (range 2–240, n=122) . No patient in that study experienced a new lacune within the first 2 months. The odds of getting bilateral, almost identical infarctions from two different and small vascular territories within 2 weeks as in the present case must thus be considered microscopic.
A good outcome after a pure sensory stroke is reported in the literature to be the rule [3, 9, 14], and reassurances are often given. However, the occurrence of central post-stroke pain is particularly high when the VPN and the trigeminal spinal nucleus and tract are involved (as in lateral medullary syndrome) . Amitriptyline is the best documented drug for central post-stroke pain, gabapentin is an alternative, but any drug may give poor results . Despite a combination of the two mentioned above and an opioid analgesic, the patient experienced severe pain. However, her pain was reduced when she was in activity, and actually buprenorphine in double dose (10μg/hour) made here tired and inactive. Due to this, she did not want to change her medical regime, but supplied the treatment with 300 to 600mg gabapentin when needed.
Combining clinical information with radiological findings, one may outline a rather precise location of the infarctions in the present case. Hypersensitivity to pinprick and thermal stimuli indicates increased activity in the spinothalamic tracts. The ventroposterolateral nuclei relay the stimuli from the trunk and extremities, and the ventroposteriomedial nuclei relay the stimuli from the head and face to the primary sensory cortex of the parietal lobe . The small ataxic component and the normal position sense found provide reasons to believe that the infarctions also may have affected areas that convey cerebellar fibers to motor-related cortices . Studying the MRI images, using both an atlas  and reference structures , there is no doubt that there is good correlation between clinical and radiological findings. This case thus shows that the thalamus plays a pivotal role in central pain, but whether it acts as a generator or as a defect modulator of ascending inputs is not clear. Spontaneous pain with an intact spinothalamic tract may support the former, but hypersensitivity to touch may support the latter.