Cerebral deep venous thrombosis presenting as acute micrographia and hypophonia

Case report.

A 62-year-old woman developed small handwriting and a soft voice in the setting of a new, dull vertex headache. She was brought to the hospital after an episode of decreased responsiveness associated with a low-frequency tremor of the right arm that resolved spontaneously. She had no significant prior medical history and was taking estrogen and progesterone hormone replacements. She occasionally used a nonsteroidal anti-inflammatory medication for rare muscle contraction-type headaches. A few days before presentation, she resumed taking this medication for her new head discomfort. A brain CT 6 months before presentation for evaluation of the headaches had shown no abnormalities.

On examination, she was afebrile and had a blood pressure of 108/53. Her general examination was normal, and she was well hydrated. On mental status examination, her digit span was normal, and her speech was fluent without paraphasic errors. However, she was hypophonic and unable to yell. Her writing was smaller than normal (figure 1). When asked to copy items, she drew them significantly smaller than the model (see figure 1). On attempting to copy a spiral, she was unable to “swing out” and became stimulus bound, despite continued reminders to get “bigger” (see figure 1). Her affect was flat as confirmed by direct questioning about her mood. She was appropriately concerned about her situation when she was told that she had a stroke. She was bradyphrenic in her conversation. The patient’s cranial nerves were normal. Her motor examination showed no pronator drift, but slight right upper motor neuron weakness (5-/5 strength in deltoid, triceps, finger extensors, and hamstrings). There was no cogwheel rigidity or tremor. The patient was slightly bradykinetic in facial movements, limb movements, and gait. She reported some transient slight tingling on her right face, arm, and leg, but she had normal sensation to primary modalities as well as stereognosis and graphesthesia. Her coordination was normal. Gait was notable for reduced arm swing and small steps. Reflexes were symmetrically brisk and toes were downgoing.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1. Items above the scale were drawn at presentation, and those below the scale were drawn 1 month later. Two tick marks represent 1 cm. (A) Writing samples. (B) Copy of a cube demonstrating initial marked reduction in size. (C) Copy of a spiral figure.

A brain CT was normal except for slightly increased density at the midline edge of the upper midbrain tegmentum, which may have reflected thrombus in the vein of Galen. There were multiple abnormalities on a diffusion-weighted MRI scan including increased signal in bilateral thalami (left more than right), bilateral putamenae, and caudate nuclei (more confluent on the right), as well as areas suggestive of punctate hemorrhage in the right caudate nucleus and left thalamus on the susceptibility-weighted images (figure 2). Head magnetic resonance venogram (MRV) (see figure 2C) showed multiple venous blockages, including absent flow in the internal cerebral vein, vein of Galen, straight sinus, left transverse sinus, left sigmoid sinus, and left internal jugular vein. The left subclavian vein and superior vena cava were patent on chest MRV.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2. (A) MRI at presentation showing multiple T2 abnormalities. (B) Diffusion-weighted sequence demonstrating marked signal changes in deep structures. (C) Magnetic resonance venography (MRV) showing absent flow in the internal cerebral venous system. (D) A T2-weighted MRI 1 month later showing a few areas of infarction. (E) Susceptibility-weighted sequences demonstrating multiple areas of punctate hemorrhage. (F) MRV 1 month later demonstrating improved flow in the internal cerebral venous system.

Routine blood chemistry and hematology were normal. An extensive workup for malignancy was unremarkable. Her thyroid-stimulating hormone and erythrocyte sedimentation rate were normal. No clotting abnormalities were identified (i.e., normal protein C, S, factor V Leiden, homocysteine). The patient’s antithrombin 3 was slightly lower than normal. The patient’s anticardiolipin immunoglobulin G was elevated, but antinuclear antibody was negative, as was the lupus anticoagulant.

The patient’s hormonal replacement was discontinued, and she was started on phenytoin for seizure prophylaxis and heparin, which was later converted to warfarin, to inhibit clot propagation. The MRV was repeated 3 weeks later, and flow in the internal cerebral vein and vein of Galen had improved (see figure 2F). Furthermore, not all of the areas that had abnormal diffusion signal had evolved into infarcts. There was residual T2 hyperintensity in the right basal ganglia and centrum semiovale on the right and on the left thalamus (see figure 2D). The patient’s handwriting had returned to its normal size, and she was able to copy items closer to their original dimensions (see figure 1). Her attempts to draw a spiral were also clearly improved (see figure 1).

Discussion.

Our patient shows that micrographia and hypophonia can occur acutely and may suggest cerebrovascular disease. Often these findings will only be found with careful observation because they may be dwarfed by more dramatic findings. However, occasionally, a patient may spontaneously report only small handwriting and “soft speech.”

Micrographia is believed to be a consequence of inappropriately “weak” movement commands in the execution of an overlearned, automatic action. It is interesting that in this patient external auditory cues (instructing the patient to draw a bigger spiral loop) were not helpful in increasing the size of the handwriting, although this did seem to assist in follow-up. Oliveira et al. reported that when patients with PD were given auditory reminders (“write big”), performance improved.1 These authors hypothesize that micrographia is the consequence of the supplementary motor area dysfunction found in PD (“open loop” dysfunction), which can be overcome by making the task less automatic. They also reported that visual cues were helpful in activating the “closed loop” of the lateral premotor cortex (by decreasing task automaticity). Our patient’s performance did not improve when provided with the direct cue of her last drawn item on a serial copying task. However, this difference could be related to differences in the underlying pathology (vascular in our patient versus an underlying degenerative process).

Micrographia and hypophonia are common symptoms of parkinsonian patients presenting to movement disorders or behavioral neurology clinics. Although common, the exact frequency of these phenomena are unknown.2 In a retrospective chart review one series reported that 24.6% of patients with vascular parkinsonism noted the onset of their symptoms with stroke.3 Micrographia and hypophonia may be present in a subset of these patients. It has been reported previously that the dysarthria–clumsy hand syndrome can occur with micrographia.4 Other reported causes of this syndrome include lenticular hematoma,5 thalamomesencephalic infarction,6 thalamocapsular tumor,7 and MS.8 Our study demonstrates that micrographia and hypophonia might be presenting complaints and practically the sole finding of extensive cerebral vein thrombosis.

Deep venous systems receive inflow from both thalami and basal ganglia. Infarctions have been noted to preferentially occur in these regions with venous thrombosis.9

The clinical course of this patient demonstrates a successful outcome of anticoagulation for cerebral venous thrombosis, despite punctate hemorrhagic lesions. This supports the safety findings of a recent study with low–molecular-weight heparin.10

It is surprising that our patient had so few other findings on her neurologic examination given the severity of lesions suggested by the neuroimaging. The lack of progression of areas of restricted diffusion to infarction indicates that the initial abnormalities were most likely caused by edema and not brain ischemia. MRV is now able to detect dural sinus and cerebral vein occlusions even in patients with relatively minor clinical findings.