Cognitive and Behavioral Abnormalities in a Case of Central Nervous System Whipple Disease FREE

WHIPPLE disease (WD) was first described by Whipple in 1907.¹ It is a rare and chronic systemic illness caused by a gram-positive bacillus, Tropheryma whippelii.² It usually affects middle-aged men and is typically characterized by a variety of digestive symptoms, including chronic diarrhea, weight loss, and malabsorption.^{3 - 4} Digestive symptoms may be preceded for several months by other symptoms, most often arthralgia, although other systems may also be affected during the course of the disease.⁵

Central nervous system (CNS) involvement occurs during the course of the disease in 10% to 43% of patients with WD.^{6 - 7} However, CNS involvement is rarely the initial manifestation (in fewer than 5% of patients).^{8 - 9} When the CNS is involved, the most common symptoms are cognitive and psychiatric changes, supranuclear gaze palsy, upper motor neuron signs, and hypothalamic dysfunction.⁷ Psychiatric symptoms can include depression, paranoia, and psychosis.¹⁰

Only 11% of patients with CNS WD have demonstrated cognitive impairment in the absence of other associated neurologic signs and symptoms.⁷ To date, no studies have reported serial evaluations by standardized neuropsychological methods in such patients to quantify what domains are involved and to what extent. We describe a patient who was examined because of a several-week history of changes in mental status before the diagnosis of CNS WD. We examined him on several occasions during a 7-month period by means of carefully standardized, quantitative measures to document his cognitive and behavioral functioning.

In June 1995, a 41-year-old, right-handed man with a high school education who was employed as a heavy equipment operator saw his local physician because of a 1-month history of diarrhea and a 9-kg weight loss. He had further signs of conjunctival inflammation, peripheral edema, and pleural effusions, but had no neurologic signs at that time. Laboratory investigations determined that he was anemic and had hepatitis C. Inflammatory bowel disease of unknown etiology was diagnosed, and the patient was treated with sulfasalazine, 500 mg every 6 hours. His gastrointestinal tract symptoms improved with this treatment, and he was apparently doing well until early April 1998.

During a 3-week period beginning in April 1998, he again suffered a 9-kg weight loss and also had marked and progressive mental status changes, which included apathy, confusion, paranoia, visual and auditory hallucinations, and decreased memory. It was initially thought that he was severely depressed, because he had been arrested for driving while under the influence of alcohol on February 28, 1998. He was treated with zolpidem tartrate, with no relief of his symptoms. He was admitted on April 23, 1998, to a local hospital for assessment of a possible overdose of zolpidem. Physical examination showed the patient to be slender and disoriented, but results were otherwise unremarkable and without evidence of neurologic or gastrointestinal signs. Computed tomographic scans with and without contrast were normal, and an electroencephalogram indicated diffuse slowing. Results of drug and alcohol screenings were negative. He was then transferred to The University of Iowa Hospitals and Clinics, Iowa City, for further evaluation on April 28, 1998.

On transfer, the patient was drowsy and oriented only to self. He often confabulated. Recall was 0/3 after a brief delay. Bedside testing at that time showed that language was grossly intact, and he was able to copy a figure of intersecting pentagons, although it took him 10 minutes to complete. His visual fields were full and his pupils were normally reactive. His extraocular movements were normal. Strength, tone, reflexes, and sensation were all normal. Results of coordination testing were normal. Gait was normal for heel, toe, and tandem walking. Romberg sign was absent. He was initially thought to have herpes simplex encephalitis and was treated with intravenous acyclovir. Analysis of the cerebrospinal fluid showed an opening pressure of 265 mm H₂O, a total protein level of 1.22 g/L (reference range, 1.5-4.5 g/L), glucose concentration of 3.1 mmol/L (56 mg/dL) (reference range, 3.6-6.1 mmol/L [65-110 mg/dL]), 0 ×10¹²/L red blood cells, and 5 ×10⁹/L white blood cells (0.4 histiocytes and 0.6 lymphocytes). Results of microbiological studies were all normal, as were those of all other serological tests. Polymerase chain reaction analysis of the cerebrospinal fluid obtained on May 4, 1998, was negative for herpes simplex virus. The patient's condition continued to deteriorate; on May 5, 1998, he developed marked hypothermia, with a temperature of 31.4°C, and became responsive only to painful stimuli.

A magnetic resonance (MR) image of the brain and brainstem obtained on May 7, 1998, showed areas of low signal intensity on T1-weighted images and high signal intensity on T2-weighted and fluid attenuated inversion recovery (FLAIR) images within the midbrain and mesial temporal lobes, orbitofrontal cortex, caudate, putamen, and globus pallidum bilaterally, with abnormal signal extending superiorly periventricularly. There was relative sparing of the thalamus. A biopsy of the right temporal lobe of the brain was performed the same day. The specimen showed areas of periodic acid–Schiff–positive foamy histiocytes with abundant coccobacillary organisms identified by Steiner silver stain. Polymerase chain reaction–based analysis on paraffin-embedded tissue from this biopsy confirmed the presence of T whippelii. Subsequent duodenal endoscopy and biopsy showed edema and inflammation in the atrial bulb and second portion of the duodenum with superficial erosions and patchy gastritis. Sigmoidoscopy gave normal results. Pathological examination of the duodenum showed occasional periodic acid–Schiff–positive foamy histiocytes consistent with the diagnosis of WD.

THE PATIENT was treated with ceftriaxone sodium, 2 g intravenously every 12 hours for 2 months (May 12 through July 11, 1998), followed by improvement in his overall condition. He became increasingly alert and interactive with family and staff, without any significant physical impairments. Of note, at no point during his illness did he manifest ocular motility abnormalities or involuntary movements such as a myorhythmia. His severe cognitive and behavioral impairments persisted, however, and are discussed below.

A follow-up MR image obtained on May 24, 1998 (Figure 1), showed changes related to the right temporal lobe biopsy but was otherwise virtually identical to the MR image from May 7. The mesial temporal cortices and orbitofrontal cortices continued to be involved, along with the caudate and inferior aspects of the lentiform nuclei and, to a lesser degree, the midbrain. An MR image obtained November 24, 1998 (Figure 1), showed a normal-appearing midbrain. All other previously involved structures showed a lesser degree of abnormal signal, though the mesial temporal lobes remained prominently abnormal, especially on the right.

Our patient was examined briefly on 3 occasions during his hospitalization (on May 18, May 29, and June 10, 1998). Standardized neuropsychological tests were used to assess intellect, memory, speech and language, visual perception, constructional praxis, visuomotor functioning, and executive functioning. These tests included subtests from the Wechsler Adult Intelligence Scale III,¹¹ the Wechsler Memory Scale III,¹² the Multilingual Aphasia Examination,¹³ and the Boston Diagnostic Aphasia Examination¹⁴ ; the Benton Visual Retention Test,¹⁵ Rey-Osterreith Complex Figure Test,¹⁶ Boston Naming Test,¹⁷ Trail-Making Test,¹⁸ and Facial Recognition Test¹⁹ were also administered. His performances were compared with findings from appropriate normative groups. The results of these evaluations are given in Table 1.

The patient's neuropsychological manifestations were notable for mild improvement during the course of his hospitalization. He had severe deficits in several domains of cognition, including orientation to time and personal information, sustained attention, constructional praxis, speeded information processing, and executive function. He also demonstrated a marked anterograde amnesia, accompanied by prolific confabulation of the type often observed in basal forebrain and diencephalic amnesia.^{20 - 21} His basic language skills were well preserved, and he was able to understand test instructions and conversed fluently with the examiners. Visual perception and confrontation naming were also relatively preserved.

Apart from his severe cognitive defects, the patient exhibited strikingly odd behavior, and rampant confabulation and delusional thinking. He frequently started conversations with staff and visitors, the content of which would range from relatively benign topics, such as the weather, to those with prominently sexual overtones. When asked why he was in the hospital, his replies varied greatly, examples of which included "I'm here for research," "I'm part of the construction crew and they're housing us here," and "I decided to take a few days off." The patient was often observed to mimic the doctors' routines, and managed to disassemble a hospital computer during the night shift.

He returned for follow-up testing in the chronic epoch of the illness on August 31 and November 24, 1998. At that time he had not returned to work and had moved from a local convalescent center to live with his father. More extensive neuropsychological evaluations on these dates included the aforementioned tests as well as the Rey Auditory Verbal Learning Test,¹⁸ Wisconsin Card-Sorting Test,²² Judgment of Line Orientation,¹⁹ Grooved Pegboard Test,²³ and a rotor pursuit task.²⁴ His neuropsychological findings were notable for considerable improvement in attention and emotional control; however, he continued to demonstrate marked impairments in anterograde memory, working memory, and executive function, with little discernible change from August to November. The Wisconsin Card-Sorting Test had to be discontinued during the August evaluation because of his frustration with the task. He was able to complete the test during the November session, albeit defectively. On the rotor pursuit task he demonstrated a relatively flat learning curve across 5 learning trials but had normal retention after a 1-hour delay. On the Grooved Pegboard Test he had a borderline performance with his left (nondominant) hand, in contrast to an average score with his right hand. Verbal intellect and language were within normal limits. His ex-wife noted that, since his discharge from the hospital in July, he had exhibited marked disinhibition, lack of initiative and planning, poor frustration tolerance, and a lack of insight into his condition. The results of these evaluations are also shown in Table 1.

Whipple disease is a rare systemic disease that is generally manifested in gastrointestinal tract symptoms. Many of the features of WD are similar to those of inflammatory bowel diseases; although neurologic symptoms, such as dementia, supranuclear palsy, and gait ataxia, are not uncommon during the course of the illness, these features are rarely the predominant or initial features of the disease. Our patient had an unusual manifestation in that the predominant factor in his being referred for further medical evaluation was a severe cognitive disturbance in the absence of other associated neurologic signs or symptoms and in the apparent absence of active gastrointestinal tract symptoms (on admission, abdominal findings were normal and he had normal bowel function).

The pattern of dysfunction shown herein is reminiscent of some other acute neurologic conditions (eg, herpes simplex encephalitis, aneurysmal rupture of the anterior communicating artery). Patients with similar conditions are often left with severe memory and behavioral impairments as a result of dysfunction in mesial temporal lobes and basal forebrain structures.^{20 ,25}

His neuropsychological and behavioral profile was entirely consistent with the areas of abnormality noted by neuroimaging. Serial MR images demonstrated prominent abnormal signal in the mesial temporal lobes and basal forebrain bilaterally as well as heavy involvement of the striatum, pallidum, and hypothalamus with relative sparing of the thalamus.

These findings are consistent with previously published reports detailing the pathological findings in patients with CNS WD. In these studies, prominent involvement of the anterior and mesial temporal lobes, caudate, putamen, pallidum, amygdala, hypothalamus, and periaqueductal gray matter have been noted,^{26 - 29} in addition to variable reports of more widespread involvement of the cortical gray matter.^{30 - 31}

To our knowledge, relatively few cases of CNS WD including MR imaging findings with varying results have been reported. Rajput and McHattie³² described a patient with CNS WD and normal MR images of the brain and brainstem on 2 occasions 4 years apart. By contrast, Wroe et al³³ reported a case in which MR imaging showed widespread contrast-enhancing lesions scattered throughout the hemispheres. Pollock et al³⁴ presented a case with similar findings shown by computed tomography. Erdem et al³⁵ reported a case in which MR imaging showed areas of abnormal signal on T2-weighted images in both frontal lobes and the left temporal and occipital regions and multiple widespread gadolinium-enhancing foci. Others have reported MR imaging findings more consistent with our own case, with lesions involving the upper brainstem, mesial temporal lobes, striatum, hypothalmus, and frontal lobes.^{9 ,36 - 38} Verhagen et al⁶ described a patient with a 1-cm right thalamic lesion that halved in size with antibiotic (sulfamethoxazole-trimethoprim) treatment. Detailed neuropsychological data were not reported, however, in this patient.

In our case, we have presented serial neuropsychological and MR imaging findings. While our patient's cognitive profile fits well with the areas of abnormality noted on MR images, the improvement noted in the appearance of his most recent MR image is clearly much greater than that observed clinically. Our interpretation of these findings is that, while the CNS infection has responded to antibiotic therapy, a substantial amount of neuronal death may have occurred, leaving the patient with permanent impairments.

Finally, we would like to reiterate the diagnostic challenges this case presented. Several salient features of the clinical appearance were reminiscent of herpes simplex encephalitis or of a paraneoplastic syndrome. As the findings and conclusions presented herein indicate, the clinician should also keep in mind the possibility of CNS WD when confronted with signs and symptoms such as those manifest in our case.

Accepted for publication July 14, 1999.

This study was supported by program project grant NS 19632 from the National Institute on Neurological Disorders and Stroke, Bethesda, Md.

Reprints: Daniel Tranel, PhD, Department of Neurology, 200 Hawkins Dr, Iowa City, IA 52242.