Complex Partial Seizures Present Diagnostic Challenge

Temporal lobe epilepsy (TLE), now more commonly called complex partial seizure disorder so as to include seizures that originate in the frontal foci, straddles the borderland between psychiatry and neurology. Since the condition may involve gross disorders of thought and emotion, patients with temporal lobe epilepsy frequently come to the attention of psychiatrists. But since symptoms may occur in the absence of generalized grand mal seizures, physicians may often fail to recognize the epileptic origin of the disorder. Indeed, misdiagnosis and failures of diagnosis are common in TLE. Fortunately, the illness is marked by certain "signature" symptoms that can aid in its identification.

John Hughlings Jackson observed in the late 1800s that seizures originating in the medial temporal lobe often result in a "dreamy state" involving vivid memory-like hallucinations sometimes accompanied by déjà vu or jamais vu (interpreting frequently encountered people, places or events as unfamiliar). Jackson wrote of "highly elaborated mental states, sometimes called intellectual aura," involving "dreams mixing up with present thoughts," a "double consciousness" and a "feeling of being somewhere else." While the "dreamy state" can occur in isolation, it is often accompanied by fear and a peculiar form of abdominal discomfort associated with loss of contact with surroundings, and automatisms involving the mouth and GI tract (licking, lip-smacking, grunting and other sounds).

Experiential Illusions

In the 1940s and '50s, Wilder Penfield, a neurosurgeon at the Montreal Neurological Institute, artificially elicited "dreamy states" by cortically stimulating the lateral temporal neocortex, the anterior hippocampus or the amygdala in awake epileptic patients prior to their surgical resections. During these operating room experiments, the patients experienced what Penfield referred to as "experiential illusions."

These illusions involved an alteration, sometimes subtle, of the person's relationship to his or her environment, as well as emotional response to it. In contrast to psychotic persons, Penfield's patients remained fully aware that their altered interpretation was an illusion. A friend's voice may sound remote, or a well-known living room may appear unfamiliar, but the meaning is preserved, the voice does not become depersonalized, nor does the living room lose its identity. Even those patients describing feelings of unreality state that they know at the same time what reality is, observed Penfield. This is an important distinction from schizophrenia and other psychotic states.

For example, a patient of mine with temporal lobe epilepsy often feels compelled to stare for brief moments at a coffee table in her living room since, as she put it, "It just doesn't look exactly like my coffee table." After a few seconds, the feeling disappears. At no time does she think that the table has actually changed; the only thing that varies is her perception and "interpretation" of it.

Altogether Penfield (Mullan and Penfield 1959) divided the illusions of interpretation into four groups:

Auditory illusions accompanied by the perception that sounds were louder or clearer, fainter or more distinct, nearer or farther away;

Visual illusions where things seemed clearer or blurred, nearer or farther away, larger or smaller; fatter or thinner;

Illusions of recognition where present experience seemed familiar, strange, altered or unreal; and

Illusions of emotion consisting of feelings of fear, loneliness, sorrow or disgust.

None of these groups of symptoms are unique to epilepsy. Migraine sufferers regularly experience illusions of sound, sight, taste and smell.

True hallucinations-those without external stimulus-may occur in complex partial seizures, especially the classic olfactory or gustatory hallucination seen with uncinate fits.

In a psychiatric or neuropsychiatric practice, the most commonly encountered illusions of interpretation are those of emotion. Typically, these are sudden in onset and unrelated to conscious experiences or anything in the environment:

A 40-year-old patient with a 20-year history of mental illness had been diagnosed at various times as having either schizophrenia or borderline personality disorder. She regularly experienced sudden episodes of dread that occurred without warning and without relationship to any inner experiences or anything happening around her.

On one occasion while waiting for a cab to come to her home to take her and her husband to an event they had both been looking forward to attending, she experienced such an overwhelming sense of dread she sank to the floor with a compulsion to hurt herself. Because of the strong visceral component to her symptoms and the inexplicable sudden change in mood, a neurological workup was ordered. An electroencephalogram (EEG) "showed the presence of an intermittent epileptiform disturbance confined to the right anterior and mid-temporal regions. EEG findings would be compatible with partial complex seizure disorder." On carbamazepine (Tegretol), she has been free of the seizures for 10 years.

(Drug information on carbamazepine)

In most instances, the emotion experienced as part of the seizure is a disturbing one variously described as dread or a feeling of impending doom; in others, the emotion may be experienced as pleasant or euphoric, as Dostoyevsky described. Since the feelings can arise de novo without any identifiable precipitant, an incorrect diagnosis of an acute panic attack may be entertained. Almost always, however, the patient will describe additional experiences that will help in the differential diagnosis. Included here is a strong visceral component to the symptoms: a feeling or sensation, almost always unpleasant, traveling upward from stomach to head. In an attempt to explain the experience, the patient will sweep his or her hands upward starting at the abdomen. Descriptions such as "a wave," "something flowing upward" are often employed.

Often such details must be elicited by careful, tactful questioning, because the patient will be reluctant to describe the experience; its intensity and bizarre nature arouses fears of insanity. The physician frequently can sidestep this reaction by asking: "Have your episodes ever involved anything strange?" with a lack of emphasis on the word strange, thereby suggesting that strange experiences are not at all unusual with these kinds of seizures. Another approach is to say: "On occasion persons who have experienced some of the things you have told me about have described other experiences they have been reluctant to discuss because they were afraid other people, even their doctors, might think them crazy." Questions about specific epileptic experiences should be delayed until the end of the interview to avoid suggestibility.

TLE Personality?

Controversy continues as to the validity of a so-called temporal lobe personality. Certainly, many of the patients tend to be obsessive and over-inclusive in their thinking, often satisfying some or all of the requirements for obsessive-compulsive personality: hyperphogia may be seen in some patients. Their speech and thinking is "viscous" and ponderous with a tendency toward loquacity and the insistence on the elaboration of fine and often tedious distinctions. Outbursts of irritability, rather than frank violence, are hallmarks of TLE.

When interviewing suspected TLE patients, it's important to inquire about their birth and any complications of the pregnancy. Forceps deliveries, now almost unheard of, were quite common years ago and led to compressive injuries of the brain, anoxic damage to Ammon's horn in the hippocampus and the subsequent temporal lobe epilepsy. Also ask about generalized seizures, head injuries, concussions, temper tantrums and, with males, a history of aggression, fire setting, truancy and impulsive behaviors. Has the patient experienced frequent déjà vu, jamais vu, depersonalization, autoscopy or sudden mood swings accompanied by visceral or oral sensations? Do others complain that the patient often doesn't seem to be listening, appears to be daydreaming or otherwise preoccupied? Often the patients are aware of their lapses, and almost all of them experience some form of memory disturbance, even if nothing more than a vague inability to grasp things with sufficient precision.

Other rare presentations include anorexia nervosa (Signer and Benson 1990), multiple personality (Schenk and Bear 1981) or compulsive water drinking (Remillard, et al. 1981). Spitting and embarrassment have been described as the aura of a complex partial seizure (Devinsky and colleagues 1982; Hecker and colleagues 1972).

Finally, the clinician should inquire as to a family history of migraine, since migraine is overrepresented in families with TLE and can mimic the majority of TLE symptoms.

Tactful inquiry may result in anecdotal reports of sexual disturbances in some patients with TLE. Most common is a global hyposexuality affecting both libidinal and genital arousal. In individual instances, such patients may be mistakenly diagnosed as exhibiting hypoactive sexual desire disorder. These two can be distinguished by eliciting on history other hallmarks of temporal lobe epilepsy. Although rarer than hyposexuality, a great variety of other sexual disorders may be encountered in TLE. These include fetishism, transvestic fetishism, sadomasochism, pedophilia, frotteurism and voyeurism. During the seizures, the patients may also experience genital sensations, even feelings of sexual excitement evoked by the epileptic discharges.

Episodes of frank psychosis can be the initial presentation of TLE: a 38-year-old businessman with a history of childhood staring spells and petit mal epilepsy confirmed by EEG, came under considerable work pressure and began an around-the-clock work marathon lasting two days. Suddenly, early in the morning of his second "all-nighter," he experienced a "realization" that his difficulties must be conveyed to the president of the United States. After his arrest at the White House and transfer to the psychiatric ward of a local teaching hospital, an EEG was ordered for him because clinicians uncovered a history of childhood epilepsy. It showed bursts of generalized spike/wave discharges consistent with a generalized seizure disorder.

A magnetic resonance imaging (MRI) scan showed "two foci of increased signal noted in the medial aspect of the right temporal horn associated with a widening of the right temporal horn." A neurology consult suggested the use of anticonvulsants only if the patient became "symptomatic." The psychiatrist concluded: "Since the patient had no observable physical symptoms of seizures, we would withhold the anticonvulsant at this time." On low doses of antipsychotics the patient improved, another tip-off that his psychosis was atypical in etiology. (Clinicians should note, however, that antipsychotics tend to lower the seizure threshold, thereby increasing patient's incidence of seizure.)

At discharge he left the hospital uncertain and worried about what had happened to him. When I saw him in initial consultation several months later, the psychosis had cleared and he was no longer on any medications. Nonetheless, he continued to experience difficulties with memory and "getting my thinking exactly right." Neuropsychological testing showed "deficits consistent with the local or remote effects of a right temporal lesion." When placed on carbamazepine, the patient reported improvement in thinking and memory.

In retrospect, the most likely explanation was this: The lack of sleep and food coupled with stress led to the onset of a schizophreniform episode in a person with a latent seizure disorder. Because the episode was not associated with any signs of generalized epilepsy, it was not recognized as an example of complex partial seizure disorder by either psychiatrist or neurologist . After the psychosis had cleared, the patient was still left with a problem involving memory and focus originating from the hippocampus, a secondary effect of the abnormality in the right temporal region. The memory disturbance improved with anticonvulsant administration.

TLE also may be responsible for chronic rather than just acute psychoses. While any of the symptoms of schizophrenia may be encountered, paranoid traits are the most common. TLE patients can be distinguished from schizophrenic patients by the maintenance, when not acutely ill, of warm affect and good rapport. In addition to the history, the diagnosis of complex partial seizure disorder can be aided by EEG. However, since such diagnosis remains a clinical one, it should be noted that several negative EEGs do not rule out the diagnosis of TLE in a given patient. Other diagnostic aids include MRI, single photon emission computed tomography (SPECT) and positron emission tomography (PET). Interictal SPECT of cerebral blood flow is not nearly as helpful as ictal SPECT. Even more sensitive, although not generally available, is PET imaging of interictal cerebral metabolism. PET permits greater spatial resolution and versatility. Only MRI can image the structural changes associated with the underlying epileptic process. Quantitative evidence of hippocampal volume loss is correlated with seizure onset in medial temporal structures.

The treatment of TLE is complicated by the fact that many times improved seizure control via anticonvulsants leads to deterioration of the neuropsychiatric status. Schizophrenia-like epileptic psychoses often emerge when anticonvulsants are normalizing or improving the seizure activity. If this antithesis isn't recognized, the psychosis will soon become more of a problem than the seizures. One expert, Dietrich Blumer, M.D., has gone so far as to claim: "It is probable that the modern schizophrenia-like epileptic psychoses are largely iatrogenic in nature, caused by modern ability to control seizures."

TLE management presents a conundrum. While the illness is an epileptic one and treated by neurologists, many neurologists remain unfamiliar with and even uninterested in its neuropsychiatric components. But by ignoring the experiential symptoms, the neurologist deprives the patient of the opportunity to coherently integrate all aspects of the epilepsy. It may also cement the patient's misconception that in addition to the epilepsy, he or she suffers from a "mental illness."

Total management of TLE by a psychiatrist is also not without problems. Although temporal lobe epileptic patients are particularly intriguing to psychiatrists because of the nature of the symptoms, these "psychic" seizures can generalize at any time into psychomotor status or grand mal attacks. What's more, neither the timing nor the seriousness of grand mal episodes can be predicted; the initial generalized seizure sometimes occurs many years after the first manifestations of the illness and may culminate in status epilepticus and death.

For these reasons, a physician should undertake the treatment of TLE patients only if he or she has sufficient training and experience in the overall management of epilepsy. When this isn't the case, close collaboration between psychiatrist and neurologist offers the best venue for successful management of this fascinating "bridge" between neurology and psychiatry.

References 1. Bancaud J, Brunet-Bourgin F, Chauvel P, Halgren E. Anatomical origin of déjà vu and vivid "memories" in human temporal lobe epilepsy. Brain. 1994(Feb);117(Pt 1):71-90. 2. Devinsky O, Hafler DA, Victor J. Embarrassment as the aura of a complex partial seizure. Neurology. 1982; 32(11):1284-1285. 3. Ellison JM. Alterations of sexual behavior in temporal lobe epilepsy. Psychosomatics. 1982;23(5):499-500, 505-509. 4. Hecker A, Andermann F, Rodin EA. Spitting automatism in temporal lobe seizures. Epilepsia. 1972;13(6):767-772. 5. Jackson JH. Brain. 1898;21:580-590. 6. Kotagal P, Luders HO, Williams G, et al. Psychomotor seizures of temporal lobe onset: analysis of symptom clusters and sequences. Epilepsy Res. 1995;20(1):49-67. 7. Markand ON, Wheeler GL, Pollack SL. Complex partial status epilepticus (psychomotor status). Neurology. 1978;28(2):189-196. 8. Meiners LC, van Gils A, Jansen GH, et al. Temporal lobe epilepsy: the various MR appearances of histologically proven mesial temporal sclerosis. Am J Neuroradiol. 1994(Sept);15(8):1547-1555. 9. Mullan S, Penfield W. Illusions of comparative interpretation and emotion. Archives of Neurology and Psychiatry. 1959;80:269-284. 10. Pritchard PB III, Lombroso CT, McIntyre M. Psychological complications of temporal lobe epilepsy. Neurology 1980;30(3):227-232. 11. Remillard GM, Andermann F, Gloor P, et al. Water-drinking as ictal behavior in complex partial seizures. Neurology. 1981;31(2):117-124. 12. Remillard GM, Andermann F, Testa GF, et al. Sexual ictal manifestations predominate in women with temporal lobe epilepsy: a finding suggesting sexual dimorphism in the human brain. Neurology. 1983;33(3):323-30. 13. Schenk L, Bear D. Multiple personality and related dissociative phenomena in patients with temporal lobe epilepsy. Am J Psychiatry. 1981;138(10):1311-1316. 14. Signer SF, Benson DF. Three cases of anorexia nervosa associated with temporal lobe epilepsy. Am J Psychiatry. 1990;147(2):235-238. 15. Slater E, Beard AW, Glithero E. The schizophrenia-like psychoses of epilepsy. Br J Psychiatry. 1963;109:95-150. 16. Spencer SS. The relative contributions of MRI, SPECT and PET imaging in epilepsy. Epilepsia. 1994;35 (Suppl 6):S72-89. 17. Spencer SS, Spencer DD, Williamson PD, Mattson RH. Sexual automatisms in complex partial seizures. Neurology. 1983;33(5):527-33. 18. Stoudemire A, Nelson A, Houpt JL. Interictal schizophrenia-like psychoses in temporal lobe epilepsy. Psychosomatics. 1983;24(4):331-333;337-339. 19. Umbricht D, Degreef G, Barr WB, et al. Postictal and chronic psychoses in patients with temporal lobe epilepsy. Am J Psychiatry. 1995;152(2):224-231.

Health effect of β-Carotene

Amelia Fossetta Manatar - Samratulangi University

Plasma β-Carotene is commony used as a reliable measure of dietary fruit and vegetable intake, giving support to possibility that β-Carotene might be causative in the decrease in CVD, cancer incidence, and all causes of age-related deaths, which has been reported for diets high in fruits and vegetables (Doll and Peto, 1981). Epidemiologic studies consistently support this association. For example, a recent study on prostate cancer in younger man found that diets rich in β-Carotene were associated with a lower incidence of prostate cancer (Wu et al., 2004). The potential prevention of chronic diseases associated with oxidative damage by β-Carotene was supported by the finding that plasma β-Carotene is decreased in smokers (Fukao et al., 1996). Because of this multiple conjugated double bonds, β-Carotene is an excellent antioxidant. Unfortunately, after many studies of the antioxidant effects of β-Carotene in cell culture and in animals and human following dietary exposure, the findings are inconclusive. Furthermore, of three key clinical trials evaluating a role for supplemental β-Carotene (20 or 30 mg/day or 50 mg on alternate days) in prevention of lung cancer, two were halted early because smokers in the β-Carotene arm had a greater incidence of cancer. The third (Physicians Health Study) showed no effect of β-Carotene on lung cancer (Clarke and Armitage, 2002). The positive side to these findings has been the insight that a reevaluation is needed in the way in which epidemiologic relationships between whole-food diets and disease incidence are translated into dietary intervention studies (Greenwald, 2002). Although trials of β-Carotene supplementation and CVD have proved negative (Clarke and Armitage, 2002), there is a strong relationship between plasma carotenoids, when used as a marker for fruit and vegetable intake, and reduced risk for CVD (Hak et al., 2004). It remains to be determined if this is due to the presence of carotenoids, or wheter carotenoids, as a marker for a plant-based diet, also act as a marker for exposure to fiber or other plant-based bioactive components.

AHA/ASA Guidelines for the Management of Spontaneous Intracerebral Hemorrhage-2010 Update

Despite the absence of a specific treatment for spontaneous, nontraumatic, intracerebral hemorrhage, early, aggressive medical management can improve patient outcome. To update physicians on state of the art care, the American Heart Association (AHA) and American Stroke Association (ASA) have published a new evidence-based guideline that covers diagnosis, hemostasis, blood pressure management, inpatient and nursing management, prevention of medical comorbidities, surgical treatment, prognosis, rehabilitation, prevention of recurrence, and other considerations (Morgenstern et al 2010). All of the recommendations summarized below are Class I ("conditions for which there is evidence for and/or general agreement that the procedure or treatment is useful and effective").

Early diagnosis is important, as >20% of patients deteriorate from the time they are seen by prehospital emergency medical services and arrival in the emergency room. While many hospitals have critical pathways for acute ischemic stroke, they may be less adept at managing the less frequently seen acute hemorrhagic stroke. Emergency management includes neurosurgery and/or neurology consultation, and may include blood pressure management, intubation, hematoma evacuation, external ventricular drainage, intracranial pressure monitoring, and reversal of coagulopathy.

Initial assessment includes prompt neuroimaging with CT or MRI. CT angiography and contrast-enhanced CT may be considered to help identify patients at risk of hematoma expansion. In addition, MRI/angiogram/venogram and CT angiogram/venogram are reasonable to identify underlying causes of hemorrhage such as arteriovenousmalformations, cerebral vein thrombosis, moyamoya disease or tumors.

Patients with underlying coagulopathies need to be identified, whether due to oral anticoagulants, acquired or congenital factor deficiencies, or quantitative or qualitative platelet deficiencies. Coagulopathies should be corrected if possible. Patients should be initially managed in the intensive care unit, and should receive elastic stockings and intermittent pneumatic compression to prevent venous thromboembolism. Normoglycemia should be maintained, blood pressure controlled, and clinical epileptic seizures treated with antiepileptic medications. Electrographic seizures should also be treated in patients with altered mental status. Prompt surgical removal of the hemorrhage is indicated for patients with cerebellar hemorrhage with neurological deterioration, brainstem compression, and/or hydrocephalus from ventricular obstruction.

Conclusions

Predicting prognosis for patients with intracerebral hemorrhage remains imperfect, justifying aggressive initial treatment. These updated recommendations provide a solid framework for a comprehensive clinical approach. Information regarding ongoing clinical trials for intracerebral hemorrhage.

References

Morgenstern LB et al. Guidelines for the management of spontaneous intracerebral hemorrhage. A guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke 2010;41:00-00. DOI:10.1161/STR.ObO13e3181ec611b

Case 28-2011 — A 74-Year-Old Man with Pemphigus Vulgaris and Lung Nodules

CASE RECORDS OF THE MASSACHUSETTS GENERAL HOSPITAL

Richard C. Cabot, Founder, Nancy Lee Harris, M.D., Editor, Jo-Anne O. Shepard, M.D., Associate Editor, Eric S. Rosenberg, M.D., Associate Editor, Alice M. Cort, M.D., Associate Editor, Sally H. Ebeling, Assistant Editor, Christine C. Peters, Assistant Editor

PRESENTATION OF CASE

Dr. Omobolaji T. Campbell (Infectious Diseases): A 74-year-old man was seen in the surgery clinic at this hospital for evaluation of lung nodules.

A diagnosis of pemphigus vulgaris had been made 5 months earlier after a 2-month history of skin lesions that had progressed from erosions on the patient's lips to multiple bullae and crusted erosions on his face, trunk, scalp, lips, and oral mucosa. The erosions and bullae were itchy and painful and associated with difficulty swallowing, anorexia, and weight loss of 1.4 kg. The patient had a history of diabetes mellitus, hypertension, and hyperlipidemia. Medications included enalapril, nifedipine, loratadine, simvastatin, hydrochlorothiazide, metformin, and acetylsalicylic acid; all these except metformin had been discontinued by the patient's primary care physician when the rash first developed, before evaluation by the dermatologist. On examination by the dermatologist at that time, there were multiple pink crusted plaques and flaccid blisters on the trunk, face, and scalp and erosions on the oral mucosa. Nikolsky's sign (blistering of the skin and mucosa as a response to slight rubbing or pressure) was positive on the left upper back. The serum level of IgG antibodies to desmoglein 1 was 180.7 U (reference range, <14.0), and to desmoglein 3 was 154.6 U (reference range, <9.0). Pathological examination of a skin-biopsy specimen of the lesions on the lower back showed intraepidermal blistering that was consistent with pemphigus vulgaris.

The administration of prednisone (60 mg daily) was begun; the lesions rapidly improved, and the dose was reduced to 40 mg daily. Hyperglycemia developed, and treatment with glipizide was begun. Two weeks later, the serum sodium level was 131 mmol per liter (reference range, 135 to 145), the chloride level 97 mmol per liter (reference range, 100 to 108), and the glucose level 213 mg per deciliter (11.8 mmol per liter) (reference range, 70 to 110 mg per deciliter [3.9 to 6.1 mmol per liter]); serum levels of potassium, carbon dioxide, calcium, total protein, albumin, and globulin were normal, as were the results of tests for liver and renal function and thiopurine methyltransferase activity. Urinalysis showed glucose 2+ (300 to 500 mg per deciliter [16.7 to 27.8 mmol per liter]); results were otherwise normal. Three weeks later, a skin test for tuberculosis was negative. The dose of prednisone was gradually increased to 60 mg daily, and treatment with azathioprine (initial dose, 50 mg daily) was begun as a glucocorticoid-sparing agent.

At follow-up in the dermatology clinic 1 month later, the patient reported hoarseness and a cough with yellow sputum of 1 week's duration. Despite resolution of many of the skin lesions, the oral lesions persisted. He reported eating and drinking well; however, his weight had decreased from 60.0 kg to 56.4 kg. The administration of trimethoprim–sulfamethoxazole for prophylaxis against infection was begun, and the patient was referred to an otolaryngologist; fiberoptic laryngoscopic examination reportedly revealed involvement of the vocal cords with pemphigus. Testing for paraneoplastic pemphigus IgG autoantibodies was negative. The dose of azathioprine was increased to 100 mg daily.

At follow-up 5 months after the onset of symptoms, the patient reported less discomfort of his oral mucosa and resolution of his hoarseness, but he reported new painful lesions on his back and burning with urination, without urinary frequency, hematuria, or blood in his stool. Physical examination revealed new crusted erosions on his back. The white-cell count was 11,900 per cubic millimeter (88% neutrophils). Urinalysis revealed a glucose level greater than 1000 mg per deciliter (55.5 mmol per liter), trace amounts of blood, and 3 to 4 red cells, 10 to 15 bacteria, and 3 to 10 epithelial cells per high-power field. A urine culture was sterile. Ultrasonography of the kidneys performed at another hospital reportedly revealed bilateral, nonobstructing renal calculi. A computed tomographic (CT) scan of the abdomen and pelvis reportedly showed nephrolithiasis and prostatic hypertrophy, without obstruction, with incidental findings of a pulmonary consolidation and lung nodules in the right lower lobe. Treatment with insulin was begun for persistent hyperglycemia.

Five weeks before this evaluation, the dose of azathioprine was increased to the goal dose of 150 mg daily. The patient reported soreness of his tongue; examination revealed white plaques on the lateral edges of the tongue. Fluconazole (100 mg orally, once weekly) was initiated for presumed thrush. CT of the chest reportedly revealed enlargement of parenchymal and pleural densities (predominantly in the right lower lung) and nodules in both lower lobes, which had not been present 4 years earlier. ¹⁸F-fluorodeoxyglucose–positron-emission tomography (FDG-PET) reportedly revealed FDG avidity of the nodules in both lower lobes and of the pleura of the right lung. Seventeen days before this evaluation, the white-cell count was 5200 per cubic millimeter (90% neutrophils); the dose of azathioprine was decreased to 50 mg daily. The patient was referred to the surgery clinic at this hospital for evaluation of the lung nodules.

The patient reported a weight loss of 11.3 kg associated with his illness; he had regained 4.5 kg after the symptoms had improved. He had persistent fatigue and dysuria; he did not have back pain, difficulty swallowing, cough, sore throat, hoarseness, blood in the urine or stool, headache, fever, chills, or sweats. He had had an episode of chest pain with normal cardiac perfusion 6 months earlier, mild-to-moderate left ventricular hypertrophy according to echocardiography, and prostatic hypertrophy; he had also undergone an appendectomy. Oral medications included prednisone, azathioprine, metformin, glipizide, acetylsalicylic acid, acetaminophen, ergocalciferol, calcium carbonate, and trimethoprim–sulfamethoxazole, with tramadol as needed for pain. Topical medications included mupirocin ointment, gentian violet, and halobetasol propionate ointment for active skin lesions, and viscous lidocaine as needed for painful mouth sores. The patient was allergic to rosiglitazone maleate, which had caused a rash. He was from Brazil and had immigrated to the United States more than 10 years earlier. He lived in New England with his wife and was retired. He had stopped smoking 40 years earlier, after 5 pack-years, and did not drink alcohol or use illicit drugs. He was monogamous with his wife and had traveled only to Florida; he had not returned to South America. There was no history of exposure to blood products or animals. His father had died of stomach cancer at 69 years of age, and his mother had died during childbirth. One of his eight children had died of an “immune problem”; there was no family history of skin diseases.

On examination, the height was 157.5 cm and the weight 54.9 kg. The vital signs were normal. There were multiple healed pink scars on the chest, abdomen, inguinal folds, and scalp; several active ulcers on the back in various stages of healing; and several broad erosions on the hard and soft palate. The remainder of the examination was normal. Pulmonary-function tests revealed a forced vital capacity of 2.42 liters (76% of the predicted value) and a forced expiratory volume in 1 second of 1.83 liters (73% of the predicted value), and the carbon monoxide diffusing capacity was 80% of the predicted value. The white-cell count was 9800 per cubic millimeter, and the serum electrolyte levels were 134 mmol of sodium per liter, 5.4 mmol of potassium per liter (reference range, 3.4 to 4.8), and 95 mmol of chloride per liter. The remainder of the complete blood count was normal, as were tests of coagulation and liver and kidney function.

A diagnostic procedure was performed.

DIFFERENTIAL DIAGNOSIS

Dr. Eleftherios Mylonakis: This 74-year-old man with pemphigus vulgaris who was taking glucocorticoids and azathioprine presented with multiple lung nodules. In order to focus our differential diagnosis, I would like to ask Dr. Muse to review the radiographic findings and to specifically describe the location, size, and number of lesions and the presence or absence of calcification or cavitation. It would also be informative to know whether the margins of the lesions are well demarcated or poorly defined (“fuzzy”) and whether the lesions have a peribronchovascular distribution.

Dr. Victorine V. Muse: Limited images from a whole-body combination FDG-PET–CT scan obtained 3 weeks before admission were provided for review. CT images of the chest without the administration of contrast material reveal multiple ill-defined coalescing pulmonary nodules, bilaterally but primarily in the peripheral right lower lobe. The lesions are of varying sizes, up to approximately 3 cm in diameter, with air bronchograms and no evidence of calcification or cavitation (Figure 1A and 1BFIGURE 1

Imaging Studies.

). The larger nodules show marked FDG avidity, which is a marker of glucose metabolism that exceeds the expected baseline. In addition, there is diffuse increased metabolic activity of the right posterior pleura as evidenced by marked FDG uptake without notable correlation on the CT images (Figure 1C). There is no evidence of mediastinal or hilar lymphadenopathy, and there are no important findings in the abdomen or pelvis.Dr. Mylonakis: The differential diagnosis is broad and includes malignant conditions, inflammatory diseases, and infection (Table 1TABLE 1

Select Differential Diagnosis of Lung Nodules in This Patient.

). However, the clinical and imaging characteristics of the lung nodules help us rule out some diagnoses and focus on others.

Neoplasms of the Lung

Metastatic disease can present as pulmonary nodules that are variable in size and location and are often round with sharply demarcated borders. In this patient, the nodules have poorly defined borders; malignant nodules of this type have a tendency to hemorrhage. Lymphoma should also be considered; however, this patient has no enlarged or metabolically active lymph nodes and the other organs are normal. The patient is negative for human immunodeficiency virus (HIV), making HIV-associated malignant conditions such as Kaposi's sarcoma unlikely.

Inflammatory Diseases of the Lung

This patient is taking glucocorticoids and azathioprine. A lack of clinical response while the patient is receiving these immunosuppressive medications argues against inflammatory diseases. Many of the characteristic findings of pulmonary granulomatous polyangiitis (formerly known as Wegener's granulomatosis) are absent in this patient, including upper-airway involvement, hilar lymphadenopathy, and cavitation. He has no signs of rheumatoid arthritis and no findings suggestive of amyloidosis. Finally, sarcoidosis is not likely, because of the radiographic findings and the lack of lesions of the skin, joints, and eyes. We should also consider lymphomatoid granulomatosis, although fever, neurologic abnormalities, chest pain, subcutaneous nodules, and cavitation are absent, and there is no peribronchovascular distribution. There is also no history of exposure that would suggest pneumoconiosis.

Infection

Evaluation of pulmonary nodules should include consideration of septic emboli from an endovascular source. This patient had negative blood cultures, and the echocardiogram did not show vegetations. Moreover, pulmonary lesions from septic emboli often cavitate, and we do not have evidence of cavitation in this case. Some parasites such as Paragonimus westermani typically migrate to or through the lung, but this patient has no history of travel to Asia to suggest this type of infection.

The clinical and radiographic findings are consistent with a fungal infection of the lung. Histoplasmosis is endemic is some areas of Brazil, where this patient had spent most of his life. However, features that might be suggestive of histoplasmosis are absent, such as lymphadenopathy, cavitary pneumonia, pericarditis, esophageal narrowing, erythema nodosum, or a history of treated histoplasmosis that could have been reactivated. Coccidioidomycosis should also be considered, and it does occur in Brazil, but this diagnosis cannot be supported, because the patient does not have a previous history of infection or any recent travel to an area where coccidioidomycosis is endemic. Infection with a filamentous fungus cannot be ruled out, but pulmonary lesions associated with invasive aspergillosis and other invasive molds commonly have a surrounding halo of ground-glass attenuation due to local hemorrhage and cavitation, which is absent in this case.

Cryptococcosis

Cryptococcosis is most often seen in the United States in patients with advanced HIV infection, but a number of other conditions can confer a predisposition to cryptococcal infection, including idiopathic CD4 lymphocytopenia,1 transplantation, connective-tissue disease, malignant tumors, glucocorticoid therapy, chronic obstructive pulmonary disease, cirrhosis, and sarcoidosis. This infection can also develop without a clear predisposing factor. Glucocorticoid therapy is associated with up to 30% of cases of cryptococcal infection in persons who are HIV-negative. 2,3 Cryptococcal infection has also been reported in patients who take azathioprine.4 In addition, the infection has been reported in persons with pemphigus vulgaris,5 but there may be no association with underlying disease.

Cryptococcal infection begins with inhalation of the basidiospores (or cells) of the fungus. Cryptococcal cells are yeasts with a small capsule. The survival of cryptococcal cells in the environment and inside mammalian and nonmammalian hosts depends on the presence of the capsule and the production of the antioxidant melanin.6-8 As is the case with most patients who have clinically symptomatic cryptococcal infection, this patient was most likely exposed to cryptococcus sometime in the past and developed a primary infection that was asymptomatic. The infection probably persisted in a latent state until the patient's immune system was suppressed, at which point active infection ensued.9 However, primary infection is also possible.

If this patient has cryptococcosis, he is most likely infected with Cryptococcus neoformans var.grubii, which has worldwide distribution and is more common in immunosuppressed patients than in those who are not immunosuppressed. C. neoformans var. neoformans also causes disease in immunocompromised hosts but is more common in parts of Europe. C. gattii may cause disease in persons without well-defined predisposing factors; although it is typically endemic in areas with tropical climates, such as Australia and Papua New Guinea, it has spread to other areas, including Vancouver Island, the U.S. Pacific Northwest, and Brazil.

In order to establish the diagnosis of cryptococcosis, a lung biopsy should be performed. A serum cryptococcal-antigen assay is not adequate to make the diagnosis in this case, since among HIV-negative patients, it is positive in only 56% of patients with pulmonary disease.3 Also, a lumbar puncture, with evaluation of the cerebrospinal fluid (CSF) for dissemination of cryptococcal infection, should be performed in this immunocompromised patient.9,10

Dr. Eric S. Rosenberg (Pathology): Dr. Gandhi, would you tell us how this patient was evaluated?

Dr. Rajesh T. Gandhi (Infectious Diseases): This patient was evaluated in the surgical clinic, where it was decided to proceed with a lung biopsy. The patient underwent examination with a flexible bronchoscope, bronchoalveolar lavage, and right-sided video-assisted thoracoscopy with pleural and diaphragmatic biopsies and wedge resection of the mass in the right lower lobe. He was referred to me after a diagnostic result was received.

CLINICAL DIAGNOSIS

Pulmonary cryptococcosis.

DR. ELEFTHERIOS MYLONAKIS'S DIAGNOSIS

Pulmonary cryptococcosis, with possible central nervous system involvement.

PATHOLOGICAL DISCUSSION

Dr. Mari Mino-Kenudson: A punch-biopsy specimen from a blister that had developed on the patient's back 5 months before the current evaluation showed acantholysis, leading to detachment of the epidermis from the dermis (Figure 2AFIGURE 2

Skin-Biopsy Specimen.

). The epidermal separation took place suprabasally. The basal keratinocytes, although separated from one another, remained firmly anchored to the dermis like a row of tombstones (Figure 2B). An immunofluorescence study highlighted intercellular, basilar deposition of IgG and C3 but no deposition of IgA, IgM, or fibrin. The findings are consistent with pemphigus vulgaris.The diagnostic procedure consisted of a thoracoscopic wedge biopsy of one of the lung nodules, as well as a pleural biopsy. The wedge-biopsy specimen of the lung showed that a broad area of lung parenchyma had been replaced by a process of alveolar filling that consisted of granulomatous inflammation that contained a large number of multinucleated giant cells and fibrosis (Figure 3AFIGURE 3

Lung-Biopsy Specimen.

). Numerous yeast-form microorganisms, ranging in size from 2 μm to 15 μm, were seen within the giant cells or in the stroma (Figure 3B). Occasional narrow-based budding forms were also present (Figure 3C). Mucicarmine stain defines the capsule of the yeast forms (Figure 3D), and the Fontana–Masson stain highlights melanin precursors within cell walls (Figure 3E). These features are characteristic of cryptococcus and are associated with its virulence.11

The pleural lesions had essentially the same features as the lung nodules. Thus, the final diagnosis of the lung and pleural lesions was granulomatous pneumonitis and pleuritis due to cryptococcus infection. The culture of lung tissue grew C. neoformans, confirming the diagnosis.

Cryptococcus can produce several forms of disease in the lung, depending in part on the degree of the patient's immunosuppression.12 In patients with a low degree of immunosuppression, the infection results in fibrocaseous cryptococcoma with central necrosis surrounded by a fibrous capsule, or as in this patient, it results in less well circumscribed, granulomatous pneumonia. Histiocytic pneumonia or mucoid pneumonia with few inflammatory cells may be seen in patients who have a high degree of immunosuppression. In a profoundly immunosuppressed host, yeast forms may plug the capillaries as a manifestation of cryptococcal sepsis, with essentially no cellular reaction in alveoli. 12,13

Dr. Rosenberg: Dr. Gandhi, would you tell us how you treated this patient and what happened to him?

Dr. Gandhi: This patient had no symptoms or findings on physical examination to suggest involvement of the skin, bones, or prostate gland. He did report an intermittent headache that he associated with drinking tea; because of the possibility of cryptococcal meningitis, he underwent a lumbar puncture. The opening pressure was not elevated; examination of the CSF revealed a total protein level of 56 mg per deciliter and a glucose level of 95 mg per deciliter (5.3 mmol per liter); there were 13 white cells, predominantly lymphocytes. The result of a CSF cryptococcal-antigen test was 1:1024, and India ink staining of the CSF showed yeast forms with a capsule. The result of a serum cryptococcal-antigen test was 1:4096. The HIV-antibody test was negative, and the CD4 count was 500 cells per cubic millimeter. A culture of the CSF ultimately grew cryptococcus; blood cultures remained sterile.

The patient was treated with intravenous liposomal amphotericin and oral flucytosine for approximately 3 weeks, with clinical improvement. He also underwent a prednisone taper and was started on intravenous immune globulin for management of his skin disease. A repeat CSF examination at the end of his intravenous therapy showed a decreased white-cell count, and the culture was sterile. The antifungal therapy was changed to oral fluconazole.

Dr. Stephen B. Calderwood (Infectious Diseases): This patient had a long history of dysuria, and we know that cryptococcus lives in the prostate, particularly in persons who are infected with HIV. Is it possible that this HIV-negative patient had involvement of his prostate?

Dr. Gandhi: I did have some concerns about cryptococcal involvement of the prostate because of the patient's dysuria, but his prostate examination did not show any bogginess or tenderness and his urine cultures were negative for cryptococcus. Dr. Mylonakis, do you have any comments on the prostate as a sanctuary for cryptococcus?

Dr. Mylonakis: The prostate may be involved with cryptococcal infection. Common organs of involvement with cryptococcus in HIV-positive and HIV-negative patients are the lungs, the central nervous system, and the skin. Involvement of the prostate with cryptococcal infection is often overlooked but should always be considered.

Dr. Gandhi: Dr. Muse, this patient had a follow-up chest CT 2 months after initiation of fluconazole therapy, and the nodular lesions had not resolved. Could the PET scan be used to assess the effectiveness of therapy?

Dr. Muse: Currently, PET-CT imaging is not used to monitor the therapeutic response in cases of infection.14,15

A Physician: How is the patient's pemphigus now being managed?

Dr. Alison R. Avram (Dermatology): The pemphigus took a while to respond to the prednisone and azathioprine. After about 4 months of treatment, which is when the patient was admitted for cryptococcosis, his mucous membranes and his skin had completely cleared. Unfortunately, when this infection developed, we had to taper the dose of glucocorticoid to physiologic levels and start intravenous immune globulin. Although he had an initial response to intravenous immune globulin, he had a recurrence of skin lesions and we had to increase the prednisone dose. Ultimately, we were unable to wean the patient from prednisone, and subsequently he underwent plasmapheresis to decrease antibody titers in preparation for rituximab. The patient has completed four weekly infusions of rituximab, and we have subsequently been able to taper the dose of prednisone to 15 mg every other day, alternating with 10 mg every other day. He is clinically clear of mucosal and skin disease.

A Physician: Do you know how long you will treat the patient for the cryptococcal infection? What will indicate that he has improved?

Dr. Gandhi: Our original plan was to complete a 12-month course of antifungal therapy, using the patient's overall clinical status and radiologic response to determine the exact duration of treatment. One year after the diagnosis, he appeared clinically well and chest imaging revealed improvement in the nodular pulmonary lesions. However, we have decided to continue fluconazole while he receives additional immunosuppressive therapy, including rituximab, for his skin disease.

ANATOMICAL DIAGNOSIS

Cryptococcal infection of the lung and central nervous system.

This case was discussed at the postgraduate course Infectious Diseases of Adults, sponsored by the Harvard Medical School Department of Continuing Education, Course Directors Stephen B. Calderwood, Nesli Basgoz, and Jay A. Fishman.

Dr. Mylonakis reports receiving grant support from Astellas and T2 Biosystems, fees for board membership from Astellas, and lecture fees from Pfizer. Dr. Muse reports receiving consulting fees and lecture fees from Biomedical Systems.

Disclosure forms provided by the authors are available with the full text of this article at NEJM.org.

No other potential conflict of interest relevant to this article was reported.

We thank Drs. Alison Avram, Rajesh Gandhi, and Mayra Lorenzo for their assistance with the preparation of the case history.

SOURCE INFORMATION

From the Division of Infectious Diseases (E.M.) and the Departments of Radiology (V.V.M.) and Pathology (M.M.-K.), Massachusetts General Hospital; and the Departments of Medicine (E.M.), Radiology (V.V.M.), and Pathology (M.M.-K.), Harvard Medical School — both in Boston.