Treating the patient with human immunodeficiency virus, type 1 (HIV-1) infection presents challenges across all aspects of the biopsychosocial model. Herein, we present a case illustrating that when the central nervous system (CNS) has been infected by HIV-1, neuropsychiatric syndromes may deviate from their typical presentations. Multiple syndromes more frequently occur concomitantly, requiring the clinician to be exact in conceptualization of the related, unfolding pathologies. This is necessary to differentiate these conditions from one another as well as to discern their respective responses to treatment.
Diffuse HIV-1—induced CNS dysfunction may obscure EEG findings that would normally be present. The invasion of the CNS by HIV-1 occurs early in the infection process.1 Approximately two-thirds of adults infected with HIV-1 eventually develop neuropsychiatric diseases that are primary complications of HIV-1 infection itself.2,3 As many as 20% of otherwise asymptomatic patients can be shown both histopathologically4,5 and physiologically6,7 to have significant central and peripheral nervous system involvement by measures of neurological activity such as EEG8,9 (especially when used with power spectral analysis10,11); visual, brainstem, and somatosensory evoked potentials;6,8 and neuropsychological testing.7,10,11
HIV-1 can affect the CNS in several different ways. It can affect the CNS directly, through macrophage infection and activation5 and increased viral load.12,13 It can also affect the CNS indirectly—through apoptosis of neurons;14 through destructive effects of cytokines such as tumor necrosis factor-alpha, interleukin-1, interleukin-6, and interferon-alpha and -gamma;15 through neurotransmitter changes such as increments in glutamate levels and decrements in dopamine,16 serotonin,17 and (less well established) acetylcholine levels;18 and through production of reactive oxidative intermediates associated with increases in intraneuronal free calcium levels and nitric oxide production.19 Still other ways that HIV-1 can affect the CNS are through vasculitis and, sometimes, cerebrovascular accident;20 through secondary CNS opportunistic infections such as toxoplasmosis or cryptococcal and tuberculous meningitides; and through CNS tumors (e.g., lymphoma). Neuropsychiatric presentations, therefore, may reflect cortical damage that is localized or diffuse as well as the well-known subcortical damage associated with infection of the brain by HIV-1. These variegated effects may cause DSM-IV Axis I disorders such as HIV-1—associated dementia21 and minor cognitive motor disorder,22 delirium,23 and mood disorder (mania24 as well as depression25,26). In addition, other neuropsychiatric complications may occur, such as new-onset generalized or partial seizures27 with postictal manifestations. Furthermore, metabolic encephalopathies and psychoneurotoxicities due to prescribed medications (e.g., zidovudine)28 and substances of abuse are also observed. In some cases, neuropsychiatric pathological entities may coexist and manifest simultaneously, as in the case we present.
The patient is a 27-year-old homosexual white male who had first presented for psychiatric treatment 2 years earlier following an intentional overdose of prescribed medication. The date of initial presentation was his birthday as well as the 1-month anniversary of his notification of positive HIV-1 serostatus. At the time, he had no opportunistic infections; however, his CD4 cell count was 2 cells/mm3—indicative of long-standing infection with HIV-1 and severe immunosuppression, qualifying this patient for the diagnosis of AIDS.29 While normally a CD4 cell count of less than 50 cells/mm3 has been used to characterize endstage immunological progression associated with a 6-month prognosis, this patient had survived longer than 2 years—before the advent of highly active antiretroviral therapy regimens that employ triple and quadruple antiretroviral drug combinations including a protease inhibitor.
In his initial presentation, the patient was admitted to a local hospital and released after 24 hours. He was nonadherent with attendance at psychiatric follow-up visits. His next psychiatric contact was at our university-affiliated hospital 10 months later, when he was hospitalized for an infected right ankle joint. At that time, he was found to have depressed mood, decreased interest and energy, and frequent crying spells. He was diagnosed as having major depressive disorder without psychotic features. He was started on trazodone 50 mg po qhs, but this was discontinued shortly thereafter because of restlessness. Sertraline was then prescribed but was discontinued 3 months later because of increased scores on liver function tests—alkaline phosphatase, 492 U/L (normal: 25—140 U/L); serum glutamicoxaloacetic transaminase, 157 U/L (normal: 0—40 U/L); lactate dehydrogenase, 399 U/L (normal: 100—225 U/L); and γ-glutamyl transpeptidase, 465 U/L (normal: 0—65 U/L). On discharge, the patient again failed to adhere to his outpatient follow-up visit schedule—purportedly because of his subjective feeling of mood improvement.
Six months later, the patient was admitted with a 3-day history of fever, headache, photophobia, nausea, generalized weakness, and cough. He was determined to have the following diagnoses: bilateral of Mycoplasma pneumoniae pneumonia, Pseudomonas aeruginosa bacteremia, an acute bullous myringitis in his left ear, the normochromic/normocytic anemia of HIV-1 disease, and cytomegalovirus (CMV) retinitis. In addition, he was found to have low levels of follicle-stimulating hormone (<1 MIU/ml; normal mean: 0.9—15), luteinizing hormone (<1 MIU/ml; normal range for males: 2.4—9.4), free testosterone (3.6 pg/ml; normal range for males ages 20—29: 19—41 pg/ml), and total testosterone (13 ng/dl; normal range: 260—1250 ng/dl). Psychiatric evaluation determined him to have adjustment disorder with mixed emotional features. No psychotropic medications were thought to be necessary at that time; he was treated with individual supportive psychotherapy and was discharged shortly thereafter. However, subsequently the patient again became nonadherent to his psychiatric outpatient follow-up visit schedule—due, again, to his subjective feeling of mood improvement.
Six and a half months thereafter, the patient was brought to the hospital by his family for emergency psychiatric evaluation because—according to the attending primary care physician—he had been "confused, disorganized, and disoriented for 3 days." According to his mother, the patient was "awake but acting like he did not know that we were there." She stated that he had not slept for 3 days. At the time, the only medication he was prescribed was Rondec-DM, a pseudoephedrine- and dextromethorphan-containing medication, for sinusitis—an early and frequently missed complication of HIV-1.30
On mental status examination, the patient was fully awake and alert; however, no determination of orientation could be made. The patient responded to all questions with nonspecific responses such as "I don't know" or "I forget." It is of note that the patient previously showed normal speech output and was capable of identifying both the examiner and the hospital by their proper names. However, on this examination, his orientation had deteriorated to the point that automatic answers, such as that for the name of his mother, were not given; yet he responded to his mother in his usual fashion. The reaction of the patient to his deficits was to cry in frustration upon recognizing them.
Memory testing of three objects after 1 minute—"telephone," "Broadway," and "red"—showed confabulation ("ATM"). Three-step commands were completed normally. The patient was able to read a written sentence directing him to close his eyes. Testing of object naming was remarkable for his inability on most objects—"thumb" being called "figure" and "fingernail" being responded to with "I don't know."
The psychiatric impression at the time was that this represented an expressive aphasia (formerly called Broca's) with anomia, implicating frontal lobe dysfunction. A manic episode secondary to pseudoephedrine or dextromethorphan was a diagnostic possibility that still needed to be entertained, since he had not slept for 3 days. Both dextromethorphan and pseudoephedrine have been known to precipitate mania.31 Their administration in combination, therefore, compounded the possibility that this was a manic episode due to psychoneurotoxicity. A manic episode secondary to HIV-1 infection itself was also entertained. However, the patient's mother denied that the patient had demonstrated psychomotor agitation, and there was no objective evidence for rapid or pressured speech. Other differential diagnostic possibilities accounting for mania in HIV-1 infection—neurosyphilis26,32 and meningoencephalitis from other causes (e.g., cryptococcal)33—were also entertained. However, there was neither the history of gradual onset one would expect with neurosyphilis nor the changes in level of mental status one would expect with other meningoencephalitides (or with psychoneurotoxicity).
On further questioning, the patient's mother noted that several months previously he had had an episode of upper-extremity "shaking," which was noted to be tonic-clonic in nature. No treatment had been sought, since the episode passed after a few minutes. This episode was interpreted as an indication that focal CNS damage had likely occurred at some point. Given the rapidity of onset of mental status change at this time, the current inability of the patient to attend to his environment, and his prior presentation with focal neurological deficits, our preliminary diagnostic impression was one of delirium on Axis I and a concomitant expressive aphasia (as previously noted) on Axis III. It was felt that the latter could be secondary to a new CNS event or that it could represent a postictal state of a complex partial seizure disorder, rather than an undifferentiated "postictal confusion." Either possibility would explain the description of "acting like he didn't know the family was there" and "not relating" to them.
Given that delirium represents a psychiatric emergency demanding an immediate medical investigation, it was recommended that the patient be admitted to the internal medicine service for a delirium workup. The blood tests ordered were complete blood count, serum chemistry, thyroid function tests, B12 level, folate level, cryptococcal antigen level, and toxoplasma antibody titers (immunoglobulin G and M [IgG, IgM]). CT of the brain with double-dose, delayed contrast was also ordered. Following these, a lumbar puncture (LP) and a neurology consult were planned.
That night the patient, who had been previously quiet and complacent, became agitated without provocation. He evidenced bizarre behavior and was climbing out of the bed, attempting to wander around the ward. The patient was not responding to limit-setting or to directions from the nursing staff. He was administered 1.0 mg haloperidol orally, and he slept.
On the following day, the patient was found to be in good control, oriented in all three spheres, and fully organized and logical in thought process. His mood was euthymic and stable, and his affect was congruent with his circumstances and normal in range. He showed no evidence of psychosis. In addition, he understood the nature and consequences of his actions and the events occurring at that point in time, and he displayed intact judgment. Moreover, regarding expressive aphasia, he communicated freely, used words properly, and named objects without difficulty. The patient stated that during the prior period, he had fully understood all verbal communications but was frustrated and frightened because he was unable to communicate his thoughts.
CT scan of the head was done and demonstrated no lesions. Mild generalized atrophy consistent with HIV-1—associated CNS involvement was seen, however. Because of his dramatic clinical improvement, the patient refused the LP. The medical workup revealed the patient to have a normochromic/normocytic anemia associated with HIV-1 infection, leukopenia, a low testosterone level, and orthostatic hypotension. In addition to a preexisting cardiomyopathy, he was also found to have HIV-1—associated retinopathy and peripheral neuropathy, as well as a CMV gastroenteritis. It should be noted that although HIV-1—associated retinopathy is often found in HIV-1—infected patients, it is rarely of clinical significance.34 On the second day of this hospitalization, the patient was started on digoxin 0.125 mg/day for cardiomyopathy, albuterol 2 puffs qid for shortness of breath, erythropoietin 10,000 units intravenously three times per week for anemia, nandrolone decanoate 50 mg intramuscularly every 2 weeks and testosterone 200 mg intramuscularly every 2 weeks for decreased lean body weight; G-CSF three times per week intravenously for neutropenia; and oral ganciclovir 500 mg tid for 2 days, then 1,000 mg tid for CMV gastroenteritis.
For the next 3 days, the patient remained symptom-free psychiatrically. During that time, an EEG showed generalized slowing with no focal or other abnormalities. His body temperature had been consistently elevated since admission; after this point, he showed lower daily spikes superimposed on an otherwise normal temperature. At no time during the course of this hospitalization were results of the mental status examination or any elevations in body temperature noted.
On the fifth day of hospitalization at 4:00 a.m., the nursing staff noted a sudden onset of bizarre behavior: rapid, extreme shifts in mood, climbing out of bed, and agitation upon approach, followed by a prolonged period of what was described as "confusion." Although an emergency EEG would have been useful to differentiate postictal seizure as opposed to other CNS dysfunction, it was not available. The next morning, the patient was again reported by the nursing staff to be "confused." On examination at that time, the patient was found to have symptoms consistent with an expressive aphasia, in addition to a labile affect. He was not pressured in speech, did not show psychomotor agitation, nor did he show any other symptoms of mania. On questioning, he admitted to headache. He was able to follow only two-step commands. The patient was presumed to be in a postictal state following a complex partial seizure and was started on carbamazepine.
Because of a persistent cough and a chest X-ray showing a right upper lobe infiltrate, a bronchoscopy was done. A set of three blood cultures was reported as showing no bacterial growth. The patient was never hypoxemic, nor did he ever show signs of respiratory discomfort or distress, rendering pulmonic encephalopathy unlikely. LP was again refused.
As his ability to focus his attention increased, the patient improved in other aspects of his cognitive function as well. In addition, he demonstrated greater stability of mood and a gradual reduction in bizarre behavior. However, he did still exhibit a circumstantial thought process to a minor extent.
Ten days later the patient was again reported by the nursing staff to be "confused." He was found to have a marked change on mental status examination with psychomotor agitation, a labile affect, and a disorganized thought process. He manifested neither aphasia nor anomia. To address these symptoms, his carbamazepine was increased to tolerance (with a blood level of 5.38 μg/ml). Following this dosage increment, the patient's symptoms slowly resolved. Three days later the patient was without any symptoms, other than mild tangentiality of thought process. Carbamazepine levels stabilized at 9.7 μg/dl.
Two days following this, an LP was done. Cerebrospinal fluid cell counts and protein levels were in the normal range. Bacterial, fungal, and viral CSF cultures showed no growth, and an acid-fast bacilli stain was negative. Toxoplasma gondii IgM antibody titers (by enzyme-linked immunoassay) in CSF were negative (0.05, normal:<0.9). Although CMV IgG antibody titer was elevated in the blood (CMV IgG, 3.9; normal: <0.9), the IgM titer was negative—suggesting chronic infection—and the CSF CMV IgG titer was negative as well.
The patient's mental status remained normal, with the exception of a tangential thought process. There were no fluctuations in level of consciousness. His mood was euthymic and affect appropriate. There was no evidence of any other psychiatric symptoms.
Three days later the patient was started on intravenous hydrocortisone 80 mg every 8 hours (a total of three doses) to treat pulmonary inflammation and edema. Three days following this, the patient was again reported to be "confused." On mental status examination, the patient was found to be completely disorganized and to have motoric (i.e., physical) agitation, which extended to psychic (i.e., cognitive/emotive) agitation with minimal external stimulation—as opposed to his more typical state of psychomotor (i.e., physical and cognitive/emotive) retardation. He was also found to have poor judgment. For example, he chose to leave the unit with no footwear for the express purpose of buying candy. He was also found to be speaking rapidly with clang associations, but he demonstrated no evidence of aphasia or anomia. Taken together, this constellation of symptoms was judged to be consistent with a "manic episode."35 The patient was started on 7 mg of haloperidol per day orally and clonazepam 0.5 mg twice a day orally. There was no evidence of either autonomic instability or a movement disorder that could have been indicative of neuroleptic malignant syndrome.
Three days later, his only remaining symptom was a mildly tangential thought process, and he was judged to be hypomanic. Forty-eight hours following, his manic symptoms had further decreased. Because of a concern that discharge was being planned in advance of the resolution of psychiatric symptomatology, haloperidol was increased to 5 mg orally twice daily. The following day, the patient was fully oriented with no evidence of the previously observed tangential thought process. He showed no evidence of anomia, nor any other symptoms of an aphasic syndrome. On short-term memory testing, he was able to recall each of three objects after a delay of 20 minutes. He was completely symptom-free and was discharged.
In summary, this patient had three distinct neuropsychiatric episodes, two of which included symptoms of rapid onset of change in mental status and deficits in executive function, and a third that was diagnosed as a manic episode secondary to corticosteroid therapy.
HIV-1 is known to invade the CNS early in the course of the infection.1,36 Focal damage occurs with CNS opportunistic infections and tumors, as well as with HIV—1 infection alone. As opposed to the immunocompetent individual, the presentation of focal brain damage can be atypical in the HIV-1 infected.
This patient's initial presentation of abrupt onset of nonresponsiveness to the environment was indicative of a delirium—defined by the DSM-IV as a disturbance of consciousness with reduced ability to focus, sustain, or shift attention35—rather than a mood disorder secondary to HIV-1 or HIV-1—associated dementia (although these had been considered in the differential diagnosis). Mood disorder secondary to a general medical condition was ruled out because persistent mood disturbance was not predominant in the clinical picture; mood changes of only relatively brief duration were observed here. The patient's aphasia was judged to be secondary to neuronal postictal exhaustion, relating to the underlying diagnosis of complex partial seizure disorder. This diagnosis was supported by the history given by the patient's mother, which was consistent with prior cortical dysfunction manifesting as a focal motor seizure with a jacksonian march. The diagnosis of a seizure disorder on Axis III, therefore, was supported.
The Axis I diagnosis, with its aforementioned disturbance of consciousness as well as rapid onset and presence of cognitive deficit (followed by complete symptom resolution) was consistent with a diagnosis of delirium by DSM-IV criteria,35 rather than a diagnosis of HIV-1—associated dementia (another diagnosis associated with global cognitive dysfunction). The patient had had interictal periods free of memory deficits and any other signs or symptoms of upper cortical dysfunction, ruling out dementia—which requires consistent neurocognitive impairment for at least 1 month.22 The CT scan did not show a focal lesion, LP yielded no specific diagnosis, and the workup for metabolic (including pulmonic) encephalopathy was negative. Although the patient did have distinct periods of affective lability, formal thought disorder with frank psychosis, and an inability to focus his attention, these signs and symptoms were always rapid in their onset and termination. Rapid onset is inconsistent with a diagnosis of mood disorder due to a general medical condition.36 It should be highlighted that the final, multiaxial, diagnosis was arrived at only after detailed clinical observation of each of the foregoing episodes was taken into account. However, the unusual presentation observed here could also be explained by the possibility that there were many other similar episodes that the nursing staff did not detect.
Regarding Axis III, partial seizures are more common than generalized, tonic-clonic seizures.37 Complex partial seizures are remarkable for the heterogeneity of their clinical presentation.37 The altered state of consciousness observed in this patient is a hallmark of this disorder.38 There were two unusual aspects to this presentation of complex partial seizure disorder. The first was the lack of an identifiable seizure focus on the EEG report, and the second was an abnormally long period of time required for the complete resolution of the postictal periods. Although this could be accounted for by numerous unobserved episodes, we believe it may be due instead to the pathophysiology of HIV-1 infection in the brain. That is, it is known that in HIV-1 infection of the brain, abnormal neuronal membrane polarization occurs and is related to increased calcium influx into the neuronal cytosol. The latter is associated with neuronal dysfunction.39 Therefore, the therapeutic utility of calcium channel blockers might be explored in patients with seizure disorders in the context of HIV-1 infection. It is notable that nimodipine, a calcium channel blocker, has been reported—when administered with zidovudine—to have therapeutic effects on sustained cognitive-motor dysfunction in the HIV-1 infected.40 Although it should be cautioned that the aforementioned therapeutic effect was observed outside of the context of a seizure disorder, we believe this observation supports the potential therapeutic utility of nimodipine to reduce the duration of postictal confusion in the setting described herein. It should also be cautioned that the use of other calcium channel blockers may not only prove to be ineffective, but also may be deleterious.26
The issue of the reliability of the correlations between clinical manifestations of a seizure, on one hand, and both lateralization and localization of an epileptogenic focus, on the other, is of paramount importance in the neurosciences.41 This issue of reliable correlation becomes critical when there is more than one process occurring concomitantly, as is the case here. As demonstrated, EEG alone is often insufficient for diagnostic purposes; it proves itself useful only in conjunction with accurate and thorough clinical observation. The final episode discussed, although described as "confusion" like other episodes described herein, was vastly different. Here, the patient had demonstrated a distinct period of abnormally and persistently irritable mood of sufficient severity in and of itself to warrant psychiatric hospitalization; he displayed the symptoms of decreased need for sleep, being more talkative than usual, a subjective experience of racing thoughts, distractibility, and psychomotor agitation. Hence, this episode was most appropriately diagnosed as mania,36 secondary to hydrocortisone treatment—a known precipitant of mood disorders in general and of mania specifically.42 It should be noted that mood disorders associated with partial complex seizure disorder do occur, but rarely with a manic presentation.43
In summary, this case brings together a number of clinically relevant points. Of clinical importance was the failure of the law of parsimony. This patient had three distinct disorders over the time course reported herein: there was 1) a delirium on Axis I, due to 2) an associated Axis III condition (complex partial seizure disorder) followed by psychoneurotoxicity, and 3) another Axis I disorder—substance-induced mood disorder (mania) due to prescribed corticosteroid use. The delirium associated with complex partial seizure disorder was followed by a prolonged postictal period. The substance-induced mood disorder was iatrogenic. The identification of both Axis I disorders was further complicated by the ongoing and fluctuating psychomotor retardation occurring secondary to physical debilitation associated with HIV-1 infection. The aforementioned observed psychomotor retardation subsequently improved with the resolution of the attendant physical symptoms associated with his concomitant, systemic disorders.
Given that there were three distinct disorders (two on Axis I and one on Axis III) present, with a superimposed fourth factor (fluctuating psychomotor retardation), this case represents a complex clinical situation. In order to adequately approach clinical situations of this complexity, one must conceptualize all the data from a truly integrated neurological and psychiatric perspective. Target symptoms must be identified by specific and objective terms. Such an approach fosters symptom control by dictating specific neuropsychopharmacologic intervention. Not only does this case demonstrate the need for a truly integrated neuropsychiatric conceptualization of the patient, but it also demonstrates the need to maintain a low threshold for consultation to physicians of other specialties and other care providers in HIV-1—infected patients. Another prominent feature of this case is the considerable variability of presentations over time, particularly common in evaluations of HIV—1—infected patients. When approaching the HIV-1—infected patient using an integrated neuropsychiatric conceptual paradigm, the clinician should consider that the setting is one of brain infection with HIV-1, which can alter the presentation and course of neuropsychiatric and other disorders in this population.
A final point to be made about neuropsychiatric conceptualization pertains to description and communication. Inaccurate use of terminology in documenting symptoms can complicate the clinical picture greatly and needlessly. The term "confusion" was used by various health professionals herein to denote disorientation, aphasia, mood lability, tangentiality of thought process, and circumstantial thought process. "Disorganization of thought process" and "disorientation" are terms that should not be used in the presence of expressive aphasia. However, within the context of expressive aphasia, it is possible to comment on disorganized or bizarre behavior. The inaccurate use of terminology described in this case unnecessarily complicated the care of the patient and could have resulted in improper diagnosis and treatment. The need for exactness in clinical terminology is paramount in this setting wherein concomitant neuropsychiatric syndromes frequently do occur.