A within‐subject consideration of the psychotic spectrum disorder concept in a patient in remission associated with cortical gray matter recovery

Summary

Introduction

Psychotic spectrum disorder (PSD) links the syndromes of bipolar disorder, psychotic depression, and schizophrenia, often viewed as unique disorders.

Aims

Application of the PSD concept to a single patient rather than across groups of patients and demonstration of a remarkable remission of schizophrenia phenotype with recovery of gray matter in specific brain regions.

Results

We report a woman who experienced discrete, nonoverlapping periods of each of the above syndromes, in the order noted, over a 30‐year period, followed by abrupt ending of psychosis and full remission lasting at least 7 years. This patient had 2 episodes of Bipolar 1 mania, followed by a 20‐year period of psychotic depression. From ages 35‐48, she manifested severe, paranoid schizophrenia with marked functional decline. She became refractory to antipsychotic drugs, including oral risperidone and clozapine. At age 48, while participating in a double‐blind, 6‐month clinical trial of long‐acting injectable risperidone (Consta®, 100 mg IM biweekly) for treatment‐resistant schizophrenia, at week 23, upon awakening, complete disappearance of psychosis and marked improvement in function was noted, which persisted until the present (approximately 7 years). Remarkably, cognitive test performance in most domains improved beginning at 6 weeks and reached normal levels in executive function, despite minimal improvement in psychosis until week 23. MRI studies before and after remission revealed unique and substantial increases in gray matter of the cingulate and parietal cortex, and subthalamic nucleus, not seen in other patients in this study.

Conclusions

The 3 discrete periods of psychopathology support the diagnosis of PSD. The unusual course and outcome, including remarkable improvement, in executive function and enhanced cortical gray matter in selective brain regions may have been the result of unique endogenous genetic and epigenetic factors and effect of medication.

1. INTRODUCTION

The concept that schizophrenia, schizoaffective disorder, bipolar disorder, and psychotic depression are a continuum based on extensive shared biological abnormalities, the unitary psychosis concept (“einheitspsychose”), was first postulated by Guislain, von Zeller, and Griesinger in the 18th and 19th century.1 However, the dominant nosological hypothesis for most of the 20th century was that at least bipolar disorder and schizophrenia are discrete disorders, which is attributed to Kraepelin.2 However, the unitary psychosis concept, now labeled psychotic spectrum disorder (PSD), has achieved considerable support from biological, including genetic, imaging, and biomarker studies, and cognitive assessment, which show overlap as well as some specificity among these 3 major forms of psychopathology.3, 4, 5, 6, 7, 8, 9 Treatment response data have provided evidence for and against the PSD concept in that some drugs are useful for patients who present with history and symptoms unique to one type of disorder, for example, particularly lithium for bipolar disorder, while other treatments, for example, antipsychotic drugs (APDs), are effective across the spectrum, although there are many nonresponders to either treatment within each disorder. At the current time, it is not possible to reject either of the 2 concepts each of which may be valid for certain purposes. What is clear, however, is the PSD concept has been applied across groups of patients, rather than within a specific patient, as this report is doing.

In support of the utility of the PSD concept, the symptomatic and functional outcome of most patients with schizophrenia and a subgroup of bipolar disorder patients, despite best available treatments, is generally poor, especially with regard to cognitive function, a heterogeneous deficit present in both bipolar and schizophrenia patients, although in various forms and levels of severity.10 It is the phenotype which is least responsive to current pharmacologic and nonpharmacologic interventions, for example, APDs, mood stabilizers, antidepressants, transcranial magnetic stimulation, electroconvulsive treatments, or cognitive remediation. However, exceptions to the dismal prognosis, especially for those with the schizophrenia phenotype, have been reported, even in persistently severely ill patients.11 The percentages of fully recovered patients with either disorder are unknown because most such individuals are lost to follow‐up. However, there is some evidence suggesting it is not uncommon to have major remissions, with partial or full recovery in even later stages of schizophrenia.12

Little is known about what enables prolonged remission in psychotic disorders. It may be related to disease modification based on changes in gene expression during aging, including epigenetic processes, psychotropic medications, other types of somatic treatments, effective psychotherapy, diminished stress, and many other factors.13, 14, 15 While remission in schizophrenia has occurred with APD treatment, it has also been suggested that APDs may damage neurons, interneurons, or glia, and, thus, be a contributor to poor outcome.16 The study of patients with exceptional remissions provides an opportunity to understand the interplay between endogenous biology and the potentially powerful neuroplastic effects of current psychotropic drugs which may be disease modifying.15

The purpose of this article was to report on an exceptional middle‐aged patient who manifested severe symptoms of treatment‐resistant schizophrenia, with prior history of periods exclusively characterized by bipolar disorder and then psychotic depression who manifested remarkable improvement in cognitive impairment, psychosis, and function over a 6‐month period. This remission has lasted for at least 7 years and was accompanied by improvement in gray matter in the anterior and posterior cingulate cortex, parietal cortex, and subthalamic nucleus. Understanding recovery in this exceptional case may provide critical information about the nosology, pathophysiology, and treatment of PSD as well as the basis for remission in one or more of these disorders.

2. CASE PRESENTATION

Information about history of illness, prior treatment, and treatment response was obtained from the referring physician who had treated her for 3 years prior to referral, multiple hospital, and clinic records covering a 30‐year period, from her common‐law husband of approximately 15 years, and the patient herself. There was no history of psychiatric disorders in her parents, 2 siblings, 2 adult children, and other relatives. At no time, did she engage in substance abuse nor was there any head trauma or somatic disease.

2.1. Onset of bipolar disorder and psychotic depression periods of illness

Early growth and development were reported to be normal, including school achievement, with no evidence of developmental delay or behavioral disturbances. Between ages 16 and 18, she was hospitalized twice because of acute onset of grandiose delusions, erratic, hyperactive behavior, rapid speech, lack of need for sleep, and suicidal thoughts. She was treated with mood stabilizers. Each of these typical manic episodes lasted more than a week. No further manic or hypomanic episodes were reported by the patient or described in medical records over the next 30 years. However, beginning at age 19, she experienced multiple depressive episodes, was frequently suicidal, and over the next 2 decades had 8 documented hospitalizations. These were often accompanied by delusions of a depressive nature which responded to antidepressants and APDs and recovered sufficiently to complete 4 years of college with average grades, obtain an elementary school teaching position which she held for 8 years, marry, and have 2 children. Medical records from this period, from multiple treatment venues, list diagnoses of major depressive disorder with or without mood‐congruent psychotic symptoms. Psychotic periods without mood symptoms and hypomanic or manic episodes were not present during this time, according to available medical records. Although she was able to care for herself and her children, she was unable to continue as a school teacher and received psychiatric disability support. Treatment during this period consisted of multiple courses of antidepressant drugs and APDs, which controlled mood and psychotic symptoms for varying periods of time, with relapses occurring when drug treatment was stopped. She was adherent to treatment according to family informants and medical records.

2.2. Clinical presentation at recruitment

At the time of referral for participation in a clinical trial of long‐acting injectable risperidone (CONSTA^®, to be referred to as LAIR) as a therapy for treatment‐resistant schizophrenia (TRS; Meltzer et al. 2014), the patient was a 48‐year‐old Caucasian woman, who was grossly psychotic and severely dysfunctional.17 Because of increasing impairment despite drug treatment and intensive supportive care in a skilled community mental health center not connected to the risperidone research, her partner in a common‐law marriage was considering long‐term placement outside their home. The burden on him and the level of her symptoms included persistent auditory and visual hallucinations throughout the day, minimal interest in social interaction, suicidal ideation of a chronic nature but without serious attempts, and periods of high anxiety. The diagnosis of the referring psychiatrist was chronic paranoid schizophrenia, as was that of 4 other psychiatrists who had treated her in the preceding decade as either an outpatient or inpatient. Both the patient and her husband stated she had been consistently compliant with antipsychotic and other psychotropic drugs, and psychosocial treatments. No substance abuse, major medical illnesses or exceptional stressors, was evident from review of medical records or interview of the patient, her husband, and referring medical staff.

2.3. Schizophrenia phase leading to enrollment in treatment‐resistant schizophrenia trial

At age 36, some twenty years after the first hypomanic episode and following recurrent psychotic depressive episodes, medical records, and self‐report indicate the onset of very frequent, accusatory auditory hallucinations, with less frequent visual hallucinations of paranoid, threatening encounters, in the absence of mood symptoms. During this period, multiple outpatient visits and inpatient admissions at various treatment centers included only the diagnoses of chronic paranoid or undifferentiated schizophrenia. She was treated primarily with a variety of oral APDs, and several courses of ECT when APDs failed to control positive symptoms. ECT was more effective than APDs alone. In 2006, during one hospitalization, risperidone, 8 mg/day, led to an amelioration of auditory and visual hallucinations. After several months, the patient elected to try a drug‐free period, leading to a return of auditory and visual hallucinations within 3 months. Resumption of treatment with risperidone, 8 mg/day, was ineffective. Multiple trials of various APDs at least 3 months in duration (eg, haloperidol 10 mg + ziprasidone 160 mg, olanzapine 20 mg + perphenazine 8 mg, and perphenazine 20 mg + quetiapine 150 mg) led to partial control of psychotic symptoms, but not their complete disappearance. Visual and auditory hallucinations increased in severity and were not accompanied by symptoms of bipolar disorder of psychotic depression. A course of ECT led to partial improvement but auditory and visual hallucinations returned within a few months. Another course of ECT was ineffective. She was then treated with clozapine, (900 mg/day for 3 months) with minimal improvement, followed by paliperidone, 12 mg/day, olanzapine, 30 mg/day + risperidone, 6 mg/day, olanzapine 40 mg/day + quetiapine 100 mg/day, and loxapine 250 mg/day + quetiapine 800 mg/day. According to medical records, the patient and her husband, none of these regimens diminished the auditory or visual hallucinations. The auditory hallucinations included voices commanding her to commit suicide or to kill her husband. During this time, she made several serious suicide attempts for which she was hospitalized. The lack of efficacy of medications was not due to adherence problems, as she was voluntarily compliant and her husband supervised her medication utilization. At the time of referral to the LAIR clinical trial for TRS patients, despite receiving quetiapine 1500 mg/day, and aripiprazole 20 mg/day, citalopram 60 mg/day, and carbamazepine 400 mg/day for nearly 6 months, she had markedly severe auditory and visual hallucinations which led to referral to the LAIR clinical trial as a last resort to avoid nursing home placement.

2.4. Treatment with long‐acting injectable risperidone

2.4.1. Minimal change in positive symptoms

In 2010, after referral by her treating physician and assent by her husband, the patient provided written informed consent to be evaluated for participation in a 24‐week double‐blind, randomized, clinical trial comparing 2 doses (50 and 100 mg biweekly) of intramuscular LAIR. The results of this 160 patient, 2 arm study have been published elsewhere.17 After meeting all entry criteria and providing written informed consent, with the assent of her husband, she was randomized to receive 100 mg LAIR q2 weeks IM. She was noted to have mild‐moderate oral‐buccal tardive dyskinesia (TD) which had been stably present for at least 5 years. Her PANSS‐total score of 113, PANSS‐positive symptom subscale score of 25, and PANSS‐negative symptom subscale score of 32 were the highest of any of the 160 patient who participated in this multicenter trial. Initial treatment was LAIR, 75 mg IM every 2 weeks for 6 weeks, which was increased per protocol to 100 mg IM every 2 weeks. For the first 2 weeks, supplementary oral risperidone, 8 mg/day, was administered in addition to the 75 mg LAIR. Quetiapine and aripiprazole were discontinued during this period, but treatment with citalopram, 60 mg/day, and carbamazepine, 400 mg/day, was continued per protocol. Thereafter, she received LAIR 100 mg biweekly along with the other 2 drugs for the duration of the study.

Over the next 3 months, PANSS‐total score decreased from 113 to 100 due to improvement in items unrelated to psychosis, for example, general psychopathology and negative symptoms (Figure 1). There was a slight, clinically nonsignificant, improvement in ratings of the PANSS‐positive subscale. The patient reported continuous, persistent, and severe hallucinations, minimally changed from the time of study entry, to her husband and to the blinded study rater. There was also minimal change in overall function, as noted in the CGI.

Figure 1.

PANSS total and subscales and BPRS anxiety depression ratings from baseline to week 24 ratings. To illustrate the abrupt nature of the improvement in psychopathology, the week 24 rating is not connected to week 18 and an estimate of week 23 based on the slope of the week 12‐18 ratings is provided, connected by a dotted line to week 18

2.4.2. Gradual improvement in cognitive test performance

At baseline, all domains of cognitive function test scores were below the means of the group in the study and indicated severe impairment on tests of attention, executive function, working memory, verbal memory, speed of processing, and semantic memory. Her performance was the worst of all 160 patients in this study (see Meltzer et al. 2014 for group data).14 However, testing showed partial improvement beginning at the first retest week 6 of treatment starting oral risperidone and LAIR, with continuation of citalopram and carbamazepine (Figure 2). The improvement was known only to the research assistant who did the testing.

Figure 2.

Time course of cognitive test scores between baseline and week 24 for case and rest of cohort.17 Note remarkable improvement in executive function (WCST‐categories and % perseveration) and working memory (Auditory Consonant Trigram Test and Digit Span) at week 24

Attention/working memory (assessed by the Peterson's Consonant Trigram Test, Digit Span), executive function (Wisconsin Card Sorting Test, Categories and Percent Perseveration) improved markedly by week 6 and even more at weeks 12, and remained at high level at week 24. Semantic memory/verbal fluency (Controlled Word Association Test and Category Instance Generation Test) also improved at week 6, but only Controlled Word Association Test showed further improvement between weeks 12 and 24. The most remarkable improvement was in the Wisconsin Categories test of executive function which is resistant to improvement with practice in patients with schizophrenia. At the end of study, performance was above the means of the other patients with schizophrenia in the study (Meltzer et al. 2014) and at the level reported in the general population.17, 18 None of the other patients in this study showed improvement in this measure to this extent.

2.4.3. Abrupt end of psychotic symptoms at week 24

For the first 5 and ½ months of the study, neither the patient or her husband, as well as the attending clinician, noted any significant change in severity or frequency of auditory and visual hallucinations. However, PANSS ratings indicated modest improvement, as reflected in the small decreases in the PANSS‐total and subscales at the scheduled ratings. Remarkably, a complete cessation of both auditory and visual hallucinations occurred overnight 5 1/2 month after the first LAIR injection. The patient and her common‐law husband reported to the research staff that she noted all auditory and visual hallucinations were absent upon awakening from an overnight sleep which had been preceded by an unremarkable day, with regard to the type, severity, and functional consequences of her customary high level of positive symptoms. Her reaction to noting the loss of the psychotic symptoms was to search her thoughts to find them but could not. She was fearful they would return over the next few days but they did not. When questioned by the clinical study staff within the first days thereafter, and then by one of us (HM) at a videotaped visit 2 weeks later, she emphasized the fear they would return and her awareness of being able to perform mental functions requiring memory, attention, planning, and concentration in a manner that she said was impossible during the preceding periods because of interference from the auditory and visual hallucinations. She expressed the belief that the positive symptoms were the major cause of her cognitive and functional impairment and that once they ended; she could function at a much higher level, which included independent shopping and visits with family members (Figure 3).

Figure 3.

CGI severity, GAF, and PSP total ratings from baseline to Week 16. Note the marked improvement between baseline and week 12. There were no week 18 ratings. The change in improvement on function between the end of the psychosis at week 23 and the week 24 assessment is not reflected in the rating because there was no opportunity for the blinded rater to rate her following the ending of her psychotic symptoms

This functional improvement was confirmed by her husband and clinicians involved in her care. However, it should be noted from Figure 2 that her cognition was much improved even when her psychosis was still severe. Nevertheless, her common‐law husband confirmed that her previous inability to function independently ended along with the cessation of positive symptoms. With a brief exception described below, she remained free of psychotic symptoms over the next 7 years, while staying on the same dose of LAIR, without asking for, or suggestions from the nonresearch clinicians who treated her after the time limited protocol ended, to alter her medication regime. There has been no progression in the severity of her tardive dyskinesia symptoms during this period. During one stressful period related to her marital relationship, approximately 1 year after the remission of her psychosis and onset of functional improvement, she had a mild return of the auditory hallucinations. She recovered from that within a month and has maintained her global improvement in function. She now enjoys improved relationships with members of her biological family. Information about the patients’ current symptoms and function has been obtained through direct contact with the patient and with the current clinical staff where she receives LAIR. She has not been hospitalized during this follow‐up period and has reported minimal mood symptoms

2.4.4. Change in motor symptoms

During the course of the study, there was a clinically significant increase in Simpson‐Angus extrapyramidal ratings, peaking at week 18, with an 8 point (40%) increase in Simpson‐Angus Total score with a decrease at 24 weeks of 5 points but still above baseline levels. The AIMS score showed mild TD at baseline, remaining fairly constant throughout the study and an appreciable drop at 24 weeks (Figure 4).

Figure 4.

Note increase in EPS at week 16 and decrease in dyskinetic symptoms at end of study following the improvement in psychosis and slight decreases throughout the study

2.4.5. Changes in brain gray matter at 6 months

High‐resolution structural MRI images (T1‐w, 1 mm isotropic resolution) adequate for analysis were obtained from this patient and 22 other patients in the study at baseline and 6 months, or study endpoint, using a 3T Philips Achieva scanner. The reasons for the limited neuroimaging data were twofold: Funding by the study sponsor was limited to 30 subjects at the Vanderbilt University site, and 7 subjects had at least one poor quality scan, or lacked one scan, because of unwillingness to participate. The details of the methods of acquisition are available upon request. Briefly, images were segmented and normalized to Montreal Neurological Institute standardized space using the voxel‐based morphometry (VBM) toolbox in SPM5.19, 20 Changes in gray matter between “pre‐” and “post”timepoints in the case of interest were compared to changes in gray matter of the rest of the group by calculating a z‐score on a voxel‐by‐voxel basis using the equation $z=\frac{x_{\mathrm{case}}-{\mathrm{\mu }}_{\mathrm{pop}}}{\mathrm{\sigma }}$ where $x_{\mathrm{case}}$ represents the gray matter change in the case of interest in a given voxel, ${\mathrm{\mu }}_{\mathrm{pop}}$ represents the population's mean gray matter change in that voxel, and σ is the population's standard deviation. Maps were thresholded at z = 2. These maps revealed greater gray matter volume change in the case of interest near the cingulate gyrus (Figure 5) where peak z = 4.61 in the right posterior cingulate gyrus (Brodmann area 30) and z = 4.16 in the left anterior cingulate gyrus (Brodmann area 24), bilaterally in the subthalamic nucleus where the peak z = 2.33 (Figure 6), and in the right parietal postcentral gyrus (Brodmann area 3; MNI coordinates (33 ‐32 49)) with peak z = 4.23 (map not shown). Hence, the patient's improvements in clinical symptoms and cognition were accompanied by significant increases in gray matter volume in these regions. The subject of this case report was the only one of the 22 subjects with pre‐post MRI scans who showed significantly increased gray matter density. Each of these brain regions contributes to cognitive function.

Figure 5.

Gray matter volume increases in the left anterior cingulate (left panel) and the right posterior cingulate (right panel), displayed on the group‐average structural scans (color scale indicates F‐statistic)

Figure 6.

Bilateral gray matter volume increase in the subthalamic nucleus (STN), displayed on the patient's structural scan (STN location given in MNI and Talairach coordinates)

3. DISCUSSION

We report a case of PSD with an exceptional clinical course, including remarkable improvement in cognitive and functional impairment, and remission in psychotic symptoms, in a patient, we propose is best diagnosed as PSD, despite manifesting quintessential schizophrenia psychopathology of great severity at the time of presentation, 6 months before remission, and the absence of uniquely mood disorder symptoms which had been characteristic of the early course of an illness which first manifested itself at age 16, some 34 years prior to remission.

3.1. Diagnosis of psychotic spectrum disorder vs schizophrenia, bipolar disorder or psychotic depression

The case for diagnosis of PSD rather than schizophrenia, bipolar disorder, or psychotic depression, is based on the perspective of long‐term course rather than cross‐sectional presentation at the time of study entry or her earlier history of mood disorder symptoms. From a cross‐sectional clinical perspective, the diagnosis of schizophrenia would be most justified while considering her as “poor outcome” bipolar disorder or depression with psychotic features does not appear warranted because of the absence of any features of these disorder for 2 or 3 decades. Notably, all 3 types of major psychotic disorders were present at different times, without admixtures.

The illness in this patient was characterized by 3 distinct periods each characteristic of PSD in our view. First, was a bipolar/mania period which included 2 manic episodes between ages 16‐18. This was followed by a 20+ ‐year period of exclusively depressive episodes, with frequent depressive delusions, a rarity in bipolar disorder (discussed below), most consistent with the diagnosis of psychotic depression. The third period was a 10‐year period of positive symptom schizophrenia, without episodes of mood‐congruent delusions. During the 6 years, as the cessation of the delusion and hallucinations, while remaining on risperidone, she experienced one brief, stress‐related period of psychosis which resolved quickly with increased dose of risperidone. Nevertheless, this indicates persistent vulnerability to become psychotic. The onset of a schizophrenia phenotype at approximately age 36 is uncommon in schizophrenia. Less than 5% of females with neuroleptic resistant schizophrenia have an age at onset of 38 or later.21

3.2. Improvement in cognition preceding improvement in psychotic symptoms

Improvement in cognition to the extent that it occurred in this patient, and its onset when psychosis was still severe, is unique, not only compared to the other patients in this study, but in our experience assessing these dimensions of schizophrenia in over 500 patients with schizophrenia. The improvement within 6 weeks of starting LAIR in most, but not all cognitive domains, is unlikely to be a practice effect, because the magnitude of practice effects is usually within 0.25 standard deviations of the baseline mean scores.22 That improvement did not occur for all domains suggests that improvement was selective for regions of the brain which is consistent with the imaging results. Particularly, marked improvement was observed in working memory and executive function. These are predominantly frontal lobe‐based cognitive functions and involve cingulate cortex, where improvement in gray matter was most marked. The onset of improvement while positive symptoms were severe is inconsistent with the patient's belief that positive symptoms prevented her from successful test performance, but consistent with lack of improvement in activities of daily living until the psychotic symptoms ended. It is noteworthy that the improvement in overall function required improvement in both cognition and the cessation of positive symptoms. This runs counter to general theory of greater dependence of functional ability on cognitive function than positive symptoms, which may still be the case for some patients with schizophrenia. There is substantial evidence that in schizophrenia patients, it is customary for positive symptoms to respond to APD treatment more rapidly than does cognitive impairment.23 Indeed, despite the considerable doubt that APDs are effective in treating the cognitive impairment associated with schizophrenia (Keefe and Harvey, 2012), there is extensive evidence that it does respond in some patients.23, 24 Thus, it is noteworthy that in this patient, the initial recovery of cognitive function, especially in working memory and executive function, occurred first, followed by significant improvement in PSP and CGI ratings beginning at 6, and becoming quite robust, by 12 weeks, with minimal improvement in positive symptoms, until their abrupt disappearance at week 23 of LAIR treatment. The reverse order of improvement is more common.25

Remission in schizophrenia has been frequently reported, although not usually in patients with duration of illness as long as this patient, and rarely in such an abrupt manner.25 For someone with early onset bipolar disorder, her initial type of psychopathology, the absence of other manic episodes over the next 35 years is an exceptionally long time without recurrence of hypomanic or manic episodes (Robert Post MD, personal communication, December 2017).26 This raises the possibility that she experienced a spontaneous or treatment‐induced remission of bipolar disorder sometime after her second manic episode. The same may be said of the period of frequent psychotic depressions, which lasted nearly 2 decades, to be followed by the onset of schizophrenia symptomatology. It is possible that the biological basis for the mood disorder phases ceased spontaneously or was the result of drug treatment, or some combination of both. Finally, a decade‐long period of severe paranoid schizophrenia ended, during the close monitoring of a clinical trial, with no subsequent return of any type of psychopathology other than a brief period of return of psychotic symptoms. Thus, this patient may be considered to have experienced 2 or 3 remissions, which may be biologically unique or the result of a common process.

3.3. Improvement in gray matter

The possibility that the recovery from the schizophrenia phase and remission for 7 years was based, in part, on neuronal structural recovery which occurred between the 2 scans, 6 months apart, in cingulate and parietal cortex, or subthalamic nucleus gray matter, is intriguing. These changes were identified by comparing the scan of the subject of interest with that of 21 other patients in the study who did not experience remission or recovery of the same magnitude. We have been unable to find a report of comparable changes associated with clinical improvement in schizophrenia. Indeed, review of pre‐post treatment with APDs suggests that loss of gray matter in patients with schizophrenia or bipolar disorder is the most frequent change with the anterior cingulate and related areas, for example, the insular cortex being the most frequent finding.27 Of the 4 regions showing increased gray matter, the posterior cingulate cortex which forms a central node in the default mode network (DMN) of the brain which communicates with multiple brain networks simultaneously and is involved in various brain functions, including working and episodic memory and social cognition would appear to be of special interest in this case.28, 29 In a study of remitted first‐episode patients with schizophrenia, connectivity analysis suggested mild deficits in the DMN were present in the remitted patients.30 Additional evidence for the DMN and remission has been reported by Manoliou and colleagues.31 Loss of key synaptic proteins that are essential for spine formation and plasticity in cortical and hippocampal areas, including the subthalamic nucleus, is likely to be critical for the pathophysiology of schizophrenia and response to APD treatment.15, 32

There was nothing remarkable about the baseline brain MRI in this patient, including total gray matter compared with other patients in the study, or specific regional shapes or sizes, to suggest some comparable improvement in gray matter as the basis for the remission of the bipolar and psychotic depressive periods previously noted. Conceivably, gray matter renewal in other regions of the brain may have been responsible for the termination of those periods. However, there is some evidence, linking improvement in gray matter, lithium treatment, and bipolar disorder with the anterior cingulate cortex.33

The increased gray matter is more likely to reflect enlargement of the neuropil for both principal neurons and interneurons rather than neurogenesis which is not known to occur in regions of adult primate brain outside of the hippocampus.34 However, it is possible that neurogenesis in the hippocampus might induce changes in other parts of the brain as connectivity began to be restored, through gamma and theta oscillations. There is extensive evidence that reduced gamma oscillations, generated by a variety of inhibitory neurons reported to be lost in patients with schizophrenia and mood disorders, may be the cause of cognitive impairment in these disorders.35, 36 It is possible that the increased gray matter included restored interneuron synaptic structure and function, enabling gamma oscillations. Gamma oscillations are particularly prominent during nonrapid eye movement (non‐REM) sleep and are important for integration of information across distant brain regions involved in memory consolidation.37 It is also of interest that neurogenesis has been linked to sleep (Akers et al. 2018) since recovery from psychosis was achieve while the patient was asleep.38 Neurotrophic factors such as brain‐derived neurotrophic factor (BDNF) and various neuregulins are likely to be key to the regrowth of neuropil or neurogenesis.39, 40 Both of these neurotrophins have been shown to be stimulated by atypical APDs, as well as many other mechanisms.40 Neuregulin has been linked to sleep and recognition memory.41 Further study of neurons which can be derived from this patient's lymphocytes or other cells might be useful to determine a propensity for synaptic plasticity, including neurogenesis.

3.4. Another case in point

Bigelow et al. (1981) reported a case of a schizoaffective woman hospitalized continuously for 40 years because of lack of response to APDs or lithium alone, who on no treatment during a research protocol, manifested a 50 day cycle of hyperactivity, other manic symptoms, including delusions and thought disorder, and which was then unresponsive to lithium treatment alone. Treatment with the combination of lithium and acetophenazine, a typical APD, led to gradual response over a four month period, including complete cessation of thought disorder and delusions. Cessation of the lithium led to a return of the 50 day cycle within the next 50 days. Restoration of the drug combination blocked the cycle, enabled discharge to a halfway house, and prevented recurrences for the duration of follow up, at least two years. The authors initiated a placebo‐controlled, double blind crossover study in which lithium carbonate was added to neuroleptic treatment in 11 schizophrenia and 7 schizoaffective patient with ‘significant improvement in 8 patients but not remission in any (Carmen et al. 1993).

What is most striking about these 2 cases is that the response to treatment began gradually, took 6 months before it was complete, and, based on the published cycle data in the second case, the end of the cycling also occurred abruptly. A significant difference between the 2 cases is that mood symptoms and schizophrenia psychopathology were concomitant rather than sequential.

3.5. Changes in gray matter in relationship to psychopathology

Loss of gray matter, due to loss of neurons and thinning of the neuropil, has been described in both schizophrenia and bipolar disorder.33, 44, 45, 46, 47 The loss of gray matter in schizophrenia has been reported in many brain areas, including bilateral insular cortex, anterior cingulate cortex, left parahippocampal gyrus, left middle frontal gyrus, postcentral gyrus and thalamus.45 Bipolar patients, in general, manifest less gray matter loss compared to normal controls, than do schizophrenia patients, with loss concentrated in cingulate cortical areas.47 Gray matter loss in schizophrenia and bipolar disorder has been attributed to excessive pruning and inflammation during early development, as well as during the prodromal period, with genetic factors and stress considered to be the major basis for the loss of gray matter.48, 49, 50 Longitudinal MRI studies indicate progressive loss of brain volume in some patients with schizophrenia or bipolar disorder.51, 52 On the other hand, the possibility that drug treatment, particularly APDs, may be the cause of gray matter loss, has been extensively investigated with some concluding that there is a dose‐related reduction of gray matter due to APDs.16, 53 A recent meta‐analysis of 18 longitudinal studies of patients reported that typical APDs, such as haloperidol and phenothiazines, were associated with greater loss of gray matter than atypical APDs.54 There have also been some reports of increased gray matter volume early in the course of treatment with APDs, which include speculation that dopamine (DA) D₂ receptor blockade produces neuroplastic changes which contribute to their efficacy in treating psychosis.55 Atypical APDs such as clozapine and olanzapine have been reported to be more likely to have positive effects on synaptic morphology and function than typical APDs.56

4. CONCLUSIONS AND COMMENT

The variability in schizophrenia and bipolar disorder, as well as the overlap in biological measures and response to some APDs, ECT, mood stabilizers, and antidepressant drugs, in those diagnosed with these disorders, demonstrates that shared phenotypes may result from diverse combination of etiologic process. As mentioned above, in terms of causation of gray matter changes over time, genetic and epigenetic factors, stress, inflammation, drugs of abuse, especially cannabis, stimulants, and psychotomimetics, contributing to the diversity of phenotypes, trajectories of illness, and treatment outcomes. For this reason, group data reflect modal results of diverse pathologic processes. Exceptional deviations from the modal course of illness, in terms of form, severity and outcome, may be the result of multifaceted disease processes which can be identified by intensive investigation of outliers in psychopathology or response to treatment.

Remission of symptoms in schizophrenia is rare in patients with TRS.53, 57, 58, 59 Among TRS patients, in fact, most such patients respond to APDs in early episodes, and become persistently treatment refractory only much later in the first decade of illness.60 While clozapine is often effective in such patients, improvement to the extent noted here is rare.61 It is of interest that a 3‐month trial of high dose clozapine in this patient's TRS phase produced no improvement in psychosis or function, suggesting her improvement may not be related to the processes that lead to response and remission with clozapine. This suggests that understanding the biological basis for the remission in this patient lead to information about recovering synaptic function even after many years of illness, information that could lead to novel treatments for others.

CONFLICT OF INTEREST

HYM has received grant support from Janssen Pharmaceuticals, Allergan, Eli Lilly, Lundbeck, Neurocrine, Takeda, Sumitomo Dainippon, Sunovion and by a donation from the Weisman family.

Meltzer HY, Sim MY, Anderson A, et al. A within‐subject consideration of the psychotic spectrum disorder concept in a patient in remission associated with cortical gray matter recovery. CNS Neurosci Ther. 2018;24:641–651. 10.1111/cns.12986