Letter: Psychosis Following Hormonal Therapy in NAA15-Related Neurodevelopmental Disorder

We describe the case of a 16-year-old female with an NAA15-related neurodevelopmental disorder who developed acute-onset psychosis after a single dose of an oral contraceptive. Given her history of no previous psychiatric illness and the temporal relationship between the drug intake and her symptoms, we presume this is a case of medication-induced psychosis, which will be described in detail. Additionally, we report the successful treatment of her symptoms. This case underscores the potential for gene-hormone interactions in psychiatric vulnerability and highlights the need for caution when prescribing hormonal agents in adolescents with neurodevelopmental conditions.

The patient is a female born to a 37-year old mother, who was treated with fluoxetine during the first weeks of pregnancy. She was delivered at 36 weeks and resided in the neonatal intensive care unit (NICU) for 4 hours. As an infant (7 months), global developmental delays were observed. Motor issues included atonia of the lower extremities, hypertonia of the upper extremities, muscle cramping, and dystonia in her right ankle, which caused adduction of the right foot. She was unable to bear weight on her legs, nor was she able to crawl. Her language and speech were also delayed as indicated by a battery of assessments. including the Battelle Developmental Inventory—2nd Edition (BDI-2) and the Hawaii Early Learning Profile (HELP), which demonstrated lower-than-average test scores. The patient was then deemed eligible for a number of early-intervention services. Although HELP does not provide traditional scores, BDI-2 scores from the time can be found in the Supplementary Table S1.

The family was never given a definite diagnosis, with initial considerations including cerebral palsy and the possibility that she would never walk or talk. The patient’s developmental trajectory improved over time, possibly due to early intervention therapies and hyperbaric oxygen therapy, which she started around the age of 1. Intelligence Quotient (IQ) testing (Wechsler Adult Intelligence Scale) at 4.5 years old demonstrated low average results (Supplementary Table S2). Currently, the patient demonstrates age-appropriate developmental functioning, as reflected by a Vineland-3 score of 109 (Sparrow et al., 2016). The patient also ambulates independently and with a normal gait. While formal IQ testing was not performed recently, the patient is presumed to be of normal intelligence and has recently been accepted to several colleges.

Early life metabolic testing showed abnormally high levels of creatine kinase (300 uL, normal range: 24–170 uL). In combination with the patient’s motor deficits, the possibility of a mitochondrial or neuromuscular disorder was suggested, although neither could be confirmed without a muscle biopsy, which was declined by the family. In 2023, genetic testing identified mutations in the LZTR1 and NAA15 genes. The patient’s clinical presentation most closely aligns with NAA15-related neurodevelopmental disorder (Lyon et al., 2023; Makwana et al., 2024). The LZTR1 variant, c.2387T > C (p.Ile796Thr) (hg19: 22: 21351236), is a heterozygous, autosomal recessive mutation located on chromosome 22q11.21, with a population frequency of 95 in 1,613,560 individuals according to gnomAD v4. Classified as likely pathogenic, LZTR1 is known to be involved in Noonan syndrome 2. This variant has been previously reported (Chen et al., 2014; Yamamoto et al., 2015) and is associated with ocular malformation, short stature, hypotonia, joint laxity, and congenital heart disease (Roberts, 1993). In the case of this proband, the clinical presentation included nystagmus, hypotonia, and a prenatal ultrasound suggesting cardiac abnormalities that resolved after fetal screening. However, besides hypotonia, the clinical features did not align with the published data for individuals with autosomal recessive Noonan syndrome 2 caused by two LZTR1 mutations, as previous studies also showed that 19 out of 26 individuals with this variant had hypertrophic cardiomyopathy (Pagnamenta et al., 2019).

The second vriant identified was in the NAA15 gene, c.1947 + 2T > C (hg19: 4:140291560), a heterozygous, autosomal dominant mutation located on chromosome 4q31.1. This variant, with a population frequency of zero in gnomAD v4, is classified as likely pathogenic and is associated with NAA15 neurodevelopmental syndrome. The clinical presentation of this proband appears consistent with NAA15-related neurodevelopmental syndrome, including developmental delays, hypotonia, syndactyly/clinodactyly, and prenatal cardiac abnormalities (Lyon et al., 2023; Makwana et al., 2024). However, the proband’s later development, with improvement followed by the onset of psychosis—potentially linked to birth control use and hormonal fluctuations—is a novel manifestation that has not been previously reported in the literature.

The NAA15 c.1947 + 2T > C (4-140291560 T > C, GRCh37) variant has not been reported in gnomAD. This variant affects the splicing donor site, also known as a canonical splice donor site variant, and is predicted to be deleterious by computational algorithms. Following American College of Medical Genetics (ACMG) guidelines, the variant is classified as likely pathogenic based on three criteria. (1) Null variant (nonsense, frameshift, canonical ±1 or 2 splice sites, initiation codon, and single or multiexon deletion) in a gene where loss of function is a known mechanism of disease (Pathogenic Very Strong [PVS1]). This variant is located at the canonical +2 splice site. (2) De novo variant not found in neither parent (Pathogenic Strong [PS2]). (3) Absent from general population (Pathogenic Moderate [PM1]). These three criteria provide strong to very strong evidence for pathogenicity according to ACMG standards (Richards et al., 2015).

Theoretically, this variant may result in many possibilities, including exon-skipping, intronic inclusion, activation of other splice sites, etc., but such may not be confirmed without functional studies. At this time, it is not feasible to perform functional studies due to the lack of available cell lines from the proband.

The patient was 16 years old with no previous psychiatric history when she ingested a single dose of instant release drospirenone, a progestin-only contraceptive pill, on day 20 of her menstrual cycle. It was prescribed by her gynecologist after complaints of heavy menstrual bleeding. She then experienced diplopia, emotional lability, night terrors, insomnia, and psychotic symptoms, including auditory hallucinations and paranoia, within 24 hours of taking the medication. Her symptoms gradually improved over the course of 5 days following the commencement of her period and with the drug being cleared from her system.

One month from the day of ingesting the contraceptive pill, the patient’s symptoms recurred, this time including catatonia, intermittent failure to recognize family members, fear of ingesting both liquids and solids, and alternating periods of lucidity. Subsequently, she was admitted to a medical center due to altered mental status. Extensive medical evaluation and testing yielded normal or inconclusive results. Continuous EEG was performed for >12 hours without epileptiform activity. Brain magnetic resonance imaging with and without contrast showed a Rathke’s cleft cyst. The patient’s Rathke’s cleft cyst compressing the optic chiasm could explain some symptoms of double vision but would not traditionally be associated with signs of psychosis or the bilateral stabbing pains in the breast that the patient was experiencing directly after ingestion of the pill. Despite an elevated level of dehydroepiandrosterone sulfate, low estradiol, and low luteinizing hormone levels, no other definitive endocrine abnormalities were identified. Cerebrospinal fluid (CSF) studies revealed normal cell count, glucose, and protein. An autoimmune encephalitis panel run on the serum and CSF was negative for NMDA, EBV, Bartonella, Varicella, and oligoclonal banding. A pelvic ultrasound was performed to rule out a teratoma. The patient was administered a lorazepam challenge given observed rigidity, waxy flexibility, and stereotyped movements but fell asleep and showed no improvements. Additional test results are provided in Supplementary Table S3. During hospitalization, the patient required a feeding tube for adequate nutrition due to persistent refusal of oral intake.

The patient was then treated with a single dose of 5 mg of olanzapine. The following day, which aligned with the end of the patient’s menstrual period, there was an improvement in the patient’s psychosis. While the once-daily olanzapine regimen was up-titrated to 7.5 mg and continued for the duration of her stay, the psychiatric team’s consensus was that the patient’s symptom improvements were more likely due to the cessation of her period than due to the administration of olanzapine. Additional management included the initiation of hormonal therapy with a transdermal patch (combined progesterone/estrogen) (Xulane) to mitigate hormonal fluctuations. A detailed summary of recent medications is available in Supplementary Table S4.

The patient was discharged from the hospital 2 weeks after admission, following symptom stabilization. In addition to an intensive outpatient program, ongoing care was arranged with a psychiatrist for continued monitoring and symptom management. Olanzapine tapering was initiated and completed over 4 months, and she was started on fluoxetine 10 mg daily to help manage concomitant social anxiety. Despite having tapered both the olanzapine and combined hormonal therapy for over a year, she has not experienced a recurrence of symptoms to date.

In this letter, we presented the case of a 16-year-old female who developed psychotic symptoms for the first time following the administration of an oral contraceptive. This patient’s presentation is the first documented case of psychosis in an individual with NAA-15 neurodevelopmental syndrome and warrants further consideration of other contributing factors, including the patient’s Rathke Cleft Cyst and abnormal hormonal levels, which are explored in detail in the supplementary discussion. Endocrine abnormalities in patients with NAA-15 neurodevelopmental disorder are still under current investigation and cannot be applied nor ruled out at this moment. Despite these considerations, the temporal relationship between the medication intake and the rapid onset of symptoms continues to strongly support a case of medication-induced psychosis. Overall, this case emphasizes the importance of a multidisciplinary and individualized approach when evaluating complex neurodevelopmental and neuropsychiatric presentations, particularly in the context of rare genetic variants.

Authors’ Contributions

G.J.L. and R.H. conducted all virtual interviews with participants and were responsible for primary Vineland data collection, with data curation conducted by E.M. C.C., G.J.L., and R.M. were responsible for data analysis and project conception. The first draft of the article was written by C.C. and R.M., with critical revision performed by E.M. and G.J.L. at several points. F.S.K.A. and C.M. were the clinicians involved. H.C. was consulted for genetics expertise.

Ethics Statement

Patient consent was obtained for research and publication, with approval of protocol #7659 for the Principal Investigator G.J.L. of the George A. Jervis Clinic by the New York State Psychiatric Institute—Columbia University Department of Psychiatry Institutional Review Board.

Data Availability

All data are deidentified to protect patient privacy, and the underlying data cannot be shared due to these same privacy restrictions.