Genetic variants associated with post-traumatic stress symptoms in patients with gynecologic cancer

Abstract

Objective:

Patients with cancer experience symptoms of post-traumatic stress disorder (PTSD) more commonly than the general population. The objective of this study was to identify single nucleotide polymorphisms (SNPs) associated with increased risk of post-traumatic stress disorder (PTSD) in patients with gynecologic cancer.

Methods:

A prospective cohort study recruited 181 gynecologic cancer survivors receiving care at the University of Minnesota between 2017 and 2020 who completed PTSD DSM-V surveys to self-report their symptoms of PTSD and provided saliva samples. DNA samples were genotyped for 11 SNPs in 9 genes involved in dopaminergic, serotonergic, and opioidergic systems previously associated with risk of PTSD in populations without cancer.

Results:

Most participants had either ovarian (42.5%) or endometrial (46.4%) cancer; fewer had cervical (7.7%) or vaginal/vulvar (3.3%) cancer. Two SNPS were identified as statistically significantly associated with higher PTSD scores: rs622337 in HTR2A and rs510769 in OPRM1.

Conclusions:

Genetic variation likely plays a role in development of PTSD. HTR2A is involved in the serotonin pathway, and OPRM1 is involved in the opioid receptor pathway. This information can be used by oncologic providers to identify patients at greater risk of developing PTSD and may facilitate referral to appropriate consultants and resources early in their treatment.

Introduction

Approximately 1.3 million people were living in the United States (US) with a gynecological cancer diagnosis in 2019. This number is projected to increase to more than 1.6 million in 2030(1). Gynecologic cancer morbidity can be high, including multiple physical and emotional sequelae, many of which are chronic (2,3). As the number of cancer survivors has increased over time, long-term negative effects of treatment and the disease have become a substantial public health concern. In order to develop preventive measures, better tools to identify underlying risk for long-term cancer outcomes must be developed (4).

Post-traumatic stress disorder (PTSD) is a debilitating disorder related to stress, trauma, and anxiety(5). The lifetime risk of PTSD is estimated to be 5–10% in the general US population (6,7). For some, a cancer diagnosis, treatment, and downstream sequelae are experienced as traumatic. PTSD as a psychosocial outcome of cancer has been described in cancer survivorship studies, with prevalence of up to 20% (8). Despite known increased incidence, PTSD symptoms are not routinely measured among patients with cancer and often remain unidentified and untreated. PTSD is associated with outcomes such as substance abuse, suicide, and long-term adverse psychological and physical health problems (9,10). Identifying individuals at higher risk for developing PTSD is also important because some existing interventions for preventing PTSD are more effective if administered shortly after a person experiences a traumatic event.

In the general population, up to 90% of individuals will encounter at least one traumatic event during their lifetime; however, only a small fraction will develop PTSD (7,9). Known risk factors for PTSD include female sex, experience of previous trauma, low socioeconomic status, racial/ethnic minority status, mental health comorbidities, adverse childhood events, lack of social support, and co-occurring stressors. Additionally, genetic risk factors have been suggested in populations without cancer and could explain some of the discrepancy between trauma exposure and onset of PTSD. Previous studies suggest that genetic risk may account for up to 30–40% of the predisposition to PTSD (11). Potential genetic variants that may be related to susceptibility for PTSD have been identified in the general population, but their potential effect in patients with a diagnosis of gynecologic cancer has not been assessed (12–14).

Genome-wide association studies (GWAS) and twin studies have shown genetic variation to be associated with higher risk of PTSD, with most identified genes involved in the hypothalamic-pituitary-adrenal (HPA) axis (15,16) and the dopaminergic system, including the biosynthesis, transport, degradation, transmission, and signaling transduction of dopamine (15,17). Disrupted dopamine signaling has been associated with PTSD symptoms such as reduced attention, hypervigilance, increased arousal, and poor sleep (18). Other genes implicated in the development of PTSD involve the serotonin pathway and opioid receptors (17,19). The role for opioidergic systems in PTSD is supported by the finding that administration of morphine, an opioid agonist, shortly after trauma exposure (e.g., combat injury) reduces the incidence of PTSD in military veterans (20).

Among patients with gynecologic cancer, there is no current standard for identification or subsequent treatment of PTSD despite estimated prevalence being 1.6–26%, depending on diagnosis(21). The objective of this study was to examine the associations between self-reported PTSD symptoms in a gynecologic cancer population with 10 genes (12 single nucleotide polymorphisms [SNPs]) involved in dopaminergic (COMT, MAO-A, DBH, DRD3, BDNF), serotonergic (TPH2, SLC6A, HTR2A), and opioidergic function (OPRM1).

Methods

Study Design and Population

We conducted a prospective cohort study of gynecologic cancer survivors receiving care at the University of Minnesota. Adults aged 18 years and older with a diagnosis of a gynecologic cancer (ovarian, uterine, cervical, vaginal, or vulvar cancer) who could read/write in English were recruited at time of diagnosis or during surveillance between 2017 and 2020. Detailed methods of this prospective cohort study are described elsewhere (22). All participants who were alive and still active in the cohort as of July 2019 (N=338) were mailed information (introductory letter and Oragene 600 collection kit) and given the option of providing a saliva sample for future studies. Participants provided written informed consent at the time of enrollment in the cohort and again at the time of saliva sample collection. The study received approval from the University of Minnesota Institutional Review Board (#1612S01581).

Survey Measures

Study participants completed surveys at study entry and semiannually thereafter for 3 years. Participants provided demographic information including age, education, annual household income, and partner status. Clinical data abstracted from electronic medical records included date of diagnosis, disease site, International Federation of Gynecology and Obstetrics (FIGO) stage, and treatments received (chemotherapy, surgery and/or radiation). The surveys also included validated measures of quality of life (QOL), psychosocial outcomes, and physical symptoms. Self-reported symptoms of PTSD related to cancer were collected using the validated PTSD checklist for DSM-V (PCL-V), which was modified to specifically ask about cancer-related PTSD symptoms (23). Total PCL-V scores ranged from 0–80. PTSD symptoms were measured at multiple time points, however, the total score used for the analysis here was the maximum PTSD score reported per individual. If an individual reported the same maximum PTSD score at more than one survey time point, we used the score from the first survey at which the maximum score was reported.

SNP Selection, Genotyping, and Quality Control

We identified 9 genes (11 SNPs) involved in dopaminergic (COMT, MAO-A, DBH, DRD3, BDNF), serotonergic (TPH2, SLC6A, HTR2A), and opioidergic function (OPRM1) for analysis. Genomic coordinates were aligned to GRCh38 assembly hg38.

Saliva samples were processed by the University of Minnesota Genomics Center. Salivary DNA isolation was performed per manufacturer’s instructions (DNA Genotek, USA) and quantified by Nanodrop UV/VIS spectrophotometry (ThermoFisher, USA) and fluorometry using PicoGreen staining (BioTek, USA). Genotyping was performed using the iPLEX Gold method. iPLEX reagents and protocols for multiplex PCR, single base primer extension (SBE) and generation of mass spectra were used, per manufacturer’s instructions (Agena, San Diego). Multiplexed PCR for the 12 SNPs was performed in 5-μl reactions on 384-well plates containing 10 ng of genomic DNA. Reactions contained Taq polymerase (QIAGEN), primers, Taq buffer, MgCl2, and dNTPs. Following enzyme activation at 95 °C for 5 min, DNA was amplified with 45 cycles of 95 °C × 20 sec, 56 °C × 30 sec, 72 *C × 1 min, followed by a 3-min extension at 72 °C.

Unincorporated dNTPs were removed using shrimp alkaline phosphatase (0.3 U, Agena, San Diego). Single-base extension was carried out by addition of SBE primers at concentrations from 0.625 μM (low MW primers) to 1.25 μM (high MW primers) using iPLEX enzyme and buffers (Agena, San Diego) in 9-μl reactions. Reactions were desalted and SBE products measured using the MassARRAY system, and mass spectra analyzed using TYPER software (Agena, San Diego) to generate genotype calls.

The average genotyping call rate was 98.5% across these 12 SNPs. Samples were considered to have passed genotyping analysis by producing genotype calls in greater than 70% of the SNPs; 23 samples were dropped for falling below this metric.

Statistical Analysis

Nearly all study participants who returned saliva samples self-reported their race/ethnicity as non-Hispanic white, therefore analyses were limited to those who self-identified as non-Hispanic white as we were underpowered to investigate associations in other populations (24). Participant demographic and clinical characteristics were summarized using descriptive statistics. Mean PTSD scores were compared by age (years), gynecologic cancer type (cervical, endometrial, ovarian, vaginal/vulvar), stage (I/II, III/IV), education (no college degree, at least a college degree), annual household income (<$50,000, $50,000-$99,999, $100,000+, prefer not to say), partner status (partnered, not partnered), and residential status (based on rural-urban commuting area [RUCA] codes; rural, urban) using t-tests, analysis of variance and linear regression models as appropriate. SNPs were dichotomized as carrier status, i.e., having at least one effect allele versus no effect allele. Mean maximum PTSD scores during the study follow-up were compared for each SNP separately by carrier frequencies using linear regression models adjusting for age at the survey at which the PTSD score was reported. Finally, we considered a multivariable model to include the demographic/clinical characteristics and SNPs with p-values < 0.05 in one combined model. Due to sample size limitations, we did not adjust for multiple comparisons and observed differences at p<0.05 were considered statistically significant. Analyses were conducted using SAS 9.4 (Cary, NC).

Results

Saliva samples from 213 participants (63.0% of those approached) were provided; 190 (89.2%) passed quality checks. Of these, 181 were non-Hispanic White and had non-missing PTSD outcome data and were included in this analysis.

The median age in our study population was 63.5 years (range: 24.7–88.6 years; Table 1). Of 181 participants, 178 self-reported female gender, with 3 participants with missing information on self-reported gender. Most participants had either ovarian (42.5%) or endometrial (46.4%) cancer, fewer had cervical (7.7%) or vaginal/vulvar (3.3%) cancer. Just over half of participants (59.3%) had stage I or II cancer. Almost all participants had surgery related to their diagnosis (96.1%), 64.6% received chemotherapy, and 27.1% received radiation therapy.

Table 1.

Demographic and clinical characteristics of study participants, N=181

Variable	N	Median (Min, Max)
Age at time of survey, years	181	63.5 (24.7, 88.6)
	N	%
Diagnosis
Cervical	14	7.7
Endometrial	84	46.4
Ovarian	77	42.5
Vaginal/Vulvar	6	3.3
Stage
I or II	105	59.3
III or IV	72	40.7
Missing	4
Treatments received
Surgery
No	7	3.9
Yes	174	96.1
Chemotherapy
No	64	35.4
Yes	117	64.6
Radiation
No	132	72.9
Yes	49	27.1
Annual household income
<$50,000	53	29.9
$50,000–99,999	66	37.3
≥$100,000	40	22.6
Prefer not to say	18	10.2
Missing	4
Education
No college degree	103	57.9
At least college degree	75	42.1
Missing	3
Partner status
Not partnered	68	38.6
Partnered or married	108	61.4
Missing	5

When comparing PTSD scores by demographic characteristics, we found that younger age at the time of survey was associated with higher PTSD scores, as was lower annual household income, and being diagnosed with vaginal or vulvar cancer (Table 2). We found no statistically significant differences in PTSD scores by education, partner status, urban-rural residential status or disease stage.

Table 2.

PTSD scores by demographic characteristics, N=181.

Characteristic	PTSD score
	N	Mean (SD)	P-value
Diagnosis			0.004
Cervical	14	11.00 (9.47)
Endometrial	84	7.96 (8.56)
Ovarian	77	10.13 (9.53)
Vaginal/Vulvar	6	22.00 (15.28)
Stage			0.74
I or II	105	9.39 (9.50)
III or IV	72	9.88 (9.93)
Missing	4
Education			0.97
No college degree	103	9.66 (9.24)
At least a college degree	75	9.72 (10.18)
Missing	3
Household income			0.002
<$50,000	53	13.83 (11.64)
$50,000–99,999	66	8.24 (7.31)
≥$100,000	40	7.05 (8.57)
Prefer not to say	18	9.11 (9.64)
Missing	4
Partner status			0.08
Partnered	108	8.62 (9.03)
Not partnered	68	11.22 (10.33)
Missing	5
Residential status			0.61
Rural	19	8.53 (6.11)
Urban	162	9.71 (9.91)

Two SNPs [rs622337 in HTR2A, effect carrier frequency=89.4%, Beta=4.91 (95% CI: 0.61, 9.21), p=0.03; and rs510769 in OPRM1, effect carrier frequency=40.7%, Beta=3.37 (95% CI: 0.62, 6.11), p=0.02] were statistically significantly associated with higher PTSD scores in our cohort (Table 2). When considered in a model also including annual household income and gynecologic cancer type in addition to age, the estimates were similar, though only rs510769 in ORPM1 remained statistically significant [rs622337: Beta=2.75 (95% CI: −1.45, 6.96), p=0.20; and rs510769: Beta=3.35 (95% CI: 0.75, 5.95), p=0.01].

Discussion

Identifying patients at risk of developing PTSD during the course of their cancer treatment is challenging. As such, many cancer survivors with PTSD symptoms are unidentified and untreated. This has the potential to have many long-term adverse outcomes after their oncologic treatment has ended. Biomarkers that have been suggested in order to identify PTSD include measuring serum catecholamines and/or cortisol under the hypothesis that aberrant activation of the HPA axis may result in higher rates of PTSD (25). Other approaches include having patients complete the Posttraumatic Stress Disorder Checklist, developed from DSM-V (PCL-V) criteria for PTSD prior to treatment (23,26). As approaches to personalized medicine continue to evolve, understanding the role of genetics in development of PTSD in patients with cancer is growing. A systematic review by Chair et al. in 2021 describes studies identifying SNPs associated with increased risk of PTSD in several different cancers populations including breast, thyroid, prostate, gastric, and lung (27). Another study also found genetic predisposition for PTSD symptoms in patients with hepatocellular cancer (28). In patients with breast cancer, SNPs associated with inflammatory cytokines, neuronal, and signal transduction pathways have been implicated to predispose risk for PTSD (29,30).

Our data suggest that considering genetic risk of PTSD through routine evaluation of blood or saliva samples may be useful in identifying patients at higher risk of developing PTSD symptoms during the course of their cancer diagnosis and treatment. This work identified two common SNPs that were significantly associated with PTSD: rs622337 in the HTR2A gene, associated with serotonin pathway, and rs510769 in the OPRMI gene, associated with opioid pathways. Further, these SNPs, particularly rs510769, remains statistically significant after taking into account age, household income and cancer type. This finding has potential for significant clinical benefit to patients because prior work shows intervention with mental health resources near the time of the traumatic event is more beneficial than interventions at greater time intervals after the traumatic event (37).

Regarding the individual SNPs that were identified, the intronic rs622337 variant is in the HTR2A gene, which encodes the 5-hydroxytryptamine receptor. This is a serotonin receptor which is associated with withdrawn behavior (38). Downregulation of this post-synaptic receptor has been observed when patients with PTSD are treated with selective serotonin reuptake inhibitors (SSRIs), resulting in an improvement in symptoms (39). HTR2A has been identified as a promising target for future pharmacotherapies to treat PTSD. It is also associated with other psychiatric diagnoses including depression and anxiety (39).

The second of the SNPS identified was intronic rs510769 in the OPRM1 gene, which encodes the mu-opioid receptor and is associated with opioid use disorder and is also implicated in the development of PTSD (40). Opioid administration near a traumatic event reduces risk of PTSD (37,41). It is believed that downregulation of mu-opioid receptors contributes to dysregulation the HPA axis, a central component in development of PTSD (42). OPRM1 has additionally been linked to decreased healthcare-related quality of life (43). Further, OPMR1 has also been shown to predict differential experience of cancer and postoperative pain in multiple cancer types (31). Approximately 30% of people with cancer report chronic daily pain (32,33). OPMR1 polymorphism has been associated with preoperative pain sensitivity in patients with cancer and confers increased oxycodone requirement (35–37). One showed polymorphisms were associated with oxycodone “strong-responders” vs. “non-responders” (36). This may be particularly useful for individualizing and optimizing the pain management plan for patients with cancer. It is reasonable to hypothesize that experience of pain in relation to cancer diagnosis may affect development of PTSD, however further research is needed.

While the focus of this analysis was genetic risk factors of PTSD, we acknowledge that there are numerous known demographic risk factors for PTSD, including experience of previous trauma, low socioeconomic status, and lack of social support. We observed greater reported symptoms among individuals with vulvar or vaginal cancers in this cohort. Other have observed that those with vulvar or vaginal cancer are a particularly vulnerable group (22), with many of these characteristic overlapping with risks for PTSD (44).

This study is limited to a single institution with a small sample size for a genetic study, which also limits ability to allow analysis of interaction of multifactorial components affecting development of PTSD. This led to focused choice of genes/SNPs to explore; however, ongoing research in this area using large-scale GWAS analyses is expected to identify new potential risk variants. This study is additionally limited by the racial and ethnic composition of study participants, which were almost exclusively non-Hispanic White; results may differ in different populations. Data regarding PTSD symptoms prior to diagnosis and previous experience of trauma were not available. It is possible that some participants had previously diagnosed PTSD or previously undiagnosed PTSD symptoms pre-dating cancer diagnosis and treatment.

In conclusion, we found two common SNPs, rs622337 in the HRT2A gene and rs510769 in the OPRM1 gene, that were associated with self-reported symptoms of PTSD in a population of patients with gynecologic cancer. These results suggest further research should focus on developing a comprehensive risk model for PTSD for cancer patients that captures complex pathways of genetic, demographic, and clinical risk in PTSD, which can be applicable in clinicians’ daily work to improve patient care. Knowing which patients are most at risk for PTSD-related symptoms and who may benefit from early intervention is crucial as the oncologic patient population increases.

Table 3.

Individual SNP associations (any effect allele present versus none) with PTSD scores, N=181.

Pathway	Gene	rs ID	Chromosome	Non-Effect Allele	EfTect Allele	Effect carrier frequency, N (%)	Beta*	SE	P-value
Dopamine	COMT	rs4680	22q11.2	G	A	132 (75.9)	−0.98	1.63	0.55
	MAO-A	rs6323	Xp11.3	G	T	169 (94.9)	−3.66	3.10	0.24
	MAO-A	rs6609257	Xp11.3	A	G	118 (67.1)	0.91	1.48	0.54
	DBH	rs2873804	9w34	T	C	139 (78.1)	1.22	1.68	0.47
	DRD3	rs6280	3q13,31	T	C	86 (48.0)	−0.78	1.38	0.57
	BDNF	rs6265	11p14.1	T	C	173 (96.1)	2.46	3.53	0.49
Serotonin	TPH2	rs4570625	12q21.1	T	G	165 (93.8)	3.85	2.88	0.18
	SLC6A4	rs4251417	17q11.2	C	T	47 (26.0)	1.95	1.54	0.21
	HTR2A	rs2296972	13q14.2	C	A	97 (54.2)	0.06	1.38	0.97
	HTR2A	rs622337	13q14.2	A	G	160 (89.4)	4.91	2.19	0.03
Opioid	OPRM1	rs510769	6q25.2	C	T	72 (40.7)	3.37	1.40	0.02

^*adjusted for age (years) at survey, difference in mean PTSD score among those with any effect allele (alone or in combination) versus none

Highlights.

• Patients undergoing cancer treatment are more likely than the general population to experience PTSD.

• Prior research suggests genes involved in dopaminergic, serotonergic, and opioidergic systems may predispose risk of PTSD.

• We found two SNPs in genes HTR2A and OPRM1 were associated with greater risk of developing PTSD symptoms in gynecologic cancer survivors.