Prostate Cancer, Use of Androgen Deprivation Therapy, and Cognitive Impairment: A Population-Based Study

Abstract

Introduction:

The association of prostate cancer and androgen deprivation therapy (ADT) use with the odds of developing mild cognitive impairment (MCI) was determined in men from the population-based Mayo Clinic Study of Aging (MCSA).

Methods:

The study included 2513 men (mean age of 73.1 years) enrolled in the MCSA. A history of prostate cancer, ADT use, and length of ADT exposure prior to their first MCSA visit was abstracted using the Rochester Epidemiology Project medical records-linkage system. MCI was diagnosed at the baseline visit. Logistic regression was used to determine whether prostate cancer and ADT use was associated with odds of MCI.

Results:

Of the 2513 participants, 349 (13.9%) had a history of prostate cancer; among whom 99 (28.3%) were treated with ADT prior to MCSA enrollment. There were 382 (15.2%) with a diagnosis of MCI. In the univariate logistic regression models, prostate cancer (OR 1.50; 95% CI, 1.12–2.00) and ADT exposure (OR 1.57; 95% CI, 0.96–2.58) were associated with higher odds of MCI. These associations were greatly attenuated and not significant in multivariable models.

Conclusions:

Neither a diagnosis of prostate cancer nor ADT use was associated with odds of MCI in this cross-sectional population-based study.

Prostate cancer is the most common cancer in men after skin cancer.¹ Circulating serum androgens, such as testosterone, promote the growth and development of prostate cancer cells. Thus, lowering levels of these hormones is a common and effective treatment strategy for many patients with this disease. It is estimated that approximately 50% of patients diagnosed with prostate cancer will receive androgen deprivation therapy (ADT) at some point following their initial diagnosis.² Historically, ADT was reserved mostly for patients who had developed metastatic, incurable disease. However, it is now often utilized in earlier stages of prostate cancer, as there is a survival benefit when given in conjunction with radiation therapy.^3,4

ADT is associated with a number of adverse effects including fatigue, erectile dysfunction, hot flashes, and osteoporosis.^5,6 Some studies,^7–9 but not others,^10,11 suggest that ADT use is associated with cognitive impairment. Cognitive impairment among patients with any type of cancer may be due to a combination of biological, psychosocial, and treatment factors. Biological processes in cancer such as DNA damage, oxidative stress, and immunological alterations share a role with the natural aging process, and may contribute to cognitive impairment seen in cancer patients. Men with prostate cancer who receive ADT have lowered testosterone levels, serving as an added factor that may precipitate them to cognitive impairment. Cancer-related cognitive impairment has a significant impact on the ability of patients to make informed choices about their treatment and on activities of daily living and quality of life.¹² The main objectives of this study were to determine, among men enrolled in the Mayo Clinic Study of Aging (MCSA): 1) the association of prostate cancer and the odds of mild cognitive impairment (MCI), and 2) whether ADT use among the prostate cancer patients was associated with odds of MCI.

METHODS

This was a case control study which included all men at their visit of enrollment into the MCSA, a prospective population-based study that began in 2004 in Olmsted County, Minnesota.¹³ The study initially recruited Olmsted County residents between the ages of 70 and 89 using an age- and sex-stratified random sampling design within the context of the Rochester Epidemiology Project (REP) medical records-linkage system. Since 2004, the population has been re-enumerated several times and extended to cover the ages of 50–90+ years following the same sampling strategy. Olmsted County residents were randomly selected from the population and were invited to participate in the MCSA. The baseline MCSA visits occurred between 2004 and 2018.

The baseline MCSA visit included a physician examination, an interview by a study coordinator, and neuropsychological testing administered by a psychometrist.¹⁴ The physician examination included a neurological examination and administration of the Short Test of Mental Status.¹⁴ Study coordinator interviews included participant demographic information, assessments of depression and anxiety, and the participant and informant Clinical Dementia Rating scale.¹⁵

The neuropsychological battery included nine tests covering four domains: 1) memory (Auditory Verbal Learning Test Delayed Recall Trial¹⁶; Wechsler Memory Scale-Revised Logical Memory II & Visual Reproduction II¹⁷); 2) language (Boston Naming Test¹⁸ and Category Fluency¹⁹); 3) executive function (Trail Making Test B²⁰ and WAIS-R Digit Symbol subtest²¹); and 4) visuospatial (WAIS-R Picture Completion and Block Design subtests).²¹ Using the mean and standard deviation (SD) from all participants included in this study, test scores were converted to z-scores. Global cognition was calculated using the z-transformed averages of the four cognitive domains.

For each participant, performance in a cognitive domain was compared with the age-adjusted scores of cognitively unimpaired (CU) individuals previously obtained using Mayo’s Older American Normative Studies.²² This approach relies on prior normative work and extensive experience with the measurement of cognitive abilities in an independent sample of volunteers from the same population. Participants with scores around 1.0 SD below the age-specific mean in the general population were considered for possible cognitive impairment. A final decision was made after considering education, occupation, visual or hearing deficits, and reviewing all other participant information. The diagnosis of MCI was made by a consensus agreement between the study coordinator, examining physician, and neuropsychologist using published criteria.²³ The diagnosis of dementia and AD²⁴ was based on published criteria. Participants who performed in the normal range and did not meet criteria for MCI or dementia were deemed CU.

A diagnosis of prostate cancer, use of ADT and other therapies were abstracted using the REP medical records linkage system. Nurse abstractors looked through the medical records of all MCSA participants and confirmed all cancer diagnoses. The variables collected for prostate cancer included: date of diagnosis (including recurrence, staging, and treatment type (surgery, radiation, chemotherapy, hormonal treatment, other). Methods of androgen deprivation therapy were also identified (gonadotropin-releasing hormone [GNRH] agonists, antiandrogens, orchiectomy, other) and length of exposure to these interventions were recorded along with time from initial ADT exposure to inclusion in MCSA.

Assessment of Covariate

Demographic variables (eg, age and education) were collected by self-report during the in-clinic exam. Medical conditions (eg, hypertension, diabetes) and the Charlson comorbidity index²⁵ were determined for each participant by REP medical record abstraction. Participants’ blood samples were used to determine APOE genotypes.

Statistics

Logistic regression was used to examine the association between prostate cancer, with and without ADT use, and odds of MCI at the baseline visit.

Multivariable models adjusted for age, education, APOE genotype, Charlson comorbidity index, anxiety and depression. All analyses were conducted using JMP Pro 13 software.

RESULTS

Amongst the 2513 men included in this study, the mean age was 73.1 years and the mean years of education was 14.5 (Table 1). Of the 349 prostate cancer patients, 99 (28.3%) used ADT prior to or at their first MCSA visit. There were 382 (15.2%) participants with a diagnosis of MCI. A further description specific to participants with a history of prostate cancers and their treatment modalities is included in Table 2. Of note, the mean survival for patients with metastatic disease was 71 months with a median of 46.5. A history of prostate cancer was associated with increased odds of MCI in univariate models (OR 1.50; 95% CI, 1.12–2.00; Table 3). Further, there was a modest dose response relationship with prostate cancer and ADT use in univariate analyses such that compared to men without prostate cancer, those with prostate cancer who did not have a history of ADT use had a 1.32 increased odds of MCI (95% CI, 0.94–1.86), and those with a history of ADT use had a 1.57 increased odds (95% CI, 0.96–2.58). However, the difference between these two groups was not significant.

TABLE 1.

Participant Characteristics by Mild Cognitive Impairment Status at the Baseline Mayo Clinic Study on Aging Visit

	Cognitive Impairment
	No		Yes
	N	Mean (SD) / N (%)	N	Mean (SD) / N (%)	P-value
Age at first visit	2131	72.05 (9.49)	382	78.5 (7.42)	<0.0001
Education	2130	14.73 (2.96)	382	13.07 (3.50)	<0.0001
Charlson index	2109	3.39 (3.09)	379	5.24 (3.49)	<0.0001
BAI	2127	2.37 (3.61)	381	4.63 (5.71)	<0.0001
BDI (>13)	2093	148 (7.1)	364	64 (17.5)	<0.0001
Any APOE4	2023	555 (26.8)	362	116 (32.0)	0.0563
Diabetes mellitus	2112	419 (19.8)	380	88 (23.2)	0.0467
Hypertension	2112	1438 (68.1)	380	309 (81.3)	<0.0001
Prostate cancer	2131	282 (13.2)	382	67 (17.5)	0.0064

APOE4 indicates apolipoprotein E ε4; BAI, Beck Anxiety Index; BDI, Beck Depression Inventory.

TABLE 2.

Description of the Primary Treatment Setting and ADT use Among Men with a History of Prostate Cancer^*

Primary Treatment Setting	Total PCa Patients, N (%)	Received ADT, N (%)	Avg ADT Exposure Months, Mean (SD)	ADT > 12 Months, N (%)
Surgery	225 (64.8)	44 (44.4)	88.4 (77.7)	31 (57.4)
Radiation	138 (39.8)	60 (60.6)	41.1 (64.0)	22 (40.7)
Metastasis	41 (11.8)	26 (26.3)	71.0 (70.6)	19 (35.2)

ADT, Androgen deprivation therapy; PCa, Prostate cancer

^*The treatment groups are not mutually exclusive.

TABLE 3.

Cross-sectional Univariate Associations Between Participant Characteristics and Odds of Mild of Cognitive Impairment

Variable	N	Odds Ratio	Low 95% CI	High 95% CI
Age (y) at first visit	2513	1.09	1.08	1.11
Education (y)	2512	0.82	0.80	0.86
Charlson index	2488	1.18	1.14	1.21
BAI	2508	1.11	1.09	1.13
BDI (>13)	2457	2.74	2.01	3.75
Any APOE4	2385	1.26	0.99	1.61
Diabetes mellitus	2492	1.30	1.00	1.69
Hypertension	2492	2.22	1.71	2.89
Prostate cancer	2513	1.50	1.12	2.00
Age at diagnosis	347	1.03	1.00	1.07
ADT exposure	99	1.57	0.96	2.58

ADT indicates androgen deprivation therapy; BAI, Beck Anxiety Index; BDI, Beck Depression Inventory.

All associations were greatly attenuated in multivariable models, specifically after adjusting for age (p-value for prostate cancer and ADT use > 0.70). We conducted a sensitivity analysis stratifying ADT exposure by ≤ 12 months or > 12 months, but we still found no association with MCI in multivariable analyses.

DISCUSSION

Survival rates for prostate cancers continue to improve, resulting in an older population of men at risk of cognitive impairment. In the present study, we examined the relationship between prostate cancer and odds of MCI in the population-based MCSA, as well as the use of ADT. Using a case-control study design, we did not find an association between a history of prostate cancer and odds of MCI among men aged 70 and older. We also did not find an association between a history of ADT use, including the length of exposure, and odds of MCI.

The association between cancer and dementia has been controversial, with some studies showing that a history of cancer is associated with a reduced risk of dementia whereas other studies show an increased risk or no association.^26–29 Many of these studies did not examine specific cancer subtypes nor did they examine cancer in relation to MCI, a prodromal stage of dementia. In the current study, we did not find an association between a history of prostate cancer and odds of MCI among men aged 70 and older after adjusting for covariates.

Low levels of testosterone may have a negative impact on neuronal connectivity and neurotransmitter modulation, eventually leading to impairment in the executive and visuospatial functions.³⁰ This has led to the hypothesis that prostate cancer patients using ADT medications might be at risk of cognitive impairment. However, studies examining the association between ADT use among prostate cancer patients and cognitive impairment or dementia have been equivocal, with some studies showing increased risk and others not finding an association.^{7–11,31–33} The conflicting results at least partly reflect the heterogeneity of study designs and cognitive endpoint definitions. The majority of studies utilized diagnostic codes and did not have objective clinical assessments that included a neuropsychological battery. In the current study, the cognitive outcome was well-defined. MCI was diagnosed after an extensive clinical evaluation, including neuropsychological testing performed as part of the MCSA. Using this protocol, we did not observe an association between ADT use among participants with a history of prostate cancer and odds of MCI. Transient changes may be possible during ADT medication use, which are subsequently improved after return of circulating testosterone. Thus, it is possible that many of the participants’ cognition improved after stopping ADT. It is also possible that the effect of ADT use on cognition is more subtle, and therefore we did not observe a difference when comparing MCI and cognitively unimpaired participants.

Strengths of the study include the large sample size, the population-based sampling frame, and the confirmation of prostate cancer and ADT use through medical record abstraction as compared to self-report. However, limitations also warrant consideration. First, is the concern of survival bias such that those with a history of prostate cancer might disproportionately not participate or may have died before having a chance to be included in the MCSA. Because prostate cancer often has a more indolent course than other cancers, the effect of survival bias may be less, but it is still present. Second, the MCSA originally focused on the prevalence and incidence of MCI and therefore excluded those with a diagnosis of dementia at baseline. This eliminates the inclusion of potential participants who had a history of prostate cancer and already carried a diagnosis of dementia at the time of invitation to the study, and would underestimate the true association. Future prospective studies would better alleviate survival bias and by performing a prospective study beginning at the time of initial ADT exposure with regular follow-up to assess cognition. Lastly, the proposed study may not be generalizable to other, more racially and ethnically diverse, populations. However, participants enrolled in the MCSA are representative of residents of Olmsted County and of the upper Midwest.

In the current study, we did not find an association between a history of a diagnosis of prostate cancer, or use and length of ADT, and odds of MCI. However, given the limitations of this study and the conflicting evidence in the existing literature, it is important for physicians to discuss the risk and benefits of ADT with prostate cancer patients when proposing use of the effective therapy.

TABLE 4.

Cross-sectional Multivariate Associations Between Prostate Cancer, ADT use and Odds of Mild of Cognitive Impairment (N=2,385)

Variable	Odds Ratio	Low 95% CI	High 95% CI
Age (y) at first visit	1.06	1.04	1.08
Education (y)	0.88	0.84	0.91
Charlson index	1.09	1.05	1.14
BAI	1.08	1.05	1.11
BDI (>13)	1.53	1.02	2.29
Any APOE4	1.28	0.98	1.67
Diabetes mellitus	0.85	0.62	1.16
Hypertension	1.19	0.87	1.63
Prostate cancer	0.73	0.51	1.04
ADT exposure	0.95	0.20	4.41

ADT indicates androgen deprivation therapy; BAI, Beck Anxiety Index; BDI, Beck Depression Inventory.