Published online Oct 23, 2023.
https://doi.org/10.4111/icu.20230103
A review of the objective cognitive function measurements in males receiving hormonal therapy for prostate cancer
Abstract
Purpose
Prostate cancer (PC) is more common in the older population and the use of hormonal therapy in PC can increase medical frailty and cognitive decline. This narrative review examines the impact of androgen deprivation therapies (ADTs) and next-generational hormonal therapies (NGHT) on cognitive function outcomes amongst patients with hormone-sensitive or castrate-resistant PC.
Materials and Methods
Six electronic databases were searched from January 2000 to June 2022 for quantitative studies to evaluate the impacts of hormonal therapies (ADT, combined androgen blockade, and NGHT) on cognitive functions in men with PC.
Results
Of the 36 studies identified, 20 studies reported no effect of hormonal therapies on any cognitive domain while 16 studies found possible declines in at least one domain. The domains assessed were highly variable and objective assessment measurements were not standardized or widely adopted. While the results have been inconsistent, a relationship between declining androgen levels and poorer performances in the visuospatial and visual memory domains has been highlighted. It was not possible to distinguish the degree of cognitive parameter changes between the populations of hormone-sensitive and castrate-resistant PC.
Conclusions
While the exact impact of ADT and NGHT on cognitive function in men with PC remains controversial, appropriate care should be undertaken especially in older and frail individuals, specifically in those with progressive or established visuospatial or visual memory deficits.
INTRODUCTION
Prostate cancer (PC) is one of the commonest cancers in the world and hormonal therapy is commonly prescribed in patients with advanced PC [1]. Recent data suggests the proportion of frail men treated with traditional androgen deprivation therapy (ADT) and next-generational hormonal therapies (NGHT) is growing [2]. Although there are survival advantages with the use of ADT and NGHT in hormone-sensitive PC (HSPC) and metastatic castrate-resistant PC (mCRPC) [3, 4], these drugs are not without adverse effects [2, 5]. Health domains affected include cardiometabolic, psycho-social, sexual, and cognitive to name a few, which can be significant, especially in the elderly [6, 7]. The use of ADT may accelerate physical changes experienced by older adults and increase medical frailty especially cognitive decline which invariably exacerbates physical weakness, disability, and falls [6, 7, 8, 9]. PC patients receiving ADT have a 41% higher risk of depression and a 47% higher risk of dementia especially those who are older age, have greater comorbidity, and have a previous history of depression [10]. The use of novel NGHT also negatively impacts cognitive function [11, 12] and studies found these agents are associated with increased seizures, falls, and dementia risk [13, 14, 15, 16].
Androgens significantly modulate specific aspects of cognition, and androgen depletion, either through normal ageing or pharmacological action can result in specific cognitive impairments and is associated with a higher incidence of neurodegenerative diseases and a worse prognosis after brain injury [17, 18]. Studies have linked cognitive impairment with declines in testosterone leading to decreases in overall cognitive ability [11, 12, 13, 14, 17]. Hence, the adverse impact of the use of hormonal therapy in PC patients across various visual and cognitive domains such as visuospatial ability, visuomotor ability and visual memory can significantly affect the quality of life (QOL) for PC men [11, 12, 13, 14, 15, 16, 17]. Nonetheless, earlier literature has been inconclusive due to several confounders including small sample sizes, diverse patient populations, subjective measures, different time points of data collection, and heterogeneous hormonal therapy interventions, while the exact impacts of ADT and NGHT on cognition remain to be established [19, 20, 21].
Given the direct causal relationship between PC-related hormonal therapy and cognitive decline in PC patients are difficult to be established [22], this narrative review aims to explore the current evidence surrounding the objective cognitive impacts of hormonal therapies in patients with HSPC or CRPC. This paper compares the degree of cognitive changes between those not on androgen therapy in any form at baseline and those with HSPC or CRPC on hormonal therapies.
MATERIALS AND METHODS
PubMed, Embase, Ovid MEDLINE, SCOPUS, CINAHL, and Cochrane Review databases were searched from January 2000 to June 2022. The search strategy incorporated a combination of keywords, Boolean operators, and medical subject heading (MeSH) terms with the following keywords “prostate cancer”, “androgen deprivation therapy”, “androgen targeted therapy”, “hormonal therapy”, and “cognition” utilized. Additional derivatives for these keywords were also used to improve the search too.
Studies were included if these were original studies published in the English language with a quantitative design and evaluates cognition outcomes in PC males on ADT and NGHT. While the detailed inclusion and exclusion criteria for these studies are quite heterogenous, the main similarities are these patients diagnosed with metastatic disease and are either HSPC or CRPC. Importantly, none of these patients included in the studies have metastatic PC deposits in the brain that can adversely impact cognitive function on de novo.
Data gathered from selected studies included study design, population, time points of analysis, and cognition assessment tools. Following de-duplication using endnotes, titles and abstracts were screened by two authors (TN and BN). Eligibility and screening of the remaining articles were conducted by two authors (TN and BN) with disagreements resolved by the senior authors (HR and EC). Quality assessment was performed using the risk of bias tool for randomized controlled trials and non-randomized interventions. A narrative synthesis of information was produced to present outcomes instead of the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) method due to the significant variations in study designs and methods as well as the heterogeneous populations (Table 1).
Table 1
Search strategy summary
RESULTS
1. The impact of ADT and/or NGHT on visuospatial ability, visuomotor ability and visual memory amongst men with HSPC (Tables 2, 3)
Table 2
The impact of ADT on visuospatial ability, visuomotor ability, and visual memory amongst men with HSPC
Table 3
The impact of ADT on other domains of cognition amongst men with HSPC
Alibhai et al. [23] in a prospective cohort study examined the impact of ADT on cognitive function amongst men with non-metastatic HSPC. Using a battery of 14 neuropsychological tests covering 8 domains of cognition they compared changes in cognitive domain scores between baseline and 12 months amongst patients on continuous ADT (n=66), patients with PC not receiving ADT (n=73) and healthy controls (n=72). Patients on ADT experienced significantly smaller scores in visuospatial ability at 12 months than either control group (mean scores: 5.7, vs. 8.5 and 12.1 respectively, p=0.034), as well as in the executive functions as assessed by spatial span backwards (mean scores: -0.08, 0.19, and 0.32 respectively, p=0.031).
Another prospective study conducted by Jenkins et al. [24] assessing the impacts of neoadjuvant ADT prior to radiotherapy for early PC noted a decline in visuospatial ability amongst 32 patients commenced on goserelin. A cognitive test battery covering seven domains was utilized and at 9 months the number of patients who significantly declined on visuospatial cognitive tasks was 13 of 32 (40.6%) compared with the controls 5 of 18 (27.8%).
Cherrier et al. [25] assessed the influence of combined androgen blockade (CAB) on cognitive function during the first cycle of intermittent androgen suppression for men with HSPC. They found those on leuprolide plus flutamide (n=19) displayed a smaller increase in scores on the Route Test when compared with controls (n=15). Their results revealed a significant group-by-time interaction for the route test where the control group improved compared to the baseline but the intervention group did not (F[1,18]=7.04, p<0.05). Further, using the Mental Rotation test 69% of participants demonstrated a clinically significant decline. However, this decline did not continue upon withdrawal of treatment where only 15% of men within the intervention group continued to display a significant decrease in scores compared to baseline.
Green et al. [26], using the Trails B Test found the leuprolide group (n=50) was significantly slower at 6 months compared to baseline when examining group-by-time interaction. This was compared to controls (n=15) whose scores did not change significantly (p=0.012). Similarly, four additional studies have shown similar results with spatial and visual function impairment noted [27, 28, 29, 30].
2. The impact of ADT and/or NGHT on other domains of cognition amongst men with HSPC
In a comparative analysis of cognition amongst 58 patients on ADT versus controls (n=82), Gonzalez et al. [12] found that at 12 months the ADT group was more likely to demonstrate impaired performance on the Color Trails 2 test of executive function (16/58 vs. 4/82, p=0.05). Salminen et al. [31] in their prospective study examining 23 men on leuprolide for 12 months observed significant slowing in choice reaction time tasks demanding working memory at 12 months (p=0.006) as well as slower recognition speed of numbers (p=0.01). Further significant declines were noted in the vigilance task of sustained attention (p=0.005) [31].
Ceylan et al. [32] investigated 72 patients with locally advanced PC receiving CAB for HSPC continuously for 12 months and found significantly lower mean total Montreal Cognitive Assessment (MoCA) scores in the intervention group versus controls (n=78). In the subgroup analyses of the MoCA test they found language ability and short-term memory capacity were most significantly impacted by ADT treatment. This outcome was replicated in a similar prospective study by Gunlusoy et al. [33] that analyzed the effect of bicalutamide with goserelin or leuprolide on MoCA scores at 6 and 12 months compared to baseline, showing that patients in the intervention group (n=72) achieved significantly lower mean total scores than controls (n=72). Again, the declines were especially prominent at twelve months in the areas of language ability (p<0.001) and short-term memory capacity (p<0.001).
3. Hormonal therapies for men with HSPC may not always be associated with cognitive function alteration
Alibhai et al. [27] in their prospective controlled trial discussed above found no consistent evidence of overall cognitive decline at 6 or 12 months of ADT monotherapy for men with HSPC. Apart from the visuospatial changes noted above, no significant changes were identified between the groups from baseline to 6 months among ADT users (n=77) compared with controls (n=82, n=82) in any of the fourteen cognitive tests (all p-values >0.05). Further, when measuring the proportion of participants in each cohort who declined ≥1 standard deviation on ≥2 tests, the authors found no significant differences between the groups at both 6 and 12 months (p=0.88).
In a cross-sectional study of 57 patients with nonmetastatic PC treated with ADT and 51 age-matched controls, Joly et al. [34] found no between-group difference in cognitive function. Using a battery of seven cognitive tests Joly et al. [34] reported similar findings. Assessing the impact of ADT on patients commenced on neoadjuvant hormone therapy for early PC prior to radiotherapy (n=32) versus healthy controls (n=18), results showed no difference between groups on any cognitive task at 9 months.
Holtfrerich et al. [35] in a cross-sectional study examined 15 men receiving GnRH agonists (GA) and nine men on nonsteroidal anti-androgen (AA) comparing these groups to 22 PC controls and 22 healthy controls. Significantly, lower testosterone concentrations were associated with worsening visuospatial performance as measured by the Block Design test (F[1,62]=4.29, p=0.042, ηp2=0.07). ADT patients exhibited a positive correlation between testosterone levels and worsening visuospatial performance whilst the control groups were unaffected (r=0.474; p=0.019). When considering the two hormonal therapy groups separately, patients receiving GA had significantly lower testosterone levels (mean±standard deviation, 46.25±20.74) than those receiving AA (mean±standard deviation, 88.58±53.73) suggesting that the global suppression of testosterone was strongest in patients who received GA.
Several studies concluded no influence of ADT on cognitive function [36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53], and these are listed in the table below (Table 3) [27, 32, 34, 36, 37, 39, 40, 42, 43, 44, 45, 50, 52, 53, 54, 56].
4. Impact of NGHT on CRPC
Matousek and Sherwin [54] in their prospective, randomized, controlled trial found cognitive scores remained stable when compared to baseline amongst a cohort of patients (n=25) on 12 weeks of CAB for CRPC with no significant declines detected. Additionally, with a population of patients receiving leuprolide and flutamide (n=37), the naturalistic study conducted by Almeida et al. [55] found that upon the removal of ADT mean Word List Total Recall scores (30.5 at week 36 end of intervention and 35.1 at week 54, p=0.001), as well as mean Verbal Paired Associates Total Recall scores (17.2 at week 36 and 18.8 at week 54, p=0.002), improved significantly, suggesting ADT and not NGHTs themselves induce possible cognitive impairments [55].
Investigating abiraterone amongst a cohort of 254 patients with mCRPC, Gotto et al. [56] concluded patients experienced no significant changes in mean MoCA scores over 72 weeks of therapy. On the contrary, men receiving cyproterone acetate (CYP) (n=62) had significantly worse scores on the digit symbol test than the control group (n=15) (p=0.064) [57]. Further, there were significant improvements in the Stroop Color Words test (executive functions) amongst all groups (p=0.087) except the CYP cohort.
DISCUSSION
Numerous studies have shown a negative impact of hormonal therapies on various visual domains of cognition such as visuospatial ability, visuomotor ability and visual memory amongst PC men. Visuospatial cognition requires the perception, selection, organization, and utilization of location and object-based information [43], visuospatial impairment is often one of the earliest symptoms of cognitive decline [58, 59, 60, 61]. Similarly, a decline in visuomotor ability is experienced by those with cognitive decline and its negative impact on overall physical functioning as shown in earlier studies is consistent with our study findings. McGinty et al. [11] in their systematic review found ADT significantly reduced visuomotor ability (effect size -0.67, 95% confidence interval-1.17 to -0.17; p=0.008) although the authors noted most studies assessed patients within 9 months of commencing ADT, emphasising that the onset of cognitive decline and the impacts of ADT beyond 9 months could not be properly determined. In our study, most of the included prospective studies included analysis at 3, 6, and 12 months, with one study reporting a significant impact of ADT on visuospatial cognition although within 6 months duration [30]. Cherrier et al. [25] found significant impacts upon 9 months of analysis while Gonzalez et al. [12] showed found deteriorations in cognition at 6 and 12 months of follow-up analysis. These findings suggest a minimum of 6 months of ADT can produce a significant decline in cognitive function.
There was inconsistency amongst domains of cognition being impacted. These include executive functions, verbal memory, language ability, short-term memory, working memory, sustained attention, recognition speeds, and global cognitive decline as measured by the MoCA. Furthermore, these findings can be isolated and unlike the visual domains, are not consistently identified as being significantly affected. Two recent reviews have noted similar findings and instead of drawing definitive conclusions suggest clinicians follow the International Society of Geriatric Oncology (SIOG) 2019 guidelines encouraging discussion regarding the risk of cognitive dysfunction with older PC patients considering the use of ADT prior to initiation [8, 62]. Similarly, contemporary reviews such as those by Andela et al. [9] have also concluded the significant variability in the actual impacts of hormonal therapies on cognitive function amongst men with PC.
The degree of cognitive decline amongst patients with NGHT is yet to be clearly elucidated at this time and few studies have identified the possible influence of NGHT on cognitive performance. Shore et al. [43] in the real-world study of enzalutamide and abiraterone acetate with prednisone tolerability study investigated the tolerability of enzalutamide and abiraterone acetate in patients with mCRPC and found an increase in patient-reported adverse events such as fatigue, asthenia, and vertigo with enzalutamide compared to abiraterone while four patients on enzalutamide experienced clinically significant cognitive decline. However, all patients had received prior ADT and showed no meaningful decline in cognition at follow-up assessment.
Most studies investigating NGHT report fatigue (as part of QOL assessments) and patient-reported adverse effects. Ryan et al. [63] in their review exploring the cognitive and central nervous system changes found fatigue to be the most common reported adverse effect amongst included studies examining men on enzalutamide, apalutamide and abiraterone acetate. Mental impairment was also commonly reported as a side effect of enzalutamide. While this study provides valuable insights, most studies focused on QOL questionnaires and patient-reported outcomes (PRO) endpoints. Since these are not objective assessments of cognition, the reliability of such responses may be limited by subjectivity.
Amongst the papers analyzed, the exact impact of NGHT on cognition is often difficult to be differentiated from those seen with ADT. Testosterone levels are directly reduced by ADT whilst most forms of NGHT block androgen receptor functions such as transcription and translation without an actual decline in total testosterone level. Hence, a possible hypothesis for the minimal effects of NGHT on cognition could be related to the negligible reduction in the androgen levels. The findings of Holtfrerich et al. [35] and Shore et al. [43] suggested the addition of enzalutamide to ADT amongst men with CRPC has a minimal effect on further cognitive performance decline in a population already impacted by the androgen-lowering effects of monotherapy ADT (approximately 20% with mild cognitive decline as baseline). In another study, Almeida et al. [55] showed that ADT serves as the catalyst for cognitive impairment among men on hormonal therapies with low testosterone levels. Nonetheless, Ryan et al. [63] suggest this hypothesis may not apply to all NGHTs since both enzalutamide and apalutamide readily cross the blood-brain barrier and patients can be more vulnerable to potential adverse impacts on the central nervous system.
1. Clinical implications
A major goal for the prevention of cognitive decline follows the adage “use it or lose it.” An increasing number of studies have shown that maintaining cognitive engagement has a beneficial impact on the prevention or delaying of the onset of age-related cognitive decline, especially in the setting of hormonal therapies for PC men [8, 9, 10, 11, 12, 64, 65]. The current medical landscape in PC treatment is dynamic and evolving based on newly accrued evidence, and new therapeutic agents are trialled without any long-term data on cognitive decline [66, 67, 68, 69]. While these medical therapies could potentially increase PC survival rate, their adverse effects especially on cognition and physical functioning would come at a great cost to patients. Similarly, psychological issues such as pre-existing memory impairment, dementia, depression, and/or suicidal thoughts can play a significant role across the cognitive domains [10, 32, 39, 63, 68] and these variables should be investigated further to determine if these variables are associated with worse prognosis in the setting of ADT and/or NGHT use. Future clinical studies must clearly define their populations and ideally, incorporate a validated and comprehensive standardized cognitive battery to assess individual cognitive domains.
2. Study limitations
We acknowledged several limitations in our study. Most studies failed to specifically identify cohorts as HSPC or CRPC. While the ADT use was clinically indicated and likely involved men with non-metastatic HSPC, the ambiguity of this assumption potentially biased the quality of our study. A total of 6 studies had a follow-up period of only six months or less, and most of the studies included small study populations of less than 100 individuals. The lack of a control group for comparison in 12 studies also affects the analysis regarding the influence of an intervention program. While most studies accounted for pre-intervention cognitive impairment with pre-test screening either by mini-mental state examination (MMSE) or MoCA, baseline education was rarely accounted for, and higher levels of baseline intelligence could serve as a confounding protective factor against cognitive decline [51]. Moreover, the overall impact of additional therapies beyond hormonal treatments such as chemotherapy is beyond the scope of this article. The MMSE and MoCA play an important role in cognitive screening although these measures are designed to assess global cognition and may lack the sensitivity to detect minor changes in individual cognitive domains. Furthermore, the lack of cognitive testing standardization may limit the significance of our findings. While most studies covered a breadth of cognitive domains, most domains were only assessed with single tests. A limited number of studies utilized multiple tests to cover a single domain and the sensitivity of assessment may have been negatively influenced too. Moreover, these observation studies cannot provide an exact causal relationship on whether ADT and/or NGHT use directly cause cognitive decline through adverse impact on brain neuroactivity and vascular parameters.
The possible reasons for inconsistent results among the studies on the association between ADT and cognitive impairment include the lack of definition and standardization in the terminology of cognitive impairment, use of validated questionnaires, variations in research designs and methods, the stage of PC, pharmaceutical-sponsored vs. real-world evidence clinical data, and that impairment in cognitive domains were not the designated primary endpoint of the clinical trial. Future research into this important field will require adequate addressing of the above-mentioned limitations and incorporating appropriate interventions to optimize the various cognitive domains to safeguard the needs and functions of PC patients as well as minimise clinical frailty related to ADT and NGHT use.
CONCLUSIONS
Hormonal therapies such as ADT and NGHT may result in the deterioration of cognitive domains especially in visuospatial performance, visuomotor ability, and visual memory. There is a need to develop and incorporate a standardized cognitive assessment to assess the true impact, especially the long-term consequences of these hormonal therapies in frail and older PC men.
CONFLICTS OF INTEREST:The authors have nothing to disclose.
FUNDING:None.
AUTHORS’ CONTRIBUTIONS:
Research conception and design: all authors.
Data acquisition: all authors.
Data analysis and interpretation: all authors.
Drafting of the manuscript: all authors.
Critical revision of the manuscript: all authors.
Approval of the final manuscript: all authors.
References
-
Kumar S, Shelley M, Harrison C, Coles B, Wilt TJ, Mason MD. Neo-adjuvant and adjuvant hormone therapy for localised and locally advanced prostate cancer. Cochrane Database Syst Rev 2006;2006:CD006019
-
-
Pilon D, Behl AS, Ellis LA, Robitaille MN, Lefebvre P, Dawson NA. Assessment of real-world central nervous system events in patients with advanced prostate cancer using abiraterone acetate, bicalutamide, enzalutamide, or chemotherapy. Am Health Drug Benefits 2017;10:143–153.
-
-
Parimi S, Eigl B, Sunderland K, Zulfiqar M, Finch D, Oja C, et al. Effects of abiraterone (ABI) and enzalutamide (ENZA) on cognitive impairment and depressive symptoms in patients (pts) with metastatic castration-resistant prostate cancer (mCRPC). J Clin Oncol 2016;34 15_suppl:5059
-
-
Białek M, Zaremba P, Borowicz KK, Czuczwar SJ. Neuroprotective role of testosterone in the nervous system. Pol J Pharmacol 2004;56:509–518.
-
-
Shah SIA, Minhas U, Nasir N. Impact of the duration of androgen deprivation therapy for prostate cancer on cognition: a study using mini-mental state examination. PJMHS 2018;12:796–798.
-
-
Okamoto K, Sekine Y, Nomura M, Koike H, Matsui H, Shibata Y, et al. Effects of a luteinizing hormone-releasing hormone agonist on cognitive, sexual, and hormonal functions in patients with prostate cancer: relationship with testicular and adrenal androgen levels. Basic Clin Androl 2015;25:3
-
-
Cinar O, Turunc T, Kazaz IO, Yildirim O, Deliktas H, Cihan A, et al. Effects of androgen deprivation therapy on cognitive functions in patients with metastatic prostate cancer: a multicentric, prospective study of the Society of Urological Surgery Andrology group. Int J Clin Pract 2021;75:e14095
-
-
Morote J, Tabernero ÁJ, Álvarez Ossorio JL, Ciria JP, Domínguez-Escrig JL, Vázquez F, et al. ANAMEM Investigator Group. Cognitive function in patients with prostate cancer receiving luteinizing hormone-releasing hormone analogues: a prospective, observational, multicenter study. Int J Radiat Oncol Biol Phys 2017;98:590–594.
-
-
Gotto G, Drachenberg DE, Chin J, Casey R, Fradet V, Sabbagh R, et al. Real-world evidence in patient-reported outcomes (PROs) of metastatic castrate-resistant prostate cancer (mCRPC) patients treated with abiraterone acetate + prednisone (AA+P) across Canada: final results of COSMiC. Can Urol Assoc J 2020;14:E616–E620.
-
-
Klímová B, Vališ M. Nutritional interventions as beneficial strategies to delay cognitive decline in healthy older individuals. Nutrients 2018;10:905
-