Psychological predictors of delayed active treatment following active surveillance for low‐risk prostate cancer: The Patient REported outcomes for Prostate cARE prospective cohort study

Abstract Objectives In a prospective, comparative effectiveness study, we assessed clinical and psychological factors associated with switching from active surveillance (AS) to active treatment (AT) among low‐risk prostate cancer (PCa) patients. Methods Using ultra‐rapid case identification, we conducted pretreatment telephone interviews (N = 1139) with low‐risk patients (PSA ≤ 10, Gleason≤6) and follow‐up interviews 6–10 months post‐diagnosis (N = 1057). Among men remaining on AS for at least 12 months (N = 601), we compared those who continued on AS (N = 515) versus men who underwent delayed AT (N = 86) between 13 and 24 months, using Cox proportional hazards models. Results Delayed AT was predicted by time dependent PSA levels (≥10 vs. <10; HR = 5.6, 95% CI 2.4–13.1) and Gleason scores (≥7 vs. ≤6; adjusted HR = 20.2, 95% CI 12.2–33.4). Further, delayed AT was more likely among men whose urologist initially recommended AT (HR = 2.13, 95% CI 1.07–4.22), for whom tumour removal was very important (HR = 2.18, 95% CI 1.35–3.52), and who reported greater worry about not detecting disease progression early (HR = 1.67, 1.05–2.65). In exploratory analyses, 31% (27/86) switched to AT without evidence of progression, while 4.7% (24/515) remained on AS with evidence of progression. Conclusions After adjusting for clinical evidence of disease progression over the first year post‐diagnosis, we found that urologists' initial treatment recommendation and patients' early treatment preferences and concerns about AS each independently predicted undergoing delayed AT during the second year post‐diagnosis. These findings, along with almost one‐half undergoing delayed AT without evidence of progression, suggest the need for greater decision support to remain on AS when it is clinically indicated.


| INTRODUCTION
The use of active surveillance (AS) to manage low risk prostate cancer (PCa) has increased over the past decade, resulting in fewer men receiving surgery or radiation immediately following the diagnosis. [1][2][3] This change is the result of several factors, including data indicating that active treatments are associated with substantial treatment side effects that can impair quality of life (QOL) [4][5][6][7] and the observational [8][9][10][11][12][13] and randomized studies 14,15 that have not found a mortality benefit of active treatment (AT) over observation for men with localized PCa. 7,16 However, decisions about selecting and then continuing on AS remain challenging for men with low-risk PCa, as they must weigh the harms of potentially unnecessary treatment against their anxiety about not actively treating the cancer. [17][18][19] Formal AS protocols include monitoring the cancer via periodic PSA tests, digital rectal exams, prostate biopsies, and MRI. 4,5,13,[20][21][22][23] Monitoring provides the option to undergo curative treatment and is based on evidence of disease progression and on patient and physician preferences. Several studies have shown that disease progression results in switching from AS to surgery or radiation. 13,20,23,24 Among the few studies that have conducted a longitudinal assessment of the role of patient preferences and anxiety in switching to AT, there is evidence of discontinuing AS due to anxiety or personal preference and without evidence of disease progression. [23][24][25] However, not all studies have found strong evidence showing that PCa-related anxiety results in being more likely to opt out of AS. 26,27 Similarly, two recent reviews reached differing conclusions regarding the role of anxiety among men who discontinue AS. 18,19 In a meta-analysis, Simpkin and colleagues 18 concluded that an average of 20% of patients discontinue AS due to anxiety or choice in the absence of disease progression.
However, Kinsella and colleagues 19 concluded that fear of progression has not been definitively shown to contribute to discontinuation of AS in the absence of progression.
Based on these differing conclusions, additional investigation of the role that psychological factors may play in delayed treatment decisions is needed. We have addressed several of the limitations that have been present in earlier studies on switching from active surveillance to active treatment. This study included the following strengths: clinical progression measures were included in multivariable models, only low-risk cases were included (intermediate cases were excluded), a large sample size, and a prospective assessment of psychological variables and reasons for discontinuing AS.
We conducted the Patient REported outcomes for Prostate cARE (PREPARE) study, a prospective, comparative effectiveness study conducted within an integrated health system. The primary objective was to assess decision-making factors and patient-reported outcomes among men with low-risk PCa. [28][29][30][31] Here we present the demographic, clinical, and psychological predictors of undergoing AT after having been on AS for at least 12 months. We hypothesized that, after accounting for baseline disease characteristics and subsequent disease progression, increased PCa-related anxiety, physician treatment recommendations, decisional uncertainty, and personal preferences regarding disease-related dysfunction would predict switching from AS to AT by 24 months. We also conducted exploratory analyses comparing men who switched to AT without a clinical progression versus those who remained on AS in the presence of a clinical progression.

| Participants
We enrolled subjects from Kaiser Permanente Northern California Exclusion criteria were (1) already started PCa treatment; (2) diagnosis via transurethral resection of the prostate, with no subsequent biopsy; (3) KPNC membership ending without evidence of treatment (excluded to avoid potential misclassification of patients who were no longer KPNC patients during the study period); and (4) physician refusal (see below). Details of the exclusions and refusals have been presented previously. 29

| Procedures
We used an ultra-rapid identification process that electronically identified putative cases by twice weekly reviewing pathology data for evidence of prostate biopsies and surgeries ( Figure 1). All cases were subsequently linked with the KPNC Cancer Registry to remove prevalent cases and then reviewed to ensure that they met study eligibility criteria. After confirming that patients had been informed of the diagnosis by the treating urologist, we mailed an invitation letter with a return postcard to provide the opportunity to decline participation.
We sought to conduct the baseline telephone assessment within 30 days of the patient's notification of his diagnosis, and all were completed prior to treatment. The baseline assessment required 30-40 min and men received a $20 gift card.
The follow-up assessment was completed 6-10 months postdiagnosis (M = 6.9, SD = 0.8; 6-7 months (68%), 8 months (24%), and 9-10 months (8%)). These assessments were completed by telephone interview (51%) or by patients on a web-based platform (49%) and they required 20-30 minutes to complete. Participants received a $10 gift card. Participants also completed a 24-month follow-up assessment (not presented here). IRB approval was received from the Kaiser Foundation Research Institute. All patients provided informed consent for study participation.   During the study period, a uniform surveillance protocol for AS was not yet in place across all 21 KPNC medical centers, and thus surveillance procedures were determined by individual clinicians. The diagnostic Gleason score was from the biopsy immediately preceding the PCa diagnosis. The surveillance biopsy that occurred closest to 24 months post-diagnosis was considered the final Gleason score for the Continued AS group, and the biopsy immediately preceding treatment was the final Gleason score for the Delayed AT group. A Gleason score increase of ≥1 was classified as disease progression.
Diagnostic and surveillance biopsies included a minimum of 12 cores.
The baseline PSA (ng/ml) was measured immediately preceding the diagnosis. PSA doubling time (<36 months vs. ≥36 months) was calculated using a minimum of the last two PSAs prior to 24 months post-diagnosis for the Continued AS group, and prior to treatment for the Delayed AT group.

| Prostate-specific anxiety
At each assessment, participants completed five items from the Cancer Control Subscale of the Health Worry Scale (alpha = 0.77). 35,36 The response scale for each item was 0-4 ('not at all' to 'very much').
A higher score indicates greater prostate-related anxiety (range = 0-20). We assessed whether the total score and the individual items were associated with undergoing delayed AT.

| Health concerns associated with the treatment decision
At baseline, men indicated the importance ('very', 'somewhat', or 'not at all') of several health concerns influencing their treatment decision, including cancer control (N = 7; e.g., wanting the cancer removed), treatment-related quality of life (N = 5; e.g., avoiding problems with sexual function) and treatment burden (N = 3; e.g., out-of-pocket costs).

| Decision making variables
At baseline, we measured decisional certainty with the SURE Test, 37 a four-item version of the Decisional Conflict Scale (alpha = 0.71).
Response choices were 'true', 'false' or 'do not know', with 'do not know' scored as incorrect. Correct items were summed for the total score (higher indicates greater knowledge).

| Descriptive analyses
We compared the two treatment groups (Continued AS vs. Delayed AT) on demographic and clinical characteristics using chi square tests for categorical variables, and t-tests for continuous variables (Table 1). Table 2 includes the surveillance procedures and results, Table 3 includes descriptive statistics for the psychological variables, and Table S1 includes the health concerns data. We had very little missing data at each assessment (<1% with the exception of income) and high retention at 6 months ( Figure 1). Cox proportional hazard models included men who completed both assessments and for whom we had complete EHR data at 24 months post-diagnosis. Finally, we explored the characteristics of men who switched to AT without a clinical progression and those who remained on AS in the presence of a clinical progression ( Figure 2).

| Outcome models
To assess the predictors of treatment group (Continued AS vs. Delayed AT) at 24 months post-diagnosis, we used two Cox proportional hazard models to estimate adjusted hazard ratios and construct 95% confidence intervals (Table 4). Model 1 included the two time dependent covariates of PSA and Gleason scores, in addition to age, race, the Elixhauser Comorbidity assessment, and men's baseline self-report of their urologist's treatment recommendation. Model 2 added the two decisional and psychological variables found to have a significant bivariate association with treatment group: wanting the cancer removed, and worry that changes in one's medical condition would not be detected early.

| Power calculations
With the sample size of 515 (Continued AS) and 86 (Delayed AT) and using categorical measures of the psychological and decisional T A B L E 1 Baseline demographic and clinical characteristics  (Table 2).
Regarding the psychological variables, there were no significant group differences at baseline or follow-up on the total scores of the PROMIS anxiety and depression scales or on the prostate-specific anxiety scale (Table 3). We also evaluated each of the prostatespecific anxiety items, observing that at follow-up those with greater worry about changes in their medical condition not being detected early were significantly more likely to switch to AT, compared to those with less worry (p = 0.008).
Decision-making variables, including baseline health concerns (cancer control, treatment-related quality of life, and treatment burdens), indicated that men who reported greater importance of 'wanting the cancer removed from my body' were more likely to undergo delayed AT (p < 0.05; Table 3). The other health concern items did not predict delayed AT (Table S1). Decisional certainty, PCa knowledge, and baseline treatment preference were not significantly associated with treatment group (Table S2). Finally, prostate-related QOL measured at baseline or follow-up 38 was not associated with switching to AT (data not shown).

| Cox Proportional Hazards Models
The Cox models assessed the likelihood of undergoing delayed AT after 12 months of AS (

| Exploratory analyses (Figure 2)
We explored whether the decision to remain on AS vs. undergo delayed AT was concordant with surveillance biopsy results ( Figure 2

| DISCUSSION
In this prospective cohort study of men undergoing AS for low-risk PCa, 14% underwent delayed AT by 24 months post-diagnosis, which is similar to other AS cohorts. 13 that disease progression of low-risk PCa predicts switching from AS to AT, 13,20,23,24 this is one of few longitudinal studies investigating the role that decisional and psychological factors may play in this decision while accounting for evidence of disease progression. 23,[25][26][27] In multivariable analyses, adjusting for disease progression and urologists' initial treatment recommendation, men's baseline desire to have their cancer removed and their subsequent worry that disease progression would not be detected early each independently predicted undergoing delayed AT during the second year post-diagnosis. Importantly, there were no differences on general anxiety or depression or overall prostate-related anxiety between those who continued AS vs. underwent delayed AT, indicating that delaying AT was not associated with greater distress while undergoing AS.
These findings support prior longitudinal studies that have found that fear of disease progression was associated with undergoing delayed AT when the disease had not progressed. 23,25 In a meta-analysis that included 26 AS cohorts, Simpkin 18 concluded that 20% of men discontinue AS due to anxiety. However, not all studies have found that switching to AT was associated with fear of disease progression. 26,27 The conflicting findings may be associated with analytic differences, including the adjustment for clinical progression in multivariable models versus limiting the analysis to men whose disease had not progressed. Our findings indicate the clinical importance of understanding men's specific prostate-related anxieties associated with the initial choice to undergo AS, given the potential for its subsequent impact on the decision to discontinue AS in the absence of disease progression. More research is needed to understand the role of psychological factors in discontinuing AS among men with low-risk PCa, as most of the work on treatment decisions for low-risk PCa has addressed the initial treatment decision. As an example, the desire to remove the cancer has been associated with selecting AT as the initial treatment, 39 but we are unaware of studies that have included this variable when assessing delayed AT.
Among men in the Delayed AT group, despite the greater likelihood of having received an initial recommendation for AT and of having an initial preference for the cancer to be removed, they nonetheless remained on AS for a minimum of 12 months. Importantly, the Delayed AT group did not report greater general anxiety or depression compared to the Continued AS group at either the baseline or six-month follow-up assessment. These results provide important data for clinicians when discussing the treatment decision with men who are considering AS. Among men who ultimately switch to AT, the likelihood of experiencing increased anxiety or depression during the AS period is low. Providing education about the fact that switching to AT is an option, with or without disease progression, may help men feel comfortable when considering AS as a management option.
These findings confirm the importance of the urologist's recommendation on the treatment decision. 2,40,41 What is notable is that the recommendation, as reported by the patient, continued to have a significant impact on the treatment decision 1-2 years later, after adjusting for disease progression and PCa-related concerns. Although we did not measure men's perception of their urologists' subsequent recommendations, this finding provides new information on the long-term treatment implications of the urologist's initial recommendation, which may not include the patient's treatment preferences. 40 In exploratory analyses assessing whether treatment decisions were concordant with surveillance results, we found that 31% of men who underwent delayed AT did so without clinical evidence of pro-  For some men, in lieu of delayed AT, a more aggressive AS regimen or newer tools (e.g., MRI and/or genetic testing among those with a significant family history of prostate, breast, or ovarian cancer), may assist physicians with risk stratification. Finally, a method is needed to assist clinicians in identifying men who may benefit from additional resources to remain on AS when it is clinically indicated.