Abstract
Background
This study assesses the incidence and timing of undetectable prostate‐specific antigen (PSA) after radiotherapy (RT) ± androgen deprivation therapy (ADT) and its association with prostate cancer mortality.
Methods
This is a population‐based study including 5299 men undergoing RT (2006–2020) in the Stockholm County, Sweden with all their PSA tests until death or emigration. The authors calculated incidence and timing of undetectable PSA (PSA ≤0.1 ng/mL) and used competing risk regression to evaluate the association of detectable PSA at 6 and 12 months with prostate cancer mortality (PCSM).
Results
Median follow‐up for survivors was 81 (32, 130) months. PSA nadir values were reached before 6 and 12 months in 233 (23%) and 470 (36%) patients undergoing RT and 2871 (85%) and 3612 (90%) patients undergoing RT+ADT. No significant association was found between PSA at nadir, 6 and 12 months and PCSM in the radiotherapy group and evidence of association with higher PCSM in RT+ADT group (PSA at nadir: subdistribution hazard ratio [sHR], 2.23 [2.01–2.49]; PSA at 6 months: sHR, 6.91 [5.17–9.23]; and PSA at 12 months: sHR, 37.9 [23.0–62.5]). At 12 years after RT + ADT, PCSM rates were 27%, 15%, 13%, and 5% for patients with PSA at 6 months of ≥0.5, 0.2–0.5, 0.1–0.2, and ≤0.1 ng/mL, respectively; corresponding rates by PSA at 12 months were 34%, 18%, 12%, and 5%.
Conclusions
Most patients undergoing RT+ADT reached a PSA ≤0.1 ng/mL within 6 and 12 months. PSA>0.1 ng/mL at these time points indicated a higher risk of PCSM, emphasizing the need for timely restaging and intensified salvage treatments.
Keywords: androgen deprivation therapy, biochemical recurrence, prostate cancer mortality, prostate‐specific antigen, radiotherapy, treatment failure
Short abstract
The authors evaluated prostate‐specific antigen (PSA) levels after radiotherapy and hormonal treatment for prostate cancer, finding that men with detectable PSA post‐treatment had a significantly higher risk of prostate cancer death compared to those with undetectable levels.
INTRODUCTION
Prostate‐specific antigen (PSA) is an imperfect biomarker for the diagnosis of prostate cancer (PCa) as it is often elevated due to nonneoplastic prostatic conditions. 1 However, after radical treatment, PSA becomes a sensitive marker to detect persistent or recurrent disease with high sensitivity. A detectable PSA following curative treatment may indicate residual benign or malign prostate tissue, lymph node involvement (pN1), or preexisting metastases (M1), potentially signaling early treatment failure and warranting further investigation. 2
Although PSA levels become rapidly undetectable following surgical removal of the prostate, 3 , 4 those treated with radiotherapy (RT) may still produce small amounts of PSA, even when the treatment is considered radical. PSA levels gradually decline, eventually reaching their lowest point, referred to as the PSA nadir. According to the Phoenix criteria, PSA relapse is defined as any increase of ≥2 ng/mL above the nadir. 3 , 5 This definition may delay early identification of treatment failure, potentially missing a window for curative salvage interventions.
Although several studies have shown that both PSA nadir values and the time to nadir after RT are associated with poorer survival outcomes 6 , 7 international guidelines do not mention any definition of persistent PSA levels following primary RT. However, a recent study including data from 16 randomized trials pointed out that detectable PSA defined as PSA ≥0.1 within 6 months after RT are prognostic for long‐term outcomes in patients treated with RT ± androgen deprivation therapy (ADT) for localized PCa. 8
Introducing a definition of persistent PSA after RT may be especially important in current clinical practice for two main reasons. First, prostate‐specific membrane antigen (PSMA)‐positron emission tomography (PET) computed tomography may help differentiate between local and distant recurrence much earlier than conventional imaging, potentially identifying candidates for adjuvant or salvage therapies within a curative window. 9 , 10 Second, recent evidence suggest a survival benefit of early treatment intensification with novel anti‐androgen therapies in high‐risk patients with PSA relapse. 11
The aim of this study was to evaluate in a population‐based real‐word cohort with complete data on ADT prescriptions, PSA tests, cause of death, and the incidence and timing of undetectable PSA after RT ± ADT, and to evaluate the risk of prostate cancer mortality (PCSM) in patients with detectable PSA levels.
MATERIALS AND METHODS
The present study relies on the Stockholm PSA and Biopsy Register (STHLM0), a population‐based register that contains data on every PSA test and prostate biopsy taken in Stockholm County from 2003 to end of 2021. The register was linked to the National Prostate Cancer Register of Sweden, the Prescribed Drug Register, the inpatient and outpatient registries, and the Swedish Cause of Death Register. 12 Methods for assessment of cause of death in Swedish studies have been previously described. 13
We included all men diagnosed with clinically localized PCa (cT1–3, M0–Mx) who underwent RT with curative intent with at least 1 year of follow‐up. All PSA measurements were performed in three centralized laboratories. We excluded patients who underwent treatment before 2006 since the Prescribed Drug Register started in July 2005 to allow every patient to have at least 6 months of prescribed therapies before treatment initiation. Additionally, we excluded patients with no PSA taken over the first year and patients with a pretreatment PSA ≥50 ng/mL given the high risk of metastatic PCa at diagnosis. Follow‐up for all patients was until death, emigration, or end of the study period. Study end‐date was set to the last available update of the Swedish Death Register (December 31, 2021). The study was approved by the regional ethics board in Stockholm, Sweden (2012/438‐31/3).
Radiotherapy protocols, adjuvant ADT, and salvage therapies
All patients underwent RT in one of two hospitals in Stockholm (Karolinska University Hospital and Stockholm South General Hospital). Two different RT protocols were used: moderately hypofractionated dose escalated (29 × 2.5 Gy) external beam radiotherapy (EBRT) or EBRT combined with high dose‐rate (HDR) brachytherapy (25 × 2 Gy + 2 × 10 Gy). ADT treatment information was extracted from the prescribed drug register. Specifically, we extracted information on dispensed medications used in the treatment of prostate cancer, including gonadotropin‐releasing hormone analogues and antagonists, androgen receptor pathway inhibitors, and androgen synthesis inhibitors. ADT length was computed based on the number of packages dispensed and assuming one defined daily dose per day. 14 The time for identification of ADT prescriptions was set to 6 months before RT. ADT initiated after RT without prior ADT, as well as all treatments given after the discontinuation of first‐line ADT in patients undergoing RT+ADT, were considered salvage therapies.
Statistical analysis
Primary outcome of the study was PCSM. Statistical analyses were performed in four steps.
First, descriptive statistics was performed for the overall cohort and according to initial treatment. Probability of overall survival (OS) was estimated with the Kaplan–Meier method and probability of cancer‐specific survival was estimated with the competing risks method (competing risk: other‐cause mortality; time zero was set to the time of start of radiation) as previously described. 15
Second, we evaluated the incidence and timing of undetectable PSA values, defined as PSA ≤0.1 ng/mL. Although ultrasensitive PSA assays have become available in recent years, we chose this threshold due to the high variability of very low PSA levels and their limited clinical relevance. To conduct this analysis, we identified the PSA nadir for each patient and calculated the time from treatment initiation to nadir. As a subgroup analysis, we examined these outcomes separately in patients treated with RT alone, RT with short‐term ADT (4–6 months), and by RT modality (EBRT alone vs. EBRT combined with HDR brachytherapy).
Third, we investigated the association of time to nadir and nadir values with cumulative incidence of PCSM using Fine‐Gray subdistribution hazard models. 16 Population‐averaged cumulative incidence of PCSM was graphed according to PSA nadir values (PSA nadir ≤0.1 vs. 0.1–0.2 vs. 0.2–0.5 vs. ≥0.5 ng/mL). Two models were built, one including PSA nadir values as continuous variable and one using PSA nadir categories as previously defined. To avoid forcing continuous PSA values to have linear effects, we considered restricted cubic splines with knots at the 10th, 50th, and 90th percentile.
Selection of adjusting covariates was guided by a combination of prior literature, clinical knowledge, and statistical considerations and the models ultimately included age at diagnosis, treatment year, PSA at diagnosis, Gleason Grade Groups, clinical stage, and ADT use at time of PSA evaluation.
Fourth, we repeated step 3 analysis using the lowest PSA value at 6 and 12 months. In this case, time zero was set to the date when this PSA was taken. These time points were chosen as previously done by Kwak et al. 8 Step 3 and step 4 were performed in patients undergoing RT with and without ADT and then repeated in patients undergoing RT+short term ADT (4–6 months). Finally, to further investigate the prognostic value of early PSA levels, we performed Fine‐Gray competing risks regression (with death as the competing event) to model the risk of biochemical recurrence (BCR), defined as PSA ≥ nadir + 2 ng/mL. We then plotted population‐averaged cumulative incidence curves of BCR according to post‐treatment PSA values. The models were adjusted for the same set of covariates used in previous analyses.
Statistical analyses were performed using STATA 16. The study was reported following the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines for cohort studies.
RESULTS
Study population characteristics
Study flowchart with total number of patients excluded and reasons for patients’ exclusion is shown in Figure S1. Demographic and disease characteristics of 5299 patients ultimately included for the analysis are presented in Table S1 with a substratification based on the initial treatment received. Adjuvant ADT was used in 3994 (75%) patients. Among those who underwent RT+ADT, treatment duration was between 4 to 6 months in 1166 (29%) patients, 7 to 18 months in 1376 (34%) patients, and >18 months for 1452 (36%) patients. We stratified patients based on NCCN risk stratification, in the RT‐only group, 75% had very low to favorable intermediate‐risk disease, whereas only 25% were classified as unfavorable intermediate or higher risk. In contrast, most patients receiving RT+ADT had higher‐risk disease: 43.2% were high‐risk and 8.9% were very high‐risk. High dose rate brachytherapy was performed in 384 (29.4%) patients in the RT group and 2041 (51.1%) in the RT+ADT group.
Median follow‐up length for survivors was 71.1 (interquartile range [IQR], 32.0–107.5) months and 92.9 (IQR,43.2–137.1) months for patients who underwent RT and RT+ADT.
By the end of follow‐up, 1067 (20%) patients died. Of these, 286 patients (27%) died due to PCa. Probability of OS estimated with the Kaplan–Meier method and probability of cancer‐specific survival, estimated with the competing risks method, are presented in Figure S2.
PSA response after radiotherapy: PSA nadir, PSA at 6 months, and PSA at 12 months
PSA nadir values and time to nadir, PSA at 6 months, and PSA at 12 months are shown in Table 1 according to initial treatment modality. The histogram of time from treatment initiation to nadir PSA value is show in Figure 1.
TABLE 1.
PSA nadir values and lower PSA values before 6 and 12 months in patients undergoing RT and RT+ADT.
| RT Only (N = 1305) | RT+ADT (N = 3994) | |
|---|---|---|
| Time to nadir, months (median, IQR) | 19.4 (9.1–34.8) | 4.9 (4.2–6.3) |
| PSA nadir value, ng/mL (median, IQR) | 0.1 (0.1–0.4) | 0.1 (0.1–0.1) |
| PSA nadir, No. (%) | ||
| ≤0.1 ng/mL | 657 (50.3) | 3713 (93.0) |
| 0.1–0.2 ng/mL | 153 (11.7) | 105 (2.6) |
| 0.2–0.5 ng/mL | 231 (17.7) | 82 (2.1) |
| ≥0.5 | 264 (20.2) | 94 (2.4) |
| PSA at 6 months, No. (%) a | ||
| ≤0.1 ng/mL | 138 (13.4) | 2786 (82.6) |
| 0.1–0.2 ng/mL | 54 (5.2) | 179 (5.3) |
| 0.2–0.5 ng/mL | 73 (7.1) | 204 (6.0) |
| ≥0.5 | 765 (74.3) | 203 (6.0) |
| PSA at 12 months, No. (%) | ||
| ≤0.1 ng/mL | 222 (17.0) | 3439 (86.1) |
| 0.1–0.2 ng/mL | 83 (6.4) | 179 (4.5) |
| 0.2–0.5 ng/mL | 204 (15.6) | 213 (5.3) |
| ≥0.5 | 796 (61.0) | 163 (4.1) |
Abbreviations: ADT, androgen deprivation therapy; IQR, interquartile range; PSA, prostate‐specific antigen; RT, radiotherapy.
Computed on the total number of patients with at least one PSA before 6 months (RT, 1030; RT+ADT, 3372).
FIGURE 1.
Histogram of time from treatment initiation to nadir PSA value in patients undergoing RT (blue) and RT+ADT (green). Reference lines are placed at 6 and 12 months, time points selected for the analysis. ADT indicates androgen deprivation therapy; PSA, prostate‐specific antigen; RT, radiotherapy.
RT
In patients undergoing RT without ADT, a total of 657 (50%) patients reached a PSA nadir ≤0.1 ng/mL, 153 (12%) a PSA nadir between 0.1 and 0.2 ng/mL, 231 (18%) between 0.2 and 0.5 ng/mL, and 264 (20%) never dropped to PSA values below 0.5 ng/mL. Median time to nadir was 19 months with wide IQRs (IQR, 9–35), and 233 (23%) and 470 (36%) patients who underwent RT only reached a PSA nadir value before 6 and 12 months, respectively.
PSA at 6 months was undetectable in 138 (13.4%) patients whereas 765 (74.3%) patients had PSA ≥0.5 ng/mL. The corresponding numbers at 12 months were 222 (17%) and 796 (61%) (Table 1).
When stratified by RT treatment type, 38% of patients receiving EBRT alone reached a PSA nadir ≤0.1 ng/mL, compared to 79% of those treated with EBRT combined with HDR brachytherapy. However, the median time to nadir was longer in the HDR brachytherapy group (24.8 months; IQR, 10.8–42.4) compared to the EBRT‐only group (17.0 months; IQR, 8.2–29.9). At 6 months, PSA ≤0.1 ng/mL was observed in 11% of EBRT‐only patients and 20% of HDR brachytherapy patients. Similarly, by 12 months, PSA ≤0.1 ng/mL was achieved by 15% and 21% of patients in the EBRT‐only and HDR brachytherapy groups, respectively (Table S2).
RT with ADT
In patients undergoing RT+ADT, the rate of patients reaching lower PSA nadir values was higher (Table 1). Specifically, 3717 (93%) patients reached a PSA nadir ≤0.1 ng/mL, 105 (3%) a PSA nadir between 0.1 and 0.2 ng/mL, 82 (2%) between 0.2 and 0.5 ng/mL and 94 (2%) never dropped to PSA values below 0.5 ng/mL. Median time to nadir was 5 months with narrow interquartile ranges (IQR: 4, 6).
At the 6 months’ time‐point, 2871 (85%) patients undergoing RT+ADT reached the PSA nadir.
PSA at 6 months was ≤0.1 ng/mL in 2786 (83%), 0.1–0.2 ng/mL in 179 (5%), 0.2–0.5 ng/mL in 204 (6%), and ≥0.5 ng/mL in 203 (6%). At the 12 months’ time‐point, 3612 (90%) patients undergoing RT+ADT reached the PSA nadir. PSA at 12 months was ≤0.1 ng/mL in 3439 (86%), 0.1–0.2 ng/mL in 179 (5%), 0.2–0.5 ng/mL in 213 (5%). and ≥0.5 ng/mL in 163 (4%).
When stratified by RT type, PSA nadir values ≤0.1 ng/mL were achieved by 89% (n = 1732) of patients receiving EBRT alone and 97% (n = 1981) of those treated with EBRT combined with HDR brachytherapy (Table S2). Similarly, patients undergoing EBRT combined with HDR brachytherapy were more likely to achieve PSA ≤0.1 ng/mL at 6 months (87%, n = 1465) compared to EBRT alone (78%, n = 1321). At 12 months, PSA ≤0.1 ng/mL was observed in 89% (n = 1815) of patients in the HDR brachytherapy group compared to 83% (n = 1624) of patients in the EBRT‐alone group
Association of detectable and undetectable PSA values after RT with PCSM
RT
In the RT group, a total of 194 patients died and 21 (11%) died from PCa. In the competing risk regression, there was no evidence of association between PSA nadir values and PSA values at 6 and 12 months with PCSM (Table 2).
TABLE 2.
The sHR and the corresponding CIs for the association of PSA nadir, PSA at 6 months, and PSA at 12 months after radiotherapy and prostate cancer–specific mortality among patients who underwent RT with and without ADT.
| RT | RT+ADT | |||
|---|---|---|---|---|
| sHR a | 95% CI | sHR a | 95% CI | |
| PSA nadir value | ||||
| PSA treated as continuous variable | ||||
| PSA, per unit | 2.85 | 0.21, 38.0 | 2.23 | 2.01, 2.49 |
| Spline term | 0.43 | 0.04, 4.27 | 0.87 | 0.86, 0.89 |
| PSA treated as categorical variable | ||||
| PSA, ng/mL | ||||
| ≤0.1 ng/mL | Ref. | Ref. | ||
| 0.1–0.2 ng/mL | 6.36 | 1.92, 20.8 | 4.74 | 2.93, 7.68 |
| 0.2–0.5 ng/mL | 2.40 | 0.60, 9.65 | 8.31 | 5.74, 12.0 |
| ≥0.5 | 4.29 | 1.42, 13.0 | 15.6 | 10.3, 23.4 |
| PSA at 6 months after RT | ||||
| PSA, per unit | 0.85 | 0.37, 1.92 | 6.91 | 5.17, 9.23 |
| Spline term | 1.02 | 0.96, 1.08 | 0.42 | 0.37, 0.48 |
| PSA treated as categorical variable | ||||
| PSA, ng/mL | ||||
| ≤0.1 ng/mL | Ref. | Ref. | ||
| 0.1–0.2 ng/mL | 4.92 | 0.84, 28.7 | 2.97 | 1.92, 4.59 |
| 0.2–0.5 ng/mL | 1.02 | 0.21, 5.04 | 3.41 | 2.29, 5.09 |
| ≥0.5 | 1.70 | 0.34, 8.41 | 6.96 | 5.07, 9.57 |
| PSA at 12 months after RT | ||||
| PSA, per unit | 0.97 | 0.33, 2.86 | 37.9 | 23.0, 62.5 |
| Spline term | 1.02 | 0.86, 1.22 | 0.03 | 0.02, 0.05 |
| PSA treated as categorical variable | ||||
| PSA, ng/mL | ||||
| ≤0.1 ng/mL | Ref. | Ref. | ||
| 0.1–0.2 ng/mL | 6.29 | 1.24, 31.8 | 2.84 | 1.80, 4.48 |
| 0.2–0.5 ng/mL | 0.97 | 0.15, 6.12 | 4.38 | 3.07, 6.25 |
| ≥0.5 | 2.82 | 0.66, 12.0 | 9.38 | 6.46, 13.6 |
Note: In the analysis for PSA at 6 months, patients with no PSA measurement within 6 months after RT start were excluded, which resulted in 1030 patients in the RT group and 3372 in the RT+ADT group.
Abbreviations: ADT, androgen deprivation therapy; CI, confidence interval; PSA, prostate‐specific antigen; RT, radiotherapy; sHR, subdistribution hazard ratio.
Adjusted for age, treatment year, PSA at diagnosis, Gleason Grade Group, cT stage, and ADT use at time of PSA evaluation.
RT with ADT
PSA nadir values and PSA values at 6 and 12 months were significantly associated with PCSM in patients who underwent RT+ADT (Table 2).
Subdistribution hazard ratios (sHR) for PSCM were 2.97 (95% CI, 1.92–4.59), 3.41 (95% CI, 2.29–5.09), and 6.96 (95% CI, 5.07–9.57) for patients who reached PSA at 6 months values 0.1–0.2, 0.2–0.5, and ≥0.5 ng/mL, respectively, compared to patients who reached PSA values ≤0.1 ng/mL. Similar results were found in the model including the lowest PSA at 12 months after treatment. Specifically, in reference to patients who had a PSA at 12 months ≤0.1 ng/mL, sHR were 2.84 (1.80–4.48), 4.38 (3.07–6.25), and 9.39 (6.46–13.6) for patients who reached PSA values 0.1–0.2, 0.2–0.5, and ≥0.5 ng/mL, respectively. Figure 2 graphically presents adjusted cancer‐specific survival probability after PSA at nadir, PSA at 6 months, and PSA at 12 months according to PSA values in patients who underwent RT+ADT. Adjusted cumulative incidence rates of PCSM according to PSA values at nadir and PSA values at 6 and 12 months after treatment are presented in Figure 3 (using the same PSA cutoffs of the competing risk models).
FIGURE 2.
Adjusted cancer‐specific survival probability after PSA at nadir (A), PSA at 6 (B) and 12 (C) months according to PSA values in patients who underwent RT+ADT. ADT indicates androgen deprivation therapy; PSA, prostate‐specific antigen; RT, radiotherapy.
FIGURE 3.
Population averaged cumulative incidence function curves demonstrating prostate cancer–specific mortality after (A) PSA nadir and according to PSA nadir value; (B) after 6 months since RT according to PSA values at 6 months; and (C) after 12 months since RT according to PSA values at 12 months. Adjusted for age at diagnosis, treatment year, PSA at diagnosis, Gleason Grade Group, cT stage, and ADT use at time of PSA evaluation. ADT indicates androgen deprivation therapy; PSA, prostate‐specific antigen; RT, radiotherapy.
Twelve‐year PCSM rates among patients undergoing RT+ADT were 27%, 15%, 13%, and 5% in patients with PSA at 6 months ≥0.5 ng/mL, 0.2–0.5, 0.1–0.2, and ≤0.1 and 34%, 18%, 12%, and 5% in patients with PSA at 12 months ≥0.5 ng/mL, 0.2–0.5, 0.1–0.2, and ≤0.1, respectively.
The subanalysis in patients who underwent RT with a short‐term ADT (4–6 months) did show the same association between PSA nadir value, PSA at 6 months, and PSA at 12 months after RT and PCSM (Table S3).
Association of detectable and undetectable PSA values after RT with PCSM
RT
In the RT group, 181 patients experienced BCR. In the competing risk regression, there was no evidence of association between PSA nadir values and PSA values at 6 and 12 months with BCR.
Radiotherapy with ADT
A total of 676 BCR occurred in the RT+ADT group. PSA nadir values, PSA values at 6 months, and PSA values at 12 months were significantly associated with BCR. As shown in Figure S3, the risk of BCR increased progressively with higher PSA values at all time points. Patients with PSA >0.5 ng/mL at 6 months had a 5‐year cumulative incidence of BCR of 38%, compared to 24% for those with PSA 0.2–0.5 ng/mL, 22% for PSA 0.1–0.2 ng/mL, and 11% for PSA ≤0.1 ng/mL. Similar gradients in BCR risk were observed when stratifying by PSA at nadir and at 12 months.
DISCUSSION
In this large population‐based study, we found that up to 85% of patients undergoing RT+ADT achieved an undetectable PSA value within 6 to 12 months of treatment. Conversely, incomplete early PSA response (PSA >0.1 ng/mL at 6 or 12 months) after RT+ADT was associated with worse oncological outcomes. Similar results have been previously reported. D’Amico et al. 17 showed that in patients receiving EBRT and 6 months of ADT, if the nadir PSA or the PSA at 6 months was >0.5 ng/mL, oncological outcomes were significantly worse. The authors suggested that patients with a detectable PSA nadir and a quick time to nadir have a higher likelihood of clinically silent castrate‐resistant PCa and are more likely to develop castrate‐resistant disease. Bryant et al. 18 proved the prognostic significance of the PSA nadir at 3 months after RT. Furthermore, a PSA nadir 0.5 ng/mL showed potential as a surrogate end point of overall mortality meeting the Prentice criteria. 19 Finally, using individual patients’ data from 16 randomized trials evaluating RT±ADT for localized prostate cancer, Kwak et al. 8 recently pointed out that PSA ≥0.1 ng/mL within 6 months after RT completion was prognostic for long‐term outcomes in patients treated with RT±ADT for localized PCa.
We took a step further. Indeed, for patients who did not reach PSA ≤0.1 ng/mL, it is always unsure if they will have further PSA drops at later follow‐ups. This hinders the clinical utility of the concept of PSA nadir value. To overcome such limitation, we looked in a population‐based study with long‐term follow‐up at the probabilities to reach the PSA nadir over time and we noticed that most patients undergoing RT+ADT had a complete PSA response (PSA ≤0.1 ng/mL) before 6 (83%) and 12 (86%) months after treatment. By using the lowest PSA taken at these time points, clinicians can select patients at high risk of progressing and PCSM at an earlier stage. 20 Indeed, a growing body of evidence shows that PSMA‐PET is often positive at PSA values below Phoenix criteria. 21 , 22 , 23 Waiting for the PSA to rise to such levels may potentially result in diagnostic delays whereas early imaging may detect recurrences at a less advanced disease stage, allowing potential salvage treatments. 21 Our findings emphasize the valuable information contained in the PSA response following RT. Future randomized clinical trials are crucial to establish whether therapy intensification or deintensification based on early PSA response can improve survival outcomes as opposed to relying on noncurative long‐term ADT and “wait and see” approaches.
Limitations of this study include the retrospective nature, the lack of information on serum testosterone levels, and restaging with PSMA PET in patients with detectable PSA after RT. Moreover, the sub analysis in patients undergoing RT without ADT is limited by the low number of PCa deaths during the follow‐up. Nevertheless, we demonstrated that PSA response is less predictable in patients undergoing RT without ADT, 5 and early PSA values are not associated with BCR in patients undergoing RT alone. Finally, our database includes patients treated over a timeframe that may precede widespread use of novel imaging modalities and novel systemic therapies. Thus, the historical nature of these data could account for higher PCSM rates in some subsets compared to what might be observed in contemporary clinical practice.
The main strength is the long‐term follow‐up and the population‐based nature. We have access to a virtually complete follow‐up for PSA, prescribed medications, and causes of death for patients included in the study. Additionally, our results reflect outcomes of patients treated outside clinical trials, in the real‐world setting.
In conclusion, this study highlights the importance of early PSA response as a predictor of long‐term outcomes in prostate cancer patients undergoing RT±ADT. Most RT+ADT patients achieved a PSA nadir ≤0.1 ng/mL within 6 to 12 months, a milestone significantly associated with reduced PCSM risk. Conversely, patients with PSA levels >0.1 ng/mL at these time points exhibited a heightened risk, emphasizing the need for timely restaging and potentially intensified salvage treatments.
Future research should explore the role of early PSA monitoring in tailoring treatment strategies, including de‐escalation for low‐risk patients and intensification for those at high risk. Randomized trials are necessary to validate the prognostic value of early PSA response and evaluate the efficacy of novel therapeutic interventions for improving survival outcomes.
AUTHOR CONTRIBUTIONS
Ugo Giovanni Falagario: Conceptualization, data curation, formal analysis, funding acquisition, methodology, writing–original draft, and writing–review and editing. Francesco Pellegrino: Formal analysis, methodology, writing–original draft, and writing–review and editing. Ahmad Abbadi: Data curation and writing–review and editing. Lars Björnebo: Data curation, methodology, and writing–review and editing. Alexander Valdman: Supervision and writing–review and editing. Marie Hjelm Eriksson: Supervision and writing–review and editing. Renata Zelic: Methodology and writing–review and editing. Giuseppe Carrieri: Supervision and writing–review and editing. Giorgio Gandaglia: Supervision and writing–review and editing. Alberto Briganti: Supervision and writing–review and editing. Francesco Montorsi: Supervision and writing–review and editing. Markus Aly: Supervision and writing–review and editing. Thorgerdur Palsdottir: Data curation and writing–review and editing. Martin Eklund: Methodology and writing–review and editing. Tobias Nordström: Conceptualization and writing–review and editing. Henrik Grönberg: Conceptualization, supervision, and writing–review and editing. Olof Akre: Conceptualization, supervision, and writing–review and editing. Peter Wiklund: Conceptualization, funding acquisition, supervision, and writing–review and editing. Anna Lantz: Conceptualization, funding acquisition, supervision, and writing–review and editing.
CONFLICT OF INTEREST STATEMENT
Markus Aly reports consulting fees from Astellas Pharma Europe. Martin Eklund reports fees for professional activities from A3P Biomedical AB. Henrik Grönberg holds stock in A3P Biomedical and he holds five patents related to prostate cancer diagnostics. Anna Lantz reports fees for professional activities from Karolinska Institutet. Tobias Nordström holds stock in A3P Biomedical. Alexander Valdman reports fees for professional activities from Karolinska Institutet. The other authors disclose no conflicts of interest.
Supporting information
Supplementary Material
ACKNOWLEDGMENTS
This study was supported by a grant from the European Urological Association scholarship program awarded to Ugo Giovanni Falagario. This study was approved by the Swedish Ethical Review Authority (2012/438‐31/3). Informed consent was waived due to the retrospective, registry‐based nature of the study. The waiver was granted by the Swedish Ethical Review Authority as the research involved minimal risk and could not practicably be performed otherwise.
Falagario UG, Pellegrino F, Abbadi A, et al. Detectable prostate‐specific antigen after radiotherapy for clinically localized prostate cancer. Cancer. 2025;e70045. doi: 10.1002/cncr.70045
DATA AVAILABILITY STATEMENT
The data underlying this study are available from the Swedish National Prostate Cancer Register and other national health registries. Access to these data is subject to ethical approval and data‐sharing agreements with the relevant authorities and is not publicly available. Qualified researchers may apply for access through the appropriate channels in accordance with Swedish data protection regulations.
REFERENCES
- 1. Stamey TA, Yang N, Hay AR, McNeal JE, Freiha FS, Redwine E. Prostate‐specific antigen as a serum marker for adenocarcinoma of the prostate. N Engl J Med. 1987;317(15):909‐916. doi: 10.1056/NEJM198710083171501 [DOI] [PubMed] [Google Scholar]
- 2. Treacy PJ, Falagario UG, Magniez F, et al. Decipher Score predicts prostate specific antigen persistence after prostatectomy. Minerva Urol Nephrol. 2023;75(5):583‐590. doi: 10.23736/S2724-6051.23.05395-8 [DOI] [PubMed] [Google Scholar]
- 3. Tilki D, van den Bergh RCN, Briers E, et al. EAU‐EANM‐ESTRO‐ESUR‐ISUP‐SIOG guidelines on prostate cancer. Part II‐2024 update: treatment of relapsing and metastatic prostate cancer. Eur Urol. 2024;86(2):164‐182. doi: 10.1016/j.eururo.2024.04.010 [DOI] [PubMed] [Google Scholar]
- 4. Scilipoti P, Garmo H, Gedeborg R, Robinson D, Stattin P, Westerberg M. Incidence and prognostic implications of PSA persistence and relapse after radical prostatectomy. Population‐based study. J Natl Cancer Inst. 2025;117(6):1142‐1150. doi: 10.1093/jnci/djaf012 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Roach M 3rd, Hanks G, Thames H Jr., et al. Defining biochemical failure following radiotherapy with or without hormonal therapy in men with clinically localized prostate cancer: recommendations of the RTOG‐ASTRO Phoenix Consensus Conference. Int J Radiat Oncol Biol Phys. 2006;65(4):965‐974. doi: 10.1016/j.ijrobp.2006.04.029 [DOI] [PubMed] [Google Scholar]
- 6. Wenzel M, Dariane C, Saad F, et al. The impact of time to prostate specific antigen nadir on biochemical recurrence and mortality rates after radiation therapy for localized prostate cancer. Urol Oncol. 2022;40(2):57e15‐57e23. doi: 10.1016/j.urolonc.2021.06.025 [DOI] [PubMed] [Google Scholar]
- 7. Pike LRG, Wu J, Chen MH, et al. Time to prostate‐specific antigen nadir and the risk of death from prostate cancer following radiation and androgen deprivation therapy. Urology. 2019;126:145‐151. doi: 10.1016/j.urology.2018.11.056 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Kwak L, Ravi P, Armstrong JG, et al. Prognostic impact of prostate‐specific antigen at 6 months after radiotherapy in localized prostate cancer: an individual patient data analysis of randomized trials. J Clin Oncol. 2024;42(18):2132‐2138. doi: 10.1200/JCO.23.00762 [DOI] [PubMed] [Google Scholar]
- 9. Valle LF, Lehrer EJ, Markovic D, et al. A systematic review and meta‐analysis of local salvage therapies after radiotherapy for prostate cancer (MASTER). Eur Urol. 2021;80(3):280‐292. doi: 10.1016/j.eururo.2020.11.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Covas Moschovas M, Saikali S, Sandri M, et al. Outcomes of salvage robotic‐assisted radical prostatectomy: high‐volume multicentric data from the European Association of Urology Robotic Urology Section Scientific Working Group. Eur Urol. 2025;88(1):103‐113. doi: 10.1016/j.eururo.2025.03.009 [DOI] [PubMed] [Google Scholar]
- 11. LBA02‐09 EMBARK . A phase 3 randomized study of enzalutamide or placebo plus leuprolide acetate and enzalutamide monotherapy in high‐risk biochemically recurrent prostate cancer. J Urol. 2023;210(1):224‐226. doi: 10.1097/JU.0000000000003518 [DOI] [PubMed] [Google Scholar]
- 12. Aly M, Leval A, Schain F, et al. Survival in patients diagnosed with castration‐resistant prostate cancer: a population‐based observational study in Sweden. Scand J Urol. 2020;54(2):115‐121. doi: 10.1080/21681805.2020.1739139 [DOI] [PubMed] [Google Scholar]
- 13. Brooke HL, Talback M, Hornblad J, et al. The Swedish cause of death register. Eur J Epidemiol. 2017;32(9):765‐773. doi: 10.1007/s10654-017-0316-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Pazzagli L, Brandt L, Linder M, et al. Methods for constructing treatment episodes and impact on exposure‐outcome associations. Eur J Clin Pharmacol. Feb 2020;76(2):267‐275. doi: 10.1007/s00228-019-02780-4 [DOI] [PubMed] [Google Scholar]
- 15. Coviello V, Boggess M. Cumulative incidence estimation in the presence of competing risks. STATA J. 2004;4(2):103‐112. doi: 10.1177/1536867X0400400201 [DOI] [Google Scholar]
- 16. Austin PC, Ibrahim M, Putter H. Accounting for competing risks in clinical research. JAMA. 2024;331(24):331. doi: 10.1001/jama.2024.4970 [DOI] [PubMed] [Google Scholar]
- 17. D'Amico AV, Chen MH, de Castro M, et al. Surrogate endpoints for prostate cancer‐specific mortality after radiotherapy and androgen suppression therapy in men with localised or locally advanced prostate cancer: an analysis of two randomised trials. Lancet Oncol. 2012;13(2):189‐195. doi: 10.1016/S1470-2045(11)70295-9 [DOI] [PubMed] [Google Scholar]
- 18. Bryant AK, D'Amico AV, Nguyen PL, et al. Three‐month posttreatment prostate‐specific antigen level as a biomarker of treatment response in patients with intermediate‐risk or high‐risk prostate cancer treated with androgen deprivation therapy and radiotherapy. Cancer. 2018;124(14):2939‐2947. doi: 10.1002/cncr.31400 [DOI] [PubMed] [Google Scholar]
- 19. Royce TJ, Chen MH, Wu J, et al. Surrogate end points for all‐cause mortality in men with localized unfavorable‐risk prostate cancer treated with radiation therapy vs radiation therapy plus androgen deprivation therapy: a secondary analysis of a randomized clinical trial. JAMA Oncol. 2017;3(5):652‐658. doi: 10.1001/jamaoncol.2016.5983 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Moreira DM, Howard LE, Sourbeer KN, et al. Predictors of time to metastasis in castration‐resistant prostate cancer. Urology. 2016;96:171‐176. doi: 10.1016/j.urology.2016.06.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. van Altena EJE, Jansen BHE, Korbee ML, et al. Prostate‐specific membrane antigen positron emission tomography before reaching the Phoenix Criteria for biochemical recurrence of prostate cancer after radiotherapy: earlier detection of recurrences. Eur Urol Oncol. 2024;8(2):417‐424. doi: 10.1016/j.euo.2024.09.015 [DOI] [PubMed] [Google Scholar]
- 22. Jansen BHE, van Leeuwen PJ, Wondergem M, et al. Detection of recurrent prostate cancer using prostate‐specific membrane antigen positron emission tomography in patients not meeting the Phoenix Criteria for biochemical recurrence after curative radiotherapy. Eur Urol Oncol. 2021;4(5):821‐825. doi: 10.1016/j.euo.2020.01.002 [DOI] [PubMed] [Google Scholar]
- 23. Thomas S, Callahan J, Conway P, et al. High detection rates for prostate‐specific membrane antigen‐avid prostate cancer recurrence at low prostate‐specific antigen levels on extended axial field‐of‐view positron emission tomography/computed tomography. Eur Urol Open Sci. 2025;71:49‐56. doi: 10.1016/j.euros.2024.11.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplementary Material
Data Availability Statement
The data underlying this study are available from the Swedish National Prostate Cancer Register and other national health registries. Access to these data is subject to ethical approval and data‐sharing agreements with the relevant authorities and is not publicly available. Qualified researchers may apply for access through the appropriate channels in accordance with Swedish data protection regulations.