Abstract
Purpose:
Men with Gleason score 9–10 prostate cancer have worse outcomes compared to those with Gleason 8 disease. Upfront treatments remain controversial for these patients. Using the Surveillance, Epidemiology, and End Results (SEER) database, we evaluated the impact of initial treatment with external beam radiation therapy (EBRT), external beam radiation therapy with brachytherapy (EBRT+BT), or surgery on prostate cancer-specific mortality (PCSM) and overall mortality (OM) in Gleason 9–10 disease.
Methods:
The SEER database was queried for men diagnosed with biopsy Gleason 9–10 prostate cancer from 2005–2014. Gathered data included demographic, pathologic, therapy received, and survival outcomes. Kaplan-Meier survival curves and crude and multivariate analyses were generated for initial therapy with EBRT, EBRT+BT, or surgery.
Results:
A total of 7,877 men were included, 4,465 (56.7%) who underwent upfront treatment with EBRT alone, 623 (7.9%) with EBRT+BT, and 2,789 (35.4%) with surgery. The 7-year PCSM rates were 29.2%, 15.0%, and 14.6% for EBRT, EBRT+BT, and surgery respectively (p < 0.001). The 7-year OM rates were 43.8%, 27.2%, and 20.0% for EBRT, EBRT+BT, and surgery respectively (p < 0.001).When controlling for age, year of diagnosis, Gleason score, clinical T stage, and PSA level on multivariate analysis, EBRT had greater PCSM and OM than surgery (HR 0.41, 95% CI 0.28 – 0.61, p < 0.001 and HR 0.44, 95% CI 0.34 – 0.57, p < 0.001 respectively), but the mortality differences was not statistically significant between EBRT and EBRT+BT.
Conclusions:
Among men with localized Gleason 9–10 disease, surgery was associated with statistically significant improved survival outcomes compared to EBRT alone.
Keywords: SEER, prostate cancer, prostatectomy, radiation
INTRODUCTION
Prostate cancer (PCa) is the second most common cause of cancer in men both in the United States and worldwide.(1) In 2020 alone, it is estimated that there will be over191,000 new cases of prostate cancer and approximately 33,000 prostate cancer deaths in the United States.(2) Men with more aggressive disease are more likely to have worse outcomes and succumb to their cancer. Currently, the National Comprehensive Cancer Network (NCCN) defines high-risk PCa as T3a, Gleason score (GS) 8–10 or prostate specific antigen (PSA) >20 ng/mL disease.(3) There is a major focus to elucidate the best treatment options for this group of high-risk patients, specifically defined by GS 8–10 disease.(4,5) Conflicting evidence exists on whether radical prostatectomy (RP) or radiation therapy (RT) is optimal in this group of men.(6,7) A recent study comparing RP versus external beam radiation therapy plus brachytherapy (EBRT + BT) in patients with high-risk localized PCa using the national cancer database (NCDB) showed that RP may offer overall survival benefit as initial therapy compared to EBRT+BT.(8) Again, this study defined high-risk according to the NCCN guidelines. Including men with Gleason score 8 PCa, however, may not be the best way to evaluate treatment outcomes for men with truly high-risk disease. Within the last few years, Epstein et al. suggested a new classification for GS stratification where GS 8 is categorized separately from GS 9–10 and showed significantly varying outcomes between the two groups.(9) Another study found that Gleason pattern 5+4 was associated with significantly lower cancer-specific survival outcomes compared to those with Gleason pattern of 4+4, suggesting that GS 8 PCa may dilute the risk of higher Gleason grade PCa.(10) Still, very few studies have adopted this classification system to assess treatment outcomes in high-risk prostate cancer. Using the Surveillance, Epidemiology, and End Results (SEER) database, we evaluated the impact of initial treatment with external beam radiation therapy (EBRT), EBRT+BT, or surgery on prostate cancer-specific mortality (PCSM) and overall mortality in localized, Gleason 9–10 disease.
MATERIALS AND METHODS
Data Source
This retrospective cohort study utilized patient data from the SEER database, a cancer-based registry sponsored by the National Cancer Institute. The database provides cancer incidence and survival outcomes for approximately 28% of the United States population. The most recent version of the SEER 18 Registries was used in this study. This data was de-identified and was exempt by the University of Maryland Baltimore Human Research Protections Program institutional review board.
Patient selection
Men with adenocarcinoma (ICD-0–3 code 8140) of the prostate (site code 61.9) from 2005–2014 were identified. We chose a 10 year time interval starting from the year the most recent data was available at the time of data extraction. Only those without nodal involvement or metastasis, N0M0 stage by the American Joint Committee on Cancer (AJCC) staging manual 6th or 7th edition, GS 9 and 10 prostate cancer diagnosed by prostate biopsy were included in this study. Men with less than 6 months of follow-up were excluded. All patients were treated with either upfront RP or RT. Radiation could have been either EBRT alone or EBRT with brachytherapy. The status of hormone therapy was unknown in this patient sample. Patients who had adjuvant or salvage RT were included in the RP cohort.
Covariates:
Covariates were chosen based on those used in similar previous studies. They were obtained using the SEER registries. These included age (<60, 60–69, 70–79, and ≥80 years), year of diagnosis, Gleason Score (9 or 10), clinical tumor (T) stage (AJCC Staging: T1, T2, T3, and T4), and pretreatment PSA concentration (<10, 10–20, or >20 ng/mL).
Outcome Measures:
The primary outcomes of interest were prostate cancer-specific mortality and overall mortality. Cause of death was defined using the SEER cause of death recode. Survival duration was defined as time from date of diagnosis to date of death or last contact. Cases who died from prostate cancer were designated as PCSM, while those who died from any cause were designated as overall mortality.
Statistical Analysis
To assess for associations between categorical variables and treatment groups, Pearson χ2 tests were used. Kaplan Meier survival curves and bivariate and multivariable analyses were generated for initial therapy with EBRT, EBRT+BT, or surgery. Cox proportional hazard regression was used to determine hazard ratios (HR) with 95% confidence intervals. Inverse probability of treatment weighting (IPTW) was conducted as another method to reduce bias in this sample population where treatment was not randomized. All p-values were acquired from two-sided tests with a significance level of p < 0.05. Analyses were performed with Stata 14.1 (Stata Corporation, College Station, TX, USA).
RESULTS
A total of 7,877 men were identified in the SEER database with biopsy GS 9–10, N0M0, diagnosed in 2005–2014 who made up our final study population. Of these men, 4,465 (56.7%) underwent initial EBRT, 623 (7.9%) underwent EBRT+BT, and 2,789 (35.6%) underwent initial surgery. Of those 2,789 men who underwent initial surgery, 564 (20.2%) received radiation at some point after surgery, and regional lymph nodes were removed for examination in 2,248 (80.6%) of them.
The median follow-up time for the entire cohort was 32 months (range, 6–114 months). Median follow-up for EBRT, EBRT+BT, and surgery were 31 months (range, 6–113 months), 35 months (range, 6–109 months), and 32 months (range, 6–114 months), respectively. The median age at diagnosis for EBRT, EBRT+BT and surgery were 72 years (range, 42–93 years), 68 years (range, 46–88 years), and 64 years (range, 39–88 years), respectively. There was a statistically significant association between age and treatment group (p<0.001). Additional patient characteristics including race, marital status, GS score, T stage and PSA level are displayed in Table 1, and all showed statistically significant associations with the treatment group.
Table 1:
Patient and treatment attributes for men with localized high-grade prostate cancer in the EBRT, EBRT+BT, and surgery treatment groups
| All Patients | EBRT | EBRT +BT | Surgery | ||
|---|---|---|---|---|---|
| Attribute | N | N (%) | N (%) | N (%) | P |
| Total | 7,877 | 4,465 (56.7) | 623 (7.9) | 2,789 (35.4) | |
| Age at diagnosis (y) | < 0.001 | ||||
| <60 | 1,183 | 389 (8.7) | 98 (15.7) | 696 (25.0) | |
| 60–69 | 3,169 | 1,404 (31.4) | 257 (41.3) | 1,508 (54.1) | |
| 70–79 | 2,787 | 2,003 (44.9) | 227 (36.4) | 557 (20.0) | |
| 80+ | 738 | 669 (15.0) | 41 (6.6) | 28 (1.0) | |
| Median (Range) | 68 (39–93) | 72 (42–93) | 68 (46–88) | 64 (39–88) | |
| Year of Diagnosis | 0.078 | ||||
| 2005–2009 | 459 | 304 (6.8) | 45 (7.2) | 146 (5.2) | |
| 2010–2014 | 7,418 | 4,161 (93.2) | 578 (92.8) | 2,643 (94.8) | |
| Race | <0.001 | ||||
| White | 6,221 | 3,491 (78.2) | 474 (76.1) | 2,256 (80.9) | |
| Black | 1,058 | 662 (14.8) | 86 (13.8) | 310 (11.1) | |
| Other | 507 | 246 (5.5) | 58 (9.3) | 203 (7.3) | |
| Unknown | 91 | 66 (1.5) | 5 (0.8) | 20 (0.7) | |
| Marital status at diagnosis | <0.001 | ||||
| Married/ Domestic Partner | 5,416 | 2,905 (65.1) | 429 (68.9) | 2,082 (74.6) | |
| Single (never married) | 744 | 434 (9.7) | 51 (8.2) | 259 (9.3) | |
| Separated/Divorced/Widowed | 1,072 | 709 (15.9) | 73 (11.7) | 290 (10.4) | |
| Unknown | 645 | 417 (9.3) | 70 (11.2) | 158 (5.7) | |
| Gleason score | <0.001 | ||||
| 9 | 7,219 | 4,042 (90.5) | 588 (94.4) | 2,589 (92.8) | |
| 10 | 658 | 423 (9.5) | 35 (5.6) | 200 (7.2) | |
| Clinical T stage | <0.001 | ||||
| T1 | 2,351 | 2,038 (45.7) | 297 (47.7) | 16 (0.6) | |
| T2 | 3,068 | 1,796 (40.3) | 249 (40.0) | 1,023 (36.7) | |
| T3 | 2,319 | 521 (11.7) | 73 (11.7) | 1,725 (61.8) | |
| T4 | 119 | 91 (2.0) | 3 (0.5) | 25 (0.9) | |
| Unknown | 20 | 19 (0.4) | 1 (0.2) | 0 | |
| PSA level | <0.001 | ||||
| < 10 | 4,118 | 2,010 (45.1) | 332 (53.2) | 1,776 (63.7) | |
| 10–<20 | 1,749 | 1,073 (24.1) | 137 (22.0) | 539 (19.3) | |
| 20+ | 1,627 | 1,198 (26.9) | 129 (20.7) | 300 (10.8) | |
| Unknown | 383 | 175 (3.9) | 25 (4.0) | 174 (6.2) |
HR- Hazard Ratio
95% CI – 95% Confidence Interval
The estimated 7-year prostate cancer-specific mortality rates were 29.2%, 15.0%, and 14.6% for EBRT, EBRT+BT, and surgery, respectively (p < 0.001) (Figure 1). The estimated 7-year overall mortality rates were 43.8%, 27.2%, and 20.0% for EBRT, EBRT+BT, and surgery, respectively (p < 0.001) (Figure 2).
Figure 1.
Kaplan Meier plot for prostate cancer-specific survival after initial therapy with EBRT alone vs. EBRT+BT vs. surgery for biopsy Gleason score 9–10 prostate cancer
Figure 2.
Kaplan Meier plot for overall survival after initial therapy with EBRT alone vs. EBRT+BT vs. surgery for biopsy Gleason score 9–10 prostate cancer
Both EBRT+BT and surgery showed improved PCSM over EBRT alone in crude analysis (HR 0.50, 95% CI 0.31–0.82, p = 0.006 and HR 0.36, 95% CI 0.27–0.49, p<0.001, respectively) (Table 2). On crude analysis, overall mortality was also improved for EBRT+BT and surgery over EBRT alone (HR 0.62, 95% CI 0.46–0.82, p =0.001 and HR 0.32, 95% CI 0.26–0.40, p<0.001, respectively) (Table 3).
Table 2.
Bivariate and Multivariable Cox proportional hazards model of prostate cancer-specific mortality for primary treatment with EBRT vs EBRT-BT vs surgery and covariates
| Bivariate | Multivariable | |||||
|---|---|---|---|---|---|---|
| Variable | HR | 95% CI | P value | HR | 95% CI | P value |
| Treatment modality | ||||||
| EBRT | 1.00 | 1.00 | ||||
| EBRT+BT | 0.50 | 0.31–0.82 | 0.006 | 0.68 | 0.41–1.12 | 0.130 |
| Surgery | 0.36 | 0.27–0.49 | <0.001 | 0.41 | 0.28–0.61 | <0.001 |
| Age at diagnosis (y) | ||||||
| <60 | 1.00 | 1.00 | ||||
| 60–69 | 0.78 | 0.54–1.13 | 0.185 | 0.71 | 0.48–1.05 | 0.083 |
| 70–79 | 1.34 | 0.94–1.90 | 0.108 | 1.10 | 0.75–1.62 | 0.632 |
| 80+ | 1.73 | 1.12–2.69 | 0.015 | 1.14 | 0.69–1.88 | 0.601 |
| Year of diagnosis | ||||||
| 2005–2009 | 1.00 | |||||
| 2010–2014 | 0.85 | 0.60–1.21 | 0.364 | 0.87 | 0.59–1.26 | 0.459 |
| Race | ||||||
| White | 1.00 | 1.00 | ||||
| Black | 1.24 | 0.89–1.71 | 0.201 | 1.16 | 0.82–1.65 | 0.410 |
| Marital status at diagnosis | ||||||
| Married/ Domestic Partner | 1.00 | 1.00 | ||||
| Single (Never Married) | 1.21 | 0.81–1.81 | 0.346 | 1.12 | 0.72–1.73 | 0.626 |
| Separated/Divorced/Widowed | 1.69 | 1.25–2.29 | 0.001 | 1.44 | 1.05–1.98 | 0.024 |
| Gleason score | ||||||
| 9 | 1.00 | 1.00 | ||||
| 10 | 2.72 | 2.01–3.69 | <0.001 | 2.41 | 1.75–3.31 | <0.001 |
| Clinical T stage | ||||||
| T1 | 1.00 | 1.00 | ||||
| T2 | 0.81 | 0.61–1.07 | 0.135 | 1.02 | 0.74–1.36 | 0.993 |
| T3 | 0.84 | 0.62–1.14 | 0.262 | 1.66 | 1.15–2.35 | 0.007 |
| T4 | 1.97 | 1.02–3.78 | 0.042 | 2.11 | 1.08–4.10 | 0.029 |
| PSA level | ||||||
| < 10 | 1.00 | 1.00 | ||||
| 10–<20 | 1.30 | 0.95–1.78 | 0.107 | 1.06 | 0.76–1.46 | 0.737 |
| 20+ | 2.06 | 1.55–2.73 | <0.001 | 1.61 | 1.20–2.16 | 0.002 |
HR- Hazard Ratio
95% CI – 95% Confidence Interval
Table 3.
Bivariate and Multivariate Cox proportional hazards model of overall mortality for primary treatment with EBRT vs. EBRT-BT vs. surgery and covariates
| Bivariate | Multivariate | |||||
|---|---|---|---|---|---|---|
| Variable | HR | 95% CI | P value | HR | 95% CI | P value |
| Treatment modality | ||||||
| EBRT | 1.00 | 1.00 | ||||
| EBRT+BT | 0.62 | 0.46–0.82 | 0.001 | 0.85 | 0.64–1.14 | 0.278 |
| Surgery | 0.32 | 0.26–0.40 | <0.001 | 0.44 | 0.34–0.57 | <0.001 |
| Age at diagnosis (y) | ||||||
| <60 | 1.00 | 1.00 | ||||
| 60–69 | 1.24 | 0.94–1.63 | 0.126 | 1.18 | 0.89–1.57 | 0.255 |
| 70–79 | 2.02 | 1.55–2.64 | <0.001 | 1.66 | 1.24–2.21 | 0.001 |
| 80+ | 3.80 | 2.83–5.09 | <0.001 | 2.59 | 1.87–3.559 | <0.001 |
| Year of diagnosis | ||||||
| 2005–2009 | 1.00 | 1.00 | ||||
| 2010–2014 | 1.06 | 0.84–1.34 | 0.627 | 1.08 | 0.84–1.39 | 0.529 |
| Race | ||||||
| White | 1.00 | 1.00 | ||||
| Black | 1.09 | 0.88–1.35 | 0.415 | 1.15 | 0.92–1.44 | 0.230 |
| Marital status at diagnosis | ||||||
| Married/ Domestic Partner | 1.00 | 1.00 | ||||
| Single (Never Married) | 1.23 | 0.96–1.58 | 0.109 | 1.31 | 1.00–1.71 | 0.049 |
| Separated/Divorced/Widowed | 1.72 | 1.43–2.08 | <0.001 | 1.48 | 1.22–1.81 | <0.001 |
| Gleason score | ||||||
| 9 | 1.00 | 1.00 | ||||
| 10 | 1.98 | 1.61–2.44 | <0.001 | 1.65 | 1.32–2.07 | <0.001 |
| Clinical T stage | ||||||
| T1 | 1.00 | 1.00 | ||||
| T2 | 0.79 | 0.66–0.94 | 0.008 | 1.02 | 0.85–1.23 | 0.822 |
| T3 | 0.61 | 0.50–0.75 | <0.001 | 1.24 | 0.97–1.57 | 0.087 |
| T4 | 1.70 | 1.11–2.60 | 0.015 | 1.79 | 1.14–2.82 | 0.012 |
| PSA level | ||||||
| < 10 | 1.00 | 1.00 | ||||
| 10–<20 | 1.59 | 1.32–1.92 | <0.001 | 1.35 | 1.12–1.63 | 0.002 |
| 20+ | 1.70 | 1.41–2.04 | <0.001 | 1.40 | 1.15–1.69 | 0.001 |
HR- Hazard Ratio
95% CI – 95% Confidence Interval
Older age at diagnosis and higher Gleason score, clinical T stage, and PSA level were all associated with worse PCSM and overall mortality (Tables 2–3). On multivariate analysis, having a marital status of separated, divorced, or widowed was associated with worse PCSM (HR 1.44, CI 95% 1.05–1.98, p=0.02) and overall mortality (HR 1.48, CI 95% 1.22–1.81, p<0.001) compared to those who were married, while being single and never married was associated with overall mortality (HR 1.31, CI 95% 1.00–1.71, p=0.05), but not associated with worse PCSM (Tables 2–3).
When controlling for age, year of diagnosis, Gleason score, clinical T stage, and PSA level on multivariate analysis, both PCSM and overall mortality were significantly improved in surgery compared to EBRT (HR 0.41, 95% CI 0.28–0.61, p<0.001 and HR 0.44, 95% CI 0.34–0.57, p<0.001, respectively), while the PCSM and overall mortality comparing EBRT+BT to EBRT were not significantly different (Table 2–3). We achieved similar results using an IPTW model, with only surgery showing a statistically significant improvement in PCSM and overall mortality over EBRT (HR 0.37, 95% CI 0.22–0.62, p<0.001 and HR 0.54, 95% CI 0.28–1.01, p=0.05, respectively) (Table 4).
Table 4.
Inverse probability treatment weighted Multivariate Cox proportional hazards models for prostate cancer-specific mortality and overall mortality
| Prostate Cancer-Specific Mortality | Overall Mortality | |||||
|---|---|---|---|---|---|---|
| Variable | HR | 95% CI | P value | HR | 95% CI | P value |
| Treatment modality | ||||||
| EBRT | 1.00 | 1.00 | ||||
| EBRT+BT | 0.89 | 0.55–1.43 | 0.624 | 0.96 | 0.70–1.33 | 0.820 |
| Surgery | 0.37 | 0.22–0.62 | <0.001 | 0.54 | 0.28–1.01 | 0.054 |
| Age at diagnosis (y) | ||||||
| <60 | 1.00 | 1.00 | ||||
| 60–69 | 1.65 | 0.93–2.93 | 0.087 | 1.82 | 1.09–3.05 | 0.022 |
| 70–79 | 2.29 | 1.22–4.26 | 0.009 | 2.61 | 1.61–4.24 | <0.001 |
| 80+ | 3.16 | 1.51–6.62 | 0.002 | 5.50 | 3.00–10.10 | <0.001 |
| Year of diagnosis | ||||||
| 2005–2009 | 1.00 | 1.00 | ||||
| 2010–2014 | 0.52 | 0.31–0.89 | 0.016 | 0.62 | 0.36–1.06 | 0.079 |
| Gleason score | ||||||
| 9 | 1.00 | 1.00 | ||||
| 10 | 1.88 | 1.06–3.33 | 0.030 | 1.50 | 0.99–2.29 | 0.054 |
| Clinical T stage | ||||||
| T1 | 1.00 | 1.00 | ||||
| T2 | 1.48 | 0.90–2.43 | 0.135 | 1.18 | 0.74–1.88 | 0.492 |
| T3 | 2.37 | 1.34–4.20 | 0.002 | 1.12 | 0.66–1.89 | 0.672 |
| T4 | 3.42 | 1.85–6.31 | <0.001 | 1.37 | 0.76–2.46 | 0.291 |
| PSA level | ||||||
| < 10 | 1.00 | 1.00 | ||||
| 10–<20 | 2.59 | 1.56–4.19 | <0.001 | 2.41 | 1.72–3.38 | <0.001 |
| 20+ | 2.18 | 1.32–3.59 | 0.002 | 1.53 | 1.09–2.14 | 0.013 |
HR- Hazard Ratio
95% CI – 95% Confidence Interval
DISCUSSION
The best treatment option for men with high risk prostate cancer remains unclear. As more information becomes available on outcomes for patients with varying disease characteristics, the definition of high risk prostate cancer continues to evolve. Our study sought to compare surgery, EBRT, and EBRT+BT as initial therapies in a cohort of men with localized high-risk prostate cancer, defined by GS 9 and 10. Using data from 7,877 men within the SEER database, our analysis showed that radical prostatectomy is associated with improved PCSM rates and overall mortality rates for men with localized GS 9 and 10 disease compared to radiation therapy with and without brachytherapy. When controlling for age, year of diagnosis, GS, clinical T stage, and PSA, radical prostatectomy showed a significantly lower risk of mortality than both types of radiation. Our findings are consistent with a study looking at the same treatment groups in Gleason score 8–10 patients; however, they looked specifically at healthy men age 65 years and younger.(11) Within our study, there was a significant association between age and treatment group, with the surgery group having the youngest median age (64 versus 68 and 72 for EBRT and EBRT+BT, respectively).
Our decision to include men with GS 9–10 PCa was to highlight an alternate grading system where Gleason 8 prostate cancer is categorized separately from Gleason grades 9 and 10. In 2016, Epstein et al. proposed a five group grading system (GS 6, 3+4, 4+3, 8, and 9–10) that was found to have the highest prognostic discrimination, and was more accurate than the current grading system.(9) However, not all GS 8 are the same. Gleason 5+3 disease has been shown to behave like Gleason 9 disease in terms of patient survival outcomes.(12) However, most Gleason 8 are 4+4, and only about 900 of 25,000 (<5%) GS 8 men were 5+3 in that study. This suggests the importance of the major pattern of tumor grade in prognosis.
A few prior studies have reported no differences in survival between radical prostatectomy and EBRT+BT. Wang et al., who also gathered patient data from the SEER database, reported that the two treatment modalities offer similar PCSM in patients with Gleason pattern 5 disease.(13) This study included patients with GS 5 + 5, 5 + 4, 5 + 3, 4 + 5, and 3 + 5 on biopsy, with overall GS 8–10. Ennis et al. looked at overall survival in men with Gleason Score 8–10 disease, and also showed no difference in mortality.(14)
One study reported an opposite finding to ours. Kishan et al. used a study design that was most similar to ours: a multi-center, retrospective study comparing disease progression and mortality after RP, EBRT and EBRT+BT in 1,809 patients with GS 9–10 PCa. They found EBRT+BT was associated with better PCSM than EBRT or RP.(15) Compared to our study, this study had longer follow up data and all patients who received radiation were known to have received androgen deprivation therapy (ADT), a data element missing from our data source. Androgen deprivation therapy has been shown to prolong survival and is recommended to be used with radiation.(3) While it is likely the majority of patients in the SEER database between 2005–2014 who received radiation were treated with ADT as well, the exact proportion is not known. Our study, on the other hand, has many strengths and advantages. In addition to our large sample size, our study included a higher proportion of GS 10 patients in all treatment arms.
There are many potential explanations for the differences observed between our results and those previously reported. The sample populations used in each study may have been fundamentally different. The populations within each treatment group were likely biased because none of the trials were randomized. There is always a chance of data or statistical error, and the presence of unknown covariates that cannot be controlled for. Finally, while all the studies focused on high risk PCa patients, there were slight differences in how that was defined with respect to grade. Using GS 8–10, Gleason pattern 5, which includes some men with GS 8, or GS 9 and 10 may drastically change the study population, and could have accounted for the variation in results between studies.
An interesting finding in our data set was the association of marital status with mortality. We found a significant association of worse mortality outcomes in men who have been separated, widowed, or divorced compared to men who are married, but no difference between those who were never married and those who are married. There have been multiple SEER database studies that have showed that unmarried men had an increased risk of both PCSM and overall mortality than married men.(16,17) Both studies, however, do not distinguish never-married men from men who were single by separation, divorce, or death of a spouse. Our results suggest that the worse survival outcomes in unmarried men may not just be due to marital status alone, but also by the potentially stressful event of undergoing separation from a spouse.
LIMITATIONS:
The major limitation in our study, as in all studies that have investigated this topic, is the observational nature of this retrospective cohort design. Such an experiment poses challenges in the form of selection bias that go into selecting treatments for patients. This is apparent by significant difference in patient variables between the treatment groups. Particularly, surgical patients were lower stage than the two other treatment groups, as evidenced by the percent of patients with PSA <10 ng/mL was higher and PSA >20 ng/mL was lower in the surgery group. Typically, healthier patients are selected for surgery, and this may have had an effect on our results. It is also important to note that since 2005, there have been changes in technique and approaches to radiation and surgery.(18) We attempted to address population differences in the treatment arms by using multivariate models and IPTW in our analysis. As stated above, the absence of ADT status, as well as the status of other neo-adjuvant and adjuvant therapies, in the SEER database was another limitation to the study. While we know that about 20% of patients who underwent initial surgery had subsequent radiation, our database did not provide information about whether it was adjuvant or salvage. In addition, patients who received initial radiation may also have undergone salvage radiotherapy, and that is not captured in the database as well.
CONCLUSION:
The need to optimize treatment options for men with high risk prostate cancer is evidenced by the growing literature on the subject. A consensus on high risk disease must be reached through epidemiological studies to find clear patterns in prognosis and through molecular studies to understand the biological basis that drive these patterns. Our study suggests that men with clinically localized, Gleason 9–10 prostate cancer may benefit more from initial surgery than from radiation. These results may sway providers to consider surgery more strongly when the patient is wavering between radiation and surgery. However, in order to properly compare the currently available therapies, surgery and radiation, specifically EBRT+BT based on the literature, randomized controlled trials in this high risk patient population are necessary.
Acknowledgments
FUNDING: No funding was received.
Footnotes
ETHICAL STATEMENTS:
This research was given IRB exemption by the University of Maryland Baltimore Human Research Protections Program (UMB HRPP) and received additional SEER-Medicare exemption, as anonymous data were used.
CONFLICT OF INTEREST:
The authors declare that they have no conflict of interest.
Publisher's Disclaimer: This Author Accepted Manuscript is a PDF file of an unedited peer-reviewed manuscript that has been accepted for publication but has not been copyedited or corrected. The official version of record that is published in the journal is kept up to date and so may therefore differ from this version.
REFERENCES:
- 1.Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA: A Cancer Journal for Clinicians. 2018;68(6):394–424. [DOI] [PubMed] [Google Scholar]
- 2.Cancer statistics, 2020 - Siegel - 2020 - CA: A Cancer Journal for Clinicians - Wiley Online Library [Internet]. [cited 2020 Mar 18]. Available from: https://acsjournals-onlinelibrary-wiley-com.proxyhs.researchport.umd.edu/doi/full/10.3322/caac.21590 [Google Scholar]
- 3.Mohler JL, Armstrong AJ, Bahnson RR, D’Amico AV, Davis BJ, Eastham JA, et al. Prostate Cancer, Version 1.2016. J Natl Compr Canc Netw. 2016. January 1;14(1):19–30. [DOI] [PubMed] [Google Scholar]
- 4.Sundi D, Tosoian JJ, Nyame YA, Alam R, Achim M, Reichard CA, et al. Outcomes of very high-risk prostate cancer after radical prostatectomy: Validation study from 3 centers: VHR Criteria Identify Aggressive PCa. Cancer [Internet]. 2018. November 13 [cited 2019 Jan 11]; Available from:http://doi.wiley.com/10.1002/cncr.31833 [DOI] [PubMed] [Google Scholar]
- 5.Andic F, Izol V, Gokcay S, Arslantas HS, Bayazit Y, Coskun H, et al. Definitive external-beam radiotherapy versus radical prostatectomy in clinically localized high-risk prostate cancer: a retrospective study. BMC Urology [Internet]. 2019. December [cited 2019 Jan 11];19(1). Available from: https://bmcurol.biomedcentral.com/articles/10.1186/s12894-018-0432-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Tewari A, Divine G, Chang P, Shemtov MM, Milowsky M, Nanus D, et al. Long-term survival in men with high grade prostate cancer: a comparison between conservative treatment, radiation therapy and radical prostatectomy--a propensity scoring approach. J Urol. 2007. March;177(3):911–5. [DOI] [PubMed] [Google Scholar]
- 7.Merino T, San Francisco IF, Rojas PA, Bettoli P, Zúñiga A, Besa P. Intensity-modulated radiotherapy versus radical prostatectomy in patients with localized prostate cancer: long-term follow-up. BMC Cancer. 2013. November 8;13:530. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Berg S, Cole AP, Krimphove MJ, Nabi J, Marchese M, Lipsitz SR, et al. Comparative Effectiveness of Radical Prostatectomy Versus External Beam Radiation Therapy Plus Brachytherapy in Patients with High-risk Localized Prostate Cancer. European Urology. 2019. April;75(4):552–5. [DOI] [PubMed] [Google Scholar]
- 9.Epstein JI, Zelefsky MJ, Sjoberg DD, Nelson JB, Egevad L, Magi-Galluzzi C, et al. A Contemporary Prostate Cancer Grading System: A Validated Alternative to the Gleason Score. Eur Urol. 2016. March;69(3):428–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Moschini M, Sharma V, Soligo M, Psutka S, Rangel L, Boorjian SA, et al. Heterogeneity of riskwithin Gleason 4 + 4, 4 + 5 and 5 + 4 prostate cancer. Scandinavian Journal of Urology. 2019. January 2;0(0):1–9. [DOI] [PubMed] [Google Scholar]
- 11.Berg S, Cole AP, Krimphove MJ, Nabi J, Marchese M, Lipsitz SR, et al. Comparative Effectivenessof Radical Prostatectomy Versus External Beam Radiation Therapy Plus Brachytherapy in Patients with High-risk Localized Prostate Cancer. Eur Urol. 2018. November 9; [DOI] [PubMed] [Google Scholar]
- 12.Mahal BA, Muralidhar V, Chen Y-W, Choueiri TK, Hoffman KE, Hu JC, et al. Gleason score 5 + 3 = 8 prostate cancer: much more like Gleason score 9? BJU International. 2016;118(1):95–101. [DOI] [PubMed] [Google Scholar]
- 13.Wang C, Kishan AU, Kamrava M, Steinberg ML, King CR. External Beam Radiation Therapy Witha Brachytherapy Boost Versus Radical Prostatectomy in Gleason Pattern 5 Prostate Cancer: A Population-Based Cohort Study. Int J Radiat Oncol Biol Phys. 2017. 01;98(5):1045–52. [DOI] [PubMed] [Google Scholar]
- 14.Ennis RD, Hu L, Ryemon SN, Lin J, Mazumdar M. Brachytherapy-Based Radiotherapy and RadicalProstatectomy Are Associated With Similar Survival in High-Risk Localized Prostate Cancer. JCO. 2018. February 28;36(12):1192–8. [DOI] [PubMed] [Google Scholar]
- 15.Kishan AU, Cook RR, Ciezki JP, Ross AE, Pomerantz MM, Nguyen PL, et al. Radical Prostatectomy, External Beam Radiotherapy, or External Beam Radiotherapy With Brachytherapy Boost and Disease Progression and Mortality in Patients With Gleason Score 9–10 Prostate Cancer. JAMA. 2018. March 6;319(9):896–905. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Tyson MD, Andrews PE, Etzioni DA, Ferrigni RG, Humphreys MR, Swanson SK, et al. Maritalstatus and prostate cancer outcomes. Can J Urol. 2013. April;20(2):6702–6. [PubMed] [Google Scholar]
- 17.Knipper S, Preisser F, Mazzone E, Mistretta FA, Palumbo C, Tian Z, et al. Contemporary analysis of the effect of marital status on survival of prostate cancer patients across all stages: A populationbased study. Urologic Oncology: Seminars and Original Investigations [Internet]. 2019. May 15 [cited 2019 May 26]; Available from: http://www.sciencedirect.com/science/article/pii/S1078143919301565 [DOI] [PubMed] [Google Scholar]
- 18.Chen J, Oromendia C, Halpern JA, Ballman KV. National trends in management of localized prostatecancer: A population based analysis 2004–2013. The Prostate. 2018;78(7):512–20. [DOI] [PubMed] [Google Scholar]