The role of cytoreductive radical prostatectomy and lymph node dissection in bone‐metastatic prostate cancer: A population‐based study

Abstract Backgrounds The role of cytoreductive radical prostatectomy (cRP) for bone‐metastatic prostate cancer (bmPCa) remains controversial. We aimed to figure out whether cRP and lymph node dissection (LND) can benefit bmPCa. Methods 11,271 PCa patients with bone metastatic burden from 2010 to 2019 were identified using SEER‐Medicare. Overall survival (OS) and cancer‐specific survival (CSS) rates were visualized using Kaplan–Meier plots. Multivariable Cox regression analyses were constructed to examine the effects of cRP and LND on survival, after stratifying to age, prostate specific antigen (PSA), clinical stages, Gleason score, metastatic burden, radiotherapy, and chemotherapy status. Results 317 PCa patients underwent cRP and cRP was increasingly performed for bmPCa from 2010 (2.2%) to 2019 (3.0%) (p < 0.05). In multi analyses, cRP was predisposed to a better OS or CSS in patients with age < 75, PSA < 98 ng/mL, bone‐only metastatic sites or patients not receiving chemotherapy (all p < 0.05). For the patients undergoing cRP, LND especially extended LND was associated with a better OS or CSS (all p < 0.05). Conclusions cRP might benefit OS or CSS in young patients with low PSA and bone‐only metastatic sites not receiving chemotherapy. And a clear OS or CSS benefit of LND especially extended LND was observed in patients undergoing cRP.

Radical prostatectomy (RP), first described by Hugh Hampton Young in 1905, has evolved remarkably for treating PCa, 3,4 and now it is a proven cancer therapy for lowmoderate risk diseases.The operative concepts of RP have evolved rapidly since 21st century.Currently men with high-risk diseases are often offered surgery for a favorable prognosis, which could not be imagined decades ago. 5,6or bone-metastatic prostate cancer (bmPCa), multidisciplinary therapy based on androgen deprivation therapy (ADT) is the standard treatment. 7As for the therapeutic role of cytoreductive radical prostatectomy (cRP), there were studies reporting its beneficial role for bmPCa especially for those with proven low-volume or occult metastatic disease. 8,9However, National Comprehensive Cancer Network (NCCN) and European Association of Urology (EAU) guidelines are silent on this topic. 10Characteristics including tolerance, age, comorbidities, PSA, grade, Gleason score, tumor stage, and metastatic burdens are often considered when determining whether to perform cRP.No previous study has investigate the efficiency of cRP on characteristic-specific survival for bmPCa.Therefore, we carried out this research to verify the effect of cRP and lymph node dissection (LND) on overall survival (OS) and cancer-specific survival (CSS) stratified to prostate specific antigen (PSA), age, clinical stages, Gleason score, metastatic burden, radiotherapy, and chemotherapy status.

| Study design
bmPCa patients were selected within the Surveillance, Epidemiology and End Results (SEER) Medicare database from 2010 to 2019.The eligibility criteria included: (1) bone metastases; (2) tumor sequence number labeled "one primary only"; (3) PCa confirmed by histology; (4) treated without surgery or with cRP (code 0 or 50); and (5) patients with clear clinical tumor stage and metastasis information.Finally, 11,271 patients with bmPCa were identified (Figure 1).317 (2.8%) PCa patients receiving cRP were identified.When assessing the effect of LND, 311 of the 317 patients were selected since 6 patients with unclear LND status were excluded.LND with 1-12 or more than 12 regional lymph nodes (LNs) removed was defined as limited or extended.

| Covariates
Covariate data were obtained from Medicare using codes.For surgical primary site, code 0 (no surgery) and 50 (radical prostatectomy) were identified.For scope of regional lymph node surgery, none, 1-12 regional LNs removed and more than 12 regional LNs removed were selected.

| Outcomes
The primary outcomes were CSS and OS as coded by Medicare.
In multi analyses, low age, recent year of diagnosis, other race, married, and cRP were predisposed to a better OS or CSS while high PSA, high Gleason score, high T stage, N1, and multiple metastases were opposite (all p < 0.05).
The 311 PCa patients undergoing cRP were divided into two groups according LND status (No LND vs. LND) and the results are showed in Table 3. LND was mainly performed in young, low PSA, high Gleason score, high grade, high T stage, N1 diseases, and bone-only metastases (all p < 0.05).In multi analyses, LND including limited or extended LND was predisposed to a better OS (all p < 0.05) or CSS (all p < 0.05) in patients especially in T1/T2, N0, bone-only metastases, and no radiotherapy.Multivariable cox analyses predicting OS for the 311 patients were described in Table S3 and the findings of LND on CSS were shown in Table S4.Details of the multiple metastases group are shown in Table S5.In Figure S2, LND (limited or extended LND) was predisposed to a better OS or CSS (all p < 0.05).And patient confirmed negative LN by LND had a better OS or CSS compared with those harboring >3 positive LNs or no LND (all p < 0.05).

| DISCUSSION
In this study, cRP was only seen in 2.8% of the bmPCa from 2010 to 2019.Notably, urologists were increasingly willing to perform cRP for bmPCa patients (from 2.2% in 2010 to 3.0% in 2019), which indicated the increasing awareness of the therapeutic role of cRP for bmPCa patients.We observed that age, race, marital status, PSA, Gleason score, T/N stage, metastatic sites, and chemotherapy were significant factors contributing to the decision-making of cRP.Using multivariable cox regression analyses, we identified low age, recent years, other races, married, and cRP were prone to a higher OS and similar findings were observed for CSS.However, elevated PSA, worse Gleason score, advanced T stage, N1, and multiple metastases were related to a worse OS or CSS.Moreover, in multivariable analyses, cRP was related to a preferable survival in patients with age ≤ 64 or 65-74, PSA < 10 or 10-50 or 50-98 ng/mL, T2/3 stage, N0/1, Gleason 3+4/4+3/9-10/unknown, bone-only metastatic sites or patients not receiving chemotherapy.For the patients undergoing cRP, LND especially extended LND was associated with a preferable survival.These findings suggest cRP might be acceptable in bmPCa especially for young patients with low PSA, bone-only metastatic sites, and not receiving chemotherapy.Futhermore, LND especially extended LND should be considered when performing cRP.
There were researchers studying the role of cRP for bmPCa but a consensus seemed far off.Culp et al. reported that cRP or brachytherapy (n = 245) could bring a 5-year OS (67.4% vs. 52.6%)and CSS (75.8% vs. 61.3%)benefit for M1a-c PCa compared to no local treatment (n = 7811). 9Gratzke et al. reported a 5-year OS of 55% for PCa undergoing cRP compared to 21% for those treated with no surgical resection. 11A study of 11 No patients with oligometastatic PCa performed by Gandaglia et al. showed a 7-year CSS of 82% and cancer progression-free survival of 45%. 12 According to STAMPEDE trial 8 and TROMBONE 13 study, cRP might be an effective first step for PCa with a limited metastatic burden before a multidisciplinary approach.And our study showed patients receiving cRP had lower HR for both OS (HR 0.45, 95%CI 0.30-0.67)and CSS (HR 0.41, 95%CI 0.26-0.64),which are consistent with the aforementioned findings.However, an essential question in evaluating the feasibility of cRP for bmPCa is whether the detrimental effects can be offset by the benefits.Complications and functional outcomes 50-98 ng/mL 18 ( ≥ 98 ng/mL 17 ( are main concerns prior to surgery. 146][17] Thus, cRP might be a feasible and effective option for bmPCa.Whereas, it is still unclear which kind of patients may benefit from it.Rajwa et al. reviewed the recent prospective studies analyzing the survival of cRP in metastatic hormone-sensitive prostate cancer and concluded that cRP is effective and safe in selected patients. 18owever, they did not clarify the target population.Therefore, we performed the multivariable analyses stratified to the characteristics including age, PSA, T stage, N stage, Gleason score, metastatic sites, radiotherapy and chemotherapy status, and we observed that the OS or CSS benefit was evident in patients with age < 75, PSA < 98 ng/ mL, bone-only metastases or patients not receiving chemotherapy.Based on these findings, we speculated that cRP might benefit OS or CSS for young patients with low PSA, bone-only metastatic sites, and patients not receiving chemotherapy.The molecular biological mechanism supporting cRP's role remains being investigated.Increased expression of epithelial to mesenchymal transition (EMT) pathways were observed in the primary carcinoma at cRP despite 1-year chemohormonal therapy. 19Src signal might be minimized by removing the primary tumor. 20][23] Currently evidences are insufficient to confirm the value of LND when performing cRP on bmPCa patients.First we performed a multi-analysis for the 311 patients undergoing cRP and found LND especially extended LND may improve OS or CSS for them.Many attempts have been made to investigate the therapeutic role of LND but there is still no clear recommendation on this topic, [24][25][26] largely due to the insufficient detectability of the multiplicity of nodal metastatic drainage and the LND templates needing to be improved. 27,28ur study has limitations.The details such as the size and quantity of bone metastases were lacking.
We defined 1-12 LNs removed as limited LND and 12 or more LNs removed as extended LND as a proxy as the extent of LND lacks standardization in SEER.Furthermore, only 317 of the 11,271 patients received cRP, which needs larger sample prospective randomized clinical trials (RCT) to validate.Up to now there have been nine ongoing RCTs (SWOG, IP2-ATLANTA, SIMCAP, etc.) focusing on this topic, 29 and we are looking forward to their findings.
To summarize, cRP might benefit OS or CSS in young patients with low PSA and bone-only metastatic sites not receiving chemotherapy.And a clear OS or CSS benefit of LND especially extended LND was observed in patients undergoing cRP.

F I G U R E 3
Kaplan-Meier plots depicting OS and CSS after stratification according to cRP status when stratified to age (A), PSA (B), metastatic sites (C) or chemotherapy status (D) in 11,271 patients.cRP, cytoreductive radical prostatectomy; CSS, cancer-specific survival; OS, overall survival.| 16703 ZHAI et al.
Multivariable Cox regression analyses predicting overall survival and cancer-specific survival stratified by cytoreductive radical prostatectomy.
Characteristics for 311 prostate cancer patients with bone metastasis undergoing radical prostatectomy stratified by lymph node dissection.
T A B L E 3 This research was supported by Shenzhen High-level Hospital Construction Fund and Sanming Project of Note: T and N stage refer to AJCC Cancer Staging Manual, 7th ed.LND, lymph node dissection.Six patients of the 317 patients undergoing cRP were excluded due to the lack of LND information.LND, lymph node dissection.Fisher's exact test or chi-squared test, except.
a With percentages in parentheses.bc Student's t test.