During the last 30 years, there has been a major shift in initial staging in prostate cancer (CaP) in Western countries, with the incidence of metastases at diagnosis decreasing from over 50% in the 1970s to currently less than 10%. Yet, CaP is still the second cause of cancer death in men. We used two monthly curated databases of patients with castration-resistant prostate cancer (CRPC) to describe the natural history of patients dying of CaP in the modern era.
The outcome of 190 men with metastatic CRPC treated from 2008 to 2011 was studied. The characteristics of the patients who died from CaP (n=113 patients, 61%) were analyzed.
All 113 patients who died of CaP were assessable for the presence of metastases at diagnosis. Sixty-three patients (56%) had detectable metastases at diagnosis: 67%, 11% and 43% had bone, visceral and lymph node metastases, respectively. The median time to CRPC was 16 months and median overall survival (OS) was 5.2 years.
Among the patients with localized CaP at diagnosis (n=50, 44%), 46% had T stage⩾3 and 38% had a Gleason score⩾8. Overall, 64% of patients were classified as having a high-risk CaP. Only 26% who died from CaP had a Gleason score⩽6. Median OS was 8.8 years.
In the modern era, approximately half of the patients who die from CaP have metastases at diagnosis. The paradigm of progression from localized disease to metastasis and eventually death is only represented in the other half, although possible initial screening and staging errors ought to be taken into consideration. More efforts are needed to conduct trials in patients with newly diagnosed metastatic CaP.
The introduction of systematic PSA screening resulted in a significant rise in the incidence of prostate cancer (CaP) in Western societies. It peaked in the early 2000s, then declined after the initial enthusiasm given the side effects of over-diagnosis and overtreatment, and is currently relatively stable with moderate fluctuation.1 As a result, there has been a major shift in initial staging for men with CaP in Western countries, with the incidence of detectable metastases at diagnosis decreasing from more than 50% of the cases in the 1970s to currently less than 10%.2, 3, 4
Despite the changes in the epidemiological profile of CaP in Western countries with an increased percentage of early-stage, low-risk disease at diagnosis, the survival benefit of PSA-based screening remains controversial, because this early detection and treatment of CaPs, which do not threaten life expectancy, leads to unnecessary side effects, impaired quality of life and health expenses. Updated survival data from the largest European PSA screening randomized trial suggest a relative reduction of 21% in the risk of death from CaP, increasing to 29% after adjustment for selection bias and non-compliance (but without a benefit in overall survival (OS)); on the other hand, screening of over 1000 men is necessary to detect 37 cancers and prevent one death from CaP.3
Accounting for approximately 9% of total estimated cancer deaths, CaP remains the second cause of cancer-related mortality in men in the US,1 and the third cause in Europe.5 CaP-specific mortality is approximately 12% in the US 1 and 17% in Europe,5 although percentages as high as 35% are observed in specific populations such as Sweden.6,7 It is therefore important to better characterize the profile of patients who will eventually die of the disease, and accurately distinguish them from the majority of patients with indolent, slow-growing disease for whom death will most likely be attributed to comorbidities.
This study aimed to describe the characteristics and natural history of patients dying of CaP in the modern era.
Materials and methods
The outcome of patients with metastatic castration-resistant CaP (CRPC) treated in clinical trials from 2008 to 2011 in two institutions in France (Institut Gustave Roussy, Villejuif, and Institut Jean-Godinot, Reims) was studied in a retrospective pooled analysis. Patients were defined as metastatic if they had radiologically and/or histologically detected regional lymph node metastasis (N+) and/or distant metastasis (M+). A total of 190 patients with metastatic CRPC were identified. After a median follow-up of 6.8 years from initial diagnosis, 113 patients had died of CaP and were assessable for the presence of metastasis at diagnosis (Figure 1).
The characteristics of patients who died from CaP were analyzed. Patient records were accessed through hospital electronic databases. Data collection pertained to demographics, initial PSA values, Gleason score and staging data, risk groups, primary treatment modalities, duration of response to androgen-deprivation therapy (ADT), metastatic sites, docetaxel treatment (duration, best response, clinical benefit, time to progression (TTP)), pre- and post-docetaxel treatment lines, terminal PSA values (last available PSA prior to death) and OS. CRPC was defined as a progressive CaP despite castrate levels of testosterone. Progression was defined according to the PCWG2 criteria.8 Risk groups for patients with localized disease were defined according to the D’Amico classification.9 Objective responses were measured according to RECIST criteria. Clinical benefit was defined as a decrease in the use of analgesics and improvement of the performance status. The interval from the diagnosis to the last available follow-up or death was used to calculate OS. Information on the completeness of data is provided in Tables 1 and 2.
Statistical analysis was carried out using the Prism statistical software, version 6 (GraphPad Software). Descriptive statistical methods were used. OS and TTP post docetaxel were calculated using Kaplan–Meier estimates and compared using a log-rank test.
Of the 113 patients who died of CaP, 63 patients (55.7%) had detectable metastases at diagnosis: 42 (67%), 27 (43%) and 7 patients (11%) had bone, lymph node and visceral metastases, respectively. Overall, nine patients (14%) had multiple metastatic sites. The majority of tumors (89%) had a Gleason score of 7 or above. The median PSA at diagnosis was 81 ng ml−l (95% confidence interval (CI) 45–150 ng ml−1), with the vast majority of patients diagnosed with a PSA over 20 ng ml−1. No locoregional treatment was performed in 59% of the patients with de novo metastatic CaP, whereas data were not available for 19% of patients. Only 8% and 14% had surgery and radiotherapy, respectively (five patients underwent radical prostatectomy, combined with a pelvic lymph node dissection in two cases). One patient had both surgery and radiotherapy. The median time to ADT initiation was 0.6 months. The median time to CRPC was 16 months (95% CI 13–19 months). The median number of docetaxel cycles was 8, and the median TTP post docetaxel was 7 months. Median PSA at docetaxel initiation was 145 ng ml−1 and median nadir PSA post docetaxel was 27.35 ng ml−1. A PSA reduction of >50% was seen in 77.5% of patients, while a >30% reduction was seen in 84.5%. A nadir PSA of <4 ng ml−1 was seen in 18.9% of patients, while no patient had a nadir PSA of <0.2 ng ml−1. A clinical benefit from docetaxel treatment was seen in 55 patients (87%), while the median number of post-docetaxel treatment lines was 1 (range: 1–4). Median terminal PSA was 455 ng ml−1 (range: 20–29 620 ng ml−1). Median OS was 62 months (5.2 years) (95% CI: 53–79 months) (Table 1).
Among patients with localized CaP at diagnosis (n=50; 45%), 23 patients (46%) had a T stage⩾3 and 9 patients (18%) had a T stage ⩽2, whereas precise T staging data were not available for a significant proportion of patients (36%). This can be attributed to the fact that the initial diagnosis and management had not been performed in the study centers for a large number of the localized CaP cohort; these patients were in fact referred to our centers after entering the CRPC stage. Nevertheless, specific M staging information was available for all the patients for whom the T stage was not available, and specific N staging information was available for all but two of them. Approximately two-thirds of the patients (32 patients, 64%) had a Gleason score⩾7, while 13 patients (26%) had a Gleason score⩽6, and for 5 patients (10%), the Gleason score was not available. The median PSA at diagnosis was 16 ng ml−1 (95% CI 10.9–24 ng ml−1). Overall, 32 patients (64%) had high-risk CaP, and 8% of patients had a low-risk tumor at diagnosis. Only 4% of the patients did not receive any locoregional treatment, while 40%, 60% and 2% underwent surgery, radiotherapy and brachytherapy, respectively. The median time to ADT initiation was 23.1 months. The median time to CRPC was 19 months (95% CI: 12–32 months). The median number of docetaxel cycles was 7, and the median post-docetaxel TTP was 7 months (95% CI: 5.9–8 months). The median PSA at docetaxel initiation was 78 ng ml−1, and the median post-docetaxel PSA nadir was 14.4 ng ml−1. A PSA reduction of >50% was seen in 65.7% of the informative patients, while a >30% reduction was observed in 74.3% of patients. A nadir PSA of <4 ng ml−1 was seen in 14.3% of informative patients, while no patient had a nadir PSA of <0.2 ng ml−1. A clinical benefit from docetaxel therapy was seen in 36 patients (72%). An average of two post-docetaxel lines was administered (range: 1–5). Median terminal PSA was 1540 ng ml−1 (range: 55–14 960 ng ml−1). Median time from local treatment to diagnosis of metastatic disease was 82 months (range: 29–186 months). Median OS was 102 months (8.5 years) (95% CI: 76–118 months) (Table 2).
As expected, patients with localized disease at diagnosis had a better OS than those with de novo metastatic cancer (P=0.0001) (Figure 2). There was no difference in the age at diagnosis, but there were expected differences in the initial PSA level (P=0.03) and the percentage of tumors with a low Gleason score, reflecting a more aggressive biology and heavier associated tumor burden. A borderline difference was found in time to CRPC between the two groups (P=0.06). There was no significant difference in the post-docetaxel TTP (P=0.695) or the post-docetaxel nadir PSA (P=0.3679), despite a significantly lower pre-docetaxel PSA level in the non-metastatic tumors.
Despite effective initial management and improved therapeutic options over recent years, patients with relapsed CaP under to long-term castration will inevitably reach the castration-resistance phase, develop metastatic disease and eventually die of cancer progression, unless death due to an unrelated cause occurs. The need to identify these patients early, if possible at diagnosis, is underlined by the finding of one of the largest screening trials, wherein the majority (74%) of deaths from CaP detected by screening occurred in men whose cancer was diagnosed at their first PSA screening.3 This indicates that the disease that will eventually be lethal is aggressive from the very beginning. It also possibly explains why the effect of screening does not further increase during extended follow-up but that, conversely, the rate of CaP deaths even from initially localized cancers may increase during a follow-up period of 15–20 years.3,10
One of the interesting features of our analysis is the high percentage (50%) of patients with de novo metastatic CaP in this cohort of patients who died of the disease. Data from a large UK CaP registry also evidenced that 44.5% of CaP-specific deaths were patients with upfront metastatic disease, despite the fact that the latter corresponded to only 14% of the overall registry.4 It is noteworthy, however, that information for stage at diagnosis was not available for 23% of CaP deaths in the UK study. A similar caveat could be put forward regarding our study: men who present with de novo metastatic disease may have been ineffectively screened, and might have presented with curable disease had they been diagnosed earlier. Conversely, it could be argued that our cohort of localized CaP might, owing to inefficient screening, harbor patients with under-staged, undetected de novo metastatic cancer. Indeed, the fact that there was no significant difference in the time to CRPC between the two cohorts further raises this question. However, the median age at diagnosis did not differ between the two cohorts, which are therefore likely to have been stochastically submitted to the same level of ineffective screening. Furthermore, the differences in the PSA level at diagnosis, time to ADT initiation and time to docetaxel use (which is usually a hallmark of symptomatic metastatic CRPC) reflect the difference in the initial tumor burden and metastatic status of the two cohorts.
The analysis period of this study ended just prior to the widely available applications of newer treatment options, notably abiraterone and enzalutamide. Conversely, however, all study patients had participated in clinical trials, the significant majority involving these two drugs. The fact that these are clinical trial population patients could also be interpreted as a study limitation, in the sense that it may not be representative of the community non-trial setting (at least not for the time period the analysis pertains to) or the developing world setting. On another issue, the Gleason grading system was modified in 2005 and 2010, and this fact ought to be taken in consideration in the interpretation of our results, especially in regard with Gleason 6 tumors.
The overall percentage of patients with de novo metastatic CaP in the literature ranges from less than 10% to a little over 20% in screening studies and population-based registries.2, 3, 4 In the pool of patients who eventually die of the disease, this percentage seems to increase twofold or even threefold. What our results therefore indicate is that the standard notion of an initially indolent disease slowly progressing to the metastatic phase and death might be true only for half of the patients who die of CaP.
It is obvious that a greater effort is needed for this subset of patients with upfront metastatic disease, who probably merit different management strategies. One such direction could be to administer chemotherapy earlier, notably before reaching the CRPC stage. The early use of regimens other than docetaxel has as yet not been proven of benefit in de novo metastatic patients in terms of time to CRPC progression or in survival.11,12
The only randomized phase 3 study with available results on early docetaxel therapy for metastatic non-castrate CaP is the GETUG 15 trial.13 Over two-thirds of the patients in both arms were metastatic at diagnosis. Although clinical (23.5 vs 15.4 months, P=0.015) and biochemical progression-free survival (22.9 vs 12.9 months, P=0.005) were significantly longer with docetaxel, the combination of ADT/docetaxel did not significantly increase OS compared with ADT alone (58.9 vs 54.2 months, P=0.955). Similar trials testing docetaxel (STAMPEDE and CHAARTED-ECOG E3805 trials) 14,15 are conducted in the US and the UK. The results of the phase 3 CHAARTED trial were very recently reported: with a median follow-up of 29 months, docetaxel improved OS compared with ADT alone (52.7 vs 42.3 months, P=0.0006, hazard ratio 0.63 (95% CI: 0.48, 0.82)).15 This benefit was mainly seen in the high-volume disease group, as there was no survival difference in the low-volume disease patients. Indeed, the fact that this was a larger study and included more high-volume disease patients, combined with the fact that less patients (31%) in the ADT-alone group received docetaxel upon progression compared with the GETUG 15 trial (62%) might be accountable for the difference in OS outcome between the two studies
An intriguing option that emerged is the early addition of novel endocrine therapies. An ongoing phase III European trial (PEACE-1)16 is designed to address this issue in metastatic hormone-naïve patients, by testing combinations of ADT, local radiotherapy and abiraterone acetate in four arms. The STAMPEDE trial also included an additional abiraterone arm to its complex design,14 while the pivotal Latitude phase III trial is testing the addition of abiraterone to ADT in patients with de novo metastatic CaP and adverse features.17 A phase 2 trial is similarly evaluating neoadjuvant enzalutamide with ADT in intermediate and high-risk patients.18
Perhaps the most controversial approach for upfront metastatic patients would be to treat the primary tumor itself. Currently, locoregional approaches are ignored in CaP in the presence of metastatic disease. This attitude is based on the commonly held notion that the prognosis is influenced by the presence of metastasis, and as such employing local treatment is deemed useless or futile in this context. A significant tumor burden is thus left in place that systemic treatments are called upon to address. Locoregional treatment, however, especially surgery, would serve as an important immediate cytoreductive approach, not allowing tumor cells the time to acquire or activate adaptive mechanisms. In addition, primary tumors are known to be responsible for shedding disseminated cancer cells even at very early stages, associated with poor prognosis and later development of metastasis,19, 20, 21 even after long periods of tumor dormancy.22 This effect is amplified in the majority of metastatic patients who currently do not receive any locoregional treatment at all, but are instead submitted to long-term ADT, which, despite an initial response, will inevitably lead to castration-resistance and progression.
Evidence from other solid tumors, such as breast, colorectal and renal cancer have educated us on the importance of treating the primary tumor and even of resecting carefully selected metastases in the context of multimodality treatment. In CaP, although the benefit of administering ADT to patients with pN1 disease has been clearly suggested, 23 studies on clinically N+ disease treated with a combination of surgery and ADT are limited and retrospective; they do, however, point in the same direction in terms of survival benefit.24,25 Regarding locoregional treatment in patients with distant metastasis at diagnosis, no real evidence exists.
In parallel, efforts are needed to reinforce our ability to appropriately treat those patients with localized CaP who are likely to die of the disease, possibly using modalities other than extended surgery and conventional adjuvant modalities. The GETUG 12 trial addressed the impact of neoadjuvant docetaxel in addition to ADT in localized high-risk patients including 25% with node-positive disease. The reported results on response, tolerance and quality of life are encouraging,26 while progression-free survival data are pending. The ongoing European PEACE 2 phase III trial is investigating the addition of cabazitaxel to pelvic radiotherapy and hormonotherapy in the same population.27
Newer endocrine therapies have also entered the arena of localized high-risk CaP. Recent studies on neoadjuvant abiraterone acetate reported promising biochemical and pathological response rates.28,29 Such results argue in favor of testing combinations of neoadjuvant abiraterone in phase III trials.
Our results indicate that approximately half of the patients who die of CaP have detectable metastases at diagnosis, although initial screening and staging errors ought to be taken into consideration. Further evidence is clearly needed to confirm these results. Similarly, large randomized trials are warranted to determine the appropriate management of patients with de novo metastatic CaP, including the evaluation of combined of systemic and locoregional modalities.
Siegel R, Naishadham N, Jemal A . Cancer Statistics, 2012. CA Cancer J Clin 2012; 62: 10–29.
Postma R, van Leenders A, Roobol MJ, Schröder FH, van der Kwast TH . Tumor features in the control and screening arm of a randomized trial of prostate cancer. Eur Urol 2006; 50: 70–75.
Schröder FH, Hugosson J, Roobol MJ, Tammela TLJ, Ciatto S, Nelen V et al. Prostate-cancer mortality at 11 years of follow-up. N Engl J Med 2012; 366: 981–990.
Chowdhury S, Robinson D, Cahill DD, Rodriguez-Vida A, Holmberg L, Møller H . Causes of death in men with prostate cancer: an analysis of 50 000 men from the Thames Cancer Registry. BJU Int 2013; 112: 182–189.
Ferlay J, Steliarova-Foucher E, Lortet-Tieulent J, Rosso S, Coebergh JWW, Comber H et al. Cancer incidence and mortality patterns in Europe: Estimates for 40 countries in 2012. Eur J Cancer 2013; 49: 1374–1403.
Riihimaki M, Thomsen H, Brandt A, Sundquist J, Hemminki K . What do prostate cancer patients die of? Oncologist 2011; 16: 175–181.
Epstein M, Edgren G, Rider JR, Mucci LA, Adami HO . Temporal trends in cause of death among Swedish and US men with prostate cancer. J Natl Cancer Inst 2012; 104: 1335–1342.
Scher HI, Halabi S, Tannock I, Morris M, Sternberg CN, Carducci MA et al. Design and end points of clinical trials for patients with progressive prostate cancer and castrate levels of testosterone: recommendations of the Prostate Cancer Clinical Trials Working Group. J Clin Oncol 2008; 26: 1148–1159.
D’Amico AV, Whittington R, Malkowicz SB, Schultz D, Blank K, Broderick GA et al. Biochemical outcome after radical prostatectomy, external beam radiation therapy, or interstitial radiation therapy for clinically localized prostate cancer. JAMA 1998; 280: 969–974.
Johansson JE, Andrén O, Andersson SO, Dickman PW, Holmberg L, Magnuson A et al. Natural history of early, localized prostate cancer. JAMA 2004; 291: 2713–2719.
Millikan RE, Wen S, Pagliaro LC, Brown MA, Moomey B, Do KA et al. Phase 3 trial of androgen ablation with or without three cycles of systemic chemotherapy for advanced prostate cancer. J Clin Oncol 2008; 26: 5936–5942.
Wang J, Halford S, Rigg A, Roylance R, Lynch M, Waxman J . Adjuvant mitoxantrone chemotherapy in advanced prostate cancer. BJU Int 2000; 86: 675–680.
Gravis G, Fizazi K, Joly F, Oudard S, Priou F, Estemi B et al. Androgen-deprivation therapy alone or with docetaxel in non-castrate metastatic prostate cancer (GETUG-AFU 15): a randomized, open-label, phase 3 trial. Lancet Oncol 2013; 14: 149–158.
NLM Identifier NCT00268476. STAMPEDE: Systemic Therapy in Advancing or Metastatic Prostate Cancer: Evaluation of Drug Efficacy: A Multi-Stage Multi-Arm Randomized Controlled Trial.
Sweeney C, Chen YH, Carducci MA, Jarrrad DF, Eisenberg MA, Wong YN et al. Impact on overall survival (OS) with chemohormonal therapy versus hormonal therapy for hormone-sensitive newly metastatic prostate cancer (mPrCa): An ECOG-led phase III randomized trial. J Clin Oncol 2014; 32: 5 s (suppl; abstr LBA2).
NLM Identifier NCT01957436. A prospective randomized phase III study of androgen deprivation therapy with or without local radiotherapy with or without abiraterone acetate and prednisone in patients with metastatic hormone-naïve prostate cancer.
NLM Identifier NCT01715285. A study of abiraterone plus low-dose prednisone plus androgen-deprivation therapy (ADT) versus ADT alone in newly-diagnosed patients with high-risk, metastatic hormone-naïve prostate cancer (mHNPC).
Montgomery RB, Joshua A, Hannah AL, Peterson AC, Lopez C, Gleave ME et al. A randomized, open-label, phase II study of MDV3100 alone or in combination with leuprolide and dutasteride as neoadjuvant therapy to prostatectomy in intermediate and high-risk prostate cancer. J Clin Oncol 2012; 30 (suppl; abstr TPS4695).
Haffner MC, Mosbruger T, Esopi DM, Fedor H, Heaphy CM, Walker DA et al. Tracking the clonal origin of lethal prostate cancer. J Clin Invest 2013; 123: 4918–4922.
Kruck S, Gakis G, Stenzl A . Circulating and disseminated tumor cells in the management of advanced prostate cancer. Adv Urol 2012; 2012: 135281.
Weckermann D, Polzer B, Ragg T, Blana A, Schlimok G, Arnholdt H et al. Perioperative activation of disseminated tumor cell in bone marrow of patients with prostate cancer. J Clin Oncol 2009; 27: 1549–1556.
Osisami M, Keller ET . Mechanisms of metastatic tumor dormancy. J Clin Med 2013; 2: 136–150.
Messing EM, Manola J, Yao J, Kiernan M, Crawford D, Wilding G et al. Immediate versus deferred androgen deprivation treatment in patients with node-positive prostate cancer after radical prostatectomy and pelvic lymphadenectomy. Lancet Oncol 2006; 7: 472–479.
Grimm MO, Kamphausen S, Hugenschmidt H, Stephan-Odenthal M, Ackermann R, Vögeli TA . Clinical outcome of patients with lymph node-postive cancer after radical prostatectomy versus androgen deprivation. Eur Urol 2002; 41: 628–634.
Gjertson CK, Asher KP, Sclar JD, Goluboff ET, Olsson CA, Benson MC et al. Local control and long-term disease-free survival for stage D1 (T2-T4N1-N2M0) prostate cancer after radical prostatectomy in the PSA Era. Urology 2007; 70: 723–727.
Fizazi K, Lesaunier F, Delva R, Gravis G, Rolland F, Priou F et al. A phase III trial of docetaxel-estramustine in high-risk localized prostate cancer: a planned analysis of response, toxicity and quality of life in the GETUG 12 trial. Eur J Cancer 2012; 18: 209–217.
NLM Identifier NCT01952223. A Phase III of Cabazitaxel and Pelvic Radiotherapy in Localized Prostate Cancer and High-risk Features of Relapse (PEACE2).
Taplin M, Montgomery RB, Logothetis C, Bubley GJ, Richie JP, Dalkin BL et al. Effect of neoadjuvant abiraterone acetate (AA) plus leuprolide acetate (LHRHa) on PSA, pathological complete response (pCR), and near pCR in localized high-risk prostate cancer (LHRPC): Results of a randomized phase II study. J Clin Oncol 2012; 30 (suppl; abstr 4521).
Efstathiou E, Davis JW, Troncoso P, Titus MA, Hoang A, Wen S et al. Cytoreduction and androgen signaling modulation by abiraterone acetate (AA) plus leuprolide acetate (LHRHa) versus LHRHa in localized high-risk prostate cancer (PCa): preliminary results of a randomized preoperative study. J Clin Oncol 2012; 30 (suppl; abstr 4556).
The authors declare no conflict of interest.
About this article
Cite this article
Patrikidou, A., Loriot, Y., Eymard, JC. et al. Who dies from prostate cancer?. Prostate Cancer Prostatic Dis 17, 348–352 (2014). https://doi.org/10.1038/pcan.2014.35
Endocrine consequences of treatment with the new androgen receptor axis-targeted agents for advanced prostate cancer
Efficacy and safety of abiraterone acetate plus prednisone in Japanese patients with newly diagnosed, metastatic hormone-naive prostate cancer: final subgroup analysis of LATITUDE, a randomized, double-blind, placebo-controlled, phase 3 study
Japanese Journal of Clinical Oncology (2020)
Glycoconjugate Journal (2020)
Performance of clinicopathologic models in men with high risk localized prostate cancer: impact of a 22-gene genomic classifier
Prostate Cancer and Prostatic Diseases (2020)