NATURE REVIEWS | UROLOGY VOLUME 10 | DECEMBER 2013 | 687

NEWS & VIEWS

secondary therapies before they develop metastases, or prevent the premature dis-continuation of treatments that might slow the rate of increase of PSA.7 The use of PSA kinetics would also take into account the fact that immunotherapies, such as s ipuleucel- T and PSA-TRICOM, can provide clinical ben-efits without inducing absolute PSA declines, with no significant accompany ing toxi city.8 For example, PSA kinetics evaluated at 3 months or beyond might reveal an abso-lute increase in the PSA value, but they might also demonstrate a slowing of the PSA kinet-ics, perhaps indicating a decrease in tumour growth-rate. Ultimately, randomized studies are needed to prospectively evaluate the association between changes in PSA kinet-ics and harder clinical outcomes, such as MFS, overall survival, and the development of symptomatic disease.

PSA kinetics assessment is just one tool for assessing clinical benefit in men with nmPC. Modern imaging technologies have greatly improved the detection of metasta-ses to bone, an important development given that the vast majority of patients with nmPC will ultimately develop osseous metastatic disease. In preliminary studies, a sodium fluoride (NaF) PET scan has been shown to detect occult osseous lesions in men with nmPC and biochemical recur-rence.9 Such imaging studies could be used to determine clinical benefit that could serve as an intermediate end point in nmPC clinical trials. However, such trials would ultimately require confirmed correlations between changes in novel imaging results and MFS or overall survival. Although NaF PET imaging has been approved for the evaluation of meta static prostate cancer, care should be taken not to overinterpret findings or to use a positive NaF PET result to automatically trigger aggressive, toxic therapies, such as chemotherapy, until more is known about how the lead time for this imaging modality c ompares with c onventional imaging techniques.

Recent and ongoing studies of modern therapies present many opportunities for evaluating PSA kinetics, and possibly NaF PET imaging, as indicators of MFS and predictors of overall survival. Trials of enzalutamide (NCT01664923), abira-terone (NCT00473512, NCT01751451, and NCT01314118), and sipuleucel-T10 in men with nmPC have been completed or are

ongoing. Data gleaned from these studies could further support the use of PSA kinet-ics, NaF PET imaging, or conventional MFS as intermediate clinical trial end points.

In recent years, therapy for prostate cancer has advanced significantly, giving patients with metastatic disease many promising treatment options. The challenge ahead is to develop intermediate clinical trial end points, beyond conventional measures of MFS and overall survival, enabling practitioners to deploy these new therapies earlier in the disease process. Clinical benefit might be even greater in patients with nmPC, whose tumour burden is smaller than those with metastatic disease. Now that the pieces are in place to help patients with nmPC, we must agree on the rules of the game as we move forward, and acknowledge that the game is indeed changing.

Laboratory of Tumor Immunology and Biology, Center for Cancer Research, National Cancer Institute, 10 Center Drive, Room 8B09, Bethesda, MD 20892, USA (R. A. Madan, J. L. Gulley). Correspondence to: J. L. Gulley gulleyj@mail.nih.gov

Competing interestsThe authors declare no competing interests.

1. Schweizer, M. T. et al. Metastasis-free survival is associated with overall survival in men with PSA-recurrent prostate cancer treated with deferred androgen deprivation therapy. Ann. Oncol. http://dx.doi.org/10.1093/annonc/mdt335.

2. Smith, M. R. et al. Denosumab and bone-metastasis-free survival in men with

‘‘…evaluating PSA kinetics …might increase the utility of PSA measurement’’

castration-resistant prostate cancer: results of a phase 3, randomised, placebo-controlled trial. Lancet 379, 39–46 (2012).

3. Sonpavde, G. & Palapattu, G. S. Neoadjuvant therapy preceding prostatectomy for prostate cancer: rationale and current trials. Expert Rev. Anticancer Ther. 10, 439–450 (2010).

4. Eastham, J. A., Kelly, W. K., Grossfeld, G. D., Small, E. J. & Cancer & Leukemia Group B. Cancer and Leukemia Group B (CALGB) 90203: a randomized phase 3 study of radical prostatectomy alone versus estramustine and docetaxel before radical prostatectomy for patients with high-risk localized disease. Urology 62 (Suppl. 1), 55–62 (2003).

5. Antonarakis, E. S. et al. Changes in PSA kinetics predict metastasis-free survival in men with PSA-recurrent prostate cancer treated with nonhormonal agents: combined analysis of 4 phase II trials. Cancer 118, 1533–1542 (2012).

6. Stein, W. D. et al. Tumor regression and growth rates determined in five intramural NCI prostate cancer trials: the growth rate constant as an indicator of therapeutic efficacy. Clin. Cancer Res. 17, 907–917 (2011).

7. Bubley, G. J. et al. Eligibility and response guidelines for phase II clinical trials in androgen-independent prostate cancer: recommendations from the Prostate-Specific Antigen Working Group. J. Clin. Oncol. 17, 3461–3467 (1999).

8. Madan, R. A., Gulley, J. L., Fojo, T. & Dahut, W. L. Therapeutic cancer vaccines in prostate cancer: the paradox of improved survival without changes in time to progression. Oncologist 15, 969–975 (2010).

9. Jadvar, H. et al. Prospective evaluation of 18F-NaF and 18F-FDG PET/CT in detection of occult metastatic disease in biochemical recurrence of prostate cancer. Clin. Nucl. Med. 37, 637–643 (2012).

10. Beer, T. M. et al. Randomized trial of autologous cellular immunotherapy with sipuleucel-T in androgen-dependent prostate cancer. Clin. Cancer Res. 17, 4558–4567 (2011).

KIDNEY CANCER

Characterizing late recurrence of renal cell carcinomaVincenzo Ficarra and Giacomo Novara

Very few studies have followed patients with renal cell carcinoma for long enough to characterize late relapses. Newly published research suggests that late‑recurring tumours are more likely to be of low nuclear grade and show a better response to first‑line targeted therapy than early relapsing cancers.Ficarra, V. & Novara, G. Nat. Rev. Urol. 10, 687–689 (2013); published online 29 October 2013; doi:10.1038/nrurol.2013.239

Over the past decade, several multicentre international studies have demonstrated that approximately 20–40% of patients who undergo partial or radical nephrectomy for clinically localized renal cell carci-noma (RCC) subsequently develop local or distant recurrence.1,2 Currently, a 5-year

threshold is most commonly used to distin-guish early from late recurrences; although most relapses occur within the first 5 years after surgery, around 10% of local or distant recurrences develop later. Considering the low rate of late relapses, very few retro-spective small series or cases reports of

© 2013 Macmillan Publishers Limited. All rights reserved

688 | DECEMBER 2013 | VOLUME 10 www.nature.com/nrurol

NEWS & VIEWS

such patients have been published to date. Thus, further information is needed regard-ing the clinical and pathological features that predict late relapse, the pattern of late metastases, the response of patients who experience late relapse to targeted thera-pies, and their survival outcomes. As such, multicentre collaborative analyses of large numbers of patients who have undergone partial or radical nephrectomy for RCC and are followed for a sufficient period to character ize late relapses are welcome a dditions to the literature.

In one such recently published multi centre Canadian study, Kroeger and colleagues3 retrospectively evaluated survival outcomes and treatment responses to targeted thera-pies for a series of 1,212 patients with early (≤5 years) or late (>5 years) recurrence after nephrectomy for clinically localized RCC. Between August 2008 and February 2013, 315 patients (26%) experience d late recurrence. Patients who experi enced late relapse were generally younger and their tumours were more likely to display clear-cell histo logy than those who experi-

enced early relapse. Moreover, tumours that underwent late recurrence were character-ized by lower nuclear grade and lower prevalence of sarcoma toid features. Finally, patients with late relapse were more likely to harbour metastases at atypical sites—such as the adrenal gland, soft tissue, the pancreas, and the thyroid—and carry a higher meta-static burden. The most frequent metastatic sites for late relapses were lung (60.3%), lymph nodes (39.4%), bone (28.9%), and liver (19.9%).

Treatment response to first-line targeted therapy was significantly better in patients

with late relapse than in those with early relapse. To be specific, progressive disease after first-line therapy was documented in just 12% of patients with late recur-rence, but in 20.5% of those with early relapse (P = 0.004). No differences were observed when response to second-line therapy was evaluated. Specifically, pro-gressive disease after second-line therapy was documented in 36.1% of patients with early relapse and in 34.8% of those with late relapse. Importantly, univariable analyses showed better median progression-free sur-vival (PFS) and overall survival in patients with late relapse than in those with early recurrence; median PFS was 10.7 months in patients with late relapse compared with 8.5 months in those with early relapse (P = 0.005), and median overall survival was 34 months and 27.4 months, respectively (P = 0.004). However, multivariable analyses adjusting for the International Metastatic RCC Database Consortium (IMDC) risk factors failed to confirm the indepen-dent predictive role of late relapses for either PFS or overall survival. Conversely, anaemia, Karnofsky Performance score <80, thrombophilia, and time from diagnosis to treatment of <1 year were shown to be indepen dent predictors of survival outcome in patients with late relapses. Thus, the study authors concluded that patients with late relapse have better treatment response and survival outcomes than patients with early relapse owing to more favourable p rognostic features.

In a previous international study, Brookman-May et al.4 identified lympho-vascular invasion, Fuhrman nuclear grade 3–4, and pathological tumour (pT)

stage >1 as independent predictors of late recurrence. This multicentre CORONA/SATURN project study included more than 13,000 patients who underwent partial or radical nephrectomy for RCC at 28 urology departments worldwide over a period of 19 years. The investigators proposed the PRELANE score—whereby two points are awarded for the presence of lymphovascular invasion, another two points are added for pT stage >1, and Fuhrman grade 3–4 adds a further point—to stratify patients without metastases into three different risk categor-ies, namely low risk (score 0), inter mediate risk (score 1–3), and high risk (score 4–5), for the development of late distant recur-rence. With a predictive accuracy of 70%, the prevalence of late recurrence was 3.1%, 8.4%, and 22.1% for patients in the low-risk, intermediate-risk, and high-risk categor-ies, respectively. These data suggest that patients classified as intermediate-risk or high-risk should be followed for >5 years after surgery in order to detect and treat late recurrence as soon as possible. Interestingly, the majority of distant metastases can be detected using the imaging techniques routinely included in the scheduled follow-up appointments (for example, chest and abdominal CT).

Notably, patients with late relapse in the Canadian study3 showed different clini-cal and pathological features to patients in the CORONA/SATURN study.4 For example, patients in the Canadian study were younger, more likely to have cancer with clear-cell histology and higher nuclear grade, and, most importantly, at a signifi-cantly greater risk of developing multiple metastases. Unfortunately, however, a lot of relevant pathological information was missing from the Canadian study. On the other hand, laboratory parameters and performance status were lacking in the CORONA/SATURN study (Table 1). This discrepancy might be explained by the fact that urologists (such as the investigators of the CORONA/SATURN study) are more focused on the initial character istics of the primary tumour, whereas oncologists (such as the clinicians who led the Canadian IMDC study) concentrate on laboratory f eatures at the time of recurrence detection.

‘‘Treatment response to first-line targeted therapy was significantly better in patients with late relapse…’’

Table 1 | Patients with late relapses described in the CORONA/SATURN4 and IMDC3 studies

Study parameter CORONA/SATURN study IMDC study

Study period 1992–2010 2008–2013

Study population (n) 310 315

Median age at surgery (years) 61.8 53.1

Female gender (%) 41.6 25.9

Partial nephrectomy (%) 16.8 NR

Clear-cell histology (%) 83.5 93.4

Median tumour size (cm) 6 NR

Fuhrman grade 3–4 (%) 27.4 38.4

Lymphovascular invasion (%) 29 NR

pT >2 (%) 43.5 NR

pN+ (%) 1 NR

Multiple metastases (%) 37.1 81.2

Abbreviations: IMDC, International Metastatic RCC Database Consortium; NR, not reported.

© 2013 Macmillan Publishers Limited. All rights reserved

NATURE REVIEWS | UROLOGY VOLUME 10 | DECEMBER 2013 | 689

NEWS & VIEWS

Although the IMDC study is similar to the CORONA/SATURN study with respect to its limitations (including its retro spective design, noncentralized imaging and patho-logy review, missing information on metas-tasectomy, and nonstandardized follow-up protocols), an important difference between the studies by Brookman-May et al.4 and Kroeger et al.3 relates to the use of targeted therapy. The majority of patients with late relapses included in the CORONA/SATURN study were treated before the tar-geted therapy era, whereas all patients in the Canadian study received targeted therapy.

Although the currently available data are from series with heterogeneous character-istics, studies such as CORONA/SATURN and IMDC provide valuable informa-tion regarding predictors of response to modern targeted therapies and survival outcomes for patients with localized RCC. However, further prospective multicentre inter national studies are needed to iden-tify the predictors of response to medical therapies, as well as predictors of survival, using prognostic models that include both pathological and laboratory parameters. Moreover, further studies are needed to better identify the most appropriate thresh-old for distinguish ing between early and late recurrences. For example, a recent

multicentre study demonstrated that the time to recurrence is a reliable predictor of cancer-specifi c survi val in patients with RCC relapse, identifying 48 months as the best threshold for defining early and late recurrences.5 Further studies should test whether this cut-off value is better than 60 months, which is the threshold currently adopted in the majority of series published in the literature.

In conclusion, recent data confirm that approximately 10% of patients with RCC experience late relapse. Pathological f eatures—such as nuclear grade, local exten-sion of primary tumour, and presence of lymphovascular invasion—can help us to identify patients to follow up for >5 years, in order to detect late recurrences as soon as possible. Chest, abdominal, and bone imaging techniques usually planned during the follow-up period are able to identify the majority of late recurrences. Patients with late relapses showed a good response to first-line treatment with the available tar-geted therapies and seem to survive longer than patients who develop early recur-rences after nephrectomy. Molecular and cyto genetic features of late relapsing RCC should be investigated to better under-stand the biology of this special subgroup of patients.

Department of Experimental and Clinical Medical Sciences, Urology Unit, University of Udine, Piazzale Santa Maria della Misericordia 15, Clinica di Urologia, Padiglione 5, 33100 Udine, Italy (V. Ficarra). Department of Surgery, Oncology and Gastroenterology, Urologic Clinic, University of Padua, Via Giustiniani 2, 35100 Padua, Italy (G. Novara). Correspondence to: V. Ficarra vincenzo.ficarra@unipd.it

Competing interestsThe authors declare no competing interests.

1. Ficarra, V. et al. Prognostic and therapeutic impact of the histopathologic definition of parenchymal epithelial renal tumors. Eur. Urol. 58, 655–668 (2010).

2. Sun, M. et al. Prognostic factors and predictive models in renal cell carcinoma: a contemporary review. Eur. Urol. 60, 644–661 (2011).

3. Kroeger, N. et al. Survival outcome and treatment response of patients with late relapse from renal cell carcinoma in the era of targeted therapy. Eur. Urol. http://dx.doi.org/10.1016/ j.eururo.2013.07.031.

4. Brookman-May, S. et al. Features associated with recurrence beyond 5 years after nephrectomy and nephron-sparing surgery for renal cell carcinoma: development and internal validation of a risk model (PRELANE score) to predict late recurrence based on a large multicenter database (CORONA/SATURN project). Eur. Urol. 64, 472–477 (2013).

5. Brookman-May, S. et al. Time to recurrence is a significant predictor of cancer-specific survival after recurrence in patients with recurrent renal cell carcinoma—results from a comprehensive multi-centre database (CORONA/SATURN-Project). BJU Int. http://dx.doi.org/10.1111/bju12246.

© 2013 Macmillan Publishers Limited. All rights reserved