original articles
Annals of Oncology 11. Joseph JD, Lu N, Qian J et al. A clinically relevant androgen receptor mutation confers resistance to 2nd generation anti-androgens enzalutamide and ARN-509. Cancer Discov 2013; 3: 1020–1029. 12. Taplin ME, Bubley GJ, Shuster TD et al. Mutation of the androgen-receptor gene in metastatic androgen-independent prostate cancer. N Engl J Med 1995; 332: 1393–1398. 13. Nakazawa M, Antonarakis ES, Luo J. Androgen receptor splice variants in the era of enzalutamide and abiraterone. Horm Cancer 2014; 5: 265–273. 14. Dehm SM, Tindall DJ. Alternatively spliced androgen receptor variants. Endocr Relat Cancer 2011; 18: R183–R196. 15. Hu R, Dunn TA, Wei S et al. Ligand-independent androgen receptor variants derived from splicing of cryptic exons signify hormone-refractory prostate cancer. Cancer Res 2009; 69: 16–22. 16. Plymate SR, Luo J. The expression signature of androgen receptor splice variants and their distinctive transcriptional activities in castration-resistant prostate cancer. In Androgen-Responsive Genes in Prostate Cancer. New York: Springer 2013; 201–213. 17. Guo Z, Yang X, Sun F et al. A novel androgen receptor splice variant is upregulated during prostate cancer progression and promotes androgen depletionresistant growth. Cancer Res 2009; 69: 2305–2313. 18. Hu R, Isaacs WB, Luo J. A snapshot of the expression signature of androgen receptor splicing variants and their distinctive transcriptional activities. Prostate 2011; 71: 1656–1667. 19. Li Y, Chan SC, Brand LJ et al. Androgen receptor splice variants mediate enzalutamide resistance in castration-resistant prostate cancer cell lines. Cancer Res 2013; 73: 483–489.
20. Mostaghel EA, Marck BT, Plymate SR et al. Resistance to CYP17A1 inhibition with abiraterone in castration-resistant prostate cancer: induction of steroidogenesis and androgen receptor splice variants. Clin Cancer Res 2011; 17: 5913–5925. 21. Antonarakis ES, Lu C, Wang H et al. AR-V7 and resistance to enzalutamide and abiraterone in prostate cancer. N Engl J Med 2014; 371: 1028–1038. 22. Antonarakis ES, Lu C, Luber B et al. Androgen receptor splice variant 7 and efficacy of taxane chemotherapy in patients with metastatic castration-resistant prostate cancer. JAMA Oncol 2015 June 4 [epub ahead of print], doi:10.1001/ jamaoncol.2015.1341. 23. Lorente D, Mateo J, Perez-Lopez R et al. Sequencing of agents in castrationresistant prostate cancer. Lancet Oncol 2015; 16: e279–e292. 24. Bianchini D, Lorente D, Rodriguez-Vida A et al. Antitumour activity of enzalutamide (MDV3100) in patients with metastatic castration-resistant prostate cancer (CRPC) pre-treated with docetaxel and abiraterone. Eur J Cancer 2014; 50: 78–84. 25. Loriot Y, Bianchini D, Ileana E et al. Antitumour activity of abiraterone acetate against metastatic castration-resistant prostate cancer progressing after docetaxel and enzalutamide (MDV3100). Ann Oncol 2013; 24: 1807–1812. 26. Nadal R, Zhang Z, Rahman H et al. Clinical activity of enzalutamide in docetaxel‐ naïve and docetaxel‐pretreated patients with metastatic castration‐resistant prostate cancer. Prostate 2014; 74: 1560–1568. 27. van Soest RJ, de Morree ES, Kweldam CF et al. Targeting the androgen receptor confers in vivo cross-resistance between enzalutamide and docetaxel, but not cabazitaxel, in castration-resistant prostate cancer. Eur Urol 2015; 67: 981–985.
Annals of Oncology 26: 1865–1870, 2015 doi:10.1093/annonc/mdv254 Published online 2 June 2015
Treatment outcome and patterns of relapse following adjuvant carboplatin for stage I testicular seminomatous germ-cell tumour: results from a 17-year UK experience C. Chau1,2,3, R. Cathomas4, M. Wheater3, D. Klingbiel5, M. Fehr6, J. Bennett7, H. Markham8, C. Lee1,3, S. J. Crabb1,3 & T. Geldart7* 1 Cancer Sciences Unit, University of Southampton Faculty of Medicine, Southampton; 2NIHR Wellcome Trust Clinical Research Facility, University of Southampton, Southampton; 3Department of Medical Oncology, University Hospital Southampton NHS Foundation Trust, Southampton, UK; 4Department of Medical Oncology, Kantonsspital Graubünden, Chur; 5Swiss Group for Clinical Cancer Research Coordinating Center, Bern; 6Department of Medical Oncology, Kantonsspital St Gallen Hospital, St Gallen, Switzerland; 7Dorset Cancer Centre, Poole Hospital NHS Foundation Trust, Poole; 8Department of Histopathology, University Hospital Southampton NHS Foundation Trust, Southampton, UK
Received 4 February 2015; revised 19 May 2015; accepted 22 May 2015
Background: Following inguinal orchidectomy, management options for patients with stage I seminoma include initial surveillance or treatment with adjuvant radiotherapy or chemotherapy. The anticipated relapse rate for patients followed by surveillance alone is ∼15%, with adjuvant treatment this risk is reduced to ∼4%–5% at 5 years. After carboplatin treatment, follow-up strategies vary and there are no validated, predictive markers of relapse. Patients and methods: We conducted a retrospective analysis of all patients presenting with stage I seminoma who received a single cycle of adjuvant carboplatin in South Central England between 1996 and 2013. We report on outcome and the results of univariate and multivariate analysis evaluating possible risk factors for post carboplatin relapse. *Correspondence to: Dr Thomas Geldart, Dorset Cancer Centre, Poole Hospital NHS Foundation Trust, Poole BH15 2JB, UK. Tel: +44-1202-665511; E-mail: tom.geldart@ rbch.nhs.uk
© The Author 2015. Published by Oxford University Press on behalf of the European Society for Medical Oncology. All rights reserved. For permissions, please email:
[email protected].
original articles
Annals of Oncology
Results: A total of 517 eligible patients were identified. All underwent nuclear medicine estimation of glomerular filtration rate before treatment with carboplatin (dosed at area under the curve × 7). With a median follow-up of 47.2 months (range 0.4–214 months), 21/517 patients have relapsed resulting in a 5-year estimated relapse-free survival of 95.0% (95% confidence interval 92.8% to 97.3%). Median time to relapse was 22.7 months (range 12.5–109.5 months). Relapse beyond 3 years was rare (4/517; 0.8%). Twenty of 21 (95%) relapsed patients had retroperitoneal lymph node metastases. The majority (16/21; 76%) of patients had elevated tumour markers at relapse. Twenty of 517 (3.9%) patients developed a new contralateral testicular germ-cell cancer. There were no seminoma-related deaths. Tumour size was the only variable significantly associated with an increased risk of relapse. Conclusions: Overall results for this large cohort of patients confirm an excellent prognosis for these patients with outcomes equivalent to those seen in prospective clinical trials. Increasing tumour size alone appears to be associated with an increased risk of post chemotherapy relapse. Key words: stage I seminoma, relapse pattern, adjuvant carboplatin, chemotherapy, treatment outcome
introduction Seminoma accounts for 60% of testicular germ-cell tumours (GCTs) with 80% of patients presenting with stage I disease [1]. Radical inguinal orchidectomy, staging with computed tomography (CT) of chest, abdomen and pelvis along with measurement of tumour markers: α-fetoprotein (AFP), β-human chorionic gonadotropin (β-HCG) and lactate dehydrogenase (LDH) represents standard initial management. Following orchidectomy, management options for stage I seminoma comprise of active surveillance or adjuvant treatment. Regardless of approach, virtually all patients with seminoma are cured given the effectiveness of salvage chemotherapy at relapse [2]. For patients managed with active surveillance, a recent pooled international retrospective study of 1344 patients reported a relapse rate of 13% after a median follow-up of 52 months, with no seminoma-related deaths [3]. A separate retrospective analysis suggested tumour size >40 mm in combination with rete testis invasion was associated with a significantly higher relapse rate (up to 32%), although prospective analyses have not confirmed these risk categories [4, 5]. The MRC TE19/EORTC 30982 clinical trial evaluated adjuvant therapy in 1447 patients treated with a single dose of carboplatin AUC 7 (area under the curve × 7; n = 573) compared with para-aortic radiotherapy (n = 904). The trial demonstrated noninferiority of carboplatin with an overall relapse rate of 5.3% versus 4% after a median follow-up of 6.5 years. Given the favourable toxicity profile and the avoidance of risk of radiotherapy-induced secondary malignancies, carboplatin has become the preferred adjuvant treatment [6–9]. Several trials have used a combined, risk-adapted approach in which patients with tumour size >40 mm and rete testis invasion were offered adjuvant carboplatin, and low-risk patients managed with surveillance [10, 11]. Within the TE19 trial, no results relating to possible risk factors of relapse were reported; as such, it remains unclear to what extent adjuvant chemotherapy reduces risk in patients at higher risk of relapse when compared with lower risk patients.
methods patients and adjuvant treatment Detailed methods are available in supplementary material, available at Annals of Oncology online. All consecutive patients with stage I seminoma treated with
| Chau et al.
a single dose of adjuvant carboplatin AUC7 within a single supra-network regional team between July 1996 and October 2013 were identified. Carboplatin dose was calculated on radioisotope measured glomerular filtration rate (GFR) uncorrected for body surface area using the Calvert formula [12]. Tumour characteristics, time to adjuvant chemotherapy, clinical outcomes, incidence of contralateral GCT and secondary malignancies, pattern of relapse and salvage treatment given at relapse were recorded. Cut-off for follow-up was 28 April 2014.
follow-up and treatment at relapse Twenty-nine of 517 (5.6%) patients within the cohort were treated within the TE19 trial and were followed up 3 monthly for the first 2 years, 6 monthly in year 3–5 and annually thereafter for up to 10 years. At each visit, patients underwent clinical examination, chest X-ray and measurement of AFP, β-HCG, ± LDH. Patients outside the clinical trial were followed up similarly for a minimum of 3 years (3 monthly in year 1, 4 monthly in year 2, 6 monthly in year 3). CT abdomen was carried out routinely at 1 year post orchidectomy and if clinically indicated thereafter. After discharge, followup was taken over by the general practitioner (GP). GPs were provided with written guidance requiring referral back to the regional specialist team on suspicion of relapse. At relapse, patients with stage IIA relapse underwent dog-leg radiotherapy. Patients of good prognosis disease according to International Germ-cell Cancer Collaborative Group (IGCCCG) classification, with stage IIB or higher relapse, had three cycles of standard dose BEP (bleomycin, etoposide, cisplatin) chemotherapy [13]. Patients with intermediate prognosis received four cycles of BEP. All salvage chemotherapy was co-ordinated via a single regional specialist centre.
statistical methods Detailed statistical methods are available in supplementary material, available at Annals of Oncology online. The primary end point was 5-year relapse-free survival rate. Secondary objectives included overall survival, cancer-specific survival, incidence of contralateral GCT, risk factors predictive of relapse and incidence of secondary malignancies. Since all deaths are reported back to NHS records, all relapses occurring within region are referred back to participating centres, and all systemic treatment co-ordinated from a single regional centre (UHS), follow-up is effectively much longer. To account for this, we calculated ‘virtual’ follow-up as time from date of orchidectomy to data cut-off date (28 April 2014) or death. Virtual follow-up was used to calculate a ‘real world’ follow-up time but was not used for Kaplan–Meier analysis of relapse. Time-to-event end points were analysed by Cox regression models and Kaplan–Meier methods.
Volume 26 | No. 9 | September 2015
original articles
Annals of Oncology
results
1
patient characteristics
relapse rate and assessment of risk factors Twenty-one of 517 (4%) patients have relapsed giving an estimated relapse free survival (RFS) of 97.4% [95% confidence interval (CI) 96.0% to 98.9%] at 2 years and 95.0% (92.8% to 97.3%) at 5 years (Figure 1 and Table 3). Twenty of 21 (95%) with good prognosis metastatic disease (IGCCCG classification) and 1/21 (4.8%) with intermediate prognosis disease (Table 2). The median time from orchidectomy to relapse was 22.7 months (range 12.5–109.5 months). Twelve of 21 (57%) relapses occurred within 2 years and 17/21 (80%) within 3 years. The relapse rate for the whole population after 3 years was 0.8% (4/517; after 49, 50, 53 and 108 months). There were no seminoma-related deaths. Three quarters of relapsing patients, 16/21 (76%) had elevated tumour markers at relapse (9 raised βHCG and LDH, 4 raised βHCG alone and 3 raised LDH alone) and/or an abnormal protocol scheduled CT. The main site of relapse was in the retroperitoneum (19/21; 90%) with 100% of patients with abdominal/pelvic involvement at relapse. Thirteen of 21 (62%) patients were asymptomatic at relapse. Twelve of 21 (57%) asymptomatic patients were detected in routine follow-up, and one was discovered incidentally during follow-up of a gastrointestinal stromal tumour (GIST). Eight of 22 (36%) patients (including all relapses beyond 3 years) were symptomatic at relapse, with the most common presentation Table 1. Patient characteristics (N = 517) Age (years) Median (range) Tumour size (mm) ≤40 >40 Unknown Rete testis invasion No Yes Unknown Tumour >40 mm and rete invasion No Yes Unknown Extra-testicular invasion No Yes Unknown Time to adjuvant carboplatin (days) ≤30 31–60 >60 Unknown
Volume 26 | No. 9 | September 2015
38 (18–73) 333 (64%) 169 (33%) 15 (3%) 331 (64%) 166 (32%) 20 (4%) 427 (82%) 65 (13%) 25 (5%) 397 (77%) 67 (13%) 53 (10%) 103 (20%) 284 (55%) 112 (22%) 18 (3%)
0.8 Estimated RFS
A total of 517 eligible patients were identified (Table 1). Median age at diagnosis was 38 years (range 18–73 years). Hospital median follow-up was 3.9 years (range 0–17.8 years). Virtual median follow-up was 6.5 years (range 0.3–17.8 years).
0.6
0.4
0.2 £ 40 mm > 40 mm
Log-rank P = 0.12 0 0 # at risk £ 40 mm > 40 mm
2
4
6
8
10
12
14
31 13
8 4
5 0
Time (years) 333 168
287 135
163 69
90 36
55 27
Figure 1. Relapse-free survival by size of primary tumour.
Table 2. Patient characteristics at relapse (N = 21) Setting of relapse detection Detected at routine follow-up 13 (62%) Unplanned 8 (38%) First indicator of relapse in asymptomatic patients (N = 14) Abnormal CT 11 (79%) Elevated tumour markers 2 (14%) Other abnormal imaging 1 (7%) (PET—unplanned and asymptomatic) Elevated LDH or β-HCG at relapse Yes 16 (76%) No 4 (19%) Unknown 1 (5%) Site of relapse Retroperitoneal LN only 17 (81%) Iliac LN only 1 (5%) Multiple nodal sites 2 (9%) 1× retroperitoneal + mediastinal 1× inguinal + iliac Multiple nodal and visceral sites 1 (5%) Retroperitoneal, mediastinal, liver Salvage therapy at first relapse Chemotherapy 17 (81%) Dog-leg radiotherapy 4 (19%) Salvage chemotherapy regime at first relapse BEP ×3 7 (41%) BEP ×4 7 (41%) Modified BEP 3 (18%) IGCCCG classification Good prognosis 20 (95%) Intermediate prognosis 1 (5%) Response to salvage therapy Complete response, marker negative 6 (29%) Partial response, marker negative 14 (66%) Unknown 1 (5%)
doi:10.1093/annonc/mdv254 |
original articles
Annals of Oncology
being abdominal pain (6/22; 27%). One patient (1/22; 5%) presented with a deep vein thrombosis and one (1/22; 5%) with a pulmonary embolus. When continuous values for tumour size were used, we found some evidence that larger size was associated with higher risk of relapse. For each centimetre increase in primary tumour size, there was an increase in risk of relapse (univariate 19%, 95% CI 0% to 43%, P = 0.05; multivariate 24%, 95% CI 1% to 52%, P = 0.04; Table 3). Using a 4-cm cut-off, 5-year RFS was nearly double (5.9% versus 3.3%) in those patients with larger primary tumours (Figure 1). There was no statistically significant association between relapse rate and the other variables assessed.
treatment of relapse Treatment at relapse is summarized in Figure 2. Four of 21 (19%) patients (those with stage IIA relapse) were treated with dog-leg radiotherapy, the remainder with salvage chemotherapy.
Eighteen of 21 (86%) remain alive and disease free following initial salvage treatment alone. Three of 21 (14%) relapsing patients suffered a second relapse (time from first to second relapse: 7.6, 11.2 and 12 months). One first relapsed with stage IIA disease and was treated with salvage radiotherapy to complete response. His 1-year planned CT scan identified a second relapse with disease in the abdomen and mediastinum. He received successful salvage EP chemotherapy but died from a myocardial infarction 7 years later. The remaining two patients had stage IIC disease at first relapse and received standard dose, salvage BEP chemotherapy. Within 12 months, both had a second relapse, one identified by planned CT at 1 year, one identified by PET/CT at 11 months). The former patient received three cycles of paclitaxel, ifosfamide, cisplatin (TIP) chemotherapy and remains alive and well without further relapse after a follow-up of 11 years. The latter received salvage treatment with involved field retroperitoneal radiotherapy to a dose of 40 Gy in 20 fractions, but suffered a
Table 3. Uni- and multivariate analysis of possible risk factors for relapse No of relapse All patients 21/517 Tumour size ≤4 cm 11/333 >4 cm 10/169 Unknown 0/15 Tumour size (continuous*) Rete testis invasion No 13/331 Yes 7/166 Unknown 1/20 Size >4 cm AND rete testis invasion No 17/427 Yes 3/65 Unknown 1/25 Age ≤29 6/86 30–59 14/418 ≥60 1/11 Lymphovascular invasion No 15/378 Yes 3/52 Excess smear artefact 2/50 Unknown 1/37 Extra-testicular invasion No 15/397 Yes 3/67 Unknown 3/53 Time to adjuvant carboplatin ≤30 days 4/103 31–60 days 14/284 >60 days 3/112 Unknown 0/18
Relapse rate
5-year RFS
4%
95.0%
3.3% 5.9% 0%
95.5% 93.6%
HR (95% CI), P value univariable
HR (95% CI), P multivariable**
1 (ref) 1.9 (0.8–4.6), P = 0.13
1 (ref) 2.2 (0.9–5.8), P = 0.10
1.2 (1.0–1.4), P = 0.05
1.2 (1.0–1.5), P = 0.04 1 (ref) 0.9 (0.3–2.6), P = 0.89
3.9% 2.2% 5.0%
95.1% 94.9%
1 (ref) 1.2 (0.5–2.9), P = 0.75
4.0% 4.6% 4.0%
95.1% 94.0%
1 (ref) 1.3 (0.4–4.5), P = 0.67
7.0% 3.3% 9.1%
89.9% 96.1% 88.9%
2.2 (0.9–5.9), P = 0.10 1 (ref) 2.9 (0.4–22.0), P = 0.30
4.0% 5.8% 4.0% 2.7%
95.2% 93.5% 94.3%
1.0 (ref) 1.4 (0.4–4.9), P = 0.58 0.9 (0.2–4.0), P = 0.90
1 (ref) 0.7 (0.1–3.4), P = 0.65 0.7 (0.2–3.2), P = 0.64
3.8% 4.5% 5.7%
95.4% 94.7%
1 (ref) 1.2 (0.3–4.0), P = 0.82
1 (ref) 1.1 (0.3–4.0), P = 0.88
3.9% 4.9% 2.7% 0.0%
95.4% 93.9% 96.4%
1 (ref) 1.4 (0.4–4.1), P = 0.59 0.7 (0.2–3.1), P = 0.63
The values shown are from the model with the continuous version of tumour size, but are similar in the model with the 40-mm cut-off. Bold value is statistically significant. *per cm increase of tumour size. **Variables without HR have not been included; tumour size has been included either continuous or with the 40 mm cut-off.
| Chau et al.
Volume 26 | No. 9 | September 2015
original articles
Annals of Oncology
517 patients
21 patients relapsed First relapse
17 chemotherapy
1
4 radiotherapy
1
18 alive and disease free
GCT (one suicide, one myocardial infarction, two bronchopneumonia and one rectal cancer). Six of 517 (1.2%) patients were diagnosed with non-germ-cell secondary malignancy during follow-up. Three of 6 within 12 months of receiving their adjuvant carboplatin chemotherapy ( plasmacytoma, renal cell carcinoma, GIST) and 3/6 between 6 and 8 years after their adjuvant chemotherapy (rectal cancer, malignant melanoma, mantle cell lymphoma).
1
discussion Second relapse
1 radiotherapy
Third relapse
1 chemotherapy (TIP)
2 chemotherapy (TIP,EP)
2 alive and disease free, 1 died of myocardial infarction
1 alive and disease free
Figure 2. Treatment at relapse.
third relapse within 3 months of treatment with enlarging pelvic nodal disease outside his radiotherapy field. He was treated with TIP chemotherapy and remains alive and disease free after a follow-up of 5 years.
contralateral testicular germ-cell tumour Twenty of 517 (3.9%) patients were diagnosed with a separate contralateral testicular germ-cell tumour (CTGCT). None of these patients relapsed from their initial seminoma. Three of 517 (0.6%) patients were diagnosed with synchronous bilateral seminoma (treated with bilateral orchidectomy followed by one cycle of adjuvant carboplatin). Seventeen of 517 (3.3%) patients developed a new, metachronous CTGCT (9 seminoma, 8 non-seminoma) during follow-up, with a median time to CTGCT of 8.8 years (range 0.8–22 years). Two patients developed CTGCT more than 20 years after their first diagnosis. In the 9/517 (2%) patients who developed contralateral seminoma (all stage 1 disease), 2 underwent orchidectomy alone without additional adjuvant therapy (diagnosed within 2 years of the first GCT), 3 received another single dose of adjuvant carboplatin and 4 received adjuvant radiotherapy (the diagnosis of these GCT were made before 2000). In the 8/517 (1.5%) patients who developed contralateral non-seminoma, 4 with stage 1 disease underwent orchidectomy alone and 4 with metastatic disease received BEP chemotherapy.
secondary malignancies and overall survival As of 28 April 2014, 511/517 (99%) patients are alive and 6 patients have died. There were no seminoma-related deaths. All deaths occurred after 5 years and the 5-year cancer-specific survival rate was 100%. One patient (0.2%) died from metastatic disease from a second, contralateral non-seminomatous GCT primary, the other five patients died of causes unrelated to
Volume 26 | No. 9 | September 2015
To our knowledge, we present the largest published series describing the clinical outcome and relapse data of patients with stage I seminoma treated with a single cycle of postoperative adjuvant carboplatin chemotherapy dosed exclusively according to nuclear medicine estimation of GFR. We report real-world experience and confirm the low relapse rate and highly favourable prognosis for this large cohort of patients. Outcome overall is excellent with all relapsing patients responding to salvage treatment and no seminoma-related deaths. Our results are in keeping with the majority of patients (355/ 573; 62% patients in the chemotherapy arm) treated prospectively in the TE19 trial whose carboplatin dose was determined by nuclear medicine estimation of GFR and prospectively followed for up to 10 years. Our group and subgroup analysis of TE19 have previously emphasized the importance of accurate estimation of GFR and the potential for underdosing and higher relapse rate when GFR estimation formulae and lower doses of carboplatin are used [7, 12]. In contrast to the very low rates of CTGCT seen in the pivotal phase III TE19 trial, the development of a contralateral primary in our series; 20/517 (3.9%), was as common as the development of systemic disease relapse following adjuvant carboplatin (21/ 517, 4%), but with a notably longer median time to event (median 8.8 versus 1.9 years). In keeping with contemporary published series, our findings do not support the conclusion drawn by the TE19 trial that the risk of developing contralateral testicular tumours is reduced by adjuvant carboplatin [14, 15]. It should be noted however, that patients in our series did not routinely undergo biopsy of their contralateral testis at initial diagnosis unless there were clear risk factors for contralateral intratubular germ-cell neoplasia; namely age
