REVIEW URRENT C OPINION
Update in germ cell tumours Darren R. Feldman a,b
Purpose of review The purpose of this study is to update the reader on advances in postpubertal male germ cell tumours (GCTs) over the last 18 months. Recent findings Single nucleotide polymorphisms, including in four sex-determination genes, have been identified as additional genetic susceptibility loci to testicular GCT development. New insights into cisplatin resistance implicate the PDGFR-PIK3CA-AKT and RAS pathways. Circulating tumour cells and circulating microRNAs are potential new biomarkers. In clinical stage I (CS-I) GCT, two large studies have confirmed the excellent outcomes achieved with surveillance, which is now the management option of choice for CS I-A nonseminoma and all CS-I seminomas; CS I-B nonseminoma remains controversial. First-line trials of dosedense multidrug regimens reported promising results but have not yet supplanted BEPx4. Survivorship issues, including secondary malignancies from chemotherapy, remain important in this disease and are a continuing focus of ongoing research. Summary Important research questions remain across all aspects of GCT. The next decade is likely to produce many new and exciting discoveries that will benefit GCT patients. Keywords cisplatin resistance, germ cell tumours, surveillance, survivorship, testicular cancer
INTRODUCTION Germ cell tumours (GCTs) represent the most common cancer diagnosed in men aged 15–40 years in the United States [1]. They comprise over 95% of testicular neoplasms and a small proportion of mediastinal, pineal gland and ovarian malignancies. Considering GCT of all primary sites, nearly 10 000 new cases will be diagnosed in 2014 in the United States [1]. Although 95% of testicular GCTs are expected to be cured, GCTs remain the focus of research for several reasons, including a low cure rate among certain patient subsets such as those with International Germ Cell Cancer Collaborative Group (IGCCCG) poor-risk advanced GCT, particularly primary mediastinal nonseminomas; the array of treatment options available for early-stage disease; and the significant increase in late toxicities such as cardiovascular disease and secondary malignancies among patients treated with chemotherapy or radiation. This review will provide an update on recent important publications in the field and ongoing research.
GENETIC EPIDEMIOLOGY GCTs are one of the few tumours to harbour a nearly universal genetic alteration, copy number gain of
chromosome 12p, most commonly in the form of an isochromosome [i(12p)] [2]. Despite identification of this pathognomonic alteration more than 30 years ago [3], the pathogenesis of GCT has remained largely elusive. A major recent advance was the identification of several single nucleotide polymorphisms (SNPs) that increase the risk of GCT diagnosis [4–10]. Compared with other malignancies, the effect size of these SNPs is much greater, and interestingly, most are within genes involved in gonocyte development, including ATF7IP, BAK1, DMRT1, KITLG, SPRY4 and TERT. Involvement of DMRT1 in sex determination led Koster et al. [11 ] to postulate that SNPs in other sex determination genes might also confer GCT susceptibility. They evaluated SNPs in 32 sex determination &
a Genitourinary Oncology Service, Department of Medicine, Memorial Sloan Kettering Cancer Center and bDepartment of Medicine, Weill Medical College of Cornell University, New York, New York, USA
Correspondence to Darren R. Feldman, MD, Genitourinary Oncology Service, Memorial Sloan Kettering Cancer Center, 353 East 68th Street, New York, NY 10065, USA. Tel: +1 646 422 4491; fax: +1 212 988 0701; e-mail: [email protected] Curr Opin Oncol 2015, 27:177–184 DOI:10.1097/CCO.0000000000000179
1040-8746 Copyright ß 2015 Wolters Kluwer Health, Inc. All rights reserved.
www.co-oncology.com
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Genitourinary system
KEY POINTS SNPs in genes associated with sex determination have been identified as additional susceptibility loci for the development of testicular GCT.
(HCG) and alpha-fetoprotein (AFP) play an integral role in the management of GCT, they are imperfect and additional noninvasive biomarkers could be useful to confirm remission status and improve screening for recurrence. With this goal, Nastaly et al. [14 ] developed a method to detect circulating tumour cells (CTCs) from GCT patients by combining the CellSearch assay, which detects epithelial cells using antibodies against keratins and EpCAM, with antibodies against GCT proteins, OCT3/4 (seminomas and embryonal carcinoma) and SALL-4 (most GCT histologies). CTCs were detected in 25 (17%) of 143 patients, including 10 of 15 with distant metastatic disease and some with normal AFP and HCG [14 ]. Demonstration of [i(12p)] in several CTCs confirmed their GCT origin and highlighted the potential for genomic profiling of these cells to noninvasively inform tumour molecular status [14 ]. Further methodological optimization and demonstration of their prognostic significance is necessary before CTCs can be incorporated into clinical management. Micro-RNAs (miRs), small noncoding RNA regions involved in posttranscriptional gene regulation, represent another promising circulating biomarker, as they are highly stable in serum, easy to detect and appear to have specificity for certain malignancies. Gillis et al. [15 ] expanded upon previous studies demonstrating higher serum expression of five miRs (302, 367, 371, 372 and 373) in GCT patients vs. controls [16,17], by developing a more standardized, quality-controlled and calibrated assay, termed TSmiR. Serum samples from 80 patients with GCT, 47 healthy controls and 12 patients with non-GCT testicular masses were evaluated. In preorchiectomy serum samples, TSmiR, using four of the miRs (367, 371–3), detected 98% of tumour samples with only one false-negative (tumour sample incorrectly classified as normal). This sensitivity was significantly higher than preoperative AFP/HCG (36% for seminoma, 57% for NSGCT). In addition, among clinical stage I (CS-I) patients, there was near normalization of miR levels following orchiectomy. Finally, a trend towards higher miR levels for patients with metastatic vs. CS-I disease was observed [15 ]. &
A method of capturing circulating GCT cells has recently been described and microRNAs represent another promising circulating GCT biomarker with a high sensitivity; however, both will require further study before they can be integrated into daily practice. Activation of the PDGFR-PI3K-AKT and RAS pathways and induction of cellular differentiation are two potential mechanisms by which GCT may become resistant to cisplatin. Active surveillance has been validated as effective and the least toxic treatment strategy for most CS-I GCTs; however, management of CS-IB nonseminoma remains controversial. Novel initial chemotherapy regimens for intermediate and poor-risk GCT, including one that incorporates tumour marker decline rates into treatment decisions, have shown promising results but have not supplanted BEPx4. NSGCT patients treated with chemotherapy are at an increased risk of second malignant neoplasms, particularly soft tissue sarcomas and neoplasms of the kidney and thyroid gland.
genes including DMRT1 using datasets from three previously reported independent genome-wide association studies. Significant association with GCT diagnosis was demonstrated for the gene set in all three cohorts, even after removing DMRT1 from the analysis. Although there was heterogeneity in the effect size of specific SNPs between cohorts, four candidate SNPs ranked consistently high, including DMRT1, GATA4, IGF1R and ZFPM2 [11 ]. Using a three-stage experimental approach, Litchfield et al. [12 ] also identified a novel susceptibility locus. This SNP, at 3q25.31, resides within an intergenic region, and therefore, unlike the above loci has an unclear biologic function [12 ]. With declining costs of high throughput technology and more rapid turn-around time, additional genetic foci are likely to be identified in the coming years and the functional significance of these findings will be more intimately explored. &
&
&
CIRCULATING BIOMARKERS Serum tumour markers (STMs) are elevated in approximately 70% of nonseminomatous GCT (NSGCT) and 20–30% of seminomas at diagnosis [13]. Although human chorionic gonadotropin 178
www.co-oncology.com
&
&
&&
&&
CISPLATIN RESISTANCE Elucidating the mechanisms underlying the unique sensitivity of most GCT to cisplatin as well as the determinants of resistance remains a research priority. Juliachs et al. [18 ] compared the cisplatin-sensitive cell lines GCT27 and SuSa with their cisplatin-resistant derivative cell lines, GCT27R and SuSaR (generated by continuous exposure to &
Volume 27 Number 3 May 2015
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Update in germ cell tumours Feldman
cisplatin). They detected increased expression of phosphorylated AKT (p-AKT) in the resistant derivative clones, GCT27R and SuSaR, as compared with their sensitive counterparts. Increased p-AKT seemed to result from upstream activation of platelet-derived growth factor receptor-b (PDGFR-b) and the authors demonstrated that cisplatin resistance could be partially reversed by its inhibition with pazopanib or sunitinib. They further demonstrated that overexpression of PDGFR-b in a cisplatin-sensitive cell line could induce resistance [18 ]. These data, together with prior reports demonstrating p-AKT influences cisplatin-resistance via regulation of p21 localization [19], suggest this pathway as a potential future therapeutic target in resistant patients. Feldman et al. [20 ] used a Sequenome assay to evaluate for common mutations in seven genes within 46 cisplatin-resistant and 24 cisplatin-sensitive GCT. Mutations in four genes, KRAS, PIK3CA, AKT1 and NRAS, were exclusively identified in resistant tumours. No mutations were found in HRAS or BRAF, whereas FGFR3 mutations were distributed equally among sensitive and resistant patients [20 ]. These findings suggest mutational differences between sensitive and resistant tumours, which may explain clinical behaviour. Future studies with more comprehensive tumour genomic analysis may provide further insight into this important aspect of GCT biology. Finally, Abada and Howell [21 ] evaluated the embryonal carcinoma cell line, NT2/D1, which is known to differentiate along a neuronal lineage after treatment with retinoic acid. NT2/D1 cells treated with cisplatin exhibited differentiation similar to retinoic acid and resulted in cisplatin (and paclitaxel) resistance. Paclitaxel treatment did not exert this effect. Cisplatin resistance correlated with decreased pluripotency markers, and overexpression of the pluripotency gene, NANOG, blunted the ability of cisplatin to induce resistance [21 ]. These three studies add to our knowledge of the determinants of platinum sensitivity in GCT. Not only will further insights likely lead to improved outcomes for advanced GCT patients, but they may also be applicable to other cisplatin-treated malignancies. &
&
&
&
&
compared with active surveillance and increasing recognition of late toxicities resulting from adjuvant treatment has shifted preferred management towards active surveillance. Two recent studies [22 ,23 ] (Table 1) evaluated active surveillance in large populations of CS-I patients. Daugaard et al. [22 ] reported 15-year disease-specific survival (DSS) and OS rates of 99 and 95%, respectively, within a nationwide, consecutive cohort of 1226 patients with CS-I NSGCT. Relapses, observed in 31% of patients, occurred early (median 5 months) and were more common among patients with lymphovascular invasion (LVIþ, 43%) vs. those without (LVI, 26%). Unfortunately, information on LVI was missing in 543 (44%) cases. Despite excellent outcomes with active surveillance, burden of care was still considerable in this study. Patients who recurred received three to four cycles of systemic chemotherapy, and 6% relapsed with IGCCCG intermediate or poor-risk disease. Nearly 1200 cycles of BEP were administered, at least 400 more than expected with a risk-adapted strategy [surveillance for LVI patients; retroperitoneal lymph node dissection (RPLND) or one cycle of bleomycin, etoposide and cisplatin (BEPx1) for LVIþ patients] in a comparable population. Furthermore, 26% of patients required operations (primarily RPLND) following chemotherapy [22 ], whereas with BEPx1, almost no patients need surgery [24]. Kollmannsberger et al. [23 ] evaluated 2483 patients with CS-I GCT (seminoma and NSGCT) managed with active surveillance, culled from five groups in North America and Europe. Of 1139 CS-I NSGCT patients, 221 (19%) relapsed, including 44% of LVIþ patients and 14% of LVI patients. Of note, only 16% of the 1139 patients had LVIþ disease due to one group (SWENOTECA) using a risk-adapted policy [25]. Most relapses occurred early (median 6 months), particularly for LVIþ patients (median 4 months) with over 90% of relapses detected within 2 years. Approximately 10% of relapsing patients were intermediate or poor-risk, but only five (0.4%) died from GCT or GCT treatment. Treatment burden with active surveillance in this study is not clear because the number of chemotherapy cycles administered was not reported; however, at least 194 patients received some chemotherapy [23 ]. For CS-I seminoma, 13% of 1344 patients relapsed with a longer median time to recurrence of 14 months. Only two relapsing patients had intermediate-risk disease; all others were good-risk. Two-thirds of relapsing patients received chemotherapy at relapse while 31% underwent radiation; 11% of radiation-treated patients had a second relapse requiring chemotherapy. No patient died &&
&&
&&
&&
&&
&&
CS-I GERM CELL TUMOUR Approximately 70% of GCTs are initially diagnosed as CS-I, including 80% of seminomas and 60% of nonseminomas. Historical management has involved aggressive adjuvant strategies to achieve high rates of progression-free survival (PFS). However, a lack of benefit in overall survival (OS)
1040-8746 Copyright ß 2015 Wolters Kluwer Health, Inc. All rights reserved.
www.co-oncology.com
179
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Genitourinary system Table 1. Recent large series of active surveillance for clinical stage I germ cell tumour Nonseminoma Kollmannsberger et al. [23 ] (N ¼ 1139)
Kollmannsberger et al. && [23 ] (N ¼ 1344)
Nationwide consecutive cohort in Denmark
Pooled retrospective data from 5 independent datasetsa
Pooled retrospective data from 5 independent datasetsa
&&
Source population
Seminoma
Daugaard et al. [22 ] (N ¼ 1226)
&&
Median follow-up time
180 months
62 months
52 months
LVIþ/LVI
25%/75%b
16%/84%
NA
5-year relapse rate (LVIþ/LVI)
31% (43%/26%)
19% (44%/14%)
13% (NA)
Median time to relapse
5 months
6 months
14 months
Relapses 2 years
94%
92%
75%
Sites of relapse (Abd/lung/both)
59%/16%/7%
NR
NR
Marker elevation at relapse (HCG/AFP)
65% (44%/44%)
48% (NR/NR)
3% (3%/0%)
Method of relapse detection (marker/CT Abd)
65%/56%c
48%/44%
3%/87%
IGCCCG group at relapse (Good/Int/Poor)
94%/5%/1%
90%/8%/2%
99%/1%/NA
No. of chemotherapy cycles
1170
NR
NR
No. of surgeries
98 (8%)
NR
NR
Disease-specific survivald
99.1%
99%
99%
Overall survival
94.5%
99%
99%
Lost to follow-up
48 (4%)
NR
NR
Abd, abdomen; DSS, disease-specific survival (includes treatment-related deaths); GCT, germ cell tumour; IGCCCG, International Germ Cell Cancer Collaborative Group; Int, intermediate; LVIþ, lymphovascular invasion present; LVI, lymphovascular invasion absent; NA, not applicable; No., number; NR, not reported. a Sites: Swedish Norwegian Testicular Cancer Group (n ¼ 1022); Princess Margaret Hospital (n ¼ 603), Bart’s Cancer Institute (n ¼ 277), British Columbia Cancer Agency/Oregon (n ¼ 535) and University of Oxford/Churchill Hospital (n ¼ 45). b Percentages taken of those who had LVI information. Note 543 patients were missing this information. c Patients may have been detected by both methods. d Both deaths due to GCT and treatment for GCT.
of GCT, but one died of treatment for relapse (DSS 99%) [23 ]. In summary, these two studies demonstrate excellent outcomes with active surveillance for Stage I GCT and support its use as standard management for CS-IA NSGCT and all CS-I seminoma. Timing and patterns of relapse have important implications to active surveillance programmes. CS-IB (LVIþ) NSGCT remains controversial because of the substantial risk of recurrence (>40%) and significant treatment burden for relapsing patients. Therefore, risk-adapted strategies (BEPx1 or primary RPLND for LVIþ patients) remain valid alternatives to uniform active surveillance for Stage I NSGCT. &&
ADVANCED DISEASE Despite many studies in IGCCCG intermediate and poor-risk GCT, BEPx4 has remained the standard of care, as nearly every experimental regimen has resulted in increased toxicity without any improvement in efficacy [26]. Four cycles of etoposide, 180
www.co-oncology.com
ifosfamide and cisplatin (VIPx4) are an established alternative for patients at a high risk for bleomycin pulmonary toxicity, as it achieves comparable outcomes and avoids bleomycin, with increased haematologic toxicity as its only major drawback [27]. Recent studies (Table 2) [28 ,29 ,30 ,31 ] reporting promising results have begun to challenge this established paradigm. In a randomized phase II study, 89 patients with IGCCCG poor-risk NSGCT received either BEPx4 or four cycles of CBOP/BEP (carboplatin, bleomycin, vincristine and cisplatin alternating with bleomycin, etoposide and cisplatin) [28 ]. The primary endpoint, favourable response rate (FRR), consisting of either complete response (CR) or partial response (PR) with negative tumour markers (PR-negative markers), was achieved in 74% of CBOP/BEP patients vs. 61% of BEP patients. Two-year PFS (58 vs. 43%) and 5-year OS (65 vs. 58%) were also superior with CBOP/BEP, although these differences were not statistically significant [28 ]. Despite meeting its primary endpoint, the study did not segue into a phase III trial as &
&&
&
&
&
&
Volume 27 Number 3 May 2015
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
1040-8746 Copyright ß 2015 Wolters Kluwer Health, Inc. All rights reserved. P (29)
TIP [30 ] (Phase II)
G (15)
I (16)
P (12)
43
AccBEP
TIP
BEPx4
98
44
vs.
vs.
BEPx4
46
BEPx1 ! DIR
vs.
vs.
105
CBOP/BEP
Regimens
43
N
89%
83%
49%
vs.
(G 93%; I 94%; P 50%)
(G 100%; I 93%; P 100%)
(I 100%; P 93%) 98%c
(I 87%; P 76%)
95%
81%c
79%
65% (P ¼ 0.34)
48%
vs.
73%
N/A
(P ¼ 0.05)
vs.
59%
N/A
N/Aa 50%
56%b
38%b
61%
64%b
3-year OS, % (P)
vs.
56%b
3-year PFS, % (P)
vs.
vs.
74%
FRR, %
Acceptable safety and promising efficacy has led to a phase II/III RCT vs. standard BEPx4 in Australia/New Zealand
Promising efficacy has led to a phase II RCT vs. BEPx4 in U.S.
Neuropathy worse with DIR
DIR patients received 5 total cycles vs. 4 with standard arm.
DIR is complex, uses nearly all drugs with antiGCT activity and may decrease salvage options
No OS benefit with DIR
Wide HR for PFS benefit (0.44–1.0)
However, switch to DIR has not been widely adapted for poor-risk patients with slow STM decline due to:
First RCT to meet 18 endpoint vs. BEPx4
Poor accrual prevented planned segue into phase III randomized trial
>95% with G3 tox on CBOP/BEP
6 (14%) toxic deaths on CBOP/BEP
Comment
AccBEP, accelerated BEP (etoposide plus cisplatin given every 2 weeks x 4 cycles with bleomycin given weekly 12 weeks); CBOP/BEP, carboplatin, bleomycin, vincristine and cisplatin alternating with bleomycin, etoposide and cisplatin; DIR, dose-intense regimen (see text); FRR, favourable response rate; G, good-risk; GCT, germ cell tumour; I, intermediate-risk; IGCCCG, International Germ Cell Cancer Collaborative Group; OS, overall survival; P, poor-risk; PFS, progression-free survival; RCT, randomized controlled trial; TIP, paclitaxel, ifosfamide and cisplatin. a Randomized phase II study but not designed to compare the two arms; rather, primary endpoint was to exclude with 90% confidence that CBOP/BEP would lead to an FRR of
Update in germ cell tumours Darren R. Feldman a,b
Purpose of review The purpose of this study is to update the reader on advances in postpubertal male germ cell tumours (GCTs) over the last 18 months. Recent findings Single nucleotide polymorphisms, including in four sex-determination genes, have been identified as additional genetic susceptibility loci to testicular GCT development. New insights into cisplatin resistance implicate the PDGFR-PIK3CA-AKT and RAS pathways. Circulating tumour cells and circulating microRNAs are potential new biomarkers. In clinical stage I (CS-I) GCT, two large studies have confirmed the excellent outcomes achieved with surveillance, which is now the management option of choice for CS I-A nonseminoma and all CS-I seminomas; CS I-B nonseminoma remains controversial. First-line trials of dosedense multidrug regimens reported promising results but have not yet supplanted BEPx4. Survivorship issues, including secondary malignancies from chemotherapy, remain important in this disease and are a continuing focus of ongoing research. Summary Important research questions remain across all aspects of GCT. The next decade is likely to produce many new and exciting discoveries that will benefit GCT patients. Keywords cisplatin resistance, germ cell tumours, surveillance, survivorship, testicular cancer
INTRODUCTION Germ cell tumours (GCTs) represent the most common cancer diagnosed in men aged 15–40 years in the United States [1]. They comprise over 95% of testicular neoplasms and a small proportion of mediastinal, pineal gland and ovarian malignancies. Considering GCT of all primary sites, nearly 10 000 new cases will be diagnosed in 2014 in the United States [1]. Although 95% of testicular GCTs are expected to be cured, GCTs remain the focus of research for several reasons, including a low cure rate among certain patient subsets such as those with International Germ Cell Cancer Collaborative Group (IGCCCG) poor-risk advanced GCT, particularly primary mediastinal nonseminomas; the array of treatment options available for early-stage disease; and the significant increase in late toxicities such as cardiovascular disease and secondary malignancies among patients treated with chemotherapy or radiation. This review will provide an update on recent important publications in the field and ongoing research.
GENETIC EPIDEMIOLOGY GCTs are one of the few tumours to harbour a nearly universal genetic alteration, copy number gain of
chromosome 12p, most commonly in the form of an isochromosome [i(12p)] [2]. Despite identification of this pathognomonic alteration more than 30 years ago [3], the pathogenesis of GCT has remained largely elusive. A major recent advance was the identification of several single nucleotide polymorphisms (SNPs) that increase the risk of GCT diagnosis [4–10]. Compared with other malignancies, the effect size of these SNPs is much greater, and interestingly, most are within genes involved in gonocyte development, including ATF7IP, BAK1, DMRT1, KITLG, SPRY4 and TERT. Involvement of DMRT1 in sex determination led Koster et al. [11 ] to postulate that SNPs in other sex determination genes might also confer GCT susceptibility. They evaluated SNPs in 32 sex determination &
a Genitourinary Oncology Service, Department of Medicine, Memorial Sloan Kettering Cancer Center and bDepartment of Medicine, Weill Medical College of Cornell University, New York, New York, USA
Correspondence to Darren R. Feldman, MD, Genitourinary Oncology Service, Memorial Sloan Kettering Cancer Center, 353 East 68th Street, New York, NY 10065, USA. Tel: +1 646 422 4491; fax: +1 212 988 0701; e-mail: [email protected] Curr Opin Oncol 2015, 27:177–184 DOI:10.1097/CCO.0000000000000179
1040-8746 Copyright ß 2015 Wolters Kluwer Health, Inc. All rights reserved.
www.co-oncology.com
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Genitourinary system
KEY POINTS SNPs in genes associated with sex determination have been identified as additional susceptibility loci for the development of testicular GCT.
(HCG) and alpha-fetoprotein (AFP) play an integral role in the management of GCT, they are imperfect and additional noninvasive biomarkers could be useful to confirm remission status and improve screening for recurrence. With this goal, Nastaly et al. [14 ] developed a method to detect circulating tumour cells (CTCs) from GCT patients by combining the CellSearch assay, which detects epithelial cells using antibodies against keratins and EpCAM, with antibodies against GCT proteins, OCT3/4 (seminomas and embryonal carcinoma) and SALL-4 (most GCT histologies). CTCs were detected in 25 (17%) of 143 patients, including 10 of 15 with distant metastatic disease and some with normal AFP and HCG [14 ]. Demonstration of [i(12p)] in several CTCs confirmed their GCT origin and highlighted the potential for genomic profiling of these cells to noninvasively inform tumour molecular status [14 ]. Further methodological optimization and demonstration of their prognostic significance is necessary before CTCs can be incorporated into clinical management. Micro-RNAs (miRs), small noncoding RNA regions involved in posttranscriptional gene regulation, represent another promising circulating biomarker, as they are highly stable in serum, easy to detect and appear to have specificity for certain malignancies. Gillis et al. [15 ] expanded upon previous studies demonstrating higher serum expression of five miRs (302, 367, 371, 372 and 373) in GCT patients vs. controls [16,17], by developing a more standardized, quality-controlled and calibrated assay, termed TSmiR. Serum samples from 80 patients with GCT, 47 healthy controls and 12 patients with non-GCT testicular masses were evaluated. In preorchiectomy serum samples, TSmiR, using four of the miRs (367, 371–3), detected 98% of tumour samples with only one false-negative (tumour sample incorrectly classified as normal). This sensitivity was significantly higher than preoperative AFP/HCG (36% for seminoma, 57% for NSGCT). In addition, among clinical stage I (CS-I) patients, there was near normalization of miR levels following orchiectomy. Finally, a trend towards higher miR levels for patients with metastatic vs. CS-I disease was observed [15 ]. &
A method of capturing circulating GCT cells has recently been described and microRNAs represent another promising circulating GCT biomarker with a high sensitivity; however, both will require further study before they can be integrated into daily practice. Activation of the PDGFR-PI3K-AKT and RAS pathways and induction of cellular differentiation are two potential mechanisms by which GCT may become resistant to cisplatin. Active surveillance has been validated as effective and the least toxic treatment strategy for most CS-I GCTs; however, management of CS-IB nonseminoma remains controversial. Novel initial chemotherapy regimens for intermediate and poor-risk GCT, including one that incorporates tumour marker decline rates into treatment decisions, have shown promising results but have not supplanted BEPx4. NSGCT patients treated with chemotherapy are at an increased risk of second malignant neoplasms, particularly soft tissue sarcomas and neoplasms of the kidney and thyroid gland.
genes including DMRT1 using datasets from three previously reported independent genome-wide association studies. Significant association with GCT diagnosis was demonstrated for the gene set in all three cohorts, even after removing DMRT1 from the analysis. Although there was heterogeneity in the effect size of specific SNPs between cohorts, four candidate SNPs ranked consistently high, including DMRT1, GATA4, IGF1R and ZFPM2 [11 ]. Using a three-stage experimental approach, Litchfield et al. [12 ] also identified a novel susceptibility locus. This SNP, at 3q25.31, resides within an intergenic region, and therefore, unlike the above loci has an unclear biologic function [12 ]. With declining costs of high throughput technology and more rapid turn-around time, additional genetic foci are likely to be identified in the coming years and the functional significance of these findings will be more intimately explored. &
&
&
CIRCULATING BIOMARKERS Serum tumour markers (STMs) are elevated in approximately 70% of nonseminomatous GCT (NSGCT) and 20–30% of seminomas at diagnosis [13]. Although human chorionic gonadotropin 178
www.co-oncology.com
&
&
&&
&&
CISPLATIN RESISTANCE Elucidating the mechanisms underlying the unique sensitivity of most GCT to cisplatin as well as the determinants of resistance remains a research priority. Juliachs et al. [18 ] compared the cisplatin-sensitive cell lines GCT27 and SuSa with their cisplatin-resistant derivative cell lines, GCT27R and SuSaR (generated by continuous exposure to &
Volume 27 Number 3 May 2015
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Update in germ cell tumours Feldman
cisplatin). They detected increased expression of phosphorylated AKT (p-AKT) in the resistant derivative clones, GCT27R and SuSaR, as compared with their sensitive counterparts. Increased p-AKT seemed to result from upstream activation of platelet-derived growth factor receptor-b (PDGFR-b) and the authors demonstrated that cisplatin resistance could be partially reversed by its inhibition with pazopanib or sunitinib. They further demonstrated that overexpression of PDGFR-b in a cisplatin-sensitive cell line could induce resistance [18 ]. These data, together with prior reports demonstrating p-AKT influences cisplatin-resistance via regulation of p21 localization [19], suggest this pathway as a potential future therapeutic target in resistant patients. Feldman et al. [20 ] used a Sequenome assay to evaluate for common mutations in seven genes within 46 cisplatin-resistant and 24 cisplatin-sensitive GCT. Mutations in four genes, KRAS, PIK3CA, AKT1 and NRAS, were exclusively identified in resistant tumours. No mutations were found in HRAS or BRAF, whereas FGFR3 mutations were distributed equally among sensitive and resistant patients [20 ]. These findings suggest mutational differences between sensitive and resistant tumours, which may explain clinical behaviour. Future studies with more comprehensive tumour genomic analysis may provide further insight into this important aspect of GCT biology. Finally, Abada and Howell [21 ] evaluated the embryonal carcinoma cell line, NT2/D1, which is known to differentiate along a neuronal lineage after treatment with retinoic acid. NT2/D1 cells treated with cisplatin exhibited differentiation similar to retinoic acid and resulted in cisplatin (and paclitaxel) resistance. Paclitaxel treatment did not exert this effect. Cisplatin resistance correlated with decreased pluripotency markers, and overexpression of the pluripotency gene, NANOG, blunted the ability of cisplatin to induce resistance [21 ]. These three studies add to our knowledge of the determinants of platinum sensitivity in GCT. Not only will further insights likely lead to improved outcomes for advanced GCT patients, but they may also be applicable to other cisplatin-treated malignancies. &
&
&
&
&
compared with active surveillance and increasing recognition of late toxicities resulting from adjuvant treatment has shifted preferred management towards active surveillance. Two recent studies [22 ,23 ] (Table 1) evaluated active surveillance in large populations of CS-I patients. Daugaard et al. [22 ] reported 15-year disease-specific survival (DSS) and OS rates of 99 and 95%, respectively, within a nationwide, consecutive cohort of 1226 patients with CS-I NSGCT. Relapses, observed in 31% of patients, occurred early (median 5 months) and were more common among patients with lymphovascular invasion (LVIþ, 43%) vs. those without (LVI, 26%). Unfortunately, information on LVI was missing in 543 (44%) cases. Despite excellent outcomes with active surveillance, burden of care was still considerable in this study. Patients who recurred received three to four cycles of systemic chemotherapy, and 6% relapsed with IGCCCG intermediate or poor-risk disease. Nearly 1200 cycles of BEP were administered, at least 400 more than expected with a risk-adapted strategy [surveillance for LVI patients; retroperitoneal lymph node dissection (RPLND) or one cycle of bleomycin, etoposide and cisplatin (BEPx1) for LVIþ patients] in a comparable population. Furthermore, 26% of patients required operations (primarily RPLND) following chemotherapy [22 ], whereas with BEPx1, almost no patients need surgery [24]. Kollmannsberger et al. [23 ] evaluated 2483 patients with CS-I GCT (seminoma and NSGCT) managed with active surveillance, culled from five groups in North America and Europe. Of 1139 CS-I NSGCT patients, 221 (19%) relapsed, including 44% of LVIþ patients and 14% of LVI patients. Of note, only 16% of the 1139 patients had LVIþ disease due to one group (SWENOTECA) using a risk-adapted policy [25]. Most relapses occurred early (median 6 months), particularly for LVIþ patients (median 4 months) with over 90% of relapses detected within 2 years. Approximately 10% of relapsing patients were intermediate or poor-risk, but only five (0.4%) died from GCT or GCT treatment. Treatment burden with active surveillance in this study is not clear because the number of chemotherapy cycles administered was not reported; however, at least 194 patients received some chemotherapy [23 ]. For CS-I seminoma, 13% of 1344 patients relapsed with a longer median time to recurrence of 14 months. Only two relapsing patients had intermediate-risk disease; all others were good-risk. Two-thirds of relapsing patients received chemotherapy at relapse while 31% underwent radiation; 11% of radiation-treated patients had a second relapse requiring chemotherapy. No patient died &&
&&
&&
&&
&&
&&
CS-I GERM CELL TUMOUR Approximately 70% of GCTs are initially diagnosed as CS-I, including 80% of seminomas and 60% of nonseminomas. Historical management has involved aggressive adjuvant strategies to achieve high rates of progression-free survival (PFS). However, a lack of benefit in overall survival (OS)
1040-8746 Copyright ß 2015 Wolters Kluwer Health, Inc. All rights reserved.
www.co-oncology.com
179
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Genitourinary system Table 1. Recent large series of active surveillance for clinical stage I germ cell tumour Nonseminoma Kollmannsberger et al. [23 ] (N ¼ 1139)
Kollmannsberger et al. && [23 ] (N ¼ 1344)
Nationwide consecutive cohort in Denmark
Pooled retrospective data from 5 independent datasetsa
Pooled retrospective data from 5 independent datasetsa
&&
Source population
Seminoma
Daugaard et al. [22 ] (N ¼ 1226)
&&
Median follow-up time
180 months
62 months
52 months
LVIþ/LVI
25%/75%b
16%/84%
NA
5-year relapse rate (LVIþ/LVI)
31% (43%/26%)
19% (44%/14%)
13% (NA)
Median time to relapse
5 months
6 months
14 months
Relapses 2 years
94%
92%
75%
Sites of relapse (Abd/lung/both)
59%/16%/7%
NR
NR
Marker elevation at relapse (HCG/AFP)
65% (44%/44%)
48% (NR/NR)
3% (3%/0%)
Method of relapse detection (marker/CT Abd)
65%/56%c
48%/44%
3%/87%
IGCCCG group at relapse (Good/Int/Poor)
94%/5%/1%
90%/8%/2%
99%/1%/NA
No. of chemotherapy cycles
1170
NR
NR
No. of surgeries
98 (8%)
NR
NR
Disease-specific survivald
99.1%
99%
99%
Overall survival
94.5%
99%
99%
Lost to follow-up
48 (4%)
NR
NR
Abd, abdomen; DSS, disease-specific survival (includes treatment-related deaths); GCT, germ cell tumour; IGCCCG, International Germ Cell Cancer Collaborative Group; Int, intermediate; LVIþ, lymphovascular invasion present; LVI, lymphovascular invasion absent; NA, not applicable; No., number; NR, not reported. a Sites: Swedish Norwegian Testicular Cancer Group (n ¼ 1022); Princess Margaret Hospital (n ¼ 603), Bart’s Cancer Institute (n ¼ 277), British Columbia Cancer Agency/Oregon (n ¼ 535) and University of Oxford/Churchill Hospital (n ¼ 45). b Percentages taken of those who had LVI information. Note 543 patients were missing this information. c Patients may have been detected by both methods. d Both deaths due to GCT and treatment for GCT.
of GCT, but one died of treatment for relapse (DSS 99%) [23 ]. In summary, these two studies demonstrate excellent outcomes with active surveillance for Stage I GCT and support its use as standard management for CS-IA NSGCT and all CS-I seminoma. Timing and patterns of relapse have important implications to active surveillance programmes. CS-IB (LVIþ) NSGCT remains controversial because of the substantial risk of recurrence (>40%) and significant treatment burden for relapsing patients. Therefore, risk-adapted strategies (BEPx1 or primary RPLND for LVIþ patients) remain valid alternatives to uniform active surveillance for Stage I NSGCT. &&
ADVANCED DISEASE Despite many studies in IGCCCG intermediate and poor-risk GCT, BEPx4 has remained the standard of care, as nearly every experimental regimen has resulted in increased toxicity without any improvement in efficacy [26]. Four cycles of etoposide, 180
www.co-oncology.com
ifosfamide and cisplatin (VIPx4) are an established alternative for patients at a high risk for bleomycin pulmonary toxicity, as it achieves comparable outcomes and avoids bleomycin, with increased haematologic toxicity as its only major drawback [27]. Recent studies (Table 2) [28 ,29 ,30 ,31 ] reporting promising results have begun to challenge this established paradigm. In a randomized phase II study, 89 patients with IGCCCG poor-risk NSGCT received either BEPx4 or four cycles of CBOP/BEP (carboplatin, bleomycin, vincristine and cisplatin alternating with bleomycin, etoposide and cisplatin) [28 ]. The primary endpoint, favourable response rate (FRR), consisting of either complete response (CR) or partial response (PR) with negative tumour markers (PR-negative markers), was achieved in 74% of CBOP/BEP patients vs. 61% of BEP patients. Two-year PFS (58 vs. 43%) and 5-year OS (65 vs. 58%) were also superior with CBOP/BEP, although these differences were not statistically significant [28 ]. Despite meeting its primary endpoint, the study did not segue into a phase III trial as &
&&
&
&
&
&
Volume 27 Number 3 May 2015
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
1040-8746 Copyright ß 2015 Wolters Kluwer Health, Inc. All rights reserved. P (29)
TIP [30 ] (Phase II)
G (15)
I (16)
P (12)
43
AccBEP
TIP
BEPx4
98
44
vs.
vs.
BEPx4
46
BEPx1 ! DIR
vs.
vs.
105
CBOP/BEP
Regimens
43
N
89%
83%
49%
vs.
(G 93%; I 94%; P 50%)
(G 100%; I 93%; P 100%)
(I 100%; P 93%) 98%c
(I 87%; P 76%)
95%
81%c
79%
65% (P ¼ 0.34)
48%
vs.
73%
N/A
(P ¼ 0.05)
vs.
59%
N/A
N/Aa 50%
56%b
38%b
61%
64%b
3-year OS, % (P)
vs.
56%b
3-year PFS, % (P)
vs.
vs.
74%
FRR, %
Acceptable safety and promising efficacy has led to a phase II/III RCT vs. standard BEPx4 in Australia/New Zealand
Promising efficacy has led to a phase II RCT vs. BEPx4 in U.S.
Neuropathy worse with DIR
DIR patients received 5 total cycles vs. 4 with standard arm.
DIR is complex, uses nearly all drugs with antiGCT activity and may decrease salvage options
No OS benefit with DIR
Wide HR for PFS benefit (0.44–1.0)
However, switch to DIR has not been widely adapted for poor-risk patients with slow STM decline due to:
First RCT to meet 18 endpoint vs. BEPx4
Poor accrual prevented planned segue into phase III randomized trial
>95% with G3 tox on CBOP/BEP
6 (14%) toxic deaths on CBOP/BEP
Comment
AccBEP, accelerated BEP (etoposide plus cisplatin given every 2 weeks x 4 cycles with bleomycin given weekly 12 weeks); CBOP/BEP, carboplatin, bleomycin, vincristine and cisplatin alternating with bleomycin, etoposide and cisplatin; DIR, dose-intense regimen (see text); FRR, favourable response rate; G, good-risk; GCT, germ cell tumour; I, intermediate-risk; IGCCCG, International Germ Cell Cancer Collaborative Group; OS, overall survival; P, poor-risk; PFS, progression-free survival; RCT, randomized controlled trial; TIP, paclitaxel, ifosfamide and cisplatin. a Randomized phase II study but not designed to compare the two arms; rather, primary endpoint was to exclude with 90% confidence that CBOP/BEP would lead to an FRR of
![[Non-seminomatous germ cell tumours].](/assets/img/hold.png)
