Phase II Study of Oxaliplatin and Gemcitabine Salvage Chemotherapy in Patients with Cisplatin-Refractory Nonseminomatous Germ Cell Tumor

1. Introduction 

Germ cell tumor (GCT) represents a model of solid tumor curable with chemotherapy. Cisplatin-based chemotherapy induces long-term remissions in approximately 80% of patients with advanced GCT as primary treatment, while in only 25–30% of patients at relapse/progression [1], [2]. Patients with cisplatin-refractory GCT have an extremely poor prognosis with long-term survival achieved in less than 5% of cases with standard-dose cisplatin-based chemotherapy [3]. Several treatment options have been investigated in this setting. Tandem high-dose chemotherapy based on carboplatin and etoposide showed impressive results with a 2-year failure-free survival of 32%, when used as the first salvage modality with the exception of mediastinal primary GCT [4]. As consequence, tandem high-dose chemotherapy is a possible option for these patients [4], [5]. However, in most cases multicycle high-dose chemotherapy is not effective or not feasible, so that the identification of new drugs and/or new combinations remains a priority. Several chemotherapeutic agents have been investigated in patients with cisplatin-refractory GCT, but only oral etoposide, paclitaxel, gemcitabine, and oxaliplatin have shown significant activity as single agents [3], [6], [7], [8], [9]. In the last years, a few phase II studies have investigated the combination of these chemotherapeutic agents in order to improve the outcome of these patients [10], [11], [12].

We investigated the feasibility and activity of a schedule with full doses of oxaliplatin and gemcitabine in patients with cisplatin-refractory GCT.

3. Results 

Between May 2002 and June 2005, eighteen patients with cisplatin-refractory GCT were enrolled onto the study. The trial was closed after the first step because of slow recruitment. All patients were assessable for toxicity and response. All patients were male with a median age of 32 years (range 19–51). Primary site was testis in twelve cases, retroperitoneum in four, and mediastinum in two. All patients were heavily pretreated with a median number of eight (range, 3–12) cisplatin-containing chemotherapy cycles before gemcitabine plus oxaliplatin therapy. Moreover, four patients (22%) had received a single high-dose chemotherapy course. In addition, two patients (11%) had experienced relapse later than 2 years (late relapse). Seven patients (39%) were classified as cisplatin-refractory, while eleven (61%) as absolutely cisplatin-refractory. Patient characteristics are listed in Table 1.

Table 1. Patient characteristics at study entry

	Characteristics	No of Patients		%
	Total patients	18
	Age, years
	Median (range)	32 (19–51)
	Location of primary tumor
	Testis	12		67%
	Retroperitoneum	4		22%
	Mediastinum	2		11%
	Sites of disease
	Lung	16		89%
	Retroperitoneum	12		67%
	Mediastinum	8		44%
	Liver	3		17%
	Bone	2		11%
	Other	4		22%
	Elevation of tumor markers
	AFP	9		50%
	HCG	5		28%
	LDH	2		11%
	No elevated markers	4		22%
	Oxaliplatin–gemcitabine chemotherapy setting
	Second-line	1		5%
	Third-line	13		72%
	Fourth- Fifth-line	4		22%
	Number of cisplatin-based cycles before therapy
	Median (range)	8 (3–12)
	Pretreated with high-dose chemotherapy	4		22%
	Late relapse*	2		11%
	Platinum-refractory disease
	Cisplatin-refractory**	7		39%
	Absolutely cisplatin-refractory***	11		61%

AFP, alpha-fetoprotein; HCG, beta-human chorionic gonadotropin; LDH, lactate dehydrogenase.

* Relapse occurring later than 2 years after previous chemotherapy. ** At least stable disease or better, but evidence of disease progression within four weeks of the last cisplatin-based chemotherapy. *** Progressive disease during cisplatin-based chemotherapy.

The total number of cycles administered was 60, with a median of 3 cycles (range 1–6) for each patient. No toxic death occurred. No patients terminated therapy because of toxicity. Dose reductions and treatment delay were required in five cases. Toxicity is summarized in Table 2. Overall, hematological toxicity was the most prevalent side-effect consisting of grade 3–4 neutropenia in seven patients (one complicated with pneumonia), grade 3 thrombocytopenia in four, and grade 3 anemia requiring packed red blood cell transfusions in two patients. Non-hematological toxicities occurred in three patients who presented grade 3 peripheral sensory neurotoxicity, two with grade 3 asthenia, and one with grade 3 vomiting.

Table 2. Grade 3–4 toxicity according to NCI-CTC version 2.0 (n=18)

	Toxicity	No of Patients	%
	Neutropenia	7	39%
	Thrombocyto penia	4	22%
	Anemia	2	11%
	Peripheral neurotoxicity	3	17%
	Asthenia	2	11%
	Vomiting	1	5%
	Pneumonia	1	5%

NCI-CTC, National Cancer Institute Common Toxicity Criteria.

Overall, an objective response was obtained in 3 cases (17%), SD in three, while PD in twelve. One patient achieved a clinical CR, one a PRm− in whom complete surgical resection of residual masses yielded mature teratoma only (pCR), and one a PRm+ in whom complete surgical resection of residual masses yielded viable tumor cells (sCR). The last patient had previously undergone high-dose chemotherapy and experienced relapse later than 2 years (Table 3, patient no 1). These three patients were characterized by testicular primary embryonal carcinoma. They remained disease-free at 44+, 20+, and 18+ months of follow-up. Results are presented in detail in Table 3.

Table 3. Oxaliplatin–gemcitabine combination in patients with cisplatin-refractory germ cell tumor

	Pt no	Primary Site	Histology	Previous Treatments (Response)	Late Relapse	Refractory Status	Sites of Disease before Oxa–Gem	Markers before Oxa–Gem	Response	DFS (mo)	OS (mo)
	1	Testis	EC, CC	PEBx6(sCR); VeIPx3+CarboPECx1(CR); VIPx1(PD)	Yes	Absol Refr	Rp, Ps	AFP=121	PR+ (sCR)	44+	44+
	2	Med	EC, IT, YST	PVBx3(PD)	No	Absol Refr	Med	None	SD	0	4
	3	Testis	CC, MT, S, AC	PEBx3(PR+); VeIPx4(CR); VIPx2+CarboPECx1(SD)	Yes	Refr	Rp, Lu, Med	AFP=5207	SD	0	24
	4	Testis	EC, S	PEBx4(PR−); VeIPx2(PD)	No	Absol Refr	Rp, Lu, Med	LDH=4xN	PD	0	3
	5	Rp	YST	PEBx6(PR−); VeIPx2(PD)	No	Absol Refr	Rp, Ps, Lu, Med, Li, Bo	AFP=1500	PD	0	7
	6	Rp	CC	PEBx4(PR+); VIPx1(PD)	No	Absol Refr	Rp, Ps, Lu, Med, Li, Bo	HCG=55581; LDH=3xN	PD	0	2
	7	Rp	CC	PEBx6(PD); TIPx4(PR+)	No	Refr	Lu	HCG=1477	PD	0	11
	8	Testis	EC	PEBx4(CR); PEBx2(PD); TIPx1(PD)	No	Absol Refr	Rp, Lu	AFP=26828	CR	20+	20+
	9	Testis	EC	PEBx6(PD); VeIPx4(PD)	No	Absol Refr	Rp, Lu, Med, Pleura	None	PD	0	1
	10	Rp	EC, YST	PEBx9(PR−); TIPx3(PD)	No	Absol Refr	Rp, Lu, Li	AFP=52500	SD	0	14+
	11	Testis	CC	PEBx3(PR+); TIPx2(PD)	No	Absol Refr	Rp, Lu	HCG=1500	PD	0	4
	12	Testis	EC	PEBx4(PR−); VIPx6(SD)	No	Refr	Rp, Lu	None	PR−(pCR)	18+	18+
	13	Med	YST, UC	PEBx4(PR−); CarboPECx1(PD); Vbl-Ifox2(PD); Tax-Gemx1(PD)	No	Absol Refr	Lu, Med	AFP=1345	PD	0	7
	14	Testis	EC, CC	PEBx4(PR+); VeIPx4(PD)	No	Absol Refr	Lu	HCG=24801	PD	0	2
	15	Testis	EC, S	PEBx4(PD); Tax-Ifox2+CarboPEI(PR−)	No	Refr	Lu	HCG=17; AFP=20	PD	0	8
	16	Testis	EC, IT, YST	PEBx3(PR+); VeIPx4(PR+)	No	Refr	Rp, Lu, Med	AFP=30	PD	0	8+
	17	Testis	YST, S	PEBx4(PR−); TIPx4(SD)	No	Refr	Lu	AFP=80	PD	0	7+
	18	Testis	YST, IT	PEBx4(PR+); VeIPx4(SD)	No	Refr	Rp, Lu	None	PD	0	6+

Pt, patient; Oxa-Gem, oxaliplatin–gemcitabine chemotherapy; DFS, disease-free survival; OS, overall survival; mo, months; Rp, retroperitoneum; Med, mediastinum; Ps, Psoas; Lu, lung; Li, liver; Bo, bone; EC, embryonal carcinoma; CC, choriocarcinoma; YST, yolk sac tumor; IT, immature teratoma; MT, mature teratoma; S, seminoma; AC, adenocarcinoma; UC, undifferentiated carcinoma; PEB, cisplatin, etoposide, bleomycin; VeIP, vinblastine, ifosfamide, cisplatin; CarboPEC, high-dose chemotherapy with carboplatin, etoposide, cyclophosphamide; VIP, etoposide, ifosfamide, cisplatin; PVB, cisplatin, vinblastine, bleomycin; TIP, paclitaxel, ifosfamide, cisplatin; Vbl-Ifo, vinblastine, ifosfamide; Tax-Gem, paclitaxel, gemcitabine; Tax-Ifo, paclitaxel, ifosfamide; CarboPEI, high-dose chemotherapy with carboplatin, etoposide, cyclophosphamide; Absol Refr, absolutely refractory; Refr, refractory; HCG, beta-human chorionic gonadotropin; AFP, alpha-fetoprotein; LDH, lactate dehydrogenase; N, normal value; CR, clinical complete remission; pCR, pathological complete remission; sCR, surgical complete remission; PR− partial remission with negative marker; PR+, partial remission with positive marker; SD, stable disease; PD, progressive disease.

After treatment with gemcitabine and oxaliplatin, most of patients were subjected to further chemotherapy at progression. One patient achieved a short-lasting PRm+ after high-dose epirubicin. Two patients received the epirubicin-cisplatin combination (1 SD, 1 PD); two daily oral etoposide (1 SD, 1 PD), two paclitaxel, ifosfamide and cisplatin (TIP) (both PD), two weekly paclitaxel (both PD), one taxotere (PD), one weekly gemcitabine (PD).

After a median follow-up of 7 months (range, 1 to 44+ months) for all patients, three (17%) are continuously disease-free at 44+, 20+, and 18+ months, four are alive with disease (14+, 8+, 7+, 6+), while eleven died of their disease. Median overall survival for all patients was 7 months (range, 1 to 44+ months).

4. Discussion 

The prognosis for patients with cisplatin-refractory GCT remains poor, in spite of the positive results obtained with multicycle high-dose chemotherapy in a subset of these patients [3], [4], [5]. New chemotherapeutic regimens with significant antitumor activity are needed to offer an opportunity to achieve long-term survival for this group of young patients. Only few agents have been found active in patients with cisplatin-refractory GCT [15]. Paclitaxel, gemcitabine, and oxaliplatin as single-agents showed objective responses in only approximately 10% to 20% of these patients, but long-term remissions were not reported [7], [8], [9]. Instead, the two-drug combination of paclitaxel and gemcitabine showed a response rate of 30%, with one durable clinical CR, and the combination of paclitaxel and oxaliplatin which obtained a PRm− with a sCR after post-chemotherapy surgery [10], [16].

We previously reported a preliminary experience with a chemotherapeutic regimen consisted of a weekly administration of the three-drug combination of gemcitabine, paclitaxel and oxaliplatin in patients with cisplatin-refractory GCT [17]. Only one partial response (ORR=11%) was achieved, lasting 5 months. Median overall survival for all patients was 5 months (range, 1 to 15 months). However, we need to consider that in most cases dose reductions and treatment delay were required due to grade 3–4 hematological toxicity. Pizzocaro et al. are reported preliminary results in abstract form of another weekly administration of a three-drug combination regimen including paclitaxel, cisplatin and gemcitabine as third-line treatment in GCT patients [12]. Toxicity was substantial, while activity in cisplatin-refractory patients was not clearly assessable because no information on prognostically relevant patient characteristics was reported. However, the toxicity of the weekly three-drug combinations seems considerably high for this heavily pretreated group of patients.

Other authors showed that oxaliplatin at full doses, but not as weekly administration possess activity in cisplatin-refractory GCT [9]. Paclitaxel has been recently incorporated into second-line chemotherapeutic regimens and is currently investigated in phase III trials as part of a primary chemotherapy regimen [18], [19]. On the basis of these observations, we planned the phase II study with gemcitabine and oxaliplatin at full doses in a three-week schedule.

In the present study, the combination of gemcitabine and oxaliplatin showed three cases (17%) of a prolonged CR lasting 44+, 20+ and 18+ months. In one case, the surgical resection of residual masses played a fundamental role to achieve a disease-free status (Table 3, patient no 1). The achievement of a response in these patients is very important, because the induction of a partial remission may allow the resection of residual masses and subsequently may offer a chance for some patients to achieve long-term survival [15].

These results are encouraging considering the poor prognostic characteristics of the study patient group (Table 3). Moreover, they appear better than those obtained with paclitaxel, gemcitabine and paclitaxel as single agents and those achieved with the combination therapies of gemcitabine–paclitaxel and oxaliplatin–paclitaxel (Table 4) [6], [7], [8], [9], [16]. Our results are in line with those reported by Bokemeyer et al. [20] and Pectasides et al. [21] in two recent studies with the same combination regimen and similar patient characteristics (Table 4). Overall, of sixty-eight patients reported in these three clinical studies of the gemcitabine–oxaliplatin combination in cisplatin-refractory GCT, seven (10%) achieved prolonged CR, in all cases characterized by testicular primary GCT.

Table 4. Gemcitabine and/or oxaliplatin based salvage treatments for cisplatin-refractory germ cell tumor

	Author	Treatment	No of patients with cisplatin-refractory germ cell tumor	Objective Response Rate (%)	Persistent Complete Remission (%)
	Einhorn [7]	Gem 1200mg/m2 (days 1, 8, 15; q28)	13	0	0
	Bokemeyer [8]	Gem 1000mg/m2 (days 1, 8, 15; q28)	17	3 (18%)	0
	Kollmansberger [9]	Oxa 130mg/m2 (days 1, 15; q28) or Oxa 60mg/m2 (days 1, 8, 15; q28)	27	4 (15%)	0
	Hinton [10]	Pac 110mg/m2 and Gem 1000mg/m2 (days 1, 8, 15; q28)	10	3 (30%)	1 (10%)
	Theodore [16]	Pac 175mg/m2 and Oxa 130mg/m2 (day 1; q21)	16	1 (6%)	1 (6%)a
	De Giorgi [17]	Oxa 50mg/m2, Gem 800mg/m2 and Pac 70mg/m2 (days 1, 8, 15; q28)	9	1 (11%)	0
	Kollmansberger [20]	Gem 1000mg/m2 (days 1, 8) and Oxa 130mg/m2 (day 1); q21	22	8 (36%)	1 (5%)
	Pectasides [21]	Gem 1000mg/m2 (days 1, 8) and Oxa 130mg/m2 (day 1); q21	28	9 (32%)	3 (11%)
	Current series	Gem 1250mg/m2 (days 1, 8) and Oxa 130mg/m2 (day 1) q21	18	3 (17%)	3 (17%)a

Note: Patients with cisplatin-refractory germ cell tumor have been selected from studies in relapsed and/or cisplatin-refractory germ cell tumor.

a Including one case in whom complete surgical resection of residual masses yielded viable tumor cells (surgical complete remission).

Recently, irinotecan as monochemotherapy every three weeks has showed no antitumor activity in patients with relapsed and cisplatin-refractory GCT [22], while the same agent in combination with a platinum-derived drug (cisplatin, nedaplatin or oxaliplatin) has demonstrated a very high response rate in these patients [23], [24]. In addition, the combination of epirubicin, a drug abandoned in GCTs, and cisplatin has recently showed an interesting activity for patients with cisplatin-refractory GCT patients [25]. Even if differences in the patient characteristics and in the definition of clinical resistance to cisplatin in these studies need to be considered, these results support a synergistic interaction between platinum-derived drugs and other agents including irinotecan and epirubicin in cisplatin-refractory GCTs, All these regimens need to be considered in the armamentarium of the oncologists. In fact, the natural history of cisplatin-refractory GCT is often characterized by multiple lines of chemotherapy, because of young age of patients and the residual chemosensitivity (mainly platinum-sensitivity) in multiple-relapse GCT, so that it is important to have more therapeutic chances for these patients.

Gemcitabine and oxaliplatin combination may offer an opportunity for patients with testicular cisplatin-refractory GCT to achieve long-term survival. Further investigation with other agents and/or drug combinations is needed in patients with extragonadal cisplatin-refractory disease, in particular in cases with mediastinal primary site.

Editorial Comment

Philippe Beuzeboc

The treatment of cisplatin-refractory germ cell tumours remains a major therapeutic challenge. Long-term survival in these patients is rarely achieved. The results of the French multicentre sequential high-dose chemotherapy TAXIF trial [1] confirm that highly refractory patients clearly do not benefit from high-dose chemotherapy. The evaluation of new drugs and combinations represents a priority.

Both single-agent gemcitabine and oxaliplatin have been identified as active with response rates of approximately 15–20% but no long-term remission was reported. The association of gemcitabine and oxaliplatin appeared to be one of the most interesting approaches. This study is the third report on GEMOX in the literature. The results subsequently confirm the data published by Kollmansberger [2] and Pectasides [3] using the same combination in a comparable patient population. The most important message, however, is that some patients in all three studies remain in continuous remission with or without additional surgery. In conclusion, the combination of these clinical reports demonstrates the feasibility of this salvage association with an acceptable toxicity profile despite the prior use of high-dose chemotherapy in these patients. Based on these highly promising results, this association should be incorporated earlier in the treatment of cisplatin-refractory germ cell tumours.

References 

[1]. [1]Lotz JP, Bui B, Gomez F, et al. Sequential high-dose chemotherapy protocol for relapsed poor prognosis germ cell tumors combining two mobilization and cytoreductive treatments followed by three high-dose chemotherapy regimens supported by autologous stem cell transplantation. Results of the phase II multicentric TAXIF trial. Ann Oncol. 2005;16:411–418. MEDLINE | CrossRef

[2]. [2]Kollmannsberger C, Beyer J, Liersch R, Schoeffski P, Metzner B, Hartmann JT, et al. Combination chemotherapy with gemcitabine plus oxaliplatin in patients with intensively pretreated or refractory germ cell cancer: a study of the German Testicular Cancer Study Group. J Clin Oncol. 2004;22:108–114. CrossRef

[3]. [3]Pectasides D, Pectasides M, Farmakis D, Aravantinos G, Nikolaou M, Koumpou M, et al. Gemcitabine and oxaliplatin (GEMOX) in patients with cisplatin-refractory germ cell tumors: a phase II study. Ann Oncol. 2004;15:493–497. MEDLINE | CrossRef