A phase II study of cisplatin, etoposide and gemcitabine in an unfavourable group of patients with carcinoma of unknown primary site

C. Balaña1,+, J.-L. Manzano1, I. Moreno3, B. Cirauqui1, A. Abad1, A. Font1, J.-L. Mate2 and R. Rosell1

1 Medical Oncology Service, 2 Pathology Service, Hospital Germans Trias i Pujol, Barcelona; 3 Hospital Municipal de Badalona, Barcelona, Spain

Received 12 February 2003; revised 28 March 2003; accepted 14 April 2003


    Abstract
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Background:

The aim of this phase II study was to determine toxicity, response rate, time to progression, and overall survival of cisplatin, etoposide and gemcitabine in patients with carcinoma of unknown primary tumour site.

Patients and methods:

Thirty patients with no previous chemotherapy and not belonging to a treatable group were treated with cisplatin 70 mg/m2 on day 1, etoposide 70 mg/m2 on days 1 and 2, and gemcitabine 700 mg m2 on days 1 and 8, administered every 3 weeks. Stable or responding patients received a maximum of eight cycles. Twenty patients (67%) had more than three affected sites, and 25 patients (84%) had adenocarcinomas.

Results:

Overall response rate was 36.6% (11 patients), including four complete responses (13.3%) and seven partial responses (23.3%), with a 95% confidence interval of 19.9–56. Median survival was 7.21 months and eight patients remained alive for >1 year. Myelosuppression was the most important toxicity, with grade 3–4 neutropenia in 18 patients (60%) in 32% of the cycles: eight patients had neutropenic fever and 10 patients had thrombopenia in 11% of cycles. No non-haematological grade 4 toxicity occurred.

Conclusions:

Cisplatin, etoposide and gemcitabine is an active combination, inducing objective responses in a subset of heavily advanced disease patients with carcinoma of unknown primary site. The role of adding gemcitabine to cisplatin and etoposide remains to be resolved as to the best schedule to diminish toxicity for the three-drug combination.

Key words: chemotherapy, gemcitabine, unknown primary site


    Introduction
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Three per cent of cancer patients are diagnosed with widespread metastatic disease with no identifiable site of origin in spite of an extensive workout, so they are allocated to the ‘carcinoma of unknown primary site’ (CUPS) entity. Diagnostic studies should be targeted to identify those primary malignancies that can benefit from treatment, such as breast or prostate cancer, as well as those whose particular treatments have shown positive clinical results in certain subgroups. These include women with adenocarcinoma involving axillary lymph nodes that should be treated as breast carcinoma [1, 2], women with peritoneal carcinomatosis that should be treated with platinum-based chemotherapy [3], as well as poorly differentiated carcinomas with lymph nodes and midline distribution [4], and squamous cell carcinoma involving cervical lymph nodes that must be treated as head and neck cancer [5, 6]. Pathology diagnosis with immunochemistry should classify the disease into one of the major pathological entities: squamous cell, adenocarcinoma, poorly differentiated carcinoma or neuroendocrine carcinoma, ruling out lymphoma, sarcoma or melanoma origins.

In spite of a reasonable diagnostic approach to CUPS, 68% of patients remain with an unidentified primary site [7], and even autopsy studies fail to identify it in 15–25% of cases [8]. Classically, these patients are treated with empirical chemotherapy. Monochemotherapy for CUPS yields a small percentage of responses (0–19%). With the advent of cisplatin, a slight increase in objective responses (20–33%) with median survivals of 4–7 months was obtained. The combination of cisplatin or carboplatin with etoposide has shown activity, and has been used extensively in this setting with responses of 23–32% (none complete) and median survival of 5–8 months [911]. New drugs such as taxanes have been added to this combination with a slight improvement in results, such as in the combination of taxol, carboplatin and etoposide proposed by Hainsworth et al. [12], which yielded objective responses of 47% with median survival of 13.4 months. Gemcitabine, an antimetabolite, has shown activity in lung [13, 14], pancreatic [15, 16] and biliary tract carcinomas [17], frequent origins of CUPS in necropsy series [8]. We designed this phase II study using cisplatin, etoposide and gemcitabine in order to test the addition of gemcitabine to cisplatin–etoposide in previously untreated CUPS.


    Patients and methods
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Eligibility criteria
The present study was carried out at the University Hospital ‘Germans Trias i Pujol’ and the Municipal Hospital of Badalona (both in Barcelona, Spain) from December 1997 to March 2002. All consecutively admitted patients with a diagnosis of unknown primary carcinoma, fulfilling the eligibility criteria, entered the study.

Patients were eligible if they were <70 years old, with multifocal not locally treatable disease and measurable metastatic tumour of an unidentifiable primary origin in spite of a complete examination. This included: physical examination including gynaecological in women, and global head and neck in men; chemistry profile; computed tomography scan of the chest, abdomen and pelvis; mammography in women; and investigation of any symptomatic areas such as colonoscopy, fibrogastroscopy and bronchoscopy, as indicated by patient symptoms. All pathological diagnoses were eligible, as well as differentiated adenocarcinoma, poorly differentiated adenocarcinoma (PDA), poorly differentiated carcinoma (PDC) and squamous cell carcinoma.

Treatment had to be the first-line therapy for their illness. Patients had to have a ECOG performance status (PS) <3, and adequate bone marrow function (white blood cell count >=4000/µl and platelets >=100 000/µl), normal liver function (bilirubin <=1.5 mg/dl) and normal renal function (creatinine <=1.5 mg/dl). Patients with brain metastasis were also eligible for the protocol if they had undergone surgery and/or radiotherapy treatment previous to chemotherapy. Painful bone lesions could be treated with palliative radiotherapy before the beginning of the treatment. Written informed consent was obtained prior to their participation and the study was approved by the hospitals’ ethics committees.

Patients were excluded if they had treatable tumours, such as women with adenocarcinoma involving only axillary lymph nodes or the peritoneal cavity, patients with squamous cell carcinoma involving only cervical lymph nodes, patients with a suspected germ cell tumour indicated by clinical characteristics and patients with prostate-specific antigen elevated in their plasma or stained in their tumour. Patients with only massive liver disease were not included in the protocol. Patients were not eligible if they had had either a recent heart attack, congestive heart failure or other severe coexistent medical illness; a history of previous malignancy within 5 years, with the exception of skin carcinoma or cervical carcinoma in situ; or pregnancy or lactation.

Treatment
Chemotherapy was as follows: cisplatin 70 mg/m2 on day 1, etoposide 70 mg/m2 on days 1 and 2, and gemcitabine 700 mg/m2 on days 1 and 8. Cycles were repeated every 21 days. Hydration with mannitol was administered on day 1. Ondasentron and dexamethasone were administered on day 1, and oral treatment for delayed emesis was recommended for four consecutive days after cisplatin. Haematological control was performed weekly, and biochemistry with creatinine clearance on day 15. The blood counts on day 1 of each course were used to modify doses as follows: if neutrophil count was >2.0 x 109/l and platelet count was >=75 x 109/l, all drugs were given at full dose. If not, treatment was delayed 1 week up to a maximum of 2 weeks, and then patients were once again treated with a full dose of all agents. In the case of no recuperation within 2 weeks, the patient was removed from the protocol. Dose reductions based on the day-8 blood counts were as follows: gemcitabine was omitted if the neutrophil count was <2.0 x 109/l and the platelet count was <75 x 109/l; doses of cisplatin on day 1 were reduced to 75 or 50% in the case of an alteration of the creatinine clearance.

The use of granulacyte colony-stimulating factors (G-CSFs) was recommended for subsequent cycles in the case of neutropenic fever, and eritropoetin was used in the case of anaemia that required blood support. Patients who were hospitalised for treatment of neutropenia and fever in spite of G-CSF received 75% doses of all drugs. Patients were evaluated every cycle for non-haematological toxicity. Toxicity was graded according to the World Health Organisation common toxicity criteria version 2.0 (1999).

Response evaluation was performed after three cycles. Patients continued on treatment for six cycles in the case of stable disease and good tolerance, and for eight cycles in the case of an objective response and the absence of grade 4 toxicities.

End points
End points of the study were response rate, time to progression and overall survival. Efficacy has been analysed on an intention-to-treat basis. Complete response (CR) was defined as the complete disappearance of all clinical and radiological detectable disease for at least 4 weeks. Partial response (PR) required at least a 50% reduction in the size of all measurable lesions, as measured by the product of the greatest length and greatest width, with no new lesions appearing. Stable disease (SD) required a reduction of <50% or an increase of <25% in the size of measurable lesions, with no new lesions appearing. Patients with progressive disease (PD) had developed either any type of new lesions, or an increase of >=25% in the size of any existing measurable lesion.

Time to progression was measured from day 1 of the first cycle until progression or death. Overall survival was measured from day 1 of the first cycle until death. Follow-up at the end of treatment was performed every 2 months until progression, at which time, if the primary tumour was discovered or the patient had a good PS, treatment was left to the physician’s criteria. A statistical package (SPSS version 10.0) was used for statistics and descriptive factors. Actuarial survival curves were constructed using the Kaplan–Meier method. Quality of life end points and cost analyses were not included as part of this phase II trial.


    Results
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Thirty patients were treated from December 1997 to July 2002. Patient characteristics are listed in Table 1. PS was 1 in 80 patients (60%), and four patients exhibited symptomatic disease with a PS of 2. Three patients had one or two affected sites: two patients with diffuse lymph node involvement and one patient with multiple lung metastases. Approximately 70% of patients had three or more sites of disease, including nine patients (30%) with bone metastasis, six patients (20%) with liver metastasis and three patients (10%) with multiple brain metastases. Twenty-five patients (84%) had adenocarcinomas, 10 (34%) of them poorly differentiated.


View this table:
[in this window]
[in a new window]
 
Table 1. Patient characteristics (n = 30)
 
The total number of cycles was 139, with a median of four cycles per patient. Three patients received only one cycle, as they presented a rapid progression of disease. All three patients had an initial PS of 1 and three or more metastatic sites. One of them died of liver dysfunction due to massive disease, one died of brain metastasis progression and the other died of a cardiac tamponade. They have been considered as PD in the analysis results.

Efficacy
Overall response rate was 36.6% (11 patients), which included four CR (13.3%) and seven PR (23.3%) with a 95% confidence interval (CI) of 19.9% to 56%. Moreover, seven patients (23.3%) had SD and 12 (40%) progressive tumours. Complete response patients all had three or more metastatic sites and a PS of 1 at the beginning of treatment. Figure 1 shows actuarial survival for all 30 patients. Time to progression for patients with an objective response (PR, CR) was 9.01 months (95% CI 8.02–10.0), significantly better than patients with progression (2.65 months, 95% CI 2.23–3.08). Time to progression for CR patients was 8.95 months (95% CI 4.91–12.98). Median survival was 7.21 months (95% CI 2.59–11.83) for all patients, and eight patients (26%) remained alive for >1 year, ranging from 12 to 29 months (Figure 1).



View larger version (11K):
[in this window]
[in a new window]
 
Figure 1. Overall survival. Median survival was 7.21 months.

 
Toxicity
The most important toxicity was myelosuppression, as 18 patients (60%) presented grade 3–4 neutropenia in 32% of cycles, and eight patients (27%) had neutropenic fever (Table 2). Six patients (20%) needed G-CSF to continue treatment in 14% of cycles. Ten patients (33%) presented asymptomatic grade 3–4 thrombopenia in 11% of cycles. Moreover, anaemia was a common event and was grade 3–4 in 14 patients (47%); all required blood support or eritropoietin. There were no treatment-related deaths.


View this table:
[in this window]
[in a new window]
 
Table 2. Treatment-related toxicity by patient (n = 30)
 
Non-haematological toxicity was controllable with no grade 4 events (Table 3). Seven patients (23%) presented grade 3 nausea and vomiting, and five patients (16%) grade 3 diarrhoea in six cycles. Asthenia was a common event in 21 patients (70%) but only grade 1–2. Two patients had grade 1–2 hearing loss, and one patient reversible liver toxicity of grade 3.


View this table:
[in this window]
[in a new window]
 
Table 3. Haematological toxicity
 
Two patients with a PR had a primary site identified: one at follow-up (renal carcinoma) and the other at necropsy (pancreatic carcinoma).


    Discussion
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
The combination of cisplatin, etoposide and gemcitabine is active in patients with an unknown primary site and poor prognostic features. Thirty-three per cent of our patients had more than three metastatic sites and 13.3% had a PS of 2. Furthermore, we included in this group three patients (10%) with brain metastasis, nine (30%) with bone metastasis and six (20%) with liver metastasis. In spite of this non-selective group of patients, we found seven PR (23.3%) and four CR (13.3%), for an overall response rate of 36.6% (95% CI 19.9% to 56%). Moreover, seven patients (23.3%) had SD. Time to progression for patients with an objective response (PR, CR) was 9.01 months. In spite of this result, median survival for the entire group was 7.21 months because three patients died prematurely due to rapid progression of their illness. Twenty-six per cent of patients remained alive for >1 year.

These results seem to be worse than those from the combination of carboplatin, etoposide and taxol studied by Hainsworth et al. [12]. However, in that study of 55 patients, brain metastasis patients were excluded and, although they attained a 47% overall response rate and 13% CR, 18 patients (33%) in their study had only one metastatic site and 14 (25%) had only two sites; only two patients had bone metastasis. The group of patients with only one or two metastatic organs show the best survival (40 months) in retrospective studies, especially if they have no liver, bone, and pleural or adrenal lesions [18]. The patients included in our study had worse prognostic characteristics, as described previously.

The toxicity pattern of our study was slightly better than that of Hainsworth et al. [12]. We had no grade 4 non-haematological toxicity, and only grade 3–4 leucopenia in 37% of patients and 13% of cycles. This was in contrast to Hainsworth et al.’s 76% of patients and 41% of cycles, although 60% of our patients had grade 3–4 neutropenia. Thrombopenia was similar: 33% of patients and 11% of cycles in our study, as compared with 26% of patients and 10% of cycles with the addition of taxol to carboplatin and etoposide. Forty-seven per cent of our patients needed blood support, as compared with 38% in the Hainsworth study. A re-analysis of the same cohort of patients with more cases and longer follow-up reproduced similar results [19].

Briasoulis et al. [20] studied a carboplatin and taxol treatment of CUPS, and attained an overall response rate of 38.7%. However, they included a subset of patients with good prognostic features: 16 women (68.4%) with peritoneal carcinomatosis had responses and contributed to the better results. This group of patients is commonly treated as ovarian cancer, with similar good response and prognosis with chemotherapy.

Other authors have not reproduced the results obtained with paclitaxel and platinum compounds with or without etoposide, in unpublished small reports [21, 22].

Another taxane, docetaxel, has also been used in this setting: 26 patients were treated with docetaxel and cisplatin, and 45 with docetaxel and carboplatin, in a recent non-randomised selective patients trial. The activity of this combination seems to be inferior to paclitaxel and carboplatin (overall responses of 26% and 22%, respectively) with median survivals of 8 months in the two groups. Once again, no brain or meningeal metastases were included, and 81% and 75%, respectively, of patients had only one or two metastatic sites [23].

The role of taxanes in this setting has not been established, and the absence of activity in pancreatic and colorectal tumours makes its use controversial in this heterogeneous group of patients.

Gemcitabine, a recent new antimetabolite compound, with a narrow antitumour activity, could be of use for these tumours. As a second-line treatment in pre-treated CUPS patients, Hainsworth et al. [24] observed PR of 8% and minor responses or SD of 25% with an improvement in symptoms and good tolerance. Recently, The Minnie Pearl Cancer Research Network Study evaluated the combination of gemcitabine, carboplatin and paclitaxel in 120 patients. Fifty-one per cent of them had multiple liver metastases and 25% had objective responses; survival was 42% at 1-year follow-up and 23% at 2 years [25]. The characteristics of the patients included in this study seem worse than in previous studies of the same authors, where only 16–19% of patients presented liver metastasis—a known poor prognostic characteristic in this setting [26]. Toxicity was also more manageable than in the previous studies and similar to our study: 46% grade 3–4 leucopenia in 32% of cycles, and 43% thrombopenia in 14% of cycles.

Gemcitabine, cisplatin and etoposide appears to be a good combination for use in this pathology. The best schedule to be used needs further study. A previous study in small-cell lung cancer used the same drugs at different doses, as etoposide was administered orally; haematological toxicity was also dose-limiting [27]. The use of gemcitabine seems convenient in this subset of patients, as it has demonstrated activity in pancreatic and lung carcinoma. However, in spite of these encouraging results, further research into the optimal number of cycles, and the impact of chemotherapy on quality of life, is warranted. Furthermore, the identification of origin of the disease by the analysis of gene expression will probably help us to identify and individualise the best treatment for each patient [28].


    Acknowledgements
 
We thank Toni de la Riva for data management and Elizabeth Breedlove for assistance with the manuscript.


    Footnotes
 
+ Correspondence to: Dr C. Balaña, Medical Oncology Service, Hospital Universitari ‘Germans Trias i Pujol’, Carretera Canyet s/n, 08916 Barcelona, Spain. Tel: +34-93-497-89-25; Fax: +34-93-497-89-50; E-mail: cbalana{at}ns.hugtip.scs.es Back


    References
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
1. Baron PL, Moore MP, Kinne SW et al. Occult breast cancer presenting with axillary metastases. Updated management. Arch Surg 1990; 125: 210–214.[Abstract]

2. Ellerbroek N, Holmes F, Singletary E et al. Treatment of patients with isolated axillary nodal metastases from an occult primary carcinoma consistent with breast origin. Cancer 1990; 66: 1461–1467.[ISI][Medline]

3. Strand C, Grosh WW, Baxter JW. Peritoneal carcinomatosis of unknown primary site in women. Ann Intern Med 1989; 111: 213–217.[ISI][Medline]

4. Hainsworth JD, Dial TW, Greco FA. Curative combination chemotherapy for patients with advanced poorly differentiated carcinoma of unknown primary site. Am J Clin Oncol 1988; 11: 138–145.[ISI][Medline]

5. Nguyen C, Shenouda G, Black MJ et al. Metastatic squamous cell carcinoma to cervical lymph nodes from unknown primary mucosal sites. Head Neck 1994; 16: 58–63.[ISI][Medline]

6. Wang RC, Goepfert H, Barber AE, Wolf P. Unknown primary squamous cell carcinoma metastatic to the neck. Arch Otolaryngol Head Neck Surg 1990; 116: 1388–1393.[Abstract]

7. Abbruzzese JL, Abbruzzese MC, Lenzi R et al. Analysis of a diagnostic strategy for patients with suspected tumors of unknown origin. J Clin Oncol 1995; 13: 2094–2103.[Abstract]

8. Le Chevalier T, Cvitkovic E, Caille P et al. Early metastatic cancer of unknown primary origin at presentation: a clinical study of 302 consecutive autopsied patients. Arch Intern Med 1988; 148: 2035–2039.[Abstract]

9. Hainsworth JD, Johnson DH, Greco FA. Cisplatin-based combination chemotherapy in the treatment of poorly differentiated carcinoma and poorly differentiated adenocarcinoma of unknown primary site. Results of a 12-year experience. J Clin Oncol 1992; 10: 912–922.[Abstract]

10. Warner E, Goel R, Chang J et al. A multicentre phase II study of carboplatin and prolonged oral etoposide in the treatment of cancer of unknown primary site. Br J Cancer 1998; 77: 2376–2380.[ISI][Medline]

11. Voog E, Merrouche Y, Trillet-Lenoir V et al. Multicentric phase II study of cisplatin and etoposide in patients with metastatic carcinoma of unknown primary. Am J Clin Oncol 2000; 23: 614–616.[CrossRef][ISI][Medline]

12. Hainsworth JD, Erland JB, Kalman LA et al. Carcinoma of unknown primary site: treatment with 1-hour paclitaxel, carboplatin and extended schedule etoposide. J Clin Oncol 1997; 15: 2385–2393.[Abstract]

13. Kelly K. The role of single-agent gemcitabine in the treatment of non-small-cell lung cancer. Ann Oncol 1999; 10 (Suppl 5): 53–56.[Abstract]

14. Manegold C, Bergman B, Chemaissani A et al. Single-agent gemcitabine versus cisplatin–etoposide: early results of a randomised phase II study in locally advanced or metastatic non-small-cell lung cancer. Ann Oncol 1997; 8: 525–529.[Abstract]

15. Rothenberg MI, Moore MJ, Cripps MC et al. A phase II trial of gemcitabine in patients with 5-FU-refractory pancreas cancer. Ann Oncol 1996; 7: 347–353.[Abstract]

16. Burris HA, Moore MJ, Andersen J et al. Improvements in survival and clinical benefit with gemcitabine as first-line therapy for patients with advanced pancreas cancer: a randomized trial. J Clin Oncol 1997; 15: 2403–2413.[Abstract]

17. Scheithauer W. Review of gemcitabine in biliary tract carcinoma. Semin Oncol 2002; 29: 40–45.

18. Hess KR, Abbruzzese MC, Lenzi R et al. Classification and regression tree analysis of 1000 consecutive patients with unknown primary carcinoma. Clin Cancer Res 1999; 5: 3403–3410.[Abstract/Free Full Text]

19. Greco FA, Burris HA, Erland JB et al. Carcinoma of unknown primary site. Long term follow-up after treatment with paclitaxel, carboplatin and etoposide. Cancer 2000; 89: 2655–2660.[CrossRef][ISI][Medline]

20. Briasoulis E, Kalofonos H, Bafaloukos D et al. Carboplatin plus paclitaxel in unknown primary carcinoma: a phase II Hellenic Cooperative Oncology Group Study. J Clin Oncol 2000; 18: 3101–3107.[Abstract/Free Full Text]

21. Lastra E, Muñoz A, Rubio I et al. Paclitaxel, carboplatin and oral etoposide in the treatment of patients with cancer of unknown primary site. Proc Am Soc Clin Oncol 2000; 19: 579a (Abstr 2279C).

22. Elkordy M, Berrry W, Khan M et al. A Phase II study of weekly paclitaxel (taxol) and carboplatin in advanced carcinoma of unknown primary origin. Proc Am Soc Clin Oncol 2001; 20: 402a (Abstr 1603).

23. Greco FA, Erland JB, Morrissey LH et al. Carcinoma of unknown primary site: phase II trials with docetaxel plus cisplatin or carboplatin. Ann Oncol 2000; 11: 211–215.[Abstract]

24. Hainsworth JD, Burris HA, Calvert SW et al. Gemcitabine in the second-line therapy of patients with carcinoma of unknown primary site: a phase II trial of the Minnie Pearl Cancer Research Network. Cancer Invest 2001; 19: 335–339.[CrossRef][ISI][Medline]

25. Greco FA, Burris HA, Litchy S et al. Gemcitabine, carboplatin and paclitaxel for patients with carcinoma of unknown primary site: a Minnie Pearl Cancer Research Network Study. J Clin Oncol 2002; 20: 1651–1656.[Abstract/Free Full Text]

26. Ayoub JP, Hess KR, Abbruzzese R et al. Unknown primary tumors metastatic to liver. J Clin Oncol 1998; 16: 2105–2112.[Abstract]

27. Earle CC, Stewart DJ, Cormier Y et al. A phase I study of gemcitabine/cisplatin/etoposide in the treatment of small-cell lung cancer. Lung Cancer 1998; 22: 235–241.[CrossRef][ISI][Medline]

28. Dennis JL, Vass JK, Wit EC et al. Identification from public data of molecular markers of adenocarcinoma characteristics of the site of origin. Cancer Res 2002; 62: 5999–6005.[Abstract/Free Full Text]