Phase II study of cisplatin preceding gemcitabine in patients with advanced oesophageal cancer

J. R. Kroep1,*, H. M. Pinedo1, G. Giaccone1, A. Van Bochove2, G. J. Peters1 and C. J. Van Groeningen1

1 Department of Medical Oncology, VU Medical Centre, Amsterdam; 2 Department of Internal Medicine, De Heel-Zaans Medisch Centrum, Zaandam, The Netherlands

Received 5 July 2003; revised 28 September 2003; accepted 6 October 2003;


    ABSTRACT
 Top
 ABSTRACT
 Introduction
 Patients and methods
 Results
 Discussion
 REFERENCES
 
Background:

For oesophageal cancer there is no effective standard therapy. We studied the feasibility and efficacy of the cisplatin–gemcitabine combination chemotherapy in patients with unresectable oesophageal cancer.

Patients and methods:

Thirty-six chemonaïve patients with unresectable or metastatic oesophageal adenocarcinoma (24) or squamous-cell-carcinoma (12) were treated with cisplatin (50 mg/m2, days 1 and 8), followed by gemcitabine (800 mg/m2, days 2, 9 and 16), every 28 days. Feasibility and efficacy were studied.

Results:

Toxicity was substantial but manageable. A median number of four therapy cycles was given. The most frequent grade ≥3 toxicities were leukopenia (75%) and neutropenia (83%). Three patients developed neutropenic fever. Grade 3/4 thrombocytopenia occurred in 24 out of 36 patients (67%), but did not result in serious bleeding disorders. Myelotoxicity was cumulative and required omission of gemcitabine on day 16 in 63% of cycles. Anaemia required treatment with erythropoietin, red blood cells or both in 81% of patients. Non-haematological toxicity consisted mainly of grade 1/2 nausea/vomiting or fatigue. Fourteen out of 34 evaluable patients had a major objective response (41%; two complete and 12 partial responses). The median actuarial survival was 9.8 months.

Conclusion:

This cisplatin–gemcitabine regimen was feasible, with myelosupression being the main toxicity, and had significant activity in patients with advanced oesophageal cancer.

Key words: adenocarcinoma, advanced oesophageal cancer, cisplatin, gemcitabine, phase II, squamous cell carcinoma


    Introduction
 Top
 ABSTRACT
 Introduction
 Patients and methods
 Results
 Discussion
 REFERENCES
 
Treatment results for patients with advanced or metastatic oesophageal cancer are unsatisfactory. For patients with locoregional oesophageal cancer, survival is poor after surgery due to local and systemic recurrence, while most patients already have advanced disease at diagnosis [1, 2]. Oesophageal cancer is increasing in incidence, with a shift in histological type from squamous cell carcinoma (SCC) to adenocarcinoma [35]. The few studies that have been reported in metastatic adenocarcinoma of the oesophagus indicate a similar chemosensitivity for adenocarcinoma and SCC [6]. Cisplatin-based chemotherapy regimens appear to be the most active ones in advanced oesophageal cancer, with response rates of 25%–50% [7]. A 35% to 40% response rate has been reported for the combination of cisplatin and fluorouracil (5FU) [8, 9]. However, responses to chemotherapy are generally short lived, lasting only 4–6 months.

These poor results underscore the need to identify new agents in the treatment of oesophageal cancer. Gemcitabine (2',2'-difluorodeoxycytidine), a deoxycytidine analogue, is attractive for combination chemotherapy due to its activity in various solid tumours and mild toxicity profile. Gemcitabine acts by incorporation of its triphosphate (dFdCTP) into DNA, subsequently leading to inhibition of DNA replication and repair. Several self-potentiating mechanisms have been described [1012], enhancing the incorporation of dFdCTP into DNA and possibly also into RNA.

Preclinical studies showed schedule-dependent additive and synergistic interactions between cisplatin and gemcitabine [1315]. Cisplatin–gemcitabine combination chemotherapy is recognized as an active treatment in various solid tumours [12]. In oesophageal cancer this combination has not been studied previously. Based on results of our schedule finding phase I/II study, comparing four alternating sequences with a 4- and 24-h interval of cisplatin and gemcitabine, the schedule with cisplatin (50 mg/m2, days 1 and 8) followed after 24 h by gemcitabine (800 mg/m2, days 2, 9 and 16) was chosen for phase II evaluation [16, 17].

The aim of this study was to determine the efficacy and toxicity of patients with metastatic or unresectable oesophageal cancer treated with cisplatin and gemcitabine.


    Patients and methods
 Top
 ABSTRACT
 Introduction
 Patients and methods
 Results
 Discussion
 REFERENCES
 
Patient selection
Patients were considered eligible for this study based on the following criteria: a histologically confirmed diagnosis of unresectable or metastatic SCC, or adenocarcinoma of the oesophagus; no prior chemotherapy; evaluable or measurable disease; age ≥18 years; a performance status <2 according to the World Health Organisation (WHO) scale; life expectancy of ≥3 months; adequate bone marrow function [white blood cells (WBC) ≥3 x 109/l, absolute neutrophil count ≥1.5 x 109/l, platelet count ≥100 x 109/l]; adequate renal function (serum creatinine <120 µmol/l or creatinine clearance ≥60 ml/min); adequate liver function (serum bilirubin <25 µmol/l); adequate cardiac function; no second tumour, other than curatively treated non-melanoma skin cancer or carcinoma-in situ of the cervix; and no brain metastasis. All patients had to give written informed consent before entering the study. The study was approved by the ethical review board of both participating hospitals.

Pretreatment evaluation included medical history and clinical examination. In addition, haematology and blood chemistry tests, urinanalysis, electrocardiogram and chest radiograph were performed. Pretreatment staging consisted of computed tomography (CT) of the chest and abdomen. Before each treatment cycle, a complete blood count and chemistries were repeated. Complete blood cell counts and serum creatinine were measured weekly. Repeat radiographic studies were performed every other cycle.

Study design and drugs
Cisplatin 50 mg/m2 was administered as a 1 h infusion on days 1 and 8, and gemcitabine (Gemzar®; 2',2'-difluoro-2'-deoxycytidine) 800 mg/m2 over 30 min weekly on days 2, 9 (24 h after cisplatin) and 16 on a 4-weekly schedule. Before cisplatin, patients received i.v. hydration with 1000 ml normal saline plus 20 mmol potassium chloride and 2 g magnesium sulfate over 2 h. After cisplatin infusion, 4000 ml normal saline plus 80 mmol potassium chloride and 8 g magnesium sulfate were given over 24 h. Prophylactic antiemetics, 8 mg ondansetron and 8 mg dexamethasone, were administered twice on the day of cisplatin infusion. No prophylactic use of granulocyte colony-stimulating factor (G-CSF) was planned.

Toxicity was evaluated before each course and graded according to the National Cancer Institute Common Toxicity Criteria (NCI-CTC; version 2.0). Treatment was delayed if grade ≥3 WBC toxicity and/or grade ≥2 thrombocytopenia occurred. If blood counts did not permit gemcitabine administration on day 16, this was omitted. If treatment delay was >2 weeks, the patient went off study. Dose modifications were not applied for both gemcitabine and cisplatin. Treatment was discontinued for disease progression, unacceptable toxicity or the patient’s wish to stop the treatment. Six cycles were given to responding patients, and patients with stable disease when feasible.

Response to therapy was assessed, according to the WHO criteria, every two courses. To qualify for response analysis, patients had to complete at least one cycle of treatment. Duration of response was defined as time from documentation of the response to documentation of disease progression. Time to progression (TTP) was defined as time from start of treatment to documentation of disease progression. Survival was defined as time from start of treatment until death.

Toxicity analysis and statistics
The primary end point of the study was response rate. The study was designed using the Gehan design [18]. Assuming a response rate of at least 20%, initially 14 patients were accrued. If the response rate was at least 20%, then the probability of obtaining no responses in 14 patients is <0.05. Accrual was planned for a total of 35 patients.

Correlations were determined using the Pearson or Spearman analysis for normal and non-parametric data, respectively. The computer program SPSS (Version 7.5; SPSS Inc., Chicago, IL) was used for statistical analysis.


    Results
 Top
 ABSTRACT
 Introduction
 Patients and methods
 Results
 Discussion
 REFERENCES
 
Patient characteristics
Patient characteristics are summarized in Table 1. Thirty-six patients entered the study between October 1997 and September 2001. Twenty-four patients had adenocarcinoma and 12 had SCC. Thirty-three patients (92%) had metastatic disease, with the majority having distant lymph node [26/33 (79%)] and liver metastasis [12/33 (36%)]. Three patients had an unresectable primary tumour. Prior oesophagectomy with curative intent had been performed in five patients. No patients had prior chemotherapy while three patients (SCC) had prior radiotherapy.


View this table:
[in this window]
[in a new window]
 
Table 1. Patient characteristics
 
Thirty-five patients were assessable for response. One patient was not evaluable for response due to delay of treatment and response evaluation after a gastric bleeding requiring surgery. Of the 35 evaluable patients, 33 had bidimensionally measurable disease. The remaining two patients (both with stable disease) could only be evaluated unidimentionally due to prior surgery and tumour localization, respectively.

Toxicity
All patients were evaluable for toxicity. Toxicity was manageable and no treatment-related deaths were observed. Myelosuppression was the most important toxicity. Table 2 lists the haematological toxicity of the first two and of all delivered treatment cycles. Over all cycles, grade 3 or 4 thrombocytopenia was observed in 24 out of 36 patients (67%). No serious bleeding disorders occurred due to thrombocytopenia. Four patients with platelets <10 x 109/l received a platelet transfusion. One patient had a bleeding gastric ulcer requiring surgery without thrombocytopenia, which was therefore probably not treatment related. Over all cycles, grade 3 and 4 leukopenia was observed in 27 out of 36 patients (75%) and neutropenia in 24 out of 29 (83%) evaluable patients. The neutrophils were not always measured on day 15 and were therefore evaluable in a lower number of patients. Three patients received secondary G-CSF after two (two patients) or four (one patient) cycles. Neutropenic fever was observed in three patients. Anaemia occurred commonly and required treatment with erythropoietin, red blood cells or both in 81% of patients. Myelotoxicity [mean ± standard error of the mean (SEM)] was cumulative for both platelets (Figure 1A), with day 15 values decreasing from 122 ± 19 in the first cycle to 52 ± 8 in the fourth cycle (P = 0.0001), and for WBCs (Figure 1B) with nadir values from 2.9 ± 0.6 in the first cycle to 2.4 ± 0.2 in the fourth cycle (P = 0.22).


View this table:
[in this window]
[in a new window]
 
Table 2. Haematological toxicity during the first two cycles and overall toxicity
 


View larger version (21K):
[in this window]
[in a new window]
 
Figure 1. (A) Accumulation of thrombocytopenia during cisplatin–gemcitabine treatment (mean of all patients ± SEM) for days 1, 8, 16 and 22 of treatment cycles I–VI (CI–CVI). Day 16 values decreased significantly from 129 ± 19 in the first cycle to 58 ± 10 in the fourth cycle (P <0.0001). (B) Accumulation of WBC toxicity during the cisplatin–gemcitabine treatment (mean of all patients ± SEM) for days 1, 8, 16 and 22 of treatment cycles I–VI. Day 16 values decreased from 3.1 ± 0.7 in the first cycle to 2.5 ± 0.3 in the fourth cycle (P = 0.22).

 
A summary of the observed non-haematological toxicity is shown in Table 3. Non-haematological toxicity was mostly mild to moderate, with only a few cases of grade 3 or 4 toxicity. Non-haematological toxicity consisted mainly of grade 1 or 2 nausea/vomiting or fatigue. Most patients had mild hair loss, with only two patients experiencing complete alopecia. Median creatinine clearance, according to Cockroft–Gault, decreased slightly from a pretreatment median value of 90 ml/min to 89.5 ml/min and 81 ml/min after four and six cycles, respectively. One patient with pretreatment dyspnoea on exertion developed shortness of breath at normal levels of activity due to pulmonary embolism. No cardiac toxicity was observed. The onset of neurosensory toxicity and hearing loss were both observed after a median of three cycles. Peripheral oedema was seen in two patients, of which at least in one patient this was tumour related due to thrombosis and lymph node metastases in the pelvis.


View this table:
[in this window]
[in a new window]
 
Table 3. Overall non-haematological toxicity
 
Omissions of gemcitabine on day 16 due to myelotoxicity occurred in 100 out of 159 cycles (63%). Overall, a median number of four treatment cycles was given, ranging from two to six. Twelve patients (33%) completed six therapy cycles. Treatment discontinuation was most related to progression of the disease (10 patients) or treatment toxicity, both myelotoxicity (three) and nephrotoxicity (three). The other patients stopped their treatment due to surgery (three), patient’s wish to stop treatment (four) and a bleeding gastric ulcer requiring surgery (one).

Response
Fourteen out of 34 evaluable patients responded (41%; 95% confidence interval 25% to 59%), including two complete responses and 12 partial responses. One additional patient also responded; however this was not confirmed after 4 weeks due to the patient’s wish to stop treatment. Stable disease was seen in 17 patients and three patients had progressive disease. Responses were observed in five out of 12 patients with SCC (42%) and nine out of 22 patients with adenocarcinoma (41%). Responses were obtained in both metastatic (13/26) and locally advanced disease (1/3). The response by disease site was as follows: primary tumor, 14 patients (all responders); lymph nodes, 11; liver metastasis, five; lung metastasis, two; and ascites, one. The median response duration and time to progression were 5.5 and 6 months, respectively. Responses were seen after a median of two cycles (ranging from one to four cycles).

One complete remission was obtained in a patient with an adenocarcinoma metastasised to the liver. CR was determined by CT, repeated endoscopy and positron emission tomography. The other patient had a T3N1M1 SCC with a positive lymph node at the truncus coeliacus. After two and four cycles, no tumour could be observed by CT or endoscopy, however an endoscopic biopsy of the lymph node at the truncus coeliacus was still positive. Three more patients underwent surgery after chemotherapy. The first patient with a partial response was radically resected, while for the second and third patients, both with stable disease (minor responses), an oesophagectomy was not possible due to locally advanced disease and an abscess in the operation area, respectively. Survival was determined by Kaplan–Meier analysis and is depicted in Figure 2. The median actuarial survival was 9.8 months, ranging from 3 to ≥28 months, with five patients alive after a median follow-up of 11 months.



View larger version (9K):
[in this window]
[in a new window]
 
Figure 2. Kaplan–Meier analysis of patients with unresectable and metastatic adenocarcinoma and SCC of the oesophagus treated with the cisplatin-preceding-gemcitabine regimen. The median actuarial survival was 9.8 months (range 3 to ≥28 months), with five patients still alive (filled diamonds).

 

    Discussion
 Top
 ABSTRACT
 Introduction
 Patients and methods
 Results
 Discussion
 REFERENCES
 
In oesophageal cancer, a combination chemotherapy regimen that includes gemcitabine has not been studied before. We observed substantial activity for the cisplatin-preceding-gemcitabine regimen, with a 41% response rate and a median survival of 9.8 months. This compares favourably to the cisplatin–fluorouracil standard therapy [8, 9, 19] and equals that of the newer combinations with etoposide [20, 21], irinotecan [22] or paclitaxel [23]. However, the median response duration remains disappointing and compares to previous studies.

The epidemiology of oesophageal cancer has changed over the past two decades. Adenocarcinomas are currently the predominant subtype of oesophageal cancer, with a 67% incidence in the current study [1, 24]. Response rates and median survivals for patients with adenocarcinoma and SCC were comparable, in accordance with previous studies [22, 23].

Although gemcitabine administered as a single agent did not show activity in metastatic oesophageal cancer [25], we observed encouraging results for the combination with cisplatin. The activity of the combination might be explained by the synergistic interactions between gemcitabine and cisplatin [1214]. Cisplatin increased gemcitabine incorporation in DNA, as well as DNA strand break formation induced in a gemcitabine-resistant cell line [14]. In addition, gemcitabine also increased the formation of Pt-DNA adducts and decreased the repair of Pt-DNA adducts in a cisplatin-resistant cell line.

There is no clear evidence for the best sequence of administration of the two drugs. Preclinical studies indicated a small advantage in administering the gemcitabine prior to the cisplatin [13, 26]. However, a pharmacological schedule-finding study with cisplatin administered 24 h prior to gemcitabine produced the best pharmacological profile [16], but resulted in more severe leukopenia [17]. In that study we also observed several responses in oesophageal cancer. Based on the pharmacokinetics, we decided to investigate further the cisplatin-preceding-gemcitabine regimen, in contrast to previous clinical studies using the alternative schedule.

Dysphagia is one of the most common symptoms of esophageal cancer. Many of our patients had dysphagia at the start of therapy, and in case of serious dysphagia a stent was placed. We did not consequently record the reponse of dysphagia on therapy, but in general relief of dysphagia seemed to occur in the responding patients after one to four cycles of chemotherapy.

Toxicity was substantial but manageable. There were no treatment-related deaths and only three patients had febrile neutropenia. Toxicity was mainly haematological, with the nadir seen at day 16, as expected [17]. Omissions of gemcitabine administration on day 16 due to myelotoxicity were necessary in 63% of administered cycles. This toxicity profile might be improved by using a 3-weekly schedule. In non-small-cell lung carcinoma, the 3-weekly schedule with gemcitabine administered prior to cisplatin showed good activity and was well tolerated [27, 28].

The development of effective systemic therapy is also a high priority for patients with locally advanced oesophageal cancer eligible for treatment with curative intent. Relative small gains in advanced disease may translate into remarkable improvement in more localized disease. Based on the activity of the gemcitabine–cisplatin combination in advanced oesophageal cancer, we are currently studying the combination as part of combined modality therapy in patients with locally advanced oesophageal cancer.


    Acknowledgements
 
This study was supported financially by Eli Lilly and Co. (International and The Netherlands).


    FOOTNOTES
 
* Correspondence to: Dr J. R. Kroep, Department of Medical Oncology, VU Medical Centre, PO Box 7057, 1007 MB Amsterdam, The Netherlands. Tel: +31-20-444-4300; Fax: +31-20-444-4355; E-mail: jr_kroep{at}hotmail.com Back


    REFERENCES
 Top
 ABSTRACT
 Introduction
 Patients and methods
 Results
 Discussion
 REFERENCES
 
1. Kok TC. Chemotherapy in oesophageal cancer [Review]. Cancer Treat Rev 1997; 23: 65–85.[Medline]

2. Webb A, Cunningham D, Scarffe JH et al. Randomised trial comparing epirubicin, cisplatin, and fluorouracil versus fluorouracil, doxorubicin, and methotrexate in advanced oesophagogastric cancer. J Clin Oncol 1997; 15: 261–267.[Abstract]

3. Blot WJ, Devesa SS, Kneller RW, Fraumeni JF Jr. Rising incidence of adenocarcinoma of the oesophagus and gastric cardia. JAMA 1991; 265: 1287–1289.[Abstract]

4. Macfarlane GJ, Boyle P. The epidemiology of oesophageal cancer in the UK and other European countries. J Soc Med 1994; 87: 334–337.[ISI]

5. Heitmiller RF, Sharma RR. Comparison of prevalence and resection rates in patients with oesophageal squamous cell carcinoma and adenocarcinoma. J Thorac Cardiovasc Surg 1996; 112: 130–136.[Abstract/Free Full Text]

6. Ilson DH, Bains M, Ginsberg RJ et al. Neoadjuvant therapy of oesophageal cancer. Surg Oncol Clin North Am 1997; 6: 732–740.

7. Enzinger PC, Ilson DH, Kelsen DP. Chemotherapy in oesophageal cancer [Review]. Semin Oncol 1999; 26 (Suppl 15): 12–20.

8. Ilson D, Kelsen D. Systemic therapy in oesophageal cancer. In Aisner J, Arriagada R, Green M et al. (eds): Comprehensive Textbook of Thoracic Oncology. Baltimore, MD: William and Wilkins 1996; 630–649.

9. Bleiberg H, Conroy T, Paillot B et al. Randomised phase II study of cisplatin and 5FU versus cisplatin alone in advanced squamous cell oesophageal cancer. Eur J Cancer 1997; 33: 1216–1220.[CrossRef][Medline]

10. Heinemann V, Hertel LW, Grindey GB, Plunkett W. Comparison of the cellular pharmacokinetics and toxicity of 2',2'-difluorodeoxycytidine and 1-ß-D-arabinofuranosylcytosine. Cancer Res 1988; 48: 4024–4031.[Abstract]

11. Huang P, Chubb S, Hertel LW et al. Action of 2',2'-difluorodeoxycytidine on DNA synthesis. Cancer Res 1991; 51: 6110–6117.[Abstract]

12. Peters GJ, van der Wilt CL, van Moorsel CJ et al. Basis for effective combination cancer chemotherapy with antimetabolites [Review]. Pharmacol Ther 2000; 87: 227–253.[CrossRef][ISI][Medline]

13. Bergman AM, Ruiz van Haperen VWT, Veerman G et al. synergistic interaction between gemcitabine and cisplatin in vitro. Clin Cancer Res 1996; 2: 521–530.[Abstract]

14. Van Moorsel CJA, Pinedo HM, Veerman G et al. Mechanisms of synergism between cisplatin and gemcitabine in ovarian and non-small cell lung cancer cell lines. Br J Cancer 1999; 80: 981–990.[CrossRef][ISI][Medline]

15. Van Moorsel CJA, Pinedo HM, Veerman G et al. Scheduling of gemcitabine and cisplatin in Lewis Lung tumour bearing mice. Eur J Cancer 1999; 35: 808–814.[CrossRef][ISI][Medline]

16. Van Moorsel CJ, Kroep JR, Pinedo HM et al. Pharmacokinetic schedule finding study of the combination of gemcitabine and cisplatin in patients with solid tumours. Ann Oncol 1999; 10: 441–448.[Abstract]

17. Kroep JR, Peters GJ, van Moorsel CJ, Catik A et al. Gemcitabine-cisplatin: a schedule finding study. Ann Oncol 1999; 10: 1503–1510.[Abstract]

18. Gehan EA. The determination of the number of patients required in a preliminary an a follow-up trial of a new chemotherapeutic agent. J Chronic Dis 1961; 13: 346–353.[ISI]

19. Lizuka T, Kakegawa T, Ide H et al. Phase II evaluation of cisplatin and 5-fluorouracil in advanced squamous cell carcinoma of the oesophagus: a Japanese oesophageal oncology group trial. Jpn J Clin Oncol 1992; 22: 172–176.[ISI][Medline]

20. Spiridonidis CH, Laufman LR, Jones JJ et al. A phase II evaluation of high dose cisplatin and etoposide in patients with advanced oesophageal adenocarcinoma. Cancer 1996; 78: 2070–2077.[CrossRef][ISI][Medline]

21. Kok TC, Van der Gaast A, Dees J et al. Cisplatin and etoposide in oesophageal cancer: a phase II study. Rotterdam oesophageal tumour study group. Br J Cancer 1996; 74: 980–984.[ISI][Medline]

22. Ilson DH, Saltz L, Enzinger P et al. Phase II trial of weekly irinotecan plus cisplatin in advanced oesophageal cancer. J Clin Oncol 1999; 17: 3270–3275.[Abstract/Free Full Text]

23. Ilson DH, Ajani J, Bhalla K et al. Phase II trial of paclitaxel, fluorouracil, and cisplatin in patients with advanced carcinoma of the oesophagus. J Clin Oncol 1998; 16: 1826–1834.[Abstract]

24. Gammon MD, Schoenberg JB, Ahsan H et al. Tobacco, alcohol, and socio-economic status and adenocarcinomas of the oesophagus and gastric cardia. J Natl Cancer Inst 1997; 89: 1247–1249.[Free Full Text]

25. Sandler AB, Kindler HL, Einhorn LH et al. Phase II trial of gemcitabine in patients with previously untreated metastatic cancer of the oesophagus or gastroesophageal junction. Ann Oncol 2000; 11: 1161–1164.[Abstract]

26. Braakhuis BJ, Ruiz van Haperen VW, Welters MJ, Peters GJ. Schedule-dependent therapeutic efficacy of the combination of gemcitabine and cisplatin in head and neck cancer xenografts. Eur J Cancer 1995; 31: 2335–2340.[CrossRef]

27. Castellano D, Lianes P, Paz-Arez L. A phase II study of novel gemcitabine plus cisplatin regimen administered every three weeks for advanced non-small cell lung cancer. Ann Oncol 1998; 9: 457–459.[Abstract]

28. Cardenal F, Lopez-Carbrerizo MP, Anton A, Alberola V et al. Randomised phase III study of gemcitabine-cisplatin versus etoposide-cisplatin in the treatment of locally advanced or metastatic non-small cell lung cancer. J Clin Oncol 1999; 17: 12–18.[Abstract/Free Full Text]