Paclitaxel and gemcitabine in advanced non-nasopharyngeal head and neck cancer: A phase II study conducted by the Hellenic Cooperative Oncology Group

Annals of Oncology 10: 475-478, 1999. © 1999 Kluwer Academic Publishers Printed in the Netherlands.

Short report Paclitaxel and gemcitabine in advanced non-nasopharyngeal head and neck cancer: A phase II study conducted by the Hellenic Cooperative Oncology Group G. Fountzilas,1 G. Stathopoulos,2 C. Nicolaides,3 A. Kalogera-Fountzila,1 H. Kalofonos,4 A. Nikolaou,1 C. Bacoyiannis,5 E. Samantas,6 C. Papadimitriou,7 P. Kosmidis,5 J. Daniilidis1 & N. Pavlidis3 1

'AHEPA'Hospital, Aristotle University oj Thessaloniki, Thessaloniki, 2'Ippokration'Hospital, Athens; 3University of loannina, Ioannina, RIO'Hospital, University of Patras, Patras; 5'Metaxa'Cancer Hospital, Piraeus; 6Agii Anargyri' Cancer Hospital, Athens; 7 Alexandra'Hospital, Athens, Greece

4

Results: Twenty-four (55%) patients completed all six cycles of treatment. A total of 205 cycles were administered, 165 Background: Paclitaxel as monotherapy or in combination (81%) of them at full dose. The median relative dose intensity with other drugs has demonstrated significant activity in pa- (DI) of gemcitabine was 0.93 and of paclitaxel 0.95. Except tients with squamous cell carcinoma of the head and neck for alopecia, which was universal, grade 3-4 toxicities inregion (SCCHN). Preclinical studies have shown gemcitabine cluded neutropenia (21%), thrombocytopenia (5%), anemia (5%), infection (5%), flu-like syndrome (5%) and peripheral to be highly active in SCCHN cell lines. Purpose of the study; To evaluate the activity and toxicity of neuropathy (2%). Five (11%) patients achieved complete and the combination of paclitaxel by three-hour infusion and 13 (30%) partial responses, for an overall response rate of gemcitabine as first-line chemotherapy in patients with recur- 41%. After a median follow-up of 13 months, the median time to progression was four months and median survival nine rent and/or metastatic head and neck cancer (HNC). Patients and methods: From September 1996 until May months. 1998, 44 patients with non-nasopharyngeal recurrent and/or Conclusions: The combination of paclitaxel and gemcitabine metastatic HNC entered the study. There were 37 men and is active and well tolerated in patients with recurrent and/or seven women with a median age of 61 years (range 35-79) and metastatic HNC - randomized studies comparing this combia median performance status of 1 (range 0-2). The location of nation with other regimens are warranted. the primary tumor in the majority of them was either the larynx or the oral cavity. Treatment consisted of six cycles of gemcitabine 1100 mg/m2 over 30 min on days 1 and 8 immediately followed on day 1 by paclitaxel 200 mg/m2 by three-hour Key words: chemotherapy, gemcitabine, head and neck tumors, infusion. The treatment was repeated every three weeks. paclitaxel Summary

Introduction The prognosis of patients with recurrent or metastatic head and neck cancer (HNC) is poor. Cisplatin-containing regimens are the most widely used and produce an overall response rate (ORR) in the range of 30%-35% [1, 2]. Combination chemotherapy produces higher response rates than single agents but median survival does not exceed six months in either case [1]. Taxanes in the treatment of HNC have been recently introduced. Paclitaxel and docetaxel have shown significant activity in patients with SCCHN [3,4]. In a phase II study [3] conducted by the Eastern Cooperative Oncology Group (ECOG), paclitaxel at a dose of 250 mg/m2 over 24 hours produced an ORR of 40%. Gemcitabine is a new deoxycytidine analog with unique activity against a wide range of solid tumors.

Experience with the use of gemcitabine in advanced HNC is limited. In a phase II trial [5] the drug was given initially at a dose of 800 mg/m2 on days 1, 8 and 15 every three weeks in patients who had previously undergone chemotherapy. In that study the dose was increased to 1250 mg/m2 in chemotherapy-naive patients. The ORR was 13%. Results concerning the cytocidal effect of the combination of gemcitabine and paclitaxel in vitro have been contradictory: less than additive in selected human tumor cell lines [6] and synergistic in others [7]. Motivated by in vitro data and the need for the identification of effective regimens in patients with HNC, the Hellenic Cooperative Oncology Group (HeCOG) conducted a phase II study to evaluate the activity and toxicity profile of the combination of paclitaxel and gemcitabine in patients with recurrent or metastatic HNC.

476 Table 1. Patients and tumor characteristics.

Number of patients Age (in years) Median Range

44 61

35-79

Sex

Male Female Performance status 0 1 2

Primary site Oropharynx Hypopharynx Larynx Oral Cavity Paranasal sinuses Salivary glands SCC histologic type Tumor grade I II III IV

Unknown Primary treatment Induction chemotherapy RT

Surgery Site of disease Locoregional Primary Nodes Skin Primary and nodes Distant Lung Mediastinal nodes Locoregional and distant

37 7 12 25 7 4 2 18 15 3 2 44 9 13 14 1 7

The chemotherapy regimen consisted of gemcitabine at a dose 1100 mg/m2 in 500 ml N/S over 30 min on days 1 and 8 immediately followed on day 1 by paclitaxel at a dose of 200 mg/m2 by three-hour infusion, with standard premedication. Treatment was repeated every three weeks. Methyprednisolone 16 mg orally twice daily was prescribed for two days after gemcitabine infusion. Ondansentron was administered as antiemetic treatment in all cases. Clinical examination, CBC and biochemical analysis were performed on the first day of each cycle. CBC were repeated on day 8. Examination with imaging techniques was repeated every two cycles. All imaging material pertinent to tumor response to chemotherapy was scanned (by a scanner model Mirage II, maximum resolution 9800 x 9800 dpi, UMAX Data Systems Inc., Hsinchu, Taiwan) and evaluated after completion of the study by one of the authors (A. Kalogera-Fountzila) and one independent radiologist. The ANC had to be 5= 1500/ul and the platelet count > 100,000/ul prior to the beginning of the next treatment cycle. Dose modifications of paclitaxel have been described in detail previously [8]. G-CSF (Filgrastim, 5 ug/kg/d) was administered in instances of ANC <500/ul or of febrile neutropenia. If ANC on day 8 was between 1.01.5 x 109/l then gemcitabine could be administered also with G-CSF to maintain DI. In such cases G-CSF was given prophylactically in all subsequent cycles If ANC was less than 1.0 x 1O9/1 then treatment was postponed for one week. In instances of grades 2 or 3 toxicity the dose of gemcitabine was reduced by 25% or 50%, respectively, in all subsequent cycles. In case of any grade 4 toxicity the patient was taken off the study. Toxicity criteria were those adopted from the World Health Organization. Standard ECOG criteria were used for evaluation of tumor response. Duration of response, time to progression (TTP) and survival were calculated as previously described [8].

10 35 15 40 26 29 6 16 8 8 1 7

Results Compliance with treatment, and toxicity

Twenty-four (55%) patients completed all six cycles of treatment. Reasons for treatment discontinuation were progression of the disease in 12 patients, early death in three (due to the disease, myocardial infarction and nonspecified in one patient, respectively), voluntary withdrawal in two, physician decision in two non-responding patients and grade 4 neutropenia in one patient. Patients and methods A total of 205 cycles were administered, 165 (81%) of Patients with histologically proven recurrent or metastatic non-naso- them at full dose (Table 2). Except for alopecia, the most pharyngeal SCC of the head and neck region, aged 3s 18 years with a common grades 3-4 toxicities were neutropenia (21%), performance status (PS) <2 of the ECOG scale and a life expectancy thrombocytopenia (5%), anemia (5%), infection (5%), of > 3 months were eligible for the study. Measurable or evaluable flu-like syndrome (5%) and peripheral neuropathy (2%). disease outside of preirradiated areas was required for inclusion, G-CSF was administered in 11 patients at some time on unless a subsequent progression of the lesion had been documented. Previous chemotherapy prior to or concomitantly with radiation was trial because of neutropenia, red blood cell transfusions allowed, if administered ;& I year before entry in the study. All patients in eight, platelet transfusion in one and antibiotics in should have had leukocyte and platelet counts greater than 4000/ul seven patients for febrile neutropenia (two patients), and 100,000/ul, respectively, serum creatinine concentration less than infection (two) and local inflammation (three). 0.3 mg/dl above the institutional upper limit of normal and normal total and direct bilirubin levels. A detailed informed consent was obtained from all patients before their entry into the study. The protocol was approved by the Protocol Review Committee (PRC) of HeCOG and by the Ethics Committees of all participating institutions. Pretreatment evaluation included a complete medical history, clinical examination, EKG, complete blood count (CBC), complete biochemistry survey, chest X-ray, bone scan and computed tomography (CT) scan, as indicated. From September 1996 until May 1998, 44 patients entered the study. Selected patient and tumor characteristics are listed in Table 1.

Best response and survival In four patients response could not be assessed because three of them died prior to first evaluation of tumor response and one decided to stop treatment after the second cycle. Five patients (11%, 95% CI: 4%-25%), all of them with locoregional relapse (primary in three patients, nodes in four and skin in one patients), achieved CR

477 Table 2. Selected treatment characteristics.

Number of cycles per patient

Total number of cycles Number of cycles at full dose Number of cycles with a delay Delay on day 1 Delay on day 8 Interruption on day 8 DI of paclitaxel Planned Delivered Median Range Relative DI of paclitaxel Median Range DI of gemcitabine Planned Delivered Median Range Relative DI of gemcitabine Median Range

Number of cycles

Number of patients

1 2 3 4 5 >6

3 5 5 5 2 24 205 165 61 35 32 7 66.6 63 39-73 0.95 0.6-1.1 733 683 314-788 0.93 0.4-1.1

Dl-dose intensity (mg/m 2 /week).

In the present study we used slightly higher doses than in the previous study in order to maximize activity without compromising the DI of the two drugs. The response rate of 41% reported here is interesting. In our previous study in such patients, using the combination of paclitaxel and carboplatin, we reported an ORR of only 23% [10], which could be partially attributable to patient selection. Except for alopecia the grade 3-4 side effect most frequently seen was neutropenia, observed in one-fifth of our patients. Phase I studies [reviewed in Ref. 11], with the combination of paclitaxel and gemcitabine have repeatedly shown the principal dose-limiting toxicity to be grade 4 neutropenia. Even though episodes of sepsis are infrequently seen with this combination, probably because of the transient nature of the neutropenia, the prophylactic use of G-CSF may be necessary in order to maintain DI. Androulakis et al. [12], administering, as second-line treatment in patients with NSCLC, the combination of gemcitabine 900 mg/m2 on days 1 and 8 and paclitaxel 175 mg/m2 over three hours on day 8, with the prophylactic use of G-CSF, every three weeks, reported that grades 3-4 neutropenia occurred in only 2 of 26 (8%) patients. The incidence of grades 2-3 peripheral neuropathy reported in this study - higher than in ours (32% vs. 16%) - is probably due to the fact that in the latter study twice as many patients had previously received cisplatin-based chemotherapy. In conclusion, the combination of paclitaxel and gemcitabine is active and well tolerated in patients with recurrent and/or metastatic HNC and deserves further exploration in randomized studies.

and 13 (30%, 95% CI: 17%-45%) PR, for an overall response rate (ORR) of 41% (95% CI: 26%-57%). The five CRs lasted for 20.5, 24, 38, 43+ and 55 weeks. Furthermore, 14 patients (32%, 95% CI: 19%-48%) References demonstrated stabilization of the disease. Median duration of response was eight months. As of 15 October 1. Vokes EE, Athanasiadis I. Chemotherapy for squamous cell 1998, after a median follow-up of 13 months (range 0.7carcinoma of head and neck. The future is now. Ann Oncol 1996; 7: 15-29. 21+) 36 patients (82%) had demonstrated disease pro2. Jacobs C, Lyman G, Velez-Garcia E, Sridhar K.S et al. A phase 111 gression and 23 (52%) had died. Median TTP was four randomized study comparing cisplatin and fluorouracil as single months (range 0.8-19+) and median survival nine months agents and in combination for advanced squamous cell carci(range 0.7-21+). Ten patients were treated with addinoma or the head and neck. J Clin Oncol 1992; 10: 257-63. tional chemotherapy after they had progressed. Among 3. Forastiere AA, Shank D, Neuberg D et al. Final report of a phase II evaluation of paclitaxel in patients with advanced squamous 10 patients who received induction chemotherapy there cell carcinoma of the head and neck: An Eastern Cooperative were two CRs and one PR. Six of eight partial responders Oncology Group Trial (PA 390). Cancer 1998; 82: 2270-4. demonstrated PR of the primary tumor within a pre4. Catimel G, Verweij J, Mattijssen Vet al. Docetaxel (Taxotere): An irradiated area. Responses were seen in all metastatic active drug of the treatment of patients with advanced squamous sites except bone. cell carcinoma of the head and neck. Ann Oncol 1994; 5: 533-7. Discussion The optimal dose of paclitaxel in combination with gemcitabine has not yet been defined. In a phase I study [9] in patients with ovarian cancer, in which gemcitabine and paclitaxel was delivered every three weeks in a manner similar to that used in our study, the maximum tolerable dose of gemcitabine was 1 g/m2 on days 1 and 8 and of paclitaxel 175 mg/m2 on day 8.

5. Catimel G, Vermorken JB, Clavel M et al. A phase II study of gemcitabine (LY 188011) in patients with advanced squamous cell carcinoma of the head and neck. Ann Oncol 1994; 5: 543-7. 6. Theodossiou C, Cook JA, Fisher J et al. Interaction of gemcitabine with paclitaxel and cisplatin in human tumor cell lines. Int J Oncol 1998; 12: 825-32. 7. Jensen PB, Holm B, Sorensen M. In vitro cross resistance and collateral sensitivity in seven resistant small-cell lung cancer cell lines. Preclinical identification of suitable drug partners to Taxotere, topotecan and gemcitabine Br J Cancer 1997; 75: 869-77. 8. Fountzilas G, Papadimitriou V, Bafaloukos D et al. Paclitaxel and carboplatin as first-line chemotherapy for advanced breast cancer. Oncology 1998; 12 (Suppl 1): 45-8.

478 9. Poole CJ, Perren T, Hogberg Tet al. Phase I study to investigate alternate sequencing of the combination of gemcitabine and paclitaxel in ovarian carcinoma. Eur J Cancer 1997: 33 (Suppl 8): SI 21 (Abstr). 10. Fountzilas G, Skarlos D, Athanassiadis A et al. Paclitaxel by three-hour infusion and carboplatin in advanced carcinoma of nasopharynx and other sites of the head and neck. A phase II study. Ann Oncol 1997; 8: 451-5. 11. Dombernowsky P, Giaccone G, Sandier A et al. Gemcitabine and paclitaxel combinations in non-small-cell lung cancer. Semin Oncol 1998; 25: 44-50. 12. Androulakis N, Kouroussis C, Kakolyris S et al. Salvage treatment with paclitaxel and gemcitabine for patients with non-

small-cell lung cancer after cisplatin-or docetaxel-based chemotherapy A multicenter phase II study. 1998; 9: 1127-30. Received 23 November 1998; accepted 2 February 1999. Correspondence to: George Fountzilas, MD 1st Department of Internal Medicine Oncology Section AHEPA Hospital Aristotle University of Thessaloniki Thessaloniki, Macedonia Greece