mesna in patients with malignant mesothelioma

Annals of Oncology 5: 653-655, 1994. © 1994 Kluwer Academic Publishers. Printed in the Netherlands.

Short report A phase II trial of dose-escalated doxorubicin and ifosfamide/mesna in patients with malignant mesothelioma L. Y. Dirix,1 J. van Meerbeeck,2 D. Schrijvers,1 B. Corthouts,3 A. Prove,1 E. van Marck,4 P. Vermeire2 & A. T. van Oosterom1 ^Department of Medial Oncology; 2Department of Respiratory Medicine; ^Department of Radiology; 'Department of Pathology, University Hospital Antwerp, Belgium

partial response was observed, resulting in a response rate of 32% (95% confidence interval 13%-51%). Median response Background: This study investigated the feasibility and effi- duration was 6 months (range 1-13) and median survival cacy of doxorubicin dose-escalated chemotherapy in combi- was 7 months (range 1-18). nation with ifosfamide in patients with malignant mesotheConclusions: The high-dose regimen with growth factor lioma. support is feasible in this group of patients and leads to an Patients and methods: In this single institution phase II interesting response rate. The limiting toxicity for further study, 24 chemotherapy-naive, eligible patients were entered. dose increments and more courses of treatment, was cumulaThe chemotherapy regimen consisted of doxorubicin 75 mg/ tive thrombocytopenia. There seems to be a subgroup of m2 in combination with ifosfamide 5 gr/m2 given as a con- patients with malignant mesothelioma which is less susceptinuous 24 hour infusion, every 21 days with either rhG-CSF tible to develop thrombocytopenia. However, the overall (5 ng/kg) or rhGM-CSF (250 |ig/m2) as haematopoietic sup- toxicity and the poor response duration limit the use of this schedule in the treatment of malignant mesothelioma. port from d3 to dl4. Cycles were repeated every 3 weeks. Results: We treated 24 patients, of whom 22 are evaluable for tumour response. One of the two inevaluable patients died from a cerebral haemorrhage during a period of grade Key words: growth factor support, high-dose chemotherapy, HI thrombocytopenia after the second course. In 7 patients a malignant mesothelioma Summary

Introduction

Patients and methods

Malignant mesothelioma (MM) is a highly lethal tumour of the pleural and/or the peritoneal cavity. Radical surgery can cure only a minority of early stage disease patients. Neither radiotherapy nor chemotherapy contribute to an increase in survival compared to supportive care. The treatment of MM with cytotoxic agents is unsuccessful, with few agents, either used as single agent or in combination regimens, achieving response rates in the range of 15 to 20% [1]. Doxorubicin is currently considered to be the most active single agent with an overall response rate of 15% [2]. Ifosfamide has some activity as a single agent with reported response rates varying between 8% and 24% [3, 4]. An alternative approach is to investigate the role of increasing the dose-intensity of the currently available agents in order to overcome intrinsic drug-resistance. We performed a single-institution phase II study in patients with MM, with doxorubicin at escalated dose combined with ifosfamide and haematopoietic growth factor support.

The patients had not been pretreated and had histologically proven MM with progressive and measurable disease. Their WHO performance status was <2 and bone marrow reserve had to be normal (platelets >125 x lO'/L, white cell count >4.0 x 109/L). Liver and kidney functions had to be within 1.5 times the upper limit of normal. All patients had to be free of prior cardiovascular disease with normal left ventricular ejection fraction (LVEF). Patients with prior or other malignancies were ineligible, as were those with CNS disease. All patients gave a detailed medical history, underwent a physical examination and a complete staging procedure including blood counts, coagulation parameters, routine biochemical tests including electrolytes, renal and liver tests, urine collection with creatinine clearance measurement, a chest radiography, a computed tomographic (CT) scanning of both the chest and the abdomen, a bone scan, and an ECG and LVEF measurement.

Treatment protocol All patients were entered and treated between March 1990 and December 1992. Chemotherapy consisted of the combination of doxorubicin 75 mg/m2 as an i.v. bolus and ifosfamide 5 gr/m2 given as 24 hour infusion. As pretreatment hydration, 2 litres of saline were given over 12 hours. Prior to the ifosfamide, an i.v. bolus injec-

654 tion with mesna was given at a dose of 800 gr/m2 and 2.4 gr/m2 of mesna was given with the ifosfamide. All patients received posttreatment hydration and mesna infusion for a period of 12 hours. Standard antiemetic therapy consisted of ondansetron 24 mg/24 h combined with methylprednisolone 125 mg. The first 12 patients received E. coli derived recombinant human granulocyte-macrophage colony stimulating factor (rhGM-CSF) (Behringwerke, Marburg, Germany) at a dose of 250 ng/m2, the latter 12 received recombinant granulocyte-colony stimulating factor (rhG-CSF) (Amgen, Thousand Oaks) at a dose of 5 (ig/kg. The CSFs were administered as a daily s.c. injection starting 24 hours after the end of the ifosfamide. This was done for 14 consecutive days or until the white cell count (WCC) had risen above 10 x 109/l. Chemotherapy cycles were repeated three weekly if toxicity permitted. Patients were seen for toxicity assessment and bloodcount analysis twice weekly. Routine biochemistry tests including renal function, electrolytes and liver tests were analysed weekly. Patients were re-evaluated for response every 6 weeks. Toxicity analysis included lung function measurement, ECG and LVEF measurement after every two cycles. WHO criteria were used for response and toxicity analysis. No dose reductions were allowed; if by day 22 the WCC < 3.0 x 109/L and/or PTL < 100 x 109/L was not reached, treatment was postponed for 1 week. All patients stopped treatment if progression was documented. If patients with no change or partial and/or complete response, treatment continued or up to six cycles. The duration of survival was calculated from the day of study entry, duration of response from the moment response was first documented.

Results We entered 24 patients (M: 18, F: 6) of whom 22 are evaluable for response and all 24 are evaluable to toxicity. Details of patients characteristics are given in Table 1. These 24 patients received 84 evaluable cycles of combination chemotherapy. One patient received 1 cycle, 9 patients received 2 cycles, 2 patients received 3 cycles, 6 patients had 4 cycles, 1 patient received 5 cycles and 5 patients completed 6 cycles. One patient was not evaluable for tumour response, because he develTable 1. Patient characteristics. Total eligible patient number Total evaluable for toxicity Total evaluable for response Sex male/female Median age (range) Performance status 0 1 2 Known asbestos exposure Primary site of disease Pleural Peritonea] Both Histological subtype Epithelial Sarcomatous Mixed Unknown Initial platelet count >200 x 109/L >300 x 109/L >400 x 1071

24 24 22 17 59 (38-67) 1 13 10 20

oped symptomatic brain metastases after one cycle. One patient developed a cerebral hemorrhage while having a grade HI thrombocytopenia after the second cycle. Details on toxicity are given in Table 2. No skin toxicity was observed, except for local erythema due to the CSFs. Most courses were accompanied by nausea. In 6 patients grade En vomiting occurred in 14 cycles. Oral mucositis was observed in 16 patients and all grade HI toxicity was encountered in patients receiving more than 2 cycles. One patient was readmitted because of grade HI mucositis. No acute cardiac or renal toxicity was observed. Neurological toxicity developed in 4 patients with a grade I or II encephalopathy in 13 cycles. The major toxicity was myelosuppression. Grade HI/TV leukopenia and/or neutropenia occurred in 64 out of 84 cycles. Grade IV leukopenia and/or neutropenia occurred in 40 out of 84 cycles. All patients receiving 4 or more courses developed grade IV neutropenia. The duration of grade IV neutropenia increased with the number of courses, with a median duration of 2 days after cycle 1, to 4 days after cycle 4, up to 5 days after 6 cycles. Median onset of grade IV neutropenia and/or leukopenia was by day 8. Eighteen patients were readmitted with febrile neutropenia, with a documented infection in 4. There is a linear correlation between initial platelet count and platelet nadir both after the first cycle (r = 0.811, p = 0.0001) and the second cycle (r 0.821, p = 0.0001). Grade HI/TV thrombocytopenia occurred in 24 out of 84 cycles again with evidence of cumulative toxicity. In 5 of 6 patients receiving 6 courses, grade IV thrombocytopenia was observed from the fifth cycle onwards. The median duration of grade HI thrombocytopenia increased from 3 days after 2 cycles, 5 days after 4 cycles to 6 days (range 4 to 17 days) after 6 cycles. Bleeding episodes were observed during 7 courses in 5 patients; 1 patient had a period of melena after an episode of grade HI thrombocytopenia and 1 patient suffered a cerebral bleed. In total, 108 units of platelets were transfused to 6 patients in order to keep the platelet count >20 x 109/L. The CSFs were administered with a median duration of administration of 12 days (range 7-14) after each cycle. The use GM-CSF at this dose level was accompanied by more side-effects; fever, local irritation and bone pain. Table 2. Number of patients with specific toxicities*. Toxicity

17 3 4 15 2 6 1 22 17 15

Alopecia Nausea/vomiting Mucositis Diarrhoea Hemoglobin Platelets Leukocytes Cardiac Infections Encephalopathy

0

1

2

3

4

0 1 8 20 3 6 1 24 20 20

1 0 8 2 7 4 1 0 1 2

1 17 4 2 10 4 1 0 3 2

22 6 4 0 4 2 7 0 0 0

0 0 0 0 0 8 14 0 0 0

Only the highest WHO grade is reported for each patient.

655 The intentional cycle length of 21 days was held in use of superior haematological support and/or protec56 of the 84 cycles. In only 17 cycles was treatment tion. postponed for 1 week due to toxicity, a haematologic or Although this trial was not randomised and historiother. The median cycle length was 21 days (range 2 1 - cal comparisons are not a reliable means to anwer the 33) with a mean of 22.9 days. questions of superiority of any particular regimen, it Seven patients had an objective response resulting in seems possible to obtain a superior response rate in a response rate of 32% (CI 13%-51%). All were partial this resistant tumour by this increased dose intensity. responses, although in 2 patients only minor abnor- There are however no arguments in our results to sugmalities remained. Responses were seen in both pleural gest that this was clinically significant. (5) and peritoneal deposits (3) and in lymph-nodes (2). Median response duration was 6 months (range 1-13). The median overall survival of the 22 evaluable pa- Acknowledgement tients was 7 months (range 2-18), in non-responders median survival was 6 months (2-10), in responding The authors are grateful to dr. J. Soedirman of Behringpatients 7 months (range 7-18). werke, Belgium and Mrs. M. Volckaert of AmgenRoche, Belgium for providing the hematopoietic growth factors. We are indebted to drs. Bockaert, de Beukelaer, Discussion Delanote, De Schepper, Goetstouwers, Goossens, This trial investigated the feasibility and efficacy of Moorkens, Ortmanns, Vandemaele, Van Mulders, Van doxorubicin dose-intensive therapy in MM. This regi- Roelen and Verhaert for their cooperation in this trial, men was selected based on knowledge of their single and to the Medical Oncology Nursing Staff for their agent activity in MM and their combined activity in soft outstanding patient care. tissue sarcomas [5]. In MM, Carmichael et al. used this combination with doxorubicin at 40 mg/m2 and ifosfamide at 5 gr/m2 also as an 24 hour infusion. This led References to a response rate of 12.5% in 16 evaluable patients [6]. We have doubled the dose intensity of the most active 1. Krarup-Hansen A, Hansen HH. Chemotherapy in malignant mesotelioma: A review. Cancer Chemother Pharmacol 1991; 28: component. An identical regimen has been investigated 319-30. in soft issue sarcomas with encouraging results [7]. We 2. Lerner HJ, Schoenfeld DA, Martin A et al. Malignant mesotheobserved a response rate of 32%, which compares lioma. The Eastern Cooperative Oncology Group (ECOG) experience. Cancer 1983; 52:1981-5. favourably with the normal dose regimen [6]. These responses lead to subjective improvement but were 3. Zidar BL, Metch B, Balcerzak SP et al. A phase II evaluation of ifosfamide and mesna in unresectable diffuse malignant mesoshortlasting with a median duration of response of 6 thelioma. Cancer 1992; 70: 2547-51. months. 4. Alberts AS, Falkson G, Van Zyl L. Malignant pleural mesotheThe toxicity of this regimen is important. The sublioma: A phase II pilot study of ifosfamide and mesna. J Natl Cancer inst 1988; 80:698-700. jective toxicity is also highlighted by the fact that 4 patients preferred to stop therapy after either 2 or 3 5. Dirix LY, Van Oosterom AT. The role of ifosfamide in the treatment of adult soft tissue sarcoma, Ewing's sarcoma, osteosarcourses in spite of having stable disease in cases with coma: A review. Sem Oncol 1990; 17: 50-7. known progression prior to entry. Mucositis and 6. Carmichael J, Cantwell BMJ, Harris AL. A phase II study of haematological suppression, which both showed a ifosfamide/mesna with doxorubicin for malignant mesothelioma. Eur J Cancer Clin Oncol 1989; 25:911 -2. tendency of increasing severity during successive cycles, are severe. The neutropaenia and/or leuko- 7. Steward WP, Verweij J, Somers R et al. Granulocyte-macrophage colony stimulating factor allows for safe escalation of dose-intenpaenia were surmountable; however with the observasity of chemotherapy in metastatic adult soft tissue sarcomas; a tion of increasing duration of neutropenia after increasstudy of the EORTC soft tissue and bone sarcoma group. J Clin ing number of courses, this toxicity had become a Oncol 1993; 11:15-21. problem by the time patients had received six cycles. 8. Schmitter D, Lauber B, Fagg B et al. Hematopoietic growth factors secreted by seven human pleural mesothelioma cell lines: The major problem, however, is cumulative toxicity on Interleukin-6 production as a common feature. Int J Cancer megakaryopoiesis. The one toxic death was due to this 1992; 51: 296-301. toxicity. The observation of some degree of protection offered by initial thrombocytosis is interesting. This Received 20 January 1994; accepted 6 April 1994. could be due to the production of cytokines with proliferative activity on platelet production such as interleukin-6 [8]. An important conclusion of this trial is Correspondence to: that it is possible to deliver this dose-intensive regimen Dr. L. Y. Dirix Department of Medical Oncology, University Hospital Antwerp to patients with MM for up to six cycles but it seems Wilrijkstraat 10 unlikely that more than six cycles or further increase in 2650 Edegem dose intensity is possible with these drugs, without the Belgium