A comparative study of cisplatin and vinblastine versus ifosfamide, cisplatin and vinblastine in non-operable non-small-cell lung cancer*

Annals of Oncology 5: 159-162, 1994. O 1994 Kluwer Academic Publishers. Printed in the Netherlands.

Original article A comparative study of cisplatin and vinblastine versus ifosfamide, cisplatin and vinblastine in non-operable non-small-cell lung cancer* A study of the Hellenic Co-operative Oncology Group for Lung Cancer Trials P. Kosmidis, N. Mylonakis, D. V. Skarlos, E. Samantas, M. Beer, D. Theocharis, E. Zachariadis, A. Gioti, E. Papadakis, N. Poulakis, S. Parastatidis & G. Fountzilas From the Hellenic Co-operative Oncology Group for Lung Cancer Trials, Medical Oncology Department; 'Metaxa' Cancer Hospital, Piraeus, Greece (see page 162 for list of participating institutions)

with stage HI disease, good PS and no visceral involvement had better survivals. Nausea/vomiting and alopecia were Purpose: To evaluate the efficacy and toxicity of ifosfamide in more pronounced in the VIP group, although both chemocombination with cisplatin and vinblastine in non-operable therapies were well tolerated. non-small-cell lung cancer (NSCLC). Conclusions: The addition of ifosfamide improved the Methods: A total of 136 patients with stage El or IV response rate, but a survival advantage cannot be proven. NSCLC were randomized to either PV (cisplatin, 120 mg/ The prognostic value of stage, PS and metastatic site is conm2, and vinblastine, 6 mg/m2) or VIP (PV with the addition firmed in this trial; further studies are required to select subgroups of patients who may have a survival benefit with of ifosfamide, 3 g/m2) every 3 weeks. Results: Patients receiving VIP had a higher response rate combination chemotherapy. The response rate elicited by (31% vs. 10%), but the performance status (PS) was signifi- VIP makes it a candidate for neoadjuvant treatment. cantly worse in those receiving PV. No difference in survival can be demonstrated between the two treatment groups. The Key words: non-small-cell lung cancer, ifosfamide, cisplatin, median survival was 8.4 months. In both groups, patients chemotherapy Summary

Introduction

The main treatment modality for localized non-smallcell lung cancer (NSCLC) is surgical resection [1, 2]. In locally confined but inoperable cancer, radiotherapy has proven efficacy [3]. Chemotherapy has a limited impact but recent studies have outlined its role more clearly. Thus, more specific information is available about individual factors, survival benefits and recently, neoadjuvant administration schedules [4-7]. Cisplatin, vindesine, vinblastine and mitomycin are among the active single agents in this disease [8-13]. The combination of cisplatin and vindesine was reported to yield improved response rates over those of single-agent treatment [11]. Combination chemotherapy with cisplatin and vindesine or vinblastine elicited response rates of over 25% [6,10,12]. Studies with cisplatin, vinblastine and mitomycin have reported response rates over 30%, and a few have indicated higher rates (50%-60%) with doses and schedules similar to those used in phase II studies [5]. Ifosfamide as a single agent has been reported to elicit responses in 15%-30% [14]. Combinations containing ifosfamide together with one * Presented in part at the Annual Meeting of the American Society of Clinical Oncology, San Diego, CA, 1992.

or two other cytostatic agents have yielded overall response rates of around 30%-45%, with 5%-10% complete remissions [14,15]. In order to evaluate the role of ifosfamide in NSCLC, and in particular to investigate if a possible major improvement in response and survival rate in patients with advanced NSCLC can justify an expected increase in toxicity, we undertook to randomize patients to receive either a combination of cisplatin and vinblastine or all three drugs.

Methods From October 1989 to June 1991,136 patients with advanced nonoperable (stages Ula, DJb, TV) non-small-cell lung cancer, confirmed by biopsy or cytology, were entered into the study. Patients were required to have measurable or evaluable disease, a performance status of less than 3 on the W.H.O. scale [17], age less than 75 years, normal renal, hepatic and cardiac function and an absence of CNS metastases. All had given informed consent in accordance with the Helsinki declaration. No previous chemotherapy was allowed and evaluable lesions had not been irradiated. Patients were randomized to receive either the PV regimen, consisting of cisplatin, 120 mg/m2 on day 1 given by intravenous infusion over 1 hour with pre- and post-hydration as well as mannitol diuresis, and vinblastine, 6 mg/ m2 i.v. on day 1 of each course, or the VIP combination, which consisted of the same combination with the addition of ifosfamide, 3

160 g/m2 on day 1 of each cycle. Mesna, 2.4 g/m2, was infused i.v. over 12 hours following the ifosfamide administration, for urothelial protection. Treatment courses were repeated every 3 weeks. A treatment delay of at most 3 weeks was allowed for the recovery of white blood cell and platelet counts. No dosage modifications were permitted. Chemotherapy was continued until either disease progression or the completion of 6 courses of treatment Patients were staged and response was assessed by clinical examination, chest x-rays and computerized tomography and, as indicated, abdominal computerized tomography, liver or adrenal ultrasound and bone scan. Disease parameters were measured at least every 8 weeks; chest x-rays were repeated monthly. The 104 available histopathology slides were reviewed by the same group of pathologists. The usual criteria were followed for definition of response [17]. Tune to progression (TTP) was defined as the time elapsed from the start of treatment to renewed progression, and survival from initiation of chemotherapy until death. Toxicity was evaluated according to the W.H.O. grading system. The two trial populations were compared and cross-correlations of all available variables checked by means of a chi-square test, with Yates' correction as appropriate, or by Fisher's exact test where necessary for discrete variables, and, for continuous variables, with Student's t-test or the Mann-Whitney nonparametric test, after logarithmic transformation if needed. Risk ratio in subgroup analyses of response was estimated with a Mantel-Haenszel statistic. The BMDP statistical package was used for these calculations [18] along with an ad hoc program. Some of the known factors were analyzed for correlation with survival for all patients and stratified by treatment group, after cut-off points were set for the continuous variables by comparing actuarial survival curves using a log-rank test [19]. Patients still alive or recurrence-free, and those lost to follow-up were censored. In all comparisons, a test power (1-ZP) of 80% or more was requested for reporting negative results [20]; reported p-values are two-tailed.

Results

follows: pleura 30%, lymph nodes 26%, bone 48%, lung 4%, adrenals 10%. Metastatic sites were multiple in 69% of stage IV patient, visceral in 48% and, in 13% only, they were confined to pleura and/or lymph nodes. A total of 229 treatment cycles were administered in the PV group, and 325 in VIP. The median duration of treatment was 4 cycles in both group (range, 1-8). The response to treatment and the time to disease progression are indicated in Table 3. Tumor relapse has already been documented in all but one case. In 7 of the 8 patients who received one cycle only of chemotherapy, progression was documented before the second cycle was due. The proportion of responders to non-responders was significantly higher in the VIP group (p-value 0.02), whereas time to progression for all patients with a response or stable disease as shown in Fig. 1, was better in the PV group (p = 0.06). No difference between the two groups could be shown with regard to the response in stage HI versus IV, in stage IV patients with single versus multiple localizations, in patients with visceral lesions versus bone, lymph node and/or pleura involvement only, or according to rumor differentiation, age group, or sex. The performance status (0 and 1 versus 2 and 3) was correlated with response (p =• 0.05); the risk ratio comparison between the two chemotherapy groups was also significant. Only 2 patients in the PV group and 4 in VIP were still alive on the day of analysis. One patient in the first group and three in the second are lost to follow-up. Median survival for the PV group is 8 months (range, Table 2. Tumor characteristics.

The characteristics of the 136 patients entered into the trial are shown in Table 1. All but one were evaluable for tumor response and toxicity. The distribution of the different patient and tumour characteristics was comparable in the two treatment groups, except for performance status, which was significantly worse in the PV group. Patients with stage IV disease also had a comparable distribution with regard to the nature and number of metastatic disease localizations, distributed as

Stage IIIA

mB IV Histologic type Squamous-cell Adenocarcinoma Large-cell Undifferentiated

PV

VIP

7 31 23

4 40 31

21 31 5 4

40 24 7 4

Table 1. Patient characteristics. Table 3. Response andtime to progression. PV

VIP

61 33-75 61

75 39-75 63

56 5

68 7

3 33 23 2

7 56 10 2

PV Number Age Median Sex Male Female Performance status* 0 1 2 3 * Significant difference.

Complete (CR) 1 Partial (PR) 5 Stable disease (SD) 28 Progressive disease 27 Non-evaluable Time to progression1" Months Median

Proportion" 1 (0-5)% 8 (1-15)% 46 (33-58)%

1-20 8.5

Percentage, (95% confidence limits). For patients with CR, PR and SD.

VIP

6 17 28 23 1

Proportion 8 (2-14)% 22 (13-32)% 37 (26-48)%

1-17 7

161 1-

250-

PV VP

750-

750-

500-

— o \

i-

—8— PV — i — VIP

V

.500-

\

^—^ -

0-

250-

8

12 16 months

20

15

24

28

1-26), and 8.8 for the VIP group [1-35]. No difference is evident between the curves, shown in Fig. 2 (test power 24% only). The survivor function showed differences when analyzed for all patients and also separately by treatment group, according to tumor stage (m or IV, p = 0.001) and performance status (p = 0.05). A difference in survival was detectable also according to the disease localization as well (visceral metastases versus the involvement of pleura, lymph nodes and/or bones only, p - 0.005). Toxicity is indicated in Table 4. Nausea/vomiting and alopecia were significantly pronounced in the VTP group (p = 0.02 and < 0.0001, respectively). In both groups, a 5HT3 antagonist such as ondansetron was used as antiemetic. No patients were lost due to toxicity. Discussion Combination chemotherapy in advanced non-smallcell lung cancer remains a controversial issue [3]. Most of the studies generally fail to show a survival advantage for treated patients, even though some authors report a prolonged survival [7]. Since 1980 the vast Table 4. Toxicity (percentage of patients).

Bone marrow Nausea/vomiting" Alopecia* Renal Neurologic Significant difference.

PV

20

25

30

35

months

Fig. 2 Survival according to chemotherapy group.

Fig. 1. Time to progression of disease for patients with response or stable disease, by chemotherapy group.

W.H.O. grade

i

0-

\

VIP

i-n

ni-rv

i-n

ra-rv

23 62 49 28 15

3 2 8 0 0

28 87 33 17 27

9 3 45 0 9

majority of the combinations have been platinumbased, with the addition of vinca alkaloids or etoposide [5, 11]. Many efforts have been made to add a third cytostatic agent to improve response rates and possibly survival, but with equivocal results. The need to evaluate promising phase II drugs by their inclusion in existing combinations led us to investigate the role of ifosfamide in patients with non-operable NSCLC. Ifosfamide as single-agent in different doses and schedules has been reported to yield responses in approximately 15%-30% of patients [14]. In combination with a second agent, responses ranging from 24% to 34% are reported [14]. When ifosfamide was combined with mitomycin and vindesine, or with cisplatin and vindesine, mitomycin or etoposide, response rates of 45%, 28%, 51%, 40% respectively were reported, with a reasonable percentage of complete responders [14,15]. In our study, the triple combination elicited a higher response rate than the PV combination. However, it must be emphasized that the response in the PV group is rather low compared to the one expected on the basis of earlier reports [5]. This is probably due to an unfavorable distribution of prognostic factors, especially of performance status, in this group. This latter factor was shown to have a major influence both on response and survival in this study. A difference in survival between the two treatment arms cannot be shown with the numbers available in our study. The numbers in the different subgroups were also insufficient to show a difference in response rate according to different patient and tumor characteristics, except for the patients' performance status; there were also too few patients with stage nia disease. In both chemotherapy groups, the presence and localization of metastatic disease and the patients' performance status were correlated with survival. Even though our study does not allow a statement as to an advantage of chemotherapy in any subgroups, because it was not specifically designed and stratified for this purpose, it shows a survival advantage for

162 patients with a good performance status and without visceral metastases. It is now generally known that these and other prognostic factors favor better survival [5]. Based on this fact, it is reasonable to assume that the administration of chemotherapy is NSCLC should be limited to selected groups of patients. Large and adequately stratified controlled studies are still needed to define those subgroups of patients who might obtain benefit from chemotherapy. It is also particularly interesting to apply active well known regimens as well as VIP in the neoadjuvant setting where high response rates of more than 50% are desirable for a possible prolongation of survival. Welldesigned studies are already in progress to answer this intriguing question.

7.

8. 9.

10.

11.

Acknowledgments

The authors thank Mrs. E. Papakostaki and Ms. I. Benardou for their skilled assistance in administrative and clerical duties.

12. 13.

Appendix

14.

Participating Institutions of the Hellenic Co-operative Oncology Group for Lung Cancer Trials: 'Metaxas' Cancer Center, Piraeus; Agii Anargyri Cancer Hospital, Athens; AHEPA University Hospital, Thessaloniki; Sismanoglion General Hospital, 1st and 4th Departments of Medicine of the Chest Diseases Hospital, Amalia Fleming' General Hospital and Army Fund Hospital, Athens, and the Hellenic Co-operative Oncology Group Office, Greence.

15.

References

19.

1. Johnston MR. Selecting patients with lung cancer for surgical therapy. Semin Oncol 1988; 15: 246-54. 2. Mountain CF. Prognostic implications of the international staging system for lung cancer. Semin Oncol 1988; 15: 236—45. 3. Minna JD, Pass M, Glatstein E et al. Cancer of the lung. In De Vita VT, Hellmann S, Rosenberg SA (eds): Cancer. Principles and practice of oncology. Philadelphia: Lippincott 1989; 591705. 4. Hansen HH. Advanced non-small-cell lung cancer To treat or not to treat. J Clin Oncol 1987; 5:1711-2. 5. Gralla RT, Kors MG. Chemotherapy in non-small-cell lung cancen Results of recent trials. Semin Oncol 1988; 15:2-5. 6. Cellerino R, Tummarello D, Piga A. Chemotherapy or not in

16.

17. 18.

20.

advanced non-small-cell lung cancer? Lung Cancer 1990; 6: 99-109. Evans Wk. Combination chemotherapy confers modest survival advantage in patients with advanced non-small-cell lung cancer Report of a Canadian multicenter randomized trial. Semin Oncol 1988; 15: 5-10. Vogl S, Berenzweig M, Camacho F et al. Efficacy study of intensive cisplatin therapy in advanced non-small-cell bronchogenic carcinoma. Cancer 1982; 50: 24-6. Panettiere FJ, Vance RB, Stuckey WJ et al. Evaluation of single agent cisplatin in the management of non-small-cell carcinoma of the lung: A Southwest Oncology Group Study. Cancer Treat Rep 1988; 66: 305-9. Kns M, Gralla R, Kallman L et al. Randomized trial comparing vindesine plus cisplatin with vinblastine plus cisplatin in patients with non-small-cell lung cancer, with an analysis of methods of response assessment. Cancer Treat Rep 1983; 15: 6-10. Einhorn LH, Loehrer PT, Williams SD et al. Random prospective study of vindesine vs. vindesine plus high dose cisplatin vs. vindesine plus cisplatin plus mitomycin C in advanced nonsmall-cell lung cancer. J Clin Oncol 1986; 4:1037-43. Williams CT, Woods R, Lebi J et al. Chemotherapy for nonsmall-cell lung cancer: A randomized trial of cisplatin and vindesine vs. no chemotherapy. Semin Oncol 1988; 15: 6-8. Mylonakis N, Tsavaris N, Bacoyiannis C et al. A randomized prospective study of cisplatin and vinblastine vs. cisplatin, vinblastine and mitomycin in advanced non-small-cell lung cancer. Ann Oncol 1992; 3:127-30. Eberhardt W, Niederle N. Ifosfamide in non-small-cell lung cancen Areview.Semin Oncol 1992; 19 (Suppl 1): 40-8. Shepherd FA, Evans KW, Gross PE et al. Ifosfamide, cisplatin and etoposide in the treatment of advanced non-small-cell lung cancer. Semin Oncol 1992; 19 (Suppl 1): 54-8. Shirinian M, Lee JS, Dhingra HH et al. Phase II study of cisplatin, ifosfamide with mesna, and etoposide (PIE) chemotherapy for advanced non-small-cell lung cancer. Semin Oncol 1992; 19 (Suppl 1>. 49-53. World Health Organization. WH.O. Handbook for Reporting Results of Cancer Treatment WHO Offset Publication 1979; No. 48 Geneva. Dixon WJ, Brown MB, Engelman L et al. BMDP Statistical Software Manual. Los Angeles, University of California Press 1990; 135-44,145-63, 231-310, 739-68. Mantel N. Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother Rep 1966; 50:163-70. Machin D, Campbell MJ. Statistical tables for the design of clinical trials. Oxford: Blackwell Scientific Publications 1987; 10-34,54-9, 79-88,132-68.

Received 7 April 1993; accepted 5 August 1993. Correspondence to: Paris Kosmidis, M.D. Medical Oncology Department 'Metaxa' Cancer Hospital Botassi51 Piraeus, Greece