Cisplatin, gemcitabine and vinorelbine in locally advanced or ...

Annals of Oncology 8: 1045-1048, 1997. © 1997 Kluwer Academic Publishers Primed in the Netherlands.

Short report Cisplatin, gemcitabine and vinorelbine in locally advanced or metastatic non-small-cell lung cancer: A phase I study G. Frasci,1 N. Panza,2 P. Cornelia,1 G. P. Nicolella,2 M. Natale,3 C. Pacilio,2 A. Gravina,1 V. Caputi,4 G. Botti5 & G. Cornelia1 'Division of Medical Oncology A, ADepartment of Radiology. s Department of Pathology, National Tumor Institute of Naples; 2 Division of Medical Oncology, 33Jst Division of Pneitmology Cardarelli Hospital, Naples, Italy

lasting more than seven days; in five of them it occurred in the first course. Neutropenic fever was observed in four cases. Purpose: The objective of this study was to determine the Grade 4 thrombocytopenia occurred in only two patients. maximally tolerable doses (MTDs) of vinorelbine (VNR) and Non-hematologic toxicity was a minor problem in all patients gemcitabine (GEM) when combined with a fixed dose of but was never dose-limiting. No complete responses were obtained, but sixteen out of 31 (52%) patients achieved partial cisplatin (CDDP). Patients and methods: Chemotherapy-naive patients with responses. The median duration of response was 20 (range stage IIIB-IV non-small-cell lung cancer (NSCLC) received a 6-56+) weeks, while at a nine-month median follow-up, the fixed dose of CDDP (50 mg/m2) and escalating doses of VNR median survival time has not yet been reached. To date, 18 (starting from 20 mg/m2) and GEM (starting from 800 mg/ patients are still alive. The one-year projected survival for all m2) on days 1 and 8, every three weeks. The single escalation of patients was 51%. GEM alone, by 200 mg/m2 at each step, was initially planned Conclusions: Our results show that CDDP, VNR and GEM up to a dose of 1,200 mg/m2, to be followed by increments of can be safely given together without substantial reductions in the VNR dose of 5 mg/m2 at each step. their individual dose intensities. In our opinion, the dose level 2 2 Results: Thirty-one patients were enrolled at five different of GEM 1,000 mg/m + VNR 25 mg/m given in combination 2 dose levels. The escalation was stopped at level 4 (GEM 1,200 with CDDP 50 mg/m on days 1 and 8 of a three-week cycle mg/m2 and VNR 25 mg/m2) since two of six patients of this can be recommended for phase II trials, since it provides a cohort showed dose-limiting neutropenia at treatment cycle 1. better balance in dose intensity of GEM and VNR. A phase II Two different dose levels, GEM 1,200 mg/m2 + VNR 20 mg/ randomised study is underway to establish the activity of this m2, and GEM 1,000 mg/m2 + VNR 25 mg/m2' were fairly well new regimen (at the above-cited dose level) in chemo-naive tolerated. No treatment-related deaths occurred. Neutropenia NSCLC patients. was the main toxic effect, occurring in 76% of the total of 116 cycles delivered, and in 24% of them was of grades 3 or 4. A Key words: cisplatin, gemcitabine, NSCLC, phase I study, total of eight patients (26%) experienced grade 4 neutropenia vinorelbine Summary

Introduction Fifty years of cancer chemotherapy have produced notable success in some neoplasms, but only a few cytotoxic agents (cisplatin, vinka alkaloids, ifosfamide, mitomycin, taxanes, irinotecan, gemcitabine and vinorelbine) have shown an activity rate > 15% in NSCLC patients [1]. VNR has been shown to have very promising activity as a single agent, at the cost of negligible neurotoxic effects [2]. Moreover, in a recent large phase III randomized trial a significantly better response rate and median survival were reported with the CDDP-VNR combination than for a CDDP-vindesine regimen [3]. Gemcitabine (2', 2' difiuorodeoxycytidine) (GEM) is a new anticancer drug with novel metabolic properties and mechanisms of action. Although it is structurally similar to cytarabine, it has shown a much wider spectrum of antitumor activity in vitro [4]. In several phase II

trials GEM as single-agent produced a response rate exceeding 20% in advanced NSCLC patients, with a median survival in excess of eight months [5,6]. Clinical trials combining GEM and CDDP have also been carried out, reporting a more than 50% response rate and a median survival longer than one year [7,8]. In view of these considerations we undertook this phase I study in an effort to determine the maximally tolerable doses of GEM and VNR when combined with a fixed dose of CDDP.

Patients and methods Chemotherapy-naive patients ^ 75 years of age were eligible if they had histologically or cytologically confirmed locally advanced (IIIB) or metastatic NSCLC. No prior chemotherapy or thoracic radiotherapy were allowed. They were required to have adequate bone marrow, liver, and renal function. A performance status ^ 2 on the World

1046 Health Organization (WHO) scale, and a life expectancy of at least 12 weeks were also required. All patients gave their written informed consent, and the protocol was approved by the Ethics Committee for Biology Research of the National Tumor Institute of Naples. Treatment consisted of a fixed dose of CDDP (50 mg/m2) and escalating doses of GEM and VNR on days 1 and 8, with recycling every 3 weeks. The starting doses of GEM and VNR were 800 mg/m2 and 20 mg/m2, respectively. Since we hoped to reach a GEM dose of 1,200 mg/m2, we first planned to increase only the GEM dose by 200 mg/m2 increments up to that level, and thereafter to escalate the VNR dose by 5 mg/m2 at each step up to 30 mg/m2. At least three patients were enrolled in each cohort, and dose escalation was stopped if 33% or more patients of a given cohort showed dose-limiting toxicity (DLT) in treatment cycle 1. DLT was defined as grade 4 neutropenia lasting more than seven days, or grade 4 thrombocytopenia, or grade 3 or 4 nonhematologic toxicity (except for nausea or alopecia). A more than one-week delay in administering the second treatment cycle was also considered a DLT. Although this study was designed primarily to evaluate toxicity, patients underwent complete tumor response assessment after three treatment cycles. An additional three cycles were administered in patients showing complete or partial response according to the WHO response criteria.

Results Characteristics of patients A total of 31 patients entered this study in five dose escalations. There were 26 men and five women with a median age of 63 years (range 43-75). Fifteen patients had stage IIIB and 16 stage IV disease. The histotype was squamous in 14, adenocarcinoma in 11, and largecell or undefined in six patients. The ECOG performance status was 0-1 in 25, and two in six patients. Toxicity Neutropenia was the main dose-limiting toxicity in this study (Table 1). A total of five patients had grade 4 neutropenia lasting more than seven days at the first treatment cycle. Since it occurred in two of six patients treated at level 4, the preceding level (GEM 1,200 mg/ m2 and VNR 20 mg/m2) represented the MTD. However, in order to obtain a better balance in the dose intensity of GEM and VNR, we reduced the dose of GEM to 1,000 mg/m2 and tried to escalate the dose of VNR to 25 mg/m2. Only one of the first six patients at Table 1. Results of the dose escalation. Step

CDDP/VNR/

Pts

Cycles DLT Type

Response

Gem 1)

2) 3) 4) 3a)

Total

50/20/800 50/20/1,000 50/20/1,200 50/25/1,200 50/25/1,000

3 6 7 6 9

10 21 31 20 34

31

116

0/3 0/6 1/7 2/6 2/9

5/31

_ -

neutrop. 2 neutrop. 2 neutrop.

1 PR 1 PR 5 PR 3 PR 6 PR 16 PR (52%)

Abbreviations: CDDP - cisplatin; GEM - gemcitabine; VNR vinorelbine; DLT - dose limiting toxicity; PR - partial response; Pts - patients.

this level showed dose-limiting neutropenia at the first cycle. This cohort was expanded to nine patients, and only two cases of dose-limiting neutropenia occurred in the first cycle. Therefore, we considered this dose level (3a) safe and more advisable for further clinical trials. The hematologic toxicity data concerning all 116 of the delivered cycles were as follows: Neutropenia of different grades occurred in 88 of 116 cycles; it was of grade 3 or4 in 16 of 31 (51%) patients, fora total of 28 of 116 (24%) cycles. An overall eight patients (26%) showed grade 4 neutropenia lasting more than seven days. Neutropenic fever occurred in four patients, all treated at dose level 4. No toxic deaths occurred. A total of 11 patients required the administration of G- or GM-CSF at least once in the course of the treatment. The neutrophil nadir usually occurred after the day 8 administration, between days 11 and 16. The median time to recovery from grade 4 neutropenia was eight (range 3-12) days. Grade 3 or 4 thrombocytopenia was far less frequent, occurring in only four (13%) and two (6%) patients, respectively. Severe anemia requiring RBC transfusion occurred in 13 patients. A delay in chemotherapy administration or a dose reduction due to hematologic toxicity occurred in 21 patients overall, but in only six cases was it before the third cycle. The average actually delivered dose intensity of all drugs was 89%, but it rose to 93% when only the first three cycles were considered. Non-hematologic toxicity was only a minor problem in all patients, never becoming dose-limiting. Vomiting was severe in only three patients, and mild or moderate nephrotoxicity occurred infivepatients. Only one patient had a transient grade 2 elevation of creatinine concentration. Microscopic hematuria and mild proteinuria occurred in 11 and 15 patients respectively, at some time during the course of treatment. Hair loss of various grades was observed in the majority of patients, but was rarely severe. Fatigue was observed in 19 patients but was severe in only three. Five patients showed moderate transient elevations of hepatic enzymes, and four had mild neurotoxic effects consisting mainly of peripheral paresthesias. No treatment-related pulmonary toxicity was observed. Response All patients were included in the analysis of activity on an 'intent to treat' basis. Nine patients discontinued the treatment early (early progression, four; severe toxicity, three; early treatment unrelated death, one; refusal, one). There were no complete responses among the remaining 22 patients who underwent restaging. Sixteen partial responses (52%, 95% CI: 33%-70%), four stable diseases and two progressions were observed (Table 1). Responses occurred at all dose levels, but a trend towards a dose-response relationship was observed, since only two responses occurred in the nine patients (22%, 95% CI: 3%-60%), treated at the first two dose levels, as compared to 14 responses in the 22 patients entered onto the last three dose levels (63%, 95% CI: 41%-83%).

1047 Table 2. Maximum heraatological toxicity in the 116 delivered cycles Step

1 2 y 4 3a

Patients

3 6 7 6 9

Courses

10 21 3}

20 34

Thrombocytopenia

Neutropenia

Anemia

1

2

3

4

1

2

3

4

1

2

3

4

4 2 H 7 11

2 4 ^ 5 8

1 2 4 4 5

1 1 2 4 4

3 2 7 5 8

1 1 4 4 6

2 1 2-3 3

0 0 0 1 1

3 6

0 5 6 3 7

0 1 3 4 5

0 1

5 6

t 1 0

dose (keeping the VNR dosefixedat 20 mg/m2), since a clear dose-response relationship for GEM alone has not been demonstrated. A review of the single-agent studies suggested that there might be a threshold dose for GEM. In fact, almost no responses were seen in patients treated at doses ^800 mg/m2/week, while no definite doseDiscussion response correlation was seen above 1,000 mg/m2/wk [9]. Our data show that the administration of GEM and On the other hand, we believe that the increase of VNR at doses of 1,000 mg/m2 and 25 mg/m2, or 1,200 vinorelbine from 20 mg/m2 to 25 mg/m2 on days 1 and 8 mg/m2 and 20 mg/m2, respectively, in combination with could translate into substantial therapeutic benefit, since 50 mg/m2 of CDDP on days 1 and 8 is associated with a clear dose-response relationship has been reported for manageable hematological toxicity and causes only negli- vinorelbine in the past [10]. In conclusion, a CDDP-GEM-VNR combination is gible non-hematological side effects in chemotherapynaive NSCLC patients. Since the former dose level allows feasible at doses which, for the respective drugs, have a better balance in the dose intensity of GEM and VNR proven therapeutic activity in NSCLC patients. This to be obtained, we recommend it for future clinical trials. combination is usually manageable and rarely life-threatBoth of the above-cited dose levels translate into dose ening. A phase II randomised trial is underway to better intensities of GEM (670-800 mg/m2/week) similar to define the level of activity of this combination in chemothose reported by most authors who have tested the naive advanced NSCLC patients. CDDP-GEM combination, and into an even higher dose intensity of CDDP (33.3 mg/m2/week) [7,8]. Although frequent delays in chemotherapy adminis- References tration or dose reductions were also necessary after the 1. Comis RL, Friedland DM. New chemotherapy agents in the first treatment cycle, the average actually delivered dose treatment of advanced non-small-cell lung cancer: An update intensity was quite satisfactory (89%) in our study, espeincluding data from the Seventh World Conference on Lung cially if only the first three treatment cycles, which probCancer. Lung Cancer 1995; 12: S63-9 (Suppl 2). ably most affect the chance of response, are considered 2. Depierre A, Lemarie E, Dabouis G et al. A phase II study of (93%). navelbine (vinorelbine) in the treatment of non-small cell lung cancer. Am J Clin Oncol 1991; 14: 115-9. We think that the relevant doses of GEM and VNR 3. Le Chevalier T, Bnsgand D, Douillard JY et al. Randomized reached in our trial, and the quite satisfactory actually study of vinorelbine and cisplatin versus vindesine and cisplatin delivered dose-intensity in the first three cycles, are the versus vinorelbine alone in advanced non-small-cell lung cancer: consequence of the schedule we chose. In fact, the temResults of a European Multicenter Trial including 612 patients. J poral occurrence of neutrophil count nadir, usually durC1in Oncol 1994; 12: 360-6. ing the second week, and its duration of no longer than 4. Hertel LW, Broder GB, Croin JS, et al. Evaluation of the antitumor activity of gemcitabine (2', 2' difluorodeoxycytidine). Canone week in most cases during the first three cycles, cer Res 1990; 50: 4417-22. allowed us to deliver the scheduled doses on day 8, and to recycle the treatment at full doses on day 21 in all but 5. Anderson H, Lund B, Back Fetal. Single-agent activity of weekly gemcitabine in advanced non-small-cell lung cancer: A phase II six patients in the first three treatment cycles. study. J Clin Oncol 1994; 12: 1821-6. Assessment of the therapeutic activity was not a 6. Gatzmeier U, Sheperd FA, Le Chevalier T et al. Activity of main end point of this study. However, the response rate gemcitabine in patients with non-small-cell lung cancer: A multicentre extended phase II study. Eur J Cancer 1996; 32A: 243-8. achieved (52%), and the survival outcome observed 7. Crino L, Scagliotti G, Marangolo M et al Cisplatin-gemcitabine (51% projected one-year survival), suggest that this new combination in advanced non-small-cell lung cancer: A phase II regimen deserves further evaluation in phase II—III study. J Clin Oncol 1997; 15: 297-303. trials. 8. Abratt RP, Bezwoda WR, Goedhals L and Hacking DJ. Weekly We decided to adopt the dose level 3a (GEM 1,000 gemcitabine with monthly cisplatin: Effective chemotherapy for mg/m2 and VNR 25 mg/m2) for further clinical trials, advanced non-small-cell lung cancer. J Clin Oncol 1997; 15: 744-50. rather than to attempt a further escalation of the GEM The median duration of response was 20 (range 6-56+) weeks. After a 9- (range 2-16) month median follow up, the projected one-year survival was 51%.

1048 9. Eli Lilly. New Drug Submission. Canadian Comprehensive Summary Clinical Synopsis, Part 3, Vol. I, 1994, p. 938. 10. Souquet PJ, Fournel P, Bohas CH et al. Cisplatin, ifosfamide, with increasing dosages of vinorelbine (NIP) in advanced non-smallcell lung cancer (NSCLC). Lung Cancer 1994, 11:115 (Suppl 1). Received 21 July 1997; accepted 1 September 1997

Correspondence to: Dr. Giuseppe Frasci Division of Medical Oncology A National Tumor Institute via M. Semmola 80131 Naples Italy