The first recognition that advanced lung carcinoma could favourably respond to chemotherapy was presented in 1948 by Karnofsky et al., in a report describing the palliative effects of nitrogen mustard in the treatment of bronchogenic carcinoma (1).

Despite this first "promising" experience, during the two decades from 1970 to 1990 the results of chemotherapy for stage IV disease have been unimpressive. Cisplatin was demonstrated to have modest activity, and appeared to be synergistic with etoposide and vinca alkaloids. Numerous phase III studies compared different cisplatin combination chemotherapy regimens and failed to identify a "standard" program (2,3). Furthermore, phase III studies comparing chemotherapy to supportive care had mixed results, with others failing to demonstrate an advantage for chemotherapy in stage IV disease. Only from a recent meta-analysis a small advantage for the combinations containing cisplatin could be demonstrated (4).

During the past 5 years, several new agents have been evaluated in NSCLC and demonstrated improved results compared to older regimens. The first of these, vinorelbine, was developed in France. Le Chevalier and colleagues conducted a phase III study in 612 evaluable patients with NSCLC (5). Patients were randomized to vinorelbine plus cisplatin, vinorelbine alone, or vinorelbine + cisplatin. Objective responses were seen in 30%, 14% and 19% of patients. Furthermore, there was statistically significant improved survival with cisplatin + vinorelbine, with 35% one year survival versus 27% for cisplatin + vindesine. The Southwest Oncology Group (SWOG) in the U.S. evaluated single agent cisplatin (100 mg/m²) versus the same cisplatin + vinorelbine (25 mg/m²) weekly in 432 patients with NSCLC. The response rates were 10% versus 26%, and there was a statistically significant survival advantage for the vinorelbine regimen, with 16.4% versus 35.4% 1 year survival (6).

The taxanes (paclitaxel and docetaxel) have also demonstrated activity, with a 20-30% single agent response. In the U.S., paclitaxel achieved a 40% 1 year survival in studies by Eastern Co-operative Oncology Group (ECOG) and M.D. Anderson, ECOG subsequently performed a phase III study in 571 evaluable patients, randomizing patients to cisplatin (75 mg/m²) + etoposide, versus cisplatin (75 mg/m²) + paclitaxel, given as a 24 hour infusion at a dosage of 135 mg/m² or 250 mg/m² with G-CSF. Response rates were 12%, 27% and 32% and there was improved survival with the two paclitaxel arms (7).

Gemcitabine is one of the most extensively evaluated single agents in NSCLC, with response rates of 20-30% world-wide (8). Several phase II studies have evaluated cisplatin + gemcitabine, with 30-50% response rates and 40% 1 year survival (9).

Carboplatin + paclitaxel (3 hour or 24 hour infusions) are among the most popular regimens in the U.S., based entirely upon phase II studies (10). SWOG is evaluating cisplatin + vinorelbine versus carboplatin + paclitaxel and ECOG cisplatin + paclitaxel (135 mg/m²) versus carboplatin + paclitaxel versus cisplatin + docetaxel versus cisplatin + gemcitabine. Other large randomized trials comparing the new combinations are being performed in Europe.

It is reasonable to expect that one of these new regimens will achieve improved survival in advanced disease and more probably when used as adjuvant therapy or combined with radiotherapy in earlier stage disease.

References
1. Karnofsky DH, et al.: Cancer 1: 634-656, 1948.
2. Weick JK, et al.: J Clin Oncol 9: 1157-1162, 1991.
3. Crinò L, et al.: Ann Oncol 6: 347-353, 1995.
4. Stewart LA, et al.: BMJ 311: 899-909, 1995.
5. Le Chevalier T, et al.: J Clin Oncol 12: 360-367, 1994.
6. Wozniak AJ, et al.: Proc ASCO 15, 374, 1996.
7. Bonomi O, et al.: Proc ASCO 15, 382, 1996.
8. Gatzemeier U, et al.: Eur J Cancer 32: 243-248, 1996.
9. Crinò L, et al.: J Clin Oncol 15: 297-303, 1997.
10. Johnson D, and Einhorn LH: J Clin Oncol 13: 1840-1842, 1995.

________________________________________________________________________________

DRUG THERAPY OF NON-SMALL CELL LUNG CANCER: NEWS FROM THE LAB
A. Ardizzoni
Division of Medical Oncology I, Istituto Nazionale per la Ricerca sul Cancro, Genova - Italy

This review will cover two aspects regarding preclinical testing of new agents in NSCLC: 1) Preclinical results with the new generation of chemotherapeutic agents currently under clinical testing. Have the promise been confirmed? 2) Preclinical results of novel non-chemotherapeutic agents.

The new generation of chemotherapeutic agents, some of which are already approved for treating NSCLC while others are still under clinical testing, have been brought to the clinic mainly because of preclinical positive results. Taxanes, gemcitabine and camptothecin analogs have all been shown to be more potent in inhibiting the growth of a various NSCLC cell lines as compared to first-generation agents such as cisplatin (1). Preliminary results of phase II-III studies seem to confirm preclinical observations indicating possible superiority of these new agents in comparison with first-generations ones (2). In addition, in-vitro studies have shown lack of cross-resistance among these agents. Also this preclinical observation seems to be confirmed in the clinical setting. In fact, a number of studies have shown activity of these new agents in patients resistant to prior standard chemotherapy (3). Another interesting observation which has been made in the lab (1) and now seems to be confirmed in the clinical setting is that new active agents for the treatment of NSCLC may have a differential activity in squamous vs non-squamous lung tumors (4).

Among new agents which are still in development for the treatment of NSCLC, the most promising belong to categories different from that of chemotherapeutics. These include 1) antimetastatic and antiangiogenesis agents, 2) small molecules interfering with growth factors and their signal transduction pathways and 3) gene therapy.

1) Matrix metalloproteinases (MMP) such as collagenase, stromelysin, and gelatinase are proteolitic enzymes responsible for degradation of extracellular matrix involved in tumor progression and metastatic process. A series of agents which are able to inhibit MMP, and thereby to prevent tissue matrix break down induced by these enzymes during the process of metastatic spread, have been synthesized (5). The prototypes of these new anticancer agents are batimastat and the related oral compound marimastat. These drugs are low molecular-weight peptides which inhibit MMPs by interacting with the zinc atom in the active site of the enzyme. Both agents have been found to block metastatic formation and tumor neovascularization in animal experiments. Unpublished studies have shown activity of this class of compounds also in lung cancer xenograft. Preliminary data from early phase I studies indicate possible activity with minimal toxicity.

Several experimental data indicate that angiogenesis is crucial for growth and persistence of solid tumor and of their metastases (6). One of the first agent to display neoangiogenesis inhibitory properties has been suramin . Although suramin has shown, at least in vitro, to yield growth inhibition in NSCLC cell lines, probably through an anti-growth factor mechanism (7), clinical studies in NSCLC have yield negative results. Analogs of suramin, such as FCE 26644 (8), less toxic and easier to administer, have been synthesized and are currently under phase I testing. More recently, potent angiogenesis inhibitors, angiostatin and endostatin, have been identified and sequenced. Angiostatin, a 38kD fragment of plasminogen and endostatin, a 20kD C-terminal fragment of collagen XVIII, specifically inhibits endotelial proliferation inducing a balance between poptosis and proliferation of metastatic and primary tumor cells defined as "dormancy". Both agents are capable of inhibiting tumor growth in vivo in animal models. Although tumors re-grow when anti-angiogenic treatment is discontinued, experimental tumors, including the Lewis lung tumor, remain sensitive to a second cycle of treatment with the same agent. Different from conventional chemotherapeutic agents, no drug-resistance was observed even after multiple repeated cycles of antiangiogenic therapy (9). These agents hold considerable promise for the treatment of a number of tumor types including NSCLC. Other drugs which, in preclinical models, have been found to be angiogenesis inhibitors and that are currently entering clinical trials include TNP-470 (a synthetic fumagallin derivative), thalidomide, vitaxin and squalamine.

2) An anti-growth factor therapy can be accomplished through out two different strategies. One is to hamper the binding of growth factor to their receptors by attacking one of these two; the other is to interfere with the transduction of the mitotic signal which occurs after the growth factor has bound to its receptor. This latter strategy is achieved by inhibiting signal transduction molecules such as tyrosine kinase (PTK) and protein kinase C (PKC). The rationale is that the inhibition of deregulated cellular signaling pathways in malignant cells at the level of PKC o PTK, where multiple mitogenic signals converge, should lead to the inhibition of tumor growth. PKC inhibitors, in addition to reduce tumor growth, are also able to sensitize cancer cells to cytotoxic agents by reverting the MDR phenotype. A series of new drugs with anti-PTK or anti-PKC properties have been recently developed and are close to enter clinical trials in a number tumor types including lung cancer (10).

Another enzyme involved in signal transduction which has been target for anticancer drug development is farnesyl transferase. Farnesylation is necessary for the association of ras proteins to the cell membrane and therefore for their transforming activity. A number of compounds which are able to interact with farnesylation of H-ras and K-ras proteins have been developed. Probably due the presence of transforming events upstream of ras, these agents have antiprolipherative effects also on tumor lines which lack an activated ras oncogene. Experiments in human tumor xenografts, including A-549 lung cancer xenografts, have shown almost 90% growth inhibition (11). Some of these agents are presently undergoing phase I clinical evaluation in a number of tumor types including NSCLC.

3) Gene therapy, particularly gene-replacement therapy, is receiving a lot of attention as a possible novel strategy for the treatment of a number of neoplastic disease including NSCLC. The correction of tumor genetic abnormalities by gene transfer approaches may result in control of cell proliferation and increased apoptosis. In particular, the correction of a single critical alteration, as in the case of p53, has been shown to suppress the growth of human lung cancer cells bearing multiple genetic lesions (12). Interestingly, transduced cells expressing normal p53 mediate killing of non-transduced cells (bystander effect). Preclinical studies carried out in an orthotopic lung cancer model have shown that intratumoral injection of viral vectors expressing wild-type p53 is able to induce regression of established tumors. In addition, animal experiments have shown that wild-type p53 gene transfer induce chemosensitivity to DNA-damaging agents such as cisplatin. The transfer of wild type p53 by the use of retroviral or adenoviral vectors has been already attempted in NSCLC patients, with some evidence of increased apoptosis and tumor regression (13). A number of clinical trials of locally delivered wild-type p53 gene transfer, with or without chemotherapy, are in progress.

References
1. Loprevite M, Favoni RE, de Cupis A et. Pre-clinical evaluation of new antineoplastic agents in NSCLC cell lines: Correlation of antiproliferative activitywith histological subtype and molecular pattern. Lung Cancer 18 (Suppl.1: Proceedings 8th World Conference on Lung Cancer, Dublin): 18, 1997
2. Giaccone G, Postmus P, Splinter T, et al. Cisplatin/Paclitaxel vs Cisplatin/teniposide for Advanced Non-Small Cell Lung Cancer. Preliminary results. Oncology (Suppl 3) 11: 11-14, 1997
3. Fossella FV, Lee JS, Shin DM et al. Phase II studies of Docetaxel for advanced or metastatic platinum-refractory non-small cell lung cancer. J Clin Oncol 13: 645-651, 1995
4. Perez-Soler R, Fossella FV, Glisson BS et al. Phase II study of topotecan in patients with advanced non-small cell lung cancer previously untreated with chemotherapy. J Clin Oncol 14: 503-513, 1996
5. Brown PD, Giavazzi R. Matrix metalloproteinase inhibition: A review of antitumor activity. Annals of Oncology 6: 967-974, 1995
6. Folkman J. What is the evidence that tumors are angiogenesis dependent? J. Natl. Cancer Inst. 82:4-6, 1989
7. Favoni RE, Ravera F, Pirani P et al. Suramin interferes with autocrine/paracrine insuline-like growth factor-I-controlled proliferative loop on human lung cancer cell lines. Eur J Pharmacology 264: 199-206,1994
8. de Cupis A, Ciomei M, Pirani P, et al. Anti-insulin-like growth factor-1 activity of a novel polysulphonated distamycin A derivative in human lung cancer cell lines. Br J Pharmacology 120: 537-543, 1997
9. Boehm T, Folkman J Browder T and O'Reilly MS. Antiangiogenic therapy of experimental cancer does not induce aquired drug resistance. Nature 390: 404-407, 1997
10. Capogiro F, French RC, Kaye SB Protein kinase C: a wothwhile target for anticancer drugs? Anticancer Drugs 8: 26-33, 1997
11. Przybranowski, Vincent PW, Lath a C et al. In vivo evaluation of farnesyltransferase inhibitor, PD 169451, versus a panel of human tumor xenografts. Proceedings AACR 39: 269 (abstr. 1841), 1998
12. Takahashi T, Carbone D, Nau MM, et al. Wild type but not mutant p53 suppresses the growth of human lung cancer cells bearing multiple genetic lesions. Cancer Res 52: 2340-2343, 1992
13. Roth. JA, Swisher SG, Lawrence DD et al. Gene replacement for lung cancer. Lung Cancer S2: 76, 1997.

________________________________________________________________________________

ACTIVITY OF DRUGS RECENTLY INTRODUCED IN THE MARKET FOR THE TREATMENT OF NON-SMALL CELL LUNG CANCER
J.P. SCULIER, A.P. MEERT, T. BERGHMANS, P. BRANLE, F. LEMAÎTRE, C. MASCAUX, E. RUBESOVA, P. VERMYLEN, M. PAESMANS
Service de Médecine, Institut Jules Bordet, Brussels, Belgium.

During the last decade, chemotherapy has been shown to significantly improve survival in patients with non-small cell lung cancer (NSCLC), whatever the disease stage - loco-regionally advanced or metastatic - is. The most active "classical" cytotoxic drugs in that indication are cisplatin, mitomycin C, ifosfamide, vindesine and vinblastine , as reported in a meta-analysis that we performed in 1990 (Lung Cancer, 7: 243-252;1991). The level of evidence of survival benefit is based on multiple randomised trials testing a cisplatin-containing chemotherapy arm versus a control arm without chemotherapy. There are very few data yet available about non-cisplatin based regimens.

A series of new active drugs against NSCLC has been identified during the last ten years and most of them are today commercially available, at least in some countries. They are gemcitabine ( a pyrimidine analogous antimetabolite like cytarabine), the taxans paclitaxel and docetaxel (inhibitors of the depolymerisation of the tubulin molecules of the microtubules, blocking the cell in phase M), the camptothecin derivatives irinotecan and topotecan ( topoisomerase I inhibitors), vinorelbine (a vinca-alcaloid inhibiting tubulin polymerisation).

Our present purpose is to report a systematic review of the literature on this topic, with a trial quality assessment . Were considered all phase II and III studies published until 1997 with one of these new drugs included into the investigated treatment regimens. The trial quality was assessed by using the European Lung Cancer Working Party (ELCWP) quality score for phase II studies, taking into account criteria for the intrinsic ("scientific") properties (internal validity: 50 points) and for the generabilisity potential (external validity: 50 points).

We identified as eligible for the analysis 55 articles, including 48 phase II, 3 phase II randomised and 4 phase III randomised trials. They concerned paclitaxel in 15, docetaxel in 6, gemcitabine in 11, vinorelbine in 21 and topotecan and irinotecan each in 2. The overall mean quality score was 62.7% (95% CI: 59.0-66.3) ranging from 33.1 to 89%, with respective internal and external mean subscores of 26.0/50 ( 95% CI: 23.6-28.4; range: 2.5-50) and of 36,9/50 (95% CI: 35.1-38.7; range: 19.8- 50). The data per drug are summarised in the following table:

Our knowledge about the new drugs according to the fully published articles can be summarized in the following way:

1.Paclitaxel
It is active as single agent therapy in non pretreated NSCLC and probably also after prior chemotherapy. Combinations with cisplatin or carboplatin have been the object of multiple phase II studies. Publication of controlled trials is waited.

2.Docetaxel
It is active as single agent therapy in non pretreated and probably also chemotherapy-pretreated NSCLC. Publication about combination regimens is waited.

3.Vinorelbine
It is active as single therapy in non pretreated NSCLC. It has been successfully combined with cisplatin. Vinorelbine has been investigated in 4 randomised trials. Administered alone, it has been shown to significantly improve survival in comparison to 5-fluorouracil (Crawford, J Clin Oncol 14:2774; 1996) and response rate (but not survival) in comparison to vindesine ( Furuse, Ann Oncol 7: 815; 1996). The combination of vinorelbine to cisplatin has been compared in two randomised trials to vinorelbine alone with significant response rates improvement with survival increase in one (Le Chevalier, J Clin Oncol 12: 360; 1994) and not in the other (Depierre, Ann Oncol 5: 37; 1994). In the first one, the comparison was also performed with a third arm treatment made of cisplatin and vindesine, with a marginal effect on survival. It should be noted that both trials have obtained high quality score in our review (89 and 79,8%).

4.Gemcitabine
It is active as single therapy in non pretreated NSCLC and has been shown in phase II randomised trials to be better tolerated than the combination cisplatin-etoposide. It has been successfully combined with cisplatin and ifosfamide in phase II studies. Results of phase III trials are waited.

5.Topotecan
Two studies assessing its effectiveness as single therapy in non pretreated NSCLC have shown contradictory results. Further data are required to specify the place of topotecan.

6.Irinotecan
The drug has been shown active against unpretreated NSCLC. Confirmatory trials are required.

In conclusion, new drugs have appeared with promising activity against NSCLC. However, the data today available in the literature are too limited to define exactly their role and their indications in the management of this disease.

________________________________________________________________________________

THE RESPONSE-SURVIVAL RELATIONSHIP
J.R. JETT, M.D.
Mayo Clinic, Rochester, Minnesota, USA

I. A) The objective of a Phase I trial is to determine a dose that is appropriate for use in Phase II trials. Phase I trials of chemotherapeutic agentstypically start at a dose of one-tenth the median lethal dose (LD10) in the most sensitive species. The dose is then escalated in a step-wise fashion with careful monitoring of toxicity.

B) The objective of Phase II trial is to determine if the drug/agent has anti-tumor activity against the tumor type in question.

1) Tumor response is not a direct measure of patient benefit. A treatment that causes a partial response is not necessarily beneficial to the patient.

2) Analyses that demonstrate that responders live longer than non- responders are invalid for concluding that the treatment extended survival (1,2).

Why?

1. a) Responders live long enough to achieve that status
2. b) Responders may have more favorable prognostic factors
3. c) Treatment may shorten survival of non-responders but not influence survival of responders

3) In order to demonstrate that treatment extends survival, it must be demonstrated that the treated group as a whole lives longer than an appropriate control group.

4) Phase II trials do not have an internal control group and, therefore, drawing conclusions about survival from such trials is problematic.

C) Phase II trial with survival as an end-point:

1) It has been observed that some combination regimens produce high response rates in Phase II trials, but do not result in improved survival in Phase III trials.

2) In some cancers, response rate is difficult to measure and many patients do not have measurable disease, i.e., brain, pancreas, ovarian.

3) Tables have been developed for planning historically controlled Phase II trials with survival or time to progression as the endpoint (3).

D) Major endpoints of Phase III trials are Survival and Symptom Control.

II. Question of relationship between response and survival in NSCLC

1. Eastern Cooperative Oncology Trial (EST 1583) (4)

III. Definition of Response
A) Measurable Disease: PR is defined as ³ 50% reduction in the sum of the longest perpendicular diameters of the indicator lesions (5).
B) Assessable Disease: Defined as a clinically apparent lesion on examination or x-ray that is not measurable in two dimensions: a regression REGR is defined as a definite decrease, but not total disappearance of the assessable lesion (5).
C) Measurable vs Evaluable: Does it matter?
D) Independent review of responses: Phase II trials of gemcitabine (6).

IV. Review of Response and Survival Relationship in Phase II/III Trials of The North Central Cancer Treatment Group

Conclusion
A) In Phase III trials survival and QOL endpoints are all that counts
B) Response rates are subject to inter-observer variability
C) Failure to monitor tumor status may result in treatment of patients with progressive disease for a longer period of time and thus expose patient to more toxicity

References
1. Anderson JR, Cain KC, Gelber RD: Analysis of survival by tumor response. J Clin Oncol 1983; 1:710
2. Simon R, Makuch RW: An nonparametric graphical representation of the relationship between survival and the occurrence of an event: Application to responder versus non-responder bias. Stat Med 1984; 3:1
3. Dixon DO, Simon R: Sample size considerations for studies comparing survival curves using historical controls.J Clin Epidemiol 1988; 41:1209
4. Bonomi PD, Finkelstein DM, Ruckdeschel JC et al: Combination chemotherapy versus single agents followed by combination chemotherapy in stage IV non-small cell lung cancer. J Clin Oncol 1989; 7:1602-1613
5. Jett JR, Su JQ, Krook JE et al: Measurable or assessable disease in lung cancer trials:Does it Matter?J Clin Oncol 1994; 12:2677-2681
6. Gwyther SJ, Aapro MS, Hatty S et al: Results of independent review of 374 patients with non-small cell lung cancer treated with gemcitabine. Lung Cancer 1997; 18(suppl 1): 10 (Abst #29)

________________________________________________________________________________

RANDOMIZED TRIALS OF BEST SUPPORTIVE CARE PLUS CHEMOTHERAPY VS. BEST SUPPORTIVE CARE (BSC) ALONE IN INOPERABLE NON-SMALL CELL LUNG CANCER (NSCLC)
G. BUCCHERI, MD
Ospedale S. Croce e Carle, Cuneo, Italy

Despite prior controversies1-5 and debates6-8, palliative purpose chemotherapy, whenever possible, has become a widely accepted treatment for inoperable NSCLC. Data in support of such a therapeutic approach are abundant and include many controlled studies, and four meta-analyses9. So far, at least eleven trials of chemotherapy vs. BSC in inoperable NSCLC have been peer-reviewed and published10-21 (Table 1). Two additional trials were presented at the 1966 American Society of Clinical Oncology (ASCO) Annual Meeting22 and during the last IASLC World Conference on Lung Cancer23. Invariably, all studies showed a prolongation of the survival times recorded after chemotherapy10-22. Confirming our prior qualitative evaluation3, all the following meta-analyses showed that chemotherapy did result in a statistically significant survival advantage24-27. The evidence was relevant to patients with inoperable NSCLC, taken as a unique and homogeneous group.

The question arises: "do these results remain valid, if we consider only patients with locally advanced disease?…" Five of the aforementioned studies were limited to metastatic patients15,18,20,21,23, and they are therefore inconsequential to this issue. In the remaining eight trials (Table 1), varying proportions of stage III patients composed the cohorts of study. In one of these trials, a subgroup analysis for non metastatic patients was accomplished16. Studies were much different as to the year of publication, drugs and regimens used, planned duration of treatment, and characteristics of the population studied. One of them, the Canadian trial reported by Rapp and al.14, showed significantly improved survivals after two different programs of chemotherapy. It was a fairly large, well-designed study, using modern drugs. owever, only twenty-seven patients with stage III disease, out of the 233 enrolled, are assessable. Two older studies, dating from the 60's and early 70's, had shown a significant difference10,11, or suggested a trend toward longer survivals in patients being treated with chemotherapy12. In these studies, patients with intrathoracic disease were sufficiently numerous. However, minimally active drugs (i.e., alkylating agents, and methotrexate) were used. One might suppose that more active regimens would have produced more remarkable results. In the early 80’s, another trial had shown a similar, statistically significant survival difference, but this was a very small study, containing only 17 non metastatic patients13. In 1990, we reported the results of a longitudinal study, where patients on BSC made up a historical control group19. Strictly speaking, this was a non randomized study. Nonetheless, it provided further evidence in favor of chemotherapy, strengthened by the rather large number of patients (175, 95 of whom had a locally advanced disease)19. Other two studies found various increases in the survival rate of patients treated with chemotherapy, but differences did not reach the statistical significance16,17. However, in the Anglo-Australian trial reported by Woods and co-workers16, the difference approached the statistical significance (log rank p=0.075), specifically in the subgroup of patients with no tumor dissemination. The authors reported a median survival of 45 weeks for the 25 stage III patients allocated to chemotherapy, and of only 26 weeks for the 39 untreated controls. To date, an additional trial of chemotherapy vs. BSC in both metastatic and locally advanced NSCLC has been published in abstract form22 and also this trial suggests a statistically significant survival advantage for patients receiving cisplatin-vindesine-mitomycin chemotherapy.

So far, no randomized study has investigated the addition of chemotherapy to the BSC, exclusively in patients with non metastatic NSCLC. However, studies containing varying proportions of stage III patients had similar or better results than studies limited to stage IV patients. Biological considerations support the idea that the beneficial effect of treatment is greater in smaller tumors, where higher response rates are more frequently observed28.

In conclusion, If we accept the idea that chemotherapy is effective in both the metastatic and non metastatic inoperable NSCLC, we should also accept the view that chemotherapy is at least equally effective in the two subgroups of patients, i.e., in the more favorable and treatable subgroup of non metastatic subjects as well as in the less prognostically favored patients with metastases.

Table 1 Studies of best supportive care plus chemotherapy vs. best supportive care alone in locally advanced or metastatic non-small cell lung cancer.

References
1. Haskell CM. Chemotherapy and survival of patients with non-small cell lung cancer: a contrary view. Chest 1991; 99: 1325-1326.
2. Vokes EE, Bitran JD, Vogelzang NJ. Chemotherapy for non-small cell lung cancer: a continuing challenge. Chest 1991; 99: 1326-1328.
3. Buccheri G. Chemotherapy and survival in non-small cell lung cancer. The old 'vexata questio'. Chest 1991; 99: 1328-1329.
4. Masters GA, Vokes EE. Should non-small cell carcinoma of the lung be treated with chemotherapy? Pro: Chemotherapy is for non-small cell lung cancer. Am J Respir Crit Care Med 1995; 151: 1285-1287.
5. Douglas IS, White SR. Should non-small cell carcinoma of the lung be treated with chemotherapy? Con: Therapeutic empiricism--The case against chemotherapy in non-small cell lung cancer. Am J Respir Crit Care Med 1995; 151: 1288-1291.
6. Johnson DH. Chemotherapy for metastatic non-small-cell lung cancer--Can that dog hunt. JNCI 1993; 85: 766-767.
7. Coates A, Forbes J. Is chemotherapy for non-small cell lung cancer worthwhile. Lancet 1993; 342: 4
8. Smith IE. Palliative chemotherapy for advanced non-small cell lung cancer. BMJ 1994; 308: 429-430.
9. Buccheri G. Chemotherapy and survival in non-small cell lung cancer: Three years later. Chest 1994; 106: 990-992.
10. Wolf J, Spear P, Yesner R, Patno ME. Nitrogen mustard in the palliative treatment of inoperable bronchogenic carcinoma. Am J Med 1960; 29: 1008-1022.
11. Green RA, Humphrey E, Close H, Patno ME. Alkylating agents in bronchogenic carcinoma. Am J Med 1969; 46: 516-525.
12. Selawry O, Krant M, Scotto J, et al. Methotrexate compared with placebo in lung cancer. Cancer 1977; 40: 4-8.
13. Cormier Y, Bergeron D, La Forge J, et al. Benefits of polychemotherapy in advanced non-small-cell bronchogenic carcinoma. Cancer 1982; 50: 845-849.
14. Rapp E, Pater JL, Willan A, et al. Chemotherapy can prolong survival in patients with advanced non-small-cell lung cancer-Report of a Canadian multicenter randomized trial. J Clin Oncol 1988; 6: 633-641.
15. Ganz PA, Figlin RA, Haskell CM, La Soto N, Siau J, for the UCLA Solid Tumor Study Group. Supportive care versus supportive care and combination chemotherapy in metastatic non-small cell lung cancer. Does chemotherapy make a difference? Cancer 1989; 63: 1271-1278.
16. Woods RL, Williams CJ, Levi J, et al. A randomized trial of cisplatin and vindesine versus supportive care only in advanced non-small cell lung cancer. Br J Cancer 1990; 61: 608-611.
17. Cellerino R, Tummarello D, Guidi F, et al. A randomized trial of alternating chemotherapy versus best supportive care in advanced non-small-cell lung cancer. J Clin Oncol 1991; 9: 1453-1461.
18. Quoix E, Dietemann A, Charbonneau J, et al. Disseminated non-small cell lung cancer (NSCLC): a randomized trial of chemotherapy (CT) versus palliative care (PC). Lung Cancer 1988; 4 (suppl.1): 127a(Abstract)
19. Buccheri G, Ferrigno D, Rosso A, Vola F. Further evidence in favour of chemotherapy for inoperable non-small cell lung cancer. Lung Cancer 1990; 6: 87-98.
20. Kaasa S, Lund E, Thorud E, Hatlevoll R, Host H. Symptomatic treatment versus combination chemotherapy for patients with extensive non-small cell lung cancer. Cancer 1991; 67: 2443-2447.
21. Cartei G, Cartei F, Cantone A, et al. Cisplatin-cyclophosphamide-mitomycin combination chemotherapy with supportive care versus supportive care alone for treatment of metastatic non-small-cell lung cancer. JNCI 1993; 85: 794-800.
22. Thongprasert S, Sanguanmitra P, Umsawasdi T. Relationship between quality of life and clinical outcomes in advanced non-small cell lung cancer: best supportive care (BSC) versus BSC plus chemotherapy. Proc Am Soc Clin Oncol 1995; 14: 358(Abstract)
23. Cullen MH, Woodroffe CM, Billingham LJ, et al. Mitomycin, ifosfamide, and cisplatin (MIC) in non-small cell lung cancer (NSCLC): 2. Results of a randomized trial in patients with extensive disease. Lung Cancer 1997; 18: 5(Abstract)
24. Souquet PJ, Chauvin F, Boissel JP, et al. Polychemotherapy in advanced non small cell lung cancer: A meta-analysis. Lancet 1993; 342: 19-21.
25. Grilli R, Oxman AD, Julian JA. Chemotherapy for advanced non-small-cell lung cancer: How much benefit is enough. J Clin Oncol 1993; 11: 1866-1872.
26. Marino P, Pampallona S, Preatoni A, Cantoni A, Invernizzi F. Chemotherapy vs supportive care in advanced non-small-cell lung cancer: Results of a meta-analysis of the literature. Chest 1994; 106: 861-865.
27. Pignon J, Stewart LA, Souhami RL, Arriagada R, NSCLC Collaborators Group. A meta-analysis using individual patient data from randomised clinical trials (RCTS) of chemotherapy (CT) in non-small cell lung cancer (NSCLC): (2) survival in the locally advanced (LA) setting. Proc Am Soc Clin Oncol 1994; 13: 334 (Abstract 1109).
28. Johnson BE. Biologic and molecular prognostic factors--Impact on treatment of patients with non-small cell lung cancer. Chest 1995; 107 Suppl. 287S-290S.
29. Anonymous. Over 200 patients now in the trial! Big Lung Trial Newsetter 3, March 1997.

________________________________________________________________________________

CHEMOTHERAPY IN NON-SMALL CELL LUNG CANCER: A META-ANALYSIS OF INDIVIDUAL PATIENT DATA FROM 11 RANDOMISED TRIALS COMPARING BEST SUPPORTIVE CARE CHEMOTHERAPY WITH BEST SUPPORTIVE CARE ALONE
J.F. TIERNEY, L.A. STEWART
MRC Cancer Trials Office, Cambridge, UK

An individual patient data (IPD) meta-analysis of chemotherapy in non-small cell lung cancer (NSCLC) has been conducted and published by the NSCLC Collaborative Group [1]. This meta-analysis investigated chemotherapy in all treatment settings (early, locally advanced and advanced disease) and looked at the effect of chemotherapy when given in addition to surgery, surgery + radiotherapy, radical radiotherapy and best supportive care. This presentation will concentrate on the results for the comparison of best supportive care + chemotherapy versus best supportive care alone. Although two other meta-analyses [2,3] have also addressed this question, these were based only on data extracted from the literature, therefore suffer from a number of potential problems and biases [4] and are consequently less reliable than the IPD approach [5] which is regarded by most as the gold standard of systematic review.

Individual randomised trials investigating chemotherapy in advanced NSCLC have been small and lacked statistical power. Consequently, although a small number of trials have reported statistically significant results, most have been inconclusive. The IPD meta-analysis was initiated in 1991 to determine whether or not chemotherapy is beneficial in treating advanced NSCLC, and if so to determine the size of any treatment effect.

Systematic literature searches were conducted to identify eligible trials completed between 1/1/65 and 31/12/91 and trial identification was supplemented by information obtained from trial registers and word of mouth. Only properly randomised and unconfounded trials were included and data from both published and unpublished trials were obtained. Updated information was sought for all individuals who had been randomised. Trials were divided into a number of pre-specified chemotherapy categories (i) those using long-term alkylating agents, (ii) those using vinca-alkaloids or etoposide and (iii) those using cisplatin-based regimens.

Updated IPD were obtained from all 11 eligible trials and included information on 1190 patients of whom 1144 had died. Two trials used long-term alkylating agents and one used etoposide as a single agent. The remaining 8 trials used cisplatin-based chemotherapy, of which 7 used a combination of cisplatin and vinca-alkaloids or etoposide. One trial was restricted to metastatic disease, the remainder randomised patients with both locally advanced and advanced disease.

The combined hazard ratio (HR) of 1.26 for the 2 trials using long-term alkylating agents suggests a detriment of this type of chemotherapy. However, the confidence intervals were wide (0.96 - 1.66) and the result did not reach conventional levels of significance (p=0.095).

The cisplatin-based trials showed a conventionally significant benefit of chemotherapy. The HR of 0.73 (p<0.0001), or 27% reduction in the relative risk of death is equivalent to a 10% absolute improvement in survival at 1 year, improving overall survival from 5% to 15%. This is equivalent to increasing median survival from 1 month to 2 ½ months.

Pre-defined subgroups were investigated to assess whether there was any evidence of a differential size of treatment effect between well-defined categories of patients. To minimise heterogeneity, only trials using cisplatin-based regimens were included in this analysis. There was no good evidence that any group of patients defined by age, sex, histology, stage or performance status benefited any more or less from chemotherapy.

These results give the most reliable estimate of the average effect of chemotherapy in the supportive care setting and is probably the best evidence on which to base future research and treatment policy. Although convincing evidence of a moderate benefit of chemotherapy over best supportive care alone has been established, essential drugs were not identified and further research may be required to establish the most effective regimens and evaluate newer agents. Ultimately the decision whether or not to use chemotherapy will be based on many factors including survival, toxicity, economic cost and, importantly in this setting, quality of life. This meta-analysis provides patients and clinicians with a reliable estimate of average treatment effect to use as part of this decision-making process.

References
1. Non-small Cell Lung Cancer Collaborative Group. Chemotherapy in non-small cell lung cancer: a meta-analysis using updated individual patient data from 52 randomised clinical trials. BMJ 311: 899-909, 1995.
2. Souquet PJ, Chauvin F, Boissel JP et al. Polychemotherapy in advanced non-small cell lung cancer: A meta-analysis. Lancet 342: 19-21, 1993.
3. Grilli A, Oxman A, Julian JA. Chemotherapy for advanced non-small cell lung cancer. How much benefit is enough? J Clin Oncol 11: 866-1872, 1993.
4. Stewart, LA, Parmar, MKB. Literature or individual patient based meta-analysis - is there a difference? Lancet 341: 418-422, 1993.
5. Stewart LA and Clarke MJ. Practical methodology of meta-analyses (overviews) using updated individual patient data. Stats in Med 14: 2057-2097, 1995.

________________________________________________________________________________

CARBOPLATIN AND PACLITAXEL IN NON-SMALL CELL LUNG CANCER (NSCLC)
R.B. NATALE
Cedars-Sinai Comprehensive Cancer Center, Los Angeles, California,

and C.P. BELANI
University of Pittsburgh, Pittsburgh, Pennsylvania, USA.

Paclitaxel(P) is widely recognized as one of the most active chemotherapeutic agents for the treatment of NSCLC. At least nine clinical trials involving over 250 patients (pts) have reported single agent activity with doses ranging from 135 to 250 mg/m2 and in several schedules of administration including 96-hour(hr), 24-hr, 3-hr and 1-hr infusions administered every three weeks (Q3wks) and 3-hr infusions administered weekly. The dose-limiting toxicity of this compound changes from primarily myelosuppression with the 24-hr and 96-hr infusion schedules to primarily peripheral neurotoxicity with the 3-hr and 1-hr infusion schedules. Despite this remarkable change in toxicity profile, its activity in NSCLC appears to remain the same with response rates of 20-40%, a median survival of approximately 10 months(mos), and a one year(yr) survival of about 40% in pts with advanced or metastatic disease.

Because of the importance of platinum compounds in the treatment of NSCLC, there has been great interest in the combination of P with cisplatin(Cis) or carboplatin(Car). The 24-hr infusion schedule of P was the first to be tested. Langer et. al. investigated the combination of Car given at a targetted AUC = 7.5 (Calvert method) + P in doses escalating from 135-215mg/m2/24-hr in 53 pts with advanced or metastatic NSCLC1. The results of this intensive approach were impressive with a 63% response rate and median survival of 12.5mos. The Eastern Cooperative Oncology Group (ECOG) conducted a phase III randomized trial comparing P (135mg/m2/24-hr or 250mg/m2/24-hour+GCSF) + Cis (75 mg/m2) Q3wks [PCis/PGCis, respectively] to standard etoposide (100mg/m2 x days 1-3) + cisplatin (75mg/m2 x day 1) Q3wks [ECis] in 108 stage IIIB and 466 stage IV NSCLC pts. There was a significantly higher objective response rate in the PCis/PGCis arms compared to the ECis arm (29% vs 12%, respectively; p < .001). Although the median survival (9.8 vs 7.7 months) and 1yr survival rate (38.5% vs 31.6%) was higher with PCis/PGCis compared to ECis, the survival differences were of borderline statistical significant (p=0.053)2.

The investigation of shorter infusions schedules of P combined with a platinum analogue followed the successful development of effective premedication regimens. At least 12 phase II trials in over 400 pts have explored P, 3-hr or in some cases 1-hr infusion, combined with a platinum-analogue. Since neurotoxicity rather than myelosuppression is the primary dose-limiting toxicity of short infusion P, most investigators have used Car rather than Cis in order to avoid overlapping neurotoxicty. In general, these trials have reported response rates in the range of 30-60%, median survivals of 10-11mos and 1yr survivals of 35-45% with P 175-225mg/m2 and Car AUCs of 5-7 or 300-350mg/m2. Dose-limiting toxicity consisted primarily of cumulative peripheral neurotoxicity, acute and self-limiting arthralgias and/or myalgias, and mild to moderate fatigue and malaise. Grade 4 neutropenia generally occurred in less than 20% of patients and febrile neutropenia was uncommon (<5%). Interestingly, grade 3-4 Car-associated thrombo-cytopenia rarely occurred.

Three multi-institutional phase III trials have been developed in the U.S. to compare PCar (using the Natale regimen of P 225 mg/m2/3-hr + Car [AUC=6] Q3wks3) to other standard or recently developed treatment regimens in pts with NSCLC; two of these studies are ongoing (SWOG and ECOG) and one has been completed as follows. From 5/95 to 7/96, 369 pts with advanced (not candidates for definitive RT or combined modality therapy) or metastatic NSCLC were entered into a randomized trial comparing PCar to standard ECis. Treatment was administered Q3wks to a maximum of 10 cycles in the absence of unacceptably severe toxicity, pt refusal or disease progression. Pt characteristics and treatment-related toxicities according to arm were as follows:

References
1. Langer et. al., J Clin Oncol 15:, 1997.
2. Bonomi et. al., Proc Amer Soc Clin Oncol 16:454a, 1997.
3. Natale RB, Semin in Oncol 23:2-6, 1996.