LONG TERM FOLLOW-UP OF NEOADJUVANT-ADJUVANT COMBINATION TREATMENT OF IIIA STAGE NON-SMALL-CELL-LUNG CANCER : RESULTS OF NEOADJUVANT CARBOPLATIN / VINORELBINE AND CARBOPLATIN / PACLITAXEL REGIMENS COMBINED WITH SELECTIVE ADJUVANT CHEMOTHERAPY ACCORDING TO IN-VITRO CHEMORESISTANCE TEST

Aim: A prospective study investigated survival of patients with stage IIIA non-small-cell-lung cancer (NSCLC) treated with a combination of neoadjuvant and adjuvant chemotherapy. Methods: Consecutive chemo-naive patients with potentially operable stage IIIA NSCLC received carboplatin-based neoadjuvant treatment. Tumor cells harvested during surgery underwent methylthiazolyl tetrazolium blue (MTT) cytotoxic assay. After surgery, adjuvant chemotherapy was selected, where possible, according to MTT results. Results: A total of 65 patients were evaluated (31 received carboplatin/vinorelbine, 34 carboplatin/paclitaxel). The overall response rate was 67.7 % (95% confi dence interval [CI]: 56.3–79.1 %) with downstaging in 52.3 % (95% CI: 40.2–64.5 %) and no signifi cant diff erences between regimens. Median follow-up was 86 months: median overall survival (OS) was 32.1 months (95% CI: 7.4–46.5), median time to progression was 25.1 months (95% CI: 15.1–34.9 months) and fi ve-year overall survival was 35.7 % (95% CI: 23.7–47.7 %). Forty-seven patients (72.3 %) underwent surgery and 43 patients received adjuvant chemotherapy. Five-year survival after tumor resection was 49.5 % (95% CI: 34.2–64.8 %), median OS was 59.0 months (95% CI: 34.2–83.1) and median disease free survival after surgery was 57.3 months (95% CI: 29.5–84.4). With MTT-directed therapy, median OS was 85.1 months (95% CI: 15.4–148.6) and the 5-year survival rate was 57.0 % (95% CI: 34.5–79.5 %); the trend for longer survival failed to reach statistical signifi cance. Conclusions: A combination of carboplatin-based neoadjuvant chemotherapy, surgical resection and adjuvant chemotherapy achieved satisfactory survival rates in stage IIIA NSCLC, especially in patients with complete resection of tumor and those given MTT-directed adjuvant treatment. Our results suggest MTT testing may help optimise adjuvant chemotherapy.


INTRODUCTION
Lung cancer has an unfavorable prognosis and remains the leading cause of cancer death worldwide.Surgery is the preferred curative method in non-small-cell lung cancer (NSCLC), which represents 75 % of all lung cancer cases, but only early-detected disease can be treated successfully.Optimal treatment of patients with locally advanced NSCLC in stage IIIA is still a matter of discussion, as 5-year survival in these patients varies from 5 % to 15 % and, even after complete resection, can only reach slightly more than 20 % (ref. 1,2 .Micrometastases at distant sites are present in almost 80 % of patients with stage IIIA NSCLC (ref. 3 ).Detection by molecular techniques shows the early spread of tumor cells to lymph nodes, whilst the presence of malignant cells in peripheral blood is probably the major cause of early death after complete resection 4,5 .For these reasons, combined treatment modalities are preferred over surgery alone 6 .There are several recommended approaches advocating neoadjuvant or adjuvant chemotherapies combined with surgery [7][8][9] .Multimodal therapy may also involve neoadjuvant chemoradiotherapy with surgery or radical chemoradiotherapy without surgery [10][11][12][13] .The results of the above mentioned therapeutic modalities are infl uenced by patients' performance status and comorbidities, lymph node involvement, staging accuracy, achievement of downstaging and the type of resection [14][15][16][17][18] .
Chemotherapy after surgery has been studied in several neoadjuvant Phase III studies investigating second generation cytotoxic drugs, as well 6,7,19 .This was considered perioperative chemotherapy, with the same drugs administered before and after surgery, and the eff ects of the postoperative part of the chemotherapy were not systematically evaluated.
In the present study, vinorelbine/carboplatin and paclitaxel/carboplatin doublets were used in the neoadjuvant protocol for IIIA NSCLC, followed by surgery and adjuvant chemotherapy selected according to the results of an in vitro test of chemoresistance.The neoadjuvant regimens were chosen because they are eff ective in fi rst line NSCLC management and have potentially less toxicity and better tolerability than cisplatin combinations 20,21 .The primary objective of this prospective trial was to assess survival parameters.The secondary end points were the eff ectiveness and tolerability of chemotherapy and the potential clinical value of in vitro chemoresistance analysis for guiding adjuvant chemotherapy.

Study Population
Chemonaive patients with potentially operable stage IIIA NSCLC were included in a prospective study.The patients were physically examined by a pneumologist, oncologist and thoracic surgeon in order to confi rm the stage of disease and confi rm the indication for medical and surgical procedures.The tumor was verifi ed by cytology or histology.The staging was performed by chest radiography, computed tomography (CT) scans of the chest and upper abdomen, bronchoscopy with fi ne needle transbronchial needle aspiration (TBNA) and bone scintigraphy for every patient.Mediastinoscopy and brain CT were optional.The eligibility criteria included age 18 to 75 years, WHO performance status 0 and 1, pulmonary function tests compatible with potential resection, acceptable blood count, serum electrolytes, urea, creatinine, liver enzymes, bilirubin and renal function test.Exclusion criteria were prior malignancy, coexisting serious nonstabilized disease, pregnancy, peripheral neuropathy, a history of psychiatric disorder, drug or alcohol abuse, or active uncontrolled infection.The study was approved by the Institutional Review Boards of the participating institutions and informed consent was obtained from all patients.

Trial Design and Treatment Plan
Consecutive eligible patients were included from January 1997 to December 2004.Patients received three cycles of inductive chemotherapy (each planned for 21 days): randomly receiving vinorelbine 30 mg/m 2 administered as a 10 minute infusion on days 1 and 8 or paclitaxel 200 mg/m 2 given as a 3 hour infusion on day 1, both with carboplatin in a dose of area under the curve (AUC) 6 administered in a 30 minute infusion.Toxicity was assessed using the standard World Health Organization (WHO) criteria 22 .
Surgery was indicated after re-staging consisting of chest radiography, CT, bronchoscopy with TBNA, bone scintigraphy and laboratory tests.Tissue from the tumor was harvested and the methylthiazolyl tetrazolium blue (MTT) cytotoxic assay 23 was used to determine the in vitro drug chemoresistance of the tumor cells to cisplatin, carboplatin, paclitaxel, docetaxel, vinorelbine, gemcitabine and etoposide.
Three cycles of adjuvant chemotherapy were planned (21 days each).When an indicative result was obtained from the MTT assay, optimal cytotoxic drugs were selected in accordance with the results of the MTT assay.The test could not be used in patients with complete remissions, in cases of only a small number of vital tumor cells isolated, due to microbial contamination or when the cultivated tumor cells were resistant to all drugs tested.Under these conditions, two diff erent drugs from the previous neoadjuvant treatment were chosen.Patients with progression after neoadjuvant chemotherapy were treated with second line chemotherapy, or chemoradiotherapy.

In vitro Chemoresistance Analysis
Isolated tumor cells were exposed to six concentrations of each drug in duplicate for three days 23 .The yellow tetrazolium salt MTT was reduced to dark blue formazan by viable cells only.Formazan crystals were dissolved in 100 μl 10% sodium dodecylsulphate/water (pH=5.0)overnight at 37 o C. The optical density (OD) was measured at 540 nm with a microplate reader.The tumor cell survival (TCS) was calculated using the following formula: TCS = (OD drug exposed well / mean OD control wells ) x 100 %.The TCS50 value, the drug concentration lethal to 50 % of the tumor cells, was calculated from the resultant dose response curves.Tumor samples with a drug TCS50 value below the median TCSC50 of lung cancers were evaluated as drug sensitive.Chemicals were obtained from Sigma Chemical Co., USA.

Clinical Evaluation
During chemotherapy, physical examination, blood counts and biochemistry were performed every 14 days.After adjuvant chemotherapy, clinical check ups and CT were carried out every 3 months for 2 years.Follow-up after the end of chemotherapy (physical examinations, laboratory tests and CT) was conducted every 6 months until death or any point when clinical symptoms had substantially changed.The end date for the survival analysis was 30 June 2007.Response to chemotherapy was evaluated by WHO criteria 22 , with a complete pathologic response defi ned as > 95 % necrosis and fi brosis.Assessment of mediastinal lymph nodes was performed using the fi fths revision of TNM classifi cation 1 .

Statistical Analysis
Eff ectiveness and tolerability of neoadjuvant and adjuvant chemotherapy, resectability and survival parameters of patients were evaluated.The data were statistically evaluated for the total series of patients and separately for each neoadjuvant regimen, as well as for patients treated according to the results of MTT assay in comparison to samples where results of in vitro chemoresistance test could not be used.Time to progression (TTP) was cal-Long term follow-up of neoadjuvant-adjuvant combination treatment of IIIA stage non-small-cell-lung cancer: Results of neoadjuvant carboplatin/vinorelbine and carboplatin/paclitaxel regimens… culated in all subjects from diagnosis to progression or recurrence, disease free survival (DFS) was calculated in those with complete resection from the date of resection to the date of recurrence or death from NSCLC.The median overall survival (OS), TPP and DFS of the total series and the subgroups were calculated using Kaplan-Meier estimates and the log-rank test; diff erences between subgroups were evaluated by the χ 2 test for categorical variables or by Fisher's exact test 24,25 .Statistical software SPSS v.14 (SPSS Inc., Chicago, USA) was used, level of signifi cance was α = 0.05.

Patient Characteristics
Sixty eight consecutive patients with stage IIIA NSCLC were included and 65 were evaluated (3 patients declined from the staging procedures); 54 (83.1 %) were men and 11 (16.9 %) were women.The mean age was 59.1 years (range 41-75); in men, 60.3 years; in women, 56.4 years.The clinical characteristics of the total series and subgroups treated with vinorelbine/carboplatin or paclitaxel/carboplatin are shown in Table 1.There were no signifi cant diff erences between subgroups in TNM classifi cation, histological types, performance status or sex ratio.

Neoadjuvant Chemotherapy
In total, 65 patients received 194 cycles (92 cycles in the vinorelbine/carboplatin subgroup and 102 cycles in the paclitaxel/carboplatin subgroup).Chemotherapy was postponed in 18 cycles (9.3 %) due to neutropenia or thrombocytopenia.The dose was reduced in three cycles (1.5 %) due to neutropenic fever.Overall grade 3 toxicity appeared in 27 cycles (13.9 %); toxicity was similar in both subgroups with a predominance of neutropenia -grade 4 in six cycles (3.1 %).Thrombocytopenia grade 3 appeared in two patients (2.1 %) in the vinorelbine subgroup; whilst myalgia and arthralgia grade 3 were observed in four patients (4.0 %) in the paclitaxel subgroup.
After chemotherapy, downstaging according to nonsurgical methods was achieved in 34 patients (52.3 %), and tumor size reduction in 10 patients (15.4 %) (Table 2).Surgery was performed in 47 patients (72.3 %), with complete resection achieved in 44 patients (67.7 %) (Table 3).Early death was noted in two patients.No statistically signifi cant diff erences were found between the subgroups treated with vinorelbine/carboplatin or paclitaxel/carboplatin.

Survival
As of the end of June 2007, the median follow-up was 86 months and 44 patients (67.7 %) had died.The median of OS was 32.1 months, as shown in survival curve (Fig. 1).Local progression occurred in 4 patients and distant metastases in 12 patients (brain in 7 patients, lung in 2 patients, liver, supra-adrenal gland and peripheral lymph nodes in one patient each); fi ve patients had both types of progression.The diff erences between the vinorelbine/carboplatin and paclitaxel/carboplatin subgroups were not statistically signifi cant for any of the survival parameters (Table 4).

Adjuvant chemotherapy and chemoresistance analysis
Adjuvant chemotherapy was administered to 43 of the 47 patients receiving surgery with a total of 129 cycles of chemotherapy.It was not applied in two patients of early death and in two patients with signifi cant deterioration of PS.Combinations with cisplatin were used in 21 patients, whilst doublets with carboplatin were chosen in 16 patients.Platinum was combined with vinorelbine   Harvesting of tumor cells for MTT assay was possible in 38 patients.The results of the MTT assay for chemoresistance are summarized in Table 5.The MTT assay was able to drive the choice of optimally eff ective drugs in 19 patients (44.2 %).However, MTT assays were not possible in 24 patients (65.8 %) due to lack of vital tumor cells or because the cultivated tumor cells were resistant to all drugs tested.Baseline characteristics for patient subgroups, whether treated or not treated with eff ective drugs according to the results of MTT assay, did not diff er signifi cantly for sex, histology of tumor, type of surgery or administered chemotherapeutic regimens.Patients treated with optimally eff ective drugs survived longer, but the diff erence did not reach statistical signifi cance (Fig. 2 and Table 6).Survival was not infl uenced by individual cytotoxic drugs used in adjuvant chemotherapy without respect to MTT assay (all p values > 0.05).

DISCUSSION
The present study analyzes experience with neoadjuvant and adjuvant chemotherapies in IIIA stage NSCLC.Complete resection was realized in 67.7 % patients and favorable survival was achieved, with a median of OS of 32.1 months.The three-year survival was 47.0 % and the 5-year survival 35.7 %.The eff ectiveness of the carboplatin-based neoadjuvant chemotherapy used in our centre was comparable with cisplatin regimens published in other studies.In phase III studies of cisplatin and older cytostatic drugs, radical resection rates were from 64 % to 74 % (ref. 6,7,19,26 ).In two studies with comparable  6,7 .Cisplatin with the third generation cytotoxic agents in phase II studies show overall response rates (ORRs) from 54 % to 74 %, but only a small number of these studies have calculated the 3-year survival 18,27,28 , which was 33 % with a docetaxel/ cisplatin neoadjuvant protocol 13 and 20 % in a gemcitabine/cisplatin study 27 .Vinorelbine in neoadjuvant treatment has been used in combination only with cisplatin 29 , or with cisplatin and gemcitabine 30 and ORRs were 53.6 % and 58.5 %, respectively. Acombination of vinorelbine/carboplatin neoadjuvant regimen with radiotherapy achieved an ORR of 70 % and a complete resection rate of 87 % (ref.31 ).Non-platinum neoadjuvant regimen vinorelbine/gemcitabine was used in a Phase II study in diff erent stages of NSCLC (ref.32 ).The regimen had low toxicity and 65 % patients with stage IIIA underwent a complete resection.
Paclitaxel/carboplatin regimen was used in stage IIIA NSCLC in an EORTC Phase II study, ORR was 63 % and the calculated median of OS was 20.5 months 33 .In several other studies, no positive eff ect of neoadjuvant treatment was found [10][11][12]34 but, despite all diff ering results a recent metaanalysis of neoadjuvant therapy in patients with resectable disease showed a hazard ratio HR 0.66 (95 % CI 0.48-0.93) in favor of adding neoadjuvant chemotherapy to the standard procedure 35 .
The eff ectiveness of adjuvant chemotherapy was confi rmed in several large studies, with vinorelbine and cisplatin used most often in the evaluated patients 8,9,36 .In the ANITA trial, the subgroup of patients with stage IIIA NSCLC achieved a median OS of 38.6 months and a 5-year survival of 42 % after complete resection and adjuvant chemotherapy 8 .These parameters can be compared with our series of patients, who after resection and combination of adjuvant and neoadjuvant chemotherapy, achieved a median of OS 61.2 months and a 5-year survival rate 49.5 %.Long term follow-up of neoadjuvant-adjuvant combination treatment of IIIA stage non-small-cell-lung cancer: Results of neoadjuvant carboplatin/vinorelbine and carboplatin/paclitaxel regimens… The present study also confi rmed a very good tolerability profi le, which enabled the administration of planned chemotherapy for the great majority of patients and there was no need to interrupt neoadjuvant therapy, delay or modify surgery due to toxicity, unlike a previous study which reported induction chemotherapy increases perioperative complications 37 .
Several studies on the neoadjuvant or adjuvant approach in IIIA NSCLC have produced comparable results to our own.It appears that the survival plateau of classic chemotherapeutic regimens without predictive markers has been reached and our results imply that multi-and cross-resistance is the major problem.Although the MTT assay has a number of technical problems with harvesting and processing of cells and refl ects the behavior of tumor cells under laboratory conditions, it may provide helpful predictive and prognostic information 38,39 .In our study, the relative improvement in the 5-year survival was 13.5 % for those treated with MTT-driven adjuvant therapy.Despite the fact that the series was small and the diff erences were not signifi cant, this trend supports the concept that customized adjuvant chemotherapy may improve the survival of some patients with stage IIIA NSCLC.1][42] ).The selection of biological agents targeting specifi c intracellular pathways related to the distinctive properties of lung cancer cells are another promising option for optimizing adjuvant chemotherapy 43 .
In conclusion, a combination of neoadjuvant and adjuvant chemotherapy in stage IIIA NSCLC is a feasible option for eff ective treatment giving a reasonable prognosis for patients with a good performance status.Carboplatin in combination with vinorelbine or paclitaxel as the induction treatment had an acceptable effi cacy with good tolerability and low toxicity.Patients can benefi t from a rational selection of drugs in the adjuvant setting, which remains one of the key problems for the future multimodal therapy of resectable IIIA NSCLC, and the MTT assay looks promising in this regard.

Table 3 .
Surgical Procedures Used After Neoadjuvant Chemotherapy Long term follow-up of neoadjuvant-adjuvant combination treatment of IIIA stage non-small-cell-lung cancer: Results of neoadjuvant carboplatin/vinorelbine and carboplatin/paclitaxel regimens… Vin: vinorelbine; CBDCA: carboplatin; Pac: paclitaxel.

Table 5 .
Chemoresistance of Tumor Cells after

Table 6 .
Subgroups Treated or Not Treated According to MTT Assay Results