A phase 1 Bayesian dose selection study of bortezomib and sunitinib in patients with refractory solid tumor malignancies.

BACKGROUND: This phase 1 trial utilising a Bayesian continual reassessment method evaluated bortezomib and sunitinib to determine the maximum tolerated dose (MTD), dose-limiting toxicities (DLT), and recommended doses of the combination. METHODS: Patients with advanced solid organ malignancies were enrolled and received bortezomib weekly with sunitinib daily for 4 weeks, every 6 weeks. Initial doses were sunitinib 25 mg and bortezomib 1 mg m(-2). Cohort size and dose level estimation was performed utilising the Escalation with Overdose Control (EWOC) adaptive method. Seven dose levels were evaluated; initially, sunitinib was increased to a goal dose of 50 mg with fixed bortezomib, then bortezomib was increased. Efficacy assessment occurred after each cycle using RECIST criteria. RESULTS: Thirty patients were evaluable. During sunitinib escalation, DLTs of grade 4 thrombocytopenia (14%) and neutropenia (6%) at sunitinib 50 mg and bortezomib 1.3 mg m(-2) were seen. Subsequent experience showed tolerability and activity for sunitinib 37.5 mg and bortezomib 1.9 mg m(-2). Common grade 3/4 toxicities were neutropenia, thrombocytopenia, hypertension, and diarrhoea. The recommended doses for further study are bortezomib 1.9 mg m(-2) and sunitinib 37.5 mg. Four partial responses were seen. Stable disease >6 months was noted in an additional six patients. CONCLUSION: Bortezomib and sunitinib are well tolerated and have anticancer activity, particularly in thyroid cancer. A phase 2 study of this combination in thyroid cancer patients is planned.

Phase I and pharmacokinetic study of ormaplatin (tetraplatin, NSC 363812) administered on a day 1 and day 8 schedule.

Ormaplatin (tetraplatin, NSC 363812) is a platinum(IV) analogue that is active against cisplatin-resistant cell lines in preclinical models. A schedule previously shown to be active and well tolerated for cisplatin was evaluated in 26 patients. Ormaplatin was administered over a dose range of 4.4-60.8 mg/m2 i.v. given over 30 min on a day 1 and day 8 schedule every 28 days. Twenty-three patients had received prior chemotherapy, and the median performance status was 1. Nausea/vomiting (> or = grade 2) occurred in 40% of patients but was well controlled with standard antiemetic therapy. One patient had grade 2 renal toxicity and 1 patient had grade 3 hepatotoxicity (grade 2 pretreatment). No toxicity limited the dose given during the first course. With repeated drug administration delayed severe neurotoxicity developed in 4 patients, manifested as a sensory polyneuropathy in 3 patients and a possible autonomic neuropathy in one. Prospective nerve conduction studies did not detect subclinical neuropathy prior to the onset of symptoms. Patients who received cumulative doses above 200 mg/m2 were at increased risk for developing neurotoxicity. Plasma elimination of ultrafilterable platinum (measured by atomic absorption spectrometry) was biphasic with a harmonic mean terminal half-life of 15.8 h. The mean total body clearance and renal clearance of ultrafilterable platinum were 173 and 29.8 ml/min/m2, respectively. Thus, renal clearance accounted for 16% of total clearance suggesting that extensive protein/tissue binding was responsible for the majority of platinum clearance. Approximately 60% of the platinum is protein bound (one-half irreversibly) at the end of the infusion. Pharmacokinetic parameters were not dose dependent. No pharmacokinetic parameters were more predictive of neurotoxicity than the cumulative ormaplatin dose. A phase II dose cannot be recommended on this schedule because severe and unpredictable neurotoxicity precludes the administration of more than three cycles at the three highest doses levels tested.

Cardiopulmonary resuscitation and the patient with cancer.

The records of 114 cancer patients suffering cardiopulmonary arrests (CPA) during a 3-year period at Memorial Sloan-Kettering Cancer Center (MSKCC) were retrospectively reviewed to identify variables predicting final outcome in these patients. Although 65.7% of the patients were successfully resuscitated, only 12 (10.5%) were discharged alive from the hospital. Median survival after discharge was 150 days. By univariate and multivariate analysis, the only variable predicting the likelihood of a patient's being discharged alive after a CPA was the performance status of the patient at the time of admission to the hospital. Thus, a patient spending more than 50% of the time in bed at the time of admission had only a 2.3% chance of being discharged alive after CPA. A thorough discussion of these findings between physicians and patients and their families is strongly recommended at the time of admission to spare cancer patients unnecessary invasive resuscitative procedures.

Phase I trial of multiple cycles of high-dose carboplatin, paclitaxel, and topotecan with peripheral-blood stem-cell support as front-line therapy.

PURPOSE: To determine the safety and feasibility of delivering multiple cycles of front-line high-dose carboplatin, paclitaxel, and topotecan with peripheral-blood stem-cell (PBSC) support. PATIENTS AND METHODS: Patients were required to have a malignant solid tumor for which they had received no prior chemotherapy. Mobilization of PBSC was achieved with either filgrastim alone or in combination with cyclophosphamide and paclitaxel. Patients then received three or four cycles of high-dose carboplatin (area under the concentration-time curve [AUC] 16), paclitaxel (250 mg/m(2)), and topotecan (10-15 mg/m(2)), with the latter two agents administered as 24-hour infusions and supported with PBSC and filgrastim. Cycles were repeated every 28 days. RESULTS: Twenty patients were enrolled onto the trial and were assessable for toxicity and clinical outcome. Dose-limiting toxicities were stomatitis and prolonged hematopoietic recovery. The maximum-tolerated dose of topotecan was 12.5 mg/m(2) when given with high-dose carboplatin and paclitaxel for three cycles. Four cycles were able to be given with a dose of topotecan of 10 mg/m(2). The pharmacokinetics of each compound were not affected by the other agents. Eleven (85%) of 13 patients with assessable disease responded. CONCLUSION: Multiple cycles of high-dose carboplatin, paclitaxel, and topotecan can be safely administered with filgrastim and PBSC support. The recommended doses for phase II study are carboplatin AUC 16, paclitaxel 250 mg/m(2), and topotecan 10 mg/m(2). Trials are currently being conducted with this regimen as front-line treatment in patients with advanced ovarian cancer and extensive small-cell carcinoma. This approach remains experimental and should be used only in the context of a clinical trial.

Evaluation of carboplatin pharmacokinetics in the absence and presence of paclitaxel.

In a clinical trial of paclitaxel (Taxol) and carboplatin in combination, the severity of thrombocytopenia was less than would be expected with an equivalent dose of carboplatin alone. To determine whether a pharmacokinetic interaction was responsible for this observation, the effect of pretreatment with Taxol on the pharmacokinetics of carboplatin was examined in 11 patients. Each patient was randomized to one of two treatment groups that determined the order of drug treatments. The treatments were carboplatin as a 30-min infusion alone or immediately following 175 mg/m2 Taxol administered as a 3-h i.v. infusion. The treatments were separated by 1 week. The carboplatin dose was chosen to produce a target area under the concentration-time curve (AUC) of 3.75 mg-min/ml according to a previously published formula (A. H. Calvert et al., J. Clin Oncol., 7: 1748-1756, 1989). The mean administered dose of carboplatin was 338 mg. Serial blood samples were collected over 24 h and analyzed for total and free platinum, and, in some patients, Taxol. The pharmacokinetics of carboplatin (i.e., total clearance and volume of distribution at steady state), was not significantly affected by pretreatment with Taxol. Total clearances of carboplatin were 67.2 +/- 28.8 ml/min and 64.6 +/- 27.9 ml/min in the absence and presence of Taxol, respectively (P = 0.56). The AUC of free carboplatin (3.45 mg-min/ml) obtained in the absence of Taxol was not significantly different from that measured in the presence of Taxol (3.27 mg-min/ml). The AUC of carboplatin in both the absence and presence of Taxol agreed with the projected target AUC of 3.75 mg-min/ml. In conclusion, the application of an individualized dosing strategy is valid for the calculation of the carboplatin dose in this combination. The pharmacokinetics of carboplatin is not altered by pretreatment with Taxol at a standard dose, and a pharmacokinetic interaction is not responsible for the altered toxicity of the combination.

Pilot study of a dual gene recombinant avipox vaccine containing both carcinoembryonic antigen (CEA) and B7.1 transgenes in patients with recurrent CEA-expressing adenocarcinomas.

Coordinated presentation of antigen and costimulatory molecules has been shown to result in the induction of an antigen-specific T-cell response rather than the development of anergy. This study evaluated the vaccine ALVAC-CEA B7.1, a canary pox virus that has been engineered to encode the gene for the tumor-associated antigen carcinoembryonic antigen (CEA) and B7.1, a T-cell costimulatory molecule. Patients with CEA-expressing tumors were immunized with 2.5 x 10(7) (n = 3), 1.0 x 10(8) (n = 6), and 4.5 x 10(8) (n = 30) plaque-forming units intradermally every other week for 8 weeks. Patients with stable or responding disease received monthly boost injections. Biopsies of vaccine sites were obtained 48 h after vaccination to evaluate leukocytic infiltration and CEA expression. Induction of CEA-specific T-cell precursors was assessed by an ELISPOT assay looking for the production of IFN-gamma. Therapy was well tolerated, without significant toxicity attributable to vaccine. All patients had evidence of leukocytic infiltration and CEA expression in vaccine biopsy sites. Six patients with elevated serum CEA values at baseline had declines in their levels lasting 4-12 weeks. These patients all had stable disease after four vaccinations. After four vaccinations, patients who were HLA-A-2-positive demonstrated increases in their CEA-specific T-cell precursor frequencies to a CEA-A2-binding peptide from baseline. The number of prior chemotherapy regimens was inversely correlated with the ability to generate a T-cell response. ALVAC-CEA B7.1 is safe in patients with advanced, recurrent adenocarcinomas that express CEA, and it is associated with the induction of a CEA-specific T-cell response.

The influence of granulocyte macrophage colony-stimulating factor and prior chemotherapy on the immunological response to a vaccine (ALVAC-CEA B7.1) in patients with metastatic carcinoma.

Granulocyte macrophage colony-stimulating factor (GM-CSF) has been shown to be an effective vaccine adjuvant because it enhances antigen processing and presentation by dendritic cells. ALVAC-CEA B7.1 is a canarypox virus encoding the gene for the tumor-associated antigen carcinoembryonic antigen (CEA) and for a T-cell costimulatory molecule, B7.1. After an initial dose escalation phase, this study evaluated vaccination with 4.5 x 10(8) plaque-forming units ALVAC-CEA B7.1 alone (n = 30) or with GM-CSF (n = 30) in patients with advanced CEA-expressing tumors to determine whether the addition of the adjuvant GM-CSF enhances induction of CEA-specific T-cells. Patients were vaccinated with vaccine intradermally every other week for 8 weeks. GM-CSF was given s.c. for 5 days beginning 2 days before vaccination. Patients with stable or responding disease after four immunizations received monthly boost injections alone or with GM-CSF. Biopsies of vaccine sites were obtained 48 h after vaccination to evaluate leukocytic infiltration and CEA expression. Induction of peripheral blood CEA-specific T-cell precursors was assessed in HLA-A2 positive patients by an ELISPOT assay looking for the production of IFN-gamma. Therapy was well tolerated. All of the patients had evidence of leukocytic infiltration and CEA expression in vaccine biopsy sites. In the patients receiving GM-CSF, leukocytic infiltrates were greater in cell number but were less likely to have a predominant lymphocytic infiltrate compared with patients receiving vaccine in the absence of the cytokine adjuvant. After four vaccinations, CEA-specific T-cell precursors were statistically increased in HLA-A2 positive patients who received vaccine alone. However, the GM-CSF plus vaccine cohort of HLA-A2 positive did not demonstrate a statistically significant increase in their CEA-specific T-cell precursor frequencies compared with baseline results. The number of prior chemotherapy regimens was negatively correlated with the generation of a T-cell response, whereas there was a positive correlation between the number of months from the last chemotherapy regimen and the T-cell response. ALVAC-CEA B7.1 is safe in patients with advanced, recurrent adenocarcinomas that express CEA, is associated with the induction of a CEA-specific T-cell response in patients treated with vaccine alone but not with vaccine and GM-CSF, and can lead to disease stabilization for up to 13 months.

Stromal fibroblast-derived miR-409 promotes epithelial-to-mesenchymal transition and prostate tumorigenesis.

Tumor-stromal interaction is a dynamic process that promotes tumor growth and metastasis via cell-cell interaction and extracellular vesicles. Recent studies demonstrate that stromal fibroblast-derived molecular signatures can be used to predict disease progression and drug resistance. To identify the epigenetic role of stromal noncoding RNAs in tumor-stromal interactions in the tumor microenvironment, we performed microRNA profiling of patient cancer-associated prostate stromal fibroblasts isolated by laser capture dissection microscopy and in bone-associated stromal models. We found specific upregulation of miR-409-3p and miR-409-5p located within the embryonically and developmentally regulated DLK1-DIO3 (delta-like 1 homolog-deiodinase, iodothyronine 3) cluster on human chromosome 14. The findings in cell lines were further validated in human prostate cancer tissues. Strikingly, ectopic expression of miR-409 in normal prostate fibroblasts conferred a cancer-associated stroma-like phenotype and led to the release of miR-409 via extracellular vesicles to promote tumor induction and epithelial-to-mesenchymal transition in vitro and in vivo. miR-409 promoted tumorigenesis through repression of tumor suppressor genes such as Ras suppressor 1 and stromal antigen 2. Thus, stromal fibroblasts derived miR-409-induced tumorigenesis, epithelial-to-mesenchymal transition and stemness of the epithelial cancer cells in vivo. Therefore, miR-409 appears to be an attractive therapeutic target to block the vicious cycle of tumor-stromal interactions that plagues prostate cancer patients.

Exploring family relationships in cancer risk counseling using the genogram.

OBJECTIVES: The genogram is a tool that has facilitated counseling in family therapy and social work for many years. It is hypothesized that genograms may also be useful in genetic counseling, because they help the counselor to acquire more objective and consistent information from the client, as well as to incorporate family dynamics and psychosocial issues into the counseling approach. MATERIALS AND METHODS: A pilot study of genograms used as an adjunct to genetic counseling was performed at Fox Chase Cancer Center's Family Risk Assessment Program. A questionnaire was developed to elicit genograms from 38 women at risk for familial breast and/or ovarian cancer. After standard pedigree expansion, a series of questions was asked about the consultand's relationship with other family members, communication patterns within the family, attitudes toward genetic testing, family reactions to cancer, roles individuals play in the family, and significant historical or anniversary events. Relationships were defined by the consultand as close, very close, conflictual, fused and conflictual, distant, or estranged. RESULTS: The majority of relationship types reported by 38 individuals was "very close" or "close." Eighty-one % reported having close/very close relationships with their spouses, 83% reported close/very close relationships with their mothers, and 70% reported close/very close relationships with their fathers. The degree of familial cohesion as depicted by the genogram correlates positively with scores obtained on the standardized Social Adjustment Scale Self-Report (P = 0.01). CONCLUSIONS: Given the family-wide implications of genetic testing, the genogram may offer important guidance in family-targeted interventions.

Mutagens, N-nitroso compounds and their precursors in gastric juice from patients with and without precancerous lesions of the stomach.

This study examined whether elevated risk of gastric cancer is associated with high levels of total N-nitroso compounds (NOC), their precursors and nitrosation-dependent genotoxins in gastric juice (GJ). An improved method for quantifying total NOC was used and genotoxicity was assayed in E. coli. Results from patients (n = 210) with or without precancerous lesions of the stomach and living in three areas with up to 8-fold variations in gastric cancer risk (U.K., France, Colombia) were compared. The level of nitrite (range < 1-472 mumol/l) was found to increase with the pH of GJ from the three countries and was dependent on country of collection. The levels of NOC (range: < or = 0.01-8.0 mumol/l) in GJ were not affected by stomach histology and country of collection. NOC levels increased linearly with nitrite concentrations, but the slope of the regression line was greater for acidic GJ (pH < or = 4). These data together suggest that chemical nitrosation contributes at least as much as other nitrosation pathways to the intragastric formation of NOC. Acid-catalysed nitrosation of GJ in vitro increased the NOC concentration (range: 7-1332 mumol/l) up to several 1000-fold but this increase was not predictive of gastric cancer risk either by country or by stomach histology. After acid-catalysed nitrosation, direct genotoxicity (SOS-inducing potency) was significantly higher in GJ with original pH > 4 and highest in samples from Colombia. The results (a) provide no support that intragastric total NOC levels are elevated in subjects with precancerous stomach lesions or living in a high risk area for stomach cancer; (b) confirm that a high nitrite level and elevated pH in GJ are strongly associated, the level of nitrite being associated with precancerous stomach conditions only in Colombia; (c) reveal the presence of precursor compounds in GJ, that after nitrosation yield direct mutagens that probably contain NOC and other substances. As their concentrations were significantly higher in achlorhydric subjects and highest in Colombian patients, these data together provide support for a role of intragastrically formed nitrite-derived direct mutagens in gastric cancer aetiology.

Risk prediction with linked markers: theory.

A Bayesian method to predict disease recurrence risks based on distance information between loci and allowing for uncertain gene location is presented. The methodology allows derivation of the posterior distribution of recurrence risks, from which point estimates and associated precision measures can be calculated. The resulting risk distribution accounts for all available information concerning location and relative distance, and can be updated as additional data become available.

Evaluating the uncertainty of risk prediction in genetic counseling: a Bayesian approach.

A general method to obtain the exact point estimate and the measure of uncertainty of recurrence risks for genetic counseling is present. This method, which follows Bayesian estimation theory, can be applied without restrictions of sample size, class of risk function, or parameter dimension. As an illustration, the recurrence risks for isolated cases in two autosomal dominant disorders with incomplete penetrance (mandibulofacial and fontonasal dyostoses) are estimated.

Risk prediction with linked markers: pedigree analysis.

Rogatko [1995: Am J Med Genet 59:14-23] has proposed a method for risk prediction with linked markers. The actual implementation of this method required that the analytical forms of likelihood and risk functions be specified. It is impractical to obtain the explicit analytical form of these functions in large phase unknown pedigrees. When large phase unknown pedigrees are encountered, the compound risk can be approximated by a transformation of the discrete distribution obtained by computing the likelihood and risk functions over a grid of points. We propose a method to compute genetic risks when the functional form of the pedigree likelihood is unknown. The method was evaluated using a simple pedigree by comparing the results when functional forms were and were not known. This method was also applied to estimate genetic risks for a single pedigree with nonsyndromal X-linked mental retardation using 3 genetic markers linked to the putative disease gene. Linkage data from an extended pedigree were combined with genome mapping data, and recurrence risk distributions were calculated for members of the pedigree. The results suggest that the proposed method provides accurate risk estimate for genetic diseases. Computer programs are available to apply this method whenever genetic markers are suspected of being linked to a disease gene.

A Bayesian method for the estimation of penetrance: application to mandibulofacial and frontonasal dysostoses.

We describe a Bayesian method of estimating penetrance from genealogical data. It consists of calculating the likelihood of the data alone to make inferences about penetrance without sample space considerations. The method is applied to mandibulofacial dysostosis giving a penetrance of 0.908 with 0.95 credible interval of [0.809; 0.972] and to frontonasal dysostosis giving a penetrance of 0.670 with 0.95 credible interval of [0.457; 0.851].

Pyruvate-kinase (PK) and creatine-kinase (CK) in normal pregnancy and its implication in genetic counseling of Duchenne muscular dystrophy (DMD).

The activities of pyruvate-kinase (PK) and creatine-kinase (CK) were measured in 50 normal pregnant women in both serum and amniotic fluid. Serum PK activity was found to be significantly higher in pregnant than in nonpregnant women, while serum CK did not differ significantly between the two population samples. In amniotic fluid, very little or no PK or CK activity was found. However, the mean PK activity in amniotic fluid obtained from women carrying male fetuses was significantly higher than in those carrying female fetuses. At the present time, it is concluded that if a woman at risk for having a son with Duchenne dystrophy is seeking genetic counseling when already pregnant her serum PK or CK should be compared with that of normal female pregnant controls. It is also suggested that determination of pyruvate-kinase in fetal blood might be used in addition to CK for prenatal diagnosis of Duchenne muscular dystrophy.

Recurrence risks for relatives in families with an isolated case of the fragile X syndrome.

The proportion of sporadic cases among affected males with fragile X-related mental impairment was reinvestigated in a new sample of family data and compared to previous studies. It was found that the estimate has increased over time from 0 in the original study to 0.24 in the present study. This difference indicated that the correction used for the ascertainment of families in the original study may not have been adequate and that the suggestion that all mothers of affected males are obligate carriers may be wrong. Based on this new information, recurrence risks for relatives in a family with an isolated case of the fragile X or Martin-Bell syndrome were calculated under different assumptions in order to investigate the effect of 1) the knowledge of the phenotype of ancestors of the proband, 2) the dependence of expression of the mutation on the sex of the carrier parent, 3) the value of the penetrance of mental impairment (MI), and 4) the equality of mutation rates in egg and sperm. The assumptions made for modelling the mutational process had the greatest effect on the recurrence risk in sibs of an isolated case, whereas small differences in penetrance parameters and assumptions based on whether the ancestors were known to be normal or of unknown phenotype made little difference. Recurrence risks for the sibs and first cousins of an isolated case calculated under different assumptions are presented.

Utility of linked markers in genetic counseling: estimation of carrier risks in X-linked ocular albinism.

We apply a method proposed by Rogatko et al. [1995: Am J Med Genet 59:24-32] to estimate carrier risks using genetic linkage data. The method is illustrated for X-linked ocular albinism. Linkage data from pedigrees were combined with genome mapping data to compute carrier risks for individuals with unknown carrier status based on pedigree data alone. We considered two situations. First, a linkage map with some ambiguity in the gene order was considered. This analysis allows us to examine the effect of incomplete genetic map information on risk computations. Second, published physical and meiotic mapping information was used to derive a linkage map that could be assumed known without ambiguity. In both situations, the mean and median estimate of carrier risk differed significantly from that obtained using pedigree relationships only, in that the computed risk was significantly different from the a priori value of 0.5. The 95% CI's associated with point estimates of risk made using the known map or an map with ambiguity did not overlap in some cases. These results suggest that the risk estimate and the confidence with which a risk estimate can be imparted may depend on the genetic map and marker data used in the risk estimation procedure. We conclude that the method presented here can be used to estimate genetic risk under a variety of analytical conditions.

Autosomal dominant late adult spinal muscular atrophy, type Finkel.

We describe clinical and genetic data from the study of two families with 80 members affected with the autosomal dominant, slowly progressive spinal muscular atrophy of late onset (average 48.8 years), first described by Finkel in 1962. Electromyography and muscle biopsy of a number of patients confirmed the neurogenic nature of the conditions. Unusual findings in this disorder were cramps, spontaneous fits of suffocation, and symptomatic myotonia. Other manifestations are slow loss of muscle strength and progressive proximal atrophy, which starts in the lower limbs and progresses to the upper limbs; hypoactive or absent tendinous reflexes; and generalized fasciculations. Sensory and cranial nerve function is unimpaired. Probabilities for genetic counseling are evaluated by means of a method adequate to the late-onset nature of the condition.

Allelic instability as a predictor of survival in Egyptian breast cancer patients.

This work was designed with the purpose of determining whether the presence of allelic imbalances (AI) such as microsatellite instability (MSI) and loss of heterozygosity (LOH) in chromosomes 2, 11, 13, and 17 in primary breast cancer could be used as prognostic indicators of patient survival. The DNA from breast cancers removed from 29 patients who were followed-up for up to five years was analyzed for MSI and LOH using a panel of 24 markers located at chromosome 2 (TPO, D2S131, D2S144, D2S171, D2S177, D2S119, D2S123, D2S147 and D2S136), chromosome 11 (C-RAS, Int-2, D11S940, D11S912), chromosome 13 (D13S289, D13S260, D13S267, D13S218, D13S263, D13S155, and D13S162), and chromosome 17 (D17S513, TP53, D17S855, and D17S785). The frequency of AI in the markers studied ranged from 30-55%, being highest for D11S912, D2S171, TP53 and D17S855. Univariate analysis showed association between overall survival rate and AI in 9 out of the 24 markers tested. Five of them were located at the area of the mismatch repair gene (MMR)-2 gene, two at 11p, one at 13q and one at 17p. Using multivariate analysis, it was observed that only pathological and clinical stage (defined as stage II or not) and AI at D2S171, D11S912, or D17STP53 generated significant predictive models for survival.

Phase II trial of induction high-dose chemotherapy followed by surgical resection and radiation therapy for patients with marginally resectable non-small cell carcinoma of the lung.

The combination of carboplatin and paclitaxel is an active regimen in non-small cell lung cancer (NSCLC). Historically, patients with stage III disease have manifested higher response rates than patients with metastatic disease, and patients achieving a pathologic complete response to induction chemoradiation therapy prior to surgery have shown better long-term outcome. Based upon our pilot data using high-dose carboplatin and paclitaxel, we designed a phase II trial in patients with marginally resectable stage IIIA NSCLC. Ten patients, with bulky nodal stage IIIA disease, initially received etoposide (2 g/m2) and granulocyte colony-stimulating factor (G-CSF) to mobilize peripheral blood stem cells (PBSC). Two cycles, 28 days apart, of carboplatin (AUC 12 in seven patients; AUC 16 in three patients) and paclitaxel (250 mg/m2) were administered with filgrastim (5 microg/kg) and PBSC support. After re-evaluation, patients underwent a thoracotomy followed by radiotherapy (44-60 Gy) if deemed resectable, or radiotherapy alone (60 Gy) if not resectable. The median age was 58.5 years (48-66) with a median ECOG performance status of 0 (0-1). Histology was adenocarcinoma in seven patients; the remainder had either squamous cell, large cell or bronchoalveolar carcinoma. Based on CT radiography, the overall response rate was 40%. Eight of ten patients underwent resection with four right pneumonectomies, three right upper lobectomies and one wedge resection of the right upper lobe. Six patients had a complete resection. Of eight patients resected, four were downstaged by induction therapy, three remained unchanged and one was found to have more extensive disease. The remaining two patients developed metastatic disease while receiving chemotherapy. The median dose of postoperative radiotherapy was 54 Gy (35-66 Gy). Actual median follow-up for all patients was 89 weeks (25 to 136+). The actuarial median overall survival was 124 weeks (25 to 136+) and time to progression was 57 weeks (17 to 136+). The median dose of carboplatin delivered expressed as mg/m2 was 779 (615-1540). Neutropenic fever occurred in two patients during the initial mobilization cycle only. The median number of units of RBC and/or platelets transfused was 0 (0-2 and 0-6, respectively). There were no significant non-hematologic toxicities. High-dose induction chemotherapy with stem cell rescue is feasible and safe with an acceptable response rate. Thoracotomy, including pneumonectomy and postoperative radiotherapy, were well tolerated by patients after undergoing high-dose induction chemotherapy with no apparent increase in peri-operative morbidity. The pathologic complete response rate was low--one out of ten patients. These results indicate that dose escalation of induction chemotherapy does not improve response rates even in this highly selected patient population. Accordingly, the complexity and potential toxicity of high-dose chemotherapy, as delivered in this trial as neoadjuvant treatment of non-small cell lung cancer, is not warranted.