Carboplatin and gemcitabine in first-line treatment of elderly patients with advanced non-small cell lung cancer: data from a retrospective study.

We retrospectively evaluated elderly patients with advanced non-small cell lung cancer (NSCLC) treated with carboplatin (AUC 4-5) and gemcitabine (1,000-1,200 mg/m²). Thirty-six patients with performance Status (pS) 0-1 and median age 73 (range 70-78 years) were considered. Histology was squamous cell carcinoma (8 patients), adenocarcinoma (22) and NSCLC not otherwise specified (6). 149 cycles of chemotherapy were administered with a median of 3 per patient (range 3-6). Grade 3 non-hematologic toxicities were dyspnea (1 patient) and fever (1). Grade 3/4 hematologic toxicities were anemia (6), neutropenia (6) and thrombocytopenia (10), with dose reduction required in 13 patients. The overall disease control rate was 44.4%. We recorded no complete response, 8 partial response, 8 stable disease and 20 progressive disease. After a medium follow-up of 11 months, median progression- free survival and median survival were 5 and 11 months, respectively. Carboplatin and gemcitabine is a safe and active regimen in elderly advanced NSCLC patients with good PS.

Renal toxicity and osteonecrosis of the jaw in cancer patients treated with bisphosphonates: a long-term retrospective analysis.

BACKGROUND: Bisphosphonates (BPs) are the mainstay of bone-directed therapy for bone metastases from multiple myelomas and a wide range of solid tumours, but some patients experience renal toxicity or osteonecrosis of the jaw (ONJ). PATIENTS AND METHODS: We reviewed data relating to 398 patients treated with intravenous BP for bone metastases, checking their serum creatinine levels throughout the treatment period in order to assess renal function, and seeking any signs and symptoms of ONJ recorded in their medical records. We also analysed other risk factors for renal toxicity and ONJ in patients who developed them. RESULTS: The median treatment period was 14 months (range 1-119); 108 patients received BP for more than 1 year, and 112 for more than 2 years. Sixteen patients (4%) developed renal toxicity after a median of 24 months of BP treatment, eight of them had been treated for more than 2 years. Ten patients (2.5%) were diagnosed as having ONJ after a median of 39 months on BP, only three of them had been treated for less than 2 years. Two patients experienced both ONJ and renal toxicity. CONCLUSIONS: The low incidence of ONJ and renal toxicity indicates the safety of BP. However, prevention and early detection are still the "first-line therapy" for decreasing their occurrence further.

Pemetrexed in combination with oxaliplatin as a first-line therapy for advanced gastric cancer: a multi-institutional phase II study.

BACKGROUND: This clinical trial assessed the efficacy of pemetrexed combined with oxaliplatin (PEMOX) in patients with advanced gastric cancer (AGC). PATIENTS AND METHODS: Forty-four patients with untreated AGC were enrolled to evaluate response rate (RR). Patients received pemetrexed (500 mg/m(2)) with vitamin supplementation and oxaliplatin (120 mg/m(2)) every 21 days for six cycles or until disease progression occurred. RESULTS: Median age was 62 years (range 26-76). The majority of patients (93%) had metastatic disease. Sixteen of the 44 patients achieved confirmed response [RR 36%; 95% confidence interval (CI) 22% to 52%]; four complete responses and 12 partial responses (complete and partial responses according to the RECIST guidelines are the confirmed-responses observed in the study population). Median time to tumor progression (TTP) was 6.2 months (95% CI 4.3-7.5) and median survival was 10.8 months (95% CI 7.7-17.2). A total of 220 cycles were administered, with a median of six cycles. Most common grade 3/4 toxic effects were neutropenia in 41% of patients (19% of cycles) and thrombocytopenia in 11% of patients (4% of cycles). Treatment delays or dose reductions for toxicity occurred in 10% and 5% of cycles, respectively. CONCLUSIONS: PEMOX is active and well tolerated in AGC. RR, TTP, and survival were comparable to those achieved in studies using different 5-fluorouracil (5-FU)-oxaliplatin combinations, without the inconvenience of prolonged 5-FU schedules.

Cytofluorimetric evaluation of DNA ploidy in lung cancer: a bronchoscopic study.

The study of the biological characteristics of lung cancer is gaining more and more interest both because of their potential role as prognostic indicators and for therapeutic reasons. The DNA content estimated by flow cytometry in surgical samples of non-small cell lung cancer (NSCLC) has already been demonstrated to be correlated with survival in these patients. From July 1990 to February 1992 we analyzed the DNA distribution of bronchoscopic biopsies from 88 patients with lung cancer (18 small cell lung cancer, SCLC, and 68 NSCLC, two unspecified histology). Twenty-eight tumors (34.6%) had a diploid DNA distribution, while 53 were aneuploid (65.4%). A correlation was found between DNA ploidy and survival. Evaluation of the DNA content in bronchoscopic samples in a large series of patients could determine the role of this analysis prior to surgery in NSCLC and its value as a marker with respect to prognosis and response to therapy in SCLC.

High pathological response rate in locally advanced esophageal cancer after neoadjuvant combined modality therapy: dose finding of a weekly chemotherapy schedule with protracted venous infusion of 5-fluorouracil and dose escalation of cisplatin, docetaxel and concurrent radiotherapy.

BACKGROUND: This phase I study was aimed at defining the toxicity profile and pathological response rate of a neoadjuvant schedule including weekly docetaxel and cisplatin, protracted venous infusion (PVI) of 5-FU and concomitant radiotherapy (RT) in locally advanced esophageal cancer. PATIENTS AND METHODS: The schedule consisted of a first phase of chemotherapy alone and a second phase of concurrent chemoradiation. Initial doses were: docetaxel and cisplatin 20 mg/m2 on days 1, 8, 15, 29, 36 and 43 plus 5-FU 150 mg/m2 PVI on days 1-21 and 29-49; RT (40 Gy) started on day 29. In the following steps the doses were escalated up to docetaxel 35 mg/m2 and cisplatin 25 mg/m2 on days 1, 8, 15, 29, 36, 43, 50 and 57 plus 5-FU 180 mg/m2 PVI on days 1-21 and 150 mg/m2 PVI on days 29-63 concurrently with RT 50 Gy. RESULTS: Forty-seven patients were enrolled and 46 completed the planned treatment. During the concomitant phase, grade 3-4 hematological toxicities occurred in three patients (6.5%) (or 3/174 cycles) and non-hematological toxicities in six patients (13%) (or 7/179 cycles). A pathological downstaging was obtained in 59.6% of the cases (28/47): complete remission (pCR) in 14 patients, near pCR (residual microfoci on the primary pN0) in eight patients, pT2 pN0 in three patients and partial response on the primary with positive lymph nodes in three patients. Six (13%) and 13 (28%) patients were considered stable and non-responders, respectively. In the last dose level, eight pCR and four near-pCR were obtained out of 15 patients. The maximum tolerable dose was not formally defined because dose escalation was stopped at the last dose level. CONCLUSION: This schedule represents a feasible treatment and the high pathological response rate is extremely encouraging; the doses found in the last dose-level are the basis for an ongoing phase II study at our institution.

Gamma knife radiosurgery in brain metastases from testicular tumors.

To our knowledge, there are no published reports on the effectiveness of radiosurgery in the management of brain metastases from testicular nonseminomatous germ cell tumor. The authors evaluate the results of gamma knife (GK) treatment in three patients with these unusual intracranial lesions. Between April 1995 and July 2001, three patients with brain metastasis from testicular nonseminomatous germ cell tumor underwent adjuvant radiosurgery at our department. The primary tumor had been surgically removed in all cases. At diagnosis, one patient was stage IB and two were stage III poor risk. Chemotherapy and whole brain radiotherapy were administered before radiosurgery in all cases. Pre-GK radiotherapy was administered with a daily fraction dosage of 1.8-2.0 Gy. The indications for radiosurgery were tumor volume <20 cm3, microsurgery too risky, refusal of surgery. All the lesions were located in eloquent brain areas. Post-GK high-dose chemotherapy with autologous peripheral-blood stem-cell rescue was administered in two cases due to systemic recurrence of the disease. All patients are still alive with a median and mean follow-up period after radiosurgery of 63 and 68.3 mo, respectively. They had no neurological deficits at the latest examination. Neuroradiological follow-up invariably showed tumor growth control (complete response in two cases and partial response in one) with typically delayed post-radiosurgical imaging changes (transient in two cases and long-lasting in one). In conclusion, GK seems to be highly effective and safe in brain metastases from testicular nonseminomatous germ cell tumor. In cases with diffuse metastatic brain involvement, the whole brain radiotherapy preceding radiosurgery should be delivered with 1.8 Gy daily fraction to prevent the risk of long-lasting post-radiosurgical imaging changes.

Gemcitabine and continuous infusion of 5-fluorouracil in locally advanced and metastatic pancreatic cancer: a phase I-II study.

BACKGROUND: Gemcitabine has been recently recognized as standard treatment in advanced pancreatic cancer. To potentiate its single-agent activity we conducted a phase I-II study with the primary objective of establishing the maximum tolererated dose (MTD) of gemcitabine and continuous infusion 5-FU in patients with locally advanced or metastatic pancreatic cancer. PATIENTS AND METHODS: Fifteen patients received a fired dose of 5-FU 200 mg/mq protracted infusion for six months. Gemcitabine was administered weekly for three out of four weeks for six cycles at escalating doses of 800 mg/mq to 1100 mg/mq. RESULTS: MTD was established at 1000 mg/mq of gem citabine. Of the 11 evaluable patients, 7 patients had stable disease, 1 had partial response and 3 had progressive disease. Of the 14 patients evaluable at follow-up, median time to progression was 5 months. Median survival was 10 months. CONCLUSION: This study confirms the good tolerability of the combination, of gemcitabine with 5-FU.

Carboplatin, doxorubicin and etoposide in the treatment of tumours of unknown primary site.

The aim of this study was to assess the activity and toxicity of a platinum-based treatment on a group of patients with unknown primary tumours (UPTs). Patients with a diagnosis of UPT underwent a standard diagnostic procedure. Treatment was started within 2 weeks from diagnosis and consisted of carboplatin 400 mg m(-2) day 1, doxorubicin 50 mg m(-2) day 1, etoposide 100 mg m(-2) days 1-3, every 21 days. Response was evaluated after three courses and treatment continued in case of objective response (OR) or symptom control. A total of 102 patients were eligible. The median age was 59 years, sex male/female 54/48, histology was mainly adenocarcinoma or poorly differentiated carcinoma. Nodes, bone, liver and lung were the most frequently involved sites. In all, 79 patients received at least three courses of treatment; 26 patients received six courses or more. Six complete responses and 21 partial responses were observed, for a total of 27 of 102 ORs or 26.5% (95% confidence interval 18.2-36.1%). The median survival was 9 months and median progression-free survival was 4 months. Toxicity was moderate to severe, with 57.8% of patients experiencing grade III-IV haematological toxicity, mainly leucopenia. The regimen employed has shown activity in tumours of unknown primary site, but was associated with significant toxicity. Such toxicity may be considered unjustified, given the large proportion of patients with tumours not likely to respond. Efforts should therefore be addressed to identify predictors of response to chemotherapy, thus limiting aggressive treatment to those patients who could benefit from it.

Correlations between family history and cancer characteristics in 2256 breast cancer patients.

A comparison of 692 early invasive breast cancer with, and 1564 without, a family history of breast cancer showed that the former were younger at diagnosis (P=0.002), had smaller tumours (P=0.012), were more frequently oestrogen receptor positive (P=0.006) and diagnosed preclinically (P<0.001).

Multicentric phase II trial of gemcitabine plus epirubicin plus paclitaxel as first-line chemotherapy in metastatic breast cancer.

In this phase II, multicentre trial, patients with metastatic breast cancer (MBC) were treated with a combination of gemcitabine, epirubicin and paclitaxel (GET). The primary objective of this study was to determine the tolerability and activity in terms of complete responce (CR) and overall response rate of the GET combination in this patient population. Patients with no prior treatment for MBC, and at least one bidimensionally measurable lesion received gemcitabine 1000 mg m(-2) intravenously (i.v.) over 30 min on days 1 and 4, followed by epirubicin i.v. at 90 mg m(-2) on day 1, and paclitaxel 175 mg m(-2) over 3 h on day 1, every 21 days, up to eight courses. From May 1999 to June 2000, 48 patients were enrolled from seven Italian institutions. A total of 297 chemotherapy courses were administered with a median of six cycles patient(-1) (range 1-8). Seven patients (15%) obtained CR and 27 patients (56%) had partial responce, for an overall response rate of 71% (95% CI: 58.3-83.7). After a median follow-up of 23.7 months (range 7.0-34.4), median progression-free survival was 10.5 months (95% CI: 9.2-11.7), and median overall survival 25.9 months. The main haematological toxicity consisted of grade 3 or 4 neutropenia that occurred in 62% of cycles (22% grade 4 and 40% grade 3). The GET combination is active and well tolerated as first-line chemotherapy for MBC.

Cisplatin versus carboplatin in combination with mitomycin and vinblastine in advanced non small cell lung cancer. A multicenter, randomized phase III trial.

BACKGROUND: In advanced not selected NSCLC chemotherapy achieved an advantage of approximately 1-2 months on median survival versus best supportive care. Chemotherapy seems to improve symptoms control, even if randomised studies with quality of life as first endpoint are lacking and often chemotherapy toxicity compromises the frail cost/benefit ratio. The aim of the present study is to evaluate the impact on QoL, substituting cisplatin, a pivot drug in NSCLC therapy, with carboplatin, an analogue with an improved toxicity profile. The combination of cisplatin with Mitomycin and Vinblastine was one of the most frequently used in the palliative setting at the time of design of our study. METHODS: Patients were randomized to receive MVP regimen (Mitomycin-C 8 mg/m2 d1, Vinblastine 4 mg/m2 d 1-8, Cisplatin 100 mg/m2 d1) or MVC regimen (Mitomycin-C 8 mg/m2 d1, Vinblastine 4 mg/m2 d 1-8, Carboplatin 300 mg/m2 d1) every 3 weeks. The QoL was evaluated by the Spitzer QL-Index and by the EORTC QLQ-C30+LC 13 questionnaires before chemotherapy, after one cycle, after three cycles, and then every 6 weeks in the first 6 months and every 3 months thenafter. RESULTS: From September 1994 to July 1997, 153 consecutive patients were randomized to MVP (75 patients) or MVC arm (78 patients). Despite difficulties in carrying out and analysing QoL items in such patients, the global QoL evaluated by the Spitzer's questionnaire suggested an advantage for MVC regimen (P=0.05) and a significant difference was observed in global health subdomain (P=0.04). The disease-related symptoms improved with time, and the benefits lasted for the entire treatment period. When evaluated with the EORTC questionnaire there was significantly less nausea and vomiting (P=0.0001), appetite loss (P=0.01), insomnia (P=0.03), constipation (P=0.01) and peripheral neuropathy (P=0.01) in favour of MVC, and a trend for less hair loss (P=0.05). The advantage lasted for all the duration of chemotherapy. No differences were observed in global quality of life subdomain (P=0.40) between the two regimen. QoL was the first endpoint and the statistical power was inadequate to assess other parameters. However, we reported a response rate of 43.1 and 38.6%, respectively, in MVP and MVC arm (P=0.59) and a median survival of 10.2 and 7.2 months, respectively, for cisplatin and carboplatin arm (P=0.39). CONCLUSIONS: The carboplatin containing regimen (MVC) has a significant better toxicity profile than the cisplatin containing (MVP) regimen as proven both by the EORTC questionnaires and by the WHO toxicity data reported by physicians. No significant differences in terms of response rate, time to progression and overall survival were observed between the two regimen. The two chemotherapy regimen showed a similar effectiveness in symptom palliation when evaluated with C30 addendum of EORTC QOL questionnaire. With the Spitzer's questionnaires a trend towards an improved quality of life index was observed during treatment with the carboplatin combination in comparison to the cisplatin combination. This difference, however, was not observed when the global quality of life was evaluated with the EORTC patients compiled questionnaires. A carboplatin containing regimen with better toxicity profile and a similar potentiality for symptoms control offers an option in comparison to similar cisplatin containing combinations in the palliative treatment of advanced NSCLC.

A phase II study of induction chemotherapy with gemcitabine (G) and cisplatin (P) in locally advanced non-small cell lung cancer: interim analysis.

BACKGROUND: Gemcitabine-cisplatin (GP) combination is one of the most active and well tolerated regimens in advanced non-small cell lung cancer (NSCLC). The aim of this study is to evaluate the activity and toxicity of the GP regimen as a 21-day schedule in patients (pts) with stage IIIAN2-IIIB NSCLC. PATIENTS AND METHODS: From October 1997 to July 2000, 47 pts entered the study: 43 were eligible (40 men and three women); median age was 61 years (range 45-73); ECOG PS 0-1; histology was squamous (20 pts), adenocarcinoma (12 pts), large cell (five pts), and undifferentiated (six pts); stage was IIIAN2 (14 pts, 32.56%), and IIIB (29 pts, 67.44%). Malignant pleural effusion or superior vena cava syndrome was criteria of exclusion. Induction treatment consisted of three cycles of GP (G 1250 mg/m(2) i.v. on days 1 and 8, and P 100 mg/m(2) on day 8 every 3 weeks). Responding and stable pts underwent surgery (S) and/or radiotherapy (RT). RESULTS: Following a minimum of two cycles, 39 pts were evaluable for response and 42 for toxicity. Two pts had complete responses (CR; 5.2%), 24 had partial response (PR; 61.5%), eight had stable disease (SD; 20.5%), and five had progressive disease (PRO; 12.8%). WHO grades 3 and 4 anaemia, neutropenia and thrombocytopenia were observed in two, four and two pts, respectively; non-haematological toxicity was moderate. After induction, stable and responding pts received either RT (18 pts) or S+RT (13 pts). Among the 16 resected pts, a radical complete resection was possible in 13 cases (81.3%), whereas tumour down-staging was observed in nine pts (56.2%). CONCLUSION: GP, as a 3-week neoadjuvant schedule, appears a safe and active regimen.

Phase II randomized study of dacarbazine, carmustine, cisplatin and tamoxifen versus dacarbazine alone in advanced melanoma patients.

This randomized phase II trial was performed to define the activity and toxicity of the combination of dacarbazine (DTIC), carmustine (BCNU), cisplatin (DDP) and tamoxifen (DBDT regimen) versus DTIC alone in patients with metastatic melanoma. Sixty patients with metastatic melanoma were randomly assigned to receive BCNU 150 mg/m2 intravenously (i.v.) on day 1, cisplatin 25 mg/m2 i.v. daily on days 1 to 3, DTIC 220 mg/m2 i.v. daily on days 1 to 3 and tamoxifen 160 mg orally daily for 7 days prior to chemotherapy (DBDT arm; arm A). Treatment cycles were repeated every 28 days, while BCNU was given every two cycles. The DTIC arm (arm B) patients received DTIC alone 1200 mg/m2 i.v. on day 1, repeated every 21 days. Patients were evaluated every two cycles; responding patients continued the treatment for a maximum of 12 cycles. The overall response rate was 26% in the DBDT arm and 5% in the DTIC arm. Complete responses were 2.5% for DBDT and 0% for DTIC. The median progression-free survival and the median survival were 4 and 9 months, respectively for DBDT, and 2 and 7 months for DTIC. DBDT was associated with significant haematological toxicity: 33% of the patients experienced a grade III or IV neutropenia and 28% a grade III or IV thrombocytopenia. In conclusion, the overall response rate obtained with DBDT was greater than that obtained with DTIC alone; however, this combination increases toxicity with limited impact on overall survival.

Use of the ratio of time to progression following first- and second-line therapy to document the activity of the combination of oxaliplatin with 5-fluorouracil in the treatment of colorectal carcinoma.

BACKGROUND: It has been proposed that the activity of a second-line treatment regimen can be documented by showing that the time to progression (TTP) following second-line therapy is longer than the TTP following first-line therapy in the same patients. PATIENTS AND METHODS: The ratio of TTP during first and second-line therapy, identified as the growth modulation index (GMI), was determined in 34 patients with advanced colorectal cancer. First-line chemotherapy consisted of one of several schedules of leucovorin (LV)-modulated 5-fluorouracil (5-FU) or raltitrexed. Second-line therapy consisted of the combination of LV-modulated 5-FU and oxaliplatin (1-OHP). Patients were switched to second-line therapy upon evidence of progressive disease following first-line therapy. RESULTS: Median TTP following first-line therapy was 13 weeks (95% confidence interval (CI): 7.6-18.7), while median TTP following second-line therapy was 31 weeks (95% CI: 21.3-41.0). Sixteen patients (47%; 95% CI: 35%-59%), showed a GMI > or = 1.33, while the remaining 18 patients (53%; 95% CI: 40%-66%) had a GMI < 1.33. Log-rank analysis of the Kaplan-Meier curves of TTP following first- versus second-line therapy demonstrated a statistically significant difference in favour of second-line therapy (P = 0.0081). CONCLUSIONS: This study demonstrates the utility of the GMI as a tool for assessing the activity of novel second-line therapeutic programs.

The gemcitabine/epirubicin/paclitaxel trials in advanced breast cancer.

Numerous trials have shown that the pharmacokinetic interferences of epirubicin (Ellence)/paclitaxel (Taxol) combinations produce less pharmacodynamic effect than doxorubicin/paclitaxel regimens. Paclitaxel is more easily combined when infused over 3 (as compared to 24) hours; the administration of optimal doses of both agents is important. Based on these findings, a phase II study was performed to evaluate the feasibility and activity of the combination of gemcitabine (Gemzar), epirubicin, and paclitaxel as first-line therapy in advanced breast cancer. Patients received gemcitabine at 1,000 mg/m2 on days 1 and 4, plus epirubicin at 90 mg/m2 on day 1, plus paclitaxel at 175 mg/m2/d on day 1 every 21 days. After six courses, patients less than 60 years old and in complete or partial remission or stable disease were treated with high-dose chemotherapy as consolidation treatment. The overall response rate was 92%, with 31% complete responses; 25 patients received high-dose chemotherapy, achieving a final overall response rate of 97%, with 47% complete responses. At a median follow-up of 25 months, median progression-free survival is 21 months. Grade 4 neutropenia was observed in 64% of patients. Other hematologic toxicities were mild. Mild to moderate peripheral neuropathy was experienced by 39% of patients; grade 2 or 3 mucositis occurred in 25% and 17% of patients, respectively. Based on these results, a multicenter trial has been started in seven Italian centers to confirm the feasibility of this regimen.

Comparative study of clinical, pathological and biological characteristics of symptomatic versus asymptomatic breast cancers.

BACKGROUND: It is well known that mammographic screening reduces breast cancer mortality. One possible explanation for this effect is that screening makes it possible to detect smaller breast cancers with fewer involved nodes, but another hypothesis is that some screening-detected tumors are in a pathologically and biologically different phase of evolution from those that are detected clinically. The aim of the present study was to compare the biological, pathological and clinical characteristics of symptomatic vs. asymptomatic breast cancers. PATIENTS AND METHODS: The study considers a series of 1916 consecutive patients who underwent surgery for stage I and II infiltrating breast cancer at Verona hospitals after having undergone ultrasound and mammography (at least one of which was positive). They were divided into two groups on the basis of why they decided to undergo the imaging examinations: group A refers to the 1247 patients with a palpable lump, and group B to the 616 who were asymptomatic. RESULTS: The patients in group A were older, and had larger tumors and a higher percentage of positive nodes than those in group B; they also had significantly higher grade tumors, higher Ki-67 levels, and a higher percentage of ER and PgR negative and c-erbB-2 positive tumors (all of the P-values were significant). A logistic regression analysis adjusted for tumor diameter and age showed a reduction in the significance of each of the considered variables, but all of them remained significantly associated with the modality of diagnosis except ER, PgR and c-erbB-2. CONCLUSIONS: Our results suggest that asymptomatic tumors are biologically different from their clinically presenting counterparts, thus confirming the hypothesis that progression towards greater malignancy may occur during the natural history of breast cancer.

Immunocytochemical assessment of bone marrow aspirates for monitoring response to chemotherapy in small-cell lung cancer patients.

Recent reports have suggested that tumour cell immunodetection in bone marrow of small-cell lung cancer patients is by far more frequent than found cytohistologically and may have clinical relevance. This study evaluates primarily the efficacy of chemotherapy as method of in vivo purging, but also the relationship of marrow involvement with survival. A total of 112 bone marrow aspirates from 30 chemo-naïve patients were stained twice using anti-NCAM antibodies, first at diagnosis and then after chemotherapy (24 patients) or at disease progression (six patients). Marrow contamination was associated with lower survival (P = 0.002), and was also detected in 7/17 patients conventionally staged as having limited disease. At multivariate analysis, marrow involvement was an independent factor of unfavourable prognosis (P = 0.033). The amount of tumour contamination, before and after chemotherapy, remained unchanged also in responders and even in the subset of patients with apparent limited disease. Following chemotherapy, bone marrow became tumour negative only in 25% of initially positive responders and in none of non-responders. Our results indicate that (i) chemotherapy is not effective in purging bone marrow even in chemo-responsive patients and (ii) a subset of patients with limited disease and negative bone marrow aspirates might have a more favourable prognosis.