Bortezomib retreatment in relapsed multiple myeloma - results from a retrospective multicentre survey in Germany and Switzerland.

OBJECTIVES: This multicenter, retrospective survey evaluated the efficacy and safety of bortezomib retreatment in patients with relapsed multiple myeloma who had responded to initial bortezomib treatment. METHODS: Clinical records of 94 patients receiving bortezomib retreatment in Germany and Switzerland were reviewed. RESULTS: Sixty patients were included according to prespecified criteria. Patients had received a mean 3.7 ± 2.3 therapies prior to initial bortezomib. Overall response rate to bortezomib retreatment was 63.3%; 8 (13.3%), 3 (5.0%) and 27 (45.0%) patients achieved complete response (CR), near-CR and partial response, respectively. Response to retreatment was associated with response to initial treatment (75.0% of patients with CR to initial treatment responded to retreatment) and treatment-free interval (TFI) after initial treatment (76.9 vs. 38.1% overall response rate for patients with TFI >6 vs. ≤ 6 months). After retreatment, median time to progression was 9.3 months. Median TFI was 5.7 months; 31.7, 25.0 and 15.0% of patients experienced a TFI longer than 6, 9 and 12 months, respectively. Reported adverse drug reactions were consistent with the known safety profile of bortezomib and most resolved completely. CONCLUSIONS: These results demonstrate that relapsed multiple myeloma patients who respond to initial bortezomib treatment have a sustained susceptibility to bortezomib and do not experience uncommon toxicity to retreatment.

Sequential high dose chemotherapy as initial treatment for aggressive sub-types of non-Hodgkin lymphoma: results of the international randomized phase III trial (MISTRAL).

INTRODUCTION: Sequential high dose (SHiDo) chemotherapy with stem cell support has been shown to prolong the event-free survival in patients with diffuse large B-cell lymphoma. METHODS: To confirm this result in a multicenter trial, we randomized patients with aggressive NHL, to receive either eight cycles of CHOP or SHiDo. The primary endpoint was overall survival. RESULTS: 129 evaluable patients were randomized to receive either CHOP or SHiDo: median age, 48 years; 62% male; stage III+IV: 73%; age adjusted International Prognostic Index 1/2/3: 21%/52%/27%. Toxicity grades 3+4 were more pronounced in the SHiDo-arm with 13% versus 3% of patients with fever; 34% versus 13% with infections; 13% versus 2% with esophagitis/dysphagia/gastric ulcer. The remission rates were similar in SHiDo and CHOP arms with 34%/37% complete remissions and 31%/31% partial remissions, respectively. After a median observation time of 48 months, there was no difference in overall survival at 3 years, with 46% for SHiDo and 53% for CHOP (P = 0.48). CONCLUSION: In this multicenter trial, early intensification with SHiDo did not confer any survival benefit in previously untreated patients with aggressive NHL and was associated with a higher incidence of grades 3/4 toxicity.

Treatment of hairy cell leukemia with cladribine (2-chlorodeoxyadenosine) by subcutaneous bolus injection: a phase II study.

BACKGROUND: To assess the activity and toxicity of 2-chlorodeoxyadenosine (cladribine, CDA) given by subcutaneous bolus injections to patients with hairy cell leukemia (HCL). PATIENTS AND METHODS: Sixty-two eligible patients with classic or prolymphocytic HCL (33 non-pretreated patients, 15 patients with relapse after previous treatment, and 14 patients with progressive disease during a treatment other than CDA) were treated with CDA 0.14 mg/kg/day by subcutaneous bolus injections for five consecutive days. Response status was repeatedly assessed according to the Consensus Resolution criteria. RESULTS: Complete and partial remissions were seen in 47 (76%) and 13 (21%) patients, respectively, for a response rate of 97%. All responses were achieved with a single treatment course. Most responses occurred early (i.e. within 10 weeks) after start of CDA therapy, but response quality improved during weeks and even months after treatment completion. The median time to treatment failure for all patients was 38 months. Leukopenia was the main toxicity. Granulocyte nadir (median 0.2 x 10(9)/l) was strongly associated with the incidence of infections (P = 0.0013). Non-specific lymphopenia occurred early after CDA treatment, and normal lymphocytes recovered slowly over several months. No significant associations were found between infections and nadir count of lymphocytes or any lymphocyte subpopulation. No opportunistic infections were observed. CONCLUSIONS: One course of CDA given by subcutaneous bolus injections is very effective in HCL. The subcutaneous administration is more convenient for patients and care providers, and has a similar toxicity profile to continuous intravenous infusion. The subcutaneous administration of CDA is a substantial improvement and should be offered to every patient with HCL requiring treatment with CDA.

[2-CDA in treatment of hairy cell leukemia: a comparison between intravenous and subcutaneous administration. Swiss Study Group of Applied Cancer Research]. / 2-CDA bei der Behandlung der Haarzell-Leukämie: ein Vergleich zwischen intravenöser und subkutaner Verabreichung. Schweizerische Arbeitsgemeinschaft für angewandte Krebsforschung.

2-chlorodeoxyadenosine (2-CDA) is very effective in the treatment of patients with hairy-cell leukaemia, with an overall response rate of 80-95%. The standard treatment is a continuous intravenous infusion for 7 days. The bioavailability of 2-CDA after subcutaneous injection is 100%, but the concentration-time profile is completely different compared to continuous intravenous administration. In the present study we compared the intravenous standard treatment (group 1, n = 22; 0.1 mg/kg/d for 7 days, civ.) with subcutaneous administration of 2-CDA (group 2, n = 62; 0.14 mg/kg/d for 5 days, s.c.) in patients with hairy-cell leukaemia. In group 1, 96% (21/22) of patients responded to 2-CDA (complete remission 73%, partial remission 23%) and in the second group 97% were responsive (complete response 76%, 47/62; partial remission 21%, 13/62). The percentage for moderate and severe infections in the trial with intravenous and subcutaneous treatment was 14% and 26% respectively (p = 0.37). We conclude that subcutaneous administration of 2-CDA in patients with hairy-cell leukaemia is feasible and economical and results in comparable responses and toxicity compared to the intravenous standard treatment.

Reduced dose of subcutaneous cladribine induces identical response rates but decreased toxicity in pretreated chronic lymphocytic leukaemia. Swiss Group for Clinical Cancer Research (SAKK).

PURPOSE: To study the efficacy and the safety of cladribine (2-chlorodeoxyadenosine, 2-CDA) administered as 24-hour infusions or as subcutaneous bolus injections at two different doses to patients with relapsing or refractory chronic lymphocytic leukaemia (CLL). PATIENTS AND METHODS: In this non randomised 2-cohort study, 20 patients with pretreated CLL received cladribine at a dose of 0.7 mg/kg/cycle as continuous i.v. infusions over seven days (group 1) and 35 patients were treated at a reduced dose of 0.5 mg/kg/cycle given as s.c. bolus injections for five days (group 2). After two cycles of four week duration, response was assessed. In the case of progressive disease, therapy was discontinued, otherwise a maximum of four additional cycles were administered until best response. RESULTS: A total of 130 cycles were administered (group 1: 41, group 2: 89). Patient characteristics in both groups were comparable. The median dose intensities were 0.172 mg/kg per week and 0.123 mg/kg per week for groups 1 and 2, respectively (P < or = 0.0001). The overall response rate for all 55 patients was 38% (95% confidence interval (95% CI): 25%-52%), with 5% CR and 33% PR. Response was similar in both patient groups (35% in group 1, 40% in group 2). No association between cladribine dose intensity and response rate was found, and there was no difference between patients relapsing after or refractory to previous therapies (11 of 24 vs. 10 of 31). Median remission duration was six months in both groups. Toxicity, in particular infections (all WHO grades, 34% in group 1 versus 7% in group 2) and myelosuppression (grade 1-4 neutropenia, 72% versus 41% of cladribine cycles) were statistically significantly more frequent in group 1. CONCLUSION: Cladribine is active in heavily pretreated patients with chronic lymphocytic leukaemias. Dose reduction by 29% led to similar response and remission duration, but to a significant decrease of myelotoxicity and risk of infection. Cladribine administered as s.c. bolus injections at 0.5 mg/kg per cycle is safe and this dose level should not be exceeded in this patient population.

Fewer infections, but maintained antitumor activity with lower-dose versus standard-dose cladribine in pretreated low-grade non-Hodgkin's lymphoma.

PURPOSE: To study the efficacy and the safety of cladribine (2-chlorodeoxyadenosine [2-CDA]) administered at two different dosages. PATIENTS AND METHODS: In this two-cohort study, patients with low-grade refractory/relapsing non-Hodgkin's lymphoma (NHL) received 2-CDA at a dose of 0.7 mg/kg per cycle as a continuous intravenous (i.v.) infusion (group 1, n = 44) or at a reduced dose of 0.5 mg/kg per cycle as a subcutaneous (s.c.) bolus injection (group 2, n = 60). Three 2-CDA cycles at > or = 4-week intervals were planned, then treatment could be pursued until six cycles. RESULTS: A total of 300 cycles were administered (group 1, 114 cycles; group 2, 186). Patient characteristics in both groups were comparable. The median dose-intensities were 0.17 mg/kg/wk and 0.13 mg/kg/wk for groups 1 and 2, respectively (P < or = .0001). The overall response rate for all 104 patients was 54% (95% confidence interval [CI], 45% to 66%; 15% complete response [CR] and 39% partial response [PR]). Response was similar in both patient groups (57% in group 1 and 53% in group 2; P = .72), and no association between 2-CDA dose-intensity and response rate was found (P = .35). Median remission duration was 7 and 12 months in groups 1 and 2, respectively (P = .21). Toxicity, in particular opportunistic infections (> or = grade 2, 30% in group 1 v 7% in group 2; P = .003) and myelosuppression (> or = grade 3 neutropenia, 33% v 8% of 2-CDA cycles, P < .0001), were more frequent in group 1. Multiple logistic regression analysis showed that the infection risk (grade > or = 2) was decreased by 81% with 2-CDA dose reduction in group 2 after adjusting for number of pretreatment regimens and time since diagnosis (P = .01). CONCLUSION: When administered as a s.c. bolus injection, 2-CDA at 0.5 mg/kg per cycle is safe and this dose level should not be exceeded in this patient population.

Cladribine (2-CDA) given as subcutaneous bolus injections is active in pretreated Waldenström's macroglobulinaemia. Swiss Group for Clinical Cancer Research (SAKK).

Clinical trials with intravenous cladribine infusions in pretreated patients with Waldenström's macroglobulinaemia have shown a response rate of 40%. Our pharmacokinetic studies revealed that the bioavailability of subcutaneous cladribine is complete but that the concentration-time profile is very different from intravenous administration. We designed this phase II multi-institutional trial to study the activity and toxicity of cladribine given as s.c. bolus injections in patients with symptomatic Waldenström's macroglobulinaemia. Between May 1993 and October 1995, 25 patients were accrued: male/female 18/7, median age 65 years (range 44-85). All except one patient had been pretreated with more than one regimen (median 2, range 0-10). 18 patients had progressed under previous therapy and six were in relapse. All patients received cladribine for a total dose of 0.5 mg/kg per cycle as s.c. bolus injections divided over 5 d at > or = 4 week intervals, for a maximum of six cycles. All 25 patients were evaluable for toxicity and response. A total of 67 cycles were administered (median 3 cycles, range 1-6). Overall response rate including disease stabilization which had been progressive under previous therapy was 68%. 10 patients (40%, 95% CI 21-61%) achieved a partial remission. Seven responders had been progressive under previous therapy. Maximum responses were reached no later than the third cycle. Median time to treatment failure and remission duration were 4.4 (range 0.5-33) and 8 months (5-29), respectively. Four patients (16%) suffered from infections W.H.O. grade > or = 2 (pneumonia grade 2, Staphylococcus septicaemia grade 3, viral encephalitis and pneumonia, both grade 4 with complete resolution). No other severe adverse events were observed. Cladribine given as s.c. 5 d bolus injections was found to be active in pretreated Waldenström's macroglobulinaemia and resulted in durable remissions.

2-chlorodeoxyadenosine (2-CDA) therapy in previously untreated patients with follicular stage III-IV non-Hodgkin's lymphoma.

PURPOSE: This phase II multi-institutional trial was designed to assess response and toxicity of 2-chlorodeoxyadenosine (2-CDA) in patients with previously untreated follicular lymphoma. The clinical significance of detecting cells carrying the t(14;18) translocation (bcl-2/JH rearrangement) in peripheral blood and bone marrow by polymerase chain reaction (PCR) before, during and after treatment was also examined. PATIENTS AND METHODS: Between May 1993 and October 1995, 37 patients were accrued: male/female: 15/22, median age 51 years (range: 20-78), stage III/IV: 9/28. Patients received a total 2-CDA dose of 0.7 mg/kg as continuous s.c. or i.v. infusions over 7 days, every 28 days for a maximum of 5 cycles. A total of 165 cycles were administered. In 25 patients, blood and bone marrow before, during and after treatment were available for PCR analysis of the bcl-2/JH rearrangements. RESULTS: All 37 patients were evaluable for response and toxicity. The overall response rate was 84% (95% confidence interval, 68%-94%) with 14% CR (n = 5) and 70% PR (n = 26) and a median time to treatment failure of 15.7 months. bcl-2/JH rearrangement in peripheral blood and/or bone marrow was found in 10/25 of patients (40%) before treatment and 5 of these became repeatedly negative after 2-CDA therapy. There was no apparent association between bcl-2/ JH result and response. In 11 patients, 2-CDA was stopped because of progressive disease (n = 4), myelotoxicity (grade 2-3, n = 4), and other causes (n = 3, pulmonary embolism, metabolic disorder, and patient's decision). Four patients (11%) suffered from infections (grade 2-3). In 6 patients, persistent thrombocytopenia of 7.5 months (range: 3-21) occurred after completion of the 5 cycles. CONCLUSION: 2-CDA is active in untreated follicular lymphomas, but time to treatment failure suggests no advantage compared with standard treatment and toxicity on haematopoietic stem cells appears to be more pronounced. Molecular remission is induced in a considerable proportion of patients with disappearance of the bcl-2/JH rearrangement, and its possible significance as a predictive factor for quality of response and relapse warrants further study.

The number of circulating CD34+ blood cells predicts the colony-forming capacity of leukapheresis products in children.

In children, only a few guidelines are available for optimizing peripheral blood progenitor cell (PBPC) harvesting. We analyzed by means of flow cytometry and clonogenic assays 60 harvest products obtained from 20 children by standardized leukapheresis after treatment with chemotherapy and CSF. In addition, 27 fresh blood samples obtained prospectively during the mobilization phase were studied. CFU-GM/kg significantly correlated with MNC/kg, CD34+ cells/kg and CD34+33- cells/kg in apheresis products (P < 0.001). In fresh blood samples, CFU-GM/ml significantly correlated with MNC/ml, CD34+ cells/ml and CD34+33- cells/ml (P < 0.001). The numbers of CD34+ cells/ml, CD34+33- cells/ml and MNC/ml in 19 blood samples taken prior to leukapheresis were compared with CFU-GM/kg harvested and thawed after cryopreservation applying multiple regression analysis with stepwise variable selection. The number of circulating CD34+ cells/ml prior to leukapheresis highly correlated with and was predictive for the number of collected CFU-GM/kg (P < 0.001). In addition, a significant correlation (P < 0.05) between the number of progenitor cells/kg reinfused and the time to myeloid and platelet recovery was found in children undergoing high-dose therapy. Our data indicate that a single leukapheresis will be sufficient to obtain a minimum number of 5 x 10(4) CFU-GM/kg if the pre-harvest number of circulating CD34+ cells is > or = 10(5)/ml. Thus, our results will help to optimize PBPC transplantation in children.

Clonality and X-inactivation patterns in hematopoietic cell populations detected by the highly informative M27 beta DNA probe.

About 80% to 90% of females are informative for X-inactivation/methylation analysis with the probe M27 beta, which would therefore seem attractive in assessing clonality in hematologic cell populations. Eighteen acute lymphoid or myeloid leukemias, three chronic lymphocytic leukemias, and three chronic myelogenous leukemias as well as 12 malignant non-Hodgkin's lymphomas mostly showed extremely skewed clonal X inactivation (median allelic cleavage ratio [ACR] of unmethylated/inactive M27 beta alleles was 50, range 1 to 100) or hypermethylation of both alleles. Two lymphomas showed random M27 beta X inactivation but clonal antigen-receptor gene rearrangements. In normal peripheral blood leukocytes from 105 healthy females aged 2 to 96 years, the median ACR was 2 (range 1 to 100). Thus, it was significantly lower than in the leukemias (P = .0001, Mann-Whitney test), but extremely skewed patterns (ACR > 10.8, ie, > 80th percentile) were seen not only in the leukemias but also in 21/105 samples (20%) of normal leukocytes, and significantly more frequent in a population of elderly women (aged 75 to 96 years) compared with healthy children (aged 2 to 8 years) and younger women (aged 20 to 58 years) (P = .00125; chi 2 test). We conclude that in a population of cells derived from the hematopoietic system where clonality is uncertain, skewed M27 beta patterns are not reliable indicators for the presence of a clonal neoplastic disorder. The basis for severe X-inactivation skewing is unclear at present, but this finding raises interesting questions regarding the composition of the hematopoietic stem cell pool.

High incidence of infections after 2-chlorodeoxyadenosine (2-CDA) therapy in patients with malignant lymphomas and chronic and acute leukaemias.

BACKGROUND: 2-chlorodeoxyadenosine (2-CDA) is a new purine analogue which has been shown to be highly active in lymphoproliferative disorders. In this clinical trial we assessed 2-CDA toxicity and response rate in patients with various haematological malignancies who were heavily pre-treated and mostly refractory to standard treatment regimens. PATIENTS AND METHODS: Twenty-two refractory patients, 51 relapsing after standard chemotherapy and seven non-pre-treated patients were treated in a non-randomized prospective phase II multicentric study. Their median age was 54 years (range 18-84) and 56 of them were male. Thirty-one had non-Hodgkin's lymphoma (NHL), 11 chronic lymphocytic leukaemia (CLL), 1 prolymphocytic leukaemia (PLL), 13 hairy-cell leukaemia (HCL), 2 mycosis fungoides (MF), 3 multiple myeloma, 7 acute myeloblastic leukaemia (AML), 2 acute lymphoblastic leukaemia (ALL), 6 chronic myeloid leukaemia (CML) in blast crisis, 2 Hodgkin's disease, 1 Waldenström's macroglobulinemia and 1 Langerhans histiocytosis. 2-CDA 0.1 mg/kg/day was given as a continuous intravenous infusion for 7 days and recycled every 4 weeks. RESULTS: One hundred thirty-two courses of 2-CDA were administered to 80 patients, 76 of whom were evaluable for response. A) Toxicity: Myelosuppression: Neutropenia to over 50% of initial value occurred in 46% of patients, thrombocytopenia in 8%, and lymphopenia < 0.5 x 10(9)/l was seen in 41%. Infections occurred in 34/80 patients (43%). The risk of severe infections (WHO grades 3-4) correlated with increasing number of years after first diagnosis (p = 0.01) and low lymphocyte counts on days 1 and/or 14 (p = 0.04); 21 infections were opportunistic. B) Response: 70% patients with lymphoproliferative disorders of low malignancy attained complete or partial response (low-grade NHL 12/16, CLL + PLL 9/12, HCL 13/13, MF 2/2, myeloma 0/3); in patients with AML, ALL, CML in myeloid (n = 4) or lymphoid (n = 2) blast crisis and high-grade lymphoma responses were seen in only 11%. Response was inversely related to the number of pretreatments (p = 0.045). In responding NHL patients the mean lymphocyte count on day 14 of cycle 1 was significantly lower (median 0.6 x 10(9)/l) than that of non-responders (1.2 x 10(9)/l, p = 0.04). CONCLUSION: 2-CDA had a high activity even in heavily pretreated and refractory patients with low-grade lymphoproliferative disorders. In contrast to previously published studies, infections, mainly opportunistic, were a serious side effect in our study. In patients with severe lymphopenia at therapy initiation, the value of prophylactic anti-infective treatment should be studied.

[Prognostic value of bone marrow examinations on day 8 and day 17 of the first induction therapy cycle in acute myeloid leukemia]. / Aussagekraft von Knochenmarkuntersuchungen am Tag 8 und am Tag 17 des ersten Induktionstherapie-Zyklus bei akuter myeloischer Leukämie.

Bone marrow examination yields valuable diagnostic information for evaluation of treatment efficacy in acute myelogenous leukemia (AML) patients. However, the ideal timing of this procedure remains to be defined. Usually the first follow-up examination is performed on day 17 of the induction course. Since an earlier analysis might be helpful in deciding on the subsequent treatment strategy, we explored whether an examination on day 8 might be equivalent, in particular in terms of quality of the specimen, as compared with that on day 17. In 10 AML patients undergoing the same induction treatment, bone marrow aspirates were examined prospectively at diagnosis (day 0), immediately after termination (day 8) and on day 17 of the first treatment course. We found that the number of cells not classifiable due to artifacts, as well as the otherwise non-evaluable cells, did not differ significantly in aspirates obtained on day 8 as compared with those on day 17. The number of leukemic cells decreased significantly from day 0 to day 8, whereas no change was noted between days 8 and 17. We conclude that a marrow examination on day 8 is not more prone to misinterpretation due to artifacts than our standard examination on day 17. Thus, an early marrow analysis immediately after completion of the first induction cycle might be a helpful tool in evaluating course of AML.

Immunomodulation during prolonged treatment with combined interleukin-2 and interferon-alpha in patients with advanced malignancy.

Treatment with combined IL-2 and alpha-IFN has resulted in synergistic antitumour efficacy in animal studies. The mechanisms responsible for this synergy remain unclear. In this study, several immune parameters which might be involved in mediating antitumour activity have been monitored serially in 15 patients with advanced malignant melanoma or renal cell cancer during treatment with concurrent IL-2 and alpha-IFN. Both drugs were given subcutaneously in low to moderate (outpatient) dosages but for a prolonged duration. This treatment resulted in remarkable immunomodulation. In vivo induction of cytotoxicity against K562 and Daudi target cells was consistently seen, and percentages of peripheral blood cells expressing CD 25 (IL-2 receptor) and CD 56 (Leu-19) increased. In vitro proliferation of lymphocytes in response to IL-2 was enhanced during the treatment periods, whereas spontaneous proliferation was inhibited. Moreover, correlations between immune parameters and subsequent clinical responses were present in the early phase of the study. Cytotoxicity levels generated in vivo as well as the percentage of CD 56+ lymphocytes were higher in patients who responded to treatment than in non-responders. In contrast, responders had lower levels of CD 25+ cells. These findings indicate that it might be possible to select patients who are likely to benefit from prolonged immunotherapy.

Clinical applications of interleukin-2.

Interleukin-2 (IL-2) is a cytokine with potent immunomodulating properties which has shown considerable antitumour activity in preclinical models. In clinical trials, the effects of IL-2 given by various routes and schedules have been investigated. IL-2 has been administered either as single drug or in combination with other cytokines and immunomodulating agents, chemo therapeutic agents, or reinfusions of ex vivo activated autologous cytotoxic effector cells. The results of published clinical studies with IL-2 based immunotherapy are reviewed in this paper.

[Syndrome of inappropriate ADH secretion (SIADH) in small-cell bronchus carcinoma]. / Das Syndrom der inadäquaten ADH-Sekretion (SIADH) beim kleinzelligen Bronchuskarzinom.

Based upon the pertinent literature, the paraneoplastic syndrome of inappropriate antidiuretic hormone secretion (SIADH) in patients with small cell lung cancer is reviewed. Small cell lung cancer is a distinct tumor with neuroendocrine features capable of producing peptide hormones amongst which the antidiuretic hormone (ADH, arginine vasopressin) is one of the most frequent. Paraneoplastic SIADH may result from ectopic ADH production or from other tumor-related mechanisms leading to increased pituitary ADH secretion. The overt SIADH is characterized by neurological and psychiatric symptoms attributable to cerebral edema. Pooled published data suggest that the average incidence of clinically manifest SIADH in patients with newly diagnosed small cell lung cancer is 4%. Cases without clinical symptoms, detectable by laboratory tests only, are more frequent: hyponatremia, serum hypoosmolality and urine hyperosmolality are present in 14%, and an inappropriately elevated level of immunoreactive ADH in 38% of all patients respectively. Successful treatment of the underlying tumor, accompanied by a restricted fluid intake in severe cases, will usually result in prompt disappearance of the paraneoplastic SIADH. During and after the tumor treatment, plasma ADH may be useful as a tumor marker.

Phase II study with mitomycin, ifosfamide and carboplatin in inoperable non-small cell lung cancer.

In a phase II study of non-small cell lung cancer a new chemotherapy combination of mitomycin 6 mg/m2 intravenously on day 1, carboplatin 400 mg/m2 intravenously on day 1 and ifosfamide with mesna 5 g/m2 intravenously over 24 hours on day 1 was evaluated. A maximum of four chemotherapy cycles was given at intervals of 4 weeks to 34 patients with progressive, inoperable disease. 1 complete and 10 partial remissions were documented, the overall response rate being 32.4%. In a further 13 patients (38.2%) the previously progressing tumours remained stable for at least 6 weeks. The median time to progression for responding patients was 184 days. The median survival time for the whole group has not yet been reached at 293 days. A considerable but easily manageable myelosuppression was the principal toxicity despite a "no dose reduction" policy. Indeed, the dose intensity of the chemotherapy actually given was extremely close (97%) to that intended on protocol. In conclusion, the regimen is active in patients with advanced non-small cell lung cancer but requires regular haematological monitoring to prevent morbidity resulting from myelotoxicity.