A phase I study of sequential administration of escalating doses of intravenous paclitaxel, oral topotecan, and fixed-dose oral etoposide in patients with solid tumors


  • Caio M. S. Rocha Lima M.D.,

    Corresponding author
    1. Department of Medicine, University of Miami Sylvester Cancer Center, Miami, Florida
    • Department of Medicine, University of Miami Sylvester Cancer Center, 1475 NW 12th Avenue (D8-4), Suite 3310, Miami, FL 33136
    Search for more papers by this author
    • Fax: (305) 243-4905

    • Dr. Rocha Lima has served as a consultant to and received speaker's honoraria from GlaxoSmithKline, and he also is a member of the GlaxoSmithKline Board of Speakers. In addition, Dr. Rocha Lima has served as a consultant to and received speaker's honoraria from Amgen Inc.

  • Carlo V. Catapano M.D., Ph.D.,

    1. Laboratory of Experimental Oncology, Oncology Institute of Southern Switzerland, Bellinzona, Switzerland
    Search for more papers by this author
  • Daniel Pacheco M.D.,

    1. Hollings Cancer Center, Medical University of South Carolina, Charleston, South Carolina
    Search for more papers by this author
  • Carol A. Sherman M.D.,

    1. Hollings Cancer Center, Medical University of South Carolina, Charleston, South Carolina
    2. Division of Hematology Oncology, Department of Medicine, Medical University of South Carolina, Charleston, South Carolina
    Search for more papers by this author
  • Greg Oakhill M.D.,

    1. Highland Oncology Group, Fayetteville, Arkansas
    Search for more papers by this author
  • Chaudhry Mushtaq M.D.,

    1. South Carolina Oncology Associates, West Columbia, South Carolina
    Search for more papers by this author
  • Kimberly D. Freeman,

    1. GlaxoSmithKline, Collegeville, Pennsylvania
    Search for more papers by this author
  • Mark R. Green M.D.

    1. Hollings Cancer Center, Medical University of South Carolina, Charleston, South Carolina
    2. Division of Hematology Oncology, Department of Medicine, Medical University of South Carolina, Charleston, South Carolina
    Search for more papers by this author



Based on preclinical findings and on the clinical antitumor efficacy of sequential paclitaxel/topotecan and topotecan/etoposide, the authors sought to define the maximum tolerated doses (MTDs) and dose-limiting toxicities (DLTs) associated with a sequential combination of paclitaxel, topotecan, and etoposide in patients with solid tumors.


The MTDs were determined through standard dose escalation in cohorts of three patients. Patients with refractory solid tumors and performance status ≤ 2 were treated with intravenous paclitaxel 50–110 mg/m2 (Day 1), oral topotecan 0.5–2.0 mg/m2 (Days 2–4), and oral etoposide 160 mg/m2 (Days 5–7) during every 21-day cycle. For dose-limiting neutropenia, granulocyte–colony-stimulating factor (G-CSF) was administered on Day 8 in subsequent cohorts. Blood samples were obtained before treatment during Cycle 1 (Days 1, 2, and 5) for topoisomerase I assessment.


Twenty-eight patients received a combined total of 129 cycles. The MTDs were paclitaxel 80 mg/m2, topotecan 1.5 mg/m2, and etoposide 160 mg/m2 without G-CSF. In minimally pretreated patients, G-CSF allowed paclitaxel dose escalation to 110 mg/m2. Three patients (11%) had radiologic partial responses, and 4 patients (14%) had stable disease. Day 2 topoisomerase I levels increased by 2–15 times relative to baseline levels in 7 of 14 patients analyzed (50%).


The novel sequential combination that was evaluated generally was well tolerated and active in patients with refractory solid tumors. Based on hematologic DLTs, the authors recommend further evaluation of paclitaxel 110 mg/m2, topotecan 1.5 mg/m2, and etoposide 160 mg/m2 with G-CSF support in minimally pretreated patients. Cancer 2004. © 2004 American Cancer Society.

Combinations of chemotherapeutic agents are used to increase therapeutic efficacy against advanced or aggressive malignancies. Because tumor cells often exhibit cross-resistance to agents with similar mechanisms of action, previous clinical trials have combined chemotherapeutic agents based on nonoverlapping toxicities and mechanisms of action.1 Although most clinical regimens have combined agents simultaneously, recent preclinical reports suggest that the sequential application of specific chemotherapeutic agents can have greater-than-additive antitumor effects.2, 3 For example, in vitro studies have demonstrated that pretreatment with taxanes and other agents that interfere with microtubules can increase the cytotoxicity of topoisomerase I inhibitors by increasing both topoisomerase I levels and the fraction of cells in S phase.2 Madden et al.4 demonstrated that preexposure of the MCF-7 and MDAH B231 human breast carcinoma cell lines to paclitaxel or docetaxel caused nearly 10-fold decreases in the 50%-inhibitory concentrations of topoisomerase I inhibitors (9-aminocamptothecin and topotecan). A 2-fold increase in topoisomerase I RNA was demonstrated 24 hours after the cells were exposed to taxanes.4 In both studies, neither simultaneous nor reverse-sequence exposures were synergistic.2, 4

Topoisomerase I inhibitors induce an increase in cellular topoisomerase II levels in human tumor cells and heighten their sensitivity to topoisomerase II inhibitors.3, 5–7 In one report, topoisomerase II was induced detectably after 2 days of topotecan treatment, reached a maximum after 4 days of exposure, and decreased to baseline 5 days after withdrawal of topotecan.3 Furthermore, preclinical data suggest that concurrent administration of topoisomerase I and II inhibitors is antagonistic.8

Thus, the sequential administration of a taxane, a topoisomerase I inhibitor, and a topoisomerase II inhibitor may optimize antitumor synergy, as each agent in the sequence may increase the sensitivity of the tumor to each subsequently administered agent. This synergy is the basis for the rational design of a novel regimen that applies the sequential combination of the taxane paclitaxel (Taxol; Bristol-Myers Squibb, Princeton, NJ), the topoisomerase I inhibitor topotecan (HYCAMTIN; GlaxoSmithKline, Philadelphia, PA), and the topoisomerase II inhibitor etoposide (VePesid; Bristol-Myers Squibb) [ETopoTax] to treat patients with treatment-refractory solid tumors or solid tumors for which no treatment of proven efficacy is available. The use of this combination is supported further by the different mechanisms of action and the lack of cross-resistance of tumor cells to these agents.1, 6, 9–11 Finally, the sequential combination of paclitaxel, topotecan, and etoposide may overcome resistance to individual chemotherapy agents and has the potential for greater efficacy, because treatment with each of these agents may cause an increase in tumor sensitivity to next agent administered.

Based on preclinical findings and on the antitumor efficacy of paclitaxel/topotecan12, 13 and topotecan/etoposide,14–16 which have been investigated as sequential combination therapies in recent clinical trials, we sought to investigate the feasibility of a sequential three-drug regimen. The treatment regimen consisted of intravenous (i.v.) paclitaxel on Day 1, followed 24 hours later by oral topotecan for 3 consecutive days, and then by oral etoposide for 3 additional days. This regimen was designed to increase the synergy that occurs between these agents, thereby maximizing the impact of chemotherapy on tumor cells. In the current trial, we also investigated the expression patterns of topoisomerase I in treated patients to determine whether preclinical observations of treatment effects can be demonstrated in vivo.


The primary objective of the study was to determine the tolerability and maximum tolerated dose (MTD) associated with i.v. paclitaxel and oral topotecan in sequential combination with a fixed dose of oral etoposide (160 mg/m2). In previous studies, severe neutropenia and thrombocytopenia occurred when dose escalations of topotecan were attempted in conjunction with administration of 135 mg/m2 paclitaxel.17 The current trial began with decreased doses of paclitaxel and involved escalation of both paclitaxel and topotecan doses with the etoposide dose held constant. The use of granulocyte–colony-stimulating factor (G-CSF; Neupogen; Amgen Inc., Thousand Oaks, CA) to manage dose-limiting neutropenia and to improve the tolerability of this treatment regimen also was investigated.

Patient Selection

Patients age ≥ 18 years with histologically documented solid tumors that were refractory to standard therapy or for which no therapy of proven efficacy was available were eligible. Patients were required to have measurable or evaluable disease; an Eastern Cooperative Oncology Group performance status ≤ 2; and adequate bone marrow, renal, and hepatic function (absolute neutrophil count ≥ 1500 cells/mm3, platelet counts ≥ 100,000/mm3, creatinine clearance > 60 mL/minute, aspartate aminotransferase and alanine aminotransferase levels ≤ 2 times the upper limit of normal, and serum bilirubin levels ≤ 1.5 mg/dL). Patients who had recent evidence of heart disease; a known allergy to topotecan, paclitaxel, or etoposide; previous radiation to the pelvis; or a preexisting gastrointestinal illness that could impair absorption of topotecan or etoposide were excluded, as were patients who were pregnant. The protocol was approved by the Institutional Review Board at the Medical University of South Carolina (Charleston, SC). All patients provided written informed consent.

Study Design and Treatments

The current study was an open-label, single-center, Phase I dose-escalation study. Oral etoposide was administered at a constant dose of 160 mg/m2. Paclitaxel and topotecan doses were escalated in consecutive cohorts according to the dose levels listed in Table 1. Dose-limiting toxicity (DLT) was defined (during Cycle 1 only) as Grade 4 neutropenia that lasted ≥ 5 days or was accompanied by fever (≥ 38.5 °C), Grade 4 thrombocytopenia or platelet transfusion, or any Grade ≥ 3 nonhematologic toxicity (except for alopecia and untreated nausea/emesis). Any toxicity that delayed Cycle 2 by > 14 days was also considered a DLT. Three patients began treatment at each dose level; if one patient experienced DLT, then up to three additional patients were evaluated at that dose level. If one or more additional patients also experienced DLT, then escalation was halted, and the previous dose level was considered the MTD. If the initial DLT was neutropenia, then accrual at that same dose level was planned with the addition of G-CSF support, and further dose escalation was attempted. The MTD was defined as the dose level immediately below the dose level at which 2 of the first 3 patients in any cohort or 2 or more of 6 patients in any expanded cohort experienced DLT.

Table 1. Dosing Levels
Dose levelPaclitaxel i.v. on Day 1 (mg/m2)Oral topotecan on Days 2–4 (mg/m2)Oral etoposide on Days 5–7 (mg/m2)
  1. i.v.: intravenous; G-CSF: granulocyte–colony-stimulating factor.

5 + G-CSF1101.5160
6 + G-CSF1102.0160

Paclitaxel 50–110 mg/m2 was administered by i.v. infusion over 3 hours on Day 1 of each 21-day treatment cycle. Standard premedication (dexamethasone 20 mg; diphenhydramine 50 mg; and cimetidine 300 mg, ranitidine 50 mg, or famotidine 20 mg) was administered 30 minutes before paclitaxel. Oral topotecan 0.5–2.0 mg/m2 was administered on Days 2–4 and was followed by a fixed dose of oral etoposide 160 mg/m2 on Days 5–7. When dose-limiting neutropenia was encountered, G-CSF (5 μg/kg per day on Day 8 of each treatment cycle until the absolute neutrophil count was ≥ 10,000 cells/μL or ≥ 5000 cells/μL on 2 consecutive measurements made 3 days apart) was added to the therapeutic regimen for subsequent cohorts.

Assessment of Safety and Tolerability

Safety and tolerability were evaluated by clinical laboratory assessments. Adverse events were graded according to the National Cancer Institute Common Toxicity Criteria (CTC). For DLT and MTD determinations, toxicity was expressed as the worst CTC grade experienced during Cycle 1. Comprehensive metabolic panels (liver and renal function tests plus electrolytes) were performed on Days 1 and 8 of Cycle 1. Complete blood counts with differential were evaluated on Days 1, 8, 11, 14, and 17 during Cycle 1. For all subsequent cycles, complete metabolic panels and complete blood counts with differential were performed on Day 1. Additional laboratory studies were performed as indicated clinically. Patients who experienced a DLT were allowed to continue treatment at the next lower dose level if they did not have progressive disease. Serious adverse events included Grade 4 hematologic toxicity and Grade ≥ 3 nonhematologic toxicity.

Tumor Assessment

Measurable disease was defined as a detectable tumor that had defined margins and measured ≥ 2 cm in greatest dimension or a skin lesion that measured ≥ 1 cm in greatest dimension. Tumors that did not meet these criteria, as well as cases of palpable hepatomegaly with no discrete liver nodules on computed tomography scanning, were considered evaluable disease. Nonmeasurable, nonevaluable disease included ascites, pleural effusion, pericardial effusion, bone or bone marrow metastases, leptomeningeal metastases, lymphangitic tumor, and previously irradiated lesions. Standard response criteria were employed. In brief, a complete response (CR) was defined as the disappearance of all measurable and evaluable disease for ≥ 4 weeks; a partial response (PR) was defined as a decrease of ≥ 50% in the total area of all lesions with no increase in evaluable disease for ≥ 4 weeks; progressive disease (PD) was defined as an increase of ≥ 25% in the area of any tumor, the worsening of evaluable disease, or the appearance of new lesions or metastases; and stable disease (SD) was defined as any response other than CR, PR, or PD. In patients with measurable or evaluable disease, tumor response was evaluated every two cycles. Potential responders were reevaluated after ≥ 4 weeks.

Measurement of Topoisomerase I

Blood samples were collected from 14 patients on Days 1, 2, and 5 of Cycle 1. For 3 patients, blood samples were also obtained during Cycle 2. Peripheral blood lymphocytes (PBLs) were isolated immediately using Ficoll-Paque (Amersham Biosciences Corp., Piscataway, NJ), as described previously.18 Cell lysis and immunoblotting were performed as described previously.19–21 Under the conditions used in the current study, complete dissociation of any topoisomerase I–DNA complex would occur during sample preparation. Therefore, total cellular topoisomerase I was detected in the immunoblots.2 A protein sample (50–100 μg) was separated electrophoretically, transferred to nitrocellulose, and assessed with human topoisomerase I antiserum (TopoGEN, Columbus, OH) or lamin B antibody (Oncogene Science, Cambridge, MA).2 Lamin B was used as an internal control for protein loading, and a T-lymphoblastoid cell line (CEM; American Type Culture Collection, Manassas, VA) was used as a positive control for topoisomerase. Blots were developed by enhanced chemiluminescence (Amersham Biosciences Corp.) and analyzed with Gel-Pro Analyzer software (Media Cybernetics, Silver Spring, MD).2 Immunoblotting was repeated at least twice for each patient.


Patient Characteristics

Between November 1998 and October 1999, 28 patients were enrolled in the current study at the Hollings Cancer Center, Medical University of South Carolina. Nineteen patients were accrued before the protocol was amended to include G-CSF. Nine other patients received treatment with G-CSF support according to the protocol. Patient demographics and baseline disease characteristics are shown in Table 2. The average patient age was 54.9 years (range, 28–77 years). Most patients had good performance status. All but four patients previously had received chemotherapy. Patients with lung carcinoma accounted for 43% of the overall treatment population, in which a variety of solid tumors were represented. The tumor types of patients enrolled at each dose level are listed in Table 3.

Table 2. Patient Demographics and Baseline Disease Characteristics
CharacteristicNo. of patients (%)
  1. NSCLC: nonsmall cell lung carcinoma; SCLC: small cell lung carcinoma; RCC: renal cell carcinoma; ECOG: Eastern Cooperative Oncology Group.

No. of patients28 (100)
 Male8 (29)
 Female20 (71)
Age (yrs) 
 ≤ 403 (11)
 41–6420 (71)
 ≥ 655 (18)
Primary tumor type 
 NSCLC (including bronchoalveolar)9 (32)
 SCLC3 (11)
 Esophageal3 (11)
 RCC2 (7)
 Pancreatic2 (7)
 Other7 (25)
 Unknown primary2 (7)
Maximum lesion diameter (cm) 
 < 23 (11)
 2 to < 58 (29)
 5–108 (29)
 > 104 (14)
 Nonmeasurable/data unavailable5 (18)
ECOG performance status 
 05 (18)
 117 (61)
 26 (21)
Previous therapy28 (100)
 Chemotherapy24 (86)
 Immunotherapy1 (4)
 Radiation17 (61)
Table 3. Dose Escalation and Tumor Types
Dose levelNo. of patientsTumor types (no.)
  • NSCLC: nonsmall cell lung carcinoma; RCC: renal cell carcinoma; SCLC: small cell lung carcinoma; G-CSF: granulocyte–colony-stimulating factor.

  • a

    Patients were treated again at Dose Level 5 with G-CSF to confirm the maximum tolerated dose.

13Skin (1), bronchoalveolar (2)
23Head and neck (1), NSCLC (2)
33Colorectal (1), bronchoalveolar (1), unknown primary (1)
44NSCLC (1), esophageal (1), RCC (1), pancreatic (1)
56Liposarcoma (1), SCLC (1), small bowel (1), NSCLC (1), esophageal (1), bronchoalveolar (1)
5 + G-CSF3Pancreatic (1), laryngeal (1), unknown primary (1)
6 + G-CSF3Spinal cord (1), NSCLC (1), SCLC (1)
5 + G-CSFa3RCC (1), SCLC (1), esophageal (1)

Two patients were not evaluable for safety during Cycle 1. At Dose Level 4, 1 patient died of an unrelated event during Cycle 1. At Dose Level 5, one patient received both oral topotecan and etoposide concomitantly on Day 2 of Cycle 1. Although no excessive toxicity was noted, this patient was not considered evaluable for safety because of the protocol violation. During Cycle 2, the medication was administered correctly and was tolerated well.

Hematologic Toxicity and Determination of the MTD

Twenty-seven patients were evaluable for toxicity in one or more cycles. Grade 3–4 hematologic toxicities are summarized in Table 4. Only three patients required study medication delays in the absence of a DLT. At Dose Level 1, one patient required a 1-week delay in Cycle 2 because of Grade 3 neutropenia on Day 22. In this patient, Grade 3 leukopenia, neutropenia, anemia, and infection occurred after Cycle 2. One patient receiving Dose Level 3 also required a 1-week delay in Cycle 2 because of Grade 3 neutropenia. At Dose Level 4, one patient required a 1-week delay in Cycle 2 because of delayed recovery from Grade 4 neutropenia and leukopenia, Grade 3 thrombocytopenia, and mild fatigue, malaise, and weakness during Cycle 1. All three patients who required dose delays had been treated previously with both radiation and chemotherapy.

Table 4. Grade 3/4 Hematologic Toxicity during Cycle 1 in Safety-Evaluable Patients
Dose levelNo. of patientsToxicity (no. of patients)
  1. G-CSF: granulocyte–colony-stimulating factor.

5 + G-CSF64312
6 + G-CSF32221
All patients26141254

At Dose Level 5, three of 5 evaluable patients had Grade 4 neutropenia for > 5 days (for 6 days, 8 days, and 8 days, respectively): all had been pretreated heavily. Therefore, without G-CSF support, Dose Level 4 represented the MTD, because ≥ 2 patients experienced DLT. Three patients were then treated at Dose Level 5 with G-CSF; none experienced DLT. Further dose escalation was not possible due to hematologic DLTs in two of three patients treated at Dose Level 6 with G-CSF; one patient had neutropenia, and the other had both neutropenia and thrombocytopenia. Therefore, three additional patients were enrolled at Dose Level 5 with G-CSF. Two of those confirming patients also developed DLTs, including one who had been treated with radiotherapy and carboplatin22 and developed Grade 4 thrombocytopenia and Grade 4 neutropenia that lasted 8 days. The second patient in this cohort developed Grade 4 neutropenia that lasted 6 days, whereas the third patient did not experience DLT. Therefore, the MTD in this heterogeneous, advanced cancer population was defined as Dose Level 4, and G-CSF did not allow further dose escalation. However, among patients who received G-CSF at Dose Level 5 and who had not previously undergone any therapy, myelosuppressive DLT did not occur. Therefore, Dose Level 5 with G-CSF is considered the recommended Phase II dose level for previously untreated patients.

Hematologic toxicities were noncumulative and reversible. The nadir neutrophil counts (typically occurring at ∼Day 14 of therapy) and platelet counts (typically occurring at ∼Day 17 of therapy) did not decrease during the first 4 courses of therapy, even in patients who were treated at higher dose levels (data not shown).

Nonhematologic Toxicities

Nonhematologic toxicities generally were mild and did not limit dose escalation. Serious nonhematologic toxicities during Cycle 1 were related primarily to disease progression. Nonhematologic toxicities included an episode of Grade 4 nausea, emesis, and dehydration and an episode of Grade 3 nausea in patients who were not medicated optimally with antiemetics. Both patients improved after optimization of antiemetic therapy. During the first administration of paclitaxel, one patient experienced transient chest tightness that did not recur during subsequent cycles. Low-grade nausea occurred in six patients, and low-grade emesis occurred in seven patients. There were ≤ 2 episodes each of ≤ Grade 2 diarrhea, loss of appetite, rash on the lower extremities, and fatigue/malaise.

Tumor Response

All 28 patients had measurable or evaluable disease, and 25 patients were evaluable for response. The best response data for each patient (intent-to-treat population) are summarized in Table 5. Three patients (11%) achieved a PR. Two PRs were reported at Dose Level 5 (with or without G-CSF) in patients with previously treated small cell lung carcinoma (SCLC). One patient had SCLC that was refractory to carboplatin/etoposide, whereas the other patient had SCLC that progressed in the chest and brain after two separate platinum-based treatments (cisplatin/etoposide and carboplatin/etoposide/paclitaxel) and radiotherapy to the chest and brain. Receiving the study regimen, this patient had a CR in the brain and a durable PR in the chest that lasted > 1 year. At this dose level, a patient with esophageal carcinosarcoma that had recurred after radiotherapy and chemotherapy had a minimal response (a decrease of > 25% but < 50% [scored as SD]). The third partial responder (Dose Level 1) had recurrent Stage IV bronchoalveolar carcinoma after treatment with paclitaxel/carboplatin, gemcitabine, and vinorelbine. In addition, at Dose Level 3, one patient with a metastatic neuroendocrine tumor of unknown origin and bone marrow metastases had a decrease in bone marrow tumor burden after Cycle 2, but increased tumor burden was detected late in Cycle 4. This patient had been treated previously with etoposide, carboplatin, and ifosfamide. Four patients (14%), including the patient who had a minimal response, had SD for ≥ 4 cycles. All remaining patients (64%) experienced PD within the first 4 cycles.

Table 5. Antitumor Response Summary
Response to treatmentNo. of patients (%)
  • a

    Includes 1 patient who had a minimal response (< 50% reduction in tumor mass).

No. of patients28 (100)
Overall response3 (11)
 Complete response0 (0)
 Partial response3 (11)
No response22 (79)
Stable disease4a (14)
Not evaluable3 (11)
Progressive disease18 (64)

Topoisomerase I Levels

Immunoreactive bands corresponding to full-length topoisomerase I and lamin B (internal control) were detected in all PBL samples (n = 14). Bands of identical size were detected in CEM cell lysates. Topoisomerase I exhibited a general upward trend after paclitaxel therapy, as predicted by experiments in human tumor cell lines.2 A representative blot is shown in Figure 1. In this patient, topoisomerase I was barely detectable in PBLs at Day 1 (before treatment) but increased on Day 2 (after paclitaxel) and decreased slightly on Day 5 (after topotecan). This patient had a durable PR while receiving the protocol. The other two responders did not have their PBLs collected. Figure 2 shows topoisomerase I levels in all 14 patients at Days 1, 2, and 5 of Cycle 1. Increases in topoisomerase I levels at Day 2 were detected in 9 patients (Fig. 2A), and no increase was detected in 5 patients (Fig. 2B). The data are presented graphically as relative changes in topoisomerase I levels in all 14 patients after the administration of paclitaxel on Day 2 (Fig. 3A) and topotecan on Day 5 (Fig. 3B). Seven patients had substantial (2–15-fold) increases in topoisomerase I levels, and 2 patients had minor increases(≤ 50%) on Day 2. The remaining 5 patients had no changes or a slight decrease in topoisomerase I at Day 2. The increase in topoisomerase I was largely reversed after topotecan treatment in the majority of patients (Fig. 3B), probably because of drug-induced DNA binding and subsequent protein degradation. This effect has been described previously both in vitro and in clinical samples.23, 24 However, topoisomerase I levels remained above baseline on Day 5 in some patients. In addition, a 2–5-fold increase was observed on Day 5 in all 5 patients who had exhibited no change or who had a decrease in topoisomerase I at Day 2.

Figure 1.

Representative blot showing topoisomerase I levels in peripheral blood lymphocytes before paclitaxel (Day 1) and after paclitaxel (Day 2) and topotecan (Day 5). A T-lymphoblastoid cell line (CEM) was used as a positive control for topoisomerase I, and lamin B was used as an internal protein-level control.

Figure 2.

Levels of topoisomerase I in peripheral blood lymphocytes from 14 patients (by pattern) on Treatment Days 1, 2, and 5 of Cycle 1. Protein levels were determined by Western blot analysis, which was followed by densitometric analysis of the blots. There were (A) 9 patients who had increased topoisomerase I levels detected on Day 2 and (B) 5 patients who did not have increased in topoisomerase I levels detected on Day 2. Each symbol represents an individual patient.

Figure 3.

Changes in topoisomerase I levels in peripheral blood lymphocytes from 14 patients (A) between Treatment Days 1 and 2 and (B) between Treatment Days 2 and 5. Diamonds represent patients with increased topoisomerase I levels after paclitaxel treatment, and circles represent patients with decreased or unchanged topoisomerase I levels.


Combination therapies are limited in the sense that concomitant administration of agents with overlapping hematologic toxicities can produce unexpectedly severe responses.25, 26 Paclitaxel, topotecan, and etoposide all are associated individually with myelosuppression, and their combination may produce severe hematologic toxicity. However, doublets of paclitaxel plus topotecan,27 topotecan plus etoposide,14, 16 and paclitaxel plus etoposide28, 29 were used successfully in previous trials. The activity and toxicity profiles reported in those trials suggest that an effective regimen with manageable toxicity can be established for the sequential administration of paclitaxel, topotecan, and etoposide.

The triplet regimen investigated in the current study was designed rationally, based on both preclinical and clinical trial data, and the objective was to maximize antitumor synergies among etoposide, topotecan, and paclitaxel. The objective of the current Phase I clinical trial was to determine the feasibility, DLT, and MTD associated with sequential paclitaxel, topotecan, and etoposide, both with and without G-CSF. Without G-CSF, Dose Level 4 represented the MTD. In previously treated patients, neutropenia and thrombocytopenia became dose limiting when further dose escalation was attempted.

The use of G-CSF to manage hematologic toxicities allowed dose escalation with paclitaxel/topotecan and paclitaxel/etoposide regimens in previous clinical trials.27, 29 In the current study, G-CSF did not increase the study-defined MTD for previously treated patients, because hematologic DLTs were not managed effectively in the confirmation cohort receiving Dose Level 5 with G-CSF. In theory, because G-CSF was administered relatively late in the cycle, treatment may not have been optimal for decreasing the severity of neutrophil nadirs. However, in the initial cohort of three patients who were treated with G-CSF at this dose level, which included one previously untreated patient and one patient who had previously received only four cycles of irinotecan/gemcitabine and no prior radiotherapy, no DLTs occurred. Furthermore, several patients continued treatment and received subsequent cycles of therapy at Dose Level 5 with G-CSF without encountering severe or cumulative toxicities.

Variations in the tolerability of Dose Level 5 with G-CSF appeared to be related to the heterogeneity in treatment history. The confirmation cohort receiving Dose Level 5 with G-CSF included a patient who had received previous chemotherapy with carboplatin/etoposide and radiotherapy, both of which can deplete bone marrow reserves.22, 30 Furthermore, the actual durations of DLT-defining Grade 4 neutropenia in the 2 patients from the confirmation cohort receiving Dose Level 5 with G-CSF were 6 days and 8 days, and cumulative toxicity did not occur. These findings suggest that Dose Level 5 with G-CSF is the appropriate recommended Phase II dose in chemotherapy-naive and radiotherapy-naive patients. Indeed, preliminary evidence from CALGB study 30002, an ongoing, multicenter, Phase II cooperative-group trial of this regimen in chemotherapy-naive patients with limited-stage SCLC, substantiates the safety of Dose Level 5 with G-CSF. Among the first 21 patients for whom study data were reported, there has been no Grade 4 febrile neutropenia or thrombocytopenia, a 29% incidence of Grade 4 neutropenia, a 14% incidence of Grade 3 febrile neutropenia, and a 5% incidence of Grade 3 thrombocytopenia (unpublished data).

Nonhematologic toxicities encountered during the current dose-escalation study generally were mild. Low-grade nausea occurred in 21% of patients and generally responded to antiemetics. Patient compliance with the study regimen was high. No patient treated at or below the MTD withdrew consent. Patient convenience associated with this three-drug regimen was notable, because only one i.v. infusion per cycle was required (paclitaxel on Day 1). Both topotecan and etoposide were then taken orally at home. The combination of manageable hematologic toxicity, limited and mild incidence of nonhematologic toxicities, ease of drug delivery, and evidence of activity in the Phase I setting makes this regimen appealing for further study.

The rationale for the current sequential combination therapy regimen is based on the hypothesis that changes in topoisomerase levels in response to therapeutic agents may contribute to treatment synergy. Indeed, ≥ 2-fold increases in topoisomerase I levels after administration of paclitaxel were detected in 50% of patients. It is unlikely that these changes represent random fluctuations in topoisomerase I levels, because similar patterns were observed during the second cycle of therapy in the three patients from whom blood samples were obtained. Tumor biopsy samples were not available for intratumoral assessments. However, pretreatment of tumor cells in vitro with microtubule-interfering agents, such as paclitaxel and vinblastine, induced topoisomerase I increases with kinetics similar to what was observed in PBLs,2 and this increase was associated with greater sensitivity to topotecan. Paclitaxel-induced increases in topoisomerase I levels may be amplified further in tumor cells, because elevated basal topoisomerase I levels have been detected in human tumor xenografts.31 In addition, topotecan pretreatment increases the sensitivity of tumor cells to subsequent etoposide treatments, generating a therapeutic synergism of antitumor effects.32–36 The most likely explanation for this synergy involves the up-regulation of topoisomerase II (the cellular target of etoposide) after the inhibition of topoisomerase I by topotecan.36, 37

In the current trial, etoposide was administered at full single-agent doses (160 mg/m2 per day × 3 days), and the doses of paclitaxel and topotecan were escalated in alternating fashion. This strategy was used because the dose intensity of oral etoposide has been well established in patients with SCLC,38 an important target population for the study regimen. The MTD of oral topotecan in the current regimen was well below the single-agent dose (2.3 mg/m2 per day on Days 1–5 of a 21-day cycle) that was recommended by a Phase II trial.39 This difference may have been caused by the overlapping hematologic toxicities of the study agents or by the novel dosing regimen. However, for i.v. topotecan, it has been shown that lower doses and shorter durations of therapy are effective for patients with SCLC or ovarian malignancy.40 In fact, using this schedule, topotecan doses as low as 1 mg/m2 per day have demonstrated single-agent activity in patients with recurrent SCLC, with an overall relative risk of 26% (including 1 CR), a median survival of 35 weeks, and a 1-year survival rate of 32% in a recent Phase II study (n = 105).41 Moreover, combining agents with synergistic antitumor activity may allow administration of lower doses with no loss in efficacy.

Although no firm efficacy conclusions can be drawn from this limited Phase I trial, the activity of the current regimen in patients with late-stage disease is encouraging. All three patients with PRs had been pretreated heavily with chemotherapy with or without radiotherapy. All either had treatment-refractory disease or developed recurrent disease after therapy. The activity in patients with SCLC is promising, and Phase II testing in patients with SCLC or with malignancies that often are sensitive to these chemotherapy agents is warranted.

Sequential paclitaxel, topotecan, and etoposide was an easily administered and well tolerated treatment regimen. No unexpected toxicities occurred, and patient compliance was high. Based on the defined DLTs in the current study, the recommended dosing regimen for Phase II trials is paclitaxel 80 mg/m2 with topotecan 1.5 mg/m2 and etoposide 160 mg/m2. However, in patients with little or no previous exposure to chemotherapy and radiotherapy, paclitaxel 110 mg/m2, topotecan 1.5 mg/m2, and etoposide 160 mg/m2, together with G-CSF, appears to represent the MTD. Indeed, the CALGB already is evaluating this dose level with G-CSF support for use as induction therapy followed by carboplatin, etoposide, and thoracic radiotherapy for previously untreated patients with limited-stage SCLC.


The authors thank Marcia Skoglund (GlaxoSmithKline, Philadelphia, PA) for assisting with the statistical analysis and Michel Martin (Charleston Cancer Center, Charleston, SC) for assisting with data collection and management.