Gemcitabine and docetaxel have significant activity as single agents or in combination with other agents in a range of malignancies, including nonsmall cell lung carcinoma (NSCLC),1–4 breast carcinoma,5, 6 pancreatic carcinoma,7, 8 bladder carcinoma,9, 10 and ovarian carcinoma.11, 12 The two agents have different cellular targets (DNA synthesis for gemcitabine, microtubules for docetaxel) and act at different phases of the cell cycle (S-phase for gemcitabine, mitosis for docetaxel).13, 14 Because of their activity as single agents, their different mechanisms of action, and their nonoverlapping toxicities, the combination of docetaxel and gemcitabine has been evaluated using a variety of treatment schedules.15–21 Activity was documented with all of these regimens, and myelosuppression was the most common dose-limiting toxicity.15–21
The weekly schedule may improve the therapeutic index of docetaxel compared with the 3-week schedule by causing less myelosuppression.22–24 Two published Phase II studies of docetaxel 36 mg/m2 administered weekly to patients with NSCLC reported a 6–8% rate of Grade 3 or 4 neutropenia or leukopenia,25, 26 compared with the 65.3% rate reported with the 3-week schedule.27 This favorable toxicity profile made that schedule the subject of further investigation of combination regimens. We undertook this trial to determine the toxicity profile and the maximum tolerated dose (MTD) of the weekly docetaxel and gemcitabine combination.
MATERIALS AND METHODS
Patients with histologically or cytologically proven malignant disease that was refractory to standard therapy and who were treated with only one or two previous chemotherapy regimens were enrolled into the study. Other eligibility criteria included age 18–75 years, an Eastern Cooperative Oncology Group (ECOG) performance status of 0 or 1, adequate hepatic function (total bilirubin ≤ 1.5 mg/dL, transaminases ≤ 2.5 times the institutional upper limit of normal [ULN], and alkaline phosphatase [ALP] ≤ 4 times ULN), adequate renal function (serum creatinine ≤ 1.5 mg/dL or measured/calculated creatinine clearance ≤ 60 mL/minute), adequate hematologic status (white blood cell count ≥ 3000/μL, absolute neutrophil count (ANC) ≥ 1000/μL, hemoglobin ≥ 9.0 gm/dL, and platelet count ≥ 100,000/μL), and measurable or evaluable disease. Exclusion criteria included prior therapy with gemcitabine or docetaxel; radiotherapy to major bone marrow areas within 4 weeks of entry; prior bone marrow transplantation or stem cell support; therapy with investigational drugs within 4 weeks of study entry; pregnant or currently lactating; refusal to use effective contraception while on study if female of child-bearing potential; inability to give informed consent; medically unstable; psychological, social, and/or geographic issues not permitting weekly medical follow-up or compliance with the study protocol; and history of hypersensitivity reaction to products containing polysorbate 80 (Tween 80). All inclusion and exclusion criteria were assessed within 14 days prior to the initiation of therapy, with the exception of hematologic and biochemical studies, which were obtained within 7 days prior to the initiation of treatment. All patients were required to give a written informed consent in accordance with U.S. Food and Drug Administration guidelines and institutional guidelines.
Therapy was administered weekly for 3 weeks, and cycles were repeated every 28 days. Each cohort at a given dose level completed one cycle of treatment before patients were entered at the next dose level. Each cycle consisted of intravenous (IV) infusion of docetaxel over 1 hour followed by a 2-hour treatment interval prior to the administration of gemcitabine on Days 1, 8, and 15. The 2-hour delay between docetaxel and gemcitabine administration was adopted initially by the ECOG secondary to a report from Europe of pulmonary toxicity observed with docetaxel and gemcitabine combination regimens.28 Prior to the administration of docetaxel, all patient received premedication with diphenhydramine 25 mg IV and a single dose of dexamethasone 20 mg IV (rather than the 3–5 doses of twice daily dexamethasone 8 mg that were administered in other weekly trials).24, 26, 29 The weekly docetaxel dose in our study ranged from 20 mg/m2 to 36 mg/m2. The weekly dosage of gemcitabine ranged from 400 mg/m2 to 800 mg/m2. Body surface area was calculated based on the actual body weight and was not increased for weight gain during this study but was decreased for weight loss. In the presence of an objective response or stable disease, patients continued treatment cycles until disease progression or unacceptable toxicity developed.
Patients were evaluated weekly for the first cycle, every 4 weeks (correlating with the end of each additional cycle), and at the end of the study. Weekly evaluations consisted of a toxicity check, physical examination and laboratory evaluation (complete blood count, urinalysis, chemistry profile (serum creatinine, aspartate aminotransferase [AST], alanine aminotransferase [ALT], bilirubin, and ALP), and coagulation parameters (prothrombin time and activated partial thrombolplastin time). Assessment of tumor response was made by objective, two-dimensional measurement of evaluable tumor according to World Health Organization criteria30 employing imaging studies (computed tomography scans or magnetic resonance imaging) within 14 days of the initiation treatment and every 8 weeks thereafter. Criteria for initiating additional cycles included responsive or stable disease, ANC ≥ 1000/μL, platelets ≥ 75,000/μL, and resolution of all Grade ≥ 2 toxicity. If the ANC or platelet count remained below the specified values for > 4 weeks, then the patient was withdrawn from study.
Three to six patients were treated sequentially at each dose level (Table 1). If dose-limiting toxicity occurred in one of three patients, then three additional patients were to be enrolled at that dose level. The MTD was defined as the dose level below the dose that produced unacceptable toxicity during Cycle I in at least two of three patients or in at least three of six patients. Additional patients were to be accrued at the MTD to further characterize the toxicity profile of the regimen. Toxicity was graded employing the National Cancer Institute Common Toxicity Criteria (version 2.0).31 Although dose escalation proceeded until the MTD was achieved, chronic and late toxicities were observed and graded carefully. Unacceptable toxicity was defined as the occurrence of any of the following: ≥ Grade 3 nonhematologic toxicity, Grade 4 neutropenia for ≥ 7 days or associated with fever or infection, Grade 4 thrombocytopenia, or any Grade 5 toxicity. Febrile neutropenia was defined as an oral temperature ≥ 38.0 °C on 3 readings in a 24-hour period or 1 reading ≥ 38.5 °C in the absence of antipyretics.
Table 1. Dosing Schema of Weekly Intravenous Administration of Docetaxel and Gemcitabine
Dose modifications were made on the basis of physical examination, blood counts, and chemistry profile on the days of scheduled treatment. Docetaxel and gemcitabine were withheld in patients with an ANC < 1000/μL or platelet counts < 75,000/μL. Treatment was resumed at the preceding dose level after patients recovered to the specified levels. Patients who developed abnormal liver function tests were managed as follows: If the AST and/or ALT levels were between 1.5 and 5.0 × ULN, or if the ALP level was between 2.5–5.0 × ULN, then treatment was administered at the preceding dose level. If AST, ALT, or ALP levels were > 5.0 × ULN, or if the bilirubin level was > ULN, then treatment was withheld until these values returned to specified levels; then, treatment was resumed at the preceding dose level. In the presence of other Grade 2 toxicity, continuation of the weekly administration of docetaxel and gemcitabine was at the discretion of the principal investigator. Any patient who had nonhematologic toxicity ≥ Grade 3 had the treatment withheld until the resolution of toxicity to ≤ Grade 1; treatment was then resumed at the preceding dose level. If treatments were held for ≥ 4 weeks, then the patient was removed from the study.
With the dose-escalation scheme used, there was a ≥ 71% chance of escalating the combination if the underlying toxicity rate was ≤ 20% and a 91% chance of escalation if the underlying rate was ≤ 10%. In contrast, there was at most a 17% chance of escalation if the underlying toxicity rate was ≥ 50% and a ≤ 8% chance if the rate was ≥ 60%. A maximum of 35 patients were needed for this study.
Thirty-three patients were enrolled in the study between February 23, 1999 and April 26, 2000. All patients were assessable for toxicity and for response to treatment. Patient characteristics are illustrated in Table 2. The median age was 62 years. Twenty-two patients (67%) had received prior therapy, 10 patients (30%) had received prior chemotherapy, 3 patients (9%) had received prior radiotherapy, and 9 patients (27%) had received both chemotherapy and radiotherapy. NSCLC was the most prevalent malignancy (n = 11 patients; 33%). All patients had an ECOG performance status of 0 or 1.
Table 2. Patient Characteristics
|Male/female||22 (67)/11 (33)|
|Median age in yrs (range)||62 (27–77)|
|ECOG performance status|| |
| 0||3 (9)|
| 1||30 (91)|
|Prior treatment|| |
| Radiotherapy||12 (36)|
| Chemotherapy and othersa||19 (58)|
|Primary site|| |
| Lung||11 (33)|
| Unknown primary||5 (15)|
| Bladder||3 (9)|
| Kidney||3 (9)|
| Adrenal||1 (3)|
| Prostate||1 (3)|
| Breast||1 (3)|
| Gastric||1 (3)|
| Gall bladder||1 (3)|
| Liver||1 (3)|
| Rectum||1 (3)|
| Mediastinum||1 (3)|
| Pleura||1 (3)|
| Thymus||1 (3)|
Treatment Administered and Toxicity
In total, 125 cycles of treatment were delivered, with a median of 3 cycles (range, 1–11 cycles). Neutropenia and thrombocytopenia were the principal dose-limiting toxicities (n = 2 patients; Dose Level V). The most common toxicities reported during the first cycle of treatment were Grade 1–2 fatigue (15 of 33 patients) and gastrointestinal toxicities, including, Grade 1–2 nausea/emesis (14 of 33 patients) and diarrhea (12 of 33 patients). Diarrhea generally was mild (Grade 1–2); however, 2 of 11 patients had Grade 3 diarrhea at Dose Level IV. Elevation of liver enzymes was common, although it was mild in most patients, occurring in 25 of 33 patients during Cycle I of therapy, and reaching ≥ Grade 3 in only 5 of 33 patients (1 patient each at Dose Levels II, IV, and V and 2 patients at Dose Level III). Despite the abbreviated steroid premedication dosing, only one patient at Dose Level III had a hypersensitivity reaction (Grade 3) to docetaxel infusion that responded to symptomatic treatment and did not result in the discontinuation of treatment.
During Cycle I of therapy, docetaxel (20–30 mg/m2 per day) and gemcitabine (400 mg/m2 per day) were well tolerated among all patients, with most patients experiencing minimal toxicity. At Dose Level I, no patients developed Grade 3 or 4 toxicities. At Dose Level II (docetaxel 30 mg/m2 and gemcitabine 400 mg/m2), one patient developed Grade 3 peripheral edema that responded to diuretics, and a second patient developed Grade 4 neutropenia and Grade 3 elevation of liver enzymes (ALT). At Dose Level III (docetaxel 30 mg/m2 and gemcitabine 600 mg/m2), two of six patients developed significant toxicity: One patient developed Grade 4 neutropenia and Grade 3 thrombocytopenia as well as Grade 3 elevated liver enzymes (ALT), and a second patient developed Grade 3 elevated liver enzymes (ALT). At Dose Level IV, two of six patients developed significant toxicity (≥ Grade 3). At this dose level, Grade 4 neutropenia and Grade 3 diarrhea occurred in one patient, and Grade 3 thrombocytopenia and Grade 3 elevation of liver enzymes (ALP) occurred in another patient. At Dose Level V, four of six patients experienced severe or life-threatening toxicity. This was predominantly hematologic (neutropenia and thrombocytopenia in two patients), elevated liver enzymes (AST and ALT in one patient), and mucositis (in one patient). It was determined that Dose Level IV (docetaxel 36 mg/m2 and gemcitabine 600 mg/ m2) was the MTD, and that dose level was expanded to include a total of 11 patients to further define the toxicity profile (Table 3).
Table 3. Selected Severe and Life-Threatening (Grade 3 or 4) Toxicities during the First and Subsequent Cycles of Therapy
During subsequent cycles, the toxicities experienced by patients remained consistent; however, the severity of toxicities appeared to increase. Two patients at Dose Level I developed Grade 3 hematologic toxicity (leukopenia or neutropenia). At Dose Levels III–V, significant toxicities included Grade 3 hematologic toxicities (neutropenia and thrombocytopenia) and nonhematologic toxicities (mainly elevated liver enzymes). Grade 4 toxicities were observed primarily in patients at Dose Levels IV and V (n = 5 patients). No patients developed ≥ Grade 3 neuropathy while they were on study. All toxicities resolved with the cessation of therapy. One patient with metastatic rectal carcinoma (Dose Level II) expired while on study (after two cycles).
Two patients had objective partial responses, and an additional 26 patients had stable disease at the first disease evaluation. Of the two patients who achieved responses, one patient (Dose Level III) had metastatic NSCLC: At the time of diagnosis, this patient had metastatic disease to the brain, bones, liver, and lymph nodes. She underwent whole-brain radiation and then was treated with one cycle of vinorelbine and carboplatin, which was discontinued secondary to disease progression. She received six cycles of gemcitabine and docetaxel at Dose Level III and had a partial response in the liver. In addition, the lung mass decreased in size, although it did not reach the criteria for a partial response. The duration of her response was 5 months. The other patient who had a response had adenocarcinoma of unknown primary origin and presented with widespread peritoneal carcinomatosis, exophytic mass involving the urinary bladder, and a mass at the tail of the pancreas, suggesting a pancreatic origin. In addition, the patient had elevated tumor markers (CA 19-9 and carcinoembryonic antigen). The patient was treated at Dose Level I and had a partial response, which was evidenced by disappearance of the exophytic mass in the bladder, stabilization of peritoneal carcinomatosis, and normalization of tumor markers. This individual received 11 cycles of gemcitabine and docetaxel on study and subsequently received the same drugs for another year off study. The partial response lasted for 2 years. Over a 21-month follow-up, 18 of 33 patients (55%) died, and 15 patients remained alive (median follow-up, 11.5 months).
There is clear rationale for the combination of docetaxel and gemcitabine in the treatment of many types of solid tumors. Both of these agents have proven activity as single agents in a variety of malignancies, including NSCLC, breast carcinoma, head and neck carcinoma, ovarian carcinoma, and renal carcinoma. Initial efforts with this combination were focused mainly on incorporating both of these agents into drug combinations with platinum compounds.32, 33 Recently, evidence has accumulated showing that these two agents in combination may be as effective as platinum-containing combinations with the added benefit of a more favorable toxicity profile.34 In a randomized multicenter trial, patients with advanced NSCLC were treated either with docetaxel plus cisplatin or with docetaxel plus gemcitabine.34 The results were similar for both groups in terms of response rate, response duration, time to tumor progression, overall survival, 1-year survival, and 2-year survival.34 The docetaxel plus gemcitabine combination had a more favorable toxicity profile, with Grade 3–4 neutropenia occurring in 22% of patients and Grade 3 nausea and emesis in occurring 2% of patients compared with 34% and 10%, respectively, of patients in the docetaxel plus cisplatin group (P = 0.01 and P = 0.001, respectively).
Docetaxel and gemcitabine have different mechanisms of action, suggesting the possibility of additive (if not synergistic) effects. Many Phase I and Phase II trials have combined these two agents employing different doses and schedules in an attempt to identify a regimen with the least toxicity with antitumor activity. Examples include 4-week regimens (gemcitabine administered on Days 1, 8, and 15 and docetaxel administered on either Day 1 or Day 15)15, 16 and 3-week regimens (gemcitabine administered on Days 1 and 8 and docetaxel administered on Day 8 every 3 weeks).15, 17, 35 These schedules were feasible, active, and tolerated relatively well. However, in most of the regimens in which docetaxel is administered every 3 weeks, the addition of other myelosuppressive agents is difficult. The docetaxel dose generally was lower than 100 mg/m2, with most regimens using docetaxel 60–75 mg/m2.36–38 In this setting, hematologic toxicity is the dose-limiting toxicity.39, 40
Weekly administration of docetaxel appears to offer several advantages. Studies of weekly schedules of docetaxel have shown that this regimen is well tolerated with minimal myelosuppression and has the added benefit of maintaining dose intensity. At least three Phase I trials of weekly docetaxel have been reported.24, 26, 29 The schedule of administration in all trials was six weekly doses of docetaxel followed by a 2-week rest period. The results of all three studies were similar, with weekly docetaxel associated with decreased myelosuppression compared with the 3-week schedules. The dose-limiting toxicities seen in these trials were fatigue/asthenia24 and nail changes (onycholysis).29 Otherwise, therapy was well tolerated. Responses were seen in all trials. The recommended dose for Phase II trials was docetaxel 35–40 mg/m2 per week, representing a higher dose intensity compared with the standard regimen of docetaxel 60–100 mg/m2 every 3 weeks. It also is interesting to note that there was an absence of Grade 4 neutropenia in those Phase I trials.24, 26, 29 In view of the modest myelosuppression produced by weekly docetaxel administration, it is likely that other agents can be added while preserving dose intensity.
In the current report, we combined weekly docetaxel with gemcitabine. The MTD of docetaxel was 36 mg/m2 combined with gemcitabine 600 mg/m2 administered weekly for 3 weeks every 28 days. This regimen was tolerated well, and the dose-limiting toxicity was predominantly hematologic (neutropenia). At the MTD, neutropenia ≥ Grade 3 occurred in 2 of 11 patients (18%), and Grade 3 thrombocytopenia occurred in 1 of 11 patients. In view of the fact that the patients who were included in this Phase I study were heavily pretreated, it is probable that this regimen would have less toxicity in the front-line setting. In a recently reported, similar Phase I study of weekly docetaxel and gemcitabine given weekly for 2 weeks followed by 1 week of rest, the overall incidence of severe neutropenia was 23.5%.41 In that study, however, Grade 3–4 neutropenia occurred in 62% of patients who had received ≥ 2 prior chemotherapy regimens and in 0 of 15 patients who had received ≤ 1 prior chemotherapy regimen.41 The MTD for patients who received ≥ 2 prior courses of chemotherapy was gemcitabine 800 mg /m2 and docetaxel 40 mg/m2, compared with an MTD of gemcitabine 1250 mg/m2 and docetaxel 40 mg/m2 in minimally pretreated patients.41 Other dose-limiting toxicities in that report also included diarrhea and mucositis.41 In another study, Grade 3–4 neutropenia occurred in 44% of patients. The MTD in that trial was docetaxel 40 mg/m2 and gemcitabine 1000 mg/m2 administered weekly for 2 weeks of 3-week cycles.42
The hematologic toxicities observed in these trials of weekly docetaxel and gemcitabine compare favorably with the toxicities observed in trials of monthly docetaxel and gemcitabine weekly times three.18, 19, 42–45 Ryan et al., in a Phase I study of weekly gemcitabine and monthly docetaxel, reported a 44% incidence of Grade 3–4 neutropenia at the MTD (docetaxel 60 mg/m2 every 4 weeks and gemcitabine 600 mg/m2 per week for 3 weeks of a 4-week cycle).43 Rischin et al.44 reported a regimen with docetaxel and gemcitabine that had a 70% incidence of Grade 3–4 neutropenia. These observations further confirm the finding of decreased myelosuppression associated with weekly docetaxel compared with other schedules, facilitating the combination with other agents.
Nonhematologic toxicity also is less common with weekly docetaxel regimens. In the current study, no patient developed Grade 3 or 4 peripheral neuropathy. This is consistent with other reported trials of weekly docetaxel as single agent or in combination with other agents. In a study by Hainsworth et al., only 1 of 38 patients developed Grade 3 neuropathy.26 No Grade 3–4 neuropathy was observed in other Phase I trials of weekly docetaxel.29, 32 Similarly, no Grade 3 neuropathy was observed in another Phase I trial of a weekly docetaxel and gemcitabine combination.42
Pulmonary toxicity has been reported with both gemcitabine and the taxanes.46, 47 Most reports have described this toxicity as mild and self-limiting. However, more recently, the literature has indicated that, in rare instances, pulmonary problems can be severe and even fatal.46, 47 A report from Europe of severe pulmonary toxicity in patients who were treated with a combination of docetaxel and gemcitabine28 resulted in modification of docetaxel plus gemcitabine protocols by the ECOG, allowing a 2-hour lapse between the administration of the two drugs. We adopted this modification and did not encounter pulmonary toxicity. Hainsworth et al.48 reported 1 possible treatment-related death secondary to bilateral pneumonitis and respiratory failure among 41 patients with NSCLC who were treated with a weekly docetaxel and gemcitabine combination. There was no mention of the sequence of administration of the two drugs in their report.48 No significant (≥ Grade 3) pulmonary toxicity was observed on other weekly docetaxel plus gemcitabine Phase I trials.42, 43, 49
Another advantage of a weekly docetaxel schedule is the low incidence of hypersensitivity reactions and peripheral edema, despite the abbreviated course of steroids used for premedication used in most studies.24, 29, 32 Hainsworth and colleagues26 used dexamethasone 8 mg orally given the evening prior to treatment and on the morning and evening the day of treatment, for a total of 3 doses. Luck and Galaubitz24 used 1 dose of dexamethasone 4 mg IV 30 minutes before treatment, and Loëffler et al.29 used dexamethasone 8 mg IV before treatment then 4 mg orally every 4 hours for 4 doses. There were no hypersensitivity reactions observed in any Phase I trials of weekly docetaxel,24, 29, 32 compared with a 15.7% incidence of hypersensitivity and a 48.5% incidence of peripheral edema compared with the standard, 100 mg/m2, 3-week administration schedule with recommended premedication (dexamethasone 8 mg orally twice daily starting the day prior to docetaxel and continuing the dexamethasone for 2–4 days).50 In the current study, we used a single dose of 20 mg of dexamethasone 30 minutes prior to docetaxel administration and observed only one hypersensitivity reaction at Dose Level III: The patient responded to symptomatic management and required no treatment interruption. These observations suggest that an abbreviated course of steroids may be appropriate premedication when administering docetaxel on weekly schedules.
Although evaluation of the antitumor activity of this regimen was not the primary objective of this study, the presence of responses in two of our heavily pretreated patients points to the potential activity of this combination. The efficacy of a weekly administration schedule of docetaxel remains an open question, however. There is evidence supporting weekly administration of some antitumor drugs. The weekly schedule is based in part on the concept of kinetics and tumor growth.51–54 The frequent exposure may affect more sensitive tumor cells and may reduce the emergence of drug-resistant clones. Weekly docetaxel has been studied in a Phase II trial of the treatment of elderly patients with advanced NSCLC. Seven of 38 patients (18%) had objective responses, and an additional 13 patients (34%) had minor response or stable disease at their first evaluation.24 Marked activity of this weekly regimen also has been documented in patients with metastatic breast carcinoma.24 In a recent Phase II study, weekly administration of docetaxel at 40 mg/m2 for 6 consecutive weeks followed by 2 weeks of rest in patients with metastatic breast carcinoma was associated with a 41% overall response rate, and no Grade 4 toxicity occurred.55 In another study of weekly docetaxel for the treatment of patients with advanced malignancies, nine patients with breast carcinoma who had previously received both taxanes and anthracyclines had tumor responses (one patient had a complete response, and eight patients had partial responses). Two of these patients had received docetaxel in combination with anthracycline within the preceding 6 months. Based on these observations, it is reasonable to hypothesize that weekly docetaxel may overcome the drug resistance observed with the 3-week schedule.49 A similar phenomenon was also observed with paclitaxel given weekly.56, 57
Another aspect that deserves further investigation is the interaction of docetaxel and gemcitabine as a function of treatment schedule. Recent studies have shown the importance of recognizing specific perturbations induced by different drugs on the cell cycle when designing combination or sequential therapies to increase additive or synergistic effects and to avoid antagonistic effects.58–60 When it was tested on a human lung cell line, it was found that docetaxel followed after 48 hours by gemcitabine was synergistic.61 Conversely, simultaneous treatment induced an antagonistic effect on the same cell line, and the sequential use of gemcitabine followed by docetaxel produced only a weak synergistic effect.61 Flow cytometric analysis showed that, in the docetaxel → gemcitabine sequence, docetaxel produced a block in G2/M-phase and, after 48 hours, provided gemcitabine with a large fraction of recovered synchronized cells in the G1/S-phase, which is the specific target for gemcitabine.61 These findings are interesting and should be tested in clinical trials.
In conclusion, the weekly schedule of docetaxel and gemcitabine employed in this trial was well tolerated, and the preliminary finding of antitumor activity is of interest. The efficacy of the weekly docetaxel and gemcitabine combination should be validated in Phase II trials for specific disease sites.