• sarcoma;
  • gemcitabine;
  • chemotherapy;
  • soft tissue sarcoma


  1. Top of page
  2. Abstract


Care for patients with advanced sarcomas is mainly palliative. Gemcitabine, a nucleoside antimetabolite, is an analog of deoxycytidine that has shown antitumor activity in several tumors. The aim of the current study was to determine the clinical activity of gemcitabine in patients with sarcomas.


The authors evaluated gemcitabine in patients with histologically confirmed sarcomas; one prior exposure to chemotherapy treatment was allowed. Prior radiation was allowed if given to non-indicator lesions. Treatment consisted of gemcitabine 1250 mg/m2 intravenously over 30 minutes, every week × three, cycles repeated q28 days.


Twenty nine of 30 patients were evaluable; one patient refused to initiate study treatment. The mean age was 50 years (range, 22-81 years); 59% were male, and 35% had an Eastern Cooperative Oncology Group performance status of 0 (vs. 1 or 2). Patients were histologically classified as leiomyosarcoma (seven gastrointestinal, four retroperitoneal, two inferior vena caval, three of the extremity, and two uterine), synovial (two patients), malignant fibrous histiocytoma (two patients), fibrosarcoma (one patient), osteosarcoma (two patients), liposarcoma (one patient), hemangiosarcoma (one patient), or giant cell (one patient). Patients received an average of two cycles (range, one to eight). Eighty three percent of patients discontinued treatment due to progression and 14% due to toxicity/refusal. Hematologic toxicities ≥ Grade 3 were seen in 32% of patients and consisted of leukopenia and thrombocytopenia. Anorexia (Grade 1/2 in 6 patients, Grade 3 in 1 patient), nausea (Grade 1/2 in 7 patients, Grade 3 in 1 patient), and lethargy (Grade 1/2 in 19 patients) were the most frequently observed nonhematologic toxicities. One patient experienced Grade 3 edema and muscle infarction. A different patient experienced unexplained Grade 3 chest pain. One partial response was observed in a uterine leiomyosarcoma patient lasting at least three months. Overall response rate was 3% (95% confidence interval [CI]: 0-15). Median time -to progression was 2.1 months (95% CI: 1.8–3.0).


The current gemcitabine regimen demonstrated acceptable levels of toxicity, but it failed to produce the number of responses needed to justify expansion of the current study. This regimen is not recommended for advanced sarcomas. Cancer 2002;94:3225–9. © 2002 American Cancer Society.

DOI 10.1002/cncr.10602

Gemcitabine, a nucleoside antimetabolite, is an analog of deoxycytidine that has shown clinical activity in a variety of solid tumors, including nonsmall cell lung, ovarian, head and neck, and pancreatic carcinomas.1 The use of gemcitabine has been well tolerated and has shown clinical benefit in the management of the fragile pancreatic carcinoma patient.2 In preclinical studies, gemcitabine has shown activity against xenografts of sarcoma cell lines.3 Recent reports suggest some activity in patients with advanced sarcomas.4

Currently, the management of patients with advanced sarcomas is largely palliative. Doxorubicin and ifosfamide are the most active agents in sarcomas, with a response rate of around 25%.5 Other agents with activity include cyclophosphamide, dacarbazine, vincristine, mitomycin, etoposide, and cisplatin.6 Combination regimens with or without doxorubicin have failed to show an improved overall survival.5, 7 Furthermore, toxicities of single agent and combination regimens for sarcomas can be substantial. There is a need to find new agents that are active against sarcomas without significant toxicities or complicated administration schedules. In the current report, we describe our single institution Phase II trial with gemcitabine in patients with advanced sarcomas.


  1. Top of page
  2. Abstract

Patient Eligibility

Patients were required to have histologically confirmed sarcoma with histologic evidence of advanced disease with no standard curative therapy. Patients were allowed to have at most one prior chemotherapy regimen for advanced sarcoma, which was to have been completed at least four weeks (nitrosoureas or mitomycin > six weeks) prior to registration of the study. Patients were to be over age 18, have a physician estimated daily oral intake of at least 1,200 calories, and have an adequate Eastern Cooperative Oncology Group performance status (PS < 2). Within 14 days of registration, acceptable hematologic and blood chemistry parameters included: leukocyte count ≥ 3500 uL, platelets ≥ 150,000 uL, total bilirubin ≤ 1.5 times the institutional upper limit of normal (UNL), creatinine ≤ 1.5 times UNL, and serum aspartate aminotransferase ≤ 3 times UNL. Pregnant or lactating women were not candidates for the current study. Prior or planned concurrent radiotherapy and central nervous system metastases were considered as contraindications to registration. Both measurable and evaluable disease were allowed.

Treatment with Gemcitabine

One course of study treatment consisted of 1250 mg/m2 of gemcitabine administered weekly via infusion over 30 minutes in 250 mL normal saline for three consecutive weeks. Patients were retreated every 28 days, continuing until either unacceptable toxicity, progressive disease, or a noncancer related illness prevented administration of gemcitabine. Patients were allowed to receive all prestudy concomitant medications, including antiemetics. For serious nonhematologic toxicity, gemcitabine was withheld until toxicity decreased to ≤ Grade 2, with subsequent treatment at a 25% dose reduction. Patients having absolute neutrophil count (ANC) between 500 and 999 uL or platelets between 50,000 and 99,999 uL had their gemcitabine doses decreased by 25% at retreatment. Gemcitabine was held in cases of patients having ANC ≤ 500 uL or platelets < 50,000 uL until resolved (i.e., ≤ Grade 2), and then gemcitabine dose was resumed at a 25% reduction. Individual doses of gemcitabine that had been delayed one week due to toxicity were subsequently not administered for that given course. The study was approved by the institutional review board, and informed consent was obtained from all patients.

Patient Evaluation

Standard response criteria were used to evaluate patients' objective tumor status ≤ 14 days prior to registration and prior to every course of treatment. A complete response (CR) required the total disappearance of known tumors in the absence of new lesions for both measurable and evaluable disease. Patients having measurable disease and showing at least a 50% reduction in the sum of the products of the largest perpendicular diameters of the indicator lesion(s), single or multiple sites, without the appearance of new lesion(s), were classified as having a partial response (PR). Progression was considered as at least a 25% increase in the size of any lesion prior to the patient's initial treatment or the appearance of new lesion(s). Patients having evaluable disease and showing a definite decrease in tumor size were classified as having regressed (REGR). A definite increase in tumor size or the development of new lesion(s) was considered progression in patients having evaluable disease. Patients were classified as having stable disease if the criteria for CR, PR, REGR, or progression were not satisfied.

Patients who died (or were lost to followup) without progression were considered to have progressed at the date of death (or last contact) unless documentation proved otherwise, in which case they were considered as having no progression at the date of last tumor evaluation. Duration of response was calculated from the date of the patient's first objective status of CR, PR, or REGR, to the date of progression. Time to progression was calculated from study entry to disease progression. Time to death (i.e., survival) was calculated from the date of study entry to death or last contact.

Statistical Considerations

The primary endpoint of the current trial was to evaluate the confirmed tumor response rate of gemcitabine. Confirmed tumor response was defined as an objective tumor response of CR, PR, or REGR, sustained for at least two consecutive evaluations. All eligible patients that initiated treatment were considered evaluable for estimating the confirmed tumor response rate. A two stage Phase II design was used to evaluate confirmed tumor response in a total of 55 patients. A minimum of 4 confirmed responses in the initial 30 evaluable patients (i.e., Stage 1) was considered sufficient to proceed to the second stage of accrual. A minimum of 10 confirmed responses in a total of 55 evaluable patients (i.e., Stage 1 and Stage 2) was considered sufficient evidence to find promising activity of the current regimen in this patient population. Using this design, we had 90% power, at 0.04 level of significance, to detect a confirmed response rate of at least 25% (i.e., clinical activity). The requirement for second stage patient accrual of exact binomial confidence intervals (CI) reported for estimates of confirmed tumor response rates was not satisfied.

Secondary endpoints included toxicity, duration of response, time to disease progression, and survival. Nonhematologic toxicity was summarized in a tabular manner as a maximum grade for a given type of event. Hematologic nadirs were reported as a minimum value per patient and course of treatment. Hematologic toxicity was summarized via summary statistics (i.e., mean, median, etc.). Kaplan-Meier methodology8 was used to describe the distributions of time to disease progression and survival. All analyses were performed using SAS (SAS, Cary, NC).9


  1. Top of page
  2. Abstract

Patient Characteristics

A total of 30 patients were enrolled between January 8, 1998, and October 29, 1999. One patient was ruled ineligible and excluded from analyses after refusing to receive the study drug, gemcitabine. Table 1 summarizes the characteristics for the remaining 29 eligible patients. At study entry, eligible patients ranged in age from 22 to 81 years (mean, 50.0), 35% (10 out of 29) had a PS of 0 (vs. 1 or 2), 97% (28 out of 29) were white, and 59% (17 out of 29) were male. Histologic subtypes included: leiomyosarcoma, 18 patients (7 gastrointestinal, 4 retroperitoneal, 2 inferior vena caval, 3 of the extremity, and 2 uterine); synovial, 2 patients; malignant fibrous histiocytoma, 2 patients; osteosarcoma, 2 patients (1 osseous, 1 extraosseous); and alveolar soft part sarcoma, liposarcoma, angiosarcoma, myofibroblastic sarcoma, sarcoma not otherwise specified (NOS), 1 patient each. Seventy nine percent (23 out of 29) of eligible patients presented with Grade 3 or 4 disease (vs. Grades 1 or 2 or NOS). Fifty two percent (15 out of 29) of patients had received prior radiation therapy for their advanced sarcomas. Sixty six percent (19 out of 29) of patients had had prior chemotherapy. A majority of patients, 69% (20 out of 29), presented with measurable disease.

Table 1. Baseline Patient Characteristics (n = 29)
 Male17 (59%)
 Female12 (41%)
Eastern Cooperative Oncology Group Performance Status 
 010 (35%)
 115 (52%)
 24 (14%)
Prior radiation therapy 
 Yes15 (52%)
 No14 (48%)
Prior chemotherapy 
 Yes19 (66%)
 No10 (35%)
Pulmonary metastases 
 Yes14 (48%)
 No15 (52%)
Liver metastases 
 Yes13 (45%)
 No16 (55%)
Other metastatic site(s) 
 Yes20 (69%)
 No9 (31%)
Primary site 
 Extremity6 (21%)
 Truncal23 (79%)
Disease Status 
 Measurable20 (69%)
 Evaluable9 (31%)
 Low (1,2)4 (14%)
 High (3,4)23 (79%)
 NOS2 (7%)


On average, patients completed two cycles (range, one to eight) of gemcitabine. Ten percent (3 out of 29) of patients completed at least five cycles of study treatment. Ninety three percent (27 out of 29) of patients received at least two thirds (i.e., two of three weeks) of their targeted gemcitabine dose during their initial cycle of gemcitabine treatment. Eighty nine percent (16 out of 18) and 60% (6 out of 10) of patients received at least two thirds (i.e., two of three weeks) of their targeted gemcitabine dose during Cycles 2 and 3, respectively. A majority (83%) discontinued due to disease progression and 14% (4 out of 29) due to toxicity/refusal. Myelosuppression consisted primarily of leukopenia and thrombocytopenia. Thirty two percent (9 out of 28) of patients experienced severe hematologic toxicity (i.e., Grade ≥ 3). Overall, 14% (4 out of 28) of patients experienced Grade 3 leukopenia, all of which occurred on their initial courses of treatment. Twenty one percent (6 out of 28) of patients reported Grade 3/4 thrombocytopenia (three Grade 4) at some time during gemcitabine treatment.

Nonhematologic toxicity included primarily anorexia, alopecia, nausea, vomiting, and lethargy (Table 2). Lethargy was the most frequently observed toxicity. Grade 3 toxicities included anorexia, cardiac, edema, phlebitis, nausea, and vomiting. One patient developed muscle infarction of the thigh after nine doses (i.e., three cycles) of gemcitabine. Another patient experienced thrombocytopenia, rigors, fever, nausea, and vomiting after receiving two doses of gemcitabine during the initial three week course of treatment.

Table 2. Maximum Severity of Toxicity, per Patient and Type (n = 29)
NCI CTC CategoryFrequency
Grade 1 and 2Grade 3
  1. NCI: National Cancer Institute; CTC: common toxicity criteria. Mayo Clinic supplement combined with CTC.

Alopecia5 (17%)0
Cardiac01 (3%)
Hepatotoxicity1 (3%)0
Anorexia6 (21%)1 (3%)
Diarrhea4 (14%)0
Nausea7 (24%)1 (3%)
Vomiting4 (14%)1 (3%)
Lethargy19 (66%)0
Edema1 (3%)1 (3%)
Phlebitis1 (3%)1 (3%)

Response, Progression, and Survival

All 29 eligible patients were considered evaluable for response. One confirmed response was observed in a uterine leiomyosarcoma patient after receiving five courses of gemcitabine; the response continued for five months. The objective (regression) response rate was 3% (1 out of 29). Out of the overall patient cohort, 90% (26 out of 29) have progressed and 48% (14 out of 29) have died, with a minimum followup of 1 month for surviving patients (range, 1-23 months). Estimated six month progression-free rate survival was 11%. The estimated one year survival rate was 43% (Fig. 1).

thumbnail image

Figure 1. Kaplan-Meier estimate of overall survival.

Download figure to PowerPoint


  1. Top of page
  2. Abstract

Chemotherapy for advanced sarcomas has been disappointing. Attempts to improve upon the single agent activity of doxorubicin or ifosfamide with dose intensification or combinations have not consistently shown an improvement in overall survival. Moreover, the chemotherapy regimens are complicated and toxic.5, 10, 11

Because gemcitabine has broad clinical activity, a favorable toxicity profile, and in vitro activity in sarcoma cell lines, several cancer research groups have studied this agent in sarcomas. Merimsky et al. reported a 5.5% response rate with gemcitabine in 18 heavily pretreated patients with sarcomas (10 of the 18 had bone sarcomas).4 Although the response rate was low, the median time to progression was over 27 weeks.

In the current study, we observed only one partial response (3% response rate) in a patient with uterine leiomyosarcoma, lasting five months. Moreover, the median time to disease progression for the study was only 2.1 months. The chemotherapy resistant gastrointestinal leiomyosarcomas made up 24% (7 out of 29) of the study population. The recent positive results of STI-571 in this subset of sarcomas will limit the number of these patients in future general chemotherapy trials against sarcomas.12, 13 In addition, 45% of our patients had liver metastases, known to be a poor predictor of response.14 Despite these poor risk factors for response, our patient population was typical for a general sarcoma practice. Although the response rates and times to disease progression were disappointing, gemcitabine was well tolerated, with 12% of patients discontinuing gemcitabine secondary to toxicity or refusal. The main toxicities were myelosuppression, anorexia, and lethargy.

Is gemcitabine an active agent in soft tissue sarcomas? The answer may depend on what subtype of sarcoma is being studied. Patel et al. had an overall response rate of 18%, with a median duration of response of 3.5 months with gemcitabine in a study population that excluded gastrointestinal leiomyosarcomas.15 In their subset of nongastrointestinal leiomyosarcomas, there was a 4 out of 10 response rate. However, in a separate arm of the study with only gastrointestinal leiomyosarcomas, they found no responses. In classical Kaposi's sarcoma, Brambilla et al. found gemcitabine to be active.16 Recently, Hensley et al. reported that, when combined with docetaxel, gemcitabine is active in advanced leiomyosarcomas.17 This suggests that gemcitabine may have activity in certain subtypes of sarcoma.

Studies using gemcitabine in a general sarcoma population with varying subtypes of sarcoma have failed to show significant activity. Spath-Schwalbe et al. reported limited activity of gemcitabine in pretreated general sarcoma patients.18 The Eastern Cooperative Oncology Group completed a study of gemcitabine in a previously untreated general sarcoma population and showed limited activity (personal communication). Due to the small number of patients with each subtype of sarcoma and inclusion of historically chemotherapy resistant sarcomas, these studies are unable to determine the true activity of gemcitabine for each subtype of sarcoma. Moreover, the infusion schedule may influence the activity of gemcitabine in sarcomas.

There is some suggestion that a longer infusion time of 150 minutes might be a better infusion duration.15 Further studies on infusion duration and activity in sarcoma are needed.

Overall, gemcitabine in the current general sarcoma population given as a short infusion was inactive. Further confirmatory studies of gemcitabine alone or in combination in select subtypes of sarcoma with particular attention to the optimal infusion schedule are needed.


  1. Top of page
  2. Abstract
  • 1
    Plunkett W, Huang P, Xu Y, et al. Gemcitabine: metabolism, mechanisms of action, and self-potentiation. Semin Oncol. 1995; 22: 310.
  • 2
    Moore M, Andersen J, Burris H, et al. A randomized trial of gemcitabine (GEM) versus 5FU as first-line therapy in advanced pancreatic cancer [abstract]. Proc ASCO. 1995; 14: A473.
  • 3
    Braakjuis BJ, Ruiz van Harpere VW, Boven E, et al. Schedule-dependent antitumor effect of gemcitabine in in vivo model systems. Semin Oncol. 1995; 22: 4246.
  • 4
    Merimsky O, Meller I, Flusser G, et al. Gemcitabine in soft tissue or bone sarcoma resistant to standard chemotherapy: a Phase II study. Cancer Chemother Pharmacol. 2000; 45: 177181.
  • 5
    Edmonson JH, Ryan LM, Blum RH, et al. Randomized comparison of doxorubicin alone versus ifosfamide plus doxorubicin or mitomycin, doxorubicin, and cisplatin against advanced soft tissue sarcomas. J Clin Oncol. 1993; 11: 12691275.
  • 6
    Brennan MR, Casper ES, Harrison LB. Sarcomas of the soft tissues and bone. In: DeVitaVT, HellmanS, RosenbergSA, editors. Cancer principles and practice of oncology. 5th ed. J.B. Lippincott Company, 1997: 7381788.
  • 7
    Edmonson JH. Needed: qualitative improvement in antisarcoma therapy. J Clin Oncol. 1995; 13: 15311533.
  • 8
    Kaplan E, Meier P. Nonparametric estimation from incomplete observation. J Am Stat Assoc. 1958; 53: 457481.
  • 9
    SAS/STAT User's Guide, Version 7. Cary, NC: SAS Institute, Inc., 1989.
  • 10
    Le Cesne A, Judson I, Crowther D, et al. Randomized Phase III study comparing conventional-dose doxorubicin plus ifosfamide versus high-dose doxorubicin plus ifosfamide plus recombinant human granulocyte-macrophage colony-stimulating factor in advanced soft tissue sarcomas: a trial of the European Organization for Research and Treatment of Cancer/Soft Tissue and Bone Sarcoma group. J Clin Oncol. 2000; 18: 26762684.
  • 11
    Antman K, Crowley J, Balcerzak SP, et al. An intergroup Phase III randomized study of doxorubicin and dacarbazine with or without ifosfamide and mesna in advanced soft tissue and bone sarcomas. J Clin Oncol. 1993; 11: 12761285.
  • 12
    Blanke CD, von Mehren M, Joensuu H, et al. Evaluation of the safety and efficacy of an oral molecularly-targeted therapy, STI571 in patients (Pts) with unresectable or metastatic gastrointestinal stromal tumors (GISTS) expressing C-KIT (CD117) [abstract]. Proc ASCO. 2001; 20: 1a(1).
  • 13
    van Oosterom AT, Judson I, Verweij J, et al. Safety and efficacy of imatinib (STI571) in metastatic gastrointestinal stromal tumours: a Phase I study. Lancet. 2001; 358: 14211423.
  • 14
    Van Glabbeke M, van Oosterom AT, Oosterhuis JW, et al. Prognostic factors in advanced soft tissue sarcoma (STS): an overview of 1742 patients treated with doxorubicin containing first line regimens by the EORTC Soft Tissue and Bone Sarcoma group (STBSG). Ann Oncol. 1994; 2( Suppl 8): 171.
  • 15
    Patel SR, Gandhi V, Jenkins J, et al. Phase II clinical investigation of gemcitabine in advanced soft tissue sarcomas and window evaluation of dose rate on gemcitabine triphosphate accumulation. J Clin Oncol. 2001; 19: 34833489.
  • 16
    Brambilla L, Labianca R, Ferrucci SM, et al. Treatment of classical Kaposi's sarcoma with gemcitabine. Dermatology. 2001; 202: 199122.
  • 17
    Hensley ML, Venkatraman E, Maki R, et al. Docetaxel plus gemcitabien is active in leiomyosarcomas: results of a Phase II trial [abstract]. Proc ASCO. 2001; 20: 1408a.
  • 18
    Spath-Schwalbe E, Genvresse I, Kaschuth A, et al. Phase II trial of gemcitabine in patients with pretreated advanced soft tissue sarcomas. Anticancer Drugs. 2000; 11: 325329.