Breast Cancer Research and Treatment

, Volume 108, Issue 3, pp 319–331

Gemcitabine in the management of metastatic breast cancer: a systematic review

Authors

  • Susan Dent
    • The Ottawa Hospital Regional Cancer Centre
    • McMaster University
  • Maureen Trudeau
    • Toronto-Sunnybrook Regional Cancer Center
Review

DOI: 10.1007/s10549-007-9610-z

Cite this article as:
Dent, S., Messersmith, H. & Trudeau, M. Breast Cancer Res Treat (2008) 108: 319. doi:10.1007/s10549-007-9610-z

Abstract

Background

A systematic review of the evidence for gemcitabine chemotherapy, alone or in combination, in women with metastatic/advanced breast cancer was undertaken in order to determine gemcitabine’s role in the first-line and/or second-line or greater setting.

Methods

MEDLINE, EMBASE, the American Society of Clinical Oncologists, San Antonio Breast Cancer Symposium proceedings, and the Cochrane Library were searched through September 2006 for randomized controlled trials and non-randomized phase two trials.

Results

Eighty-three trials were identified, including four randomized phase III trials. All of the phase III trials included first-line patients. Two of the phase III trails demonstrated clinical benefit with gemcitabine-based chemotherapy in terms of superior efficacy or less toxicity while two phase III trials found no clinical benefit based on less efficacy or increased toxicity. Although 78 phase II trials of gemcitabine alone or in combination with other chemotherapy agents were identified, few combinations showed results compelling enough to warrant randomized trials.

Conclusion

Available data do not support the acceptance of gemcitabine as a standard therapeutic option in women with metastatic breast cancer in the third-line or greater setting, nor should it be considered as first-line therapy in anthracycline naïve women. Gemcitabine appears to be most effective when administered with a taxane (docetaxel/paclitaxel) in the first- or second-line setting, with gemcitabine/taxane combinations representing a viable alternative to currently accepted taxane combinations such as capecitabine/docetaxel. There is no evidence at this time to support the use of gemcitabine triplets, given the equal efficacy to anthracycline triplets and the added toxicity.

Keywords

Metastatic breast cancerGemcitabineChemotherapySystematic review

Introduction

Breast cancer is the most common malignancy of women in developed countries. Despite the improvements in breast cancer prevention, early detection, and adjuvant therapy, many women develop metastatic breast cancer that is considered incurable, with a median survival of 2–3 years.

The anthracyclines (doxorubicin and epirubicin) have been considered one of the most active classes of chemotherapy agents in the treatment of advanced breast cancer. Over the last 10 years, the widespread adoption of anthracyclines in the treatment of early-stage disease has led to a shift in the first-line treatment of metastatic breast cancer. The anthracyclines remain an option for women with advanced breast cancer who are chemotherapy naïve.

Several studies have investigated the impact of the combination of anthracyclines and taxanes, given either sequentially [13] or in combination [4] in women with early-stage disease. The favourable results of those studies have led to the increasing use of taxanes in the treatment of early-stage breast cancer and their diminishing use in advanced disease. Several other systemic therapy agents, alone or in combination, have been shown to be active in metastatic breast cancer, including capecitabine, vinorelbine, platinums (cisplatin, carboplatin), mitoxantrone, and trastuzumab (in patients who overexpress HER2/neu). There is no single standard systemic therapy for the treatment of metastatic breast cancer, particularly in women who have received anthracyclines and/or taxanes in the adjuvant setting. Given the significant morbidity and mortality associated with advanced breast cancer, novel chemotherapeutic agents which provide clinical benefit with minimal toxicity are needed.

Gemcitabine (Gemzar) is a pyrimidine antimetabolite that interferes with DNA synthesis. Gemcitabine has shown promising activity both as a single agent and in combination in the treatment of metastatic breast cancer with minimal toxicity. A systematic review of the literature was conducted to answer the following questions:
  • What is the role of gemcitabine, alone or in combination, as first-line chemotherapy in women with metastatic breast cancer?

  • What is the role of gemcitabine, alone or in combination, as second-line or greater chemotherapy in women with metastatic breast cancer?

Methods

MEDLINE was searched in its entirety to September 2006 using a disease-specific medical subject heading (MeSH) descriptor (“breast neoplasms”) and an agent-class MeSH descriptor with qualifier (“deoxycytidine/analog and derivatives”). The Excerpta Medica database (EMBASE) was also searched in its entirety to September 2006 using a disease-specific Excerpta Medica Tree (EMTREE) term (“breast cancer”) and an agent-specific EMTREE term (“gemcitabine”). These terms were then combined with the publication-type search terms for clinical trials, systematic reviews, meta-analyses, and practice guidelines. Issue 1 (2004) of the Cochrane Library and on-line conference proceedings from the American Society of Clinical Oncology (ASCO) (http://www.asco.org/ac/1,1003,_12-002095,00.asp; 1999–2006) and the San Antonio Breast Cancer Symposium (http://www.sabcs.org/SymposiumOnline/index.asp#abstracts; 2001–2005) were also searched. The Canadian Medical Association Infobase (http://www.mdm.ca/cpgsnew/cpgs/index.asp) and the National Guidelines Clearinghouse (http://www.guideline.gov) were searched for existing evidence-based practice guidelines. Relevant articles and abstracts were selected and reviewed by two reviewers, and the reference lists from these sources were examined for additional trials, as were the reference lists from relevant review articles.

Articles were selected for inclusion if they met the following criteria:
  • Gemcitabine, alone or in combination with other systemic therapy agents, was evaluated in the metastatic setting, using any of the publication types listed in the search strategy (clinical trials, systematic reviews, meta-analyses, and practice guidelines). After August 2005, only randomized controlled trials were considered for inclusion.

  • Reported outcomes included overall response rate (ORR), time to progression (TTP), duration of response, or disease-free or overall survival.

  • Clinical trial results were reported in full papers or abstracts.

  • Trials published in a language other than English were excluded, as resources for translation were not available.

  • Reports based on trials that have not completed patient accrual and were clearly ongoing were excluded, unless the report clearly stated that the analysis was pre-planned.

  • Reports based on solely dose-finding phase I trials were excluded. Reports of combination phase I/II trials were included.

Because of the low number of randomized controlled trials identified during the literature search, no meta-analysis was considered.

Results

Eighty-three trials were identified. Of these trials, four were phase III randomized clinical studies [514], one was a randomized phase II clinical trial [15], and the remainder were single-arm phase II clinical trials (see tables for references). Thirty-nine of the studies were reported solely in abstract form. There were no notable deficiencies in trial design identified in any of the identified randomized trials, although the reporting of randomization details (sequence generation, allocation concealment, and implementation) was not complete for any of the trials. Table 1 summarizes several key characteristics of the five identified randomized trials.
Table 1

Gemcitabine alone or in combination in first-line therapy or greater for metastatic breast cancer—Characteristics of included randomized trials

Trial

Treatment arms

Pts

Type of trial and analysis

Treatment setting

Patient characteristics

Expected effect, power, and planned sample sizea

Intent to treat analysis?

Correction for interim analysis?

Chan et al [5, 6] [abstract]

GD

153

Multicentre Phase III, Final

First- or second-line

One prior anth.-containing (neo)adj. or first-line met. chemo., prior (neo)adj. taxane allowed if ≥ 6 months

PFS median 6 months versus 8.2 months, 80% power, ≥  250 events (actual 259 events)

Yes

No interim analysis reported

XD

152

O’Shaughnessy et al. [712] [abstract]

GT

267

Multicentre Phase III, Interim

First-line

Prior (neo)adj. anth.-containing chemo., unless contraindicated, no chemo. for met. disease

At final analysis, HR 0.75 for overall survival, 85% power, 440 events (actual 343 events)

Yes

Yes

T

262

Feher et al. [13]

G

198

Multicentre Phase III, Final

First-line

No prior anth. chemo., up to one prior adj. chemo. if ≥ 1 year before study, no prior chemo. for met. disease

TTP median 26 vs. 35 weeks, 80% power, 440 patients (actual 397 patients)

Yes

Yes

E

199

Zielinski et al. [14]

GET

124

Multicentre Phase III,

First-line

No prior anth. chemo., up to one prior adj. chemo. if relapsed ≥ 6 months after chemo., no prior chemo. for met. disease

TTP HR 1.50, 80% power, 260 patients (actual 259 patients)

No

No interim analysis reported

FEC

135

Final

Khoo et al. [15]

GT1b

72

Randomized phase II

First-line or greater

Prior anth. chemo., either adjuvant or metastatic. Prior taxane therapy allowed if ≥ 12 months

Rule out response rate of 40% or lower given a true response rate of 60%, 89% power, 70 patients per arm

NR

No interim analysis reported

GT2b

67

GD

65

Note: See text for complete dosage and schedule information, except as indicated below

a Significant level 0.05

b GT1 was 1250 mg/m2 gemcitabine plus 175 mg/m2 paclitaxel. GT2 was 1000 mg/m2 gemcitabine and 100 mg/m2 paclitaxel

Abbreviations: anth., anthracycline; adj., adjuvant; C, cyclophosphamide; chemo., chemotherapy; D, docetaxel; E, epirubicin; F, 5-fluorouracil; G, gemcitabine; HR, hazard ratio; met., metastatic; NR, not reported; PFS, progression-free survival; Pts, number of patients; T, paclitaxel; TTP, time-to-progression; X, capecitabine

A large number of phase II studies were also identified. The selected studies were comparable in terms of patient inclusion and exclusion criteria, where those requirements were reported. These general criteria were: women with advanced or metastatic breast cancer; adequate bone marrow, renal, and hepatic function; either a Karnofsky performance scale rating of > 60 or an Eastern Cooperative Cancer Group (ECOG)/World Health Organization (WHO)/Zubrod performance scale rating of 0–2; no second malignancy other than cervical carcinoma or basal cell carcinoma of the skin (usually within five years); and no brain or central nervous system (CNS) metastases. The dosing regimen of gemcitabine differed between studies, which could account for some of the differences in outcome measures between trials.

Randomized trials

All of the identified randomized trials were of gemcitabine, alone or in combination, in the first-line or greater setting. There were no trials that excluded first-line therapy. Data on efficacy and toxicity from these trials is summarized in Tables 2 and 3.
Table 2

Gemcitabine alone or in combination in first-line therapy or greater for metastatic breast cancer—Efficacy results of randomized trials

Trial

Treatment arms (Pts)

ORR

TTP/TTF

PFS

Response duration

Overall survival

Rate

Comp.

Median (months)

Comp.

Median (months)

Comp.

Median (months)

Comp.

Median (months)

Comp.

Chan et al. [5, 6]

GD (153)

32%

P = 0.93

4.75a

P = 0.51

8.75a

P = 0.29

9a

P = 0.31

NR

 

XD (152)

32%

4.5a

8.75a

10.5a

O’Shaughnessy et al. [712]

GT (267)

40.8%

P  <  0.0001

5.2

P  < 0.0001b

NR

NR

NR

NR

18.5

P = 0.018c

T (262)

22.1%

2.9

15.8

Feher et al. [13]

G (198)

16.4%

P  <  0.0001

3.4

P = 0.0001

NR

NR

7.8

P = 0.2137

11.8

P = 0.004

E (199)

40.3%

6.1

9.9

19.1

Zielinski et al. [14]

GET (124)

62.3%

P = 0.093

9.1

P = 0.557

NR

NR

7.8

P = 0.125

29.5

P = 0.61

FEC (135)

51.2%

9.0

8.5

24.9

Khoo et al. [15]

GT1 (72)d

48.6%

NR

7.5

NRe

NR

8.3

NR

NR

GT2 (67)d

52.2%

7.0

8.2

GD (65)

52.3%

7.4

7.8

Note: See text for complete dosage and schedule information, except as indicated below. Statistically significant comparisons are in bold text

a Originally in weeks, converted at 4 weeks per month for comparability

b HR (GT vs. T) 0.650 (95% CI 0.524–0.805)

c HR (GT vs. T) 0.775 (95% CI 0.627–0.959)

d GT1 was 1250 mg/m2 gemcitabine plus 175 mg/m2 paclitaxel, GT2 was 1000 mg/m2 gemcitabine and 100 mg/m2 paclitaxel

e HR (GT1 vs. GT2) 0.96 (95% CI 0.65–1.42), HR (GT1 vs. GD) 0.97 (95% CI 0.65–1.44), HR (GT2 vs. GD) 1.01 (95% CI 0.68–1.51)

Abbreviations: C, cyclophosphamide; CI, confidence interval; Comp., comparison; D, docetaxel; E, epirubicin; F, 5-fluorouracil; G, gemcitabine; HR, hazard ratio; NR, not reported; ORR, Objective Response Rate; PFS, progression-free survival; Pts, number of patients; T, paclitaxel; TTF, time-to-treatment-failure; TTP, time-to-progression; X, capecitabine

Table 3

Gemcitabine alone or in combination in first-line therapy or greater for metastatic breast cancer—Toxicity results of randomized trials

Trial

Treat-ment arms

Rate of grade 3/4 toxicity

Anemia

Leuko-penia

Neutro-penia

Thrombo-cytopenia

Febrile Neutro-penia

Nausea/ vomiting

Diarrhea

Mucositis

Polyneu-ropathy

Hand-foot Syndrome

Chan et al. [5, 6]a

GD

7%

NR

85%

∼11%

8%

9%

∼8%

4%

NR

0%

XD

∼3%

 

82%

3%

13%

5%

18%

17%

26%

O’Shaughnessy et al. [712]b

GT

7%

NR

48%

6%

5%

3%

NR

NR

NR

NR

T

3%

 

11%

0%

2%

3%

Feher et al. [13]b

G

5.3%

NR

25.3%

9.0%

NR

6.9%

2.1%

0%

NR

NR

E

5.7%

17.9%

1.5%

7.3%

0.5%

7.8%

Zielinski et al. [14]

GET

21.1%

74.8%

92.7%

28.5%

12.3%

13.1%

NR

26.1%

4.9%

NR

FEC

7.7%

62.3%

83.8%

3.1%

2.3%

17.7%

1.8%

0.0%

Khoo et al. [15]B

GT1c

8.3%

43.0%

63.9%

18.1%

0.0%

1,4%/2.8%

2.8%

NR

2.8%

NR

GT2c

10.3%

50.0%

57.3%

14.7%

4.4%

0%/2.9%

2.9%

1.5%

GD

23.5%

58.8%

67.7%

22.1%

11.8%

0%/1.5%

13.3%

1.5%

Note: See text for complete dosage and schedule information, except as indicated below. Statistically significant comparisons are in bold text. There were no statistical test results reported for comparisons in italics

a Specific P-values were not reported, but comparisons were reported as statistically significant

b Significance tests were not reported

c GT1 was 1250 mg/m2 gemcitabine plus 175 mg/m2 paclitaxel. GT2 was 1000 mg/m2 gemcitabine and 100 mg/m2 paclitaxel

Abbreviations: C, cyclophosphamide; D, docetaxel; E, epirubicin; F, 5-fluorouracil; G, gemcitabine; NR, not reported; T, paclitaxel; X, capecitabine

Chan et al. [5, 6] reported on a randomized phase III trial of gemcitabine and docetaxel versus capecitabine and docetaxel as first- or second-line therapy in an abstract and oral presentation at the 2005 ASCO meeting. These same results have since been summarized elsewhere [16, 17] with no new analysis. Women (n = 305) were randomized to either gemcitabine 1000 mg/m2 days one and eight and docetaxel 75 mg/m2 day one every 21 days or capecitabine 1250 mg/m2 twice per day for 14 days and docetaxel 75 mg/m2 day one every 21 days. All patients had received anthracycline-containing chemotherapy in either the (neo)adjuvant or metastatic setting. No significant differences were found between the gemcitabine/docetaxel and capecitabine/docetaxel in terms of ORR, TTP, progression-free survival (PFS), or response duration. Significantly lower rates of grade 3/4 hand-foot syndrome, mucositis, and diarrhea were reported in patients treated with the gemcitabine/docetaxel combination.

At the 2006 ASCO Annual Meeting [18, 19] Chan et al. reported on an exploratory analysis of toxicity and quality of life from the same trial. The Rotterdam Symptom Checklist (RSCL) was used to assess quality of life at baseline and with every cycle of chemotherapy. Of the randomized patients, 267 had completed checklists and were available for analysis. Over the first six cycles of chemotherapy, 79% to 88% of checklists were completed. No significant differences in quality of life scores were measured at baseline or in any cycle of therapy. However, attrition due to toxicity was higher in the capecitabine/docetaxel arm, with 28% discontinuing therapy due to serious adverse events versus 13% in the gemcitabine/docetaxel arm.

O’Shaughnessy et al., Albain et al., and Moinpour et al. [712] presented at the 2003 and 2004 ASCO meetings two different interim analyses of a randomized phase III trial of paclitaxel versus gemcitabine and paclitaxel in first-line therapy for women with metastatic breast cancer. The first [710] was a planned interim analysis of TTP, PFS, ORR, toxicity, and quality of life. The second [11, 12] was an unplanned interim analysis done as part of the United States (US) Food and Drug Administration approval process. Women were randomized to 1250 mg/m2 of gemcitabine on days one and eight and 175 mg/m2 paclitaxel on day one every 21 days, or to the same schedule of paclitaxel without gemcitabine. All patients must have received prior adjuvant or neoadjuvant anthracycline therapy. Significantly better ORR, TTP, and overall survival were reported with the gemcitabine/paclitaxel combination. There was also a statistically significant (P-value not reported) interaction between mean global quality-of-life scores compared to baseline on the RSCL instrument and cycle of therapy, with patients receiving the combination reporting higher scores near the end of treatment (cycles 5 and 6) than those receiving paclitaxel alone [9, 10]. The overall survival benefit at the unplanned interim analysis, while significant (P = 0.018), did not meet the planned stopping rule (P  <  0.0001). The combination was associated with a higher rate of neutropenia (48% vs. 11%) over those treated with paclitaxel alone.

Feher et al. [13] reported on a randomized phase III trial of gemcitabine versus epirubicin in the first-line therapy of elderly women. All patients were post-menopausal and age > 60 years. Women were randomized to epirubicin (35 mg/m2 on days one, eight, and 15) or gemcitabine (1200 mg/m2 on days one, eight, and 15) every 28 days. Epirubicin was found to be superior to gemcitabine in terms of ORR, TTP and overall survival.

Zielinski et al. [14] reported on a randomized phase III trial of gemcitabine, epirubicin, and paclitaxel (GET) versus 5-fluorouracil, epirubicin, and cyclophosphamide (FEC) as first-line therapy in women with metastatic disease. Women were randomized to 90 mg/m2 of epirubicin and 175 mg/m2 of paclitaxel on day one, and 1000 mg/m2 of gemcitabine on days one and four or 90 mg/m2 of epirubicin, 500 mg/m2 of 5-fluorouracil, and 500 mg/m2 of cyclophosphamide, all on day one. Both arms used a 21-day cycle regimen and planned eight cycles of therapy. Patients were required to have had one prior non-anthracycline-containing adjuvant chemotherapy. There were no significant differences in any of the reported outcomes between the two arms. However, there were significantly more grade three and four hematological and non-hematological toxicities in the GET arm.

Khoo et al. [15] reported on a randomized phase II trial of three schedules of gemcitabine combined with taxanes: 1250 mg/m2 gemcitabine and 175 mg/m2 paclitaxel (GT1), 1000 mg/m2 gemcitabine and 100 mg/m2 paclitaxel (GT2), and 1000 mg/m2 gemcitabine and 40 mg/m2 docetaxel (GD). All regimens were given every 21 days. ORR, TTP, and response duration were all similar across the arms. All relevant data from this trial can be found in Tables 3 and 4.
Table 4

Gemcitabine alone or in combination in first-line or greater therapy for metastatic breast cancer—Range of patient outcome measures reported by included non-randomized trials

Combination

Number of patients

ORR, percent of patients (median)

Median duration of response, in months (median)

Median TTF/TTP/ PFS, in months (median)

Median OS, in months (median)

G alone

1st-line only [20, 21]

77

14.3–37.1

5.6–8.8

5.1–5.6

15.2–21.1

1st-line and greater [22, 23]

60

25–25

13.5

6.3

11.5–51.9

G + anthracycline

G + A (1st line) [2426]

123

52–57 (55)

5.6–12

4.5–23(11.5)

16.1–27

G + A (1st line and greater) [27]

21

47.6

6

5.8

NR

G + E (1st line and greater) [28, 29]

49

28.6–60

11.2

5.8

17.1

G + taxanes

G + D (1st line only) [3033]

194

59–71 (64.3)

6.1

8.5 −13.6 (10)

 > 15, 22.10

G + D (1st line or greater) [34, 35]

79

36–79

10.3

7

12.7–24.5

G + T (1st line only) [3638]

123

41.6–71 (66.7)

11.5–18(13.4)

7.5–16.6 (11)

19

G + platinum agents

G + cisplatin (1st line only) [39, 40]

61

76.7–80

NR

NR

NR

G + cisplatin (1st line and greater) [41, 42]

36, 58

29–44.4

NR

4.4

9.7

G + other agents

G + V (1st line only) [43]

45

55.5

10.8

9.5

NR

G + V (1st line or greater) [4448]

214

24–48 (43.8)

5.3 −12 (6.2)

4.3–5.0

14–20

G + erlotinib (1st line or greater) [49]

59

14

4.6

2.8

NR

G + pemetrexed (1st line and greater) [50]

59

24

5

3.8

10.3

G + anthracycline + taxane

G + E + T (1st line only) [51, 52]

84

71–92

10.3

10.5–21

25.9

G + A + T (1st line only) [53]

41

82.9

14.1

13.9

26.2

G + A + C (1st line only) [54]

20

80

NR

NR

NR

G + T + H (1st line or greater) [55]

27

78

G + T + H (1st line only) [56, 57]

82

52.5–67

14

9–13.7

∼27

G + D + H (1st line only) [58]

34

56

12

14.6

NR

G + F + Leucovorin (1st line or greater) [59]

27

22

NR

NR

NR

Notes: Median of reported outcome is only provided if at least three studies reported that outcome. Not all studies in a row of the table reported all outcomes

Abbreviations: A, doxorubicin; C, cyclophosphamide; D, docetaxel; E, epirubicin; G, gemcitabine; H, trastuzumab; NR, not reported; ORR, objective response rate; OS, overall survival; PFS, progression-free survival; T, paclitaxel; TTF, time-to-treatment-failure; TTP, time-to-progression; V, vinorelbine

Non-randomized phase II trials

All relevant data from the identified non-randomized trials published prior to August 2005 is summarized in Tables 47.
Table 5

Gemcitabine alone or in combination in first-line therapy or greater for metastatic breast cancer—Range of grade 3 and 4 toxicities reported by included non-randomized trials

Combination

Neutropenia, percent of patients (median)

Febrile Neutropenia, percent of patients (median)

Thrombo-cytopenia, percent of patients (median)

Anemia, percent of patients (median)

Nausea/Vomiting, percent of patients (median)

Stomatitis/Mucositis, percent of patients (median)

Gemcitabine alone

G (1st line only) [20, 21]

11.9–30.3

NR

4.8–6.3

0–9.1

10.3–14.6

0

G (1st line and greater) [22, 23]

15–30.3

0

0–2

0–2

10–27.3

0

G + anthracycline

G + A (1st line) [2426]

66.6–74

0–2.0

16.7–27.5

0–12.7(2.4)

5.7–10.6 (7.3)

8.5–9.8

G + A (1st line and greater) [27]

39

3.5

NR

NR

NR

11

G + E (1st line and greater) [28, 29]

51.5–64

26.7

6–11

5.8

6–11.4

2.9–11

G + taxanes

G + D (1st line only) [3133]

13.8–44 (33)

0–6 (4.2)

6, 6.9

3.4–4

1.7–8.3 (2)

2–3.4

G + D (1st line or greater) [30, 34, 35]

29–100 (49.0)

7.7

0–5.0 (2.6)

2–10.0 (5.1)

5.1–8.0 (6)

0–5.1 (4)

G + T (1st line only) [37, 38]

13.3, 29

2

4, 13.3

0, 0

0

4, 15.5

G + platinum agents

G + cisplatin (1st line only) [39, 40]

20–20

NR

1.7

2.9

3.3–16.9

NR

G + cisplatin (1st line or greater) [41, 42]

10.2–43

NR

35.8–38

5.1

10–30.8

NR

G + other agents

G + V (1st line or greater) [4348]

0–52 (37)

6.3

0–20 (7.9)

0–15.1 (3.8)

0–12 (5)

0–4 (0)

G + erlotinib (1st line or greater) [49]

35

NR

9

7

NR

NR

G + pemetrexed (1st line or greater) [50]

81

12

25

NR

5

NR

G + anthracycline + taxane

G + E + T (1st line only) [51, 52]

92

11.1

28

20

2–8

17–31

G + A + T (1st line only) [53]

43.9

2.4

7.2

12.3

14.6

1.9

G + T + H (1st line or greater) [55]

33

0

2

12

0

0

G + T + H (1st line only) [56, 57]

25–66.7

NR

12.5–20

5–8.9

8.9

2.5

G + D + H (1st line only) [58]

64

6

3

6

NR

NR

G + F + Leucovorin (1st line or greater) [59]

27

NR

17

NR

NR

NR

Notes: Median of reported toxicities is only provided if at least three studies reported that toxicity. Not all studies in a row of the table reported all outcomes

Abbreviations: A, doxorubicin; C, cyclophosphamide; D, docetaxel; E, epirubicin; G, gemcitabine; NR, not reported; T, paclitaxel; V, vinorelbine

Table 6

Gemcitabine alone or in combination in second-line or greater therapy for metastatic breast cancer—Range of patient outcome measures reported by included non-randomized trials

Combination

Number of patients

ORR, percent of patients (median)

Median duration of response, in months(median)

Median TTF/TTP/ PFS, in months (median)

Median OS, in months (median)

G [6066]

186

0–42 (20)

8.5–9

1.9–3.6 (3)

6.8–18.6 (8.8)

G + A [67]

31

29

NR

7.0

 > 11

G + D [30, 6871]

234

43–54 (52.3)

3.6–7.8 (6.1)

6.6–8 (7.5)

15–16.5

G + T [71, 72]

169

48–55 (52.2)

8–8.3 (8.2)

7.0–7.5

12

G + carboplatin [7377]

93

21.5–69.2 (30)

8.6

4.8–5

7.5

G + cisplatin [7884]

171

10–62.5 (39.3)

5.3

3–11.2 (4.8)

9.43–11.9 (11.5)

G + V [43, 8591]

270

10–54 (31.3)

4–7.4 (6)

3–8.3 (6)

9.5 −17.8 (11.5)

G + V + H [92]

23

39.1

NR

6

9

G + H [93, 94]

89

36–38

5.8

5.8–7.8

14.7–18.7

G + C [95]

12

8

NR

NR

NR

G + capecitabine [9698]

73

21.4–52 (33.3)

4.3

4.2–10.5(8.8)

12.0–17

G + irinotecan [99]

38

18

5

8

8.5

G + mitoxantrone [100, 101]

86

24–30

 

4.4–5.1

9.8–13.3

G + F + C + leucovorin [102]

69

38

NR

6

13

Notes: Median of reported outcome is only provided if at least three studies reported that outcome. Not all studies in a row of the table reported all outcomes

Abbreviations: A, doxorubicin; C, cyclophosphamide; D, docetaxel; E, epirubicin; G, gemcitabine; H, trastuzumab; NR, not reported; ORR, objective response rate; OS, overall survival; PFS, progression-free survival; T, paclitaxel; TTF, time-to-treatment-failure; TTP, time-to-progression; V, vinorelbine

Table 7

Gemcitabine alone or in combination in second-line or greater therapy for metastatic breast cancer—Range of grade 3 and 4 toxicities (percent of patients) reported by included non-randomized trials

Combination

Neutropenia, percent of patients (median)

Febrile Neutropenia, percent of patients (median)

Thrombo- cytopenia, percent of patients (median)

Anemia, percent of patients (median)

Nausea/ Vomiting, percent of patients (median)

Stomatitis/ Mucositis, percent of patients (median)

G [6062, 64, 66]

14–63.8 (24)

NR

5–20 (9)

4 −12 (6.5)

0–14 (4.5)

0

G + A [67]

32

NR

9

NR

NR

NR

G + D [6871]

29–68 (40.5)

9–11

6–22 (17.5)

2–23 (7)

0–4(2)

0–6 (2)

G + T [71]

58–64

0–4

15–18

8–10

NR

NR

G + carboplatin [73, 7577]

50–60 (50)

3.6–10

21–40 (30)

20–38.5 (27)

NR

NR

G + cisplatin [78, 8284]

4.5–35 (22.0)

0

0–19.5 (15)

0–41.6 (6.0)

0–21 (12.1)

0

G + V [8591, 103]

8.7–48 (12.5)

0

0–6 (3)

0–6.25 (3.5)

0–5.9 (3.4)

0–6 (2.2)

G + V + H [92]

8.7

NR

4.3

4.3

NR

NR

G + H [93, 94]

11–28.1

1.6

11

7

1.6–4

NR

G + C [95]

50

0

0

0

0

16

G + capecitabine [9698]

4.3–59 (29)

NR

0–4.3

3–8.7

2.4

4.3

G + irinotecan [99]

34

8

5

NR

NR

NR

G + mitoxantrone [100, 101]

35–41

4

7.5–8.5

5–13

 

5

G + F + C + leucovorin [102]

38

NR

16

6

3

1

Notes: Median of reported toxicities is only provided if at least three studies reported that toxicity. Not all studies in a row of the table reported all outcomes

Abbreviations: A, doxorubicin; C, cyclophosphamide; D, docetaxel; E, epirubicin; G, gemcitabine; NR, not reported; T, paclitaxel; V, vinorelbine

Ongoing trials

A search was made of the National Cancer Institute’s clinical trials database on the Internet (http://www.cancer.gov/search/clinical_trials/) for ongoing randomized clinical trials of gemcitabine alone or in combination in metastatic breast cancer. Three were identified: NCT00236899, comparing two different schedules of the combination of gemcitabine with a taxane [104]; NCT00191243, comparing docetaxel to docetaxel plus gemcitabine [105]; and NCT00294385, comparing concurrent docetaxel and gemcitabine with sequential docetaxel followed by gemcitabine [106]. A search of the Cochrane Central Register of Controlled Trials did not identify any additional ongoing trials. In addition, a preliminary report of a randomized trial comparing the combination of vinorelbine and gemcitabine against capecitabine in patients with advanced breast cancer pretreated with taxane and anthracycline chemotherapy was presented at the 2006 ASCO Annual Meeting [107].

Discussion

Gemcitabine has been studied extensively in the phase II setting in women with advanced breast cancer. While gemcitabine has been shown to be efficacious and has generally been well tolerated, there appears to be no particular advantage of gemcitabine over existing chemotherapeutic agents in the third-line or greater setting.

Four randomized phase III trials of gemcitabine-based chemotherapy in women with advanced breast cancer have been reported in the literature. The results of these randomized trials, while difficult to compare directly, suggest that gemcitabine as a single agent is inferior to standard anthracycline-based chemotherapy in patients who are anthracycline naïve. The greatest benefit to be derived from gemcitabine in women with metastatic breast cancer is achieved when it is administered in the first- or second-line setting with a taxane.

In particular, the phase III study by Chan et al. [5, 6] demonstrated that gemcitabine plus docetaxel was as efficacious as the standard arm of capecitabine and docetaxel in the first- or second-line setting, with significantly reduced toxicity. The trial by O’Shaughnessy et al. [108], combining capecitabine and docetaxel, is one of the few phase III trials in the literature to have reported an overall survival advantage in women with advanced breast cancer. The utility of this regimen, however, is hampered by the significant toxicities seen clinically, especially hand-foot syndrome and mucositis. Thus, the combination of gemcitabine and docetaxel may be a better-tolerated alternative to the capecitabine–docetaxel regimen.

The results of the gemcitabine plus paclitaxel trial [712], showing the superiority of gemcitabine/paclitaxel over paclitaxel alone, led to the approval by the US Food and Drug Administration and the recent approval by Health Canada of this combination for women with metastatic breast cancer after failure of prior anthracycline-containing adjuvant therapy. Single-agent paclitaxel has generally not been considered as efficacious as single-agent docetaxel [109] in the treatment of women with metastatic breast cancer. Docetaxel has been the preferred taxane for women with metastatic breast cancer, particularly in Canada and Europe. The randomized phase II trial by Khoo et al. [15], however, suggests that the choice of taxane, paclitaxel or docetaxel, may not make any meaningful difference in the efficacy of gemcitabine plus taxane combinations.

While gemcitabine appears to be generally well tolerated when administered with a taxane doublet, the trial reported by Zielinkski et al. [14] demonstrated equal efficacy with significantly higher hematological toxicity in patients treated with a gemcitabine/anthracycline/taxane triplet (GET) over those treated with FEC chemotherapy. This trial suggests that there is no additional benefit, and more toxicity, to the addition of a third chemotherapeutic agent to a gemcitabine/taxane doublet.

The currently available data do not support the acceptance of gemcitabine as a standard therapeutic option in women with metastatic breast cancer in the third-line or greater setting. Gemcitabine should not be considered as first-line therapy in women with metastatic breast cancer who are anthracycline naïve. Gemcitabine appears to be most effective when administered with a taxane (docetaxel/paclitaxel) in the first- or second-line setting. Gemcitabine/taxane combinations represent a viable alternative to currently accepted taxane combinations such as capecitabine/docetaxel. There is no evidence at the present time to support the use of gemcitabine triplets, given the equal efficacy to anthracycline triplets and the added toxicity.

Copyright information

© Springer Science+Business Media, LLC 2007