A Randomized Phase III Study Comparing Concurrent Gemcitabine plus Cisplatin and Radiation followed by Adjuvant Gemcitabine plus Cisplatin versus Concurrent Cisplatin and Radiation in Patients with Stage IIB to IVA Carcinoma of the Cervix

Reviewer: Christine Hill-Kayser, MD
Abramson Cancer Center of the University of Pennsylvania
Last Modified: June 10, 2009

Presenter: Alfonso Duenas-Gonzalez
Presenter's Affiliation: Unidad de Investigacion Biomédica en C?ncer, Mexico City, Mexico
Type of Session: Scientific


  • Cervical cancer is the second-leading cause of cancer-related death in women around the world. Although pap testing has led to improved early detection and decreased rates of advanced disease in many countries, a large population of largely underserved patients remains at risk for life-threatening disease.
  • Concurrent chemoradiation is the standard of care for treatment of locally advanced cervical cancer, with the most commonly used regimen being weekly cisplatin with concurrent radiotherapy, generally delivered both as external beam treatment and as brachytherapy.
  • Despite aggressive treatment, 5-year overall survival rates in the US remain as low as 55%, with high rates of distant recurrence following treatment.
  • The study described here was designed in order to investigate the possible benefit of addition of gemcitabine to current regimens. Gemcitabine was chosen as an experimental agent in this study because of its recognized radiosensitization effect.
    • Phase I-II trials have confirmed a dose of 125 mg/ m2 of gemcitabine in combination with weekly cisplatin, 40 mg/ m2 and radiation to be safe in patients with FIGO stage IIA-IVA cervical cancer (Zarba, 2003).
    • Improved pathologic response rate with this regimen compared to cisplatin and radiotherapy alone has also been demonstrated previously (Duenas-Gonzalez, 2005).

Materials and Methods

  • Patients were randomized in an open-label fashion to one of two arms:
    • Arm A: Gemcitabine (125 mg/ m2 weekly) + cisplatin (40 mg/ m2 weekly) + external beam radiotherapy to 50.4 Gray (Gy) in 1.8 Gy daily fractions, followed by brachytherapy (low dose rate, 30-35 Gy), followed by two 21 day cycles of adjuvant gemcitabine (1 g/ m2 days 1 and 8) + cisplatin (50 mg/ m2 day 1) given every 3 weeks.
    • Arm B: Cisplatin (40 mg/m2 weekly) + external beam radiotherapy to 50.4 Gy in 1.8 Gy daily fractions, followed by brachytherapy (low dose rate, 30-35 Gy).
  • The primary endpoint was 3 year progression free survival (PFS).
  • Secondary endpoints were overall survival, tumor response, and toxicity.
  • Eligibility requirements included the following:
    • Histologic diagnosis of squamous cell carcinoma, adeno/squamous cell carcinoma, or adenocarcinoma of the cervix.
    • Measurable FIGO stage IIB to IVA disease
    • No evidence of disease in para-aortic lymph nodes
      • FNA performed if notes greater than 1 cm on imaging evaluation.
    • Karnofsky Performance Status (KPS) score of at least 70
    • Normal hematologic, liver, and kidney evaluation, with hemoglobin of at least 10 g/dL.
    • Age range of 18 – 70 years, with use of approved contraception during the study if of child-bearing age.
  • The study was initially designed to detect overall survival; however, this was limited to progression free survival due to slow accrual.


  • Between May, 2002 and March, 2004, 515 patients were enrolled in this study, 259 being randomized to arm A, and 256 to arm B. Patient characteristics did not differ significantly between the two groups, and were as follows:
    • Median age 46 years
    • Median KPS 90
    • Disease stage IIB (n = 316), IIIA (n = 2), IIIB (n = 188), IVA (n = 9)
    • Median hemoglobin 12 g/dL
  • Patients were recruited from Pakistan, Thailand, Argentina, Panama, Peru, Mexico and India, without significant variation in country of origin between the two arms.
  • Patients were equally likely to complete the planned course of radiotherapy, including brachytherapy, between the two arms; however, those in the experimental arm were more likely to have had a longer median duration of radiotherapy (49 total days versus 45, p < 0.0001).
  • The median number of cycles of concurrent cisplatin received was 5 in the experimental arm versus 6 in the control arm.
  • Patients randomized to receive adjuvant chemotherapy received dose 1 in 83.6% of cases and dose 2 in 76.5%.
  • 3 year PFS was significantly increased in the experimental arm, being 74.4% in the group receiving gemcitabine and 65% in the group receiving standard treatment (p = 0.029) (HR 0.68 favoring experimental arm).
  • 3 year OS was also significantly improved in the group receiving gemcitabine (78.2% vs. 69.1%, HR 0.68).
  • A non-significant trend for improved local control was seen in the group receiving gemcitabine, with 29 local failures in arm A, and 42 in arm B (p = 0.097). Decreased distant failures were also observed in arm A (n = 21) versus arm B (n = 42), p = 0.005.
  • More adverse events and toxicities were observed in the group receiving gemcitabine:
    • Two drug-related deaths were observed in arm A, versus 0 in arm B.
    • 33 serious adverse events were observed in arm A versus 12 in arm B (p = 0.005).
      • 18 adverse events led to discontinuation in arm A versus 1 in arm B (p = 0.001)
      • 30 adverse events led to hospitalization in arm A versus 11 in arm B (p = 0.003)
    • Grade 3/4 toxicities were observed in 215 patients in arm A versus 108 in arm B (p < 0.001).
      • Specific toxicities included neutropenia, anemia, thrombocytopenia, diarrhea, vomiting, and proctitis, all of which were more common in arm A.
      • During the chemoradiation part of treatment alone, hematologic toxicity and diarrhea were the most coming toxicities encountered, also increased in arm A.

Author's Conclusions

  • The authors conclude that the addition of gemcitabine to standard concurrent cisplatin and radiation improves survival outcomes in patients with locally advanced cervical cancer.
  • They note that outcomes in control arm of this study were consistent with previously published results, further validating the findings reported here.
  • They note that the significant decrease in distant failures in the group receiving gemcitabine may be important as distant metastatic disease remains an important cause of death in this population.
  • They comment that, as expected, the gemcitabine-based regimen was associated with significantly increased toxicity compared with standard treatment, but that survival outcomes remained higher with the more toxic regimen.

Clinical/Scientific Implications

  • The authors have performed an important and well-designed study. Cervical cancer remains a prominent cause of death worldwide, and continued advances in its treatment are important.
  • As the authors originally intended, overall survival data would be the most interesting outcome to observe from a study such as the one reported here. Unfortunately, this study was not powered to detect a difference in overall survival, although one appears to be present from the data presented. Progression free survival is of course more difficult to measure accurately.
  • The increased toxicity observed as part of this study is of utmost importance. In many disease sites, trials using concurrent gemcitabine and radiation have been terminated early due to unacceptable toxicity, and two of these have been trials treating cervical cancer patients (Swisher, 2006; Rose, 2007). Although the results presented here are enticing, the potential of lethal toxicity with gemcitabine delivered concurrent with radiotherapy must be kept in mind.
  • Along these lines, future trials might examine the impact of the addition of gemcitabine delivered either neoadjuvantly or adjuvantly, rather than concurrent with radiation. Potentially, such a regimen might allow the benefits of gemcitabine with regard to distant control with reduction of toxicity risk.  

See Patient Summary