Is Anemia a Cause of Radiation Treatment Failure in Patients with Squamous Carcinoma of the Cervix?

Reviewer: Voika BarAd, MD
Abramson Cancer Center of the University of Pennsylvania
Last Modified: October 19, 2005

Presenter: P.J. Eifel
Presenter's Affiliation: Department of Radiation Oncology, U.T. M.D. Anderson Cancer Center, Houston, TX, Gynecologic Oncology
Type of Session: Scientific


  • Previous studies have shown that anemia is an independent poor prognostic factor in cervical cancer.
  • In addition, anemia may affect a patient's quality of life in ways that impact upon patient compliance with and/or ability to tolerate therapy.
  • The basis for the impact of anemia on prognosis or outcome of cancer is complex. In vitro and animal models have shown that cellular hypoxia, the consequence of anemia, may help select for tumor cells that have a higher rate of mutation, ultimately resulting in increased metastatic potential, increased cellular growth, and increased therapy resistance.
  • There is also evidence to indicate that anemia itself may induce a feedback mechanism that results in angiogenesis.
  • Finally, the effect of anemia on the pharmacokinetics of chemotherapy agents may be an underestimated parameter for therapeutic outcome.
  • The purpose of the study was to determine whether the relationship between anemia and outcome of cervical cancer patients treated with radiation therapy is independent of tumor volume and other clinical factors.

Materials and Methods

  • The medical records of all the patients treated at M.D. Anderson Cancer Center between 1960 and 2001 for stage I and II squamous carcinomas of the cervix were reviewed retrospectively.
  • Patients who had initial or adjuvant hysterectomy, or who were treated with palliative intent only, were excluded.
  • The remaining 2988 patients formed the basis of this study.
  • All patients were treated with radiotherapy only, with curative intent.
  • Four anemia-related variables (pre-RT hemoglobin, or Hgb; minimum Hgb during RT; transfusion before RT; and transfusion during RT) and three descriptive tumor variables (clinical tumor diameter; clinical lymph node status; and FIGO stage) were recorded.
  • Tumor diameters were recorded in 94% of patients and were categorized as < 4 cm (38% of patients), 4- 4.9 cm (14% of patients), 5- 5.9 cm (18% of patients), 6- 6.9 cm (11% of patients), or >/= 7 cm (13% of patients).
  • 1670 patients had tumor size > 4 cm, and the size was analyzed as a continuous variable in this subgroup.
  • Comparisons between actuarial curves were made using the log-rank method.
  • A proportional hazards regression model was used to evaluate the relative importance of predictive factors.


  • The median follow up of surviving patients was 14 years.  
  • The minimum Hgb during RT was <10 in 19% of patients, 10- 10.9 in 19% of patients, 11- 11.9 in 27% of patients, 12- 12.9 in 21% of patients, >/= 13 in 10% of patients, and unknown in 3% of patients, respectively.
  • Transfusions were administered before RT in 8.3% and during RT in 12.8% of patients.
  • The rates of central disease recurrence, pelvic recurrence, distant metastasis, and disease-specific survival were all strongly correlated with each of the four anemia-related and three descriptive tumor variables.
  • However, patients with large tumors, advanced stage (IIB), or clinically positive nodes, were significantly more likely to have transfusions or low Hbg before or during RT.
  • Minimum Hgb during RT was < 10 in 9%, 15%, 26%, 32%, and 39% of patients whose tumors were <4, 4- 4.9, 5- 5.9, 6-6.9, or >/= 7cm in diameter, respectively.
  • On Cox multivariate analysis, none of the four measures of anemia had a significant independent correlation with the central, pelvic, or distant recurrence rates. Only the tumor size and lymph node status were independent predictors of central recurrence.
  • Disease-specific survival was independently correlated with tumor size, lymph node status, and FIGO stage, and marginally correlated with transfusion during RT. However, there was no significant relationship between disease-specific survival and pre-RT Hbg, pre-RT transfusion, or minimum Hbg during RT.

Author's Conclusions

  • In this large, single institutional study, anemia was not an independent predictor of poor outcome in patients with stage I or II cervical cancer.
  • In previous studies that did not consider the influence of tumor size or that dichotomized this variable, anemia may have acted as a surrogate for advanced disease.
  • The data presented here suggest that anemia is a symptom, rather than a cause, of radiation treatment failure in stage I or II cervical cancer.

Clinical/Scientific Implications

  • A large number of studies have demonstrated that anemia is an independent poor prognostic factor in several cancer sites, such as head and neck, lung, and cervical cancers.
  • Patients with carcinoma of the cervix often experience anemia before and during radiation treatment, and its origin is complex, representing a combination of hemorrhages, iron deprivation, inflammatory reactions and infection. Concomitant chemotherapy regimens with cisplatin are standard, and they can largely increase the risk of inducing anemia.
  • The frequency of anemia (hemoglobin < level 12g/dl) is correlated with the clinical stage of cervical cancer and varies from one publication to another.
  • Anemia is correlated with patient survival and appears to be one of the most powerful prognostic factors in cervical cancer, after clinical stage and tumor size, in most studies.
  • In this large, single institutional study, anemia was not a poor independent prognostic factor in stage I or II cervical cancer. The presenter concluded that the old theory, that severe anemia causes poor results of radiotherapy in cervical cancer, should perhaps be reversed to a new hypothesis, that radiotherapy causes severe anemia. Anemia seems to act as a surrogate for advanced disease (large tumor size, biologically resistant tumors, and poorly responsive tumors), rather than as an independent prognostic factor.
  • Ultimately, whether it is an independent or non-independent poor prognostic factor, it is clear to everyone that ANEMIA does not have a good augur in cervical cancer.
  • Large controlled trials are needed to clarify the role of this important confounding variable in the treatment and outcome of patients with cervical cancer.