Abstract
Background. It was the purpose of this study to investigate whether Pacific Island (PI) ethnicity, Micronesian and Polynesian, is an independent prognostic factor in the survival of cervical cancer in a health care system with minimal racial bias and few barriers to access to care.
Methods. Records from 1988 to 1999 were reviewed for the U.S. Military Health Care System. The medical records of women with the diagnosis of invasive cervical cancer were abstracted and clinical data recorded. A cohort analysis based on Pacific Island ethnicity was also performed on all patients treated at Tripler Army Medical Center (TAMC) during this time period. Significant differences in distribution of clinical factors were determined by Wilcoxon rank-sum test and survival analyses were performed using Kaplan–Meier actuarial statistics.
Results. A total of 153 patients were identified who were treated at TAMC; 74 were of PI ethnicity. An additional 1400 patients were identified throughout the military health care system during this time. Forty-eight percent of non-PI TAMC patients were Caucasian, 14% Filipino, and 13% Korean. The mean age of PI was 45 versus 40 years for their non-PI counterparts. There was no difference in the distribution of the grade of tumors among cohorts analyzed. Seventy-five percent of non-PI patients presented at an early stage while 74% of PI women presented at an advanced stage. Twenty-three percent of PI patients had positive lymph nodes, versus 7% of non-PI patients. There was no difference in the radiation dosages among patients treated with primary radiation therapy. PI patients had a significantly decreased 5-year survival, 32% versus 71%, compared to their cervical cancer patient counterparts, P < 0.001. Multivariate analysis revealed PI ethnicity to be a significant independent predictor of decreased survival, P < 0.001.
Conclusion. PI women diagnosed with cervical cancer tend to present at an advanced age and stage with metastatic disease. They have a decreased survival that remains present after adjusting for age, stage, and grade. The poor prognosis is likely due to lack of uniform screening among this population; however, molecular etiologies and human papillomavirus could also contribute to decreased survival.