Cervical cancer is the third most common gynecologic malignancy in the United Sates and the leading gynecologic cancer in the world.1,2 A majority of cervical cancer patients in the United States will present with early-stage disease. It is well established that women with early-stage cervical cancer (at least IB) may be treated with curative intent with either radical surgery or radiotherapy (RT).3 Because both treatment modalities provide an equivalent cure rate, decisions about definitive therapy can be individualized to the patient with regard to their other comorbidities and potential treatment adverse effects.4 Younger premenopausal women with stage I disease are more likely to have surgery for their primary treatment because of both the risk of second malignancy and to preserve ovarian function. Patients treated with surgery who are subsequently found to have high-risk disease (2 of the following 3 features: size of ≥4 cm, lymphovascular invasion, deep stromal invasion, or any of the following features: positive margins, parametria, or nodes) are recommended to receive adjuvant RT with or without chemotherapy.5–8
Previous studies have shown that there is significant variation in patterns of care for early-stage cervical cancer by race, especially in regard to adjuvant RT therapy.9,10 In addition, Howell et al11 noted significant differences in treatment received and overall survival (OS) based on race after controlling for other known prognostic factors in patients with all stages of cervical cancer. Also, marital status has been shown to be a significant predictor of not only receiving appropriate treatment, but also survival in various prior studies in different cancer sites.12,13 The goal of the current study was to assess the impact of race and marital status on patterns of care and survival for young women with stage I cervical cancer. A secondary objective was to examine the use of adjuvant RT treatment in women who underwent surgical resection.
FIGO (International Federation of Gynecology and Obstetrics) stage I cervical cancer data from the Surveillance, Epidemiology, and End Results (SEER) 18 database were reviewed. The SEER program sponsored by the National Cancer Institute collects patient and disease data from varied geographic areas representing more than a quarter of the US population.9 SEER database information is deidentified; therefore, approval by an ethics committee (institutional review board) and informed consent by the study participants were not necessary to perform the analyses.
Patients with a primary diagnosis of cervical cancer aged 15 to 39 years and undergoing treatment from 1988 to 2007 were identified. Only patients who received RT and/or at minimum total hysterectomy were included in this cohort. Patients with local tumor excision only such as loop electrocautery excision procedure, trachelectomy, or cone biopsy were excluded. Women with adenocarcinoma, squamous cell carcinoma, and adenosquamous carcinoma of the cervix were included. Race was categorized as white or nonwhite, which included black and other (American Indian, Alaska native, Asian/Pacific Islander). “Other” and “black” were combined because of overall small number of “other” patients. Hispanic ethnicity is tabulated separately from race in SEER, and a number of patients do not have confirmed ethnicity data. Therefore, in this study, we decided to use race data available in SEER (black, white other), which had no missing data points. Criteria for high-risk disease required for postoperative RT are 2 of the following: size 4 cm or greater, lymphovascular invasion, or deep stromal invasion, and for chemoradiation are any of the following: positive margins, parametria, or nodes. Only disease characteristics available in SEER are nodal status and size. We defined high-risk disease as size 4 cm or greater or positive nodes to have a more sensitive estimate of high-risk features, although there is a risk for overestimation of percentage of women with high-risk features with this definition. Socioeconomic, demographic, and clinical covariates were recorded by race, including marital status, median household income, percentage with high school education or more, year of diagnosis, histology, primary treatment, lymph node status, FIGO 1988 (clinical) and AJCC (American Joint Committee on Cancer) third and sixth editions (pathological) staging, size of tumor, SEER registry, and vital status. Year of diagnosis was divided in 5-year groups (1988–1992, 1993–1997, 1998–2002, 2003–2007), and SEER registry location was divided into 4 geographic regions (northeast, south, midwest, and west). Extent of disease codes used for FIGO (clinical) staging does not separate IB1 from IB2 FIGO stage in patients with IB stage disease. Patients who received both surgery and RT were included in the surgery group. Racial patterns of primary treatment and impact of marital status and other socioeconomic/disease-related characteristics were evaluated. Factors associated with adjuvant treatment and pattern of adjuvant treatment were also assessed in this group.
Between the racial groups, categorical variables were compared using the χ2 test, and continuous variables were compared using the Wilcoxon rank sum test. Covariates potentially associated with treatment including marital status, age, race, year of diagnosis, 5-year group of diagnosis, education, income, stage (FIGO for primary treatment analysis and AJCC for adjuvant treatment), histology, grade, and region were incorporated into multivariable analysis using logistic regression. Kaplan-Meier estimates were used to evaluate survival time, and a Cox proportional hazards model was used to estimate the effect of socioeconomic, demographic, and clinical variables on subsequent survival. All reported P values were 2-sided, and statistical significance was set at P < 0.05. All statistical tests were performed using SPSS version 19 statistical software.
Treatment and Disease Characteristics by Race
A total of 6586 patients with FIGO stage I cervical cancer had complete data available for review. Of these, 5447 (89.5%) were identified as white, and 1139 (18.7%) were identified as black or other (American Indian, Alaska native, Asian/Pacific Islander). Of the white women, 5080 (93.3%) underwent surgery, and 367 (6.7%) underwent RT as their primary form of therapy. Of the nonwhite women, 985 (86.5%) underwent surgery, and 154 (13.5%) underwent RT (P < 0.001). Characteristics of the patients by race are outlined in Table 1. Compared with white women, nonwhite women were more likely to be single or divorced, have RT as primary treatment, have higher FIGO stage disease, have node positive disease, and have tumors 4 cm or greater.
Predictors of RT as Primary Treatment
On multivariate analysis (MVA), women with higher than FIGO stage IA1 disease (IA2 OR 3.09 [P = 0.01]; IB odds ratio [OR], 21.41 [P < 0.001]), widowed/single compared with married status (OR, 1.39; P = 0.02), squamous histology compared with adenocarcinoma (OR, 1.69; P < 0.001), diagnosis during 1993–1997 time period compared with other time periods in the study (OR, 1.69; P < 0.001), and nonwhite race compared with white (OR, 1.95; P ≤ 0.001) were more likely to receive RT as primary treatment. Region, education level, income, age, and grade were not significant predictors of receiving RT as primary treatment.
Disease Characteristics and Patterns of Adjuvant Treatment by Race in the Surgical Group
Ninety-three percent of white women had surgery as the primary treatment versus 86.0% of nonwhite women. Table 2 outlines the type of surgery performed for these 6065 women. Of 6065 women, 4013 women (66.2%) were reported to have undergone a lymph node dissection. Of 4013, 986 patients (16.3%) had high-risk disease (either documented node positive disease or tumor size ≥4 cm). Of 6065 patients, nodal status was not available for 317 patients, and size was not available for 2584 patients.
Characteristics of surgical patients by race are described in Table 3. High-risk disease was documented in 15.4% of white women versus 20.4% in the nonwhite women (P < 0.001). Of the high-risk patients, 66% of the white women versus 71.0% of the nonwhite women received adjuvant RT (P = 0.136). Rates of adjuvant treatment in high-risk patients in both racial groups increased from 28% in 1988–1992 time period to 81.3% in 2002–2007 time period (P < 0.001). Significant predictors of receiving adjuvant RT on multivariable analysis included size greater than or equal to 4 cm (OR, 4.46; P < 0.001) and later decades of study (1998–2002 OR, 4.47; P < 0.001; 2003–2007 OR, 6.57; P < 0.001). Marital status, race, region, education level, income, nodal status, grade, and histology were not significant predictors of receiving adjuvant RT.
Predictors of worse OS in all patients were nodal positivity (hazard ratio, [HR] 2.71; P < 0.001), size of 4 cm or greater (HR, 2.35; P < 0.001), RT as primary treatment (HR, 1.89; P < 0.001), earlier year of diagnosis—1988–1992 (HR, 1.49; P = 0.001), and nonwhite race (HR, 1.6; P = 0.001). Marital status was not a significant predictor of survival.
For patients who received radiation as primary treatment, being single (HR, 1.65; P = 0.02) and nodal positivity (HR, 1.67; P = 0.03) were predictors of worse survival. Year of diagnosis, size, stage, and race were not significant predictors of survival.
For patients who received surgery as primary treatment, nonwhite race (HR, 1.93; P < 0.001), FIGO stages 1B (HR, 7.44; P < 0.001) and I not otherwise specified (HR, 9.93; P < 0.001), nodal positivity (HR, 2.24; P < 0.001), and size of 4 cm or greater (HR, 2.46; P < 0.001) were predictors of worse survival, and later year diagnosis 2003-2007 (HR, 0.61; P = 0.02) was a predictor of better survival than earlier year of diagnosis. Marital status was not a predictor of survival in this surgical group.
The overall objective of our study was to determine the impact of race and marital status on patterns of care among young women (<40 years of age) with stage I cervical cancer. In this SEER cohort of young women with stage I cervical cancer, a significantly higher percentage of nonwhite women and single/widowed women received RT for primary treatment. Although the outcomes are reportedly equivalent with surgery versus RT with stage I disease, at least with IB disease,3 younger women generally undergo surgery to avoid late RT-related toxicities and risk of second malignancies and to preserve some ovarian function. These women may, however, receive adjuvant RT if they are found to have high-risk features that would warrant postoperative RT.3 In this SEER cohort, nonwhite women are significantly more likely to have larger tumors and node positive disease, and this likely explains the lower rates of surgery. Other surgical literature has also repeatedly demonstrated that nonwhite patients are less likely to choose operative treatment than white patients.14,15 The rationale for these decisions needs to be explored in future studies.
We noted in our analysis that single and widowed women were less likely to receive surgery as primary treatment of stage I cervical cancer. This could be suggestive of the fact that more single and widowed women choose RT over surgery because of lack of social support to undergo a major surgical procedure. Marital status has been shown to have an impact on receiving appropriate cancer treatment in various studies. Goodwin et al12 have shown that married men and women are more likely to present at an early stage of disease, more likely to get treated, and overall more likely to live longer than their unmarried counterparts. Similarly, it has been noted that in men with head and neck cancer, after controlling for disease-related factors, unpartnered men have poorer survival than do married men.13 In addition, in our analysis, single or widowed women who underwent radiation had significantly inferior survival compared with married women. These results may be explained by the impact of social support on treatment completion and recovery. In a previous study by Patel et al,16 which examined marital status and survival in cervical cancer patients, improved survival among married women compared with unmarried or widowed women was demonstrated. Among women with early-stage disease, married women who did not receive radiation as their primary treatment had higher survival than did nonmarried women. This finding is different from our study likely because of inclusion criteria for women in both studies. For example, any surgical procedure less than hysterectomy was excluded in our study, whereas trachelectomy or local surgical excision was included as local surgery in the study by Patel et al.16 Furthermore, in our study, primary treatment was defined as surgery or radiation, and because there is no mechanism to find out the order of treatment in SEER, all patients who received both were included in the surgery group. There is a potential risk of misclassification here because certain patients who received both surgery and radiation could have received radiation followed by adjuvant hysterectomy. Although marital status was not a predictor of death on multivariable analysis in their study, it did have an influence on cancer stage and cancer treatment. In contrast, in a retrospective analysis by Torres et al17 at M. D. Anderson Cancer Center, single, divorced, or separated patients did not have inferior outcomes or poorer treatment completion rates compared with married patients treated with chemoradiation for locally advanced cervical cancer. Although their cohort of patients was different than those in our SEER analysis, which included community hospitals as well as academic centers, one possible explanation for the difference may be that there are significant social services and support that are available at a major academic medical center and likely vital to receiving appropriate cancer therapy. Further studies are needed in this area given that this population of patients is particularly vulnerable.
In 1999, Sedlis et al,8 and in 2000, Peters et al6 published the results of 2 important phase III studies that supported the addition of adjuvant RT or chemoradiotherapy for high-risk patients who underwent surgical resection for early-stage cervical cancer to improve local control and survival. Within our surgical group, 15% of white women and 20% of nonwhite women were identified as having high-risk disease (positive nodes or size ≥4 cm). Of the high-risk disease patients, rates of adjuvant RT were similar between the white and nonwhite women. Also, rates of adjuvant RT have gone up over time (GOG92 and GOG109 [Gynecologic Oncology Group]/SWOG8797 [Southwest Oncology Group] data).
Race has also been noted to be significantly related to survival in cervical cancer.18 We did not see racial differences in the use of adjuvant treatment as previously noted by Ghia et al.9 This could reflect that we limited our analysis to stage I patients only and did not include IIA, as was done in the analysis by Ghia et al.
The worse OS with RT as primary treatment that was seen in our study could be due to unmeasured confounders and a result of factors associated with receiving RT in our study group. Although our cohort represents a younger population, women receiving RT may have had other comorbidities, making them nonsurgical candidates. Therefore, many of these women may have other competing risks contributing to their worse survival. Given that this is a retrospective analysis of the SEER data set, there are likely biases with respect to the treatment received including physician preference. In addition, women receiving RT upfront were more likely to have higher-risk disease, which in itself is associated with worse survival.19 We also noted that nonwhites had worse OS than did white women, which could be explained by the fact they were more likely to have RT as primary treatment, in part because of a lack of social support limiting them from having large surgical procedure, but most importantly because they presented with overall higher-risk disease.
Race was not a significant predictor of survival in patients who received RT as primary treatment, whereas it was a significant predictor of survival for surgical patients. One could postulate that because there were no differences in survival by race in patients who received upfront RT, inferior survival in the nonwhite group in the surgery-alone patients is not due to differences in tumor biology between races. One possibility is that the follow-up care is worse in nonwhites versus whites because of social issues such as underinsurance and lack of support structure. Other groups have identified the challenges of follow-up in the preinvasive cervical cancer setting20; these challenges could certainly extend into invasive cancer setting. Recurrences after surgery alone are potentially salvageable (with RT) if picked up early on in follow-up versus local recurrences after RT, which may have a lower likelihood of cure as treatment options are limited. Patients who develop metastatic disease in either category will ultimately die regardless of whether metastatic disease is diagnosed early or late, given the limited salvage options. Therefore, follow-up for surgical patients may have the most impact on survival for women with early-stage disease, and the lack of regular follow-up may be one explanation as to why there is a difference in survival between whites and nonwhites.
A number of limitations of our study should be noted. The SEER registry lacks information with regard to known surgicopathologic and treatment-related factors that are important in cervical cancer including lymphovascular invasion, overall treatment time, RT dose, clinical nodal status, and hemoglobin status. In addition, the impact of chemotherapy on survival could not be examined in this study as information with regard to chemotherapy is not available in the SEER registry. Furthermore, as mentioned earlier, there is no way to determine the sequence of treatments in SEER for patients who received both surgery and radiation. Our categorization of all combination patients under the surgery group assumes that these patients underwent surgery as primary treatment followed by adjuvant RT. This could potentially lead to misclassification bias.
In conclusion, surgery is the most commonly used primary treatment modality for all young women with stage I cervical cancer; however, rates of RT as primary treatment are higher in nonwhite compared white women without any differences in survival. For patients who received surgery as primary treatment, nonwhites were more likely to have high-risk features warranting adjuvant RT. Overall, nonwhites had worse survival likely because of more high-risk disease at presentation, likelihood of other comorbidities, and lack of social support.
1. Gray HJ. Primary management of early stage cervical cancer
(IA1-IB) and appropriate selection of adjuvant therapy. J Natl Compr Cancer Netw
. 2008; 6: 47–52.
2. Greer BE, Koh WJ, Abu-Rustum NR, et al.. Cervical cancer
. J Natl Compr Cancer Netw
. 2010; 8: 1388–1416.
3. Landoni F, Maneo A, Colombo A, et al.. Randomised study of radical surgery versus radiotherapy for stage IB-IIA cervical cancer
. 1997; 350: 535–540.
4. Undurraga M, Loubeyre P, Dubuisson JB, et al.. Early-stage cervical cancer
: is surgery better than radiotherapy? Expert Rev Anticancer Ther
. 2010; 10: 451–460.
5. Keys HM, Bundy BN, Stehman FB, et al.. Cisplatin, radiation, and adjuvant hysterectomy compared with radiation and adjuvant hysterectomy for bulky stage IB cervical carcinoma. N Engl J Med
. 1999; 340: 1154–1161.
6. Peters WA 3rd, Liu PY, Barrett RJ 2nd, et al.. Concurrent chemotherapy and pelvic radiation therapy compared with pelvic radiation therapy alone as adjuvant therapy after radical surgery in high-risk early-stage cancer of the cervix. J Clin Oncol
. 2000; 18: 1606–1613.
7. Rotman M, Sedlis A, Piedmonte MR, et al.. A phase III randomized trial of postoperative pelvic irradiation in stage IB cervical carcinoma with poor prognostic features: follow-up of a Gynecologic Oncology Group study. Int J Radiat Oncol Biol Phys
. 2006; 65: 169–176.
8. Sedlis A, Bundy BN, Rotman MZ, et al.. A randomized trial of pelvic radiation therapy versus no further therapy in selected patients with stage IB carcinoma of the cervix after radical hysterectomy and pelvic lymphadenectomy: a Gynecologic Oncology Group study. Gynecol Oncol
. 1999; 73: 177–183.
9. Ghia AJ, Neeley ES, Gaffney DK. Postoperative radiotherapy use and patterns of care
analysis for node positive or parametria positive cervical cancer
. Gynecol Oncol
. 2010; 119: 411–416.
10. Lee CM, Szabo A,, Shrieve DC, et al.. Descriptive nomograms of adjuvant radiotherapy use and patterns of care
analysis for stage I and II endometrial adenocarcinoma: a surveillance, epidemiology, and end results population study. Cancer
. 2007; 110: 2092–2100.
11. Howell EA, Chen YT, Concato J. Differences in cervical cancer
mortality among black and white women. Obstet Gynecol
. 1999; 94: 509–515.
12. Goodwin JS, Hunt WC, Key CR, et al.. The effect of marital status
on stage, treatment, and survival of cancer patients. JAMA
. 1987; 258: 3125–3130.
13. Konski AA, Pajak TF, Movsas B, et al.. Disadvantage of men living alone participating in Radiation Therapy Oncology Group head and neck trials. J Clin Oncol
. 2006; 24: 4177–4183.
14. Arega A, Fau-Birkmeyer NJO, Lurie JDN, et al.. Racial variation in treatment preferences and willingness to randomize in the Spine Patient Outcomes Research Trial (SPORT). 2006; 31: 2263–2269.
15. Merrill RM, Fau-Merrill AV, Merrill AV, Fau-Mayer LS, Mayer LS. Factors associated with no surgery or radiation therapy for invasive cervical cancer
in black and white women. Ethn Dis.
2000; 10: 248–256.
16. Patel MK, Patel DA, Lu M, et al.. Impact of marital status
on survival among women with invasive cervical cancer
: analysis of population-based Surveillance, Epidemiology, and End Results data. J Low Genit Tract Dis
. 2010; 14: 329–338.
17. Torres MA, Jhingran A, Thames HD, et al.. Concurrent chemoradiation in the routine management of patients with cervical cancer
: does marital status
matter? Int J Gynecol Cancer
. 2009; 19: 1107–1112.
18. Patel DA, Barnholtz-Sloan JS, Patel MK, et al.. A population-based study of racial and ethnic differences in survival among women with invasive cervical cancer
: analysis of Surveillance, Epidemiology, and End Results data. Gynecol Oncol
. 2005; 97: 550–558.
19. Liu MT, Hsu JC, Liu WS, et al.. Prognostic factors affecting the outcome of early cervical cancer
treated with radical hysterectomy and post-operative adjuvant therapy. Eur J Cancer Care (Engl)
. 2008; 17: 174–181.
20. Eggleston KS, Fau-Coker AL, Das IP, et al.. Understanding barriers for adherence to follow-up care for abnormal pap tests. J Womens Health
. 2007; 16: 311–330.