Cervical carcinoma in the elderly

An analysis of patterns of care and outcome

Authors

  • Jason D. Wright M.D.,

    Corresponding author
    1. Division of Gynecologic Oncology, Department of Obstetrics and Gynecology, Washington University School of Medicine, St. Louis, Missouri
    • Division of Gynecologic Oncology, Department of Obstetrics and Gynecology, Washington University School of Medicine, 4911 Barnes Hospital Plaza, Box 8064, St. Louis, MO 63110
    Search for more papers by this author
    • Fax: (314) 362-2893

  • Randall K. Gibb M.D.,

    1. Division of Gynecologic Oncology, Department of Obstetrics and Gynecology, Washington University School of Medicine, St. Louis, Missouri
    Search for more papers by this author
  • Sajeena Geevarghese M.D.,

    1. Division of Gynecologic Oncology, Department of Obstetrics and Gynecology, Washington University School of Medicine, St. Louis, Missouri
    Search for more papers by this author
  • Matthew A. Powell M.D.,

    1. Division of Gynecologic Oncology, Department of Obstetrics and Gynecology, Washington University School of Medicine, St. Louis, Missouri
    Search for more papers by this author
  • Thomas J. Herzog M.D.,

    1. Division of Gynecologic Oncology, Department of Obstetrics and Gynecology, Washington University School of Medicine, St. Louis, Missouri
    Search for more papers by this author
  • David G. Mutch M.D.,

    1. Division of Gynecologic Oncology, Department of Obstetrics and Gynecology, Washington University School of Medicine, St. Louis, Missouri
    Search for more papers by this author
  • Perry W. Grigsby M.D.,

    1. Department of Radiation Oncology, Washington University School of Medicine, St. Louis, Missouri
    Search for more papers by this author
  • Feng Gao Ph.D.,

    1. Division of Biostatistics, Washington University School of Medicine, St. Louis, Missouri
    Search for more papers by this author
  • Kathryn M. Trinkaus Ph.D.,

    1. Division of Biostatistics, Washington University School of Medicine, St. Louis, Missouri
    Search for more papers by this author
  • Janet S. Rader M.D.

    1. Division of Gynecologic Oncology, Department of Obstetrics and Gynecology, Washington University School of Medicine, St. Louis, Missouri
    Search for more papers by this author

  • See related editorial on pages 1–4, this issue.

Abstract

BACKGROUND

Advanced age often is considered a poor prognostic factor for cervical carcinoma. The authors investigated the patterns of care and treatment outcomes of elderly women with cervical carcinoma.

METHODS

A hospital-based tumor registry was used to identify patients with invasive cervical carcinoma who were treated between 1986 and 2003. Patients were divided into 2 cohorts: women age < 70 years and women age ≥ 70 years. Survival was examined using the Kaplan–Meier method. Single and multivariate Cox proportional hazards modeling was used to estimate hazard ratios with 95% confidence intervals (95% CI).

RESULTS

In total, 1582 patients were identified, including 1385 patients age < 70 years and 197 patients age ≥ 70 years. The elderly patients presented with more advanced stage tumors at diagnosis (P < 0.0001) and were more likely to have nonsquamous neoplasms (P = 0.002). A marked difference in treatment was noted for the elderly cohort, even after stratifying by disease stage. Only 16% of the older patients underwent surgical treatment compared with 54% of the younger patients (P < 0.0001). Elderly women were 9 times more likely to receive no treatment (P < 0.0001). In a multivariate model of known prognostic factors, the hazard ratio for death from any cause in women age > 70 years was 2.1 (95% CI, 1.5–3.0). The hazard ratio for death from cervical carcinoma in the elderly women was 1.6 (95% CI, 1.1–2.5).

CONCLUSIONS

Age is an important factor in the allocation of treatment and survival for patients with cervical carcinoma. Elderly women with cervical carcinoma are more likely to receive primary radiotherapy, to forego treatment, and to die from their disease. Cancer 2005. © 2004 American Cancer Society.

Despite the availability of an effective screening test, cervical carcinoma remains an important cause of morbidity and mortality among women. Worldwide, cervical carcinoma is the second most common malignancy in women and accounts for 190,000 deaths annually.1 The age distribution of new cases of cervical carcinoma is bimodal, with peaks at ages 30–39 years and at ages 60–69 years.2 Although cervical carcinoma is relatively common in older women, the impact of age on survival remains uncertain. Whereas much of the literature evaluating the influence of age on survival has been conflicting, more recent population-based studies have suggested that advanced age is an adverse prognostic factor.

With the aging of the population, increased attention has been focused on the treatment of geriatric patients with cancer. Advanced age has been associated with decreased survival for a variety of tumors. In addition to its impact on survival, age has been described as an important factor in the selection and allocation of treatment for a host of malignancies.3–5 Geriatric patients in general are less likely to receive aggressive primary therapy than their younger counterparts. This trend has been observed for surgery, radiation, and chemotherapy.3–7 To date, a paucity of data to describe the influence of age in the allocation of treatment for patients with cervical carcinoma has been reported. The objective of the current study was to examine the patterns of care and outcome of elderly patients with invasive cervical tumors.

MATERIALS AND METHODS

Data were extracted from the Barnes-Jewish Hospital tumor registry after obtaining study approval from the Washington University School of Medicine Human Studies Committee. Between January 1, 1986, and July 31, 2003, 1582 patients with invasive cervical carcinoma were evaluated and managed. Patient data abstracted included age at diagnosis, race, tumor histology, International Federation of Gynecology and Obstetrics disease stage, date of treatment, type of primary and adjuvant therapy, disease status, and vital status.

Tumor histology was classified as squamous, adenocarcinoma, adenosquamous, or other. Treatment was classified as surgery, radiation, chemoradiation, other, or none. During the study period, several Phase II and III chemoradiation trials accrued patients at our institution; thus, patients received a variety of chemotherapeutic regimens. Patients whose treatment consisted of a radical hysterectomy, modified radical hysterectomy, extrafascial hysterectomy, or trachelectomy were included in the surgery category. Those patients who underwent a lymphadenectomy and an aborted hysterectomy were placed in the radiation or chemoradiation category, depending on their subsequent therapy. Patients who were placed in the “other” treatment category included patients who were managed with cervical conization or chemotherapy and patients whose treatment was not recorded. Patients who did not receive any treatment were placed in a separate category. To determine the treatment characteristics of patients whose primary therapy included radiation, data were ascertained from the Department of Radiation Oncology treatment data base. Data abstracted included the total dose to Point A and whether brachytherapy was administered. Patients are only entered into this data base after 5 years of follow-up; thus, complete radiotherapy data are available for only a fraction of the cohort.

Beginning in 1996, data on the severity of comorbid medical conditions among patients entered into the tumor registry have been collected prospectively. A modified version of the Kaplan–Feinstein Index was utilized to classify patient comorbidity as none (0), mild (1), moderate (2), or severe (3). The overall comorbidity score is determined by the number and severity of individual comorbidities. A detailed description of the modified Kaplan–Feinstein Index has been published previously.8, 9

For analysis, study participants were classified into 2 groups based on age: < 70 years and ≥ 70 years. Follow-up was calculated from the date of diagnosis to the date of last contact or death. Cervical carcinoma-specific survival was calculated from the date of diagnosis until the time of death from cervical carcinoma, whereas overall survival was calculated from the diagnosis date until the time of death from any cause.

Demographic and clinical characteristics were compared using the chi-square test and the Fisher exact test. For each characteristic, the distributions of time to event outcomes were described by Kaplan–Meier estimators and were compared using the log-rank test. Multivariate Cox proportional hazards models were fitted to compare survival experiences between the two age groups while adjusting for the effects of other demographic and clinical characteristics. A P value < 0.05 was considered statistically significant.

RESULTS

In total, 1582 patients were identified. One thousand three hundred eighty-five patients (87.5%) were age < 70 years, and 197 patients (12.5%) were age ≥ 70 years. The median follow-up for the entire cohort was 4.0 years, and 665 patients (42.0%) had been followed for at least 5 years. The median age of patients in the younger cohort was 43 years (range, 19–69 years), compared with 76 years (range, 70–94 years) for the older cohort.

The demographic and pathologic features of the study population are displayed in Table 1. Squamous histology was most common, occurring in 1072 younger women (77.4%) and in 139 of the older women (70.5%). Adenocarcinomas and nonsquamous, nonadenocarcinoma histologies were more common in the older women (P = 0.0019). Intercurrent disease was more frequent among the elderly patients, as measured by our comorbidity score. Thirty percent of the older patients had moderate-to-severe medical comorbidities (score, 2–3) compared with 10% of the younger patients (P < 0.001). Analysis of the stage distribution of the cohort revealed that the younger patients tended to present with earlier stage disease (P < 0.0001). Sixty percent of women age < 70 years had early-stage carcinomas (Stage I–IIA). Only 44% of elderly patients had Stage I–IIA neoplasms. There was a preponderance of advanced stage tumors in the elderly cohort. Nineteen percent of patients age > 70 years had Stage III tumors, and 12% had Stage IV tumors. Conversely, 15% of younger patients had Stage III tumors, and only 6% had Stage IV tumors.

Table 1. Demographic Characteristics of the Population
CharacteristicNo. of patients (%)P value
Age < 70 yrs (n = 1385)Age ≥ 70 yrs (n = 197)
  • SD: standard deviation; Gy: grays; NED: no evidence of disease.

  • a

    For patients who were treated with primary radiation or chemoradiation.

  • b

    For patients who were treated with radiation or chemoradiation.

Race  0.38
 White1014 (73.2)135 (68.5) 
 African American 358 (25.8) 60 (30.5) 
 Other  13 (0.9)  2 (1.0) 
Age (yrs)  < 0.0001
 Median4376 
 Range19–6970–94 
Histology  0.0019
 Squamous1072 (77.4)139 (70.5) 
 Adenocarcinoma 163 (11.8) 29 (14.7) 
 Adenosquamous  72 (5.2)  5 (2.5) 
 Other  78 (5.6) 24 (12.2) 
FIGO stage  < 0.0001
 IA 183 (13.2) 14 (7.1) 
 IB 616 (44.5) 54 (27.4) 
 IIA  25 (1.8) 18 (9.1) 
 IIB 175 (12.6) 38 (19.3) 
 III 202 (14.6) 38 (19.3) 
 IV  84 (6.1) 24 (12.2) 
 Unstaged/unknown 100 (7.2) 11 (5.6) 
Comorbidity  < 0.0001
 0 458 (71.1) 25 (30.9) 
 1 121 (18.8) 32 (39.5) 
 2  48 (7.5) 17 (21.0) 
 3  17 (2.6)  7 (8.6) 
Primary treatment  < 0.0001
 Surgery 753 (54.4) 32 (16.2) 
 Radiation 417 (30.1)137 (69.5) 
 Chemoradiation 188 (13.6) 19 (9.6) 
 Other  21 (1.5)  2 (1.0) 
 No treatment   6 (0.4)  7 (3.6) 
Dose to Point Aa  0.0003
 Mean ± SD (Gy)6212 ± 9835727 ± 1035 
Adjuvant treatment  0.0039
 None1161 (83.8)181 (91.9) 
 Radiationb 174 (12.6) 11 (5.6) 
 Other  43 (3.1)  3 (1.5) 
 Unknown   7 (0.5)  2 (1.0) 
Disease status  < 0.0001
 NED 982 (70.9)117 (59.4) 
 Progressive 139 (10.0) 48 (24.4) 
 Recurrent 247 (17.8) 29 (14.7) 
 Unknown  17 (1.2)  3 (1.5) 

Overall, the younger patients were more likely to undergo surgery, and the older patients were more likely to receive primary radiotherapy (P < 0.0001). Fifty-four percent of patients age < 70 years underwent surgery, whereas the primary treatment for the older women consisted of surgery in only 16%. Conversely, 605 of the younger women (44%) received primary radiation or chemoradiation compared with 156 of women age ≥ 70 years (79.1%). The mean radiation dose to Point A was greater for women age < 70 years (6212 grays vs. 5727 grays; P = 0.0003). Ninety-nine percent of patients who were treated with radiotherapy received brachytherapy. To account for the imbalance in the stage distribution between the two groups, treatment was analyzed by stage (Table 2). There was a significant difference in treatment between the two groups for each of the three stage strata. Younger women with Stage IA tumors predominately underwent surgery, whereas 50% of the older women were treated with radiation (P < 0.0001). Seventy-five percent of the younger women with Stage IB or IIA tumors underwent surgery, whereas 25% received primary radiation. Among similarly staged elderly women, only 24% underwent surgery, whereas 74% were treated with radiation (P < 0.0001). In addition to differences in primary treatment, there was also a discrepancy in the administration of adjuvant therapy. The younger women were significantly more likely to receive some form of adjuvant therapy than their older counterparts (16% vs. 7%; P = 0.0039). Finally, there was a disparity in the number of untreated women. The elderly women were 9 times more likely than the younger women to forego treatment (3.6% vs. 0.4%; P < 0.0001).

Table 2. Treatment by Disease Stage in Women with Invasive Cervical Carcinoma
StageNo. of patients (%)P value
SurgeryRadiationaOtherNone
  • a

    For patients who were treated with radiation or chemoradiation.

  • b

    Stage IIB–IV and unstaged.

Stage IA     
 Age < 70 yrs161 (88.0) 14 (7.7)8 (4.4)0 (0.0)< 0.0001
 Age ≥ 70 yrs  6 (42.9)  7 (50)1 (7.1)0 (0.0) 
Stage IB/IIA     
 Age < 70 yrs479 (74.7)157 (24.5)4 (0.6)1 (0.2)< 0.0001
 Age ≥ 70 yrs 17 (23.6) 53 (73.6)0 (0.0)2 (2.8) 
Otherb     
 Age < 70 yrs113 (20.1)434 (77.4)9 (1.6)5 (0.9)0.0008
 Age ≥ 70 yrs  9 (8.1) 96 (86.5)1 (0.9)5 (4.5) 

Disease status and survival were then examined. Fifteen percent of the elderly women had recurrent tumors, whereas 48 elderly women (24%) had persistent disease. Among the younger patients, recurrent disease was noted in 247 women (18%), and persistent disease was seen in 139 women (10%; P < 0.0001). The patterns of failure were similar between the groups. In a univariate analysis, advanced age was associated with persistent and recurrent disease (hazard ratio [HR]. 1.57; 95% confidence interval [95% CI], 1.24–2.00; P = 0.002). However, after adjusting for tumor stage, histology, and treatment, there was no statistically significant association between recurrent/persistent disease and age (P = 0.98). A univariate analysis revealed that age, disease stage, histology, comorbidity score, and treatment all were associated significantly with both overall survival and cervical carcinoma-specific survival. All of the covariates remained as significant predictors of overall survival and cervical carcinoma-specific survival in a multivariate analysis. The HR for death from any cause was 2.13 (95% CI, 1.51–3.01) among women age ≥ 70 years (P < 0.0001; data not shown). The HR for death from cervical cancer was 1.61 (95% CI, 1.05–2.49) for the elderly women (P = 0.030) (Table 3). Kaplan–Meier plots of overall survival and cervical carcinoma-specific survival are displayed in Figures 1 and 2.

Table 3. Multivariate Proportional Hazards Model of Disease-Specific Survival in Women with Cervical Cancer
VariableP valueHR95% CI
  • HR: hazard ratio; 95% CI: 95% confidence interval.

  • a

    For patients who were treated with radiation or chemoradiation.

Age   
 < 70 yrs1.00  
 ≥ 70 yrs0.0301.611.05–2.49
Stage   
 IA1.00  
 IB0.4381.620.48–5.52
 II0.1722.400.68–8.41
 III, IV, unknown0.0094.911.48–16.31
Histology   
 Adenocarcinoma1.00  
 Squamous0.0550.620.40–1.01
 Other0.0042.321.30–4.11
Comorbidity0.0241.261.03–1.54
Treatment   
 Surgery1.00  
 Radiationa< 0.00014.913.00–8.05
 Other0.6341.630.22–12.20
 No treatment< 0.000128.5911.01–74.25
Figure 1.

Kaplan–Meier plot of overall survival stratified by age. Solid line: age < 70 years; dashed line: age ≥ 70 years.

Figure 2.

Kaplan–Meier plot of cervical cancer-specific survival stratified by age. Solid line: age < 70 years; dashed line: age ≥ 70 years.

DISCUSSION

The results of the current study revealed that the patterns of care and survival of elderly women with cervical carcinoma differ distinctly from those of younger patients. These findings add to a growing body of literature suggesting that elderly patients with malignancies from a wide variety of sites are treated differently and often have inferior outcomes compared with younger patients.3–6, 10–18 Studies of breast carcinoma in the elderly have found consistently that these women are treated less aggressively than younger patients. Several reports have noted that women with breast cancer age > 70 years are less likely to receive surgery, axillary lymph node dissection, radiation, and chemotherapy compared with younger women.3, 4, 10–14 Elderly patients with leukemia and lymphoma often do not undergo treatment or are treated with less dose-intensive chemotherapeutic regimens.6, 16, 17 Similar trends toward less aggressive surgical and chemotherapeutic treatments for geriatric cancer patients have been noted in the management of other solid tumors.5, 15, 18 de Perrot and coworkers found that younger patients with lung carcinoma were treated most commonly with pneumonectomy, whereas older patients were more likely to undergo a lobectomy or a lesser resection despite the fact the elderly patients tolerated thoracotomy surprisingly well.5 We documented a similar discrepancy in the treatment of elderly patients with invasive cervical carcinoma.

Relatively little data examining patterns of care for elderly patients with cervical carcinoma have been reported. In our cohort, there was a strong bias toward treating younger patients surgically and treating older patients with radiotherapy. This discrepancy was noted for all stages of disease, but it was most striking for women with Stage IB and IIA neoplasms. Seventy-five percent of women age < 70 years underwent radical surgical resection, compared with only 24% of elderly women. There also was a strong correlation between advancing age and the decision to forego treatment. A variety of factors likely contributed to these differences in management. The elderly patients more frequently had moderate or severe medical comorbidities, likely biasing physicians against radical pelvic surgery. The presence and severity of other comorbidities has been identified as a powerful influence in the allocation of both primary and adjuvant therapy to cancer patients.7, 14, 19 In addition, older women are more likely than their younger counterparts to refuse aggressive treatment.20, 21 Newcomb and Carbone surveyed 628 women with breast or colorectal carcinoma. The authors found that participants age > 65 years were more likely to reject more intensive therapy.20 Thus, the discrepancy in treatment for women with cervical carcinoma is likely multifactorial and reflects both physician and patient preferences.

Despite assumptions that elderly patients who undergo surgery and radiation have high complication rates and impaired survival, more recent studies have found that these patients tolerate treatment remarkably well.22–31 In an evaluation of 581 patients with cervical cancer who were treated with radiation, Mitchell et al. found that age had no effect on survival after adjusting for treatment prolongation and delivery of intracavitary radiotherapy. Those authors did note that patients age > 70 years were less likely to receive brachytherapy than their younger counterparts.22 Treatment interruptions and failure to complete radiotherapy appear to be relatively common among elderly populations.23, 25, 28 In a series of 31 patients age > 75 years, 32% of patients failed to complete treatment, and treatment interruptions occurred in 42% of patients.26 When adequate treatment is delivered, both tumor control and survival in elderly women are similar to that seen in younger women.23–25, 27 Although some studies have described a higher rate of acute radiation-related complications among older women, there does not appear to be an association between age and chronic radiation-related sequelae.22, 24–28 Likewise, radical hysterectomy in elderly patients with early-stage cervical cancer is associated with acceptable morbidity.29–31 Fuchtner et al. examined 45 women age > 65 years who underwent radical hysterectomy. Although transfusions were required more frequently in the elderly patients, there were no differences in survival or in any of the other complications examined. Those authors concluded that age alone should not be a contraindication to radical hysterectomy.29 Little data are available that specific address the tolerance and outcome of elderly women who are managed with chemoradiation. In our cohort, 10% of women age ≥ 70 years were treated with combination chemoradiation. Further study of chemoradiation in geriatric populations clearly is warranted.

Although age commonly has been regarded as a prognostic factor for patients with cervical carcinoma, the literature describing the effect of age on survival has been conflicting. Many reports have not taken into account treatment differences and inequalities in stage distribution. Similar to other reports, we noted that elderly patients presented more commonly with more advanced neoplasms.25, 32 Whereas one-third of women age < 70 years had Stage IIB–IV tumors, greater than one-half of the elderly patients had Stage IIB–IV disease. Although previous studies reported that young age was associated with decreased survival, more recent population-based studies have found that advanced age is associated with decreased survival.33–39 In a multivariate analysis of > 10,000 patients, Meanwell et al. noted that survival decreased with advancing age.38 Reporting data from the Surveillance, Epidemiology, and End Results data base, Kosary stratified patients in 10-year increments and noted that survival decreased consistently with age. For patients age ≥ 70 years, the relative risk of death was 9.93.39 However, those data were not adjusted for treatment or comorbidity. In our cohort, advanced age was a strong predictor of both overall survival and cervical cancer-specific survival. After adjusting for stage, histology, comorbidities, and treatment, the relative risk of death from cervical carcinoma for women age ≥ 70 years older was 1.61 (95% CI, 1.05–2.49).

The current study has the advantage of including over 1500 consecutively treated patients from a single institution. However, our study has several limitations that must be acknowledged. First, the study included patients treated since 1986. During the past 15 years, the treatment of cervical cancer has evolved, most notably with the widespread acceptance of chemoradiation in 1999. To investigate the impact of these treatment changes, an analysis of survival was performed based on date of treatment with stratification into 5-year blocks. We were unable to identify any survival difference based on date of treatment when the entire cohort was examined (data not presented). Second, most of the data presented were ascertained from a hospital-based tumor registry. Although all data were collected prospectively, we cannot exclude the possibility that data on some of the patients were incomplete.40 Finally, complete data on radiation treatment and medical comorbidities were available for only a fraction of the study population. A further analysis of the impact of comorbidity in patients with cervical cancer currently is underway to explore this issue further.

The current study revealed that age is an important factor in the allocation of treatment and survival for patients with cervical cancer. Compared with younger patients, elderly women with cervical carcinoma are more likely to receive primary radiotherapy or to forego treatment. Even after adjusting for disease stage, medical comorbidity, and treatment, women age > 70 years are more likely to die from cervical carcinoma than their younger counterparts. Given the fact that geriatric patients often tolerate treatment well, aggressive therapy for patients with cervical carcinoma should not be withheld based on age alone.

Ancillary