Patterns of Care for Women With Ovarian Cancer in the United States By Kathryn Anne Muioz, Linda C. Harlan, and Edward L. Trimble Purpose: To characterize treatments for ovarian cancer, to determine if recommended staging and treatment were provided, and to determine factors that influence receipt of recommended staging and treatment. Methods: A total of 785 women diagnosed with ovarian cancer in 1991 were selected from the National Cancer Institute (NCI) Surveillance, Epidemiology, and End Results (SEER) program. Type and receipt of recommended staging and treatment were examined using data on surgery and physician-verified chemotherapy. Results: Most women with presumptive stage I and II ovarian cancer were treated with surgery alone (58%), while women with stage III or IV disease were treated with surgery plus platinum-based chemotherapy (75% stage III, 56% stage IV). Approximately 10% of women with presumptive stage I and 11,71% with stage III, and 53% with stage IV disease received recommended stag-

ing and treatment. The absence of lymphadenectomy and assignment of histologic grade were the primary reasons women with presumptive stage I and II cancer did not receive recommended staging and treatment, whereas for stages III and IV, it was due to older women not receiving surgery plus platinum-based adjuvant chemotherapy. Age, stage, comorbidity, "other" race/ethnicity, and treatment at a facility with an approved residency training program were associated with whether recommended staging and therapy were received. Conclusion: Older women with late-stage disease did not receive recommended treatment. The majority of women with early-stage disease did not receive recommended staging and treatment. J Clin Oncol 15:3408-3415, 1997. This is a US government work. There are no restrictionson its use.

VARIAN CANCER IS THE FIFTH most common 0J cancer among women in the United States, with an expected annual incidence rate of 26,700 for 1996.' The incidence increases with age until age 60 to 70 years,

Several investigations have evaluated chemotherapeutic and surgical treatments for ovarian cancer' 0 - '4 and current approaches for management of ovarian cancer have been evaluated by a national consensus panel.' 5 .'7 According to the National Institutes of Health (NIH) Consensus Conference statement, recommendations for earlystage ovarian cancer include accurate surgical staging and surgical removal of the tumor.'.5 2 ' Surgery may include a total abdominal hysterectomy (age dependent)/bilateral salpingo-oophorectomy, omental biopsy, tumor removal, and debulking in cases where the tumor is too large to remove completely. Accurate surgical staging requires biopsy of the pelvic and paraaortic lymph nodes. For advanced disease, surgical intervention is required, which allows for staging, accurate diagnosis, and cytoreduction. This includes removal of the omentum, total abdominal hysterectomy/bilateral salpingo-oophorectomy, debulking, or total removal of large tumors. Treatment recommendations for women with late-stage disease are systemic adjuvant chemotherapy that consists of cisplatin or carboplatin combination chemotherapy combined with

when it plateaus. Therefore, the majority (75%) of ovarian cancers are identified in women ages 65 years and older.2 It is estimated that more than 14,800 women in the United States will die of ovarian cancer in 1996.' Five-year survival rates for early-stage (presumptive stage I and II) disease range from 50% to 90%. 3 The 5year survival rate for late-stage disease (stages III and IV) is 21%. 3 The primary factors that influence survival prognosis for ovarian cancer include age, histologic grade and type, stage of disease, and tumor rupture.4 9 Increased survival is found in younger women, among women with low-grade well-differentiated tumors, and in women with earlier stages at diagnosis. Prognosis based on histologic grade ranges from excellent among women with borderline ovarian tumors to poor for undifferentiated carcinoma.4

surgery. 15-19,22-29

From the Applied Research Branch, Division of Cancer Prevention and Control,and Cancer Therapy Evaluation Program, Division of CancerTreatment, Diagnosisand Centers, National CancerInstitute, National Institutes of Health, Bethesda, MD. Submitted May 3, 1996; accepted July 1, 1997. Address reprint request to Kathryn A. Mutoz, PhD, MPH, Merck and Co, Inc, 10 Sentry Pkwy, BL2-3, Blue Bell, PA 19446; Email [email protected]. This is a US government work. There are no restrictions on its use. 0732-183X/97/1511-0012$0.00/0

3408

In this study, we report types of staging, surgery, and chemotherapy used to treat women with ovarian cancer and factors that influence recommended treatment in patients with epithelial carcinoma. METHODS The National Cancer Institute (NCI) initiated a study to determine the extent to which recommended staging and treatment were being provided to ovarian cancer patients. The purpose of the patterns-of-

Journal of Clinical Oncology, Vol 15, No 1 1 (November), 1997: pp 3408-3415

Downloaded from ascopubs.org by 37.44.207.118 on January 22, 2017 from 037.044.207.118 Copyright © 2017 American Society of Clinical Oncology. All rights reserved.

3409

OVARIAN CANCER TREATMENT care (POC) study of ovarian cancer was to describe the types of treatment provided to women based on stage and age at diagnosis and factors that affect treatment such as race, patient insurance status, and treatment center characteristics.

Data Source and Study Subjects A sample of 785 ovarian cancer cases diagnosed in 1991 was selected from the Surveillance, Epidemiology, and End Results (SEER) program using nine SEER registries that had participated in the program before 1993. The registries included San Francisco/ Oakland, Connecticut, Detroit, Hawaii, Iowa, New Mexico, Seattle, Utah, and Atlanta. Women were staged using the International Federation of Gynecology and Obstetrics (FIGO) staging procedures. Women were classified as having early-stage (presumptive stage I and II) or late-stage (stages III or IV) ovarian cancer. To obtain a larger sample of earlystage cases, stage I and II cancers were oversampled, which resulted in a total of 399 presumptive stage I and 11 (50.9%), 143 stage III (18.2%), and 242 (30.9%) stage IV ovarian cancers. The sample included 685 white women, 42 black women, 52 women of other race/ethnicity, and six women of unknown race/ethnicity.

Analytic Variables Patientinformation. Information on age at diagnosis, grade, histology, lymphadenectomy, comorbidity, and insurance status were collected for each patient. Grade and histologic classification were determined from the medical record and coded using the International Classification of Diseases-Oncology (ICD-O) coding system. Grade classification is defined by ICD-O as poorly differentiated, undifferentiated, moderately differentiated, or well differentiated. For this study, an independent pathology review was not performed. For statistical analysis, histologies were categorized into epithelial carcinoma, germ cell, sarcoma, borderline disease (tumors of low malignant potential), sex cord, and other. To asses the impact of comorbidity on treatment selection and receipt of recommended staging and treatment for epithelial carcinoma, a clinical comorbidity index was determined according to a modified method of that reported by Charlson et al.3"3' This index is used to account for the number and seriousness of comorbid conditions and is based on relative risk estimates that predict mortality. For analysis purposes, women were categorized into comorbidity groups that reflected indices of 0, 1, and 2+. Insurance status was categorized into the following groups: none, Medicare/Medicaid, Medicare plus private, private, and other. Private insurance included health maintenance organizations (HMOs) and other private insurance carriers. Treatment facility characteristics. Data collected by each registry on the characteristics of the treatment facility included hospital size (number of beds), hospital ownership (federal government, nonfederal, government, nongovernment/not for profit, and nongovernment/for profit), proportion of Medicaid and Medicare patients discharged, and whether there was an approved residency training program for any specialty. The disproportionate share of a hospital patient load is computed as the sum of the number of Medicare Part A days with social security insurance divided by all Medicare Part A days added to the number of Medicaid paid days divided by the total hospital days. This sum of proportions includes only Medicare costs associated with the prospective payment system and is adjusted for the hospital's location (rural or urban). This measure is used as

Table 1. Percentage of Women Who Received Surgery for Ovarian Cancer According to Stage at Diagnosis Type of Surgery None Biopsy and exploratory (not primary) Unilateral resection (no hysterectomy) Unilateral resection (hysterectomy) Bilateral (no hysterectomy) Bilateral (hysterectomy) Extensive surgery* Surgery, not otherwise specified

Stages I&II Stage III Stage IV In = 399) (n 143) In = 242) 2 1 15 2 5 21 54 0

1 4 0 0 1 4 90 0

7 21 2 0 2 4 64 1

NOTE. Data reflect all histologies. One subject unstaged excluded from all analyses. *Partial or total omentectomy with unilateral or bilateral salpingo-oophorectomy (with or without hysterectomy), debulking, or pelvic exenteration and surgery of regional and/or distant sites/nodes.

a surrogate for hospitals located in areas of low income or with a heavy burden of nonprivate insurance patients. Treatmentpatterns. Chemotherapy, hormone therapy, and radiation therapy were verified with the patients' physician. Reported treatments included 31 types of chemotherapy and four types of hormonal therapy. Type of cancer-directed surgery was reported according to SEER site-specific surgery codes and categorized as follows: (1) no surgery; (2) biopsy and exploratory; (3) unilateral resection without hysterectomy; (4) unilateral resection with hysterectomy; (5) bilateral salpingo-oophorectomy without hysterectomy; (6) bilateral salpingo-oophorectomy with hysterectomy; (7) extensive surgery-partial or total omentectomy with unilateral or bilateral salpingo-oophorectomy (with or without hysterectomy), debulking, or pelvic exenteration and surgery of regional and/or distant sites/nodes, and (8) surgery, not otherwise specified (Table I). For statistical analysis, combination treatments were categorized into two major subgroups based on surgery and chemotherapy (Table 2). The category of surgery alone also includes 2% of patients who received surgery plus hormone for presumptive stage I and II. In Table 1, surgical treatment refers to surgery directed at the primary cancer site.

Table 2. Percentage of Women Who Received Various Treatments for Ovarian Cancer According to Stage at Diagnosis Typeof Treatment None Hormone alone Surgery alone Surgery and radiation Carboplatin/cisplatin alone Carboplatin/cisplatin plus surgery Carboplatin/cisplatin and surgery, hormone therapy radiation Other chemotherapy Other chemotherapy and surgery and/or hormone therapy

Stages I &II (n = 399)

Stage III in = 143)

Stage IV in = 242)

1 1 58 4 0 32

1 0 11 1 2 75

7 1 13 0 14 56

2 1

4 0

5 2

1

6

2

NOTE. Data reflect all histologies.

Downloaded from ascopubs.org by 37.44.207.118 on January 22, 2017 from 037.044.207.118 Copyright © 2017 American Society of Clinical Oncology. All rights reserved.

3410

MUNOZ, HARLAN, AND TRIMBLE

For cases with epithelial carcinoma, staging and recommended therapy was determined according to guidelines established by the NIH Consensus Conference. 5 '6"7 Women were considered to have received recommended staging and therapy if the following conditions were met: stage I, grades 1 or 2-determination of grade, lymphadenectomy, and omentectomy/oophorectomy, no chemotherapy; stage I, grade 3 or 4-omentectomy/oophorectomy and any type of chemotherapy; stage II, all grades-lymphadenectomy, omentectomy/oophorectomy, and any type of chemotherapy; and stages III or IV-omentectomy/oophorectomy or debulking, and cisplatin- or carboplatin-based chemotherapy. For the purposes of this analysis, if the aforementioned criteria were not met, the patient was categorized as not receiving staging and recommended therapy.

StatisticalAnalysis Statistical analysis was conducted using SAS version 6.0 (SAS Institute, Cary, NC) and StatXact (CYTEL Software Co, Cambridge, MA) to compute exact X2 statistics. Analyses included descriptive statistics and the x2 test to determine significance where appropriate. Since some analyses resulted in cell sizes of zero, the exact X2 value was calculated using StatXact. Descriptive and x2 results represent analyses based on all histologies. Receipt of staging and recommended therapy was determined for epithelial carcinoma using logistic regression analysis. Patient and treatment facility characteristics were evaluated separately as independent variables using univariate models. Independent variables included age at diagnosis, race/ethnicity (white, black, and other), stage (I and II, III, or IV), comorbidity score (0, 1, or 2+), geographic region (region 1: Hawaii, California, Seattle; region 2: New Mexico, Utah; region 3: Iowa and Detroit; and region 4: Connecticut and Atlanta), number of hospital beds (< 200, 200 to 400, > 400), residency training approval status (approved or not approved), proportion of Medicaid and Medicare patients discharged (< .30, .30 to .49, > .50) hospital ownership (federal government, non-federal government, nongovernment/not for profit, and nongovernment/for profit), and insurance status (none, Medicare/Medicaid, Medicare plus private, private, and other). Independent variables that were significant in predicting recommended treatment were then entered into the multivariate model. The univariate and multivariate models included staging and recommended therapy as the dependent variable.

RESULTS Table 3 lists the characteristics of the sample, which includes all histologies. The mean age at diagnosis was 58.0 ± 0.6 years (range, 14 to 96). Eighty-seven percent of women were white and 5% were black. Eighty-three percent had a comorbidity score of 0, which indicates there were no significant comorbid conditions associated with increased risk of mortality at the time of diagnosis. Approximately 47% of women age 65 or older were diagnosed with stage IV disease, compared with 2% of women ages 14 to 29, 14% ages 30 to 49, and 29% of women ages 50 to 64 (P = .0001, X2). Seven percent of older women (age 65+) with stage IV disease had comorbidity scores - 2 compared with 2% of younger women. There were no significant race/ethnic differences in stage at di-

Table 3. Characteristics of 785 Women With Ovarian Cancer in the SEER POC for Ovarian Cancer Study Patient Characteristics n = 785)

Race White Black Other Unknown Stage at diagnosis I and 11 III IV Comorbidity score 0 1 2+

%

87 5 7 1 51 18 31 83 14 3

agnosis or in comorbidity scores (P > .05; data not shown). Only 20 women did not have any type of insurance. Overall, 23% of women had Medicare/Medicaid, 36% had private, and 20% had Medicare plus private insurance. The remaining women (21%) had other types of insurance or their insurance status was unknown. The majority of patients were treated at nongovernment/not for profit (79%) or government/not federal (14%) institutions. Most women (85%) received treatment at establishments with a low disproportionate share (< 30%). Twenty-four percent of black women and 29% of women of other race/ethnic backgrounds were treated at centers with Medicaid and Medicare patient discharges of 2 30%, while only 8% of white women were treated at these institutions. More than half (58%) of women were treated at hospitals that had approved residency training programs. The majority of women (77%) presented with epithelial carcinoma followed by borderline disease (17%), germ cell (3%), sarcoma (1%), other (1%) and sex cord (1%) carcinomas. Epithelial carcinoma accounted for 60% of all presumptive stage I and II, 90% of stage III, and 96% of stage IV cancers. Tumors of low malignant potential accounted for 31% of presumptive stage I and II, 7% of stage III, and 2% of stage IV cancers. The distribution of histologies did not differ significantly by race/ethnicity (P > .05), although black women had the highest percent of germ cell (5%) and sex cord (2%) histologies. For all histologies, tumor grade was undetermined for 41% of women, while 33% had tumors that were poorly or undifferentiated, 18% that were moderately differentiated, and 9% that were well differentiated. For women with epithelial carcinoma, tumor grade was not deter-

Downloaded from ascopubs.org by 37.44.207.118 on January 22, 2017 from 037.044.207.118 Copyright © 2017 American Society of Clinical Oncology. All rights reserved.

3411

OVARIAN CANCER TREATMENT

mined for 36.1% of stage I, 25% of stage II, 19% of stage III, and 24% of stage IV cancers. Lymphadenectomy was not performed in 66% of women with presumptive stage I and II disease, 57% with stage III disease, and 83% with stage IV disease. The percentage of women who received lymphadenectomy did not differ according to race/ethnicity (69.6% for white and 73.8% for black women, respectively; P > .05). Extensive surgery was the most common surgical category used for all stages of ovarian cancer. Extensive surgery was provided to 54%, 90%, and 64% of women with presumptive stages I and II, III, or IV, respectively. Table 1 lists the types of surgery used according to stage at diagnosis. A greater percentage of older women (7.2%; age - 65 years) did not receive surgery compared with younger women (0.7%; < 65 years). A consistent pattern of older women receiving less treatment was found, even after controlling for stage at diagnosis and comorbidity. For example, women aged greater than 64 with no comorbidity were more likely to receive no surgery or only a biopsy compared with younger women, and less likely to receive extensive surgery compared with 40- to 64-yearold women (P = .001). Younger women (age < 40) primarily received surgery alone ( 67%) even after controlling for comorbidity. The type of surgery performed differed based on comorbidity score (P = .001). Nearly 25% of women with a comorbidity score 2 did not receive surgery, compared with only 3% who had a score of 0 (P < .05). In addition, 64% of women with a comorbidity score of 0 received extensive surgery, compared with only 40% with a score - 2 (P < .05). The majority of patients with a score 2 received either no surgery or biopsy/exploratory surgery. The most common treatments used for all stages of ovarian cancer combined were carboplatin/cisplatinbased therapy plus surgery (47%) or surgery alone (36%). Treatments differed according to stage at diagnosis, as outlined in Table 2. The primary mode of treatment for stages I and II was surgery alone, followed by carboplatin/ cisplatin-based therapy plus surgery. Carboplatin/cisplatin-based therapy plus surgery was most commonly used for stages III and IV. After controlling for stage at diagnosis, older women were less likely to receive carboplatin/cisplatin-based therapy plus surgery. For example, only 47% of older women (age - 65 years) received carboplatin/cisplatin-based therapy plus surgery for stage IV disease, compared with 71% of women younger than 65. The type of combination treatment used also differed

Table 4. Percentage of Women Who Received Recommended Treatment According to Stage at Diagnosis Among 602 Women Diagnosed With Ovarian Epithelial Carcinoma in the POC Study Recommended Treatment Received (n = 238) Stage I 11 III IV

n 192 49 128 233

%

No.

9 14 71 53

17 7 91 123

NOTE. Recommended treatment as described in Methods.

based on comorbidity. The majority of women with a comorbidity score of 0 received carboplatin/cisplatinbased therapy plus surgery (77% and 60% of stage III and IV, respectively), compared with 39% of women with a comorbidity score of - I (P = .0001). No difference was found in the type of combination therapy used for the various race/ethnic groups even after controlling for stage at diagnosis or comorbidity score (P > .05) For epithelial carcinoma, treatment with recommended staging and treatment (defined by the NIH Consensus Conference) was determined (described earlier). Percentages of women who received recommended staging and therapy according to stage at diagnosis are listed in Table 4. Less than 10% of women with stage I and less than 15% of women with stage II disease received recommended staging and therapy. The majority (71%) of women with stage III and about half (53%) with stage IV disease received staging and recommended treatment that included carboplatin/cisplatin-based chemotherapy, as well as the recommended surgical procedures. The finding that few women received recommended staging and treatment for presumptive stage I and II is mainly due to the lack of lymphadenectomy. Lack of appropriate surgery was the main reason women were classified as not receiving the staging and recommended treatment for stages III and IV. This finding mainly reflects lack of surgery in women older than 64. More than half (56%) of white women, 74% of black women, and 76% of women from other race/ethnicity categories did not receive staging and recommended therapy (P = .064). Logistic regression was used to determine contributing factors for receiving recommended staging and treatment for epithelial carcinoma. As shown in Table 5, for all stages of disease, age at diagnosis, comorbidity score, and being of "other" race/ethnicity were negatively associated with receiving recommended staging and treatment, while receiving treatment at an institution with an approved residency training program was positively asso-

Downloaded from ascopubs.org by 37.44.207.118 on January 22, 2017 from 037.044.207.118 Copyright © 2017 American Society of Clinical Oncology. All rights reserved.

3412

MUNOZ, HARLAN, AND TRIMBLE Table 5. Effect of Patient and Treatment Facility Characteristics on Receipt of Recommended Treatment for Epithelial Carcinoma All Stages Variable

n

Age at diagnosis, years Stage at diagnosis I and 11 111 IV Comorbidity Index 0 1 2+ Residency training No Yes Race White Black Other

591 236 125 230

Stage III

OR 0.97

29.1 17.7

CI

n

OR

0.96-0.99

125

0.97

15.26-55.41 9.76-32.04

-

Stage IV CI 0.95-1.0

-

0.64 0.14

0.36-1.15 0.04-0.53t

104 18 3

237 354

1.9

1.25-2.95t

519 31 41

0.44 0.34

0.16-1.17 0.14-0.85'

OR

230

0.95

CI 0.93-0.98

-

480 89 22

n

-

0.88 0.0

0.47-1.51 0

174 41 15

30 95

1.7

0.67-4.32

106 124

111 5 9

0.35 0.15

0.16-1.16 0.07-0.43*

206 13 11

-

.89 .14

2.5

.31 .47

0.37-1.73 0.06-0.92

1.15-4.75t

0.08-1.21 0.14-1.85

Abbreviations: OR, odds ratio; Cl, confidence interval. *P < .05. tP < .01 tP < .001.

ciated with receiving recommended treatment. The effect of a comorbidity score - 2 on receipt of recommended treatment could not be determined for stage III disease, since there were no women in this category who received recommended treatment. Regression analysis for stages I and II could not be performed, because only 23 women with early-stage disease received recommended treatment. Variables tested that did not predict whether women received the recommended treatment (data not shown) included insurance status, hospital bed size, hospital ownership, and proportion of Medicaid and Medicare patients discharged. DISCUSSION The majority of younger women were diagnosed with early-stage disease, while older women were diagnosed with late-stage disease. These findings are in agreement with other studies that indicate older women are more likely to be diagnosed with later stage disease. 3 233 Since survival rates for early-stage disease are considerably higher compared with late-stage disease, it is important to obtain an early diagnosis and to implement accurate staging procedures, especially in older women who are currently more likely to be diagnosed with late-stage disease. However, survival for older women has been found to be inferior compared with younger women, even when stage of disease is taken into account. 34 Therefore, other factors, such as treatment intensity, comorbidity, involve-

ment of vital organs, or disease characteristics (ie, drug resistance or aggressiveness of disease), may be of greater importance in older women. The primary mode of treatment for presumptive stage I and II ovarian cancer was surgery alone. However, we found that older women were less likely to receive extensive surgery for early-stage cancer, even after controlling for comorbidity. In fact, 7% of older women compared with 0% of younger women did not receive any surgery for early-stage disease. In addition, we found that older patients with stage III and IV disease were treated with less extensive surgery regardless of their comorbidity status. These findings are similar to those reported by Ries 33 and Hightower et al,3 5 who found that many older women were not treated for their cancer, and those who were treated received less aggressive treatments compared with younger women. Unfavorable prognosis in older women may be related to increased comorbidity, as well as less aggressive treatment.3 6 37 Improved prognosis has been associated with successful surgical cytoreduction, tumor reduction, cisplatin-based adjuvant chemotherapy, and earlier diagnosis. 34,38,39 Consistent with recommendations from the NIH consensus conference,'4 1- 6 the main treatment for stage III and IV disease was surgery plus carboplatin- or cisplatinbased chemotherapy. However, there was a significant percentage of women who received either surgery or car-

Downloaded from ascopubs.org by 37.44.207.118 on January 22, 2017 from 037.044.207.118 Copyright © 2017 American Society of Clinical Oncology. All rights reserved.

3413

OVARIAN CANCER TREATMENT

boplatin/cisplatin-based chemotherapy. As in other studies, older women were less likely to receive aggressive surgery and adjuvant chemotherapy. 33 Current practice now includes the use of paclitaxel as well as platinumbased chemotherapy. 40 However, paclitaxel was not approved by the Food and Drug Administration for commercial use at the time of this study. The majority of women with presumptive stage I and II, and about half of women with stage IV epithelial carcinoma, did not receive recommended staging and therapy for ovarian cancer as defined by the 1994 NIH Consensus Conference. However, most women with stage III epithelial cancer received staging and recommended therapy. The absence of lymphadenectomy was the primary reason women with presumptive stage I and II did not receive recommended treatment, whereas for stages III and IV, it was due to older women (> 65 years) not receiving both surgery and platinum-based adjuvant chemotherapy. One previous study has shown that 31 of 100 (31%) patients with presumed stage I and II disease were upstaged on the basis of the more thorough surgical evaluation.4 ' Of those upstaged, 77% actually had stage III disease. Although complete anatomic staging is recommended by FIGO and the NIH, its impact on survival has not been demonstrated.4 2 As many clinicians do not treat women with stage I ovarian cancer in the same manner that they treat women with stage III ovarian cancer, accurate surgical staging seems to be important. Understaging has been implicated as one factor that may influence selection of effective treatment and thereby decrease survival.43 44 The finding that the majority of

women were not appropriately staged is in agreement with that reported by Young et al,4' who found that 75% of patients referred for early-stage disease did not have the appropriate incision to allow for adequate surgical staging. In this study, we investigated nonbiologic factors that may influence the receipt of recommended staging and treatment for ovarian cancer, including insurance status and treatment facility characteristics. We found that women with stage III or IV disease were more likely to receive recommended care than those with early-stage and low-grade disease. This was primarily due to lack of lymphadenectomy for early-stage disease. We also found that receiving treatment at an institution with a residency program generally predicted receipt of recommended treatment. This may be due to availability of specialists who are trained to perform accurate surgical staging including lymphadenectomy or due to increased availability of technology. This is supported by the findings of McGo-

wan, 45 who reported that 97% of patients operated on by gynecologic oncologists were adequately evaluated compared with approximately 50% of patients seen by a general obstetrician/gynecologist and 35% seen by a general surgeon. Nuguyen et al, 46 who analyzed treatment given to 12,316 patients with ovarian cancer in the United States in 1983 and 1988, found that gynecologic oncologists performed more extensive surgery and more biopsies at the time of initial surgery for ovarian cancer than did obstetrician/gynecologists or general surgeons. Junor et al, 47 who studied 533 patients with ovarian cancer registered in Scotland in 1987, found that women managed by a multidisciplinary team, including gynecologists and oncologists at a joint clinic, were twice as likely (P < .01) to receive platinum than those managed by gynecologists or surgeons alone. There was no significant difference in receiving staging and recommended treatment by insurance status, size and type of the hospital, proportion of Medicare/Medicaid patients, and region of the country. Some studies for other cancer sites have found treatment differences based on race/ethnicity or insurance status. 38 48 Although we did not find a significant role for insurance status in receiving recommended staging and treatment for ovarian cancer, the uniformity of insurance coverage in our sample may have limited our ability to determine if this factor might influence treatment in a more heterogeneous population. We did observe that women of "other" race/ethnicity were less likely to receive recommended staging and therapy. Within SEER, no significant difference in survival has been noted between young black and white women. Race does not appear to be a prognostic factor for survival of ovarian cancer 39; however, insurance status has been found to affect aggressiveness of treatment and use of services for some other cancers. 3 848 Therefore, further study with respect to race/ethnic and insurance differences in treatment is warranted. Several of the hospital characteristics, such as bed size and type of hospital, were not important determinants for receiving recommended care. Taken together with the finding that residency training was important, this may indicate that recommended treatment may be dependent on the availability of physicians trained in appropriate surgical management of ovarian cancer. In conclusion, this study provided for the first time information about various treatments provided to women with ovarian cancer who were selected from the SEER population-based registries. Therefore, the strength of this study is that the sample is more representative of women treated nationally compared with

Downloaded from ascopubs.org by 37.44.207.118 on January 22, 2017 from 037.044.207.118 Copyright © 2017 American Society of Clinical Oncology. All rights reserved.

3414

MUNOZ, HARLAN, AND TRIMBLE

a sample derived from a single location or hospital. Strong evidence is provided that the majority of women with early-stage disease and many older women with late-stage disease are not receiving recommended surgical staging and/or treatment. Although these findings indicate that many women are not receiving recommended staging and treatment, information regarding clinical judgment factors such as surgical risk and social factors were not available for analysis and may have played an important role in treatment decisions. The results of this investigation indicate that further research is needed to determine physician and patient factors that influence staging and treatment decisions. In

addition, the lack of lymphadenectomy for early-stage disease and the strong evidence that institutions with an approved residency training program are more likely to provide recommended care suggests that type of training may influence physician performance of lymphadenectomy. Therefore, examination of issues related to physician training and practice, such as time since completion of a residency training and type of specialty, may provide additional insights into factors that influence receipt of staging and recommended therapy. Research is essential to identify areas in which information dissemination and/ or training may improve the likelihood of receiving staging and recommended therapy.

REFERENCES 1. Parker SL, Tong T, Bolden S, et al: Cancer statistics, 1996 [see comments]. CA Cancer J Clin 46:5-27, 1996 2. Yancik R, Ries LA: Cancer in older persons. Magnitude of the problem-How do we apply what we know? Cancer 74:1995-2003, 1994 3. Wingo PA, Tong T, Bolden S: Cancer statistics, 1995 [published erratum appears in CA Cancer J Clin 45:127-128,1995]. CA Cancer J Clin 45:8-30, 1995 4. de Souza PL, Friedlander ML: Prognostic factors in ovarian cancer. Hematol Oncol Clin North Am 6:761-782, 1992 5. Schueler JA, Cornelisse CJ, Hermans J, et al: Prognostic factors in well-differentiated early-stage epithelial ovarian cancer. Cancer 71:787-795, 1993 6. Sevelda P, Vavra N, Schemper M, et al: Prognostic factors for survival in stage I epithelial ovarian carcinoma. Cancer 65:23492352, 1990 7. Sigurdsson K, Alm P, Gullberg B: Prognostic factors in malignant epithelial ovarian tumors. Gynecol Oncol 15:370-380, 1983 8. Dembo AJ, Davy M, Stenwig AE, et al: Prognostic factors in patients with stage I epithelial ovarian cancer. Obstet Gynecol 75:263-273, 1990 9. Friedlander ML, Dembo AJ: Prognostic factors in ovarian cancer. Semin Oncol 18:205-212, 1991 10. Cannistra SA: Cancer of the ovary [published erratum appears in N Engl J Med 10:448, 1994; see comments]. N Engl J Med 329:1550-1559, 1993 11. Qazi F, McGuire WP: The treatment of epithelial ovarian cancer [see comments]. CA Cancer J Clin 45:88-101, 1995 12. Trimble CL, Trimble EL: Management of epithelial ovarian tumors of low malignant potential. Gynecol Oncol 55:S52-S61, 1994 13. Colombo N, Chiari S, Maggioni A, et al: Controversial issues in the management of early epithelial ovarian cancer: Conservative surgery and role of adjuvant therapy. Gynecol Oncol 55:S47-S51, 1994 14. Hacker NF, Van der Burg ME: Advanced ovarian cancer. Debulking and intervention surgery. Ann Oncol 4:17-22, 1993 (suppl 4) 15. NIH Consensus Conference: Ovarian cancer. Screening, treatment, and follow-up. NIH Consensus Development Panel on Ovarian Cancer. JAMA 273:491-497, 1995 16. National Institutes of Health Consensus Development Conference Statement. Ovarian cancer: Screening, treatment, and followup. Gynecol Oncol 55:S4-14, 1994

17. Trimble EL: The NIH Consensus Conference on Ovarian Cancer: Screening, treatment, and follow-up. Gynecol Oncol 55:S1S3, 1994 18. Thigpen JT, Blessing JA, Vance RB, et al: Chemotherapy in ovarian carcinoma: Present role and future prospects. Semin Oncol 16:58-65, 1989 19. Thigpen T, Blessing JA: Current therapy of ovarian carcinoma: An overview. Semin Oncol 12:47-52, 1985 20. Thigpen JT, Lambuth BW, Vance RB: Management of stage I and II ovarian carcinoma. Semin Oncol 18:596-602, 1991 21. Thigpen T: Systemic therapy with single agents for advanced or recurrent endometrial carcinoma. Cancer Treat Res 49:93-106, 1989 22. Williams CJ, Mead GM, Macbeth FR, et al: Cisplatin combination chemotherapy versus chlorambucil in advanced ovarian carcinoma: Mature results of a randomized trial. J Clin Oncol 3:14551462, 1985 23. Omura GA, Bundy BN, Berek JS, et al: Randomized trial of cyclophosphamide plus cisplatin with or without doxorubicin in ovarian carcinoma: A Gynecologic Oncology Group study. J Clin Oncol 7:457-465, 1989 24. Omura G, Blessing JA, Ehrlich CE, et al: A randomized trial of cyclophosphamide and doxorubicin with or without cisplatin in advanced ovarian carcinoma. A Gynecologic Oncology Group study. Cancer 57:1725-1730, 1986 25. Williams CJ, Stewart L, Parmar M, et al: Meta-analysis of the role of platinum compounds in advanced ovarian carcinoma. The Advanced Ovarian Cancer Trialists Group. Semin Oncol 19:120128, 1992 26. Swenerton KD, Pater JL: Carboplatin in the treatment of carcinoma of the ovary: The National Cancer Institute of Canada experience. Ovarian Cancer Subcommittee. Semin Oncol 19:114-119, 1992 27. Alberts DS, Canetta R, Mason-Liddil N: Carboplatin in the first-line chemotherapy. of ovarian cancer. Semin Oncol 17:54-60, 1990 28. Mangioni C, Bolis G, Pecorelli S, et al: Randomized trial in advanced ovarian cancer comparing cisplatin and carboplatin [see comments]. J Natl Cancer Inst 81:1464-1471, 1989 29. Green JB, Green S, Alberts DS, et al: Carboplatin therapy in advanced endometrial cancer. Obstet Gynecol 75:696-700, 1990 30. Charlson ME, Pompei P, Ales KL, et al: A new method of

Downloaded from ascopubs.org by 37.44.207.118 on January 22, 2017 from 037.044.207.118 Copyright © 2017 American Society of Clinical Oncology. All rights reserved.

3415

OVARIAN CANCER TREATMENT classifying prognostic comorbidity in longitudinal studies: Development and validation. J Chronic Dis 40:373-383, 1987 31. Deyo RA, Cherkin DC, Ciol MA: Adapting a clinical comorbidity index for use with IC.-9-CM administrative databases. J Clin Epidemiol 45:613-619, 1992 32. Yancik R: Ovarian cancer. Age contrasts in incidence, histology, disease stage at diagnosis, and mortality. Cancer 71:517-523, 1993 33. Ries LA: Ovarian cancer. Survival and treatment differences by age. Cancer 71:524-552, 1993 34. Markman M, Lewis JL Jr, Saigo P, et al: Impact of age on survival of patients with ovarian cancer. Gynecol Oncol 49:236-239, 1993 35. Hightower RD, Nuguyen HN, Averette HE, et al: National Survey of Ovarian Carcinoma IV: Patterns of care and related survival for older patients. Cancer 73:377-383, 1994 36. Yancik R, Ries LG, Yates JW: Ovarian cancer in the elderly: An analysis of Surveillance, Epidemiology, and End Results program data. Am J Obstet Gynecol 154:639-647, 1986 37. Kennedy AW, Flagg JS, Webster KD: Gynecologic cancer in the very elderly. Gynecol Oncol 32:49-54, 1989 38. Kosary CL: FIGO stage, histology, histologic grade, age and race as prognostic factors in determining survival for cancers of the female gynecological system: An analysis of 1973-87 SEER cases of cancers of the endometrium, cervix, ovary, vulva, and vagina. Semin Surg Oncol 10:31-46, 1994 39. Balvert-Locht HR, Coebergh JW, Hop WC, et al: Improved prognosis of ovarian cancer in The Netherlands during the period

1975-1985: A registry-based study. Gynecol Oncol 42:3-8, 1991 40. McGuire WP, Hoskins WJ, Brady MF, et al: Cyclophosphamide and cisplatin compared with paclitaxel and cisplatin in patients with stage III and IV ovarian cancer. N Engl J Med 334:1-6, 1996 41. Young RC, Decker DG, Wharton JT, et al: Staging laparotomy in early ovarian cancer. JAMA 250:3072-3076, 1983 42. Finn-CB, Luesley-DM, Buxton-EJ, et al: Is stage I epithelial ovarian cancer overtreated both surgically and systemically? Results of a five-year cancer registry review. Br J Obstet Gynaecol 99:5458, 1992 43. Liberati A, Mangioni C, Bratina L, et al: Process and outcome of care for patients with ovarian cancer. Br Med J 291:1007-1012, 1985 44. Ayanian JZ, Kohler BA, Abe T, et al: The relation between health insurance coverage and clinical outcomes among women with breast cancer [see comments]. N Engl J Med 329:326-331, 1993 45. McGowan L: Patterns of care in carcinoma of the ovary. Cancer 71:628-633, 1993 46. Nuguyen HN, Averette HE, Hoskins W, et al: National Survey of Ovarian Carcinoma Part V: The impact of physician's specialty on patients' survival. Cancer 72:3663-3670, 1993 47. Junor EJ, Hole DJ, Gillis CR: Management of ovarian cancer: Referral to a multidisciplinary team matters. Br J Cancer 70:363370, 1994 48. Greenberg ER, Chute CG, Stukel T, et al: Social and economic factors in the choice of lung cancer treatment. A populationbased study in two rural states. N Engl J Med 318:612-617, 1988

Downloaded from ascopubs.org by 37.44.207.118 on January 22, 2017 from 037.044.207.118 Copyright © 2017 American Society of Clinical Oncology. All rights reserved.