|Home | About | Journals | Submit | Contact Us | Français|
Papillary serous carcinoma of the peritoneum (PSCP) has been recognized for almost five decades, but little is known about the etiology or pathogenesis of this uncommon malignancy. The objective of this analysis was to examine trends in the incidence of PSCP in the United States. Invasive PSCP cases (N=4,389) were identified through 24 population-based registries in the U.S. during the period 1995–2004. Incidence rates were calculated per million population. PSCP is a disease of older women, with few cases diagnosed before the age of 40 years. The incidence of PSCP was 64% lower among black women and 47% lower among Asian-Pacific Islander women compared to white women. Rates among Hispanic women were 39% lower among than among non-Hispanic women. The majority of PSCP (68%) was diagnosed at a distant stage, underscoring the difficulty of diagnosing this malignancy. The incidence of PSCP has increased dramatically during the past decade in the U.S. with the greatest rise (>13% per year) among non-Hispanic and white women. This trend was more pronounced among older women and women with early stage disease. The incidence of PSCP shows substantial racial and ethnic diversity. The increase in the rate of PSCP among all racial and ethnic groups during the 10-year observation period is cause for some alarm. Although the reason for this temporal trend is unknown, some of the increase may be attributable to reclassification of ovarian carcinoma to the peritoneum.
Papillary serous carcinoma of the peritoneum (PSCP) has been recognized as a distinct disease entity for almost five decades,1 but the etiology and pathogenesis of this malignancy is largely unknown. The epithelial layer of the ovary and the peritoneum share a common embryonal heritage, deriving from coelemic epithelium early in life.2 Clinically, women diagnosed with PSCP are treated using the same surgical and chemotherapeutic approach as epithelial ovarian cancer because of the similarities in biological behavior.3,4 In 1993, the Gynecologic Oncology Group developed specific diagnostic criteria that allowed for the differentiation of PSCP from papillary serous ovarian carcinoma,5 a histologically identical, but morphologically distinct malignancy.6 Ovarian serous tumors are generally discovered after spread to the peritoneal surfaces, but are arbitrarily assigned to the ovary if involvement of this organ is 5 mm or greater.7 This observation is consistent with the hypothesis that a proportion of papillary serous ovarian carcinomas originate in the epithelium of the peritoneal cavity, with secondary spread to the ovary.8,9
PSCP is rare so etiologic studies are difficult to conduct, requiring large collaborations among research centers. The biological and pathological similarities between PSCP and papillary serous ovarian carcinomas support the likelihood of a common molecular pathogenesis.10 Immunohistochemical analyses have demonstrated congruously altered protein expression patterns for PSPC and papillary serous ovarian carcinoma.6 PSCP appears to be part of the hereditary breast-ovarian cancer syndrome as the frequency of BRCA mutations in peritoneal and ovarian cancer cases is comparable.11,12 Furthermore, PSCP may occur in high-risk, BRCA-positive women years after prophylactic oophorectomy, suggesting a common etiology. The absence of descriptive and analytic studies of PSCP highlights the need for a comprehensive analysis of incidence that might provide leads for further research.
Incidence data for 4,389 women diagnosed with invasive PSCP from 1995 to 2004 were obtained from the North American Association of Central Cancer Registries.13 All women with microscopically-confirmed, invasive PSCP were included in the analytic file.14 Qualifying states meeting the standards for data quality that participated in the analysis through written consent were Arizona, California, Colorado, Connecticut, Florida, Hawaii, Idaho, Illinois, Iowa, Kentucky, Louisiana, Maine, Michigan, Nebraska, New Jersey, New Mexico, New York, Pennsylvania, Rhode Island, Texas, Utah, Washington, Wisconsin, and Wyoming. These population-based registries represented approximately 64% of the U.S. population during the ten-year time period, including 62% of the white population, 54% of the black population, 59% of the Alaska Native-American Indian population, 81% of the Asian-Pacific Islander population, and 88% of the U.S. Hispanic population. Although there were too few American Indian-Alaska Native cases to perform subgroup analyses (n = 9), they were included as part of the “total”. A summary stage variable was used to group cases for comparative analysis of stage of disease at diagnosis.15,16 Cases diagnosed from 1995 to 2000 were coded according to guidelines established in 1977 (SSS1977);15 cases diagnosed between 2001 and 2003 were coded according to revised guidelines established in 2000 (SSS2000).16 Because another revision of the staging system was implemented in 2004, cases diagnosed in 2004 were not included in the analyses involving stage. Population counts, used as denominators for calculating the cancer rates included in this analysis, were derived from the 2000 U.S. Census.17
SEER*Stat was used to calculate both annual and average annual age-adjusted incidence rates expressed per million population.18 All rates were age-adjusted by 5-year age groups to compare racial and ethnic groups by demographic and clinical variables of interest. Confidence intervals (CI) for the rates were calculated by the method of Fey and Feuer.19 The 2000 U.S. population was used for age standardization. For all analyses, rates were suppressed when the category had fewer than 20 cases, although counts were always provided. Incidence rate ratios (IRR), 95% confidence intervals (CI), and significance tests for the IRRs were calculated to compare patterns in cancer incidence.20,21 To assess a trend in incidence rates, the annual percent change in rates between 1995 and 2004 was calculated. The statistical comparison of trends was based on the t-test of the trend regression parameter.
Few cases of invasive PSCP were diagnosed before the age of 40 years (Fig 1). Rates increased up to ages 70–74 years and then declined. This peak was observed somewhat earlier among black women (mean age at diagnosis 66 years) and Asian-Pacific Islander women (mean age at diagnosis 64 years) compared to white women (mean age at diagnosis 67 years). Similar age-specific curves were observed among Hispanic and non-Hispanic women, although Hispanic women were diagnosed at younger ages than were non-Hispanic women (67 years versus 64 years).
Incidence rates for invasive PSCP were significantly higher among white women (4.99 per million) compared to black women (1.82 per million) and Asian-Pacific Islander women (2.65 per million) (Table 1). PSCP incidence among non-Hispanic women (4.80 per million) was also significantly greater than the incidence among Hispanic women (2.93 per million).
The majority of invasive PSCP (68%) was diagnosed at a distant stage (Table 2). Women were at three-fold greater risk of diagnosis at a distant stage compared to a localized stage. White women and non-Hispanic women were generally diagnosed with PSCP at an earlier stage compared to black women, Asian-Pacific Islander women, and Hispanic women.
The AAIR for invasive PSCP increased by 234% (Ptrend < 0.0001) between 1995 and 2004, with an annual change in incidence of 12.9% (95% CI: 10.2% to 15.7%; see Fig 2). The annual percent change was greater for whites (13.2%) than for blacks (9.1%) or Asian-Pacific Islanders (11.0%). Non-Hispanic women experienced the greatest annual percent change (13.4%) and Hispanic women the lowest annual percent change (7.9%) in incidence.
Although the rates for invasive PSCP rose dramatically between 1995 and 2004 among all age groups, the trend was more pronounced among older women than among younger women (Fig 3). The annual percent change was greater for localized/regional PSCP (23.2%) than for distant (12.9%) or unstaged (3.3%) carcinoma (Fig 4).
Before the 1990s, PSCP was an extremely uncommon disease occurring among a handful of women in the U.S. During the past decade however, the rate of increase in PSCP has been profound, although the reason for this trend is unclear. Presumably, some of the cases of PSCP would have been diagnosed as papillary serous ovarian carcinoma before the Gynecologic Oncology Group established specific criteria for differentiating the two carcinomas in 1993,5 suggesting that heightened awareness of this malignancy among oncologists and pathologists may have led to some artifactual changes in diagnostic patterns. According to Chen et al.,6 the new Gynecologic Oncology Group guidelines resulted in reclassification of 7–20% of ovarian carcinomas to the peritoneum. Nonetheless, it is uncertain why the rise in PSCP has been more dramatic among white and non-Hispanic women compared to other racial and ethnic groups. Furthermore, racial and ethnic differences in the incidence of PSCP are more pronounced than for epithelial ovarian cancer;22 and studies of the pathogenesis of ovarian carcinoma suggest a unifocal origin, whereas histopathological and molecular evidence is accumulating that a proportion of peritoneal carcinomas are multifocal.23–25
Published reports of primary peritoneal cancer have generally focused on clinical presentations and case series, few exceeding 50–100 patients.2,4,10,26–29 In the largest examination of primary peritoneal cancer incidence, Roffers et al.10 studied 139 women with invasive and borderline malignancies diagnosed between 1992 and 1997 in 26 cancer registries in the U.S. These investigators reported rates of serous cystadenocarcinoma of 2.6 per million, slightly lower than that in our analysis of invasive PSCP, but consistent with a lower rate for this disease in the mid-1990s.
The epidemiology of PSCP and papillary serous ovarian carcinoma has been reported to be similar,4,30 although analytic studies of PSCP risk factors have been underpowered and poorly controlled. Peritoneal cancer appears to be part of the hereditary breast-ovarian cancer syndrome as the frequency of BRCA mutations in peritoneal and ovarian cancer cases is comparable.11,12 In a small study of 50 women with primary peritoneal cancer and 503 women with epithelial ovarian cancer,30 no differences were found between the two groups in reproductive history, oral contraceptive pill use, history of tubal ligation, or use of hormone replacement therapy. Although women with primary peritoneal cancer were older, had later menarche, and were less likely to have used perineal talc than women with epithelial ovarian cancer, these findings were not confirmed in a larger comparison of 95 peritoneal cancer cases with 117 women with advanced ovarian cancer.4
Pelvic serous carcinomas are distinguished by tumor distribution and precursor lesions.9 PSCP is diagnosed in the presence of a large, generally disseminated, tumor mass involving peritoneal surfaces with minimal or nonexistent tubal and ovarian involvement.3,9 High-grade serous carcinomas, characterized by genetic instability and TP53 mutations, are thought to arise in the fallopian tube mucosa or the mullerian epithelium in the peritoneal cavity, with secondary spread to the ovary.8,9 The tubal fimbria, which is in close proximity to the ovarian surface and is exposed to the peritoneal cavity, has been identified as the most common site of origin for early serous carcinoma in BRCA-positive women.31 Large ovarian and peritoneal carcinomas are frequently associated with small, minimally invasive carcinomas of the distal fallopian tube, suggesting a common origin.9 Unfortunately, misclassification of pelvic serous tumors in the absence of a defined precursor lesion, such as an intraepithelial carcinoma or an endometriotic cyst, limits the validity of our investigation.
Similar to epithelial ovarian cancer,32 a high percentage of women with PSCP were diagnosed at an advanced or unknown stage, underscoring the aggressive nature and difficulty of diagnosing this malignancy. Black women, in particular, had a four-fold greater risk of diagnosis with an advanced PSCP than at a localized/regional stage, raising the possibility of racial or ethnic differences in access to medical care or health insurance coverage. The growing proportion of carcinomas diagnosed at a localized/regional stage during the 10-year study period argues for an enhanced recognition of PSCP among oncologists and pathologists.
In summary, it is unclear why the incidence of PSCP has been increasing in the U.S., particularly among white and non-Hispanic women. Changes in risk factors are a possible explanation, but it is likely that a proportion of the rapidly rising incidence is an artifact of changes in diagnostic criteria promulgated immediately before the observation period. The substantial racial and ethnic differences in the incidence of PSCP are intriguing and suggest a distinct etiology for this fatal malignancy that should be explored in analytic studies.
Research Support: Centers for Disease Control and Prevention under cooperative agreement U75/CCU515998 to the North American Association of Central Cancer Registries, and the National Cancer Institute’s Surveillance, Epidemiology, and End-Results Program under contract N01-PC-35137 from the NIH, Department of Health and Human Services.
Presentation of Study Results Elsewhere: Not applicable
Disclaimers: No conflicts of interest to disclose