Screening History Among Women with Invasive Cervical Cancer in an Academic Medical Center: Will We Miss Cancers Following Updated Guidelines?

Introduction

Before the advent of the Papanicolaou (Pap) test, cervical cancer was one of the most common and deadly malignancies diagnosed in women. Within developed countries, widespread access and acceptance of Pap test screening programs aimed at early detection and management of abnormalities has significantly decreased the incidence and mortality of invasive cervical cancer. In the United States there has been a 50% decline in the incidence of cervical cancer between 1975 and 2009 with a rate of 14.8 per 100,000 in 1975 to a rate of 6.6 per 100,000 in 2009. ¹ Mortality from invasive cervical cancer has also declined significantly over the last 40 years with a rate of 5.55 per 100,000 women in 1975 to 2.38 per 100,000 women in 2009. ¹

It is estimated that ∼60% of patients diagnosed with invasive cervical cancer in the United States today are in women who have never been screened or those who have been inadequately screened. ^{2 –5} Despite the decline of invasive cervical cancer in the United States, immigrants, the uninsured, and those with limited access to healthcare are still at an increased risk to develop the disease. ⁶

Invasive squamous cell carcinoma (ISCC) accounts for ∼80% of cervical cancer lesions, whereas invasive adenocarcinoma (IAC) accounts for ∼20% of lesions. ⁷ If diagnosed early, invasive cervical cancer is considered curable. Definitive treatment traditionally involves hysterectomy, however, there is a growing body of literature to suggest that fertility-preserving interventions such as Loop Electrosurgical Excision Procedure (LEEP), cold knife cone (CKC), and trachelectomy may be acceptable surgical options for properly selected patients. ^8,9

Cervical cancer screening guidelines have changed many times over the last 50 years. Revised guidelines for the screening of cervical cancer in the United States were published in 2012 by the American Cancer Society (ACS), the American Society for Colposcopy and Cervical Pathology (ASCCP), the American Society for Clinical Pathology (ASCP), and the United States Preventative Task Force Services (USPSTF). ^10,11 Notable changes include delayed screening until age 21, screening every 3 years between the ages of 21 and 29, and the option of extending the 3-year screening interval for women aged 30–65 to every 5 years with cytologic and human papillomavirus (HPV) combined testing. According to abnormal Pap test management guidelines published in March 2013 by the ASCCP and in December 2013 by the American Congress of Obstetricians and Gynecologists (ACOG), complex algorithms recommend less invasive treatment strategies for women aged 21–24 with preinvasive disease. ^12,13 The diagnosis of cervical lesions is largely dependent on a program of frequent and consistent Pap screening. Cure rates are dependent on the ability to effectively screen, diagnose, and manage these patients expeditiously. Increasing Pap test screening intervals may increase the risk of invasive cervical cancer. In this study, we reviewed the cases of invasive carcinoma of cervix at our institution to examine how updated Pap test screening guidelines might affect the detection of invasive cervical cancer, especially among women <30 years of age.

Materials and Methods

An institutional review board-approved retrospective cohort study was conducted at Brigham and Women's Hospital. Patients diagnosed with ISCC and IAC of the cervix between the years 2002 and 2012 were identified. The cases were collected searching the hospital's electronic medical record (EMR) database using ICD-9 codes related to malignant neoplasms of the cervix, including 180.0, 180.1, 180.8, and 180.9. Patients were also identified by a review of tumor board and pathology records that documented all invasive cervical cancer cases during the designated study period. Individual medical records were reviewed for demographic information, past medical history, Pap test screening history, and follow-up data. Information on procedures performed at the time of diagnosis and treatment were also obtained. Pathology reports were reviewed to confirm histology and surgical margins. Cohorts were divided by age comparing patients younger than 30 years of age (Age Group 1) to patients 30 years and older (Age Group 2). This cohort division was selected to closely resemble updated guideline recommendations, which rely on Pap alone for screening among women under 30 years of age. Patients diagnosed with invasive cervical cancer by cervical biopsy, LEEP, CKC, trachelectomy, and hysterectomy were included. Patients were excluded if there was inadequate or no previous Pap history available. In this study, the Pap test interval is defined as the time from the last normal Pap test to the time of invasive cervical cancer diagnosis. Continuous variables were analyzed by Student's t-test or Wilcoxon rank sum test. Categorical variables were analyzed by chi square test or Fisher's exact test. Statistical tests were all two sided and differences were considered statistically significant at p < 0.05. Statistical analysis was performed using SPSS statistical software (version 16.0; SPSS, Inc., Chicago, IL).

Results

We identified 288 patients with invasive cervical cancer. A total of 109 patients had adequate documentation on prior Pap test screening history. The remaining 179 patients were excluded because of incomplete Pap test history. A Pap abnormality was confirmed in 67 (61.47%) patients at the time of presentation. Patients who were not diagnosed by a Pap abnormality were diagnosed by biopsy or incidentally at the time of hysterectomy. IAC was diagnosed in 37 (33.94%) patients while 64 (58.72%) patients were diagnosed with ISCC. Eight patients carried the diagnosis of other (7.34%), which included mixed adenosquamous, melanoma, and neuroendocrine histologies. The diagnosis of invasive cancer was made by cervical biopsy (79, 72.47%), LEEP or CKC (25, 22.94%), and hysterectomy (5, 4.59%). Table 1 describes the demographic data obtained. A total of 13 patients were identified as younger than 30 years of age with a mean age of 26 (18–29). Two patients in this age group were diagnosed at younger than 21 years of age. Six (46.15%) patients were nulliparous, four (30.77%) had a history of oral contraceptive pill (OCP) use during their reproductive life, one (7.69%) was a smoker, and one (7.69%) was pregnant at the time of diagnosis. There were no immunocompromised patients in this age group. Of the 13 patients identified, 5 (38.46%) presented asymptomatically. Eight (61.53%) patients had a confirmed Pap abnormality at the time of presentation. High-grade cervical squamous intraepithelial lesion (HSIL) was the most common cytological diagnosis (Table 2). The majority of patients in this group were diagnosed with invasive cervical cancer by cervical biopsy; however, smaller numbers of patients were also diagnosed by CKC (Table 3). IAC was diagnosed in seven (53.85%) patients in this group and ISCC was diagnosed in five (38.46%) patients. One patient in this age group was diagnosed with invasive neuroendocrine tumor of the cervix. The majority of patients presented with stage I disease and no patients in this age group presented with stage IV disease (Table 3).

A total of 96 patients identified were 30 years or older at the time of diagnosis. The mean age in this group was 46 (30–91). Six (6.25%) patients were nulliparous, 17 (17.71%) had a history of OCP use, 45 (46.88%) were smokers, and 5 (5.21%) were pregnant at the time of diagnosis. Among these patients, 47 (48.96%) presented asymptomatically and 59 (61.46%) had a confirmed Pap abnormality at presentation. Among those with a Pap abnormality, the most common cytological diagnosis was HSIL (Table 2). Various modalities were used to make the diagnosis of invasive cancer in this age group. The most common method was by cervical biopsy, followed by LEEP and CKC. Less commonly, women were diagnosed by hysterectomy (Table 3). In this group, 30 (31.25%) patients were diagnosed with IAC while 59 (61.46%) patients were diagnosed with ISCC. Atypical diagnoses seen in this age group included neuroendocrine, melanoma, and mixed adenosquamous histologies. Approximately 63% of patients presented with stage I disease and about 5% presented with stage IV disease.

Comparing all histologies, the mean time from normal Pap to diagnosis was 22 months in women younger than age 30 (Age Group 1) compared to 70 months in women greater than or equal to age 30 (Age Group 2) (p < 0.001). The mean time from normal Pap to diagnosis of IAC was 15 months in Age Group 1 compared to 56 months in Age Group 2 (p < 0.001). The mean time from normal Pap to diagnosis of ISCC was 38 months in Age Group 1 and 82 months in Age Group 2 (p = 0.018).

Follow-up data were available for all 13 (100%) patients in Age Group 1. The mean duration of follow up was 27 months (6–60). Surgical management in this patient population varied. One patient was treated with trachelectomy plus lymphadenectomy. Two patients were treated with CKC plus lymphadenectomy. CKC alone was used as surgical management in one patient. Of the remaining patients, four underwent radical hysterectomy and the other five did not undergo surgery in the setting of advanced disease at the time of presentation. Eight patients received upfront or adjuvant treatment with chemotherapy and or radiation therapy. There was one death in this age group. Only one patient in this group had recurrent disease at 36 months, while the remaining patients had no evidence of disease on follow-up.

Discussion

When diagnosed and treated early, invasive cervical cancer is a curable disease. Pap test screening for the early detection and management of dysplasia, carcinoma in situ, and invasive lesions is an effective way to reduce the incidence and mortality of invasive cervical cancer in the Unites States. Although Pap test screening guidelines have changed over the years, the update published in 2012 by the ACS, ASCCP, ASCP, and USPTF suggest the most lenient screening recommendations to date. ^10,11 A review of The Surveillance Epidemiology and End Results database (2004–2008) showed that ∼14% of invasive cervical cancers in the United States were diagnosed between the ages of 20 and 34. Among those diagnosed in this age group, more than half of the cases were diagnosed at <30 years of age with the majority of lesions being adenocarcinoma. ¹⁴ Our study confirmed that among women who are 30 years or older, new screening guidelines are not likely to miss many cancers. However, in women younger than 30 who are now screened and evaluated at greater intervals may see an increase in invasive cancers, particularly of adenocarcinoma histology.

The incidence of invasive cervical adenocarcinoma has dramatically increased over the past few decades, particularly in younger women. ¹⁵ An increased exposure to risk factors that promote glandular neoplasia, including endogenous (obesity) or exogenous (hormonal contraception) estrogens along with an increased prevalence of HPV 16 and 18 specific variants that are more associated with adenocarcinoma, may explain this trend. ^16,17 Although there is a large body of evidence to suggest that Pap screening for precancerous lesions of squamous histology reduces invasive cancer, screening for glandular lesions is less effective. ¹⁸ This is problematic in the age <30 group in that the majority of invasive cervical cancers diagnosed are of glandular histology. Smith et al. noted that although the incidence of squamous cell carcinoma of the cervix has declined over the past 24 years, the rate of adenocarcinoma relative to squamous cell carcinoma has doubled. ¹⁹ According to Bulk et al. a review of the Netherlands Cancer Registry (1989–1998) demonstrated that the incidence of adenocarcinoma has increased in women aged 15–29 by 15.8%. ²⁰ Adenocarcinomas are more challenging for Pap screening detection given that these lesions are typically located in the glandular cells of the endocervical canal. ²¹ Krane et al. described a decreased Pap test sensitivity to adenocarcinoma lesions secondary to the underrecognition of glandular neoplasia, in that it can resemble the cells from the lower uterine segment, tubal metaplasia, or reactive endocervical cells. ²² Additionally, approximately 10%–15% of patients present with multifocal or skip lesions increasing the complexity of the disease. ²³ Despite the challenges of using Pap test screening for glandular lesions, this is currently the standard of care for women under 30. The interval between preinvasive lesion and invasive disease suggests that the natural history of preinvasive disease developing into IAC is shorter than ISCC. ^24,25 Lee et al. described that, on an average, the interval between preinvasive disease and IAC is ∼5 years in their population of cases with an average age of 40.9. Since IAC can exist in the absence of a precursor lesion, the possibility of rapid disease progression exists. ²⁶ This may be even more true for younger patients <30 years of age.

In this study, an abnormal Pap test initiated the workup of cervical disease in about half the patients younger than 30 years of age. The remaining patients were either symptomatic at the time of presentation or were diagnosed incidentally. In this age group, the mean Pap test interval for IAC was less than the currently recommended 3-year screening interval. The mean Pap test interval for ISCC was approximately equal to 3 years, an important finding in relation to the updated recommended screening guidelines (Table 4). Under prior guidelines, patients were screened earlier, more frequently, and often treated for less concerning lesions. One study, by Barroilhet et al., reports that many abnormal Pap tests preceding the diagnosis of adenocarcinoma in situ were found to be atypical squamous cells of undetermined significance (ASCUS) or low-grade cervical squamous intraepithelial lesion (LSIL), a category that would be followed expectantly under the current guidelines. ²⁷ In the past, these patients with minor abnormalities, found in an aggressively screened and managed population, may have been detected early and removed from the normal screening population. This may explain why women who are 30 years and older, most of whom have had aggressive screening and management, may have greater reassurance about increasing the screening interval. Greater reassurance in this age group is also provided in the form of cotesting with cytology and HPV screening.

Based on the data collected at our institution, patients in the Age Group 1 following the new screening guidelines have the potential to develop IAC or ISCC before their next scheduled Pap. These findings also suggest that, in practice, current screening guidelines may not accurately detect a significant proportion of adenocarcinoma in situ lesions or IAC. It is possible that by delaying the onset of screening, not evaluating minor Pap abnormalities, or delaying treatment of more significant Pap abnormalities among women with no prior screening history (women aged 21–24) as suggested by national guidelines, the incidence of IAC may increase in women younger than 30 years of age.

It remains difficult to predict which patients in this study presented with invasive disease, because of a failure of the screening process or because of disease type and burden. It is unclear how a delay in diagnosis would affect treatment and prognosis with the updated guidelines.

This study has several limitations that must be considered carefully when interpreting the data. The greatest limitation of this study is its small size. Due to the very low incidence of invasive cervical cancer in the state of Massachusetts, the high rate of gynecologic oncology providers in the area, and poor documentation in the early years of the EMR system at our institution, only a very small number of invasive cervical cancers were identified. ^28,29 Given that our institution is a referral center for cervical cancer, the majority of the patients in this study were not diagnosed at our institution, but rather referred for treatment from surrounding outside clinics. Incomplete and inaccurate Pap test history documentation allowed for fewer than half the patients identified with cervical cancer eligible for analysis. A selection bias also exists in the setting of a single cancer institution, limiting its ability to be generalized. The retrospective nature of the study limits the ability to reliably determine prior results and management, as our database has no information regarding care at nonaffiliated institutions. We also did not have reliable HPV testing results for patients referred or screened at our institution given the time frame of the study.

Our findings among women over 30 years of age closely follow earlier published findings lending credibility to our results in women <30 years of age.

Although overscreening and treatment is problematic in women who are younger than 30 years of age, this age group may be particularly interested in fertility sparing options if diagnosed early. Currently, Pap testing is the only tool available to detect adenocarcinoma lesions of the cervix in this age group, and the level of Pap abnormality may not always reflect concerning glandular abnormalities. Perhaps by increasing the Pap test screening interval as suggested by new guidelines, but evaluating mild Pap abnormalities earlier, it may be possible to detect more concerning lesions sooner. This can lead to an early interruption of invasive cervical cancer and better fertility-sparing options. This small subgroup of patients may then be followed more closely without overscreening and overtesting the general population of women in this age group.

Based on our data, we agree with current guidelines suggesting screening starting at age 21. Two patients in this study were diagnosed at 21 years of age or less. Both presented with atypical histological diagnosis (neuroendocrine tumor and adenocarcinoma with clear cell features) and presented with a bleeding cervical mass. One can argue that the traditional Pap test is not helpful in the detection of rare lesions and that these patients presented symptomatically. We also agree with current guidelines suggesting screening every 3 years in low-risk patients aged 30 years and older based on the Pap timeline results of our study. At this time we cannot comment on HPV cotesting every 5 years in this age group, as we did not have data on cotesting for all patients given the time frame of the study. The interval from normal Pap test to invasive cancer diagnosis may be shorter than currently recommended screening intervals for women ages 21–29, based on our data. In particular, young patients <30 years of age may be at increased risk of developing IAC of the cervix between recommended screening intervals. Despite our findings, further investigations are needed to determine the true clinical significance of the change in Pap test screening guidelines. Given the small sample size, this study should be considered preliminary. With improved EMR, larger and more complete studies can provide more definitive data. Future research should be geared toward the impact of the new screening guidelines and Pap tests on the diagnosis of invasive cervical cancer in this age group.