CHAPTER      21

Cervix

MARCUS E. RANDALL PAULA M. FRACASSO TAKAFUMI TOITA SEAN S. TEDJARATI HELEN MICHAEL

EPIDEMIOLOGY, ETIOLOGY, AND RISK FACTORS

Worldwide, cervical cancer is the third most common malignancy and the second most common cancer (after breast cancer) in women. Nearly one-half million new cases occur each year. In 2012, it is estimated that 12,170 women will be diagnosed with invasive cervical cancer in the United States and over a third of them (4, 220) will likely succumb to the disease (1). These statistics stand in sharp contrast to those of the developing nations where cervical cancer is the second leading cause of cancer death in women with well over 500,000 new cases annually worldwide and over 250,000 deaths. These are thought to be underestimations of the true numbers and burden of the disease. These disparate statistics and toll are due to paucity of screening programs and lack of access to life-saving treatments for women diagnosed with preinvasive and invasive cervical cancer. The disease burden in the United States is disproportionate among minority and economically burdened populations (1). Over 50% of cervical cancers diagnosed in the United States occur in women with no prior screening and around 10% are in women without any screening in the preceding 5 years.

The incidence and mortality of this disease in North America have declined during the last half century owing to both increased availability of Pap smear screening and a decrease in fertility rate. An extensive discussion of cervical cancer etiology and risk factors appears earlier in the text, particularly in Chapters 1 to 3 and 8. The reader is referred to those chapters for detailed information.

ANATOMY

The uterus, situated in the center of the pelvis, is composed of a muscular layer (myometrium) and mucosal glandular layer (endometrium), which lines a hollow cavity extending from the top of the uterus (fundus) through the lower uterine segment connecting to the endocervical canal. The endometrial cavity is surrounded by the myometrium. The three segments of the uterus include the fundus, body, and isthmus (lower uterine segment). Anteriorly, the lower uterine segment is attached to the bladder peritoneum (vesico-uterine pouch), which requires dissection and separation to expose the cervix. This compartment between the bladder and uterus is the anterior cul-de-sac. Posteriorly the peritoneum lines and separates the lower uterine segment and cervix from the rectum, creating the posterior cul-de-sac (pouch of Douglas). The cervix is an extension of the lower uterine segment and can be divided into vaginal and supravaginal components. The cervix varies in length, averaging 3 to 5 cm. Centrally located in the vaginal portion is the external cervical os, which connects to the endocervical canal, the internal cervical os, and the endometrial cavity.

The cervix is a fibrous organ lined by squamous and columnar cells. The transition from columnar cells to squamous cells occurs in the region of the cervical os in what is termed the transformation zone. Most dysplastic and malignant cervical lesions arise from this area. The endocervix contains mostly mucinous glandular epithelium with similar morphology to that of the fallopian tube.

The uterus has 5 ligamentous attachments. Specifically these include the round, broad, utero-ovarian, cardinal, and utero-sacral ligaments. The round ligaments run from both sides of the fundus anterolaterally on top of the broad ligaments to the pelvic sidewall, where they are closely related to the inferior epigastric vessels and continue on their path to leave the pelvis through the inguinal ring and canal. They contain a vessel known as Sampson’s artery as well as other small vessels and nerves.

The broad ligaments attach to the lateral margin of the uterus on either side and extend underneath the round ligaments from above the cervico-uterine junction and to the pelvic sidewall. These ligaments are covered by peritoneum anteriorly and posteriorly, and are relatively thinner above the cervico-uterine junction. Within the leaves of the broad ligament are extraperitoneal connective tissues known as the parametrium, which surround the lower uterine segment and the cervix.

The utero-ovarian ligaments originate just below the fallopian tubes and directly connect to the ovaries. The cardinal ligaments, also known as Mackenrodt’s ligaments, are thickened extensions of the broad ligaments attaching laterally from the lower uterine segment (isthmus) and the cervix. They course to the lateral vaginal wall and rectal pillars and extend to the pelvic sidewalls.

Each cardinal ligament helps create and define 2 important anatomical spaces, the pararectal and paravesical spaces, which are important in the surgical treatment of early cervical cancer. The cardinal ligament covers a portion of the pelvic ureter, which lies underneath the uterine artery and above the uterine vein about 2 cm away from the isthmus. It contains uterine, vaginal, middle rectal, and inferior vesical arteries and veins as well as lymphatic tissue and channels.

Two important anatomical landmarks during radical hysterectomy (RH) for treatment of early cervical cancer are the paravesical and pararectal spaces. The paravesical space is lateral to the bladder and can be opened by dissection underneath the insertion of the round ligament into the pelvic sidewall. This space is opened using blunt dissection underneath the round ligament dissection along the curve of the bony pelvic wall proceeding medially to open the relatively avascular space. The anterior wall is bounded by the superior pubic ramus, the medial wall is the bladder and vagina, the lateral wall is composed of the external iliac vessels and parts of the obturator fossa, and the posterior wall is formed by the cardinal ligament, separating this space from the pararectal space. This space can be continuous with the space between the bladder and the pubic bone, known as the space of Retzius. The pararectal space is found between the ureter laterally and the origin of the hypogastric artery medially and follows the curve of the sacrum. Anteriorly it is bounded by the cardinal ligament. The lateral wall is formed by the hypogastric vessels. Medially it is bounded by the ureter, and the floor is composed of the curve of the sacrum down to the levator muscles. During a RH, these two spaces allow for direct examination of the parametrium to determine extent of the disease and the resectability of the primary tumor allowing for free surgical margins.

The uterosacral ligaments connect laterally to the cervix and lower uterine segment medial to the ureters. They run from the posteriolateral segment of the cervix over the top of the anterolateral portion of the rectum and attach to the rectal wall and the sacrum. The cardinal and utero-sacral ligaments provide the bulk of support for the uterus and cervix.

The blood supply of the uterus is mainly through the uterine artery, originating from the anterior division of the hypogastric artery. An important landmark during a RH, the uterine artery spawns cervical and vaginal branches that supply the cervix, and then extends superiorly to the fundus where its tributaries supply the fundus and the body of the uterus. The uterine artery also descends toward the vagina and gives off a vaginal branch. The ovarian vessels originating from the aorta also contribute to the blood supply of the uterus. There is a rich anastomosis between the uterine and ovarian arteries. Other potential collateral blood supply to the uterus may be from the branches of the posterior division of the hypogastric, the ilio-lumbar artery, or other vessels supplying the pelvis such as middle sacral artery, and pudendal and obturator arteries. The venous drainage follows the arterial supply.

The lymphatic drainage of the cervix is rich and complex. Laterally from the cervix, the lymphatics drain into the paracervical and parametrial lymph nodes and to the internal, external, and common iliac chains. The obturator lymph nodes are the most medial portion of the external iliac nodal region and play a key role as one of the first sites of lymphatic spread of cervical cancer following the parametrial nodal tissues. Other lymphatic drainage routes include the inferior and superior gluteal lymph nodes and the superior rectal, presacral, and paraaortic lymph nodes. The innervations of the cervix are from the sacral roots (S2–4) crossing through the hypogastric plexuses.

NATURAL HISTORY

Preinvasive Disease

Cervical cancer is preceded by an interval of epithelial dysplastic changes, typically in the transformation zone, known as cervical intraepithelial neoplasia (CIN), which may progress to invasive cancer. Low-grade dysplasia (CIN 1) is confined to the basal one-third of the epithelium, whereas high-grade dysplasia (CIN 2 to 3) involves two-thirds or greater of the epithelial thickness. Full thickness involvement is known as carcinoma in situ (CIS), which is thought to be the most significant precursor lesion for invasive squamous carcinoma. Cytologic screening has helped in decreasing the incidence of cervical cancer in the United States and other developed nations dramatically. Of approximately 55 million Pap smears performed in the United States annually, 5% to 7% are abnormal, and the majority of these are atypical squamous cells of undetermined significance (ASCUS) (2). The majority of the disease burden is in developing nations where screening and treatment programs are scarce and difficult to implement.

Many studies have evaluated the natural history of cervical dysplasia. Holowaty et al. evaluated cervical dysplasia in about 17,000 women with mild, moderate, and severe dysplasia (3). The reported risk of progression from mild to severe dysplasia was 2% and 6% at years 2 and 5, respectively. However, progression from moderate to severe dysplasia was 16% and 25% at 2 and 5 years, respectively. The relative risks of progression from severe dysplasia to CIS and invasive cancer were 4.2 and 2.5, respectively, 2 years after diagnosis of dysplasia. The majority of mild dysplasias (CIN 1) regressed to normal within 2 years.

In a meta-analysis of 15 studies of 27,929 women, Melnikow et al. found 2-year progression rates to high-grade dysplasia to be 7% and 21% from ASCUS and low-grade dysplasia, respectively (4). The rate of progression to invasive cancer for ASCUS was 0.25%, as compared to low-grade dysplasia (0.15%) and high-grade dysplasia (1.4%). The regression rates to normal for ASCUS, low-grade, and high-grade dysplasias were 68%, 47%, and 35%, respectively.

The incidence of progression from CIS to invasive cancer in patients with persistent abnormal cytology after initial treatment was 24.8 times greater than in those who had normal cytology after initial treatment (5). Overall the reported rate of progression of CIS to invasive cancer ranges from 12% to 22%.

Most low-grade dysplasia will regress to normal in about 24 months. Therefore, only careful follow-up is warranted. However, women with high-grade disease or CIS should be more closely evaluated and treated aggressively as the risk of progression is higher (5). A great deal is being learned about the natural history of this disease; as there are many variations including a subset of CIS lesions that can rapidly present as invasive cancer within a couple of years. Many glandular lesions associated with HPV 18 can also herald a rapid progression to invasive disease proving evasive even in adequately screened populations. Emerging data strongly suggest that despite negative cytology, a woman who tests positive for highly oncogenic HPV 16 has a significantly higher chance of having or developing CIS. These will be discussed in detail in the screening section.

Patterns of Spread

Cervical cancer can spread through direct extension to the endocervix, lower uterine segment, parametrium, vagina, and, less often, to the bladder and/or rectum. It can also spread lymphatically to the parametrial, obturator, and internal, external, and common iliac lymph nodes. Inferior and superior gluteal, superior rectal, presacral, and paraaortic lymph nodes can also become involved. The pattern of spread is usually predictable and orderly, as the rate of paraaortic lymph node metastasis is very rare if the pelvic nodes are spared. The overall risk of pelvic lymph node metastasis in stage IB is about 9% to 17% (6,7). The risk of paraaortic nodal (PANL) metastasis is higher with advanced stage disease, with rates of 16% and 25% for stage II and III, respectively (8). Scalene nodal involvement is reported to be between 10% and 23%, and involvement of higher paraaortic nodes can suggest possible scalene nodal involvement (9).

The most common sites of hematogenous spread include lung, mediastinum, bone, and liver. Other less common sites are spleen, brain, and adrenal gland. Most recurrences occur in the first 24 months, with a median of 17 months.

CLINICAL PRESENTATION AND DIAGNOSTIC EVALUATION

Preclinical Invasive Disease

Screening and Cytology

The effectiveness of screening cytology is related to the ease of obtaining specimens, the sensitivity of the test, and a long dysplastic phase that can be detected via cytology and effectively treated with high success rates. There are currently 2 primary means of obtaining and diagnosing cervical dysplasia. One is the conventional Pap smear and the other is liquid-based thin layer preparation.

Using classification as low-grade squamous intraepithelial lesion (LSIL) as the threshold, Nanda et al. reported a large variation in the sensitivity of conventional smears, with published data ranging from 30% to 87%; specificity was much higher at 86% to almost 100% (10). Cytologic screening is a continuum in which specificity is increased through sequential testing. The most common reasons for false-negative smears are errors in sampling and preparation. In 1996, liquid-based cytology was introduced in the United States. Cells are obtained by a cytobrush and transferred into a buffered alcohol medium. The suspended cells are then plated in a thin, evenly distributed layer, improving the clarity of the slide for evaluation. Another benefit is that the specimen can be tested for HPV, gonorrhea, and chlamydia. A careful review of data suggests that liquid-based cytology may be superior to conventional cytology in minimizing the number of unsatisfactory smears and increasing detection rates of low- and high-grade abnormalities (11). Sulik et al. reported no difference between the 2 methods in a large systematic review (12). The American Cancer Society (ACS) and U.S. Preventive Services Task Force (USPSTF) have concluded that, although liquid-based technology may be more sensitive than conventional cytology, there was insufficient evidence to recommend universal adoption. The American College of Obstetricians and Gynecologists (ACOG) does not support changing screening intervals based on the method of screening (13).

The changes in recommendations for screening published in 2012 have been sweeping. Screening is not recommended in anyone under the age of 21 regardless of the age of sexual debut. For women aged between 21 and 29, cytology alone is recommended every 3 years. HPV testing is not recommended in this age group, as there is a high prevalence of infection in this population, with many being transient. The rationale is that cotesting with HPV will only increase the number of colposcopies and will result in harm without reducing the risk of CIN 3 or invasive cancer. For patients aged 30 to 65, the recommendations are for cytology alone every 3 years versus cytology and HPV co-testing every 5 years, with the latter being the preferred recommendation.

Screening should stop at age 65 for women with 3 negative consecutive previous cytology results or 2 consecutive negative cytology and HPV testings within the last 10 years, with the most recent being within last 5 years. Once screening is discontinued, it should not be resumed even if the woman has a new sexual partner (14,15). It is important to note that following regression or appropriate clinical treatment of CIN 2 or 3 and AIS, screening should continue for 20 years, even if that exceeds the threshold of 65 years of age. Annual screening results in an over 50% increase in the number of colposcopies when compared to screening every 3 years, without adversely affecting the outcomes.

Women who have undergone hysterectomy for a prior diagnosis of CIN 2 or 3 or a more severe diagnosis should continue screening for 20 years. Those with no prior history of cervical dysplasia should not undergo routine vaginal cytology. Women who have been vaccinated for HPV should follow the age-specific recommendations, since 30% of cervical cancers are caused by etiologies other than HPV 16 and 18.

If HPV is negative and cytology is ASCUS, then rescreening with co-testing in 5 years or with cytology alone in 3 years is reasonable. If HPV is positive and cytology is negative, then a 12-month follow-up with repeat co-testing or subtesting for HPV16 and 18 can be done and, if positive, should herald a referral for colposcopy given the high rate of coexistent CIN 2 or 3 with positive testing for types 16 or 18. If the subtyping is negative, then a 12-month follow-up with co-testing is reasonable. Detection of adenocarcinoma is much more difficult than detection of squamous lesions. HPV 18 positivity with negative cytology may indicate a significant underlying glandular lesion (16).

To provide a standard terminology for reporting cytologic findings, the Bethesda system set forth guidelines in 1988, with revisions in 2001 (17). This system describes the adequacy of the specimen, a general categorization of whether it falls within normal limits or not, and a description of the dysplastic cytologic abnormalities detected in 3 categories: ASCUS or atypical squamous cells—cannot rule out high-grade squamous intraepithelial lesion (ASC-H), LSIL (the same as CIN 1), and high-grade squamous intraepithelial lesion (HSIL), which is the same as CIN 2 or 3. It can also report squamous cell carcinoma directly. For glandular lesions, categories include atypical glandular cells (AGS), denoting origin if possible (endocervix, endometrial, or not otherwise specified), AGS favoring neoplastic, adenocarcinoma in situ (AIS), or frank adenocarcinoma.

It is estimated that approximately 10% to 20% of women with ASCUS or ASC-H have underlying CIN 2 or 3 and that 1 in 1,000 may have underlying invasive cancer. ASC-H is expected to be diagnosed in about 5% to 10% of all ASCUS readings. AGS is associated with a much higher percentage of high-grade disease when compared to ASCUS—as high as 39%.

The ASCUS/LSIL Triage Study (ALTS), a multicenter randomized trial of 3,488 women, compared immediate colposcopy versus triage to colposcopy based on HPV testing, and liquid-based cytology with a threshold of HSIL versus conservative management with triage based on repeat cytology with a threshold of HSIL (18,19). ASCUS was reproducible in repeat liquid-based cytology in only 32% of cases with an overall CIN 2 and 3 diagnosis of 15% when ASCUS cytology was further investigated. HPV testing was able to identify 92% of women who would ultimately have a diagnosis of CIN 3. Only 1.4% of women with negative HPV testing developed CIN 3 over 2 years. If HPV testing was positive with LSIL or ASCUS was the same around 27%. Repeat cytology for ASCUS would have referred almost 67% of women to colposcopy as compared to HPV testing, which referred about half the number of women for colposcopy. The American Society of Colposcopy and Cervical Pathology (ASCCP) endorsed HPV testing as the preferred method of triage for ASCUS smears detected by liquid-based cytology (20).

HPV Testing

HPV DNA testing is a new tool that has gained greater popularity and utility in the last few years. It is especially useful in women over 30 when combined with cytology, because it provides much greater sensitivity for detection of CIN 2–3 disease, although it is less specific (16). Methods of detecting HPV include hybrid capture 2 and polymerase chain reaction. Both tests have similar sensitivity and specificity, but the former is easier to perform technically, especially in a large screening program setting. Sherman et al. studied over 20,000 women, evaluating the role of conventional Pap smear and concurrent HPV testing to determine the risk of developing CIN 3 (21). In this study, 72% of women who developed CIN 3 had baseline abnormal cytology and positive HPV testing with a cumulative incidence of developing CIN 3 or invasive cancer of 4.5% in approximately 45 months of follow-up. This was compared with women with negative smears and HPV testing whose cumulative incidence of developing CIN 3 or invasive cancer was only 0.16%. The combination of negative cytology and HPV testing carries a high-negative predictive value of over 95% for the development of CIN 3 or invasive cancer, providing great reassurance for low-risk women and lengthening the interval of screening in that population.

Clavel et al. studied the role of HPV testing as a primary screening tool in almost 8,000 women (22). They reported 100% sensitivity of HPV testing in detecting histologically proven high-grade dysplasia compared to conventional smear (68%) and liquid-based cytology (87%). The specificity of HPV testing was 93% in women over 30, compared to conventional and liquid-based technologies with specificities of 86% and 95%, respectively. Overall it was found that HPV testing is much more sensitive in detecting high-grade disease compared to conventional smears, but with a lower specificity rate. This approach requires further study; screening women younger than 30 years may lead to unnecessary testing, as many HPV infections are transient, and most women will not develop dysplasia with HPV infection. In LSIL, more than 85% of cases test positive for HPV. Therefore, HPV testing is not encouraged. CIN 2 and 3 require immediate referral for colposcopic evaluation. The overall 2-year cumulative risk of developing CIN 2 and 3 for LSIL and ASCUS with positive HPV testing was essentially the same (27.6% and 26.7%). The authors suggested that repeat HPV testing 12 months after diagnosis of LSIL and ASCUS with positive HPV yielded the highest sensitivity and lowest rate of referral for colposcopy. Repeat cytology with HPV testing did not significantly increase sensitivity and increased referral for colposcopy. HPV testing appears to be useful in triage of ASCUS smears, possibly reducing colposcopic referrals by almost 50% (17,18). HPV subtyping for 16 and 18 may even be more helpful in further delineating the risk of CIN 3 and subsequent risk of invasive cancer. HPV 16 is found in 55% to 60% of all cervical cancers. A patient with a positive test for HPV 16 has a high risk of CIN 3 or more severe lesion of almost 30% if there is persistent HPV 16 infection for around 2 years. CIN 3 has a roughly 30% chance of progression to invasive cervical cancer over a three-decade period, and treated CIN 3 has less than a 1% risk. CIN 3 is the greatest predictor of development of cervical cancer (16).

A large study (ATHENA) that included 40,901 evaluable patients with positive HPV testing and liquid-based cytology test results was recently reported. Of these, 4,275 women (10%) tested HPV-positive and 2,617 (6%) had abnormal cytology. Four hundred and thirty-one women were diagnosed with CIN 2 or worse, and 274 with CIN 3 or worse. In the subset of women who had colposcopy, the HPV testing was more sensitive than liquid-based cytology for detection of CIN 3 or worse (92% vs. 53.3%). The addition of liquid-based cytology to HPV testing improved sensitivity for CIN 3 or worse to 96.7%; however, this contributed to a greater number of positive screenings (by 35%) compared with HPV testing alone. The rate of positive testing for HRHPV was 6.7%, and 1.5% were positive for types 16 and 18. Overall incidence of CIN 2–3 was 1.2%. The absolute risk of a CIN 2 or worse lesion was 11.4% if positive for HPV 16/18 compared to 6.1% if positive for HRHPV and 0.8% if HPV-negative. These results highlight the importance of types 16 and 18 in the development of CIN 3 and a referral for colposcopy if positive, despite negative cytology (23).

Two commercial tests for HPV DNA have been approved by the U.S. Food and Drug Administration. The hybrid capture II test has been shown to have excellent interlaboratory reliability and reproducibility (24).

Colposcopy

Colposcopy examines the lower genital tract including the vulva, vagina, and epithelium of the cervix and opening of the endocervix. The use of acetic acid and Lugol’s solution assists in highlighting abnormal and dysplastic changes. Changes such as aceto-white plaques and vascular abnormalities such as punctations, mosaicism, and abnormal branching may signify high-grade disease. Indications for colposcopy include an abnormal-appearing cervix, persistent postcoital bleeding or discharge, persistent CIN 1, 2, or 3 on cytology, in utero exposure to DES, and ASCUS smears with positive high-risk HPV testing (16 and 18). In order to have an adequate colposcopic examination the entire transformation zone must be fully visualized. Endocervical curettage (ECC) can be helpful if positive; however, its role as a routine procedure is debated. Positive predictive value of colposcopy can be increased by multiquadrant biopsies.

Conization Biopsy

Cervical conization refers to the surgical removal of the squamo-columnar junction. This may be done in the operating suite through a classical cold knife conization technique or as an outpatient procedure using thermal cautery with loop excision. Indications for conization include inadequate colposcopy, positive ECC, persistent CIN 1 (usually >1 year), CIN 2, CIN 3, or CIS, and discrepancy between cytologic, colposcopic, or pathologic findings.

Loop Electrosurgical Excision Procedure

Loop diathermy, an alternative to cold knife conization, is usually performed on an outpatient basis. Trials have demonstrated loop electrosurgical excision procedure (LEEP) to be as effective as cold knife conization, with advantages in terms of cost, anesthesia, and ease of use (25). A disadvantage is that significant thermal artifact can hamper evaluation of margin status, which can be minimized with careful technique avoiding prolonged contact with the tissue.

Clinical Disease

Symptoms and Complaints

The most common presentations of invasive cervical cancer are abnormal vaginal bleeding, postcoital bleeding, and vaginal discharge. However, many women are asymptomatic and are found to have disease upon pelvic examination or cytologic evaluation. As the tumor enlarges, it can cause local symptoms such as pelvic pain and difficulty with urination or defecation. As the disease metastasizes to regional lymph nodes, back pain, leg swelling (especially unilateral), and neuropathic pain may occur.

Physical Findings

The most common finding on physical examination is an abnormal lesion on the cervix, at times necrotic and friable. Possible extension onto the vaginal wall should be assessed. Determination of parametrial, sidewall, and uterosacral ligament involvement must be done through a rectovaginal examination. Other areas of concern are the superficial groin and femoral lymph nodes and the supraclavicular region.

Diagnostic Biopsy

Suspicious lesions on the cervix and surrounding areas should be biopsied with enough depth to assure adequate nonnecrotic tissue to render diagnosis. Biopsy on the margin tends to yield better results.

Staging

Clinical Staging Procedures

The most critical component of staging is a thorough pelvic examination, including a rectovaginal exam. If necessary, an exam under anesthesia allows for a careful visual and digital examination as well as cystoscopy and proctoscopy.

Laboratory Studies

A complete peripheral blood count is indicated to evaluate for anemia, which might warrant correction. Thrombocytosis is found in up to 30% of patients with locally advanced disease, and has been associated with larger tumor size, parametrial involvement, and poorer survival. Serum chemistries with attention to serum creatinine are essential to evaluate for renal failure or to assess renal function in patients who will receive cisplatin-based chemoradiotherapy. Additional tests should include liver function and urinalysis.

Serum Tumor Markers

Though several tumor markers are thought to be useful as prognostic and predictive markers in response to treatment, risk of relapse, and overall survival (OS) in women with cervical cancer, their role in the initial diagnostic evaluation is limited (26,27). Elevation of the most commonly studied tumor markers, squamous cell carcinoma antigen (SCCA), cytokeratin fragment 19 (CYFRA 21.1), and cancer antigen 125 (CA-125), has been observed in 20% to 88% of women with newly diagnosed cervical cancer (26). These elevated levels have been correlated with tumor stage, tumor size, cervical stromal invasion, lymphovascular space invasion, parametrial involvement, and lymph node status (26). In a retrospective study by Huang et al. (28), pre-and posttreatment levels of SCCA and carcinoembryonic antigen (CEA) were evaluated in 188 women with squamous cell carcinoma (SCC) of the cervix receiving concurrent chemoradiation. Pretreatment marker levels predicting high risk for paraaortic lymph node recurrence (66.5% at 5 years) were SCCA ≥ 40 ng/mL or the combination of CEA ≥ 10 ng/mL and SCCA < 10 ng/mL, for intermediate risk were SCCA 10 to 40 ng/mL, and for low risk were both SCCA and CEA < 10 ng/mL (28). Kotowicz et al. (29) assessed pelvic and paraaortic lymph nodes in 182 untreated patients with cervical cancer (87% SCC and 13% adenocarcinoma) and evaluated levels of the tumor markers SCCA, CYFRA 21.1, CA-125, and CEA, and the cytokines interleukin-6 (IL-6) and vascular endothelial growth factor (VEGF). SCCA, CA-125, and IL-6 were significantly higher in women with lymph node metastases than in those with no lymph node involvement. The simultaneous elevation of SCCA and CA-125 levels or SCCA and IL-6 was statistically significant in women with lymph node metastases.

Sheng et al. (30) evaluated SCCA, CYFRA 21.1, CEA, and a nuclear protein found to play a role in tumor development, growth, and metastases, high mobility group box-1 (HMGB1), in early detection of recurrent SCC of the cervix. In 112 women with recurrent SCC of the cervix, 174 women without evidence of recurrent cervical cancer, and 128 age-matched healthy controls, HMGB1 levels were significantly higher in women with recurrent cervical cancer than in women from the other groups (p = 0.027). Measurements with the combined tumor markers HMGB1, SCCA, and CYFRA 21.1 were reported to increase the diagnostic sensitivity and specificity (30). Further studies are required to confirm the use of tumor markers alone or in combination for the early detection of recurrence.

Radiographic Studies

In accordance with the International Federation of Gynecology and Obstetrics (FIGO) system for staging, clinical evaluation is the predominant tool for staging cervical cancer (31–33). Noninvasive radiographic imaging studies such as chest radiography, intravenous pyelography (IVP), cystoscopy, sigmoidoscopy, and barium enema may aid in this process but are not mandatory (31–33). Additionally, radiographic studies such as computed tomography (CT), positron emission tomography (PET), combined PET/CT, and magnetic resonance imaging (MRI) are not considered integral to formal staging, but may aid in determination of extent of disease in the pelvis and lymph nodes, evaluation of prognosis, measurement of response, and assessment of recurrence following initial treatment (34–37).

Given that evaluation of the extent of disease in the pelvis and lymph nodes is important for individualizing treatment, advanced imaging with CT, PET/CT, and MRI should be performed preoperatively when available (37). MRI and CT imaging are the most valuable imaging techniques for evaluation of disease in the pelvis and chest/abdomen, respectively. Due to its superb soft tissue delineation, high-resolution MRI with contrast and diffusion-weighted imaging is useful for the evaluation of disease in the pelvis assessing tumor size, extension into the uterine corpus, depth of stromal invasion, parametrial and pelvic wall extension (34,35). In a randomized, prospective study performed by the American College of Radiology Imaging Network (ACRIN) and the Gynecologic Oncology Group (GOG), MRI more accurately determined tumor diameter, uterine extension, and parametrial involvement than did CT or clinical examination in early invasive cervical cancer (38–41). However, in large, bulky cervical tumors, MRI may overestimate parametrial extension and may less accurately determine upper vagina, bladder, and/or rectum invasion (34).

PET/CT plays a valuable role in the imaging of nodal metastases (36,42–44). Kidd et al. reported a large prospective study to evaluate the role of pretreatment [¹⁸F] fluorodeoxyglucose PET imaging for lymph node staging in 560 women with clinical stage IA1–IVB disease (42). The frequency and pattern of lymph node metastases on PET were associated with clinical staging. Although this study did not confirm lymph node status by histopathology, disease-specific survival correlated with PET-positive lymph nodes. Choi et al. published a meta-analysis comparing the diagnostic performance of CT, MRI and PET or PET/CT for detection of nodal metastases in 41 published reports (43). PET or PET/CT showed the highest pooled sensitivity (82%) and specificity (95%), while CT showed 50% and 92%; and MRI, 56% and 91%, respectively (43). Kang et al. performed another meta-analysis assessing the diagnostic value of PET for the evaluation of PALN metastases in 10 published reports (44). Their analyses indicated assessment of PALN metastases by PET or PET/CT is acceptable for higher stage cervical cancer, but for earlier stage disease, PALN dissection is preferable. In summary, PET or PET/CT has a higher diagnostic performance for detecting lymph node metastases than CT or MRI, especially in women with higher-stage cervical cancer.

Future studies combining MRI and PET may provide additional staging information. In a retrospective study by Kim et al., lymph node status was evaluated comparing MRI/PET fusion and PET/CT prior to lymphandenectomy in 79 women with histologically confirmed invasive Stage IB–IVA cervical cancer (45). Review of lymph node status with MRI/PET fused images revealed identification of 6 additional metastatic nodal groups in 3 additional patients as compared to PET/CT images. Sensitivity and specificity of PET/CT was 44.1% and 93.9%, respectively, while for fused MRI/PET was 54.2% and 92.7%, respectively (45).

In conclusion, both MRI and PET/CT imaging modalities are recognized as being very useful in the staging and subsequent treatment planning of this disease. There is further interest in performing dynamic contrast-enhanced MRI to evaluate tumor vascularity and perfusion, and PET/CT to evaluate intratumoral metabolic activity, and using this information to predict response to therapy (46–48). As the technology of these imaging modalities continues to advance, integration of both anatomic and biologic information will permit further individualization of treatment.

FIGO and American Joint Committee on Cancer Staging

The staging system used for cervical cancer is based on the FIGO staging system, updated in 2009 (33). The main change from the 1995 staging system involves stage IIA. Stage IA is microscopic, with IA1 defined as stromal invasion to less than 3 mm in depth and no more than 7 mm wide, and stage IA2 defined as stromal invasion 3 to 5 mm deep with width no greater than 7 mm. The Society of Gynecologic Oncology (SGO) also considers negative lymphovascular space invasion (LVSI) and margin status as criteria for microinvasive disease (MID). The horizontal spread must be 7 mm or less to be considered MID according to FIGO, but SGO does not include lateral spread in defining MID. The diagnosis of MID must be made on a conization or hysterectomy specimen. Punch biopsies are not adequate. Stage IB involves the cervix only. It is divided into IB1, which includes all lesions greater than IA2 and no more than 4 cm in greatest dimension, and stage IB2 for lesions measuring greater than 4 cm. Stage II invades beyond the uterus but not to the pelvic wall or to the lower third of the vagina. It is divided into A and B. Stage IIA involves the proximal vagina and is now subdivided similarly to stage IB into IIA1, which includes tumors less than 4 cm, and IIA2 for tumors greater than 4 cm involving the proximal vagina and sparing the parametrium. Stage IIB indicates parametrial invasion. In Stage III, the tumor extends to the pelvic wall and/or involves lower third of the vagina and/or causes hydronephrosis or non-functioning kidney. Stage IIIA extends to the lower third of the vagina sparing the pelvic sidewall, while IIIB extends to the pelvic wall and/or causes hydronephrosis or nonfunctioning kidney without alternate reason for such findings.

Surgical findings and radiographically guided biopsies of suspected lesions such as lymph nodes or lung metastasis cannot be used to change or modify clinical FIGO staging. In Stage IV, the tumor extends beyond the true pelvis or has involved the mucosa of the bladder or rectum (biopsy proven). Bullous edema alone does not permit a case to be allotted to stage IVA. Stage IVB denotes distant metastasis. A tumor-node-metastasis (TNM) staging system proposed by the American Joint Committee on Cancer (AJCC) corresponds well to the FIGO clinical staging. The current criteria for the various stages are defined in Table 21.1. TNM is mainly used in documenting findings on surgical and pathologic evaluations as the pathologic stage of the disease, and FIGO is used for clinical staging. All histologic types are included. If there is ambiguity regarding the correct stage, the lower stage is assigned.

Note: The definitions of the T categories correspond to the stages accepted by the Fédération Internationale de Gynécologie et d’Obstetrique (FIGO). Both systems are included for comparison.
^aAll macroscopically visible lesions—even with superficial invasion—are T1b/IB.

Source: AJCC. Cervix uteri. In: Edge SB, Byrd DR, Compton CC, et al., eds. AJCC Cancer Staging Manual. 7th ed. New York, NY: Springer; 2010:395–402.

Pretreatment Nodal Staging

The presence of paraaortic lymph node metastases has a significantly negative impact on progression-free and overall survival. Pretreatment knowledge of paraaortic spread could potentially direct treatment in a manner that could improve outcomes. Given the limitations of imaging in reliably detecting paraaortic micrometastases, surgical staging has been used. However, the role of surgical staging for locally advanced cervical cancer remains unclear. The use of PET/CT scans has become widely adopted in Western countries with a relatively low sensitivity rate of 30% to 40%, depending on the study. However, a relatively high negative predictive value of 80% to 90% has been reported in various studies. Ramirez et al. studied 65 patients with stages IB2 to IVA comparing radiographic (PET/CT) versus surgical staging of paraaortic nodes (49). Positive PA nodes were found in 23%; 22% of those with positive histologic nodes had negative radiographic findings. PET/CT was superior to CT or MRI alone when nodes were negative on those modalities, with a specificity of 96%. Treatment was modified in 18% due to surgical findings. Kidd et al. evaluated 560 patients with PET/CT and correlated finding with prognosis (50). Almost half (47%) had positive nodes on PET/CT, which conferred a poor prognosis with a hazard ratio of recurrence based on positive nodes, depending on location of pelvic, paraaortic, and supraclavicular, of 2.4, 5.8 and 30, respectively. Gold et al. reviewed 555 patients with surgical paraaortic node sampling and 130 who had radiographic staging. For stages III and IV, the 4-year progression-free survival (PFS) was 49% versus 36%, with an OS of 54% versus 40%, favoring patients who were surgically staged (51).

With the advent of laparoscopy and the presumed lower morbidity, there has been some renewed interest in surgical nodal assessment as a method to refine treatment. Faggotti et al. reviewed 13 published series to investigate the role of pretreatment laparoscopic staging. They showed laparoscopic surgical staging can be performed safely and contribute to tailor treatment. However, the positive impact on DFS has still to be demonstrated (52).

Leblanc et al. evaluated 184 patients with stages IB2 through IVA with pretreatment extraperitoneal laparoscopic staging of paraaortic nodes (53). They found a 24% incidence of PANL involvement and an operative complication rate of 2%. They had a 13% symptomatic lymphocyst rate that decreased to about 3% with modification of techniques. Findings of surgical staging changed the management of 20% of patients who would have received only pelvic radical trachelectomy (RT) to extended field RT. Data suggested a survival advantage in patients with resected positive nodes.

Lai et al. compared clinical staging versus surgical staging (laparoscopic or extraperitoneal approach) in 61 patients (54). PANL metastasis was detected in 25% of the surgical group. The study was terminated early, as the surgical arm had a significantly worse progression-free and overall survival compared to the nonsurgical arm. The question of pretreatment surgical staging remains controversial, and ongoing prospective studies are in progress to determine the utility of surgical staging in the setting of quality PET/CT imaging (49–51).

Sentinel Lymphatic Mapping in Early Cervical Cancer

The standard surgical treatment for early cervical cancer (stages IA2 to IIA) is RH and pelvic lymphadenectomy (PL). The PL is complete and removes all nodal bearing tissues on the external iliac artery and vein (up to the circumflex iliac vein), internal iliac artery above the obturator nerve, and common iliac artery up to the iliac bifurcation. The procedure carries some risk of injury to vessels and nerves, can lead to lymphedema, and can predispose to a higher complication rate if postoperative RT is needed. Lymphatic mapping (LM) refers to the concept of delineating the path of malignant cells through the lymphatics, identifying the nodes primarily at risk for early spread, and determining if this knowledge will safely allow a limited rather than a complete regional lymphadenectomy.

Thompson and Uren described the sentinel node (SLN) as “any lymph node that receives lymphatic drainage directly from primary tumor” (55). SLN has revolutionized care for women with breast cancer and for those with melanoma. It has also been shown to be of benefit in patients with vulvar cancer. In cervical cancer many reports have shown feasibility and high detection rate of SLN in stages 1A2 to IB. Levenback et al. reported 39 patients with early cervical cancer who underwent RH and PL. Preoperative lymphoscintigraphy was performed, revealing at least one SLN in 85%. Most SLNs were iliac, obturator, and parametrial nodes; however, approximately 20% were in the common iliac and aortic chains. Eight (21%) had nodes involved with tumor, and 5 of these 8 patients had SLN as the only positive node. Sensitivity was 87.5% and negative predictive value (NPV) was 97% (56).

Plante et al. followed with a larger series of 70 patients using laparoscopy. Bilateral SLNs were found in 60% of cases. Using both a gamma probe and blue dye, the SLN detection rate was 93%. Eighty-eight percent of SLNs were in the external iliac, obturator, and bifurcation area. Fifty-one percent had 2 SLNs, and 24% had ≥SNs. The NPV was 100% and sensitivity was 93%. The rate of allergy to blue dye was 3% (57). Roy et al. studied 211 patients. One SLN was identified in 97% of patients, 16.7% of SLN were in aberrant sites, including about 4% in the paraaortic region. Of the latter group, 15% had a positive node. With ultrastaging the NPV of SLN was 100%, compared with frozen section sensitivity of 94%. The results were best in patients with tumor diameter of less than 2 cm (58). Lecuru et al. studied 139 patients and found at least 1 SLN in 98%. The NPV was 98% but the reliability was significant when bilateral SLN were identified (59). Diaz et al. evaluated their single-institution experience at Memorial Sloan Kettering Hospital and reported a 95% detection rate of SLN. Routine pathologic processing identified 71% of involved SLN with an additional 29% found on ultrastaging (60). Altgassen et al. reported a multi-institutional study of 590 patients, showing a detection rate of 93.5% when blue dye was used in conjunction with technetium-99. The sensitivity was 77% with a NPV of 94% (61).

At the present time, SLN is still investigational and not standard of care in cervical cancer. The overall SLN identification is around 90%, with a sensitivity of 92%, and an approximately 8% false negative rate, with a NPV of 97%. Challenges to adoption include the sensitivity of frozen section, availability of pathologic expertise, uniformity of technique, surgical experience, and clinical impact of tumor size (62).

PATHOLOGY

Squamous Cell Carcinoma

Stay updated, free articles. Join our Telegram channel

Join