eMedicine - Cervix, Cancer : Article by Faysal A Saksouk

HPV and Cervical Cancer Resource Center

HPV and Cervical Cancer Resource Center

View all HPV and Cervical Cancer Articles

HPV and Cervical Cancer Multimedia Library

Quick Find

Authors & Editors
Introduction
Differentials
CT Scan
MRI
Ultrasound
Nuclear Medicine
Intervention
Multimedia
References

Related Articles

Endometrium, Carcinoma

Leiomyoma, Uterus (Fibroid)

Patient Education

Cancer and Tumors Center

Procedures Center

Women�s Health Center

Cervical Cancer Overview

Cervical Cancer Causes

Cervical Cancer Symptoms

Cervical Cancer Treatment

Pap Smear Introduction

Background

Cancer of the uterine cervix is largely a preventable disease that is characterized by a long lead time; precancerous lesions gradually progress through recognizable stages before developing into invasive disease.^{1, 2, 3, 4, 5, 6, 7} The disease process is almost certainly curable if it is identified before its progression to invasive cancer. However, invasive cervical cancer remains a disease of significant morbidity, and it is a major cause of cancer deaths in women worldwide, although the incidence and mortality rates of invasive cervical cancer have declined substantially (particularly in countries that have well-developed screening programs).^{8, 9} Cancer of the cervix in its early stages is readily managed with surgery. Radiation or chemoradiation therapies are reserved for high-risk early stages or advanced disease.

Human papilloma virus (HPV) is now recognized as the most important causative agent in cervical carcinogenesis at the molecular level, although HPV may not induce many of the identified molecular alterations.^{10, 11} As many as 5% of cervical cancers may not be associated with HPV.^{12, 13} First intercourse at an early age, sexual promiscuity, high parity, race, and low socioeconomic status are presently thought to increase the risk for cervical cancer because these factors are linked to sexual behavior that increases the likelihood of exposure to HPV and/or because they are cofactors that modify the risk in women who are infected with HPV. Tobacco smoking is also a significant independent risk factor.

Pathophysiology

Cancer of the cervix typically originates from a dysplastic or premalignant lesion previously present at the active squamocolumnar junction. The transformation from mild dysplastic to invasive carcinoma generally occurs slowly within several years, although the rate of this process varies widely.

Carcinoma in situ is particularly known to precede invasive cervical cancer in most cases. In different reported series of patients with untreated carcinoma in situ who were followed up for many years, invasive carcinoma developed in about 30% of patients at 10 years and in about 80% of patients at 30 years. However, the carcinoma-in-situ lesion may regress after the initial diagnosis; such an occurrence was reported in 17 (25%) of 67 patients who were followed up for at least 3 years. Progression to invasive carcinoma becomes established and is considered irreversible once the malignant process extends through the basement membrane and invasion of the cervical stroma occurs.

Multiple local growth patterns of invasive cervical cancer have been described, with combination growth patterns being common. The patterns include the following: exophytic, nodular, infiltrative, and ulcerative.

The exophytic variety is the most common growth pattern. It usually arises from the exocervix and is often polypoid or papillary in form. Exophytic cervical cancer may result in a large, friable, bulky mass that involves only the superficial aspect of the cervix and has the tendency for excessive bleeding.

The nodular variety typically arises in the endocervix and grows through the cervical stroma into confluent, firm masses that cause the cervix and isthmus to expand. Large, nodular-type tumors that circumferentially involve the endocervical region and large, exophytic-type tumors that originate from the endocervix and extend into the endocervical canal result in what has been referred to as a barrel-shaped cervix.

The infiltrative growth pattern leads to a stone-hard cervix that may be predicated to have minimal visible ulcerations or an exophytic mass. Infiltrative exocervical lesions tend to invade the vaginal fornices and the upper part of the vagina. On the other hand, infiltrative endocervical lesions tend to extend into the corpus and the lateral parametrium.

The ulcerative growth pattern is associated with tumor necrosis and sloughing, with the formation of a cavity that is marginated by the invasive tumor. This process is usually complicated by infection that causes seropurulent discharge.

The main pathways for the spread of invasive cervical cancer consist of the following: (1) microscopic spread into the vaginal mucosa beyond a visible or palpable tumor; (2) extension into the endometrium or myometrium of the corpus; (3) direct extension into the parametrium and, in advanced stages, into the adjacent structures; and (4) spread into the regional pelvic lymph nodes and, potentially, into the retroperitoneal, inguinal, or thoracic lymph nodes.

The extrauterine spread of cervical cancer occurs primarily by means of direct extension and lymphatic invasion that initially affects the contiguous tissues in the region of the laterally positioned cardinal ligament. Later, the tumor may involve the anterior or posterior parametrium, which are separated from the cervix by the pubovesicocervical fascia. The lateral paracervical and parametrial regions are more vulnerable to tumor invasion than the anterior and posterior parametrium because of the lack of a protective fascial covering at the lateral regions and because of the natural lymphatic drainage through the lateral paracervical tissues into the cardinal ligaments.

The local spread of cervical cancer may progress through the parametrium to involve the ureters and, eventually, the pelvic sidewalls. In some patients, the sciatic plexus is involved. Hydronephrosis, pyelonephritis, and renal failure are common complications of progressive disease. Involvement of the urinary bladder and rectum can occur in advanced cases because of direct tumor extension or subsequent to invasion of the vesicouterine or uterosacral ligaments, respectively. A vesicovaginal fistula or rectovaginal fistula may or may not develop. A tumor that extends through the posterior aspect of the cervix or corpus infrequently leads to intraperitoneal spread. Adnexal metastases are uncommon in the earlystages of the disease.

Lymphatic tumor spread usually occurs in a fairly orderly pattern or sequence that first involves the regional paracervical and parametrial lymph nodes and then the internal and external iliac lymph nodes. This may then be followed by spread to the common iliac nodes, the para-aortic nodes, and eventually the supraclavicular nodes via the thoracic duct. Metastasis to the para-aortic lymph nodes without involvement of pelvic lymph nodes is unusual.

Hematogenous tumor spread may be a result of a lymphatic venous anastomosis or direct venous invasion. The most common sites of hematogenous metastases are the lungs, bones, and liver.

Squamous cell carcinoma (SCC) accounted for 80-90% of all cervical malignancies in a large case series.¹⁴ Most of the information regarding the etiology and epidemiology of cervical cancer is derived from experience and research that are related to the most common SCC lesion. The major histopathologic SCC subtypes include: (1) well-differentiated, keratinizing, large-cell SCC (25% of cases); (2) moderately differentiated, nonkeratinizing, large-cell SCC (70% of cases); and (3) small-cell undifferentiated carcinoma (about 5% of cases), which is associated with a distinctly poor prognosis.

Pure adenocarcinomas arise from endocervical-type cells and constitute 5-20% of all cervical malignancies. A trend toward an absolute increase in the incidence of adenocarcinoma has been observed during the past 20-30 years.^{12, 13} In particular, among women younger than 35 years, the incidence more than doubled between 1970 and the mid-1980s.^{12, 13} The histologic patterns include well-differentiated mucinous adenocarcinoma, papillary adenocarcinoma, and a clear-cell pattern that contains glycogen but no mucin. Many of these lesions may simulate endometrial adenocarcinoma. There is a relatively higher incidence of poorly differentiated and more aggressive histologic subtypes of cervical adenocarcinoma that are associated with a poorer prognosis than SCC.

Frequency

United States

Invasive cancer of the cervix is the third most common genital malignancy in women, after cancers of the endometrium and the ovary.¹² The American Cancer Society estimated approximately 11,150 cases of invasive cervical cancer were diagnosed in 2007.⁸ The incidence of carcinoma in situ is estimated to be about 4 times that of invasive cancer.⁸

International

Worldwide, invasive cervical cancer is the most common genital female malignancy and the second most common malignancy in women, after breast cancer.¹³ The estimated total number of new cases is 371,200 per year worldwide or 9.8% of all cancers in women.

Invasive cervical cancer is more common in economically disadvantaged developing countries, in which 78% of worldwide cervical cancers occur.¹³ These account for 15% of all cancers in women. The associated lifetime risk of invasive cervical cancer in such countries is about 3%. In developed countries, the disease accounts for 4.4% of all new cancers in women, and it is associated with a lifetime risk of about 1.1%. This disparity is attributed to the lack of effective screening programs in developing countries that have a high incidence of cervical cancer.

Reported global cancer statistics show that the incidence is highest in Latin America, the Caribbean, southern Asia, Southeast Asia, and sub-Saharan Africa.¹³

Mortality/Morbidity

The prognosis of cervical cancer is relatively good in low-risk countries, with a reported 5-year survival rate of 72% for all stages combined.⁸ The 5-year survival rate is reportedly about 48% in developing countries, where patients are likely to seek medical attention when the cancer is more advanced.¹³ The 5-year survival rate for early invasive cancer is about 92%, and that for preinvasive cervical cancer is nearly 100%.⁸

However, invasive cervical cancer is second only to breast cancer as a leading cause of worldwide cancer-related mortality in women. Approximately 190,000 deaths per year occur worldwide as a result of cervical cancer; these account for 8.5% of all cancer deaths in women. In the United States, an estimated 3670 women died from cervical cancer in 2007.⁸

The presence of metastatic adenopathy is an important factor in the prognosis of cervical cancer.

Race

Compared with non-Hispanic white women, the incidence in black women is 50% higher, and the 5-year survival rate is lower.⁸ The rate in Hispanic women is over twice that of non-Hispanic white women.⁸

Age

Cervical carcinoma in situ is most commonly detected in women aged 25-34 years. Invasive cervical cancer is most frequently diagnosed in women older than 50 years, and it is uncommon in women younger than 25 years.

Anatomy

The uterus is a pear-shaped muscular organ with thick walls and a flattened hollow cavity. It consists of the corpus, or body, and the cervix, or neck. The cervix is in the most caudal position and is a relatively narrow, cylindrical segment of the uterus that measures approximately 2-4 cm in length. The uterine corpus and cervix are externally demarcated from each other by a subtle constriction called the isthmus, and they are internally demarcated by a slight narrowing of the endocervical canal called the internal os.

The cervix enters the vagina through the anterior vaginal wall, forming an oblique attachment line that separates it into a supravaginal segment and a lower intravaginal segment (portio vaginalis and exocervix). The supravaginal segment is posteriorly covered by the peritoneum in the region of the posterior cul-de-sac (pouch of Douglas or rectovaginal pouch). Anteriorly, it is separated from the urinary bladder by fatty tissue without being covered by the uterovesical pouch of the peritoneum. Laterally, it is connected to the broad ligament and the parametrium, the extraperitoneal connective tissue through which the vasculature reaches the cervix.

The vaginal segment projects into the vaginal vault and consists of about one third of the anterior aspect and one half of the posterior aspect of the cervix. At the vaginal surface of the cervix is the external os, which is generally a small, circular aperture. However, the external os may be oval or almost linear. The usually spindle-shaped endocervical canal extends from the external os to the internal os, where it joins the endometrial cavity. The transition from endocervical to endometrial glands is the histologic landmark for the internal os.

The cervix has a preponderance of fibrous tissue, as compared with the predominantly muscular uterine corpus. The cervical stroma is primarily composed of collagenous connective tissue with a small amount of interspersed stratified muscle fibers (about 10% of the stromal tissues) and a small amount of elastic tissue.

The exocervix (portio vaginalis) is covered by stratified squamous epithelium that is essentially identical to the vaginal epithelium. A single layer of columnar epithelium, supported by a basement membrane, lines the endocervical surface. The squamocolumnar junction is located at the exocervix in young individuals and gradually migrates into the endocervical region with age as the exocervical lips atrophy. Approximately 90% of SCCs arise from the squamocolumnar junction. The mucosa of the endocervical canal forms branching intramucosal folds called the plicae palmatae.

The cervix has a rich network of lymphatics that drain principally into the paracervical lymph nodes and subsequently to the hypogastric and external iliac nodes (of which the obturator nodes are the innermost component). The pelvic lymphatics drain into the common iliac and the para-aortic lymph nodes.

Clinical Details

Early cervical cancer is usually asymptomatic, and as many as 20% of patients who have invasive cervical cancer are asymptomatic when the disease is diagnosed by means of a Papanicolaou (Pap) smear/test or routine clinical examination.

Approximately 80-90% of patients with cervical cancer experience a form of abnormal vaginal bleeding such as postmenopausal bleeding, irregular menses, heavy menstrual flow, painless metrorrhagia, or postcoital bleeding. In some case series and in geographic regions where endometrial cancer is not common, postmenopausal bleeding is the most common presenting symptom of cervical cancer. Abnormal vaginal discharge is a presenting symptom in about 10% of patients; the discharge may be watery, purulent, or mucoid. Pelvic or abdominal pain and urinary or rectal symptoms occur in advanced cases.

Related Medscape topic:
Resource Center Women's Sexual Health

Preferred Examination

The American Cancer Society guidelines for the early detection of cervical cancer recommend screening with an annual conventional Pap smear (or every 2 y with the newer liquid-based Pap test) and a pelvic examination in all women approximately 3 years after they begin having vaginal intercourse but no later than age 21 years.¹⁵ In women aged 30 years and older, the Pap test may be performed less frequently (usually every 2-3 y), at the discretion of the physician, after 3 consecutive normal results.

An alternative is for women older than age 30 years to be screened every 3 years with either the conventional or liquid-based Pap test, in addition to the HPV deoxyribonucleic acid (DNA) test.¹⁵ Women with certain risk factors (eg, prenatal diethylstilbestrol [DES] exposure, human immunodeficiency virus [HIV] infection, etc) should be screened annually. The National Comprehensive Cancer Network (NCCN) and American Society for Colposcopy and Cervical Pathology have similar guidelines.^{16, 17}

Screening for cervical intraepithelial neoplasia (CIN) with the Pap smear allows the early detection of preinvasive disease, and the test is generally accepted to be effective in reducing the incidence and mortality rate of cervical cancer. Nonetheless, the Pap smear is only a screening tool, as opposed to a definitive diagnostic procedure, and it has a 15-25% false-negative rate in the detection of cervical dysplasia.

No gross cervical abnormality may be visualized during the speculum pelvic examination if the tumor is small or if it is located in the endocervix. As many as one third of the cancers are endocervical. In this location, some fairly large tumors may escape inspection, or they might be appreciated only with bimanual rectovaginal examination. However, in most patients, a visible lesion is detected, and it may be ulcerative, plaquelike, or exophytic.

Except in pregnant patients, colposcopy-directed biopsy and endocervical curettage are advised when a cervical lesion is suspected on the basis of the clinical findings or if the Pap smear reveals a precancerous lesion or malignant cells. Punch biopsy of any gross cervical lesion should be performed, and samples of any suspicious areas in all 4 quadrants of the cervix or vagina should be obtained. Curettage of the endometrium may also be performed if tumor extension in a superior direction is suspected.

Colposcopy is usually adequate for the evaluation of the exocervix and of a segment of the endocervix near the transition of the squamous and columnar epithelium. Conization biopsy is used to evaluate a subclinical tumor when the colposcopic results are insufficient or inadequate, when the colposcopic biopsy findings suggest microinvasive cancer, when the endocervical curettage samples show dysplastic fragments, or when no gross cervical lesion is visualized and endocervical disease is suspected. Conization is usually contraindicated in patients with overt cancer because of the increased risk of hemorrhage and treatment complications. Dysplasia and carcinoma in situ may be managed with cold-knife conization or with other excision or ablation methods such as laser conization, loop electrosurgical excision procedures (LEEP), laser vaporization, and cryotherapy. Cervical conization may eventually prove to be therapeutic in many patients.

The pretreatment evaluation of patients with cervical cancer includes physical examination, chest radiography, and intravenous urography (IVU) or cross-sectional imaging (computed tomography [CT] scanning or magnetic resonance imaging [MRI]). In early-stage disease with a small tumor confined to the cervix, IVU and cross-sectional imaging are not routinely performed because of their relatively low yield.^{24, 25} IVU is not needed when cross-sectional imaging is performed because both modalities are similarly accurate in depicting urinary obstruction and because cross-sectional imaging has the additional ability to depict a gross tumor that involves the urinary tract. Barium enema examination, radioisotope bone scanning, cystoscopy, and proctosigmoidoscopy have a low yield, particularly in early disease, and these procedures are performed for only specific indications that are based on the symptoms or clinical findings.

MRI has excellent soft-tissue contrast resolution, which exceeds that of CT scanning and ultrasonography (US).^{26, 27, 28, 29, 30, 31, 32, 33} Consequently, MRI is significantly more valuable than CT and US in the assessment of the size of the tumor, the depth of the cervical invasion, and the locoregional extent of the disease (direct invasion of the parametrium, pelvic sidewall, bladder, or rectum).^{29, 34, 35, 36, 37, 38, 39} CT scanning and MRI are approximately equivalent, and both are significantly superior to US, in the detection of enlarged lymph nodes.^{40, 41, 42, 43, 44, 45, 46} Overall, CT scanning and MRI are more accurate staging modalities than US. Furthermore, US is not suited for staging of the full extent of the tumor spread because of the inability of this technique to adequately depict all the potential sites of metastasis or the anatomic regions that contain lymph nodes.

Despite the advantages of MRI, the gynecology literature mostly recommends the use of CT scanning for the pretreatment evaluation of cervical cancer.^{24, 47, 48, 49} Reportedly, the additional information provided with the excellent soft-tissue contrast resolution of MRI often has no significant effect on clinical decision making or on the choice of therapy.⁵⁰ In general, CT scanning and MRI are not warranted in patients with small-volume early disease (stage Ib disease and a cervical tumor diameter <2.0 cm) because of the low probability of parametrial invasion and nodal metastasis. Imaging with CT scanning or MRI is appropriate when the cervical tumor is larger than 2.0 cm, when the size of the tumor cannot be adequately evaluated during the clinical examination, or when the tumor is endocervical.

The International Federation of Gynecology and Obstetrics (FIGO) believes that any staging system should be universally feasible and applicable, as well as provide a worldwide standardized classification that allows various medical centers to compare results.

According to FIGO, only clinical staging fulfills these criteria, and therefore, the staging classification of cervical cancer should be entirely based on findings from the pretreatment clinical evaluation. Costly CT scanning and MRI examinations are generally not readily available worldwide, the technology and imaging techniques are variable, and the quality of the imaging interpretation is not uniform or ensured. Surgicopathologic staging is not available or feasible in patients with early disease who are receiving radiation therapy, is not indicated for patients with advanced disease, and may not be an option in countries where the economic resources and surgical expertise are limited.

The examinations permitted by FIGO for consideration in the clinicodiagnostic staging include palpation, inspection, colposcopy, biopsy, endocervical curettage, hysteroscopy, cystoscopy, proctoscopy, IVU, and radiographic evaluation of the lungs and skeleton.⁵¹ The physical examination is one of the most valuable components of the clinical staging process. A meticulous, bimanual rectovaginal examination of the pelvis should be performed (preferably with the patient under anesthesia) to evaluate potential sites of locoregional tumor spread such as the parametrium, uterosacral ligaments, pelvic sidewalls, rectum, and bladder. Any suspected tumor invasion of the rectum or bladder should be confirmed by means of endoscopically guided biopsy.

The strict FIGO clinical staging guidelines do not include the status of the lymph nodes, although the presence of metastatic adenopathy is an important factor in treatment planning and in the prognosis. Extended clinical staging with cross-sectional imaging (CT scanning and/or MRI) includes the status of the lymph nodes in the assessment of the extent of the disease. The detection of enlarged pelvic lymph nodes is considered equivalent to pelvic sidewall tumor extension (stage III), and the detection of enlarged lymph nodes in the para-aortic, paracaval, or inguinal regions is considered extrapelvic tumor spread (stage IV).

The major limitations of the FIGO clinical staging system are encountered in the estimation of the size of the primary tumor, particularly when the tumor is endocervical. The size of the tumor is significant because, in each stage, the incidence of lymph node metastases increases and the prognosis deteriorates with increased volume of the primary tumor. Other limitations occur in the evaluation of tumor extension into the parametrium and pelvic sidewalls and in the detection of metastatic lymphadenopathy or distant metastasis.

Extended clinical staging with cross-sectional imaging (CT scanning and/or MRI) and surgicopathologic staging, including pelvic and abdominal retroperitoneal lymphadenectomy, provide additional diagnostic value. Each has been proven to be superior to the conventional FIGO clinical staging system in determining the full extent of the tumor spread. However, once the clinical stage is assigned on the basis of the clinical pretreatment workup results (in compliance with the FIGO guidelines), the stage should not be altered as a result of subsequent findings. Instead, any additional information that is revealed by cross-sectional imaging or surgery is primarily used for planning treatment regimens, and they should not be used to revise the assigned clinical stage.

Limitations of Techniques

CT scanning uses ionizing radiation and the CT image quality is degraded by metallic prostheses, an extremely large body habitus, and patient or respiratory motion. The use of intravenous iodinated contrast material for CT imaging is associated with a risk of significant allergic reactions (including fatal anaphylaxis), nephrotoxicity, and complications due to its extravasation into the soft tissues at the injection site.

MRI is contraindicated in patients who have vital metallic biomedical devices or metallic objects in strategic anatomic regions. MRI is more costly and less readily available than CT scanning and requires long image acquisition times. The image quality is degraded by artifacts that are related to respiratory motion and bowel peristalsis, which are likely to occur during the long image acquisition time. No effective gastrointestinal (GI) contrast material is currently available for MRI. Claustrophobia deters some patients from undergoing MRI.

US is operator dependent. The image quality is degraded by a large body habitus, and visualization of portions of the pelvis and abdomen is precluded by bowel gas and bony structures. The transabdominal approach is also influenced by the degree of bladder filling and is impeded by the presence of surgical incisions, dressings, drains, or skin lesions. Transvaginal US and transrectal US (TRUS) probes have inherent limitations, including a small field of view, a short range of target penetration with high-frequency transducers, and occasional patient intolerance of the transvaginal or transrectal approach.

Additional limitations and pitfalls encountered in the staging of cervical cancer are described in the section for each specific imaging modality (see CT Scan, MRI, Ultrasound, and Nuclear Medicine).

Other Problems to Be Considered

Cervical leiomyoma
Cervicitis
Changes related to surgical biopsy
Complex nabothian cyst
Endocervical polyp
Lymphoma and other cervical or myometrial malignancies

Findings

CT scanning is the imaging modality that is most commonly used in clinical practice to evaluate the extent of spread of cervical cancer. The oral, rectal, or intravenous administration of contrast material is necessary for optimal CT scan evaluation (unless a contraindication exists). The intravenous injection of contrast material is particularly useful in increasing the conspicuity of the cervical tumor and in facilitating the evaluation of the parametria and the pelvic sidewalls.

The normal uterine cervix is an ovoid or round structure with homogeneous soft-tissue attenuation. The cervix shows peripheral enhancement on the earliest images that are obtained with the intravenous administration of contrast material using a power injector. The contrast enhancement rapidly becomes diffuse and uniform throughout the cervix, but it may not be as intense as the myometrial enhancement because of the preponderance of fibrous tissues in the cervical stroma. To the author's knowledge, no precise CT scan criteria for the normal size of the cervix have been published, although the cervix is anecdotally described to be generally smaller than 3 cm in diameter.

Cervical cancer and the normal cervical stroma usually have similar attenuations on CT images that are obtained without intravenous contrast enhancement. Therefore, the tumor and the normal cervical parenchyma cannot be reliably distinguished on nonenhanced CT scans, and the cervix may have a normal CT scan appearance. Alternatively, the only detectable finding may be an enlarged cervix with homogeneous attenuation and regular or irregular contours.

After the intravenous administration of contrast material, the manifestations of cervical cancer include the following: (1) a cervix with a normal CT scan appearance, (2) an enlarged cervix with normal contrast enhancement, (3) an enlarged cervix with inhomogeneous areas of hypoattenuation but without a discrete mass that is clearly delineated or definitely evident (see Images 4 and 11), and (4) an enlarged cervix with a circumscribed solid mass that has an enhancement that is less than that of the normal cervical stroma and shows a homogeneous or heterogeneous hypoattenuation (see Images 1, 8, and 12).

The hypoattenuation of the mass may be a manifestation of the inherent properties or vascularity of the tumor tissue. Areas of necrosis or ulceration are common and are depicted as foci of accentuated low attenuation in the hypoattenuating mass. Collections of air are sometimes evident in the mass, and they may be the result of a recent surgical biopsy, tumor necrosis, or infection (see Image 3). Cervical cancer frequently obstructs the endocervical canal, resulting in accumulation of serous, hemorrhagic, or purulent fluid in the uterine cavity (see Image 2).

Because CT scanning does not consistently allow the direct visualization of the primary tumor, an evaluation of the size of the tumor and the depth of stromal penetration is precluded in many patients. The evaluation of the lymph nodes is based on morphologic size criteria. Pelvic, abdominal retroperitoneal, celiac axis, and mesenteric lymph nodes greater than 1.0 cm in their short axis are considered enlarged.^{21, 52, 53} The occasional detection of intranodal necrosis is most consistent with metastasis; this necrosis is best depicted on contrast-enhanced images.

The extended clinical staging of cervical carcinoma with cross-sectional imaging (CT and/or MRI) is based on the FIGO criteria (see Clinical staging in the Preferred Examination section).

Stage Ib tumors are isoattenuating relative to the normal cervical stroma in as many as 50% of patients. The cervix may have a normal CT scan appearance, or the only detectable finding may be an enlarged cervix with homogeneous attenuation. Alternatively, the enlarged cervix may show inhomogeneous areas of hypoattenuation or a hypoattenuating solid mass (as described in General CT scan findings, above).

CT scan features of invasive cancer confined to the cervix (stage I) include the following (see Image 1): smooth, well-defined, and intact peripheral cervical contours; lack of parametrial soft-tissue strands or masses; and preservation of the periureteral fat planes.

CT scan features of parametrial tumor invasion (stage IIb) include the following (in order of increasing positive predictive value [PPV]): loss of definition, irregularity, or nodularity of the cervical contours; increased attenuation and prominent or thick soft-tissue strands in the parametrial and/or periureteral fat (see Images 4-5); confluent soft tissue that replaces the periureteral fat (see Image 9); and an eccentric 3-dimensional (3-D) parametrial soft-tissue mass (see Images 10-11).

In stage IIIa, the tumor extends into the lower one third of the vagina, without spread to the pelvic sidewall. In stage IIIb, the tumor extends into the pelvic sidewall (see Image 13), and/or a nonfunctioning kidney or hydronephrosis due to invasion of the ureter is present (see Images 15-16). The depiction of enlarged pelvic lymph nodes on CT scans is considered equivalent to pelvic sidewall tumor extension (see Images 12 and 16).

CT scan features of pelvic sidewall tumor invasion (stage IIIb) include the following: a parametrial tumor within 2-3 mm of the pelvic sidewall; irregular, thick, soft-tissue strands or confluent soft tissues that extend through the parametrium to the obturator internus muscle or piriformis muscle (see Image 13); a confluent mass that incorporates the pelvic sidewall muscles and obliterates the associated fat planes; and encasement or distortion of iliac vessels by the tumor.

CT scan features of tumor extension into the bladder or rectum (stage IVa) include the following: focal obliteration of the perivesical or perirectal fat (see Images 10 and 13); eccentric or asymmetric wall thickening that can be uniform, nodular, or serrated (see Images 10 and 13); intraluminal extension of a soft-tissue mass (see Image 17); and a vesicovaginal fistula.

The CT scan finding of enlarged inguinal and/or retroperitoneal lymph nodes is considered a feature of stage IVb (see Image 14).

The pitfalls and limitations of CT scanning in staging cervical cancer include the following:

Degree of Confidence

Clinical staging by experienced gynecologic oncologists is more accurate than CT scanning in evaluating patients with stage Ib or IIa tumors. CT scanning is not recommended for the routine staging of early cervical cancer, and the value of this modality is also limited in the assessment of parametrial tumor invasion (stage IIb tumors), with an accuracy in the range of 30-76%.

The accuracy of CT scanning increases in the setting of advanced disease, reportedly as high as 92% in evaluating patients with stage IIIb-IVb disease. The accuracy of CT scanning in the detection of enlarged lymph nodes secondary to metastasis is about 65-80% in the pelvis and about 80-98% in the abdominal retroperitoneum. However, the accuracy of CT scanning in the overall staging of cervical cancer is reportedly about 58-88%.

CT scanning is therefore most valuable (1) in evaluating patients with advanced disease that has progressed beyond the uterus (stage IIb or worse), (2) in detecting enlarged lymph nodes and in guiding biopsy of those nodes, and (3) in depicting urinary tract involvement (obviating the need for IVU). Note: CT scanning is not consistently reliable for evaluating tumor size, and it is not sufficiently accurate for detecting early invasion of the parametrium, vagina, bladder, or rectum. Vaginal extension of cervical carcinoma is best evaluated with the use of clinical methods.

False Positives/Negatives

The pitfalls and limitations of CT scan staging that are described in the sections above lead to false-negative and false-positive results.

A published report revealed that CT scanning did not depict 10 (33%) of 30 cervical cancers or 49 (49%) of 99 cervical cancers in another study because of the small size of the tumor or because of its isoattenuation relative to the normal cervical stroma and myometrium. In the assessment of parametrial tumor invasion, the reported false-positive rate of CT scanning is in the range of 35-70%. CT scanning causes understaging in 12-33% of patients with clinical stages IIb-IIIb disease because of false-negative findings in areas of microscopic tumor spread.^{22, 48}

Hricak and Yu reported diagnostic test statistics based on compiled data that they obtained from several published studies.²¹ In the evaluation of parametrial invasion, CT scanning had an overall accuracy of 76%, a PPV of 58%, and a negative predictive value [NPV] of 85%; in the detection of lymph node metastasis, the values were an 85% overall accuracy, a PPV of 61%, and an NPV of 87%. In some patients, treatment planning may require CT-guided fine-needle aspiration (FNA) biopsy of lymph nodes that appear enlarged on CT scans because of reported false-positive rates of at least 5-10%.

Findings

Exquisite high-resolution images of the female pelvis can be obtained by using current state-of-the-art MRI technology, including high-field-strength systems with torso phased-array coils or specialized endorectal or endovaginal coils. However, published reports indicate that the strength of the magnetic field and the type of coil used have no significant impact on the accuracy of MRI in detecting or evaluating the size of the primary cervical tumor, in depicting the locoregional extent of the tumor, or in staging the overall disease process.^{29, 54}

The routine use of high-resolution sagittal and axial T2-weighted fast spin-echo (FSE) sequences and axial T1-weighted spoiled gradient-echo sequences is advocated. The sagittal T2-weighted images facilitate the evaluation of the primary cervical tumor and of the tumoral extension into the uterine corpus, vagina, bladder, or rectum. The axial images are critical for evaluating the extent of stromal penetration and for detecting parametrial invasion.

The administration of an antiperistaltic agent such as glucagon may be useful in diminishing internal motion artifacts. The value and role of intravenous contrast material in the MRI staging of cervical cancer are controversial. Additional investigative studies are needed to determine the appropriate indications and optimal techniques for contrast-enhanced imaging.

The anatomic details of the normal cervix are best depicted on T2-weighted images.⁵⁵ The epithelium and the mucus in the endocervical canal appear together as a spindle-shaped central zone with a high signal intensity that is similar to that of the endometrium. The surrounding cervical stroma is depicted as a relatively wide band with a predominantly low signal intensity, which likely reflects the overall preponderance of fibrous tissue. The outermost cervical stroma is depicted as a thin layer of intermediate signal intensity, which likely reflects the predominance of smooth muscle in this region of the stroma.

High-resolution T2-weighted images obtained with specialized coils may allow visualization of the mucosal plicae palmatae as a feathery layer of intermediate signal intensity interposed between the central high signal intensity of the endocervical region and the low signal intensity of the stroma. On T1-weighted images and proton-density-weighted images, the normal cervix has a homogeneous intermediate to low signal intensity; however, the above-described internal anatomic details are not discernible.

T2-weighted images depict the vaginal mucosa with a high signal intensity and the vaginal wall with an intermediate signal intensity. T1-weighted images and proton-density-weighted images depict the vaginal mucosa and wall as a single structure with a homogeneous intermediate to low signal intensity. The intravenous administration of gadolinium-based contrast material usually causes greater enhancement of the cervical and vaginal mucosa and the outermost cervical stromal layer than of the predominantly fibrous cervical stroma or the vaginal wall.⁵⁶ Parametrial enhancement also occurs.

Cervical carcinoma has a long T1 and a long T2 that render this disease relatively hyperintense on proton-density-weighted and T2-weighted images (see Images 18-21), on which the signal intensity of the tumor is approximately similar to that of the endometrium. The hyperintense appearance on T2-weighted images accounts for the essentially consistent conspicuity of the primary tumor against the background of the hypointense normal cervical stroma.

The tumor has a medium signal intensity on T1-weighted images. Cervical carcinoma exhibits variable contrast enhancement after the intravenous administration of gadolinium chelates. Some tumors may even lose their conspicuity as they become isointense with the surrounding contrast-enhanced stroma.

Direct visualization of the primary cervical neoplasm allows MRI to be reliable in assessing the size of the tumor and the extent of stromal invasion or penetration. Partial stromal invasion is characterized by the presence of an intact layer of low-signal-intensity stroma at the periphery of the cervix, which is visualized on T2-weighted images as a dark stripe or ring that surrounds the hyperintense tumor; this finding virtually excludes tumor spread into the parametrium.

Full-thickness stromal invasion is typically depicted on T2-weighted images as disruption of the hypointense or dark-appearing stromal stripe by an abnormal high-signal-intensity tumor, which extends through the entire stromal depth to the outer cervical margin (see Images 21 and 23). However, the finding of full-thickness stromal invasion by itself should not be considered to be evidence of tumor spread beyond the cervix, although direct microscopic invasion of the parametrium or vagina is possible in this setting.

Gadolinium-based contrast agents linked to development of nephrogenic systemic fibrosis (NSF) or nephrogenic fibrosing dermopathy (NFD)

Please note that gadolinium-based contrast agents (gadopentetate dimeglumine [Magnevist], gadobenate dimeglumine [MultiHance], gadodiamide [Omniscan], gadoversetamide [OptiMARK], gadoteridol [ProHance]) have recently been linked to the development of nephrogenic systemic fibrosis (NSF) or nephrogenic fibrosing dermopathy (NFD). For more information, see the eMedicine topic Nephrogenic Fibrosing Dermopathy. The disease has occurred in patients with moderate to end-stage renal disease after being given a gadolinium-based contrast agent to enhance MRI or MRA scans.

As of late December 2006, the FDA had received reports of 90 such cases. Worldwide, over 200 cases have been reported, according to the FDA. NSF/NFD is a debilitating and sometimes fatal disease. Characteristics include red or dark patches on the skin; burning, itching, swelling, hardening, and tightening of the skin; yellow spots on the whites of the eyes; joint stiffness with trouble moving or straightening the arms, hands, legs, or feet; pain deep in the hip bones or ribs; and muscle weakness. For more information, see the FDA Public Health Advisory or Medscape.

The extended clinical staging of cervical carcinoma with cross-sectional imaging (CT scanning and/or MRI) is based on the FIGO criteria (see Clinical staging in the CT Scan section). Staging with MRI benefits from the interpretation of abnormalities in the signal intensity in addition to alterations in the morphology of the organs or anatomic regions that are involved by tumor spread.

MRI manifestations of invasive cancer confined to the cervix (stage I) include the following: normal MRI appearance of the cervix; intact layer or ring of hypointense cervical stroma that completely surrounds the hyperintense tumor on T2-weighted images; full-thickness disruption of the hypointense cervical stroma by a high-signal-intensity tumor, with preservation of the normal morphology and signal intensity of the parametrium and vagina (see Images 18-19); smooth, well-defined, and intact peripheral cervical contours; and lack of parametrial soft-tissue strands or masses.

In stage IIa, tumor spread into the vagina may be depicted as a high-signal-intensity mass that disrupts or replaces the hypointense vaginal wall, as a segmental disruption of the low signal intensity of the vaginal wall, or as a thickened hyperintense vagina. Difficulties with interpretation are encountered in determining the presence or absence of minimal vaginal invasion. Overstaging errors are caused by relatively large exophytic tumors that expand the vaginal fornix and stretch the vaginal wall without invading it. In this setting, the stretched and thin vaginal fornix may not be recognized to be intact, although it is not disrupted by tumor. Vaginal extension of cervical carcinoma can be reliably evaluated by clinical methods, and therefore, the role of MRI is not crucial in this regard.

MRI features of parametrial tumor invasion (stage IIb) include the following (in order of increasing PPV): loss of definition, irregularity, or nodularity of the cervical contours; prominent or thick soft-tissue strands in the parametrial fat; full-thickness stromal invasion (described for stage Ib) accompanied by irregularity in the interface between the tumor and the parametrium, asymmetric tumor protrusion, or encasement of parametrial vessels (In addition, for tumors arising in the vaginal cervix, the surrounding thin vaginal fornix should be disrupted before parametrial invasion is suspected.); and full-thickness stromal invasion with direct extension of an abnormal signal intensity or a mass into the parametrium (see Images 20-23).

MRI features of pelvic sidewall tumor invasion (stage IIIb) include the following: parametrial tumor that extends beyond the lateral margins of the cardinal ligament or within 2-3 mm from the pelvic sidewall; confluent soft tissues or irregular, thick, soft-tissue strands that extend through the parametrium to the obturator internus muscle or piriformis muscle; loss of the normal low signal intensity of the pelvic sidewall muscle that is adjacent to the tumor; a confluent mass that incorporates the pelvic sidewall muscles and obliterates the associated fat planes; and encasement or distortion of the iliac vessels by the tumor.

The evaluation of lymph nodes is based on morphologic size criteria. Pelvic lymph nodes larger than 1.0 cm in their short axis are considered enlarged. The occasional detection of intranodal necrosis is most consistent with metastasis and is best depicted on contrast-enhanced images. The detection of enlarged pelvic lymph nodes on MRIs is considered equivalent to pelvic sidewall tumor extension.

MRI features of tumor extension into the bladder or rectum (stage IVa) include the following (see Images 22-23): focal obliteration of the perivesical or perirectal fat; eccentric or asymmetric wall thickening that can be uniform, nodular, or serrated (The normal low signal intensity of the bladder or rectal wall is replaced by an abnormally high signal intensity.); segmental disruption of the low signal intensity of the bladder or rectal wall that is adjacent to tumor; and intraluminal extension of a soft-tissue mass.

The MRI finding of enlarged inguinal and/or retroperitoneal lymph nodes is considered a feature of stage IVb. In some patients, treatment planning may require CT-guided FNA biopsy of lymph nodes that appear enlarged on MRI.

MRI has limitations that are related to the evaluation of the morphologic alterations in the organs or anatomic regions that may be involved by tumor spread, which are similar to the CT scan limitations listed earlier (see Stage-specific CT findings in the CT Scan section). Additional pitfalls that are pertinent to MRI staging of cervical cancer include the following:

Degree of Confidence

The overall accuracy of MRI staging is reported to be 76-92%. MRI is most valuable in the direct visualization of the primary cervical tumor and in the evaluation of the locoregional extent of the tumor. Hricak and Yu reported diagnostic test statistics based on compiled data that they obtained from several published studies.²¹ In the evaluation of parametrial invasion, MRI had an overall accuracy of 90%, a PPV of 67%, and an NPV of 95%; in the detection of lymph node metastasis, the values were an accuracy of 88%, a PPV of 66%, and an NPV of 90%.

False Positives/Negatives

The pitfalls and limitations of MRI staging described in the above sections lead to false-negative and false-positive results.

Togashi et al reported an accuracy of 95% for the MRI detection of macroscopic lesions of invasive cervical cancer (stage Ib or higher) in a series of 67 patients who were referred for imaging after a histologic diagnosis was established by means of colposcopic biopsy or endocervical curettage.³³ The sensitivity of MRI was 92% and the specificity was 100%, with 3 false-negative and no false-positive results.

In this series of 67 patients, the accuracy of MRI in the evaluation of parametrial tumor spread was 89%.³³ The sensitivity was 76% and the specificity was 94%, with 4 false-negative and 3 false-positive results. In the overall evaluation, the cancer was understaged in 10 of the patients and overstaged in 6 patients.

The false-positive rate of MRI in assessing the parametrial spread of tumor is reported in the literature in the range of 8-25%. In the overall MRI evaluation of the full extent of tumor spread, the disease is understaged in 5-15% of patients because of false-negative findings in areas of microscopic spread of tumor. Overstaging occurs in 9-14% of patients because of false-positive results.¹⁹

Findings

US has been used to evaluate the size and locoregional extent of the tumor.^{25, 57, 58, 59, 60, 61} In the early stage of cervical carcinoma, the primary lesion is difficult to depict with any imaging modality, including transvaginal US and TRUS. Eventually, with disease progression, the tumor mass may appear as a hypoechoic or isoechoic region with undefined margins, or the disease may be manifested as an enlarged cervix with heterogeneous echogenicity (see Images 6-7). In general, the intracervical cancer is not clearly distinguished from the surrounding normal cervical stroma.

In a minority of patients, the tumor may be depicted as a poorly outlined focal lesion (see Images 24-26) that is readily distinguishable from the normal cervical stroma. The lesion may have a hypoechoic, hyperechoic, or mixed echogenic appearance. Fluid can accumulate in the uterine cavity as a result of the tumor obstructing the endocervical canal.

Typical US features of a grossly infiltrated parametrium include irregular or nodular lateral margins of asymmetrically protruding tumors, parametrial vascular encasement, and replacement of the vascular parametrial connective tissue by parenchymatous tissue with echotexture characteristics that are identical to those of the primary intracervical cancer. Tumor extension into the sidewall is suggested when the parametrial tumor is within 2-3 mm of the musculature. Advanced tumor may grossly involve the muscles at the pelvic sidewall, encase the iliac vessels, or directly extend into the bladder or rectal wall.

US has limitations in evaluating the tumor extent that are similar to the CT limitations discussed earlier (see Stage-specific CT findings in the CT Scan section). Additional limitations that are pertinent to the US staging of cervical cancer include the following:

Degree of Confidence

Innocenti et al reported an overall accuracy of 83% for the locoregional staging of invasive cervical cancer with TRUS in a series of 124 patients.⁵⁷ The overall diagnostic accuracy of TRUS for the evaluation of parametrial involvement was 87%, with a sensitivity of 78% and a specificity of 89%. Additional investigative studies are needed to determine if these results are reproducible, particularly because US is operator dependent.

False Positives/Negatives

The pitfalls and limitations of US in the evaluation of tumor spread, described in the above sections, lead to false-negative and false-positive results. In the study by Innocenti et al, false-negative TRUS results were encountered in 5 of 23 cases.⁵⁷ False-positive TRUS results were encountered in 10 of 92 cases.

Findings

Positron emission tomography (PET) scanning with use of fluorodeoxyglucose (FDG) has some value relative to (1) conventional imaging methods for the detection of nodal metastatic disease and recurrent cervical cancer, (2) possibly being effective in the evaluation of cases of locally advanced cervical cancer in which CT scan findings are negative, and (3) being predictive of survival outcome ^{62, 63, 64, 65}

Degree of Confidence

Reinhardt et al found a sensitivity of 91% and specificity of 100% for FDG-PET scanning in nodal staging compared with a 73% sensitivity and 83% specificity for MRI.⁶⁵ Belhocine et al found that PET scanning was most valuable in staging extrapelvic metastases and in detecting recurrence, whereas MRI was most valuable in evaluating the locoregional status of the disease.⁶⁴ PET scan staging significantly influenced management in 18% of the authors' cases. Although a positive FDG-PET study may obviate nodal sampling in patients with positive nodes, FDG-PET scanning does not depict microscopic nodal disease and cannot replace invasive procedures if microscopic disease might be clinically relevant.

False Positives/Negatives

For pelvic malignancies, urinary FDG activity can be a confounding factor, and a Foley catheter and diuretics should be used, if possible, to minimize urinary activity.

Surgery is the treatment of choice for patients with a tumor that is confined to the uterus (stages 0 and I) or with tumor spread to only the upper two thirds of the vagina (stage IIa). Patients with tumoral spread into the parametrium (stage IIb) or beyond (stages III and IV) are primarily treated with irradiation. Therefore, the differentiation between stage IIa disease and stage IIb disease or the detection of more advanced disease is critical for treatment planning.

Dysplasia and carcinoma in situ may be managed with cold-knife conization or with other excision or ablation methods, such as laser conization, LEEP, laser vaporization, and cryotherapy. Cervical conization may eventually prove to be therapeutic in many patients (see also Medical procedures in the Preferred Examination section).