Carcinoma, cervix

Approximately 90% of cervical malignancies are squamous cell carcinomas. The remaining 10% of cervical cancers consist of adenocarcinomas and sarcomas. The majority of cervical carcinomas occur at the squamocolumnar junction. Approximately one third of cases can be entirely endocervical in location, usually in women over the age of 35. Cervical carcinoma generally arises from a pre-existing dysplastic lesion. There is thought to be an orderly progression through grades of increasing dysplasia and cellular atypia, designated cervical intraepithelial neoplasm (CIN). Carcinoma in situ, the highest grade of CIN, represents extension of malignant cells to the basement membrane of the epithelium. Invasive disease occurs when the basement membrane is breached. The CIN continuum is potentially reversible up to the point of microinvasion. The time course of the progression from early dysplasia to invasive cervical carcinoma is unknown, but thought to take up to 20 years.

Early spread of invasive cervical carcinoma is typically by direct extension and lymphatic invasion. Lateral spread into the parametrium typically occurs prior to anterior or posterior extension, due to the lack of restraining fascial planes. With disease progression, there is extension into the upper vagina and/or uterine myometrium. Direct invasion of the rectum, sigmoid colon, urinary bladder, ureters, and pelvic side wall generally occurs late in the disease. Rarely, invasive cervical carcinoma can extend posteriorly into the pelvic cul-de-sac, resulting in peritoneal dissemination. Lymphatic spread typically occurs in an orderly fashion, initially involving the parametrial, obturator, internal iliac, external iliac and lateral sacral nodes. The disease eventually progresses to involve common iliac chain and para-aortic nodes. Spread into the mediastinum or supraclavicular nodes can occur with extensive disease. Haematogenous dissemination is a relatively late event in the course of invasive cervical carcinoma. These are typically secondary to direct venous invasion or anastamoses between the lymphatics and venous system. Common sites of metastatic spread in order of frequency include the lung, skeletal system, brain and liver.

Lymph node status is the single most important prognostic factor in invasive cervical carcinoma. Multiple studies have demonstrated the impact of nodal metastases on long-term survival. At the time of diagnosis, the reported frequency of node-positive disease ranges fr radiography is still routinely performed. Lung metastases from cervical carcinoma typically appear as multiple discrete pulmonary nodules. Cavitation is not uncommon with metastases from the cervix, similar to other squamous cell carcinomas. Mediastinal adenopathy can rarely be seen in advanced cases. In addition to demonstrating unexpected pulmonary metastases, chest radiography is useful in detecting comorbid pulmonary disease. Signs of rectal or colonic involvement by double contrast barium enema include mucosal serrations, crenulation, ulceration or fistula formation. BE, however, is no longer routinely employed in the staging of cervical carcinoma. There is a very low yield of positive colonic involvement in patients with clinical stage Ib or less. In addition, sigmoidoscopy and colonoscopy have become the more popular methods of evaluating patients for suspected colonic or rectal extension of disease. The demonstration of urinary obstruction classifies a patient as having stage IIIb disease. Signs of obstructive uropathy by IVP are myriad, including a delayed or persistent nephrogram, pelvocaliectasis and ureterectasis. Extrinsic compression of the ureter by tumour can often be demonstrated. Similar to BE, the routine use of IVP in the staging of cervical carcinoma has also decreased. This is most likely due to the increased use of cross-sectional imaging. Another factor is the low pretest probability of urinary obstruction in early disease. It has been shown that the prevalence of urinary obstruction in clinical stage Ib disease is 0-2%.

As previously mentioned, the presence of nodal metastases significantly impacts on long-term survival. Traditionally, lymphangiography was utilized to assess the pelvic and para-aortic lymph nodes. Recent reports suggest that lymphangiography is not a reliable method in the evaluation of pelvic nodal status. The positive predictive value for lymphangiography ranges from 14 to 80%. The number of centres routinely performing lymphangiography is decreasing, resulting in limited availability and reduced expertise. Cross-sectional imaging demonstrates comparable positive predictive values for detecting nodal metastases. In addition, CT and MRI are able to demonstrate other morphological risk factors. For these reasons, lymphangiography is not routinely performed in the evaluation of patients with cervical cancer.

Imaging

Endovaginal and endorectal ultrasonography play a limited role in the imaging of cervical cancer. Frequently, the sonographic appearance of cervical cancer is isoechoic to normal cervical tissue. In these instances, cervical enlargement is the only sign of the presence of malignancy. Large necrotic tumours may appear relatively hypoechoic. Hydrometria or pyometria can be identified with endocervical canal obstruction. Parametrial invasion can sometimes be seen as replacement of the normally hyperechoic parametrial fat by relatively hypoechoic tumour. Additional signs include vascular encasement or an irregular cervical margin. Ultrasound can demonstrate urinary obstruction with high degree of accuracy. Gross invasion of the bladder can sometimes be seen. Due to its inherent low contrast resolution, however, ultrasound is limited in its ability to differentiate between tumour and normal cervical, uterine or parametrial tissue. In addition, uterine or adnexal pathology cannot be reliably differentiated from extension of cervical disease. False positive results can be seen with leiomyomas, endometriosis and pelvic inflammatory disease. Furthermore, accurate evaluation of pelvic lymph nodes with ultrasonography has remained elusive. For these reasons, ultrasound is not recommended for cervical cancer staging.

CT has become increasingly popular in the evaluation of clinically advanced disease (stage greater than IIb). CT has the ability to demonstrate the primary tumour, as well as other important sta reliable of these signs is obliteration of the periureteric fat plane, usually a late finding with gross parametrial invasion.

CT is useful in detecting pelvic side wall invasion. Extension of tumour to within 3 mm of the pelvic side wall, encasement of the iliac vessels, or enlargement of the obturator or piriformis muscles are all signs of stage III disease. Gross invasion of the bladder or rectum (stage IV) is usually seen as loss of fat planes and irregular wall thickening. Early invasion of these organs is not reliably elucidated from CT examinations. Ureteric involvement with urinary obstruction is reliably (stage IIIb) detected by CT, largely replacing the need for IVP. Evaluation of nodal status is a primary use of CT in cervical carcinoma. Nodal size is the only morphological criteria used to determine the presence of metastases. CT has a fairly high specificity (93%) when utilizing criteria of nodal enlargement greater than 1 cm in the short axis as abnormal. The sensitivity, however, is only 44%. The overall accuracy of detecting pelvic nodal metastases is between 70 and 80%. CT is the most popular modality for guiding percutaneous biopsy of suspicious lymph nodes.

The superior contrast resolution of MRI makes it an ideal modality for evaluation of cervical carcinoma. Cervical carcinoma is identified as a high signal intensity mass on T2-weighted images (Fig.1) (Fig.2) (Fig.3) (Fig.4). This is in stark contrast to the low signal of normal cervical stroma. The cervical mass is much less conspicuous on T1-weighted images, as the mass and normal cervix often demonstrate similar signal intensity. Intravenous gadolinium chelates cause variable tumour enhancement. Contrast enhancement can differentiate between viable and necrotic tumour. The use of contrast, however, has not shown any improvement in overall accuracy of tumour depiction. MRI is very useful in the determination of tumour size. One of the seemingly universal principles of oncology is that tumour volume generally correlates with prognosis. It is for this reason that FIGO updated its staging system to include tumour size as a prognostic variable. Excellent contrast resolution between tumour and normal cervical tissue allows accurate measurements. MR imaging has been shown to be up to 93% accurate in measuring tumour size to within 5 mm of measurements obtained from surgical specimens. Lesion measurement by MRI is useful when determination of tumour size is difficult or equivocal by physical examination. In addition to determining lesion size, MR imaging is useful in the local staging of cervical cancer. MRI can identify deep stromal invasion (stage Ib) with a 94% accuracy. MRI is less reliable in the depiction of lesions with superficial stromal invasion (stage Ia). Overall, MR is 88% accurate in determining the presence of stromal invasion and 78% accurate as to the depth of stromal invasion. MRI is useful in the evaluation of parametrial invasion (stage IIb). In addition to full depth cervical stromal invasion, MR imaging findings of parametrial invasion include an asymmetrical tumour bulge, irregular tumour  parametrial interface, and vascular encasement. MRI demonstrates parametrial invasion with an accuracy of 87-94%. An intact fibrous ring of cervical stroma (low signal intensity on T2-weighted images) has a high negative predictive value (95%) for parametrial invasion. This is an important factor in determining eligibility for surgery. MRI is also accurate in the depiction of lower vaginal involvement, ureteric obstruction, and pelvic side wall invasion. MRI findings of pelvic side wall invasion include tumour proximity to the side wall less than 3 mm, vascular encasement, and increased signal of adjacent muscle on T2-weighted images. The sagittal plane is useful in detecting rectal, urinary bladder, or lower vaginal involvement. The MRI findings in bladder invasion include direct tumour invasion or increased signal within the bladder wall on T2-weighted images. Invas defined.

HH