Carcinomas of the cervix are predominantly squamous cell carcinomas (85% to 90%), and approximately 10% to 15% are adenocarcinomas.
Squamous cell carcinomas are strongly associated with human papillomavirus (HPV) infection.
Cervical carcinoma is the third most common malignancy of the lower female genital tract, after endometrial and ovarian cancer, and the second most common cause of death, after ovarian cancer.
Definitive diagnosis of microinvasive carcinoma is established only by means of cervical conization, not biopsy. The margins of the cone should be free of neoplastic epithelium before conservative therapy is undertaken.
Microinvasive carcinoma of the cervix can be effectively treated by total hysterectomy, with a 5-year survival rate of almost 100%, but recurrent neoplasia can develop after 5 years; however, a precise and reliable definition of microinvasion is controversial.
Prognosis in squamous cell cancer of the cervix is related to tumor stage and lesion size, depth of invasion, and spread to lymph nodes.
Cervical carcinomas are locally invasive tumors that spread primarily to the pelvic tissues and then to the pelvic and paraaortic lymph nodes. Less commonly, hematogenous spread to the liver, lung, and bone occurs.
The risk of the spread of cervical carcinoma to pelvic nodes is approximately 15% for stage I, 29% for stage II, and 47% for stage III. For the paraaortic nodes, the figures are 6% for stage I, 19% for stage II, and 33% for stage III.
Surgery is often used to treat stage IB and early stage IIA carcinomas of the cervix, particularly for smaller tumors and in younger patients to preserve their ovarian function. Surgery produces less scarring and vaginal fibrosis than radiation and is preferred for women with a pelvic infection or a history of conditions such as inflammatory bowel disease, which increase the risk for radiation complications.
Minimally invasive radical hysterectomies are associated with increased rates of recurrence and decreased survival rates compared with an open abdominal surgical approach.
High-stage tumors are treated by chemoradiation. Programs usually use cisplatin, 40 mg/m 2 weekly, during external treatment and with brachytherapy.
Most cancers of the cervix are treated by radiation therapy (teletherapy and brachytherapy). Radiation doses vary with tumor size and stage but are approximately 50 to 65 Gy at point B and 80 to 85 Gy at point A. Brachytherapy should be used whenever feasible because data indicate that use of brachytherapy improves survival compared with any other type of modality.
Complications after radiation are related to dose and volume of tissue treated; these include radiation inflammation of the bladder or bowel, which may lead to pain, bleeding, or, infrequently, fistula formation. The normal cervix is resistant to radiation, and the dose can be as high as 200 to 250 Gy over 2 months. The rectum should be limited to doses of 70 Gy or less and the bladder to doses of 80 Gy or less. Overall, the rate of moderate to severe radiation complications for treatment of all stages is approximately 10%.
Worldwide 5-year survival rates reported for patients with carcinomas of the cervix are as follows: stage IA, 95%; stage IB, 80%; stage II, 70%; stage III, 50%; and stage IV, 20% with radiation therapy alone.
Pregnancy does not adversely affect the survival rate for women with carcinoma of the cervix, stage for stage.
Approximately one-third of patients treated for cervical carcinoma develop tumor recurrence, and approximately 50% of these recurrences are located in the pelvis; most occur within 2 years.
Patients whose recurrences occur more than 3 years after primary therapy have a better prognosis than those with earlier recurrence.
Pelvic exenteration in carefully selected patients with central pelvic recurrence can lead to a 5-year survival rate of 50% or better.
Chemotherapy of recurrent squamous cell carcinoma of the cervix does not produce long-term cures, but results suggest that cisplatin-paclitaxel-bevacizumab should be considered the standard treatment for patients with stage IVB recurrent or metastatic cervix cancer.
The majority of cervical malignancies are carcinomas; a summary of the more common histologic types is shown in Box 31.1 . Approximately 80% to 85% of these tumors are squamous cell carcinomas, and 15% to 20% are adenocarcinomas. The incidence of adenocarcinomas has increased in most developing countries, particularly among younger women. Carcinoma of the cervix is closely associated with early and frequent sexual contact and cervical viral infection, particularly human papillomavirus (HPV), as detailed in Chapter 29 . According to the American Cancer Society, the frequency of cervical cancer has been steadily decreasing, in part because of the effect of widespread screening for premalignant cervical changes by cervical cytologic testing (Papanicolaou [Pap] smear). In the United States there will be an estimated 13,240 new cases of invasive cervical cancer diagnosed in 2018, with 4170 related deaths ( ). Racial and ethnic disparities continue in the United States in both the incidence and mortality from cervical cancer. The incidence of cervical carcinoma in the United States is higher among the Hispanic population (9.7 per 100,000) compared with white (7 per 100,000) and African American populations (9.5 per 100,000) ( ); however, the mortality rate from cervical cancer is the highest among African Americans compared with other races and ethnicities, partly because African Americans tend to be diagnosed at a later stage. Invasive cervical cancers are diagnosed at a localized stage in 47% of white women and 37% of African American women. This chapter details the various types of cervical carcinoma and considers the natural history, methods of diagnosis and evaluation, and details of therapy. Primary sarcomas and melanomas of the cervix are extremely rare and are not considered separately.
Varieties of squamous cell carcinoma of the cervix are illustrated in Fig. 31.1 . An early form, microinvasive carcinoma, is considered separately in the next section. Most squamous cell carcinomas of the cervix are reported to be of the large cell, nonkeratinizing type, but some are keratinized, and squamous pearls may be seen. The degree of differentiation of the tumors is usually designated by three grades: G1, well differentiated; G2, intermediate; and G3, undifferentiated; however, there is no consensus on the value of tumor grade as a major prognostic factor for squamous cell carcinoma of the cervix.
A rare variety of squamous cell carcinoma is the so-called verrucous carcinoma, which is morphologically similar to that found in the vulva (see Chapter 30 ). These warty tumors appear as large bulbous masses ( Fig. 31.2 ). They rarely metastasize but unfortunately may be admixed with the more virulent, typical squamous cell carcinomas, in which case metastatic spread is more likely.
Adenocarcinomas may have a number of histologic varieties. The typical variant often contains intracytoplasmic mucin and is related to the mucinous cells of the endocervix (endocervical pattern; Fig. 31.3 ); however, on occasion the cells contain little or no mucin, and then the tumor may resemble an endometrial carcinoma (endometrioid pattern). It may be difficult histologically to ascertain whether these carcinomas arise in the cervix or endometrium. Although not independently diagnostic, the immunohistochemical panel that is recommended to assist in differentiating endocervical from endometrial primary malignancies includes estrogen receptor (ER), vimentin, monoclonal carcinoembryonic antigen (CEA), and p16. Typically an endocervical carcinoma stains diffusely positive for p16 and CEA and is negative for ER and vimentin.
A rare but important virulent variety of adenocarcinoma is adenoma malignum. These microscopically innocuous-appearing tumors consist of well-differentiated mucinous glands ( Fig. 31.4 ) that vary in size and shape and infiltrate the stroma. Despite their bland histologic appearance, they tend to be deeply invasive and metastasize early. The term minimal deviation adenocarcinoma is applied to these tumors.
Clear cell adenocarcinomas of the cervix are histologically identical to those of the ovary (see Chapter 33 ) and vagina (see Chapter 30 ). They are uncommon in the cervix and can be associated with intrauterine diethylstilbestrol exposure, although they also may develop spontaneously in the absence of diethylstilbestrol exposure.
Adenoid cystic carcinomas are rare. Berchuk and Mullin summarized 88 cases reported in the literature ( ). These tumors are aggressive and may resemble cylindromas of salivary gland or breast origin and histologically may resemble basal cell carcinomas of the skin (adenoid basal, or basaloid, carcinomas). Most patients with these tumors are older than 60 years. The basaloid variety appears to be less aggressive.
Adenosquamous carcinomas, as the name implies, consist of squamous carcinoma and adenocarcinoma elements in varying proportions ( Fig. 31.5 ). They occur often in pregnant women. A particularly virulent variety is termed glassy cell carcinoma ( Fig. 31.6 ). This is an undifferentiated tumor consisting of large cells containing cytoplasm, with a ground-glass appearance. Glassy cell carcinomas tend to metastasize early to lymph nodes and to distant sites and usually have a fatal outcome.
Small cell carcinoma of the cervix is rare, comprising less than 5% of all carcinomas of the cervix. Women with small cell carcinoma are likely to be 10 years younger than those with squamous cell carcinoma. Cervical small cell carcinoma is composed of small anaplastic cells with scant cytoplasm that behave aggressively and are often associated with widespread metastasis to multiple sites, including bone, liver, skin, and brain. Efforts to treat these cancers with approaches typically used for small cell carcinoma of the lung have had mixed results.
Another variant that is not in the World Health Organization (WHO) classification is non–small cell neuroendocrine tumors. These tumors contain intermediate to large cells, high-grade nuclei, and eosinophilic cytoplasmic granules of the type seen in neuroendocrine cells. Reported survival rates for patients with these aggressive carcinomas are similar to those of patients with small cell tumors, and an optimal therapy has yet to be established.
Carcinoma of the cervix
Patients with carcinoma of the cervix characteristically present with abnormal bleeding or brownish discharge, often noted after douching or intercourse and also occurring spontaneously between menstrual periods. These patients often have a history of not having had a cytologic (Pap) smear for many years. Other symptoms, such as back pain, loss of appetite, and weight loss, are late manifestations and occur when there is extensive spread of cervical carcinoma. The patients tend to be in their 40s to 60s, with a median age of 52 years. Preinvasive intraepithelial carcinoma of the cervix (see Chapter 29 ) occurs primarily in women in their 20s and 30s and has become more common in those in their 20s, leading to a gradual increase in the incidence of invasive carcinoma in younger patients.
The diagnosis is established by biopsy of the tumor; a specimen can easily be obtained during an office examination. A Kevorkian, Eppendorf, Tischler, or similar punch biopsy instrument is convenient to use. Occasionally it is necessary to biopsy nodularity or indurations in the vagina near the cervix to ascertain the limit of tumor spread and define a correct tumor stage. If the woman’s cytologic smear suggests invasive carcinoma, with no gross lesion visible, and endocervical curettage does not demonstrate carcinoma, or if an adequate biopsy specimen to establish carcinoma cannot be obtained, cervical conization should be performed.
Historically the staging of carcinoma of the cervix depended primarily on the pelvic examination, with no changes to staging based on operative findings. In 2018, the staging system underwent revisions that allow imaging and pathologic findings from surgery, where available, to assign the stage. Table 31.1 describes the four stages of cervical carcinoma according to the International Federation of Gynecology and Obstetrics (FIGO; revised in 2018) ( ). The types of tumor distributions that may be observed in the various stages are illustrated in Fig. 31.7 .
|I||Carcinoma is strictly confined to the cervix (extension to the corpus should be disregarded)|
|IA||Invasive cancer that can be diagnosed only by microscopy, with deepest invasion <5 mm *|
|IA1||Measured stromal invasion <3 mm in depth|
|IA2||Measured stromal invasion of ≥3 mm and <5 mm in depth|
|IB||Invasive carcinoma with measured deepest invasion ≥5 mm (greater than stage IA), lesion limited to the cervix uteri †|
|IB1||Invasive carcinoma >5 mm depth of stromal invasion, and ≤2 cm in greatest dimension|
|IB2||Invasive carcinoma >2 cm and ≤4 cm in greatest dimension|
|IB3||Invasive carcinoma lesion >4 cm in greatest dimension|
|II||The carcinoma invades beyond the uterus, but has not extended onto the lower third of the vagina or to the pelvic wall|
|IIA||Involvement limited to the upper two-thirds of the vagina without parametrial involvement|
|IIA1||Invasive carcinoma <4 cm in greatest dimension|
|IIA2||Invasive carcinoma mensionn greatest dimension|
|IIB||With parametrial involvement but not up to the pelvic wall|
|III||The carcinoma involves the lower third of the vagina and/or extends to the pelvic wall and/or causes hydronephrosis or nonfunctioning kidney and/or involves pelvic and/or paraaortic lymph nodes ‡|
|IIIA||The carcinoma involves the lower third of the vagina, with no extension to the pelvic wall|
|IIIB||Extension to the pelvic wall and/or hydronephrosis or nonfunctioning kidney (unless known to be due to another cause)|
|IIIC||Involvement of pelvic and/or paraaortic lymph nodes, irrespective of tumor size and extent (with r and p notations) ‡|
|IIIC1||Pelvic lymph node metastasis only|
|IIIC2||Paraaortic lymph node metastasis|
|IV||The carcinoma has extended beyond the true pelvis or has involved (biopsy-proven) mucosa of the bladder or rectum; a bullous edema, as such, does not permit a case to be allotted to stage IV|
|IVA||Spread to adjacent pelvic organs|
|IVB||Spread to distant organs|
‡ Adding notation of r (imaging) and p (pathology) to indicate the findings that are used to allocate the case to stage IIIC. Example: If imaging indicates pelvic lymph node metastasis, the stage allocation would be stage IIIC1r, and if confirmed by pathologic findings, it would be stage IIIC1p. The type of imaging modality or pathology technique used should always be documented.
Natural history and spread
Carcinoma of the cervix is initially a locally infiltrating cancer that spreads from the cervix to the vagina and paracervical and parametrial areas. Grossly the tumors may be ulcerated ( Fig. 31.8 ), similar to carcinomas occurring elsewhere in the female genital tract, and may have an exophytic growth pattern or cauliflower-like appearance extruding from the cervix. Alternatively, they may be endophytic, in which case they are asymptomatic, particularly in the early stage of development, and tend to be deeply invasive when diagnosed. These usually start initially from an endocervical location and often fill the cervix and lower uterine segment, resulting in a barrel-shaped cervix. The latter tumors tend to metastasize to regional pelvic nodes and, because of the tendency of late diagnosis, are often more advanced than the exophytic variety. The primary path for distant spread is through lymphatics to the regional pelvic nodes. Bloodborne metastases from cervical carcinomas do occur but are less common and are usually seen late in the course of the disease.
Initially, cervical carcinoma spreads to the primary pelvic nodes, which include the pericervical node; presacral, hypogastric (internal iliac), and external iliac nodes; and nodes in the obturator fossa near the vessels and nerve. From this primary group, tumor spread proceeds secondarily to the common iliac and paraaortic nodes. Rarely the inguinal nodes are involved; however, if the lower third of the vagina is involved, the median inguinal nodes should be considered primary. The distribution of lymph node involvement in 26 cases of untreated carcinoma of the cervix was studied in detail by Henriksen ( Fig. 31.9 ) ( ). A series studying the incidence and distribution pattern of retroperitoneal lymph node metastases in 208 patients with stages IB, IIA, and IIB cervical carcinomas who underwent radical hysterectomy and systemic pelvic node dissection reported that 53 patients (25%) had node metastasis ( ). The obturator lymph nodes were the most commonly involved, with a rate of 19% (39 of 208), and the authors proposed them as sentinel nodes for cervical cancers. An important distal node that becomes involved after the paraaortic group is the left scalene node—that is, the left supraclavicular node. A biopsy of this node may be performed in the assessment of advanced cervical carcinoma to clarify whether the tumor has spread outside the abdomen. In addition to nodal spread, hematogenous spread of cervical carcinoma occurs primarily to the lung, liver, and, less often, bone (see Recurrence later in the chapter).
FIGO stage is the most important determinant of prognosis for carcinoma of the cervix ( Table 31.2 ); however, there are other factors, including tumor and patient characteristics, that are prognostic. One of the most important predictors is tumor size for local recurrence and death for patients treated with surgery or radiation therapy ( ). The 2018 FIGO staging classification for stage IB disease was further modified based on tumor diameter (i.e., IB1, <2 cm; IB2, ≥2 cm and <4 cm; and IB3, ≥4 cm). Another important prognostic factor is involvement of lymph nodes, which is now included in the staging system as stage IIIC disease. In several surgical series, after a radical hysterectomy, patients with positive pelvic lymph nodes had a 35% to 40% lower 5-year survival rate than patients with negative nodes. Patients with positive paraaortic nodes have a survival rate that is approximately 50% that of patients with similar stage disease and negative paraaortic nodes. With extended-field radiation therapy, patients with positive paraaortic nodes have approximately a 40% to 50% 5-year survival rate. There is a strong correlation between positive nodes and positive lymph–vascular space invasion (LVSI) in the tumor specimen in patients with cervical carcinoma; however, LVSI may be an independent predictor of prognosis, as shown in a number of larger surgical series.
|Stage||No. of Patients (n)||5-Year Survival|
In patients who have had a radical hysterectomy, histologic evidence of extracervical spread (≥10 mm), deep stromal invasion (>70% invasion), and LVSI are associated with a poorer prognosis. A randomized trial from the Gynecologic Oncology Group (GOG) compared observation versus adjuvant radiation therapy in patients after radical hysterectomy with a combination of two of the factors mentioned earlier; patients who received radiation therapy had better local control and improved overall survival ( ). Involvement of the parametrium in the hysterectomy specimen has been correlated with higher rates of lymph node involvement, local recurrence, and death from cancer. Uterine body involvement is associated with an increased rate of distant metastases in patients treated with radiation or surgery.
Patients with adenocarcinomas of the cervix have a poorer prognosis than patients with squamous cell carcinomas of the cervix. Investigators have found that among patients treated surgically, patients with adenocarcinomas have high relapse rates compared with rates in patients with squamous cell carcinomas. In an analysis of 1767 patients treated with radiation for FIGO stage IB disease, Eifel and associates found that independent of age, tumor size, and tumor morphology, patients with adenocarcinomas had the same pelvic control rate but twice as high a rate of distant metastasis as patients with squamous cell carcinomas of the cervix ( ). Although the prognostic significance of histologic grade for squamous carcinomas has been disputed, there is a clear correlation between the degree of differentiation and the clinical behavior of adenocarcinomas.
There has been a great interest in molecular markers for prognosis and treatment in carcinoma of the cervix. One of the most studied markers is the serum squamous cell carcinoma antigen. Studies have shown that pretreatment levels of this antigen correlate well with stage of disease, tumor histologic characteristics, grade, type of tumor (exophytic vs. infiltrative), microscopic depth of invasion, and risk of lymph node metastases in patients with early-stage disease. Possible clinical applications of this antigen may be to predict clinical outcome and as a marker for monitoring the course of disease and response to treatment in patients with cervical cancer. Several investigators have reported significantly lower survival rates in patients with elevated values compared with patients with normal baseline levels, independent of stage. For detection of tumor recurrence, serial squamous cell carcinoma antigen testing has proved to be more specific than sensitive, with specificities ranging from 90% to 100% and sensitivities ranging from 60% to 90%. Further investigation is needed in these areas. Some investigators have found a higher rate of recurrence in patients with HPV-positive nodes (although negative for malignancies) and poor prognosis with the presence of HPV messenger RNA (mRNA) in the peripheral blood of cervical cancer patients. Other markers that have been investigated include epidermal growth factor receptor, cyclooxygenase-2, DNA-ploidy, tumor vascularity, and S-phase fraction. Programmed death ligand 1 (PDL1) expression may be useful for treatment decisions in the recurrent setting ( ).
Once a woman has been diagnosed as having an invasive carcinoma, a pretreatment evaluation is conducted to determine the extent of disease, arrive at an accurate clinical staging, and plan the program of therapy. The usual evaluation consists of a thorough history and physical examination, routine blood studies, intravenous pyelogram (IVP) or computed tomography (CT), and chest radiography. Demonstration of an obstructed ureter or nonfunctioning kidney caused by tumor automatically assigns the case at least to stage III (see Table 31.1 ). A barium enema test or flexible sigmoidoscopy, as well as a cystoscopy, is sometimes performed in the case of large tumors or for patients who will be receiving radiation treatment.
The best radiographic imaging technique for detecting lymph node metastases is unclear. CT and magnetic resonance imaging (MRI) are good for identifying enlarged nodes; however, the accuracy of these techniques in the detection of positive nodes is compromised by their failure to detect small metastases, and many enlarged nodes are caused not by metastases but by inflammation associated with advanced disease. The accuracy of MRI in the detection of lymph node metastases (72% to 93%) is similar to that of CT but better than CT and physical examination for the evaluation of tumor location, tumor size, depth of stromal invasion, vaginal extension, and parametrial extension of cervical cancer; however, with regard to detecting lymph node metastases or other distant disease, positron emission tomography (PET) shows promise. Several studies from a single institution have shown that 18 F-fluorodeoxyglucose PET (FDG-PET) detects abnormal lymph nodes more often than CT, and those findings with PET are a better predictor of survival than those with CT or MRI in patients with carcinoma of the cervix. Medicare has approved PET/CT as part of the initial staging evaluation for patients with cervical carcinoma, and most insurance companies approve PET/CT for a 3-month follow-up.
Surgical sampling of lymph nodes is the most sensitive method of evaluating whether regional lymph nodes contain metastases; however, it is invasive, expensive, and may delay treatment of the primary lesion. Laparoscopic lymph node dissection may decrease the time between surgery and the start of treatment and may be associated with less late radiation-related morbidity than open transperitoneal staging. Laparoscopic extraperitoneal paraaortic lymphadenectomy has also been described and may further decrease radiation-related bowel morbidity by avoiding entrance into the peritoneal cavity. In a study by Ramirez and colleagues, 22% of patients who had positive pelvic but negative paraaortic nodes on PET/CT had histopathologically positive paraaortic nodes ( ).
Treatment for stage I
The term microinvasion has been used for years to describe patients with minimally invasive cervical cancer, but this term is not part of the FIGO staging system. Microinvasion was used to describe patients with 3-mm invasion or less and essentially no risk of metastatic spread. There is ample evidence that patients with small-volume tumors measured only by depth of invasion have a low risk of relapse and death with radical surgery or more conservative surgical approaches.
The diagnosis of microinvasive tumor cannot be made based on a biopsy specimen alone; a cervical conization must be performed. If the margin of the cervical cone specimen contains neoplastic epithelium, the risk of invasive tumor in the remaining uterus is increased. Decisions on treatment should be based on an adequate cone biopsy specimen. If a woman has positive margins, the cone can be repeated. Sometimes, deeper invasion will be uncovered and more radical treatment will be required. In some patients, conization alone is adequate.
The measured stromal invasion of stage IA1 tumors is 3 mm or less. These measurements are determined on a cone biopsy, which also determines other prognostic factors (e.g., lymph vascular space involvement, histologic subtype, grade). These factors do not alter the stage assignment, in spite of their adverse prognostic significance. In the absence of LVSI or high-risk histologic subtypes, the risk of lymph node metastases is remote and nonradical surgery is adequate. This may include cone biopsy, simple trachelectomy, or simple hysterectomy, depending on the circumstances and patient preference.
Patients with stage IA2 tumors have a measured stromal invasion of 5 mm or less. Patients in this category, even without LVSI, are at a low risk of nodal involvement. Thus radical or modified radical approaches are usually recommended, which include modified radical hysterectomy or trachelectomy and pelvic lymphadenectomy. Stage IA1 patients who have LVSI are treated in the same manner as stage IA2 patients.
There continues to be interest in determining the necessity of treating the parametrium in patients with low-stage cervical cancer. Several investigators have noted that the risk of parametrial involvement is 1% or less in patients undergoing radical hysterectomy in low-risk situations. Parametrial resection contributes significant short-term and long-term morbidity to surgery. Lymphatic mapping techniques suggest that parametrial sentinel lymph nodes can be identified and resected without a radical dissection. In the future, as cervical cancer surgery trials mature, the indications for surgery that omits the parametrial resection may grow.
Stage IB encompasses tumors that are larger than stage IA, meaning 5 mm or larger in measured stromal invasion.
Patients with stage IB1 and IB2 disease have tumor limited to the cervix that is 4 cm or less in diameter. These patients have equally good outcomes with radical surgery or radiotherapy. A major factor in favor of radical surgery is younger age, especially premenopausal patients for whom ovarian preservation is an option. In addition, there is now increasing availability of ovum transfer and pregnancy surrogates. Other factors favoring radical surgery are smaller size, desire to preserve fertility, and absence of other comorbidities that escalate the risk of surgery. Factors that favor radiotherapy are the presence of indicators that would result in postoperative radiotherapy if surgery were the primary therapy. These include larger size, extensive LVSI, suspicious findings on preoperative imaging, high-risk histologic subtypes, and deep stromal invasion on imaging or examination that increases the risk of close margins.
The decision regarding radical surgery or radiotherapy should be made with the active involvement of the woman, the gynecologic oncologist, and the radiation oncologist. Both modalities are associated with the potential for significant short- and long-term complications. Some complications, such as bladder atony, food intolerance, and loss of sexual function, are not easy to measure and can persist for years after treatment. Long-term survival data are similar for well-selected patient populations. Some data suggest more long-term patient satisfaction with outcomes related to surgery than radiotherapy.
Most gynecologic oncologists and radiation oncologists recommend concurrent chemoradiation for patients with stage IB3 cervical cancer. Reports demonstrate that up to 80% of patients with tumor size greater than 4 cm who have undergone a radical hysterectomy have clear indications for postoperative radiotherapy. Our experience is that postoperative radiotherapy results in greater toxicity than treatment with the cervix intact. The counterargument is that radical hysterectomy obviates the need for high-dose brachytherapy, which is associated with the most severe and difficult to manage complications, notably pelvic fistulas. A number of innovations in treatment targeting techniques appear to be having a beneficial impact on reduction of posttreatment fistulas. Concurrent chemoradiation is our primary recommendation for stage IB3 patients and is discussed later in this chapter.
Operative therapy: Radical hysterectomy and pelvic node dissection
Radical hysterectomy and bilateral pelvic lymphadenectomy are effective for the treatment of stage IB1 to IB2 and some early stage IIA1 cancers. It is important that the surgery removes the same volume of tissue that has received tumoricidal doses of radiation in patients for whom radiation is the sole therapy. The amount of tissue removed, particularly in the paracervical and parametrial areas near the ureter, depends on the extent and location of the tumor. Piver and colleagues defined five classes to describe the extent of the operation ( ). Class I guarantees the removal of the entire cervix and uterus. The ureter is not disturbed from its bed. In many cases this is described as an extrafascial hysterectomy, the type used after preoperative radiation for treatment of a barrel-shaped cervix (discussed later). A class II operation ( Fig. 31.10 ) removes more paracervical tissue than class I; the ureters are retracted laterally but are not dissected from their attachments distal to the uterine artery, and the uterosacral ligaments are ligated approximately halfway between the uterus and rectum. The operation, which is usually performed with pelvic lymphadenectomy and is often termed a modified radical hysterectomy, is useful to treat small microscopic carcinomas of the cervix. Magrina and colleagues used modified radical hysterectomy primarily for tumors smaller than 2 cm (median, 1.1 cm), with a 5-year survival of 96% ( ). This procedure may occasionally be used to treat small, central cervical recurrences of carcinoma that are diagnosed after radiation therapy of the primary tumor. For a class III operation, the uterine artery is ligated at its origin from the anterior division of the hypogastric artery and the uterosacral ligaments are ligated deep in the pelvis near the rectum (see Fig. 31.10 ). This operation is usually termed a radical hysterectomy (Meigs-Wertheim hysterectomy) and is performed for stage IB2 and, rarely, for stage IIA carcinomas of the cervix.
Class IV and V operations are infrequently performed. A class IV procedure involves a complete dissection of the ureter from its bed and sacrifice of the superior vesical artery. A class V operation involves resection of the distal ureter, bladder, or both, with reimplantation of the ureter into the bladder (ureteroneocystostomy). Both are designed to remove small, central recurrent disease and would be attempted to avoid an anterior exenteration (see later). Extensive data are not available, but the latter two procedures appear to have high complication rates.
Other classification systems for radical hysterectomy exist. Querleu and Morrow described a classification that is principally based on the lateral extent of resection (four types, A through D) with subtypes that address nerve preservation and paracervical lymphadenectomy ( ). Within this classification, four levels of lymph node dissection (1 to 4) are also defined according to arterial anatomy and overall radicality of the procedure.
Preoperative preparation for a woman who is to undergo a radical hysterectomy includes the same basic considerations for anyone undergoing a major operative procedure. Graduated compression, below-the-knee leg stockings, and perioperative prophylactic doses of heparin or other low-molecular-weight heparin formulations are used to reduce the risk of thromboembolism. Prophylactic antibiotics prior to incision are also recommended. During the course of the operation, care is taken not to grasp the ureters with instruments such as forceps to avoid damaging the periureteral capillary blood supply.
An important complication of pelvic lymphadenectomy is lymphocyst formation. Most gynecologic oncologists have abandoned the use of closed suction drains in patients undergoing radical hysterectomy, instead leaving the pelvic peritoneum open to allow lymph fluid to drain internally in the peritoneal cavity. Sentinel lymph node mapping and biopsies increasingly are replacing full lymphadenectomy.
Ovarian function may be preserved in younger patients if there is little likelihood of postoperative radiation. If intraoperative findings suggest that radiotherapy will be given postoperatively, the ovaries may be transposed superior and lateral to preserve their function. This technique has some liabilities, including early loss of ovarian function and abdominal pain from ovarian cysts.
In clinical patients with stage I disease treated by radical hysterectomy and node dissection, the results obtained are related primarily to the status of the pelvic nodes and the surgical resection margins around the primary tumor (ideally, >1 cm). If the pelvic nodes are free of tumor, the 5-year survival rate can be expected to exceed 90%, whereas if the nodes are found to contain tumor (surgical stage IIIC), the 5-year survival rate drops to 45% to 50%. If the woman is found to have extensive spread of gross disease to the pelvic nodes, the studies of Potter and coworkers ( ) have suggested that it is preferable to cease the operation and complete radiation therapy to improve pelvic control of tumor; however, Hacker and associates reported an estimated 5-year survival of 80% for 34 patients whose tumor-positive pelvic or paraaortic nodes were resected and the areas subsequently radiated ( ). In a GOG study, Sedlis and coworkers evaluated disease-free survival for patients treated with radical hysterectomy who had negative lymph nodes and surgical margins but with intermediate risk factors, including more than one-third stromal invasion, capillary lymphatic space involvement, adenocarcinoma, and large tumor diameter by randomly allocating patients to pelvic radiotherapy or observation ( ). Survival was improved in those who received postoperative pelvic radiation; however, there were radiation complications, including bowel obstruction and death.
Nerve-sparing radical hysterectomy is an innovation described by Hockel and others ( ). Bladder atony is a difficult-to-study outcome of radical hysterectomy. The incidence of complete bladder atony requiring self-catheterization or nerve stimulators is low, but milder forms are common. The severity of bladder atony is directly related to the trauma inflicted on the hypogastric nerves, which may be traumatized during radical hysterectomy. The impact of the nerve-sparing approach on sexual function is not known.
Dargent developed a combined laparoscopic and vaginal technique for removal of the pelvic lymph nodes, cervix, parametrium, and upper vagina ( ). Dargent trained gynecologic oncologists from around the world to perform radical vaginal trachelectomy and laparoscopic pelvic lymphadenectomy ( ). Long-term outcomes reported by Plante, Diaz, and others have confirmed that in well-selected patients, oncologic outcomes are identical to radical hysterectomy outcomes ( ; ). First-trimester pregnancy loss rates are approximately the same for radical trachelectomy patients as for the general population. Second-trimester pregnancy loss is approximately doubled in trachelectomy patients compared with the general population, presumably because of the loss of cervical stroma. Typically a permanent cerclage is placed at the time of radical trachelectomy. Approximately two-thirds of patients have a successful pregnancy after radical trachelectomy.
When fertility-sparing surgery was first described, the assumption was that it would be offered to only a small proportion of patients. From a cohort of more than 400 patients who underwent radical hysterectomy, Sonoda and colleagues determined that approximately 50% of those younger than 40 years had low-risk histologic types and tumor size smaller than 2 cm, making them candidates for radical trachelectomy ( ).
In spite of the contribution of radical vaginal surgery to fertility preservation, the technique has been difficult for gynecologic oncologists in the United States to master. Vaginal surgical skills are diminishing and there are no other indications for radical vaginal surgery. American gynecologic oncologists, unlike their counterparts in Canada and Europe, appear to have been discouraged by the long learning curve and have not invested the time to master the approach. Gynecologic oncologists in the United States have described abdominal radical trachelectomy as an alternative to the vaginal approach. Although smaller numbers have been published, it is anticipated that oncologic and fertility outcomes will be similar to the laparoscopic-vaginal approach.
Patient selection is important when considering fertility-sparing surgery. Preoperative pelvic MRI is recommended for all patients with a visible lesion. Patients should have a desire to preserve fertility, no evidence of metastatic disease to lymph nodes or distant metastases, age younger than 45, and stage IAI with LVSI, IA2, or IB1 disease (lesion size ≤2 cm) with limited endocervical extension assessed by colposcopy and MRI.
Minimally invasive surgery
Minimally invasive techniques for treatment of cervical cancer are attractive for several reasons. Minimally invasive surgery (MIS) is associated with shorter length of stay, less pain, few postoperative infections, fewer thromboembolic complications, and reduced blood loss compared with abdominal procedures.
Minimally invasive radical surgery has become more popular over time, with minimally invasive radial hysterectomies increasing from 2% in 2006 to 33% in 2010 ( ). For gynecologic oncologists in practice, the long learning curve associated with laparoscopy has been an impediment to advancement. The most recent minimally invasive technique, robotic laparoscopic surgery, offers new advantages and increased the use of MIS for radical hysterectomies; however, data have demonstrated that patients who had a radical hysterectomy by a minimally invasive approach (laparoscopic or radical) have a higher risk of recurrence and death compared with those who had an open abdominal radical hysterectomy ( ; ). An international randomized trial reported that disease-free survival at 4.5 years was 86% in the MIS arm compared with 96.5% with open surgery. The 3-year overall survival rates were also lower in the minimally invasive group (93.8% vs. 99.0%) ( ).
Sentinel node biopsy
Cervical cancer, like most solid tumors, spreads primarily lymphatically. Surgical management of solid tumors, as pioneered more than 100 years ago by Halsted, is based on the resection of all regional lymph nodes and lymphatic channels connecting the lymph nodes to the primary tumor. Implicit in this approach is that all regional lymph nodes have the same risk of containing metastatic disease. Morton, working in patients with cutaneous melanoma, demonstrated that there are sentinel lymph nodes that are the first nodes to receive lymphatic drainage from the primary tumor and are therefore the first site of metastasis ( ). Experience with thousands of patients with melanoma and breast cancer has validated this concept, which has been successfully extended to other disease sites, notably vulvar cancer.
Cervical cancer is an excellent target for the sentinel lymph node concept because the tumor is easy to inject and the regional lymph nodes can be reached through an incision. Lymphatic drainage of the cervix is complex; however, most sentinel lymph nodes of the cervix are found along the external iliac artery or vein, obturator space, or parametrium. A number of investigators have reported their experience with sentinel lymph node biopsy in patients undergoing radical hysterectomy. In one series of 188 patients, the sensitivity of sentinel lymph node mapping in cervical cancer was calculated to be 96.4%, with a negative predictive value of 99.3% and a 3.6% false-negative rate ( ). Various protocols have been used to perform sentinel lymph node mapping in cervical cancer. Common mapping strategies use technetium-99m, blue dyes (methylene blue, isosulfan blue, patent blue), and, more recently, indocyanine green (ICG).
After radical hysterectomy, many patients experience long-term complications. Montz and associates noted a 5% frequency of small bowel obstruction, which increases to 20% if radiation is used postoperatively ( ). Fistulas from the urinary tract, particularly ureterovaginal fistulas, have been reported to occur in approximately 1% of cases. The low rate appears to result from the administration of antibiotics, prevention of retroperitoneal serosanguineous collections, and avoidance of direct manipulation of the ureter to avoid injury to the periureteral blood supply. Most gynecologic oncologists do not reperitonealize the pelvis, which allows direct drainage of lymphatic fluid to the peritoneal cavity, where it is reabsorbed.
Many women suffer postoperative bladder dysfunction. In part, this appears to be caused by disruption of the sympathetic nerve supply to the bladder; however, the dysfunction may be temporary. Low and associates noted an increase in bladder pressure with a decrease in urethral pressure after radical hysterectomy ( ). There was reduced bladder compliance with detrusor instability. The bladder can develop hypotonicity, and overdistention can then become a problem. If overdistention of the bladder and infection are avoided, progressive improvement of bladder function usually occurs. Forney correlated the degree of bladder dysfunction after radical hysterectomy with the extent of resection of the cardinal ligament ( ). Those who had a complete resection of cardinal ligaments could void satisfactorily at an average of 51 days compared with 20 days for those with only partial resection of the ligaments. All patients experienced a decrease in bladder sensation. In a few patients the decrease in bladder sensation can be permanent. For patients in whom it is temporary, recovery usually occurs after continuous drainage of the bladder with an indwelling catheter. Westby and Asmussen observed that by 1 year after surgery a slight decrease in urethral pressure persists but that the decrease is not as great as that noted immediately after the operation ( ). After 1 year, the postoperative changes and bladder function usually recover. Newer nerve-sparing surgical techniques where the uterosacral ligament is transected after separation of the hypogastric nerve and preservation of the bladder branches of the pelvic plexus have been associated with improved bladder function without compromising oncologic outcomes and survival.
In a 1999 study from Sweden, Bergmark and coworkers noted compromised sexual activity, decreased lubrication, and shortened vaginas in women treated for cervical cancer by surgery or radiation ( ). During the consent process, patients should be informed regarding the potential impact of radical hysterectomy on their sexual function.
Lymphedema is another complication of radical pelvic surgery that can affect quality of life. The areas most affected are the mons, lower abdomen, and upper thighs. Lymphedema massage may help reduce this problem, but treatment options are limited and of only modest effectiveness.
Outcomes after surgical treatment
Reported 5-year survival rates for women with stage IB cervical cancer treated with radical hysterectomy and pelvic lymphadenectomy are approximately 80% to 90%. Patients with positive or close margins or positive lymph nodes have the highest risk of recurrence and poor outcome. In large prospective studies, 3-year disease-specific survival rates of 85.6% in patients with negative nodes and 50% to 74% in patients with positive nodes were reported. A randomized study found that postoperative chemoradiation improves survival in patients with positive lymph nodes and positive surgical margins ( ).
Most patients with carcinoma of the cervix are treated by radiation. The principles of external megavoltage treatment (teletherapy) and local implants (brachytherapy) are reviewed in Chapter 28 . External beam radiation is administered in fractions, usually 180 cGy/day, 5 days a week, to destroy the tumor without causing permanent damage to normal tissues. This delivers uniform doses to the entire pelvis, including the regional pelvic nodes. The local implant delivers its highest energy locally to the cervix, surface of the vagina, and paravaginal and paracervical tissues. The radiation from the implant diminishes according to the inverse square law. The uterus and cervix serve as receptacles for arranging and holding the intracavitary applicator stem (tandem) and accompanying vaginal applicators (ovoids or ring) in a fixed and optimal position for delivering the desired radiation dosimetry. Usually the tandem and ovoids or a tandem and ring are inserted and a pack is placed into the vagina to stabilize the apparatus and increase the distance from the mucosa of the bladder and rectum. After the position of the applicator has been confirmed to be satisfactory by imaging, the radioactive source, such as cesium-137 or iridium-192, is inserted (afterloading technique). Other types of applicators are available, but the principle of delivering intense radiation to the cervix and paracervical areas is the same. The goal is to increase the total dose of radiation to the maximum allowable to achieve tumor control without introducing a major risk of complications and injury to adjacent normal tissue. The specific protocols followed in various treatment centers differ; individualization for specific patients is often needed depending on the stage and size of the cervical tumor as well as the patient’s local anatomy. In general, external therapy is given first to treat the regional pelvic nodes and shrink the central tumor mass, which then is more amenable for a local implant. In some patients, external therapy can lead to excessive shrinkage of the vaginal apex, making safe, effective implantation of local radiation sources difficult. This can be a problem, particularly in older or postmenopausal patients. Occasionally, in those patients, the implantation is done first, especially for smaller stage I tumors. Intraoperative ultrasounds may be helpful especially in difficult cases for optimal implant positioning. In some cases the central pelvis is shielded during external radiation therapy to allow for subsequent higher doses from the implant. Occasionally, interstitial therapy in the form of needles implanted into the area of the tumor is needed to achieve effective local tumor control. Although criteria differ, patients with stage III disease or poor vaginal anatomy are most often considered candidates for interstitial brachytherapy.
Intracavitary radiation therapy may be delivered at either a low-dose rate or a high-dose rate. The advantage of high-dose-rate brachytherapy is that it is given on an outpatient basis and can be done with 3 to 4 hours. High-dose-rate brachytherapy and low-dose-rate brachytherapy have similar survival and toxicity and high-dose-rate brachytherapy has become the most common type of brachytherapy available throughout the world. The number of fractions varies from two to five; the most common one in the United States is 5.5 Gy to 6 Gy to point A in five fractions. Brachytherapy in combination with external beam should be used in the treatment of locally advance cervical cancer whenever feasible because data reveal that there is an increase in survival with the use of brachytherapy compared with any other modality ( ; ).
In calculating the doses of radiation, two reference points, A and B, are used ( Fig. 31.11 ). Point A is 2 cm above the external os and 2 cm lateral to the cervical canal. Point B is 5 cm lateral to the cervical canal and 3 cm lateral to point A, which places point B in the vicinity of the lateral pelvic wall (see Fig. 31.11 ). The total dose administered depends on tumor stage but, in general, at the pelvic wall, it is in the range of 50 to 65 Gy, with the higher doses used for high-stage disease. At point A, it varies, but approximately 80 Gy is given for small IB1 lesions and doses higher than 85 Gy for larger lesions. The normal cervix is particularly resistant to radiation and can tolerate doses as high as 200 to 250 Gy over 2 months, whereas the adjacent bladder and, in particular, the rectum are much more sensitive, and their exposure in general should be limited at the point of maximal radiation to 80 Gy to the bladder and 70 Gy to the rectum, with overall average doses in the range of 65 to 70 Gy. The small bowel can be damaged at doses greater than 45 to 50 Gy, especially if adhesions limit intestinal mobility and a large volume is treated.