Innovations in the treatment of invasive cervical cancer

Abstract

Invasive cervical cancer is characterized by basement membrane–invading lesions capable of metastasizing through the lymphatic and vascular systems. Treatment methods were reviewed by panelists at the Second International Conference on Cervical Cancer (Houston, TX, April 11–14, 2002), and new opportunities for translational research were discussed. Reviews encompassed hysterectomy with or without lymph node dissection or cervical conization in cases with microinvasion and radical trachelectomy with or without lymph node dissection as fertility-sparing surgery. Chemoradiation is used to treat advanced cervical malignancies, and the risks and benefits of radiotherapy are significant. Pelvic exenteration is used to treat certain types of recurrences. Use of the Miami pouch for continent urinary diversion was highlighted. Gynecologic oncologists expect novel in vivo imaging techniques currently being developed to help guide therapy choices within the next decade. The most significant research priorities are large group-randomized trials involving fertility-sparing procedures and the management of microinvasive carcinoma (MICA); better identification of candidates for chemoradiation; and the development of innovative approaches to exenteration. Improving diagnostic technologies, refining the criteria by which therapies are chosen, and preserving fertility remain challenges in selecting the most appropriate treatment for invasive cervical cancer. Research advances in both diagnosis and treatment are expected to improve therapy and outcomes. Cancer 2003;98(9 Suppl): 2052–2063. © 2003 American Cancer Society.

Invasive cervical cancer is defined histopathologically by lesions that invade the basement membrane and are capable of metastasizing through the lymphatic and vascular systems. Staging studies for invasive cancer are clinical and are based on physical examination and on the results of radiographic studies. There are two staging systems for cervical cancer: one from the International Federation of Gynecology and Obstetrics (FIGO) and another from the International Union Against Cancer (UICC) (Table 1).1, 2 Survival rates by FIGO stage for patients with cervical cancer are illustrated in Figure 1. Treatment failure is best predicted by early lymph node or distant organ involvement that may be undetectable by standard radiographic methods. Microinvasive cervical cancers typically are treated with hysterectomy. Early cervical cancers may be treated with either radiotherapy or surgery. Given the importance of lymph node metastasis, many believe that cervical cancer should be surgically staged. Advanced cervical cancers are treated with chemoradiation. Patients with localized central recurrences after primary radiotherapy who exhibit no evidence of any metastatic spread are candidates for pelvic exenteration. A panel of experts met at the Second International Conference on Cervical Cancer (Houston, TX, April 11–14, 2002) to discuss existing evidence and future challenges in cervical cancer. Many other patients may benefit from exciting opportunities for translational research in invasive cervical cancer.

Table 1. TNM and International Federation of Gynecology and Obstetrics Clinical Classifications of Cervical Cancera

TNMb	Description	FIGO
FIGO: International Federation of Gynecology and Obstetrics. a Sobin LH, Wittekind C, editors. TNM classification of malignant tumours, 6th edition. New York: John Wiley & Sons, 2002.2 b Staging method of the American Joint Committee on Cancer and the International Union Against Cancer.
T	Tumor
Tis	In situ	0
T1	Confined to uterus	I
T1a	Diagnosed only by microscopy	IA
T1a1	Depth ≤ 3 mm, horizontal spread ≤ 7 mm	IA1
T1a2	Depth 3–5 mm, horizontal spread ≤ 7 mm	IA2
T1b	Clinically visible or microscopic lesion, greater than T1a2	IB
T1b1	≤ 4 cm	IB1
T1b2	> 4 cm	IB2
T2	Beyond uterus but not pelvic wall or lower third of vagina	II
T2a	No parametrium	IIA
T2b	Parametrium	IIB
T3	Lower third of vagina/pelvic wall/hydronephrosis	III
T3a	Lower third of vagina	IIIA
T3b	Pelvic wall/hydronephrosis	IIIB
T4	Mucosa of bladder/rectum; beyond true pelvis	IVA
N	Regional lymph nodes
NX	Regional lymph nodes cannot be assessed
N0	No regional lymph node metastasis
N1	Regional lymph node metastasis	—
M	Distant metastasis
MX	Distant metastasis cannot be assessed
M0	No distant metastasis
M1	Distant metastasis	IVB

Figure 1. Survival rate by International Federation of Gynecology and Obstetrics stage for patients with carcinoma of the cervix uteri treated between 1990 and 1992. Reprinted from Benedet et al.1 (Carcinoma of the cervix uteri. J Epidemiol Biostat. 2001;6:7–43) with permission from Taylor & Francis Ltd. (Abingdon, United Kingdom; http://www.tandf.co.uk/journals) and J. Benedet.

MICROINVASIVE CERVICAL CANCER

Microinvasive cervical cancer (MICA) provides an opportunity for the development of novel therapies. The concept of microinvasive carcinoma of the cervix as a distinct clinical entity was introduced by Mestwerdt in 1947.3 The basis for this specific diagnosis was the classification of a subset of patients with early carcinomas of the cervix and a favorable prognosis. It is extremely important to remember that these lesions are defined microscopically and cannot be visualized on macroscopic examination. The definition of microinvasion has been debated for more than 50 years. The main issues affecting the definition are the maximum depth of stromal invasion, the significance of lymphatic space invasion, tumor volume, and the relation of the invasive pattern to the frequency of pelvic node metastases, vaginal recurrences, and ultimate survival.

The definition of microinvasive carcinoma of the cervix was modified by FIGO seven times between 1961 and 1994. The current FIGO definition was established in 1994. Stage IA1 disease is defined as having a depth of invasion no greater than 3 mm and a width no greater than 7 mm. Stage IA2 disease is defined as having a depth of invasion greater than 3 mm but no greater than 5 mm and a width no greater than 7 mm. Lymphatic vascular invasion should not affect staging.

The planar definitions outlined above were extrapolated from volumetric criteria for estimating lymph node metastasis without pretesting for validity. According to the staging guidelines, the diagnosis should involve complete excision (cone biopsy or hysterectomy). Colposcopy with biopsy, loop electrosurgical excision procedures, and the techniques of pathologic evaluation make the diagnosis of microinvasive carcinomas difficult. Consequently, decisions regarding appropriate treatment create a therapeutic dilemma. The application of the FIGO definition is most appropriate for microinvasive squamous cell carcinomas of the cervix; application of these definitions is secondary to measurements in relation to the definable basement membrane. The concept of microinvasive adenocarcinoma has not been generally accepted, but a limited number of small case reports attempting to define this issue exist. It is accepted that microinvasive adenocarcinoma of the cervix exists, but the criteria regarding depth of invasion are problematic; thus, the use of conservative treatment remains debatable.

Understanding the concepts behind the definition and diagnosis of microinvasive carcinoma of the cervix is important in understanding the biology and behavior of the disease. Lymph node metastasis is the most consistent and strongest predictor of survival for patients with early invasive cervical cancers. Lymph node metastases are closely related to depth of invasion. Treatment of microinvasive squamous cancers of the cervix warrants careful attention to detail and individualization of therapy based on histologic evaluation. Because the rates of parametrial and pelvic lymph node involvement are negligible, in general, patients with Stage IA1 carcinomas can be treated with simple hysterectomy without nodal dissection or, in selected cases, with conization of the cervix. Patients diagnosed with microinvasive carcinoma during pregnancy can choose to delay treatment to optimize fetal outcome. Patients with Stage IA2 disease should undergo radical hysterectomy and pelvic lymphadenectomy.

Preservation of fertility has become increasingly important as more women delay having children. For ethical, medical, and legal reasons, patients who opt for conservative management should fully understand the risks involved. Radical trachelectomy, with or without lymph node dissection, as fertility-sparing surgery recently has been described.

Fertility-sparing surgery is appealing to a select group of patients. Cervical cancer frequently affects women still in their reproductive years. For patients with early-stage disease and negative lymph nodes, the survival rate is very good (95–98%). With such a good prognosis for survival, loss of fertility becomes a prime concern for patients diagnosed with invasive cervical cancer. In fact, at most centers today, such patients are likely to be offered a radical hysterectomy as their only treatment option.

Cold-knife conization, largely used in the treatment of high-grade intraepithelial lesions, now is recommended as a fertility-preserving treatment for selected patients with FIGO Stage IA1 squamous cell cervical cancer, for which the rates of parametrial and pelvic lymph node involvement are negligible. The use of cone biopsy to treat microinvasive adenocarcinomas is believed to be similar in this regard, although data supporting it have not been published. The fertility rate is not affected by conization, but the risks of spontaneous second-trimester loss and prematurity are greater. These increased risks appear to be related to the amount of cervical tissue removed by conization. It has been demonstrated that the risks of second-trimester abortion and preterm delivery are related to cervical length.4–6

In patients with FIGO Stage IA2 disease, who carry a risk of up to 5% of developing lymph node metastasis, treatment must include pelvic lymph node dissection and parametrectomy so that all lymph node–bearing pelvic tissue is removed. With today's knowledge, conization is not sufficient to accomplish this goal, because it leaves in place lymph node–bearing parametrial tissue. In 1987, Dargent designed a fertility-preserving operation—radical vaginal trachelectomy (RVT), preceded by a laparoscopic pelvic lymphadenectomy (LPL)—that is oncologically satisfying, removing the affected part of the cervix and the parametrium, and leaving the body of the uterus intact (Table 2). In 1992, Dargent and Mathevet presented the first results from this work,7 which made it possible for treated women to have children without lowering their chances of cure (Table 3).

Table 2. Recurrence in a Study of 224 Patients Who Underwent Radical Vaginal Trachelectomy with Laparoscopic Pelvic Lymphadenectomy

Location of recurrence	No. of recurrences (%)
Data drawn from Dargent and Mathevet,7 Dargent,8 Covens et al.,9 Roy and Plante,10 and Shepherd et al.11 a Excludes two patients with small cell neuroendocrine tumors diagnosed only on final pathology. Both patients died despite aggressive postoperative adjuvant chemotherapy.
Parametrium	3 (1.3)
Pelvic sidewall	1 (0.4)
Distant site	3 (1.3)a
Total	7 (3.1)

Table 3. Obstetric Results of a Study of 224 Patients Who Underwent Radical Vaginal Trachelectomy with Laparoscopic Pelvic Lymphadenectomy

Event	No. of events
Data drawn from Dargent and Mathevet,7 Dargent,8 Covens et al.,9 Roy and Plante,10 and Shephard et al.11 a In 61 women. b Prematurity (birth at < 34 weeks) occurred in 18 cases.
Pregnancy	96a
Live birth	51b
First trimester loss	22
Spontaneous abortion	16
Therapeutic abortion	5
Ectopic pregnancy	1
Second trimester loss	12
Current pregnancy	11

When therapeutic cervical conization for FIGO Stage IA1 disease is performed, care must be taken to remove the affected specimen in a single piece to facilitate histopathologic evaluation, which is used to rule out the possibility of more invasive disease. Diathermy loop excision should not be used as a diagnostic or therapeutic modality for microinvasive lesions. This technique is known to frequently produce a fragmented specimen, which makes the margins and the exact depth of invasion difficult to evaluate. The results of therapeutic conization performed by cold knife, laser, or diathermy needle all are comparable to each other.

RVT with LPL is indicated when the patient desires preservation of fertility; when cervical cancer is in FIGO Stage IA1 (with vascular space involvement [VSI]), IA2, or IB1 (squamous cell carcinoma or adenocarcinoma); when any lesions present are ≤ 2 cm in diameter; and when there is limited endocervical involvement, as determined by magnetic resonance imaging and colposcopy. RVT is feasible when there is no evidence of lymph node metastasis, as determined by frozen section at laparoscopy, and when the upper endocervical margins are free of tumor, as determined by frozen section of the trachelectomy specimen. The results from 4 groups—1 in France (Dargent8: n = 82), 2 in Canada (Covens et al.9: n = 58; and Roy and Plante10: n = 44), and 1 in the United Kingdom (Shepherd et al.11: n = 40)—were presented at the Eighth Meeting of the International Gynecologic Cancer Society (Buenos Aires, Argentina, October 22–26, 2000). These results are presented in Tables 2 and 3. The numbers in these tables represent the aggregate data from the four studies.

In summary, conservative treatment should be offered to young patients affected by early-stage invasive cervical cancer. Conization for microinvasive cancer (FIGO Stage IA1 with no VSI) and radical vaginal trachelectomy should no longer be considered experimental. The data collected on LPL-RVT make it clear that the subsequent recurrence rate is not higher than the rate observed after radical hysterectomy. Leaving the upper cervix and the uterine body intact does not appear to lead to a greater risk of recurrence. Fertility is definitively preserved in most instances, but the risk of premature birth is increased.

CHEMORADIATION

Chemoradiotherapy has been demonstrated to be superior to radiation alone (Fig. 2). Radiotherapy has been the mainstay of treatment for locally advanced carcinoma of the uterine cervix. Because of the limitations of radiotherapy alone, other options have been considered. Chemotherapy agents have a cytotoxic effect and may decrease tumor volume, thereby improving radiation geometry. As single agents, cisplatin and 5-fluorouracil both have demonstrated activity against cervical cancer. These drugs also have been shown, both clinically and in vitro, to be radiosensitizers.

Figure 2. Kaplan–Meier estimates of disease-free survival rates for patients with cervical cancer who were treated with radiotherapy and chemotherapy compared with patients who were treated with radiotherapy alone (P < 0.001). Numbers of patients at risk at 3 years and at 5 years are indicated in parentheses.15

In 1999, the results of five large randomized trials12–16 were summarized in a National Cancer Institute (NCI) Clinical Announcement. (The NCI website contains a discussion and overview of the five studies.17) Collectively, these findings have changed the standard of care.

These five trials included patients with primary carcinoma of the cervix (Stages IIB–IVA as well as postoperative Stages IA2–IIA with high-risk factors, and bulky Stage IB). Lymph node status was documented surgically in three trials and radiographically in the other two. All protocols included pelvic radiation with or without cisplatin. Prescribed radiotherapy included external-beam radiation for all patients and intracavitary brachytherapy for some when indicated. Total doses ranged from 40.8 to 49.3 grays (Gy) for external total pelvic irradiation and from 30 to 40 Gy administered to Point A for brachytherapy. Patients on the investigational treatment arms also concurrently received cisplatin, in doses ranging from 40 to 75 mg/m², every 1–3 weeks. Primary outcome variables were progression-free survival and overall survival. Recurrences were classified as local or distant.

The trials involved a total of 1894 patients treated between 1986 and 1998. All five trials were well balanced, and adherence to radiation prescription was good in all trials. Recurrence was reported in 37% and 20% of patients enrolled in the control and experimental regimens, respectively. There were fewer pelvic recurrences reported in patients who received cisplatin regimens. Furthermore, distant recurrence was observed less often in patients on experimental regimens (10%) than in patients on control regimens (13%). The recurrence-free interval was longer in the experimental cisplatin group than in the control group (P < 0.001; log-rank test). The rates of freedom from recurrence at 3 years were 62% and 78%, again favoring the experimental regimens (Fig. 2). The relative risk of recurrence was 0.51 (95% confidence interval, 0.34–0.75) in favor of the group that received cisplatin. In all trials, concurrent chemoradiation was associated with a reduction in the relative risks of recurrence and death (Fig. 3). This improvement was largely due to a reduction in pelvic recurrence. The consistency of these results provides support for the use of cisplatin with radiotherapy for patients with locally advanced cervical cancer.

Figure 3. Relative risk estimates of recurrence and death from five clinical chemoradiotherapy trials. Squares indicate reduction in risk, and lines indicate 90% confidence intervals. GOG: Gynecologic Oncology Group; SWOG: Southwest Oncology Group; RTOG: Radiation Therapy Oncology Group; Cis: cisplatin; 5-FU: 5-fluorouracil; H: hydroxyurea.17

Considered collectively, the results of the randomized trials of cisplatin-based chemoradiotherapy to treat cervical cancer demonstrate the benefit associated with the combined-modality treatment (Fig. 4). Identification of the optimal agents and regimen is controversial and has not been studied extensively in randomized trials. Only one of the five randomized trials compared different cisplatin-based regimens. That trial14 demonstrated significantly greater hematologic and gastrointestinal toxicity associated with the combination of cisplatin, 5-fluorouracil, and hydroxyurea compared with weekly cisplatin alone. The use of 5-fluorouracil and mitomycin-C as radiation sensitizers for cervical cancer is based on their activity in combined-modality chemoradiotherapy for anal carcinoma.18 However, in patients treated with radiotherapy for cervical cancer, there was a threefold increase in serious late bowel toxicity among those treated with 5-fluorouracil and mitomycin-C compared with those whose treatment was combined with 5-fluorouracil alone. In a four-arm randomized trial performed by Thomas et al.,19 an improvement in survival rate was noted in patients with early-stage disease who received 5-fluorouracil infusion and standard irradiation.

Figure 4. Reduction in the risk of death due to cervical cancer in six clinical chemoradiotherapy trials. Squares indicate reductions in risk, and lines indicate 90% confidence intervals. GOG: Gynecologic Oncology Group; SWOG: Southwest Oncology Group; RTOG: Radiation Therapy Oncology Group; NCIC: National Cancer Institute of Canada; Cis: cisplatin; 5-FU: 5-fluorouracil; H: hydroxyurea.

If we accept that the benefit is the result of using cisplatin-containing chemotherapy, is it possible we can make further progress in the chemotherapy regimen? Carboplatin, which has less associated gastrointestinal, renal, and neurologic toxicity, has not been used extensively as a radiation sensitizer. A number of uncontrolled studies have evaluated the use of carboplatin on a variety of schedules.20–23 To date, to our knowledge, no controlled trials using carboplatin as a radiation sensitizer for cervical cancer have been performed. In metastatic and recurrent cervical cancer, a number of new drugs, including paclitaxel, gemcitabine, topotecan, and vinorelbine, recently have demonstrated modest single-agent activity, with response rates of 17%, 8%, 12.5%, and 18%, respectively.24–27 Increased activity was observed when these agents were used in combination with cisplatin, with response rates of 46%, 41%, 33%, and 64% for paclitaxel, gemcitabine, topotecan, and vinorelbine, respectively. In a randomized GOG trial that compared paclitaxel and cisplatin with cisplatin alone, the combination was associated with a significantly greater response rate (36.2% vs. 19.4%; P = 0.002). In a GOG trial that currently is under way, the combination of topotecan and cisplatin is being compared with cisplatin alone. Numerous studies using these newer agents in combination with cisplatin as radiation sensitizers have been initiated.28–36 In the Gynecologic Oncology Group trials, these regimens are being evaluated for use with both pelvic and extended (pelvic and paraaortic) irradiation. An expanding series of trials will help to define the role of these new agents in combination with radiotherapy for the treatment of cervical cancer.

RADIOTHERAPY

The risks associated with radiotherapy, alone or in combination with chemotherapy, are numerous, and all are increased by smoking and vascular occlusive diseases.37, 38 Radiotherapy plays an important role in the curative and palliative treatment of invasive cervical cancers. Irradiation of the cervix and pelvic lymph nodes affects adjacent structures. Normal tissue complications observed during or after radiotherapy result from the killing of critical target cells within the tissue—cells that also are critical to the tissue. The possible clinical manifestations of the loss of critical target cells include acute reactions that occur before therapy is completed and late effects, which occur after the completion of therapy (Figs. 5,6). The cumulative changes in normal tissue make surgery hazardous, and the management of such injuries is among the most difficult tasks faced by gynecologic oncologists.39

Figure 5. Pathologic changes during acute clinical period with loss of functional villi as demonstrated in normal rat bowel (A) and rat bowel 96 hours after 10 Gy (B).

Figure 6. Clinical (A) and pathologic (B) changes with late radiation complications. Note serosal fibrosis and adhesions.

The organs that invariably are exposed during pelvic radiotherapy are at risk for injury; the organs most frequently involved are the small intestines, urinary bladder, and rectosigmoid colon. Early effects of radiotherapy may be demonstrated by clinical manifestations such as small bowel dysfunction, including diarrhea, abdominal cramping, dehydration, electrolyte imbalance, and weight loss. The target cells are crypt stem cells of the small intestine gut epithelium. Mature cells of the villi are lost over several days, and no new cells are available to replace them. The villi begin to shorten, and there is a decrease in the total surface area. Absorption of small bowel contents is hindered. Diarrhea is aggravated by irritation caused by bile salts and other substrates. The mucosa of the small intestine can become denuded, increasing the risk of bleeding and infection.

Late effects may manifest themselves clinically 3 or more months after the completion of radiotherapy. Vascular changes, fibrosis, stricture, perforation, and necrosis characterize late effects. The small intestine is sensitive to these injuries, and fibrosis with obstruction necessitates surgical intervention. Fistulas between the small intestine and bladder and/or vagina can occur. The sigmoid colon and rectum may undergo fibrosis with stricture, and in some patients, necrosis with fistula will occur.

Regardless of the location of the injury, surgery is hazardous for patients with late effects, because these effects impair wound healing. Progressive disease can bring about symptoms that are indistinguishable from radiation injury. Preoperative evaluation must be comprehensive.

Early and late effects vary in intensity from patient to patient.40 Normal tissue from various individuals may have important differences that predispose certain patients to damage. Individual patients with identical tumors may have different tumor outcomes; some tumors may be controlled, whereas others may not. Tissue heterogeneity applies to both tumor and normal tissue. Numerous intrinsic and extrinsic factors contribute to tissue heterogeneity. Intrinsic factors include inherent radiosensitivity, biochemical repair processes, modes of cell death, and genetic instability. Extrinsic factors tend to be related to physiologic differences, such as degree of vascularity, oxygen levels, availability of nutrients, and pH. The practical ramification of tissue heterogeneity is the inability to predict which patients will experience tumor control and which patients will experience complications.

Clinical considerations, such as previous abdominal surgery, a history of peritonitis or pelvic infection, technical factors related to packing of intracavitary applicators, and preexisting medical problems, predispose patients to early and late effects. The radiotherapy parameters also influence these effects. Total dose, treatment time, size of fractionation, number of fractions, and size of field treated all have clinical significance.

It is important for gynecologic oncologists to review radiotherapy records. There is no standard-size therapy portal for external beam therapy. Fields are designed to fit each individual patient. Although there is no standard dose, a common prescription for external beam therapy is a total dose of 45 Gy, or 1.8 Gy daily for 5 days per week for 5 weeks. The superior boundary of the pelvic portal typically is the L4-L5 vertebral interspace. The external beam therapy usually is followed by brachytherapy. However, field size and design, as well as the total dose, are at the discretion of the radiotherapist. Some patients with advanced-stage disease may receive 55-Gy external beam therapy. Specific lymph node areas may receive a booster dose. The standard pelvic portal may be extended to cover common iliac and paraaortic lymph node areas with proven or suspected involvement. All information regarding previous therapy, size and extent of the primary cancer, and surgical and medical history are useful in planning a surgical procedure.

Before an operation, it is useful to evaluate the entire gastrointestinal tract for the presence of radiation-induced abnormalities and for metastatic disease. Evaluation of the urinary tract also is necessary. Ureteral obstruction necessitates cystoscopy and retrograde pyelography or percutaneous nephrostomy. Pelvic examination also can yield information concerning local control. Necrotic areas, fistula tracts, and any other suspicious areas should be evaluated with biopsy. Diagnostic radiographic procedures may identify metastatic disease or define areas of injury. It is important to note that multiple sites may be involved, because normal tissue response is not uniform within the treatment field. Injury occurs randomly.

Radiation-induced changes in normal tissue are unavoidable and unpredictable. Patients with advanced-stage disease confined to the cervix, with or without regional lymph node involvement, warrant aggressive therapeutic radiotherapy. The concomitant use of chemotherapy is considered state-of-the-art treatment, but the long-term impact of combination therapy is not well defined for patients with cervical malignancies.

Patients with localized central recurrences after primary radiotherapy who exhibit no evidence of metastatic spread are candidates for pelvic exenteration. Pelvic exenteration is an operation in which the uterus and the surrounding bladder and rectal tissue are removed and extensive lymph node dissection and vaginal reconstruction are performed. Depending on the size and distribution of the recurrent lesion, some central recurrences may allow preservation of the bladder or rectum. Many innovative bladder reconstruction procedures have been developed and are being further improved. Previously, urinary conduits constructed with stomas emptied into appliances. Colostomies, which similarly allowed drainage into an appliance, also were a routine part of pelvic exenteration. The advent of continent urinary reservoirs has allowed patients to avoid wearing ostomy bags. The stoma can be catheterized at regular intervals. In parallel, surgeons have developed rectum-sparing procedures that involve preservation of the rectum, anastomosis of healthy colon to the rectum, and temporary colostomy (usually left in place for 6–8 weeks). These advances have made what previously was a deforming operation requiring two permanent ostomy appliances into a procedure that may result in only a single, small, continent ostomy measuring approximately 1 cm.

Many advances also have been made in vaginal reconstruction. Early vaginal reconstructions often were simple skin grafts that had variable success rates. Advances in the field of reconstructive surgery involving the use of muscle-flap grafts, in which skin, fat, and muscle with an intact blood supply are used, have become standard in postexenteration vaginal reconstruction. Further advances in the field of reconstructive surgery have allowed the use of abdominal rectus muscle flaps. Overall, muscle-flap grafts have greatly improved patient satisfaction, quality of life, and sexual function for patients undergoing exenteration. Pelvic exenterations fail to cure patients when micrometastases that are present at the time of operation go undetected. Advances in in vivo molecular imaging could aid in the detection of these metastases and allow the identification of cases not suitable for exenteration and cases in which the addition of chemotherapy or secondary chemoprevention after exenteration could prolong survival.

PELVIC EXENTERATION

Exenterative procedures may be curative for patients with central recurrences of cervical cancer after radiotherapy.41–43 These procedures are increasingly palatable to patients because advances in surgical techniques have improved quality of life. Indications for urinary diversion in patients with gynecologic disorders vary. Most urinary diversion procedures are performed in conjunction with surgery for recurrent cervical cancer or for complications of pelvic radiotherapy. In some cases, women with persistent cancer and acute ureteral obstruction who otherwise are in good health may benefit from percutaneous nephrostomy. In other cases, particularly when patients have undergone multiple surgeries for urinary incontinence, supravesical diversion may be considered if severe incontinence persists. This approach also may be appropriate for women who have experienced failure after multiple bladder fistula repairs. In most of the gynecologic literature, however, urinary diversion is described as a component of pelvic exenterative surgery (Fig. 7). The most common indication for urinary diversion involves bilateral ureteral obstruction—vesicovaginal fistulas or severe bladder damage resulting from pelvic radiotherapy.

Figure 7. Pelvic exenteration procedure.

In some situations, temporary measures such as a percutaneous nephrostomy may be applicable. Percutaneous nephrostomies may be of special use for patients with cancer and ureteral obstruction who are undergoing or plan to undergo pelvic irradiation or chemotherapy. The tumor shrinkage subsequent to the therapy may allow the urinary tract to return to a normal functional status. If permanent diversion is necessary, however, one might select a urinary conduit or a urinary reservoir. In general, conduits (noncontinent urinary diversion) are easier and faster to create and carry a smaller risk of complications compared with reservoirs (continent urinary diversion). If the patient is in good general health and the procedure is performed with curative intent, we prefer to use a continent form of urinary diversion, the Miami Pouch. When the procedure is performed for palliative reasons or the condition of the patient requires a rapid procedure for any of several reasons (e.g., bleeding, extensive disease, or poor general condition), the urinary diversion of choice is a conduit, which can involve the use of the ileum, sigmoid colon, or transverse colon, or even a so-called wet colostomy. When an adult patient has no history of pelvic irradiation, ileal conduits commonly are used for diversion. Some investigators have advocated the use of sigmoid conduits in conjunction with pelvic exenteration to avoid small bowel anastomosis. However, some investigators report that when colonic segments are subjected to irradiation, the early and late complications are similar or identical to those found in irradiated ileal segments. Transverse colon conduits have gained considerable popularity with gynecologic oncologists. Their advantages, which include simplicity of performance, use of a nonirradiated bowel segment, and fewer stomal complications, have made their use the procedure of choice at many academic centers. Jejunal conduits, which are introduced in an attempt to use less heavily irradiated bowel in patients who have undergone pelvic irradiation, have been shown to be associated with significant complications, such as hyperchloremic acidosis with hyponatremia, hypokalemia, and anemia. Although this jejunal syndrome is correctable with the administration of salt, it makes the selection of these bowel segments less attractive as a permanent form of diversion. Since l988, the staff at the University of Miami (Miami, FL) has been using the continent ileocolonic reservoir, the Miami Pouch, as the preferred method of continent urinary diversion. During the past few years, the Miami Pouch has become one of the most common forms of urinary diversion for patients with gynecologic cancer in the United States. The benefits of the Miami Pouch include continence and improved renal function related to the low-pressure system (Table 4). The features of this device include a detubularized colonic pouch, a tapered ileum, a reinforced ileocecal valve, and a nontunneled ureterocolonic anastomosis. Assessment of complications has led to the conclusion that the Miami Pouch is a technically simple and reliable reservoir compared with other forms of continent urinary diversion (Table 5).

Table 4. Miami Pouch: Management of Ureteral Complications

Complication	Treatment (%)
	Conservative	Surgical
Stenosis	12/15 (80)	3/15 (20)
Obstruction	7/9 (71)	2/9 (29)
Leak	4/8 (50)	4/8 (50)
Pyelonephritis	13/13 (100)	0/13 (0)

Table 5. Miami Pouch: Management of Pouch Complications

Pouch complication	Treatment (%)
	Conservative	Surgical
Fistula	5/8 (60)	3/8 (40)
Incontinence	7/12 (56)	5/12 (44)
Difficult catheterization	15/15 (100)	0/15 (0)
Urinary calculi	3/6 (50)	3/6 (50)
Ileal stenosis	2/2 (100)	0/2 (0)

FUTURE DIRECTIONS

Novel in vivo molecular imaging techniques are under intensive development and may provide important diagnostic and prognostic information that can be used to guide new therapies in the next five years. Early cervical cancers may be treated with either radiotherapy or surgery. Microinvasive cervical cancers typically are treated with hysterectomy. Several investigators have suggested less invasive therapies, such as cone biopsy. A number of investigators have ongoing series of patients who have been treated with cone biopsies. The clear advantage of treating patients with cone biopsy is that uterine preservation allows patients to retain the ability to bear children. The long-term recurrence rate in the uterus and pelvic lymph nodes after cone biopsy has not been established. This therapy clearly should be tested in a randomized clinical trial. Small Stage I cervical cancers confined to the cervix typically are treated with radical hysterectomy and pelvic lymph node dissection or with radiotherapy. Some investigators have suggested radical trachelectomy and laparoscopic or extraperitoneal lymph node dissection as an alternative to standard therapy. Radical trachelectomy preserves the uterus and thus allows the possibility of an intrauterine pregnancy. Several investigators are collecting case reports of patients treated with radical trachelectomy with cerclage placement and laparoscopic lymph node dissection. Although the advantages of uterine preservation and the possibility of pregnancy are clear benefits, the recurrence rate of cervical cancer and the long-term survival results are not yet known. To answer questions unanswered by experience with radical trachelectomy as a treatment strategy for early invasive cervical cancer, investigators must subject this procedure to a randomized clinical trial. Surgical therapy (e.g., vaginal hysterectomy for microinvasive cancer, extrafascial hysterectomy for microinvasive cancer, radical hysterectomy for early invasive cancer, and radical trachelectomy for early invasive cancer) allows preservation of the ovaries and, thus, continued ovarian function in young patients. When radiotherapy is planned for young patients, ovarian transposition may be performed, either by laparoscopic methods or by laparotomy. Although ovarian transposition is designed to move the ovaries out of the irradiation field, many case series have demonstrated that ovarian function does not continue for as long as expected (i.e., more than the expected number of patients lose ovarian function). At present, it is not clear why these procedures are not more successful in preserving ovarian function. Advanced cervical cancers are best treated with chemoradiotherapy. The most effective chemotherapeutic agents are those that act as radiation sensitizers (e.g., cisplatin). Several promising radiosensitizing agents are being developed. Refinement of the criteria for patients likely to benefit from chemoradiotherapy is a critical problem that needs to be solved. In vivo molecular imaging may make possible the early identification of vascular metastases and metastases in the lymph vessels and nodes. The ability to identify these metastases and understand how they affect prognosis would aid physicians in determining optimal therapies, thereby maximizing the cost-benefit ratio of complications and survival. Trials involving maintenance chemotherapy could be beneficial.

The most important priorities for research on invasive cervical cancer are large group randomized trials of management strategies for microinvasive disease, development of fertility-sparing procedures, identification of candidates for chemoradiotherapy, and development of innovative approaches to exenteration. There also are many other potentially valuable areas for new research on the treatment of invasive cervical cancer. Important advances in in vivo molecular imaging could yield better diagnostic and prognostic information and thus lead to the development of more appropriate therapies that would decrease morbidity and increase survival. A thorough review of current staging procedures and their correlations with survival should be undertaken. Several investigators believe that it is incumbent upon health care professionals to consider surgical staging for cervical cancer to be the standard both for understanding the natural history of the disease's metastatic process and for better ascertaining appropriate treatment strategies.

The goal of surgical staging is to resolve any ambiguities present in clinical staging information. Fifty percent of patients with Stage IB2–IVA disease are thought to have positive lymph nodes. Surgical removal of lymph nodes is the most sensitive and specific way to detect lymph node metastasis. The information gained from lymph node dissection can be applied to designing better radiation fields as well as to removing microscopically positive lymph nodes and possibly debulking macroscopically positive lymph nodes. The use of lymphadenectomy for patients with bulky or locally advanced cervical cancer leads to improved survival rates. Future directions should include modification of the FIGO system into clinical and surgical staging formats. Future research on imaging techniques, with an emphasis on the relation between lymph node metastasis size and positive scans (e.g., positron-emission tomography scans), is warranted. In addition, centers of excellence specializing in cervical cancer treatment should be developed to provide the best possible options for evaluation and management, whether it be surgery or primary radiotherapy with concurrent chemotherapy.44–47

High priority also should be given to the funding of randomized clinical trials (RCTs) comparing cone biopsy with standard therapy combined with hysterectomy for both microinvasive squamous cell lesions and microinvasive adenocarcinoma. Such trials should be designed carefully and stratify patients by age, qualitative consensus pathologic review, quantitative histopathologic measurements, and relevant surrogate endpoint biomarkers. In addition, strong consideration should be given to an RCT comparing radical trachelectomy with radical hysterectomy for the treatment of early invasive cancer. To address ongoing concerns, investigators should carefully design such a trial to assess both recurrence and survival and stratify patients by age. Again, qualitative consensus pathologic review, quantitative histopathologic measurements, and relevant surrogate endpoint biomarkers should be taken into account. The field of cell and tissue engineering may allow for cervical reconstruction or even uterine reconstruction. Advances in these fields would make a wider range of curative therapies possible while still allowing fertility preservation. Improvements in survival due to chemoradiotherapy already have been demonstrated. Advances in chemoradiotherapy could involve agents that induce less morbidity or improve survival through novel molecular mechanisms. Several new radiotherapeutic techniques have been developed, and they should be tested in RCTs against standard radiotherapy with the goal of reducing complications and improving survival rates. There are many early and late effects associated with radiotherapy; all organs in the irradiated field, which for cervical cancer typically includes organs in the pelvis and sometimes those along the aorta, are susceptible to these effects. Increased attention to the molecular mechanisms of these complications, as well as to pharmaceuticals and changes in time-dose fractionation, which could decrease late and early complications, should be a high priority in clinical trials. Efforts also should be directed toward the development of in vivo molecular imaging techniques so that patient selection for exenteration, a highly invasive procedure, is limited to those who have the greatest probability of benefit. Although there have been several noteworthy advances in reconstruction of the bladder, rectum, and vagina, further advances are expected, and quality-of-life studies of patients undergoing these procedures are warranted. Attention also should be paid to the completion of a maintenance chemotherapy study of patients who are identified to be good candidates for exenteration but whose risk of recurrence is greater than average.