Practice nursing care for Clients with Cancer II: Nursing Interventions Related to Side Effects of Treatment

GYNECOLOGIC CANCERS

Molecular Biology of Gynecologic Cancers

ENDOMETRIAL CANCER

Although this most common gynecologic cancer carries the best prognosis, it is important to differentiate between the classic endometrioid tumors that tend to be estrogen dependent and well differentiated, from other less common high-risk uterine malignancies, such as uterine papillary serous carcinomas (UPSCs), clear cell carcinomas, mixed mullerian tumors (MMTs), or sarcomas, which appear to have a different biology.

MICROSATELLITE INSTABILITY

The replication error (RER+) phenotype is characteristic of cancers arising in HNPCC kindreds (a familial cancer syndrome with a high incidence of colon, endometrial, gastric, and a lower incidence of ovarian and pancreatic cancers), and is also found in approximately 20% of sporadic endometrial cancers. The RER+ phenotype per se does not appear to correlate with clinicopathologic features of the tumors or clinical outcome; thus the hereditary form of endometrial cancer alone does not appear to portend a worse prognosis than the sporadic form. Mutation of the transforming growth factor (TGF) beta receptor type II gene is common in RER+ colon and gastric cancers, but uncommon in RER+ endometrial cancers even those arising in HNPCC kindreds, suggesting that the genesis of RER+ tumors even within the same familial cancer syndrome is not the same. Among sporadic endometrial cancers, when present, the microsatellite instability is confined to the malignant cells, and is not seen in the adjacent normal epithelium. Although inactivation of both alleles of either hMSH2 or MLH1 (DNA mismatch repair genes) appears to underlie microsatellite instability in tumors of HNPCC kindreds, similar to the findings in sporadic colon cancers, sporadic endometrial cancers were not associated with mutations of any of the four known human mismatch repair genes. In contrast, 25% of uterine sarcomas, an entity not recognized to be part of a familial cancer syndrome, exhibits microsatellite instability that may be related to a mutation in the hMSH2 gene. Again, clinical outcome was not correlated with RER+ status in those tumors; thus this finding does not underlie the more aggressive biology of the sarcomas.

LOSS OF HETEROZYGOSITY AND TUMOR SUPPRESSOR GENES

Unlike most other cancers studied so far, endometrial cancer is not characterized by a small number of loci that are involved by loss of heterozygosity. In endometrial cancers, most chromosomes contain regions of sustained allelic loss. Chromosome 17p is frequently involved, which contains the p53 gene known to confer a poor prognosis in endometrial cancers when mutated or deleted, and which plays an important role in the transition to carcinoma from atypical hyperplasia. Other frequently involved regions include 10q, 3p, and 18q, with no significant differences found in the chromosomal regions involved in endometrial from cervical cancers. Chromosome 18q contains the DCC gene, a putative tumor suppressor gene frequently mutated in colon cancers. While no mutations were found in the exons of the DCC gene in endometrial cancers, chromosome 18 was found to be capable of suppressing tumorigenicity of endometrial cancer cells in nude mice by the procedure of microcell fusion, with DCC expression elevated in most of the suppressed hybrids. These data strongly suggest that some gene on chromosome 18, is a tumor suppressor gene in endometrial cancers.

ONCOGENES

Kirsten (Ki)-ras activating point mutations in codons 12 and 13 have been implicated in the development of atypical endometrial hyperplasias and endometrioid carcinomas in Japanese women, as well as in colon cancers. Recent studies of endometrial cancers arising in US women, however, show the prevalence of such mutations to be significantly lower (11% versus 31%) than in Japan. Endometrioid cancers arising in Japanese women tend to be high grade and deeply invasive, are less common than in the United States, and are not generally associated with hormonal exposure. These data suggest that genetic or environmental factors may influence endometrial carcinogenesis. Notable was the almost complete absence of Ki-ras mutations in an aggressive form of endometrial cancer, UPSC, when compared to the usual endometrioid tumors.

Overexpression of HER-2/neu has been associated with advanced stage, deep myometrial invasion, and poor survival in endometrial cancers in several studies. Gene amplification did not underlie all cases of HER-2/neu overexpression, although both gene amplification and overexpression were each associated with poor outcomes. When multivariate analysis was used to determine if HER-2/neu was an independent prognostic factor in endometrial cancers taking into account other molecular features such as DNA ploidy, epidermal growth factor receptor, or p53 status, HER-2/neu status failed to achieve significance. Both c-myc gene amplification and c-fms overexpression have also been associated with advanced stage and high grade endometrial cancers.

HORMONE-RELATED MOLECULAR ABNORMALITIES

This is a logical direction for endometrial cancer research, because estrogen acts as a tumor promoter for the classic endometrioid cancers. Aromatase cytochrome p450 is part of the complex responsible for conversion of C19 steroids to estrogen; its increased expression in endometrial cancers, but not in normal endometria, suggests a role in promotion of neoplastic proliferation.

The findings of both gonadotropin-releasing hormone (GnRH) (the ligand) and its receptor, as well as the luteinizing hormone (LH) receptor in endometrial cancers may serve as a rationale for the therapeutic use of GnRH analogues in the treatment of endometrial cancer (although this approach has not been shown to be clinically relevant to date). GnRH analogues may act directly on GnRH-sensitive cancers, as well as indirectly by decreasing systemic LH levels. A clue to the signal transduction pathway of GnRH comes from data in ovarian cancers, where GnRH has an antimitogenic effect by stimulating protein tyrosine phosphatase activity.

Tamoxifen has been associated with an increased risk of development of endometrial cancers, with a possible preferential risk for the high grade nonendometrioid subtypes, such as UPSC or MMTs, although this is controversial. Molecular abnormalities associated with UPSC include overexpression of p53 and c-myc gene amplification; p53 overexpression correlated with a poor prognosis. Two functional different isoforms of progesterone receptor have been described, which may account for some of the tissue-specific differences in the effects of progestins and antiprogestins on the breast as compared with the endometrium.

MOLECULES INVOLVED IN ADHESION AND INVASION

Integrin (cell adhesion molecules) expression inversely correlates with grade in endometrial cancers, with the loss of the alpha(2)beta(1)-integrin associated with lymph node metastases. Variant forms of CD44 (a molecule important for cell adhesion and migration) were less frequently expressed in endometrial cancers than in normal endometria, and absence of CD44 was significantly associated with an increased propensity for lymph-vascular space invasion. These data suggest that CD44 may play an important role in the function of the normal endometria, where it is strongly expressed near the basement membrane, and its loss may be related to invasion and metastasis.

OVARIAN CANCER

Most of the recent breakthroughs in understanding the molecular basis for this disease has been in the area of hereditary epithelial ovarian cancer syndromes, which affect up to 5% of ovarian cancer cases. Much work is still needed to understand the biology underlying sporadic ovarian cancers, which invariably present as advanced stage disease and have a poor long-term outcome.

CERVICAL AND VULVAR CANCER

The study of the role of viruses in the carcinogenesis of lower genital tract malignancies (thought to be a field effect) has focused on cervical cancer, the third most common gynecologic cancer in the United States. Extension of those studies to vulvar cancer has led to support for two separate etiologies of vulvar cancer: one related to human papillomavirus (HPV), with epidemiologic risk factors similar to that for cervical cancer, and the other that does not appear to be HPV related.

HUMAN PAPILLOMAVIRUS. That HPV is a critical factor for cervical carcinogenesis, and that the HPV E6 and E7 genes are oncogenic are clearly established. Infection of human keratinocytes by the oncogenic HPV subtypes leads to abnormalities in differentiation and growth; however, only after long-term culture of immortalized cells does an occasional clone become tumorigenic in nude mice, suggesting that HPV infection alone is not sufficient for cervical carcinogenesis. This is supported by data from transgenic mice studies, where E6/E7 genes can give rise to hyperplastic and neoplastic lesions of epithelial cell types after a latent period, however, epidermoid cervical cancers have not been noted. Cervical cancers of mesenchymal origin were noted to arise after a long latent period in some of the female progeny of transgenic mice into whom HPV-18 LCR/E6/E7 was introduced. While the majority of invasive cancers contain integrated forms of HPV, usually at fragile sites that result in cis activation of protooncogenes such as c-myc, and the large majority of dysplasias contain episomal forms of HPV, this is not always the case. In invasive cancers, HPV-18 is always found to be integrated in the host genome, while HPV-16 can be found in an episomal location one third of the time. When DNA integration occurs, it does so by preferentially disrupting the E2 open reading frame, thus, the negative effect of the E2 protein on E6/E7 transcriptional activity; or in such a way that transcriptional initiation from host sequences gives rise to overexpression of E6/E7. High level transcription of E6/E7 is seen in CIN 3 and invasive cancers when compared to CIN 1 and 2. Such overexpression has been shown in vitro to lead to radioresistance of cervical cancer cells. Among both low grade and high grade dysplasias, the presence of oncogenic HPV subtypes has been reported to be closely associated with monoclonality.

MOLECULAR COFACTORS IMPORTANT TO CERVICAL CARCINOGENESIS

Because HPV infection is not sufficient for cervical carcinogenesis, attention has focused on molecular cofactors important to this process, such as coinfection by HSV-2, and the presence of activated Ha-ras; the latter results in rearrangements and amplifications of the HPV-16 sequence. Many positive and negative transcriptional regulators of E6/E7 transcription have been identified. The presence of the glucocorticoid response element 5' of the HPV genome probably underlies the clinical progression of HPV infection seen in pregnancy. Although retinoic acid represses HPV transcription in normal and malignant cells, its induction of retinoic acid receptor beta is restricted to normal cells. Both the retinoic acid receptor beta gene and a locus on chromosome 11q23 may have tumor suppressive properties in epidermoid cancer. Loss of heterozygosity studies demonstrate allelic loss of many chromosomes including 11q, but most frequently involving 3p, 6p, and 18q.

The immune response is likely to be key in determining malignant transformation of HPV-infected cervical epithelium. The consequences of human immunodeficiency virus infection include a dramatic increase in the risk for cervical dysplasia and invasive cancer, the degree of which correlates with the level of immunosuppression. Loss of expression of HLA class I alleles along with interference with the transporter associated with antigen presentation in cervical cancers is common; such changes may influence specific immunogenic presentation by tumors. In addition, the finding of certain HLA class II haplotypes in the cancer (when compared to cervical DNA from controls), which may influence the immune response to specific HPV-encoded epitopes, may contribute to the development of cervical neoplasia. Similar findings were seen when HLA class II haplotypes in CIN were compared to controls; the HLA DQB1 haplotype was significantly more associated with CIN.

In general, more than 90% of squamous cell cervical cancers contain HPV DNA, and rarely p53 mutations are seen. p53 mutations appear to be more common in HPV-negative cancers, although HPV-negative cancers that do not contain p53 mutations exist. The latter do not contain MDM2 gene (capable of binding to p53) amplifications either. Among HPV-negative tumors, c-myc overexpression has been associated with an increased risk of metastasis in early stage disease. Among HPV-positive cancers, nuclear c-myb expression was correlated with presence of HPV; c-myb can transactivate the HPV-16 promoter, as it can HIV-1 and c-myc. In most studies, HPV status was not a strong independent prognosticator of outcome in cervical cancer patients; however, there appears to exist a trend for HPV-negative tumors to do worse.

NON-HPV-RELATED MOLECULAR ABNORMALITIES IN CERVICAL CANCERS

Expression of other cell cycle genes such as bcl-1 and bcl-2 has been studied in cervical cancer. Bcl-1 (cyclin D1) is capable of binding to the Rb protein, and is overexpressed and amplified in the majority of cervical and vulvar cancer cell lines, and its level of expression is elevated by activated CSF-1 receptor. Bcl-2 overexpression, which protects against apoptosis and differentiation, was not found to relate to HPV status, but was more likely to be seen in CIN 3 rather than low grade dysplasias; thus, its expression may be early event important to malignant transformation. In vitro, increased bcl-2 expression is noted in cervical cancer cell lines that contain an inactive p53.

Proteinases capable of degrading extracellular matrix may underlie the propensity of cervical cancer to invade adjacent tissues. Both the expression (as measured by reverse transcriptase in situ PCR and activity of metalloproteinases have been described in cervical cancers, with more activity seen than in ovarian or endometrial cancers; the opposite finding is seen for expression of TIMP-1, a specific inhibitor of the matrix metalloproteinases. In fact, the ratio of matrix metalloproteinases to TIMP-1 is increased in those cervical cancers and their surrounding stroma that have a poor prognosis.

CARCINOMA OF THE CERVIX

EPIDEMIOLOGY

The American Cancer Society estimates that there will be 15,800 new cases of invasive cervical cancer and 65,000 cases of carcinoma in situ diagnosed in the United States in 1995. Forty-eight hundred patients are expected to die of cervical cancer; this represents approximately 2% of all cancer deaths in women and 18% of deaths from gynecologic cancers. In the United States, age-adjusted death rates from cervical cancer have declined steadily since statistics were first collected in the 1930s. Although the adoption of routine screening programs with pelvic examinations and cervical cytology is primarily responsible for this improvement, the death rates from cervical cancer had begun to decrease before the implementation of Papanicolaou (Pap) screening, suggesting that other unknown factors may have played some role.

Cervical cancer continues to be the leading cause of cancer deaths for women of many economically underdeveloped countries. Incidence and death rates are particularly high in Latin America, Africa, India, and Eastern Europe. Cervical cancer is the cause of approximately 25% of cancer-related deaths in Mexican women.

Squamous cell carcinoma of the cervix and its intraepithelial precursor follow a pattern typical of sexually transmitted disease. The risk of cervical cancer is increased in prostitutes and in women who have first coitus at a young age, multiple sexual partners, sexually transmitted diseases, or who bear children at a young age. Promiscuous sexual behavior in male partners may also be an important risk factor. Other factors associated with cervical cancer include cigarette smoking, immunodeficiency, vitamin A and C deficiency, and possibly oral contraceptive use. In the United States, the incidence is greatest in Native American, African-American, and Hispanic American women.

Recent studies suggest that the number of cervical adenocarcinomas affecting young women in their 20s and 30s may have increased. Although, on the basis of epidemiologic studies, some investigators have suggested an association between adenocarcinoma and prolonged oral contraceptive use, [ref: 9] others have questioned the validity of these conclusions because of the many potential confounding risk factors.

Molecular studies have demonstrated a strong relationship between human papillomavirus (HPV), cervical intraepithelial neoplasia (CIN), and invasive carcinomas of the cervix. HPV DNA has been identified in more than 60% of cervical cancers; HPV DNA transcripts and protein products also have been identified in invasive cervical carcinomas. In high-grade CIN and invasive carcinoma, the papillomavirus DNA is typically integrated into the human genome rather than remaining in an intact viral capsid. Many of the more than 70 HPV identified subtypes have been isolated, sequenced, and cloned. Types 6 and 11 usually cause benign warts (condyloma acuminata), but are occasionally associated with invasive lesions. Types 16, 18, 31, and 33 are commonly associated with high-grade CIN and invasive cervical cancer. HPV-18 has been associated with poorly differentiated carcinomas, an increased incidence of lymph node involvement, and a high rate of disease recurrence, whereas HPV-16 has been associated with large cell keratinizing tumors and a lower recurrence rate. It has been theorized that HPV-mediated carcinogenesis results from the binding of E6 and E7 proteins (produced by HPV-16 and HPV-18) to Rb and p53 tumor suppressor proteins, interfering with the normal regulation of cell proliferation.

Although most recent investigators have concluded that HPV is probably an important factor in cervical carcinogenesis, its role and that of other viral agents continues to be controversial. Some investigators have argued that other viral agents, including herpes simplex virus type II and Epstein-Barr virus, may also be important.

Human immunodeficiency virus seropositivity has been associated with a high incidence (up to 40%) of CIN and invasive cancer. This is at least partly due to an overlap in risk factors for the two diseases, although changes in cell-mediated immunity may also play a role in the development of cervical cancer. Some investigators have suggested that cervical cancer is a more aggressive disease in these immunosuppressed patients. For these reasons, frequent surveillance with Pap smears, pelvic examination, and colposcopy (when indicated) should be part of the routine care of these patients. In 1993, the Center for Disease Control added cervical cancer to the list of AIDS-defining neoplasms.

NATURAL HISTORY AND PATTERN OF SPREAD

The junction between the primarily columnar epithelium of the endocervix and squamous epithelium of the ectocervix is a site of continuous metaplastic change; this change is most active in utero, at puberty, and during first pregnancy, and declines after menopause.

Viral-induced atypical squamous metaplasia developing in this region can progress to higher-grade squamous intraepithelial lesions. The greatest risk of neoplastic transformation coincides with periods of greatest metaplastic activity, and most carcinomas arise from this zone of metaplastic transformation in the squamocolumnar junction.

The mean age of women with CIN is 15.6 years younger than that of women with invasive cancer, suggesting a slow progression of CIN to invasive carcinoma. In a 13-year observational study of women with CIN 3, Miller found that disease progressed in only 14%, whereas it persisted in 61% and disappeared in the remainder. Syrjanen and colleagues reported spontaneous regression in 38% of high-grade HPV-associated squamous intraepithelial lesions. However, in a large prospective study, Richart and Barron reported mean times to development of carcinoma in situ of 58, 38, and 12 months for patients with mild, moderate, or severe dysplasia, respectively, and predicted that 66% of all dysplasias would progress to carcinoma in situ within 10 years. Once tumor has broken through the basement membrane, it may penetrate the cervical stroma directly or via vascular channels. Invasive tumors may develop as exophytic growths protruding from the cervix into the vagina or as endocervical lesions that can cause massive expansion of the cervix despite a relatively normal appearing cervical portio. From the cervix, tumor may extend superiorly to the lower uterine segment, inferiorly to the vagina, or into the paracervical spaces via the broad or uterosacral ligaments. Tumor may become fixed to the pelvic wall by direct extension or by coalescence of central tumor with regional adenopathy. Tumor may also extend anteriorly to involve the bladder or posteriorly to the rectum, although rectal mucosal involvement is a rare finding at initial presentation. The cervix has a rich supply of lymphatics organized in three anastomosing plexuses that drain the mucosal, muscularis, and serosal layers. The lymphatics of the cervix also anastomose extensively with those of the lower uterine segment, possibly explaining the frequency of uterine extension from endocervical primary tumors. The most important lymphatic collecting trunks exit laterally from the uterine isthmus in three groups. Upper branches originating in the anterior and lateral cervix follow the uterine artery, are sometimes interrupted by a node as they cross the ureter, and terminate in the uppermost hypogastric nodes. Middle branches drain to deeper hypogastric (obturator) nodes, and the lowest branches follow a posterior course to the inferior and superior gluteal, common iliac, presacral, and subaortic nodes. Additional posterior lymphatic channels arising from the posterior cervical wall may drain to superior rectal nodes or may continue upward in the retrorectal space to the subaortic nodes overlying the sacral promontory. Anterior collecting trunks pass between the cervix and bladder with the superior vesical artery to terminate in the internal iliac nodes. Summarizes the reported incidences of pelvic and paraaortic node involvement for patients who underwent lymphadenectomy as part of primary surgical treatment or before radiation therapy for cervical carcinomas. The incidences reported for radical hysterectomy series vary widely, probably reflecting surgeons' different criteria for selecting patients for radical surgery rather than for primary radiation treatment. Many series exclude patients with extrapelvic disease. Variations in the completeness of lymphadenectomies and histologic processing may also lead to underestimates of the true incidence of regional spread from carcinomas of the cervix. Cervical cancer usually follows a relatively orderly pattern of metastatic progression initially to primary echelon nodes in the pelvis, then to paraaortic nodes and distant sites. Even patients with locoregionally advanced disease rarely have detectable hematogenous metastases at initial diagnosis of their cervical cancer. The most frequent sites of distant recurrence are lung, extrapelvic nodes, liver, and bone. Although the lumbar spine is said to be a relatively frequent site of skeletal metastases, more recent studies using abdominal imaging demonstrate that most patients with isolated lumbar spine involvement actually have direct extension of disease from paraaortic nodes.

PATHOLOGY

Cervical Intraepithelial Neoplasia

Several systems have been developed to classify cervical cytology. Although criteria for the diagnosis of CIN vary somewhat between pathologists, the important characteristics of this lesion are cellular immaturity, cellular disorganization, nuclear abnormalities, and increased mitotic activity. The degree of neoplasia is determined from the extent of the mitotic activity, immature cell proliferation, and nuclear atypia. If mitoses and immature cells are present only in the lower one third of the epithelium, the lesion usually is designated CIN 1. Involvement of the middle or upper third is diagnosed as CIN 2 or CIN 3, respectively.

The term cervical intraepithelial neoplasia, as proposed by Richart, refers only to a lesion that may progress to invasive carcinoma. Although CIN 1-2 is sometimes referred to as mild to moderate dysplasia, CIN is now preferred over the term dysplasia. Because the word dysplasia means abnormal maturation, proliferating metaplasia without mitotic activity has sometimes been erroneously called dysplasia.

The Bethesda system of classification, designed to further standardize reporting of cervical cytology, was developed following a National Cancer Institute consensus conference in 1988 and was refined in 1991. This system, which separates condylomata and CIN 1, classified as low-grade squamous intraepithelial lesions (LSIL), from high-grade squamous intraepithelial lesions (HSIL), is meant to replace the Papanicolaou system and is now widely used in the United States.

ADENOCARCINOMA IN SITU. The diagnosis of adenocarcinoma in situ (AIS) is made when normal endocervical gland cells are replaced by tall, irregular columnar cells with stratified, hyperchromatic nuclei and increased mitotic activity, but the normal branching pattern of the endocervical glands is maintained and there is no obvious stromal invasion. About 50% of women with cervical AIS also have squamous CIN, and AIS is often an incidental finding in patients operated on for squamous carcinoma.

Colposcopy

Colposcope

Microinvasive Carcinoma

Because the definition of microinvasive carcinoma is based on the maximum depth and linear extent of involvement, it can be diagnosed only from a specimen that includes the entire neoplastic lesion and cervical transformation zone. This requires a cervical cone biopsy.

The earliest invasion appears as a protrusion of cells from the stromoepithelial junction; these cells are better differentiated than the adjacent noninvasive cells and have abundant pink-staining cytoplasm, hyperchromatic nuclei, and small to medium nucleoli. As the tumor progresses, invasion occurs at multiple sites, and its depth and linear extent become measurable. The depth of invasion should be measured with a micrometer from the base of the epithelium to the deepest point of invasion. Although lesions that have invaded less than 3 mm (FIGO stage IA1) rarely metastasize, about 5% of tumors that invade 3 to 5 mm (FIGO stage IA2) have positive pelvic lymph nodes.

Although investigators occasionally label very small adenocarcinomas as microinvasive, the term probably should not be used for these tumors. No definable, consistent method has been found to measure the depth of an invasive adenocarcinoma because it may have originated from the mucosal surface or the periphery of underlying glands. For this reason, adenocarcinomas are generally classified as either AIS or invasive carcinoma (FIGO stage IB).

Invasive Squamous Cell Carcinoma

Between 80% and 90% of cervical carcinomas are squamous. A number of systems have been used to grade and classify squamous carcinomas, but none have been consistently demonstrated to predict prognosis. One of the most commonly used systems categorizes squamous neoplasms as either large cell keratinizing, large cell nonkeratinizing, or small cell carcinoma. The latter should not be (but often is) confused with anaplastic small cell carcinoma, which resembles oat cell carcinoma of the lung because it contains small tumor cells that have scanty cytoplasm, small round to oval nuclei, small or absent nucleoli, coarsely granular chromatin, and high mitotic activity. In contrast, small cell squamous carcinomas have small to medium nuclei, open chromatin, small or large nucleoli, and more abundant cytoplasm. About 30% to 50% of anaplastic small cell carcinomas display neuroendocrine features. Most authorities believe that patients with large cell squamous carcinoma, with or without keratinization, have a better prognosis than those with small cell neoplasms and that small cell anaplastic carcinomas behave more aggressively than poorly differentiated small cell squamous carcinomas.

Invasive Adenocarcinoma

Invasive adenocarcinoma may be pure or mixed with squamous cell carcinoma (adenosquamous carcinoma). A wide variety of cell types, growth patterns, and differentiation have been observed. About 80% of cervical adenocarcinomas are made up predominantly of cells whose differentiated features resemble endocervical glandular epithelium with intracytoplasmic mucin production. The remaining tumors are populated by endometrioid cells, clear cells, intestinal cells, or a mixture of more than one cell type. By histologic examination alone, some of these tumors are indistinguishable from those arising elsewhere in the endometrium or ovary. Minimal deviation adenocarcinoma (adenoma malignum) is an extremely well-differentiated adenocarcinoma in which the branching glandular pattern strongly resembles normal endocervical glands. Because of this, the tumor may not be recognized as malignant in small biopsy specimens and the correct diagnosis may be delayed. Earlier studies reported a dismal outcome for women with this tumor, but more recently, patients have been reported to have a favorable prognosis if the disease is detected early.

Young and Scully have described a villoglandular papillary subtype of adenocarcinoma that primarily affects young women, appears to metastasize infrequently, and has a favorable prognosis. Glucksmann and Cherry first described glassy cell carcinoma, a form of poorly differentiated adenosquamous carcinoma with cells that have abundant eosinophilic, granular, ground-glass cytoplasm, large round to oval nuclei, and prominent nucleoli. Other rare variants of adenosquamous carcinoma include adenoid basal carcinoma and adenoid cystic carcinoma. The former is a well-differentiated tumor that histologically resembles basal cell carcinoma of the skin and tends to have a favorable prognosis. Adenoid cystic carcinomas consist of basaloid cells in a cribriform or cylindromatous pattern and tend to have an aggressive behavior with frequent metastases, although the natural history of these tumors may be long. Whether the prognoses of these rare subtypes are different from other adenocarcinomas of similar grade is uncertain. A variety of neoplasms may infiltrate the cervix from adjacent sites, presenting differential diagnostic problems. In particular, it may be difficult or impossible to determine the origin of adenocarcinomas involving the endocervix and uterine isthmus. Although endometrioid histology suggests endometrial origin and mucinous tumors in young patients are most often of endocervical origin, both histologic types can arise in either site. Metastatic tumors from the colon, breast, or other sites may involve the cervix secondarily. Malignant mixed mullerian tumors, adenosarcomas, and leiomyosarcomas arise occasionally in the cervix, but more often involve it secondarily. Primary lymphomas and melanomas of the cervix are extremely rare.

CLINICAL MANIFESTATIONS

Preinvasive disease is usually detected during routine screening from cervical cytology. Patients with early invasive disease may also be asymptomatic. The first symptom of invasive cervical cancer is usually abnormal vaginal bleeding, often following coitus or vaginal douching. This may be associated with a clear or foul-smelling vaginal discharge. Pelvic pain may result from locoregionally invasive disease or from coexistent pelvic inflammatory disease. Flank pain may be a symptom of hydronephrosis, often complicated by pyelonephritis. The triad of sciatic pain, leg edema, and hydronephrosis is almost always associated with extensive pelvic wall involvement by tumor. Patients with very advanced tumors may have hematuria or incontinence from a vesicovaginal fistula caused by direct extension of tumor to the bladder. External compression of the rectum by a massive primary tumor may cause constipation, but the rectal mucosa is rarely involved at initial diagnosis.

DIAGNOSIS, CLINICAL EVALUATION, AND STAGING

Diagnosis. The long preinvasive stage of cervical cancer, relatively high prevalence of the disease in unscreened populations, and the sensitivity of cytologic screening have made cervical carcinoma an ideal target for cancer screening. In the United States, screening with cervical cytology and pelvic examination has led to a decrease in the mortality rate from cervical cancer of more than 70% since 1940. Only nations with comprehensive screening programs have experienced substantial decreases in cervical cancer death rates during this period. Authorities disagree about the optimal frequency of cervical cancer screening. In a 1988 consensus statement, the American Cancer Society and other medical groups recommended annual Pap smears beginning at age 18 years or with the onset of sexual activity and added that, after three or more consecutive normal annual examinations, the cytologic evaluation could be performed less frequently at the discretion of the physician. For patients who have had repeated negative tests, the marginal gain from screening more often than every 3 years decreases sharply. The US Preventive Services Task Force has recommended that screening be discontinued after age 65 years if results have been consistently normal, and the Canadian Task Force suggests that the interval be extended to 5 years after age 35 if previous studies have been normal. Although these groups have suggested tailoring the frequency of Pap smears to patient risk, practical definitions of low and high risk remain controversial. As a result, most clinicians continue to recommend that their patients be screened more frequently than recommended by the national guidelines. The false-negative rate of the Pap smear is about 10% to 15% in women with invasive cancer. The sensitivity of the test may be improved by ensuring adequate sampling of the squamocolumnar junction and the endocervical canal; smears without endocervical or metaplastic cells are inadequate and must be repeated. Because AIS originates near or above the transformation zone, it may not be sampled with conventional cervical smears. Detection of high endocervical lesions may be improved when specimens are obtained with a cytobrush. Also, because hemorrhage, necrosis, and intense inflammation may obscure the results, the Pap smear is a poor way to diagnose gross lesions; these should always be biopsied. Patients with abnormal cytology who do not have a gross cervical lesion must be evaluated by colposcopy and directed biopsies. Following application of a 3% acetic acid solution, the cervix is examined under 10- to 15-fold magnification with a bright, filtered light that enhances the acetowhitening and vascular patterns characteristic of dysplasia or carcinoma. The skilled colposcopist can accurately distinguish between low- and high-grade dysplasia, but microinvasive disease cannot consistently be distinguished from intraepithelial lesions. If no abnormalities are found on colposcopic examination or if the entire squamocolumnar junction cannot be visualized in a patient with an atypical Pap smear, endocervical curettage should be performed. Some authorities advocate the routine addition of endocervical curettage to colposcopic examination to minimize the risk of missing occult cancer within the endocervical canal. However, it is probably reasonable to omit this step in previously untreated women if the entire squamocolumnar junction is visible with a complete ring of unaltered columnar epithelium in the lower canal. Cervical cone biopsy is used to diagnose occult endocervical lesions and is an essential step in the diagnosis and management of microinvasive carcinoma of the cervix. The geometry of the cone is individualized and tailored to the geometry of the cervix, the location of the squamocolumnar junction, and the site and size of the lesion. Cervical cone biopsy yields an accurate diagnosis and decreases the incidence of inappropriate therapy when (1) the squamocolumnar junction is poorly visualized on colposcopy and a high-grade lesion is suspected; (2) a high-grade dysplastic epithelium extends into the endocervical canal; (3) the cytology suggests a high-grade dysplasia or carcinoma in situ; (4) a microinvasive carcinoma is found on directed biopsy; (5) the endocervical curettings show high-grade CIN; or (6) the cytology is suspicious for AIS.

Clinical Evaluation of Patients With Invasive Carcinoma

All patients with invasive cervical cancer should be evaluated with a detailed history and physical examination, paying particular attention to inspection and palpation of the pelvic organs with bimanual and rectovaginal examinations. Standard laboratory studies should include complete blood count and renal function and liver function tests. All patients should have a chest radiograph to rule out lung metastases and an intravenous pyelogram to determine the kidneys' location and to rule out ureteral obstruction by tumor. Cystoscopy and proctoscopy or barium enema should be obtained in patients with bulky tumors. Many clinicians obtain computed tomography (CT) or magnetic resonance imaging (MRI) scans to evaluate regional nodes, but the accuracy of these studies is compromised by their failure to detect small metastases and because patients with bulky necrotic tumors often have enlarged reactive lymph nodes.

In a large Gynecologic Oncology Group study that compared the results of radiographic studies with subsequent histologic findings, Heller and colleagues found that 79% of the cases with paraaortic lymph node involvement were detected by lymphangiography whereas only 34% were detected by CT. MRI can provide useful information about the location and depth of invasion of tumors in the cervix, but gives less accurate assessments of parametrial involvement. Clinical Staging The International Federation of Gynecology and Obstetrics (FIGO) has defined the most widely accepted staging system for carcinomas of the cervix. The latest (1994) update of this system is summarized in Table. Since the earliest versions of the cervical cancer staging system, there have been numerous changes, particularly in the definition of stage I disease. Preinvasive disease was not placed in a separate category until 1950, and the stage IA category for "cases with early stromal invasion" was first described in 1962. Cases of early stromal invasion and occult invasion were redistributed between stages IA(i), IA(ii), and IB(occult) several times until 1985, when FIGO eliminated stage IB(occult and provided the first specific definitions of microinvasive disease (stages IA1 and IA2). In 1994, these definitions were changed again and, for the first time, stage IB tumors were subdivided according to tumor diameter. Although these changes have gradually improved the discriminatory value of the staging system, the many fluctuations in the definitions of stage IA and IB have complicated our ability to compare the outcomes of patients whose tumors were staged and treated during these periods. In addition, gynecologic oncologists in the United States have for many years stag disease using the Society of Gynecologic Oncologists' definition of a microinvasive carcinoma, that is, tumor that "invades the stroma in one or more places to a depth of 3 mm or less below the base of the epithelium and in which lymphatic or vascular involvement is not demonstrated," a definition that still differs from the current FIGO classification. FIGO stage is based on careful clinical examination and the results of specific radiologic studies and procedures. These should be performed and the stage should be assigned before any definitive therapy is administered. The clinical stage should never be changed on the basis of subsequent findings. When it is doubtful to which stage a particular case should be allotted, the case should be assigned to the earlier stage. FIGO warns that, because it is impossible to tell from clinical examination whether a smooth and indurated parametrium is truly cancerous or only inflammatory, a case should be classified as stage III only if the parametrium is nodular out on the pelvic wall or if the growth itself extends out on the pelvic wall. In the rules for clinical staging, FIGO states that palpation, inspection, colposcopy, endocervical curettage, hysteroscopy, cystoscopy, proctoscopy, intravenous urography, and radiologic examination of the lungs and skeleton may be used for clinical staging. Suspected bladder or rectal involvement should be confirmed by biopsy. Findings of bullous edema or malignant cells in cytologic washings from the urinary bladder are not sufficient to diagnose bladder involvement. FIGO clearly states that findings by examinations such as lymphangiography, arteriography, venography, and laparoscopy are of value of planning therapy but, because these are not yet generally available and the interpretation of results is variable, the findings of such studies should not be the basis for changing the clinical stage. Examination under anesthesia is desirable but not required. The rules and notes to the FIGO staging system are integral parts of the clinical staging system and should be strictly observed to minimize inconsistencies in staging between institutions. Although most clinicians use the FIGO classification system, a number of European groups use a staging system that divides stage IIB tumors according to the extent of parametrial involvement and divides stage III tumors according to whether there is unilateral or bilateral pelvic wall fixation. Until the mid-1980s, most reports from the University of Texas M.D. Anderson Cancer Center used a similar staging system that also categorized patients with bulky endocervical tumors in a special category. Although surgically treated patients are sometimes classified according to a TNM pathologic staging system, this has not been widely accepted because it cannot be applied to patients who are treated with primary radiotherapy. Surgical Evaluation of Regional Spread In the 1970s, studies of diagnostic preradiation lymph node dissection used a transperitoneal approach that led to unacceptable morbidity and mortality rates from radiation-related bowel complications, particularly after treatment with high radiation doses and extended fields. More recently, extraperitoneal dissection, which induces fewer bowel adhesions, has been recommended. With this approach, postradiotherapy small bowel complications occur in fewer than 5% of patients. Laparoscopic node dissection is advocated by some surgeons but remains investigational. The late complication rate of radiotherapy following such procedures has not yet been defined. Although the indications for surgical staging are controversial, advocates argue that the procedure identifies patients with microscopic paraaortic or common iliac node involvement who can benefit from extended-field irradiation. Some investigators have also suggested, on the basis of first principles and encouraging pelvic control rates, that debulking of large pelvic nodes before radiotherapy may improve outcome. Because patients with radiographically positive pelvic nodes are at greatest risk for occult metastasis to paraaortic nodes, they may have the greatest chance of benefiting from surgical staging. Some have advocated pretreatment blind biopsy of the scalene node in patients with positive paraaortic nodes and in patients with a central recurrence who are being considered for exenteration. The reported incidence of supraclavicular metastasis varies widely (5% to 20% or more) for patients with positive paraaortic lymph nodes.

PROGNOSTIC FACTORS

Although the survival and pelvic disease control rates of cervical cancer patients are correlated with FIGO stage, prognosis is also influenced by a number of tumor characteristics that are not included in the staging system. Clinical tumor diameter is strongly correlated with prognosis for patients treated with radiation or surgery. For this reason, FIGO recently modified the stage I category to subdivide tumors according to clinical tumor diameter (i.e., 4 cm or less or more than 4 cm). For patients with more advanced disease, other estimates of tumor bulk, such as the presence of medial versus lateral parametrial involvement in FIGO stage IIB tumors or of unilateral versus bilateral parametrial or pelvic wall involvement, have also been correlated with outcome. The predictive value of the staging system itself may, in part, reflect an association between the stage categories and the primary tumor volume. Operative findings often do not agree with clinical estimates of parametrial or pelvic wall involvement, and some authors have found that the predictive power of stage diminishes or is lost when comparisons are corrected for differences in clinical tumor diameter. Lymph node metastasis is also an important predictor of prognosis. For patients treated with radical hysterectomy for stage IB disease, survival rates are usually reported as 85% to 95% for patients with negative nodes and 45% to 55% for those with lymph node metastases. Inoue and Morita reported that survival was correlated with the size of the largest node, and several authors have reported correlations between the number of involved pelvic lymph nodes and survival. Survival rates for patients with positive paraaortic nodes treated with extended-field radiation therapy vary between 10% and 50% depending on the extent of pelvic disease and paraaortic lymph node involvement. For patients treated with radical hysterectomy, several histologic parameters have been associated with a poor prognosis. Lymph-vascular space invasion, deep stromal invasion (10 mm or more or more than 70% invasion), and microscopic evidence of parametrial extension are strongly correlated with lymph node involvement and recurrence, whereas the presence of a strong inflammatory response in the cervical stroma predicts a good outcome. Uterine body involvement is associated with an increased rate of distant metastases in patients treated with radiation or surgery. Although several investigators have reported similar survival rates for patients with squamous carcinomas or adenocarcinomas, others have drawn the opposite conclusion, noting unusually high pelvic relapse rates in patients treated surgically for adenocarcinomas and poorer survival rates in patients who undergo surgery or irradiation. In a multivariate analysis of 1767 patients treated with radiation for FIGO stage IB disease, Eifel and colleagues reported a highly significant independent correlation between histology and survival. Using Cox regression analysis, the relative risk of death from cancer for 106 patients with adenocarcinomas 4 cm or more in diameter was 1.9 times that for patients with squamous tumors (P < 0.01). Pelvic disease control rates were similar for patients with squamous or adenocarcinomas, but there was a significantly higher incidence of distant metastases in patients with adenocarcinomas. Although the prognostic significance of histologic grade has been disputed for squamous carcinomas, there is a clear correlation between the degree of differentiation and the clinical behavior of adenocarcinomas. Several studies have demonstrated a relationship between hemoglobin level and prognosis of patients with locally advanced cervical cancer. The strongest evidence that anemia plays a causative role in pelvic recurrence comes from a small 1978 randomized study conducted at the Princess Margaret Hospital. All patients were maintained at a hemoglobin level of at least 10 gm%, but those in the treatment arm were maintained through transfusions at hemoglobin levels of 12.5 gm% or more. The 25 anemic patients in the control arm had a significantly higher locoregional recurrence rate than patients in the study who received transfusions. Unfortunately, the results of this small study have never been confirmed, and subsequent studies aimed at overcoming the theoretical radiobiologic consequences of intratumoral hypoxia (hypoxic cell sensitizers, hyperbaric oxygen breathing, neutron therapy) have not been successful. Hockel and colleagues reported a relationship between intratumoral pO(2) measurements and the outcome of patients with cervical cancer, but larger studies will be needed to confirm the prognostic value of intratumoral pO(2).

In a study of sera obtained before treatment from 587 patients with cervical cancers, Duk and colleagues reported a strong correlation between the concentration of squamous cell carcinoma antigen and the stage and size of the tumor as well as the presence of lymph node metastases; multivariate analysis also showed that serum squamous cell carcinoma antigen was an independent predictor of prognosis in their study.

Other clinical and biologic features that have been investigated for their predictive power with variable results include patient age, platelet count, tumor vascularity, DNA ploidy or S phase, and HPV infection. In a preliminary study of archival material from 21 patients with histologically negative lymph node dissections, Ikenberg and colleagues recently reported a higher rate of disease recurrence when a polymerase chain reaction assay of the lymph nodes was strongly positive for HPV-16 DNA.

TREATMENT

A number of factors may influence the choice of treatment, including tumor size, stage, and histology; evidence of lymph node involvement; risk factors for surgery or radiation; and patient preference. However, as a rule, intraepithelial lesions are treated with superficial ablative techniques, microinvasive cervical cancers invading less than 3 mm (stage IA1) are managed with conservative surgery (excisional conization or extrafascial hysterectomy), early invasive cancers (stages IA2 and IB1 and some small stage IIA tumors) are managed with either radical surgery or irradiation, and locally advanced cancers (stages IB2 through IVA) are managed with radiation therapy. Selected patients with centrally recurrent disease after maximum radiation therapy may be treated with radical exenterative surgery; pelvic recurrence after hysterectomy is treated with irradiation.

Preinvasive Disease (Stage 0)

Patients with noninvasive squamous lesions can be treated with superficial ablative therapy (cryosurgery or laser therapy) or with loop excision if (1) the entire transformation zone has been visualized colposcopically, (2) directed biopsies are consistent with Pap smear results, (3) endocervical curettage findings are negative, and (4) there is no cytologic or colposcopic suspicion of occult invasion. If patients do not meet these criteria, a conization should be performed.

With cryotherapy, abnormal tissue is frozen with a supercooled metal probe until an ice ball forms that extends 5 mm beyond the lesion. Because cryonecrosis tends to be patchy and may be inadequate after a single freeze, the tissue should be frozen a second time after it has visibly thawed. Another common and equally effective technique ablates tissue with a carbon dioxide laser beam. After laser ablation, there is less distortion and more rapid healing of the cervix, but the procedure requires more training and more expensive equipment than cryosurgery.

Many practitioners now consider loop diathermy excision to be the preferred treatment. With this technique, a charged electrode is used to excise the entire transformation zone and distal canal. Loop diathermy is easily learned, less expensive than laser excision, and preserves the excised lesion and transformation zone for histologic evaluation. However, some authorities think that low-grade lesions may be overtreated with this method. Loop excision should not be considered an alternative to formal excisional conization when microinvasive or invasive cancer is suspected or for patients with AIS, because it may inadequately treat disease within the cervical canal and complicate further treatment. Cryotherapy, laser excision, and loop excision are all outpatient office procedures that maintain fertility. Although recurrence rates are low (10% to 15%) and progression to invasion rare (less than 2% in most series), lifelong surveillance of these patients must be maintained. Vaginal or type I abdominal hysterectomy currently is reserved for women who have other gynecologic conditions that justify the procedure; invasive cancer still must be excluded before surgery to rule out the need for a more extensive operative procedure.

Microinvasive Carcinoma (Stage IA)

The standard treatment for patients with stage IA1 disease is total (type I) or vaginal hysterectomy. Because the risk of pelvic lymph node metastases from these minimally invasive tumors is less than 1%, pelvic lymph node dissection is not usually recommended. Selected patients with tumors that meet the Society of Gynecologic Oncologists' definition of microinvasion (FIGO stage IA1 disease without lymph-vascular space invasion) and who wish to maintain fertility may be adequately treated with a therapeutic cervical conization if the margins of the cone are negative. In 1991, Burghardt and colleagues reported one recurrence (which was fatal) in 93 women followed for more than 5 years after therapeutic conization for minimal (less than 1 mm) microinvasion. Morris and colleagues reported no invasive recurrences in 14 patients followed for a mean of 26 months after conization for tumors invading 0.5 to 2.8 mm. However, patients who have this conservative treatment must be followed closely with periodic cytology, colposcopy, and endocervical curettage.

Diagnostic or therapeutic conization for microinvasive disease is usually performed with a cold knife or carbon dioxide laser on a patient under general or spinal anesthesia. Because an accurate assessment of the maximum depth of invasion is critical, the entire specimen must be sectioned and carefully handled to maintain its original orientation for microscopic assessment. Complications occur in 2% to 12% of patients, are related to the depth of the cone, and include hemorrhage, sepsis, infertility, stenosis, and cervical incompetence. The width and depth of the cone should be tailored to produce the least amount of injury while providing clear surgical margins.

For patients with 3 to 5 mm of stromal invasion (FIGO stage IA2), the risk of nodal metastases is approximately 5%. Therefore, a bilateral pelvic lymphadenectomy should be performed in conjunction with a modified radical (type II) hysterectomy. Modified radical hysterectomy is a less extensive procedure than a classic radical hysterectomy. The cervix, upper vagina, and paracervical tissues are removed after careful dissection of the ureters to the point of their entry to the bladder. The medial one half of the cardinal ligaments and the uterosacral ligaments are also removed. With this treatment, significant urinary tract complications are rare and cure rates exceed 95%. Although surgical treatment is standard for in situ and microinvasive cancer, patients with severe medical problems or other contraindications to surgical treatment can be successfully treated with radiation therapy. Grigsby and Perez reported a 10-year progression-free survival rate of 100% in 21 patients with carcinoma in situ and in 34 patients with microinvasive carcinoma treated with radiation alone. Hamberger and colleagues reported that all patients with stage IA disease and 89 (96%) of 93 patients with small stage IB disease (less than one cervical quadrant involved) were disease free 5 years after treatment with intracavitary irradiation alone.

Stages IB and IIA

Early stage IB cervical carcinomas can be treated effectively with combined external-beam irradiation and brachytherapy or with radical hysterectomy and bilateral pelvic lymphadenectomy. The goal of both treatments is to destroy malignant cells in the cervix, paracervical tissues, and regional lymph nodes.

Overall survival rates for patients with stage IB cervical cancer treated with surgery or radiation usually range between 80% and 90%, suggesting that the two treatments are equally effective. However, biases introduced by patient selection, inconsistencies in the definition of FIGO stage IB disease, and variable indications for postoperative radiotherapy or adjuvant hysterectomy confound comparisons about the efficacy of radiotherapy versus surgery.

In a 1976 review of 321 patients, Morley and Seski reported similar 5-year survival rates of 91.3% and 87.3% for patients treated with surgery or radiotherapy, respectively. Though treatment was assigned alternately for most patients in this series, the study was not truly randomized; exclusion of some patients found to have unfavorable findings at surgery and deviations from the alternating scheme could have led to biased results. In another review of their experience, Hopkins and Morley noted that a significant difference in survival favoring surgical treatment disappeared when the authors excluded from the radiation group patients who were selected for radiation treatment only after radical hysterectomy was aborted because of intraoperative findings of extrauterine disease. Because young women with small, clinically node-negative tumors tend to be favored candidates for surgery and because tumor diameter and nodal status are inconsistently described in published series, it is difficult to compare the results reported for patients treated with the two modalities.

Preliminary results of the first prospective trial randomizing patients with stage IB or IIA cervical cancer to radical surgery or radical radiotherapy were recently reported. In the surgical arm, findings of parametrial involvement, positive margins, deep stromal invasion, or positive nodes led to the use of postoperative pelvic irradiation in 62 (54%) of 114 patients with tumors 4 cm or smaller in diameter and in 46 (84%) of 55 patients with tumors greater than 4 cm in diameter. Patients in the radiotherapy arm received a relatively low dose of radiation to the cervix with a median dose to point A of 72 Gy. The authors reported similar survival rates for the two treatments, but the frequent use of combined modality treatment led to a significantly higher rate of complications in patients treated with initial surgery.

For patients with stage IB1 squamous carcinomas, the choice of treatment is based primarily on patient preference, anesthetic and surgical risks, physician preference, and an understanding of the nature and incidence of complications with the two treatment approaches (described in detail later). The overall rate of major complications is similar for patients with comparable tumors treated with surgery or radiotherapy, although urinary tract complications tend to be more frequent after surgical treatment, and bowel complications are more common after radiation therapy. Surgical treatment tends to be preferred for young women with small tumors because it permits preservation of ovarian function and may cause less vaginal shortening. Radiation therapy is often selected for older, postmenopausal women to avoid the morbidity of a major surgical procedure.

Some surgeons have also advocated the use of radical hysterectomy as initial treatment for patients with larger (stage IB2) tumors. However, patients who have tumors measuring more than 4 cm in diameter usually have deep stromal invasion and are at high risk for lymph node involvement and parametrial extension. Because patients with these risk factors have an increased rate of pelvic disease recurrence, surgical treatment is usually followed by postoperative irradiation, exposing the patient to the risks of both treatments. Consequently, many gynecologic and radiation oncologists believe that patients with bulky (stage IB2) carcinomas are better treated with radical radiotherapy.

RADICAL HYSTERECTOMY. The standard surgical treatment for stages IB and IIA cervical carcinomas is radical (type III) hysterectomy and bilateral pelvic lymph node dissection. This procedure involves en bloc removal of the uterus, cervix, and paracervical, parametrial, and paravaginal tissues to the pelvic sidewalls bilaterally, taking as much of the uterosacral ligaments as possible.

The uterine vessels are ligated at their origin, and the proximal one third of the vagina and paracolpos are resected. For women younger than 40 to 45 years, the ovaries usually are not removed. If intraoperative findings suggest a need for postoperative pelvic irradiation, the ovaries may be transposed out of the pelvis. Intraoperative and immediate postoperative complications of radical hysterectomy include blood loss (average 0.8 L), ureterovaginal fistula (1% to 2%), vesicovaginal fistula (less than 1%), pulmonary embolus (1% to 2%), small bowel obstruction (1% to 2%), and postoperative fever secondary to deep vein thrombosis, pulmonary infection, pelvic cellulitis, urinary tract infection, or wound infection (25% to 50%). Subacute complications include lymphocyst formation and lower extremity edema, which occurs with a risk related to the extent of the node dissection. Lymphocysts may obstruct a ureter, but hydronephrosis usually improves with drainage of the lymphocyst. Complications risks may be increased in patients who receive preoperative or postoperative irradiation.

Although most patients have transient decreased bladder sensation after radical hysterectomy, with appropriate management severe long-term bladder complications are infrequent. However, chronic bladder hypotonia or atony occur in approximately 3% to 5% of patients, despite careful postoperative bladder drainage. Bladder atonia probably results from damage to the bladder's innervation and may be related to the extent of the parametrial and paravaginal dissection. Radical hysterectomy may be complicated by stress incontinence, but reported incidences vary widely. Patients may also experience constipation and, rarely, chronic obstipation after radical hysterectomy.

RADIATION THERAPY AFTER RADICAL HYSTERECTOMY.

The role of postoperative irradiation in patients with cervical carcinoma has not yet been clearly established. Most investigators have reported that postoperative irradiation decreases the risk of pelvic recurrence in patients whose tumors have high-risk features (lymph node metastasis, deep stromal invasion, insecure operative margins, or parametrial involvement). However, because the patients who received postoperative radiotherapy in these studies were selected for the high-risk features of their tumors, it is difficult to determine the impact of adjuvant irradiation on survival.

Stages IB and IIA (Continued). In 1989, Kinney and colleagues retrospectively compared the outcome of 60 patients who had postoperative irradiation with 60 unirradiated patients who were matched for stage (stage IB versus IIA), tumor size, and number and site of positive nodes. There were fewer isolated pelvic failures in the irradiated group, but there was no significant difference in survival. However, even this study did not characterize patients in terms of all the risk factors that can influence the choice of treatment. Some authors have hypothesized that the dose of radiation that can be given safely postoperatively may be inadequate to control microscopic disease in a surgically disturbed, hypovascular site. If this is true, it would be an argument for primary radiotherapeutic management of tumors with known high-risk features.

The overall risk of major complications (particularly small bowel obstructions) is probably increased in patients who receive postoperative pelvic irradiation, but inconsistencies in the methods of analysis and the relatively small number of patients in most series make studies of this subject difficult to interpret. Montz and colleagues reported a 20% risk of small bowel obstruction requiring surgery in 20 patients treated with postoperative irradiation compared with 3 (5%) of 60 patients treated with hysterectomy alone. Bandy and colleagues reported that patients who were irradiated after hysterectomy also had more long-term problems with bladder contraction and instability than those treated with surgery alone.

RADICAL RADIATION THERAPY. Radiation therapy also achieves excellent survival and pelvic control rates in patients with stage IB cervical cancers. Eifel and colleagues reported a 5-year disease-specific survival rate of 90% for 701 patients treated with radiation alone for stage IB1 squamous tumors less than 4 cm in diameter. The central and pelvic tumor control rates were 99% and 98%, respectively. Disease-specific survival rates were 86% and 67% for patients with tumors measuring 4 to 4.9 cm or 5 cm or more in diameter, respectively. Pelvic tumor control was achieved in 82% of patients with tumors 5 cm or more in diameter. Perez and colleagues and Lowrey and colleagues reported similar excellent disease control rates for patients with stage IB tumors. Survival rates for patients with FIGO stage IIA disease treated with irradiation range between 70% and 85% and are also strongly correlated with tumor size.

As with radical surgery, the goal of radiation treatment is to sterilize disease in the cervix, paracervical tissues, and regional lymph nodes in the pelvis. Patients are usually treated with a combination of external-beam irradiation to the pelvis and brachytherapy. Clinicians balance external and intracavitary treatment in different ways for these patients, weighting one or the other component more heavily. However, brachytherapy is a critical element in the curative radiation treatment of all carcinomas of the cervix. Even relatively small tumors that involve multiple quadrants of the cervix are usually treated with total doses of 80 to 85 Gy to point A. This dose may be reduced by 5% to 10% for very small superficial tumors. Although patients with small tumors may be treated with somewhat smaller fields than patients with more advanced locoregional disease, care must still be taken to cover adequately the obturator, external iliac, low common iliac, and presacral nodes. Radiation technique and potential complications are discussed in more detail later.

IRRADIATION FOLLOWED BY HYSTERECTOMY.

In a 1969 report from M. D. Anderson Cancer Center, Durrance and colleagues reported a lower pelvic recurrence rate for patients with bulky endocervical tumors (6 cm or larger) treated with external-beam and intracavitary irradiation followed by extrafascial hysterectomy than for those treated with radiation alone. Combined treatment was subsequently adopted by many groups as a standard approach to bulky stage IB or IIA disease. However, in a 1992 update of the M. D. Anderson experience, Thoms and colleagues suggested that the differences observed in earlier reports may have resulted from a tendency to select patients with very massive tumors (8 cm or larger) or clinically positive nodes for treatment with radiation alone. When these patients were excluded, pelvic control rates were similar with the two approaches.

In 1991, Mendenhall and colleagues reported no difference in pelvic disease control or survival rates for patients treated before or after the University of Florida adopted a policy (in the mid-1970s) of using combined treatment for patients with bulky (6 cm or more) tumors. In a study of 1526 patients with stage IB squamous carcinomas, Eifel and colleagues reported central tumor recurrence rates of less than 10% for tumors as large as 7 to 7.9 cm treated with radiation alone, suggesting that the margin for possible improvement with adjuvant hysterectomy is small. Perez and Kao also found that central recurrences were rare if adequate doses of irradiation (more than 80 Gy to point A) were delivered.

There is, therefore, no clear evidence that adjuvant hysterectomy improves the outcome of patients with a bulky stage IB or IIA tumor, though many clinicians continue to advocate combined treatment. When combined treatment is planned, the dose of intracavitary irradiation is usually reduced by about 25%. A type I extrafascial hysterectomy is usually performed removing the cervix, adjacent tissues, and a small cuff of the upper vagina in a plane outside the pubocervical fascia. This procedure involves minimal disturbance of the bladder and ureters. Intrafascial hysterectomy is not used for cervical cancer because it does not remove all cervical tissue, and radical hysterectomy is avoided after high-dose irradiation because of an increased risk of urinary tract complications.

The Gynecologic Oncology Group recently completed a prospective randomized trial of irradiation with or without extrafascial hysterectomy in patients with stage IB tumors of 4 cm or more in diameter. This study should be ready for final analysis soon.

CHEMOTHERAPY FOLLOWED BY RADICAL SURGERY. Since the late 1980s, a number of investigators have reported the results of treating patients with bulky stages IB and II cervical carcinomas with a combination of neoadjuvant chemotherapy followed by radical surgery. Neoadjuvant chemotherapy has usually included cisplatin and bleomycin plus one or two other drugs. The results of these uncontrolled studies cannot be easily compared with more traditional treatments because the series are small, often have short follow-up, and the criteria for patient selection are not always clear. Some or all of the patients in each of these series received postoperative pelvic irradiation, but detailed descriptions of this additional treatment are not always given. There has been one report of a prospective randomized trial comparing radical hysterectomy followed by postoperative radiotherapy with chemotherapy followed by surgery and irradiation. In this study, Sardi and colleagues observed similar outcomes with the two treatments for patients who had tumors smaller than 60 cm3 (measured ultrasonographically), but they reported a significantly better projected 4-year disease-free survival with neoadjuvant chemotherapy for patients who had larger tumors. However, most patients had been followed for less than 3 years at the time of the report. Ultimately, the cost and morbidity of this triple-modality treatment may be justified only if it proves to be more effective than treatment with radiotherapy alone. However, studies comparing this approach with radical irradiation have not yet been reported.

Stages IIB, III, and IVA

Radiation therapy is the primary treatment for most patients with locoregionally advanced cervical carcinoma. The success of treatment depends on a careful balance between external-beam radiation therapy and brachytherapy, optimizing the dose to tumor and normal tissues, and the overall duration of treatment. Five-year survival rates of 65% to 75%, 35% to 50%, and 15% to 20% are reported for patients treated with radiotherapy alone for stages IIB, IIIB, and IVA tumors, respectively. In a French Cooperative Group study of 1383 patients treated with radiotherapy according to Fletcher guidelines, Horiot and colleagues reported 5-year survival rates of 76%, 50%, and 20.5% for patients with stages IIB, IIIB, and IVA tumors, respectively. With appropriate radiotherapy, even patients with massive locoregional disease have a significant chance for cure.

External-beam irradiation is used to deliver a homogeneous dose to the primary cervical tumor and to potential sites of regional spread. An initial course of external irradiation may also improve the efficacy of subsequent intracavitary treatment by shrinking bulky endocervical tumor (bringing it within the range of the high-dose portion of the brachytherapy dose distribution) and by shrinking exophytic tumor that might prevent satisfactory placement of vaginal applicators. For this reason, patients with locally advanced disease usually begin with a course of external-beam treatment. Subsequent brachytherapy exploits the inverse square law to deliver a high dose to the cervix and paracervical tissues while minimizing the dose to adjacent normal tissues. Although many clinicians delay intracavitary treatment until pelvic irradiation has caused some initial tumor regression, breaks between external-beam and intracavitary therapy should be discouraged, and every effort should be made to complete the entire treatment in less than 7 to 8 weeks. The favorable results documented in reports from large single institutions have been based on policies that dictate relatively short overall treatment durations (less than 8 weeks), and several studies in patients with locally advanced cervical cancer have suggested that longer treatment courses are associated with decreased pelvic control and survival rates.

EXTERNAL-BEAM TECHNIQUE. High-energy photons (15 to 18 MV) are usually preferred for pelvic treatment because they spare superficial tissues that are unlikely to be involved with tumor. At these energies, the pelvis can be treated either with four fields (anterior, posterior, and lateral fields) or with anterior and posterior fields alone. When high-energy beams are not available, four fields are usually used because less penetrating 4 to 6 MV photons often deliver an unacceptably high dose to superficial tissues when only two fields are treated. However, lateral fields must be designed with great care because clinicians' estimates of the location of potential sites of disease on a lateral radiograph may be inaccurate. In particular, "standard" anterior and posterior borders that have been described in the past may shield regions at risk for microscopic regional disease in the presacral and external iliac nodes and in the presacral and cardinal ligaments; care must also be taken not to underestimate the posterior extent of central cervical disease in patients with bulky tumors. The caudad extent of disease can be determined by placing radiopaque seeds in the cervix or at the lowest extent of vaginal disease. Information gained from radiologic studies can also improve estimates of disease extent. Lymphangiograms are helpful in tailoring blocks, particularly at the anterior border of lateral fields. MRI and CT scans can also help clinicians to design lateral field borders with an improved understanding of uterine position. In fact, some investigators have argued that these studies should be obtained routinely for patients with bulky disease to avoid errors in lateral field design. However, when all these factors are considered, differences in the volume treated with a four-field or a high-energy two-field technique may be small. For this reason, some clinicians prefer to use the simpler technique for patients with bulky tumors.

Tumor response should be evaluated with periodic pelvic examinations to determine the best time to deliver brachytherapy treatment. Some practitioners prefer to maximize the brachytherapy component of treatment and begin as soon as the tumor has responded enough to permit a good placement (with very bulky tumors this may still require 40 Gy or more). Subsequent pelvic irradiation is delivered with a central block. A somewhat higher total paracentral dose can be delivered with this approach, but greater reliance is placed on the complex match between the brachytherapy dose distribution and the border of the central shield. Other clinicians prefer to give an initial dose of 40 to 45 Gy to the whole pelvis, believing that the ability to deliver a homogeneous distribution to the entire region at risk for microscopic disease and the additional tumor shrinkage achieved before brachytherapy outweigh other considerations. In fact, both approaches have been in use for several decades and, when optimally used, appear to give excellent tumor control rates with acceptable complication rates.

THE ROLE OF PARAAORTIC IRRADIATION. The role of extended-field irradiation in the treatment of cervix cancer is still being defined. Numerous small series of patients with documented paraaortic node involvement demonstrate that some enjoy long-term survival after radiation therapy. Patients with microscopic involvement have a better survival than those with gross lymphadenopathy, but even 10% to 15% of the latter appear to be curable with aggressive management. Survival is also strongly correlated with the bulk of central disease. A 1991 study by Cunningham and colleagues reported a 48% 5-year survival rate in patients who had paraaortic node involvement discovered at exploration for radical hysterectomy (which was then aborted). This experience with patients who had small, radiocontrollable primary disease demonstrates that patients with paraaortic node metastases can often be cured if their primary disease can be sterilized. This indicates that patients may have extensive regional spread without distant metastases and provides an argument for surgical staging in high-risk patients.

Two randomized prospective trials have addressed the role of prophylactic paraaortic irradiation in patients without known paraaortic node involvement. In a recently updated study conducted by the Radiation Therapy Oncology Group (RTOG), 367 patients with primary stage IIB or stages IB and IIA tumors more than 4 cm in diameter were randomized to receive either standard pelvic radiotherapy or extended-field radiotherapy before brachytherapy. No consistent method was used to evaluate the paraaortic nodes. For the 337 evaluable patients, absolute survival was significantly better for those treated with extended fields compared with those treated with standard pelvic radiotherapy (67% versus 55% at 5 years; P =0.02). There was no significant difference in disease-free survival (P = 0.56). The most marked difference in sites of initial disease failure was a higher rate of concurrent locoregional and distant failure in the pelvic radiotherapy group. Recurrences in the paraaortic nodes were scored as distant failures, so it is difficult to explain the higher overall rate of locoregional recurrence in the control group.

Stages IIB, III, and IVA (Continued)

A second trial from the European Organization for Research and Treatment of Cancer (EORTC) involved a similar randomization between pelvic irradiation and extended fields, but had very different eligibility criteria. This study included patients with bulky stage IIB (involving distal vagina or lateral parametrium) and III disease and patients with stage I and less bulky stage IIB disease who had positive pelvic nodes on lymphangiogram or at surgery. The 4-year disease-free survival rates of 49.8% and 53.3% for patients treated with pelvic or extended fields, respectively, were not significantly different. However, the rate of paraaortic node recurrence was significantly higher in the pelvic group, and for patients in whom local control was achieved, the rate of distant metastases was 2.8 times greater if treatment was with pelvic irradiation only (P <0.01).

Both studies revealed an increased rate of enteric complications in patients treated with extended fields. In the ROTG study, most small bowel obstructions occurred in patients who had undergone pretreatment transperitoneal staging. The EORTC did not mention a relationship between surgical staging and enteric complications.

Although the morbidity of extended-field irradiation is no longer prohibitive with multiple field techniques, more modest radiation doses, and extraperitoneal staging, it is still greater than that of standard field radiotherapy. Further definition of patients most likely to benefit from prophylactic paraaortic irradiation would improve the therapeutic ratio of such treatment.

BRACHYTHERAPY TECHNIQUE. Fletcher described the following three conditions for successful cervical brachytherapy: (1) the geometry of the radioactive sources must prevent underdosed regions on and around the cervix, (2) an adequate dose must be delivered to the paracervical areas, and (3) mucosal tolerance must be respected. These factors dictate the character, intensity, and timing of brachytherapy.

Brachytherapy is usually delivered using afterloading applicators that are placed in the uterine cavity and vagina. A number of different intracavitary systems have been used; in the United States, variations of the Fletcher-Suit-Delclos system are used most commonly. The intrauterine tandem and vaginal applicators are carefully positioned, usually with the patient under anesthesia, to provide an optimal relationship between the system and adjacent tumor and normal tissues. Vaginal packing is used to hold the tandem and colpostats in place and to maximize the distance between the sources and adjacent bladder and rectum. Radiographs should be taken at the time of insertion to verify accurate placement, and the system should be repositioned if it can be improved. Encapsulated radioactive sources are inserted in the applicators after the patient has returned to her hospital bed, reducing exposure to personnel during applicator placement. Remote afterloading devices that further reduce personnel exposure are often used in departments that treat many gynecologic patients. Although **226Ra was used to treat most patients before the 1980s, it has gradually been replaced by *137Cs, which produces a similar dose distribution and avoids the radiation protection problems caused by the radon gas byproduct of radium decay.

BRACHYTHERAPY DOSE. Optimal placement of the uterine tandem and vaginal ovoids produces a pear-shaped distribution, delivering a high dose to the cervix and paracervical tissues and a reduced dose to the rectum and bladder.

Treatment dose has been specified in a number of ways, making it very difficult to compare experiences. Paracentral doses are most frequently expressed at a single point, usually designated point A. This reference point has been calculated in a number of different ways, but it is usually placed 2 cm lateral and 2 cm superior to the external cervical os, in the central plane of the intracavitary system. Point A lies approximately at the crossing of the ureter and the uterine artery, but it bears no consistent relationship to the tumor or target volume. Point A was originally developed as part of the Manchester treatment system (a modification of the earlier Paris system). It was meant to be used in the context of a detailed set of rules governing the placement and loading of the intracavitary system. Today this context is often lost.

In 1985 the International Commission on Radiation Units and Measurements recommended that reference points like point A not be used because "such points are located in a region where the dose gradient is high and any inaccuracy in the determination of distance results in large uncertainties in the absorbed doses evaluated at these points." Instead, they recommended that doses be specified in terms of the following: (1) total reference air Kerma -- expressed in muGy at 1 meter -- is an alternative to milligram-hours that allows for the use of various radionuclides; it is proportional to the dose at points distant from the system, providing an estimate of the integral dose to the pelvis; (2) description of the reference volume, that is, the volume of tissue receiving 60 Gy or more, and (3) doses to specific normal tissue reference points (bladder, rectum, vagina). Although normal tissue reference points provide useful information about the dose to a portion of normal tissue, several studies have demonstrated that they consistently underestimate the maximum dose to those tissues. Whatever system of dose specification is used, emphasis should always be placed on achieving an optimized relationship between the intracavitary applicators and the cervical tumor and other pelvic tissues. Source strengths and positions should be carefully chosen to provide optimal tumor coverage without exceeding normal tissue tolerance. However, optimized source placement can rarely correct for a poorly positioned applicator.

A detailed description of the characteristics of an ideal intracavitary system and of the considerations that influence source strength and position are beyond the scope of this chapter. However, an effort should always be made to deliver 85 Gy or more to point A for patients with bulky central disease. If the intracavitary placement has been optimized, this can usually be accomplished without exceeding a dose of 75 Gy to the bladder reference point or 70 Gy to the rectal reference point, doses that are usually associated with an acceptably low risk of major complications. The dose to the surface of the lateral wall of the apical vagina should not usually exceed 130 to 140 Gy. Suboptimal placements occasionally force compromises in the dose to tumor or normal tissues. To choose a treatment that optimizes the therapeutic ratio in these circumstances requires experience and a detailed understanding of factors that influence tumor control and normal tissue complications.

A total dose (external beam and intracavitary) of 50 to 55 Gy appears to be sufficient to sterilize microscopic disease in the pelvic nodes in most patients. It is customary to boost the dose to a total of 60 to 65 Gy in lymph nodes known to contain gross disease and in heavily involved parametria.

BRACHYTHERAPY DOSE RATE. Traditionally, cervical brachytherapy has been performed with sources that yield a dose rate (at point A) of approximately 40 to 50 cGy/hr. These low dose rates permit repair of sublethal cellular injury, normal tissues are preferentially spared, and the therapeutic ratio is optimized. To reduce the 3 to 4 days of hospitalization needed to deliver an appropriate dose of low-dose-rate irradiation, some investigators have explored the use of intermediate-dose-rate brachytherapy (80 to 100 cGy/hr). However, in a randomized trial, Haie-Meder and colleagues [ref: 275] reported a significant increase in complications when the dose rate was doubled from 40 to 80 cGy/hr, indicating that the total dose must be reduced and the therapeutic ratio of treatment may be compromised with higher dose rates.

Since the 1970s, computer technology has made it possible to deliver brachytherapy at very high dose rates (more than 100 cGy/min) with a high activity *60Co or *192Ir source and remote afterloading. High-dose-rate intracavitary therapy is now being used for radical treatment of cervical cancer by a number of groups, including several in Japan, Canada, and Europe, and more recently by some groups in the United States. Clinicians have found this approach attractive because it does not require that patients be hospitalized and may be more convenient for the patient and the physician. However, unless it is heavily fractionated, high-dose-rate brachytherapy loses the radiobiologic advantage of low-dose-rate treatment, potentially narrowing the therapeutic window for complication-free cure. Advocates of high-dose-rate treatment disagree about the number of fractions and total dose that should be delivered. Published experiences suggest that survival rates are roughly similar to those achieved with traditional low-dose-rate treatment, but are difficult to compare because of the same potential problems of selection and bias that confound other nonrandomized comparisons. Two purported randomized trials have been criticized for methodologic flaws, and the use of high-dose-rate brachytherapy for cervical cancer continues to be a source of controversy.

INTERSTITIAL BRACHYTHERAPY. Several groups have advocated the use of interstitial brachytherapy to treat patients whose anatomy or tumor distribution make it difficult to obtain an ideal intracavitary placement. Interstitial implants are usually placed transperineally, guided by a Lucite template that encourages parallel placement of hollow needles that penetrate the cervix and paracervical spaces; needles are usually loaded with *192Ir. Advocates of the procedure describe the relatively homogeneous dose distribution achieved with this method, the ease of inserting implants in patients whose uteri are difficult to probe, and the ability to place sources directly into the parametrium. Initial reports were enthusiastic, describing these theoretical advantages and high initial local control rates.

However, there have as yet been very few reports of long-term survival rates in patients treated with interstitial brachytherapy for primary cervical cancer. In 1986, Syed and colleagues reported a projected 5-year survival rate of 53% for 26 patients with stage IIIB disease. Martinez and colleagues reported an 83% local control rate in 37 patients with stages IIB to IIIB treated at Stanford and the Mayo clinic. However, the results of a recent update of the combined experiences at Stanford and the Joint Center for Radiation Therapy at Harvard were disappointing: 3-year disease-free survival rates for patients with stages IIB and IIIB tumors were 36% and 18%, respectively; local control rates were 22% and 44%, respectively; and in patients with local control, the rate of complications requiring surgical intervention was high. This series of 70 patients with cervical cancers treated with interstitial therapy is the largest that has been published to date.

Recently several groups have been exploring the use of MRI-guided needle placement, interstitial hyperthermia, and high-dose-rate interstitial therapy to improve local control and complication rates. However, outside of an investigational setting, interstitial treatment of primary cervical cancers should probably be limited to patients who cannot accommodate intrauterine brachytherapy and to those with distal vaginal disease that requires a boost with interstitial brachytherapy.

COMPLICATIONS OF RADICAL RADIOTHERAPY. During radiotherapy of the pelvis, most patients have mild fatigue and mild to moderate diarrhea that usually is controllable with antidiarrheal medications; some patients have mild bladder irritation. When extended fields are treated, patients may have nausea, gastric irritation, and mild depression of peripheral blood counts. Acute symptoms may be increased in patients receiving concurrent chemotherapy. Unless the ovaries have been transposed, all premenopausal patients who receive pelvic radiotherapy experience ovarian failure by the completion of treatment.

Complications of intracavitary therapy include uterine perforation, fever, and the usual risks of anesthesia. Thromboembolic complications are rare. In a review of 327 patients who had gynecologic brachytherapy for a variety of indications, Dusenberry and colleagues reported four (1.2%) thromboembolic complications. In a recent unpublished review of 1784 patients treated with radiotherapy for Stage IB cervical cancer at M. D. Anderson, there were 3 (0.17%) suspected cases of pulmonary embolus, none of which were fatal. Patients with bulky disease on the pelvic wall may have a somewhat greater risk of thromboembolic events.

Estimates of the risk of late complications of radical radiotherapy vary according to the grading system, duration of follow-up, method of calculation, treatment method, and prevalence of risk factors in the study population. However, most reports quote an overall risk of major complications (requiring transfusion, hospitalization, or surgical intervention) of 5% to 15%. Perez and colleagues reported a crude risk of major complications of 14.8% with a median follow-up of 12 years. In a report from the Patterns of Care Study, Lanciano and colleagues reported an actuarial risk of 8% at 3 years. In a study of 1784 patients with stage IB disease, Eifel and colleagues reported an overall actuarial risk of major complications of 7.7% at 5 years. Although the actuarial risk was greatest during the first 3 years of follow-up, there was a continuing risk to surviving patients of approximately 0.34% per year, resulting in an overall actuarial risk of 14.4% at 20 years.

During the first 3 years after treatment, rectal complications are most common and include bleeding, stricture, ulceration, and fistula. In the study by Eifel and colleagues, the risk of major rectosigmoid complications was 2.3% at 5 years. Urinary tract complications, including hematuria, fistula, and ureteral stricture, occurred less frequently during the first 3 years, but had a continuing risk with added follow-up. The actuarial risk of developing a fistula of any type was 1.7% at 5 years.

Small bowel obstruction is an infrequent complication of standard radiotherapy for patients without special risk factors. The risk is increased dramatically in patients who have undergone transperitoneal lymph node dissections. However, there appears to be little added risk if the operation is performed with a retroperitoneal approach. Other factors that can increase the risk of small bowel complications in patients treated for cervical cancer include pelvic inflammatory disease, thin body habitus, and the use of high doses or large volumes of external-beam irradiation, particularly with low-energy treatment beams and large daily fraction sizes.

Most patients treated with radical radiotherapy have some agglutination and telangiectasia of the apical vagina. More significant vaginal shortening can occur, particularly in elderly, postmenopausal women and those with extensive tumors treated with a high dose of irradiation. Vaginal function can be optimized with appropriate estrogen support and vaginal dilatation.

NEOADJUVANT CHEMOTHERAPY. Studies that have explored combinations of chemotherapy and radiation therapy for patients with locally advanced cervical cancers have generally used one of two strategies. Studies of neoadjuvant chemotherapy have attempted to exploit the high response rates of previously untreated tumors to chemotherapy by administering systemic agents before any local therapy, hopefully to reduce the volume of tumor treated with radiation. In contrast, the primary goal of studies that combine radiation therapy with concurrent chemotherapy is to use chemotherapeutic agents to sensitize tumor cells to the effects of radiotherapy.

Investigators were encouraged to explore neoadjuvant treatment because of the high response rates reported for a variety of cisplatin-containing combinations evaluated in phase II studies of previously untreated patients with cervical cancer. However, it was not possible to determine from these uncontrolled trials whether high response rates to chemotherapy would lead to improved survival rates. Five randomized trials comparing neoadjuvant chemotherapy followed by irradiation and irradiation alone have been reported. Chauvergne and colleagues randomized 151 patients with stages IIB or III disease to receive radical radiotherapy alone or preceded by chemotherapy (cisplatin, methotrexate, chlorambucil, and vincristine). Although objective responses to chemotherapy were observed in 31 (43%) of 73 patients, there was no difference in the survival or pelvic disease control rates of patients in the two treatment arms. In a randomized study of 184 patients with stages IIB to IVA squamous carcinomas, Kumar and colleagues found no significant difference in disease-free or overall survival between patients treated with bleomycin, cisplatin, and ifosfamide followed by irradiation or with irradiation alone. Tattersall and colleagues reported no difference in survival when they compared chemotherapy (cisplatin, bleomycin, and Velban) followed by pelvic irradiation with radiotherapy alone in 71 patients who had positive pelvic lymph nodes discovered at radical hysterectomy. Souhami and colleagues reported a significantly poorer survival rate (23% versus 39%; P =0.02) for patients who received neoadjuvant chemotherapy (cisplatin, bleomycin, vincristine, and mitomycin-C) in a study of 107 patients with stage IIIB cervical cancers. In another large prospective trial reported by Tattersall and colleagues, 129 patients who received neoadjuvant chemotherapy (cisplatin and epirubicin) had a significantly poorer pelvic disease control rate (P = 0.003) and survival rate (P = 0.02) than 131 patients treated with radiation therapy alone, despite a high initial response rate to chemotherapy (63%).

In summary, despite high response rates of locally advanced cervical cancers to initial chemotherapy, none of the randomized studies reported to date has demonstrated an improvement in outcome when neoadjuvant chemotherapy was added to radical radiotherapy. In many ways, this recapitulates the experience with treatment of locally advanced head and neck cancers for which it has been hypothesized that the failure to influence outcome with neoadjuvant chemotherapy may reflect cross-resistance of clonagens to drugs and radiation or accelerated repopulation of tumor clonogens induced by neoadjuvant chemotherapy.

CONCURRENT CHEMORADIATION. A number of investigators have reported high response rates and encouraging survival rates in uncontrolled phase I/II studies of patients with locally advanced cervical carcinomas treated with concurrent chemotherapy and irradiation. Drugs that are most commonly given concurrently with radiation therapy include hydroxyurea, 5-fluorouracil, mitomycin-C, and cisplatin.

Hydroxyurea, a drug that has been demonstrated in vitro to sensitize cells to radiation by arresting them in a sensitive portion of the cell cycle, was identified as an attractive subject for clinical study in the early 1970s. After a small randomized study suggested encouraging results, the Gynecologic Oncology Group conducted a larger study randomizing patients with stage IIIB or IVA disease either to receive oral hydroxyurea (80 mg/kg, up to 6 g per dose, every 3 days for 12 weeks) or placebo during radiotherapy. The authors of this study reported that patients who received hydroxyurea had a higher complete response rate (68% versus 48%), a longer median progression-free interval (13.6 versus 7.6 months), and a longer median survival (19.5 versus 10.7 months). However, this trial has been criticized because 100 (53%) of the 190 patients entered in the study were considered to be either ineligible or inevaluable and were excluded from the analysis. The minimum recommended dose of radiotherapy was also conservative for patients with theses advanced tumors (70 Gy if a combination of external-beam and intracavitary irradiation was used or 60 Gy if only external-beam irradiation was used), possibly contributing to the poor overall survival rates in both arms. A second Gynecologic Oncology Group study compared hydroxyurea (80 mg/kg, given twice per week during external-beam irradiation) with misonidazole, a nitroimidazole hypoxic cell sensitizer that has since been demonstrated to be of no benefit in several trials that compared misonidazole with a placebo. Patients with stages IIB to IIIB squamous carcinomas were eligible for this study if they had negative paraaortic lymph nodes confirmed by staging laparotomy. Of 294 evaluable patients, 178 (61%) had stage IIB disease. An updated analysis of this study in 1993 showed a marginal advantage in progression-free survival (P = 0.05) and survival (P = 0.07), which appeared to be most marked for patients with stages IIIB and IVA disease. This has been interpreted as confirmatory evidence of the value of hydroxyurea as a radiosensitizer in this setting. However, the suggestion from several studies that nitroimidazole sensitizers may actually decrease pelvic control rates in patients with cervical cancer may invalidate this conclusion.

A third Gynecologic Oncology Group study using similar entry criteria randomized patients to receive either hydroxyurea (80 mg/kg orally twice weekly during external-beam irradiation) or cisplatin (50 mg/m*2 intravenously on days 1 and 29) and 5-fluorouracil (1000 mg/m*2 as a continuous intravenous infusion on days 2 through 5 and days 30 through 33 of radiation). This study closed to patient accrual in 1990, and the results should be analyzed and reported soon. Ongoing studies that address the value of chemosensitization in patients with locally advanced cervical cancer include (1) a Gynecology Oncology Group study randomizing patients to receive either hydroxyurea, weekly cisplatin, or a combination of hydroxyurea, 5-fluorouracil, and cisplatin during external-beam irradiation and (2) a Radiation Therapy Oncology Group study comparing cisplatin plus 5-fluorouracil (administered in three cycles during external-beam and intracavitary irradiation) with extended field irradiation plus brachytherapy.

In summary, concurrent treatment of locally advanced cervical cancers with chemotherapy and radiotherapy is an approach that holds considerable promise, although additional studies are needed to confirm the benefit of this treatment. To confirm that concurrent chemotherapy improves the therapeutic ratio achievable with radical radiotherapy, future studies also need to provide careful documentation of the late complications of treatment. The morbidity of radiotherapy with and without chemotherapy has not yet been compared in a randomized study, but one uncontrolled study suggests that late gastrointestinal toxicity may be increased with concurrent administration of some drugs.

INTRAARTERIAL CHEMOTHERAPY. Intraarterial infusion of chemotherapeutic agents delivered neoadjuvantly, concurrent with radiotherapy, or as salvage treatment for recurrent disease has generated interest for some years because of the distinct arterial supply to the central pelvis. A number of drugs have been used in small pilot studies, but 5-fluorouracil and cisplatin have been the most popular in this setting. Unfortunately, this technique is difficult and invasive, the toxicity reported in some series has been substantial, and the results have been variable in several small series of patients. However, occasional optimistic reports have maintained a low level of interest in this approach, particularly for concurrent intraarterial chemotherapy and irradiation.

Stage IVB

Patients who present with disseminated disease are almost always incurable. Management of these patients must emphasize palliation of symptoms with appropriate pain medications, and localized radiotherapy. Tumors may respond to chemotherapy, but the duration of responses is usually short.

SINGLE-AGENT CHEMOTHERAPY. Fifty-two drugs have been studied in sufficient numbers of patients with carcinoma of the cervix to assess their activity. Nineteen of these have yielded response rates (partial and complete) of at least 15% and may be of therapeutic value.

Several of the platinum compounds have been evaluated in greater detail. Cisplatin has been studied in a variety of doses and schedules. These studies have demonstrated activity of the drug at a dose of 50 mg/m*2 given intravenously at a rate of 1 mg/min every 3 weeks. Although there appears to be a small but statistically significant increase in the response rate with a doubling of the dose to 100 mg/m*2, this has not resulted in a detectable improvement in the rates of progression-free or overall survival. More prolonged infusion of the same dose over 24 hours yields a similar response rate with less nausea and vomiting, although the recent development of more effective antiemetic agents reduces the clinical importance of this observation. The response rates of other platinum compounds (i.e., carboplatin and iproplatin) are lower than those observed with cisplatin, which remains the platinum compound of choice for patients with cervical carcinomas.

Ifosfamide has been studied as a single agent in patients with recurrent cervical cancer in at least five phase II trials. Response rates ranged between 33% and 50% in three studies that were conducted in patients who had received no previous chemotherapy. However, the response rates were much lower in two phase II trials that included patients who had received prior systemic chemotherapy, with only three partial responses (8%) in 36 patients.

COMBINATION CHEMOTHERAPY. Most reports of combination chemotherapy for carcinoma of the cervix have described uncontrolled phase II trials of drug combinations that have included at least some agents with known activity. Most studies have been small, with reported response rates ranging from 0% to 100%. In general, data from these phase II studies provide no firm evidence that any of the studied combinations are superior to single-agent therapy for patients with disseminated or recurrent cervical cancer. However, combinations based on ifosfamide and cisplatin and those based on 5-fluorouracil and cisplatin have attracted significant interest and deserve further discussion.

Several small phase II studies have evaluated treatment with combinations of ifosfamide and either cisplatin or carboplatinum in patients who had not received prior radiotherapy. Response rates for these combinations ranged between 50% and 63%. A number of investigators have combined bleomycin with ifosfamide and a platinum compound. Three studies included patients who had not had prior radiotherapy, and reported response rates were 65% to 100%. Reports of treatment with these drugs in previously irradiated patients have yielded mixed, but generally lower, response rates of between 13% and 72%.

Combinations of cisplatin and continuous infusion 5-fluorouracil also produce high response rates in previously untreated patients. Again, response rates decrease significantly if patients have had previous irradiation.

The Gynecologic Oncology Group recently completed a large prospective randomized trial comparing cisplatin alone with cisplatin plus ifosfamide and cisplatin plus dibromodulcitol in patients with advanced or recurrent cervical cancers. The addition of ifosfamide to cisplatin improved the response rate (33% versus 19%, P =0.02) and progression-free survival rate (4.6 versus 3.2 months, P < 0.05), but caused significantly greater toxicity (leukopenia, peripheral neuropathy, renal toxicity, and encephalopathy) and did not significantly improve the overall median survival. The addition of dibromodulcitol did not improve the response rate or survival duration.

PALLIATIVE RADIOTHERAPY. Localized radiotherapy can provide effective pain relief for symptomatic metastases in bone, brain, lymph nodes, or other sites. A rapid course of pelvic radiotherapy can also provide excellent relief of pain and bleeding for patients who present with incurable disseminated disease.

Side Effects of Chemotherapy

- Anemia

- Appetite Changes

- Bleeding Problems

- Constipation

- Diarrhea

- Fatigue (Feeling Weak and Very Tired)

- Hair Loss

- Infection

- Memory Changes

- Mouth and Throat Changes

- Nausea and Vomiting

- Nerve Changes

- Pain

- Sexual and Fertility Changes in Men

- Sexual and Fertility Changes in Women

- Skin and Nail Changes

- Swelling (Fluid Retention)

- Urination Changes

Pregnancy and contraception

Many chemotherapy drugs may cause birth defects. It is important that a woman undergoing chemotherapy avoids becoming pregnant. As most chemotherapy medications interfere with oral contraceptives it is important to use a barrier method of contraception, such as condoms, during the whole chemotherapy treatment period and for a year after treatment is completed. If you are pregnant you need to tell the medical team straight away beforehand. If you become pregnant during treatment tell the medical team straight away. What are the side effects of chemotherapy?

Most people immediately link chemotherapy with uncomfortable side effects. However, side-effect management has improved considerably over the last twenty years. Many side effects that were once inevitable can be either prevented or well controlled today.

There is no reliable way to predict how patients may react to chemotherapy. Some experience very mild side-effect, others will have none at all, while some people will report various symptoms.

Depending on the type of cancer and treatment, chemotherapy may have a bigger impact on the patient's work status than radiotherapy. Women with breast cancer who receive chemotherapy appear more likely than those treated with radiation therapy to experience a major change in work status, according to researchers at the Dana-Farber Cancer Institute.

Below is a list of the most commonly reported side effects:

Nausea and vomiting

Over half of all patients receiving chemotherapy will experience nausea and vomiting. Doctors will usually prescribe anti-emetics for this. These need to be taken even when symptoms have gone as they will prevent them from coming back. If the anti-emetics do not work the patient should contact his/her doctor who may switch to another anti-emetic.

Ginger - scientists at the Rochester University Medical Center found that taking ginger supplements with standard anti-vomiting drugs beforehand can reduce the nausea that often accompanies chemotherapy treatment by 40%.

Alopecia (Hair loss)

Some chemotherapy medications cause hair loss while others don't. If hair does start to fall out this will usually happen a few weeks after treatment starts. On some occasions the hair will just become thinner and more brittle (without falling out). Hair loss can occur in any part of the body.

Although hair loss has no physical health consequences, it may cause distress and embarrassment for some people. The psychological impact tends to be greater among women than men. If you experience hair loss and find it is causing distress and embarrassment, there are several steps you can take:

Tell your doctor, who may refer you to a counselor who can provide effective help and support.

Many people find that if they purchase a wig their quality of life improves significantly.

If there is a cancer support group in your area, go to their meetings. Meeting people who share similar experiences to yourself may help give you a boost, as well as providing you with some useful tips, and possibly an opportunity to make new friends.

Cold cap - this looks a bit like a bicycle helmet and keeps the scalp cool while the chemotherapy dose is being administered. If the scalp can be kept cool less chemotherapy medication reaches the scalp, thus preventing the occurrence or reducing the severity of hair loss. Some people cannot wear a cold cap - leukemia (blood cancer) patients need the medication to reach their scalp.

The hair loss is NOT permanent - it will grow back soon after treatment if finished.

Fatigue

Most patients receiving chemotherapy will experience some degree of fatigue. This may be a general feeling which exists most of the day, or may only appear after certain activities. Doctors say patients need to make sure they get plenty of rest and not to perform tasks which are overtiring.

While light exercise has been shown to help, it is important to remember to keep the activities 'light'.

If the tiredness becomes severe it is important to tell the doctor, as this could be caused by a significant drop in red blood cells (anemia).

Hearing impairment (deafness, ototoxicity)

Scientists from Oregon Health & Science University reported that deafness as a side effect of chemotherapy has long been underreported by the medical community, because a well-known classification system doctors use for reporting toxicities in patients does not consider high-frequency hearing loss, allowing the magnitude of ototoxicity (hearing damage) in children treated with platinum agents to be miscalculated.

Children with cancer who suffer hearing loss due to the toxic effects of chemotherapy might one day be able to get their hearing back through pharmacological and gene therapy, said researchers from St. Jude Children's Research Hospital after carrying out a study on mice.

Neutropenia (low white blood cells) - Susceptibility to infections

When receiving chemotherapy the immune system will be weakened because the white blood cell count will go down. White blood cells form part of our immune system - they fight infection. Consequently, patients become more susceptible to infections.

Some patients will be prescribed antibiotics which may reduce their risk of developing infections. The following precautions will help reduce the risk of infections:

Personal hygiene - the cleaner you are, the fewer bacteria there will be around which can infect you. Regularly wash your hands with warm water and soap, have a bath/shower at least once a day, change your clothes and bathroom towels and flannels daily. Change your bed linen regularly.

Preparing food - make sure your food is free of food borne pathogens (organisms, such as bacteria that can make you ill). If you handle raw meat make sure you wash your hands before touching plates and cutlery or work surfaces. Thoroughly cook animal sourced proteins before eating them. Wash your dishes thoroughly and always use a clean plate and cutlery - keep the kitchen clean.

Infected people - stay away from people who are ill. This may include those who just have a temperature.

Skin wounds - bacteria find it hard to get in through your skin, unless there is a cut. If you graze or cut your skin, clean the area well with warm water, dry it, and cover it with a sterile dressing.

Patients receiving chemotherapy who develop an infection need immediate treatment. This may mean being hospitalized and receiving antibiotics via an intravenous drip.

Thrombocytopenia (low blood platelet count) - Blood clotting problems

Chemotherapy may lower the patient's blood platelet count. A platelet is a type of blood cell that helps the blood to clot (coagulate). Coagulation is essential, otherwise bleeding does not stop. Lower blood platelet counts linked to chemotherapy is a risk, but less so than lower red or white blood cell counts. If you are affected you will bruise more easily, you will be more likely to have nosebleeds and bleeding gums, and if you cut yourself it may be harder to stop the bleeding.

Patient's whose blood platelet counts fall too low will need a blood transfusion.

Below are some steps you may wish to take to reduce your risk of bleeding:

Shave with an electric razor (or don't shave)

Avoid hard toothbrushes

Use kitchen utensils and gardening equipment carefully

If you are gardening, wear gloves

Anemia (low red blood-cell count)

As well as lowering you white blood cell count, chemotherapy will also lower your red blood cell count. Tissues and organs inside your body get their oxygen from the red blood cells. If your red blood cell count goes down too many parts of your body will not get enough oxygen and you will develop anemia.

People with anemia feel very tired. A patient on chemotherapy who has anemia will feel extra tired - much more tired than straightforward fatigue caused by the treatment. Dyspnea (shortness of breath) is also another symptom of anemia, as are palpitations (when the heart beat is irregular).

Anemia linked to chemotherapy requires immediate treatment. A blood transfusion will bring the red blood cell count back up immediately. Erythropoietin (EPO) is a drug that makes the body produce more red blood cells.

The following foods are rich in iron, which helps red blood cells carry more oxygen. Dark green leafy vegetables, beans, meat, nuts, prunes, raisins, and apricots.

Scientists from The Medical University of Vienna, Austria found that patients with breast cancer who developed anemia during chemotherapy had nearly three times the risk of local recurrence as those who did not.

Mucositis (inflammation of the mucous membrane)

Chemotherapy attacks rapidly dividing cells, such as blood cells, bone marrow cells, and cells of the mucous membranes that line the digestive system - this includes the mouth, esophagus, stomach, intestines, and the rectum to the anus. Chemotherapy may damage and even destroy some of those mucous membrane cells.

Oral Mucositis (in the mouth) - patients more commonly experience symptoms in their mouth.

If symptoms do appear, they will usually do so about 7 to 10 days after treatment starts. The inside of the mouth may feel like sunburn; some people say it feels as if the area had been scalded. Ulcers often appear on the lining of the mouth, the tongue, and sometimes around the lips. The severity of symptoms is closely linked to the strength of the chemotherapy dose.

Some may find it painful when they eat, drink, or even talk. If the ulcers bleed there is a risk of infection.

Caphosol is often prescribed for mucositis.

A clinical trial showed that out of 100 cancer patients that were treated with DAVANAT® and chemotherapy that included 5-FU, none developed mucositis.

As better drugs are appearing, mucositis is becoming less common. Symptoms clear up a few weeks after treatment is completed.

Loss of appetite

Loss of appetite is a common side effect of chemotherapy. It is possible that the chemotherapy, or the cancer itself, affects the body's metabolism. If the loss of appetite is just due to the chemotherapy it will come back when the treatment is finished - although this may sometimes take a few weeks.

The severity of appetite and consequent weight loss depends on the type of cancer and chemotherapy treatment.

Although this is sometimes easier said than done, it is important to keep trying to eat well and take in plenty of fluids. Many patients find that smaller and more frequent meals are easier to get down than the typical three meal-a-day regime. Also, drinking liquids through a straw may result in a better fluid intake.

Patients who become seriously affected by lack of food and liquid intake may need to be hospitalized and fed through a nasogastric tube. The tube goes into the patient's nose and down to his/her stomach.

Nails and skin

Chemotherapy can sometimes cause dry and sore skin. Nails may also become flaky and brittle. The skin may become more sensitive to sunlight. It is important to protect yourself from too much sunlight exposure. This includes staying out of the sun during peak times of the day, using sun blocks, and wearing clothes that provide maximum protection. Surprisingly, scientists at Michigan University, USA, reported that the chemotherapy drug fluorouracil appeared to reduce the appearance of sun-damaged and aging skin as well as the number of potentially pre-cancerous skin patches.

Cognitive problems

About one fifth of patients undergoing chemotherapy report some kind of cognitive problem, including attention, thinking and memory. This can sometimes have an impact on daily tasks. Patients who do experience these symptoms should talk to their doctor, and social worker.

Symptoms may include:

Shorter attention span; concentration, focus and attention problems

Memory problems; especially the short-term memory

Comprehension and understand problems

Judgment and reasoning problems

Organizational skills may be affected

Multitasking problems (performing/thinking about several things at the same time)

Mood swings

Experts are unsure how much is due to the chemotherapy, and how much is due to fatigue, stress and anxiety that comes with having cancer.

Libido (sex drive) and fertility

For a significant proportion of patients, chemotherapy may result in a lower sex drive (less interest in sex). This is temporary and usually returns after treatment is completed.

Depending on the type of medication administered, chemotherapy may also damage men's sperm. Some women may become infertile. In most cases - though not all - fertility returns after treatment is over.

Men who wish to father children and women who plan to become pregnant one day should discuss possible options with their doctors before starting treatment. It is possible to freeze sperm and embryos.

Bowel movement problems (diarrhea or constipation)

Sometimes when damaged cells in the intestinal tract are rapidly expelled from the body there is a risk of diarrhea. Constipation is also a possible risk for chemotherapy patients. You should talk to your doctor if you experience any unpleasant change in your bowel movements. Symptoms, if they do occur, will do so a few days after chemotherapy begins.

Depression

The risk of developing depression is already higher for patients with cancer. It is normal to feel distressed, anxious, sad and stressed - especially if you are concerned about what the future holds and whether treatment is going to be effective.

It is important that you talk to a member of the medical team if you feel it is all getting to be too much, or if you no longer get pleasure out of the things that you used to like. Joining a support group and talking to people who are going through the same as you and understand how you feel has helped many people with cancer. Protecting the gut from chemotherapy in cases of metastatic cancer

When cancer has spread beyond the primary tumor to other parts of the body, i.e. when it has metastasized, treatments tend to involve intensive chemoradiotherapy. These high doses of raditation therapy and chemotherapy damage good tissue and often push the patient to an early death.

Researchers from the University of Michigan School of Dentistry reported in the journal Nature that they found a way to preserve the health of the gut in laboratory mice that had been given very high doses of chemoradiotherapy.

Team leader, Jian-Guo Geng, said "All tumors from different tissues and organs can be killed by high doses of chemotherapy and radiation, but the current challenge for treating the later-staged metastasized cancer is that you actually kill the patient before you kill the tumor."

Healthy digestive tract improves survival - the scientists found that if the gut of mice with metastatic cancer stayed healthy after intensive chemoradiotherapy, their chances of survival increased exponentially.

There are stem cells in the gut of the laboratory mice that help preserve tissue after undergoing lethal doses of chemoradiotherapy. Geng and colleagues found a way of getting the gut stem cells to launch a massive healing campaign. Gene Print Predicts Who Will Survive Chemotherapy

An 8-gene "print" or "signature" can predict how long people remain cancer-free, without any relapse after undergoing chemotherapy, researchers from Academia Sinica and the National Taiwan University College of Medicine explained in the journal BMC Medicine.

Prof Ker-Chau Li and team located genes that are involved in cellular invasion - something several cancer cells do. They gathered and examined data from the National Cancer Institute's 60 human cancer cell line panel.

Prof. Li said "Our study found eight genes which were involved in invasion, and the relative activation of these genes correlated to chemotherapy outcome, including the receptor for growth factor EGF. We also found that some invasion genes had unique patterns of expression that reflect the differential cell responses to each of the chemotherapy agents - five drugs (paclitaxel, docetaxel, erlotinib, everolimus and dasatinib) had the greatest effect."

Cancer Survivors

Ø Cancer support groups

Ø Cancer information workshops

Ø Low-cost medical consultation

Ø Increased government funding to seek cures

AND

186,550 - The number of lives that could be saved each year if we ate a healthy diet and exercised regularly

186,550 - The number of lives that could be saved each year with no tobacco

American Cancer Society