Epidemiology and Etiology

Bladder cancer is estimated to have an annual incidence in the United States of 68,810 cases, accounting for 5% of all newly diagnosed cancers. Approximately 14,100 people per year will die of this disease, accounting for 2.5% of all cancer-related mortality in the United States and 3% of all cancer deaths in men.¹ The incidence of bladder cancer varies considerably among countries; the highest incidence rates are in Western countries. At diagnosis, approximately 75% of patients have localized disease only, 20% have regional disease, and 5% present with distant metastases. Two thirds of all bladder cancer diagnoses occur among persons 65 years of age or older. The male/female ratio is 3 : 1. In the United States, the risk for white men is approximately twice that for African-American or Hispanic men.² From the early 1970s to 2004 the incidence of this cancer rose slightly while mortality fell. This phenomenon is likely attributable to the increased detection of localized lesions and improved therapies.³

Cigarette smoking is the single most significant known cause of bladder cancer, as demonstrated by both cohort and case-control studies.^3–8 The risk of bladder cancer appears to be related to the number of pack-years and, as such, an individual’s risk of bladder cancer declines if smoking is ceased.⁹ It is estimated that 50% of bladder cancers diagnosed in the United States can be linked to a smoking history.¹⁰ In a case-controlled study of more than 1500 patients with bladder cancer and a large control group matched for age, sex, and race, there was a statistically higher risk of bladder cancer observed in female smokers compared with male smokers with similar smoking habits.¹¹ In addition, younger age at initiation of tobacco use correlates to a higher risk of developing bladder cancer.¹² It appears that the relationship of smoking to bladder cancer development stems from exposure to selected arylamines. Studies have demonstrated that smokers have higher levels of 3- and 4-aminobiphenyl hemoglobin adducts, which are increased 10-fold and 3-fold, respectively, compared with nonsmokers.¹¹ Cytogenetic analyses indicate a direct relationship between smoking and suppressor gene p53 nuclear overexpression.¹³

Exposure to industrial chemicals and high rates of bladder cancer in specific occupations have been a major focus of epidemiologic investigation. Individuals in contact with chemicals necessary for the production of dyes, rubber, plastics, and synthetic materials are at increased risk for development of bladder cancer.^3,13–17 Dye workers may be particularly at risk owing to exposure to aromatic amines, specifically 2-naphthylamine and benzidine, which have been shown to be bladder carcinogens.¹⁸ One study demonstrated an additive effect for bladder cancer development among smokers who use hair dyes.¹⁹ There have been recent changes in hair dye formulations that appear to have decreased this risk.²⁰ Two recent meta-analyses did not demonstrate a link to personal hair dye users and the subsequent development of bladder cancer.^21,22 Rubber workers are also exposed to 2-naphthylamine, and benzidine is found in paints.³ Dry cleaners may have an increased risk of bladder cancer from exposure to organic solvents, including perchloroethylene.²³

Schistosoma haematobium is presumed to be the primary cause of endemic squamous cell carcinoma common in Egypt.²⁴ This form of bladder cancer is recognized to occur in patients 10 to 20 years younger than patients with typical cases of transitional cell carcinoma diagnosed in the United States.²⁵ It has been hypothesized that the parasite may induce an inflammatory response that includes the release of reactive oxygen radicals, which results in genetic damage, leading to carcinogenesis.²⁶ Patients with chronic urinary tract infections or with the need to frequently self-catheterize after spinal cord injury are also predisposed to squamous cell carcinomas.^27–29 Mechanisms proposed include obstructive uropathy with recurrent bacterial superinfections leading to urine acidification, nitrosamine production, and bacteria-induced oxidative damage.³⁰

Patients with a history of upper tract urothelial cancer are at high risk for the subsequent development of synchronous or metachronous cancers of the urinary bladder as well as contralateral upper urinary tract malignancies. The incidence of developing a bladder cancer is 20% to 50% regardless of the treatment modality and typically occurs within the first 2 years after treatment of the upper tract; however, lifelong surveillance is necessary in these patients.³¹

The role of diet in the predisposition for, or reduction of risk of, developing bladder cancer remains unclear. There have been conflicting data about the role of coffee consumption and an association with bladder cancer development. A summary of 35 case-control studies of the causal role of coffee showed no evidence of a relationship between the beverage and the development of bladder cancer.^32,33 On the other hand, a recent meta-analysis for urinary tract cancers estimated a 20% increase of bladder cancers among coffee drinkers irrespective of dose or duration of consumption.³⁴ It is now believed that there is no relationship between artificial sweeteners and the development of bladder cancer.³⁵ Some studies have shown a minimal protective effect with vitamin A, fruits, and vegetables.^36,37 People who drink fresh well water were found to have lower rates of bladder cancer than those drinking chlorinated municipal water.³⁸ This may be related to trihalomethanes, by-products of chlorine or bromine disinfection processes, and currently most governments place restrictions on these chemicals. A higher incidence of bladder cancer has been noted in agricultural areas where fertilizers seep into water supplies, raising the level of nitrates in drinking water.³⁹

Iatrogenic causes of bladder cancer include radiation exposure, chronic phenacetin usage, and administration of cyclophosphamide.³ Several reports have noted an increase in bladder cancer in patients who received pelvic radiation for cervical, prostate, and testicular cancers^40–42; however, other studies have not established a clear causative relationship.^43,44 Patients with rheumatoid arthritis were shown to have an increased rate of bladder cancer for as long as 17 years after taking oral cyclophosphamide.⁴⁵

A genetic predisposition to bladder cancer in certain families has been described. It is likely that there is an interaction between hereditary predisposition of oncogenes and tumor suppressor genes that are amplified or deleted because of environmental stimuli.³ Acetylation of aromatic amines, an important mechanism of detoxification of these compounds, may be interrupted in individuals with mutations of the N-acetyltransferase genes, particularly NAT2. These “slow acetylators” have been reported to have an increased risk of bladder cancer, particularly in individuals exposed to aromatic amines and with a higher relative risk compared with “rapid acetylators” (i.e., individuals without NAT2 mutations).⁴⁶ The interplay between environment and genetic predisposition is also exemplified by the reduction of glutathione S-transferase activity with the loss of the GSTM1 gene. The glutathione enzyme appears to play a role in modulating the carcinogenic effects of tobacco smoke.^47,48

Cytogenetics

Scientific inquiry has concentrated on the molecular biology and cytogenetics of bladder cancer, with the expectation that insight into the natural history of the disease will translate into clinical advances. An attempt is under way to define independent prognostic factors that predict which of two similar superficial lesions is destined to relapse as muscle-invasive disease. Flow cytometry has shown that lesions with an increased chromosome number and aneuploidy were associated with higher-grade tumors destined to be invasive.^49–51 One study showed that the percentages of hyperdiploid cells for chromosomes 7, 11, and 17 were predictive of grade and stage.⁵² Other studies have shown that FGFR3 mutation is strongly associated with low tumor grade and low risk of progression to more advanced bladder cancer.⁵³ A model of an orderly progression of alteration in genetic material involving oncogenes and tumor suppressor genes is being elucidated.⁵⁴

An early occurrence that may be critical in distinguishing T_is and T₁ from normal bladder epithelium and the less aggressive papillary tumors (T_a) appears to involve loss of a tumor suppressor gene on chromosome 9.^54–57 Progression to muscle-invasive disease appears to require loss of the tumor suppressor gene p53 on chromosome 17, shown to be present in 65% of T_is compared with 3% of T_a tumors.⁵⁸ Other intermediate events in the progression to invasive disease likely involve deletions in 5q, 3p, and 17p, which are not present in T_a lesions.^59,60 No single oncogene has been found in the majority of tumors, although c-erb B2, H-ras, c-src, and c-jun have been expressed and likely play a role in the intermediate progression to muscle-invasive lesions.^61–63 Late events may involve loss of suppressor genes on 11p, 6q, 13q, and 18q.^48,54,64 In particular, the loss of the retinoblastoma gene on chromosome 13 likely has a role in bladder cancer similar to its role in other malignancies.^65,66 Investigators have also suggested additional loci of tumor-suppressor genes, including 4p, 8p, 14q, and 12q.^67,68 Others have recently demonstrated a strong association with 6p+ and high tumor cell proliferation rate, which was noted to be an independent variable in addition to grade and tumor stage.⁶⁹

A prognostic factor for transitional cell carcinoma confined to the bladder independent of grade, stage, and lymph node status is the accumulation of the p53 protein in the nucleus. With mutation of the wild-type p53, the protein product has a longer half-life and can be detected by immunohistochemistry.⁷⁰ After radical cystectomy, patients with stages pT₁, pT₂, and pT_3a who had p53 detected by immunohistochemistry had a shorter relapse-free and overall survival.⁷¹ Similarly, an analysis of p53 nuclear overexpression in 33 patients with carcinoma in situ (CIS) showed disease progression in 13 of 15 p53-positive patients compared with 3 of 18 p53-negative patients.⁷² The potential clinical implications of using the p53 status as a marker are far reaching and include the selection of lower-risk patients for bladder conservation and patients with early-stage disease who may benefit from adjuvant radiotherapy or chemotherapy.^73–77

Anatomy

Both for the radiation oncologist and for the surgeon, the relationship between the bladder and neighboring structures is critical in treatment planning. The bladder functions as a muscular reservoir and is anatomically divided into a fundus, apex, body, trigone, and neck. Empty, the bladder is confined to the true pelvis. As the organ fills, it expands anterosuperiorly into the abdominal cavity. The fundus or base faces posteriorly and in the male is separated from the rectum superiorly by the rectovesical pouch and inferiorly by the seminal vesicles (Fig. 44-1 and Fig. 44-2). In the female, the bladder faces the uterus and the anterior vaginal wall. The neck is located 3 to 4 cm behind the symphysis pubis. This lowest and most fixed part of the bladder is pierced by the urethra and in the male approximates the prostate. A remnant of the median umbilical ligament called the urachus links the apex of the bladder (which lies posterior to the cranial section of the symphysis pubis) to the umbilicus. The superior surface of the bladder is covered by peritoneum and may approximate the sigmoid colon or ileum and the uterus in the female. Fascia forms the pubovesical ligaments that suspend the bladder between the pubic bones, anterior abdominal wall, and pelvic side walls. An empty bladder is characterized by a mucosa with folds except for a smooth triangular region called the trigone. The superior border of the trigone is characterized by its appearance as a pale band on cystoscopy and is called the interureteral crest. This band separates the ureteral openings. At the trigonal apex is the internal urethral orifice completing the triangle.⁷⁸

Figure 44-1 • Sagittal view in the male demonstrates that the full bladder expands into the lower abdominal cavity. Tumors occur in all five regions of the bladder and are often multifocal.

Figure 44-2 • Cross-sectional view in the male of the bladder and surrounding anatomy. The proximity of the bladder to the prostate and rectum explains the occurrence of locally advanced invasion into these structures.

The arterial supply to the bladder arises from the anterior division of the hypogastric artery. Branches from the obturator and inferior gluteal arteries contribute to the rich vascular supply. In the female, additional vessels include the uterine and vaginal arteries. An important venous plexus exists between the bladder wall and the adventitial layer that drains into the hypogastric veins.⁷⁸ Lymphatics from the mucosa and muscular layer drain to the surface of the bladder and form three collecting ducts at the trigone, anterior, and posterior walls. The trigone collecting duct drains into the external iliac nodes (Fig. 44-3). Hypogastric and common iliac lymph nodes are the destination of lymph collected in the anterior and posterior ducts. Some of the lymph also flows through the route of the anterior bladder wall collecting ducts to external iliac nodes.⁷⁹

Figure 44-3 • Lymphatic drainage of the bladder. Involvement of the common iliac nodes is considered a manifestation of distant metastatic and not regional disease.

The internal anatomy of the bladder can be viewed as consisting of four layers: (1) urothelium or mucosa, (2) lamina propria, (3) muscularis propria, and (4) adventitia or serosa. The mucosa consists of an epithelium of transitional cells supported by a lamina propria of loose connective tissue. The muscularis propria layer consists of nonstriated, irregularly arranged muscle bundles that form the detrusor muscle. The outermost layer is adventitia, and for certain sections of the bladder a serosal covering of peritoneum is also present.

Innervation of the bladder can be divided into a parasympathetic nerve supply arising from S2-4 and a sympathetic supply that originates mainly in T11-12 and L1-2. Sympathetic nerves are found primarily in the trigone, and in the male these nerves stimulate closure of the bladder neck sphincter to prevent reflux ejaculation. Parasympathetic nerves play a large role in detrusor muscle contraction. A small group of autonomic nerves can be found scattered in the bladder wall. The adult male bladder has a normal capacity of 120 to 320 ml, with pain due to stretch sensations mediated by parasympathetic nerve fibers when distention beyond 500 ml occurs.^78,80

Several anatomic variations are important with respect to a predilection to develop bladder cancer. One such variation is the persistence of a patent urachal remnant. The urachus is an embryologic canal connecting the urinary bladder of the fetus with the allantois, a structure that contributes to the formation of the umbilical cord. The lumen of the urachus is normally obliterated during embryonic development, transforming the urachus into a solid cord; however, in some children this closure fails to occur and remnants can take the form of a cyst, sinus, diverticulum, or patent urachus.⁸¹ Urachal adenocarcinoma can arise in these retained remnants. The location of these tumors are typically midline, either anteriorly or in the dome, and although often discrete, can form a large mass on the anterior abdominal wall. Urachal adenocarcimona is rare and accounts for approximately 0.35% to 0.7% of all bladder cancers.⁸² Extroversion of the bladder is one of the most common congenital bladder anomalies leading to a portion of the bladder, typically the mucosal surface of the fundus, protruding from the abdominal wall surface. This defect is also associated with epispadias and pelvic bone development abnormalities, and predisposes patients to bladder cancers of variable histologic anatomies.⁸³ Adults with a history of exstrophy have a 694-fold increase in the risk of bladder cancer by the age of 40 years. Etiologic factors remain unclear, but may be related to resconstructive techniques.⁸⁴ Bladder tumors are often multifocal, reflecting a “field cancerization” defect. Of the tumors that have a subsite listed at diagnosis, the majority occur on the bladder walls. The lateral walls account for 40% of malignancies, with posterior involvement of 11% and anterior involvement of 3%. As many as 17% of tumors are known to arise adjacent to the ureteral orifice, whereas the trigone, dome, and neck account for 13%, 9%, and 7%, respectively, of sites from which the cancers arise.³

Pathologic Findings

Transitional cell carcinomas account for approximately 93% of all bladder cancers in the United States.³ Pattern of tumor growth can be either papillary or flat. Papillary lesions can be either benign (i.e., inverted papilloma) or malignant. Flat lesions can be either dysplasia, CIS, or invasive cancer. In 1998, the World Heath Organization and the International Society of Urologic Pathologists published a consensus classification for urothelial malignancies that has been accepted as the standard classification schema.⁸⁵ According to this schema, urothelial malignancies are classified as either low- or high-grade depending on the degree of anaplasia and architectural features. The vast majority of invasive cancers are high-grade. Noninvasive lesions are either flat or papillary.

CIS on cystoscopy often appears as multiple, flat, erythematous patches without a gross structural component. Histologic examination reveals an increase in the number of cell layers, lack of cohesion between cells that results in shedding into the urine, disordered architecture, and an increased number of mitotic figures. The basement membrane confines the tumor to the mucosa. CIS associated with an invasive tumor has been associated with a worse prognosis. Papillary lesions have exophytic stalks protruding into the lumen of the bladder. Invasive lesions show involvement of the muscularis propria layer and are often palpable on examination.^86,87 The predominant histologic findings associated with invasive carcinomas are transitional cell or urothelial carcinomas. Transitional cell carcinomas often contain squamous or glandular components and are classified as transitional tumors with a natural history similar to those of pure histology. Immunohistochemical analysis of microvessel density has been shown to be an independent prognostic indicator demonstrating a relationship between tumor angiogenesis, nodal metastasis, and survival.^88,89

Pure squamous cell carcinomas comprise 5% to 8% of primary bladder malignancies in the United States and Europe; however, in Egypt and other areas endemic with Schistosoma haematobium, squamous cell tumors are the predominant histologic finding. Most squamous cell carcinomas are bulky, polypoid, solid masses and often present at an advanced stage. Necrotic material and keratin debris on the surface is typical. Pure adenocarcinomas account for 1% to 2% of bladder tumors. Most often, these tumors are derived from the urothelium of the bladder (“nonurachal adenocarcinoma”) and less frequently from a urachal remnant (“urachal adenocarcinoma”). They predominantly arise in the base of the bladder or trigone, but may arise at any location.^87,90 A variety of rare histologic findings account for less than 2% of all primary bladder cancers. Extrapulmonary small cell carcinoma can occasionally originate in the bladder and, because of its aggressive and high metastatic potential, it is often treated with chemotherapy generally used for small cell carcinomas of the lung in addition to local therapy.⁹¹ These patients may be younger (median 40-60 years); however, similar to other histologic findings, small cell carcinoma of the bladder is associated with smoking.⁹² Because of the rarity of this finding, there are no prospective trials designed to define the best management of these patients. Poor long-term survival has been seen with single-modality therapy; therefore, combined-modality regimens are typically used.⁹² Rhabdomyosarcoma of the bladder occurs in children but rarely presents in adults.⁹³ Other rare histologic findings include lymphoma, leiomyosarcoma, melanoma, and primary bladder pheochromocytoma.^94–97 In addition, adenocarcinomas can metastasize to the bladder from the colon, rectum, endometrium, prostate, ovary, breast, and lung.⁹⁸

Clinical Presentation and Routes of Spread

Gross or microscopic hematuria is a presenting sign in 75% of patients with bladder cancer. Approximately 25% of bladder cancer patients present with symptoms of bladder irritation (i.e., urinary frequency, urgency, dysuria), especially patients with CIS. Typically, symptoms are intermittent and diagnosis is often delayed as these may be similar to benign disorders such as cystitis, prostatitis, or bladder stones. Advanced tumors may present as pelvic pain, ureteral obstruction, hydronephrosis, or rectal obstruction.^99,100

The regional lymph nodes at risk as defined by the current American Joint Committee on Cancer (AJCC) 2002 staging system are those in the true pelvis below the bifurcation of the common iliac arteries. These lymph nodes include the hypogastric, obturator, internal and external iliac, perivesical, pelvic, sacral, and presacral chains. Staging depends on the size and number of involved nodes, but not on whether they are unilateral or ipsilateral (Table 44-1). Spread to the common iliac nodes is considered a manifestation of distant metastases and not regional disease.¹⁰¹ Recent surgical series, however, have demonstrated long-term disease control in patients undergoing cystectomy with positive lymph nodes from extended lymph node dissections including the periaortic, common iliac, and presacral chains (see Surgical Treatment of Muscle-Invasive Bladder Cancer).^102,103 The extent of nodal dissection currently remains controversial and the staging system has not yet been altered. Skinner and colleagues¹⁰⁴ correlated the depth of bladder wall invasion to involvement of lymph nodes. In their study, 5% of patients with pT₁, 30% with pT₂, 31% with pT_3a, 64% with pT_3b, and 50% with pT₄ disease had nodal metastases. The most common sites of distant metastatic disease are the lung, bone, and liver.

Diagnostic and Staging Studies

The initial formulation of an organized staging system reproducible in predicting survival dates back to the work of Jewett and Strong in the 1940s. Primary bladder tumors were classified as stage A when involvement was limited to the submucosa, as stage B when tumor penetrated into the muscularis, and as stage C when there was evidence of perivesical fat involvement. Stage A tumors had no evidence of metastases compared with 13% of stage B tumors and 74% of stage C tumors. In 1951, Jewett proposed a division of stage B into B1 and B2 to correspond with superficial and deep muscle penetration. The following year, Marshall added a stage 0 to include tumors not infiltrating the lamina propria and stage D for metastatic disease.¹⁰⁵

The current tumor-node-metastasis (TNM) staging system standardized by the AJCC and the TNM Committee of the International Union Against Cancer continues to reflect the importance of depth of bladder wall invasion. The T stands for depth of tumor in relation to the mucosa, muscularis, and the perivesical fat. Regional nodes include those of the true pelvis, whereas M disease involves nodes above the bifurcation of the common iliac arteries or distant visceral metastasis.¹⁰¹

The current TNM staging system (see Table 44-1) recognizes that T_a papillary growths convey a significantly better prognosis than T_is (CIS) or T₁ lesions. Also recognized is that either microscopic (T_3a) or gross involvement (T_3b) of perivesical fat is more important than the distinction between superficial (T_2a) or deep muscle invasion (T_2b). Stage T₄ includes locally advanced tumors that invade the prostate, uterus, or vagina (T_4a) or directly adhere to the pelvic or abdominal wall (T_4b).¹⁰¹

Bladder cancer may be staged clinically (cTNM) or pathologically (pTNM). Clinical staging is defined by the extent of tumor on physical examination, bimanual examination under general anesthesia before and after cystoscopic biopsy or transurethral resection, urine cytologic study, biochemical studies, and radiographic imaging.

The presence of gross hematuria in a patient older than 40 should raise suspicion for urothelial carcinoma until proven otherwise, particularly in a patient with known risk factors. Approximately 20% of all patients presenting with gross hematuria and 5% to 10% of all patients with microscopic hematuria will ultimately be diagnosed with bladder cancer.^106,107 A history and physical examination should be followed by a urinalysis with urine cytologic study to detect the presence of suspicous cells. A cystoscopic examination includes documentation on a bladder diagram of tumor location, number of lesions, pattern of growth, and tumor size. A bimanual examination should be performed before and after endoscopic resection and is an important part of clinical staging. A mobile bladder mass indicates a T₃ tumor, whereas a fixed mass suggests T₄ disease. A transurethral resection of a bladder tumor (TURBT) is diagnostic and often therapeutic in superficial lesions. Muscle must be seen by the pathologist in a biopsy or transurethral resection for it to be an adequate indicator of prognosis. Biopsy specimens of surrounding and apparently unaffected areas, including the prostatic urethra, are taken because CIS may be visibly indistinguishable from normal mucosa. Persistence of a palpable mass on examination under anesthesia after TURBT identifies a poor prognostic group, usually with T_3b disease.¹⁰⁸ Patients with muscle-invasive disease require a screening workup for metastatic disease that includes chest imaging by either computed tomography (CT) or plain films as well as pelvic imaging to identify nodal involvement using CT or magnetic resonance imaging (MRI) of the pelvis. Imaging should include the abdomen if there is pelvic nodal disease or abnormal results of liver biochemistry studies. The upper urinary tract should be imaged with intravenous urography or contrast CT urogram to rule out synchronous upper tract lesions. In patients with invasive or locally advanced disease or in patients with elevation of the serum alkaline phosphatase level or bone pain, a bone scan should be performed to rule out metastases.^108,109

Imaging with CT or MRI has an important role in defining the extent of disease for locally advanced bladder lesions. Mural and mucosal lesions are distinguishable from perivesical fat and urine, allowing visualization of papillary and extravesical lesions (Fig. 44-4). Important diagnostic information that can be demonstrated with CT is the presence of gross perivesical extension or lymphadenopathy.¹¹⁰ Understaging of local disease, however, remains an important problem with clinical staging evaluations. Although not always as available as CT, MRI may be more accurate in determining the T stage of bladder tumors.¹¹¹ MRI has many advantages over other modalities for detecting and staging bladder neoplasms because of its intrinsic high soft tissue contrast, direct multiplanar imaging capabilities, and the availability of a non-nephrotoxic, renally excreted contrast agent. Tachibana and coworkers¹¹² correlated imaging with histologic staging for 57 bladder tumors. The sensitivity and specificity for differentiating superficial and invasive tumor were 96% and 83% for gadolinium-enhanced MRI, 96% and 58% for CT, and 88% and 67% for transurethral ultrasonography. The contrast between tumor and fat is critical for evaluating locally advanced lesions and is best detected on T1-weighted images. It is important to recognize in the post-treatment management of the patient that changes in signal intensity of normal bladder wall 3 to 4 months after irradiation limit the usefulness of MRI in detecting tumor recurrences.¹¹³ Reports indicate improved accuracy in staging muscle-invasive tumors when MRI is performed with gadolinium-enhanced images. Tanimoto and associates¹¹⁴ showed that for 86 tumors, MRI with gadolinium enhancement had a staging accuracy of 85%, in comparison with 55% for CT scan and 58% for conventional MRI. Detection of superficial lesions was improved and there was a decrease in the overstaging of T1 tumors. Sensitivity and specificity of MRI for staging bladder tumors have been improved with the advent of imaging with oblique planes, smaller pixel size, and three-dimensional (3-D) technology. Applying 3-D technology to MRI was associated with improved accuracy of local staging from 78% to 93% and nodal staging from 86% to 93%.^113,115,116 The detection of nodal involvement with clinical staging has obvious treatment-planning implications for the radiation oncologist. One report from the Christie Hospital in the United Kingdom demonstrated that preradiotherapy staging with MRI was helpful in providing more anatomic detail and clarifying the extent of disease. In that series, 39 of 71 patients with clinical T_2a tumors were upstaged. Among the 27 patients in that series upstaged to T_3b, this group experienced a significantly worse local control and overall survival outcomes. A multivariate analysis demonstrated that, among other variables, tumor stage based on MRI was a significant factor predicting for treatment failure and overall survival.¹¹⁷ MRI may be particularly helpful in the evaluation of direct prostatic extension of tumor (T_4a).¹¹⁸

Figure 44-4 • Axial T1-weighted images demonstrate a large left-sided bladder mass with extravesical extension into the perivesical fat. Axial T2-weighted images show the mass to be of intermediate signal intensity, but darker than the urine-filled bladder and displacing it to the right.

The role of fluorodeoxyglucose (FDG)–positron emission tomography (PET) remains controversial in patients with superficial and locally advanced bladder cancer. The presence of radiotracer in the urine makes interpretation of early stage disease problematic. For the early detection of locally advanced bladder cancer with lymph node involvement, FDG-PET shows promise. FDG-PET imaging prior to cystectomy had a sensitivity of 67% and a specificity of 86% for detection of lymph node metastases, which compares favorably with conventional CT.¹¹⁹ FDG-PET appears able to detect metastatic lesions with high frequency as well.¹²⁰ The full role of FDG-PET and other PET technology has yet to be realized in bladder cancer.

Pathologic staging requires either a partial or total cystectomy and lymph node dissection. Suspected metastatic disease on imaging studies can often be confirmed by CT-guided biopsies. Staging error between clinical and pathologic stages has been shown to be as high as 50%.¹⁰¹ Survival is lower for each stage when grouped clinically as opposed to pathologically. It is for this reason that comparing outcomes of radiotherapy series based on clinical staging with surgical series based on pathologic staging is fraught with difficulties. Nevertheless, with both modalities, survival decreases with increasing stage. This provides the clinician with the necessary prognostic information to formulate a treatment approach.¹⁰⁸

Identification of prognostic factors allows the clinician to select those patients who would potentially benefit from bladder conservation approaches or, conversely, patients better managed with upfront radical cystectomy. Advanced-stage disease has been shown to be an accurate predictor for a high risk of developing early distant metastases.¹⁰⁹ In addition to the depth of invasion and the presence or absence of nodal involvement reflected in the stage, other prognostic variables have been defined. Low-grade and papillary growth have been shown to represent good prognostic indicators for survival.¹²¹ Tumors characterized as larger than 5 cm, with associated CIS, with evidence of vascular invasion, and with failure to obtain a complete response to chemoradiation have been shown to have a poor outcome and these patients may be better suited for cystectomy.¹²² The presence or absence of hydronephrosis has also been correlated with prognosis in both cystectomy and radiation therapy series.^123,124 Researchers have indicated that a tumor with a high degree of angiogenesis may have a poorer prognosis.⁸⁹ It is likely that in the future prognostic factors will be increasingly identified at the cytogenetic level.

Treatment of Superficial Bladder Carcinoma

Three fourths of bladder carcinomas present as T_a, T₁, or T_is superficial lesions that do not extend deeper than the lamina propria. Although the majority of such tumors are controlled by transurethral resection, an estimated 40% to 80% will recur within 6 to 12 months and 10% to 25% will progress to muscle-invasive disease. Prognostic factors for recurrence include grade, depth of penetration, multifocality, and presence of CIS. The management of superficial bladder carcinoma involves TURBT and, in many cases, adjuvant therapy to prevent recurrence and potential development of muscle-invasive disease.^99,125,126

T_a papillary tumors typically present as low-grade lesions that frequently recur multiple times prior to progressing to invasive tumors. Low-grade T_a papillary lesions without multifocality and without associated CIS can be treated with a diagnostic and therapeutic resection using a resectoscope, followed by close monitoring and repeat cystoscopies. Muscle must be seen in this specimen before ruling out invasive disease and biopsies of apparently uninvolved urothelium should be obtained to rule out occult T_is. Approximately 5% to 30% of patients will ultimately progress to a higher-grade lesion or require more aggressive therapy.^127,128 For patients with persistent tumor cells on urine cytology after TURBT, multifocal lesions, grades II or III histologic findings, or evidence of CIS, adjuvant intravesical therapy is indicated. If all visualized superficial lesions are not excised, there is also a need for further therapy. Adjuvant therapy is given in the form of intravesical administration of immunotherapy or chemotherapy. Bacille Calmette-Guérin (BCG) is administered adjuvantly weekly for 6 weeks. BCG, a live attenuated form of Mycobacterium bovis, has an unknown exact mechanism of action; however, it triggers an immune-response cascade including mononuclear cells and cytokines such as interleukin (IL)-1, IL-2, IL-6, IL-8, interferon-γ, and tumor necrosis factor-α. In addition, a direct tumor-cell suppression may be involved. The chemotherapeutic agents of greatest known efficacy are mitomycin C, doxorubicin, and thiotepa. These three agents have all demonstrated efficacy in the reduction of disease recurrence.¹²⁶ Thiotepa causes the most systemic toxicity, including myelosuppression in 20% of patients, owing to its low molecular weight, which allows absorption. Doxorubicin creates the greatest local reaction, which may predispose the patient to urinary urgency. The three drugs have nearly identical efficacy in prolonging the time to recurrence.⁹⁹

A randomized trial of intravesical doxorubicin versus BCG in 262 patients with superficial transitional cell carcinoma showed an advantage in the complete response and 5-year disease-free survival rates for those treated with adjuvant BCG.⁹⁹ A classic series of 86 patients with recurrent superficial bladder cancer prospectively randomized to TURBT plus BCG or TURBT alone was reported by Herr and colleagues.¹²⁹ Patients treated with BCG had a prolonged median disease-free interval of 24 months compared with 8 months for those patients treated with TURBT alone.¹²⁹ An update of the data has revealed a 10-year disease-specific survival rate of 75% in the group that received TURBT plus BCG compared with 55% in those treated with TURBT alone. The majority of recurrences were discovered within the first 5 years. Of the 61 patients who received BCG, 33 (54%) were disease free with their bladders preserved.^130,131

In a recent review of 11 clinical trials consisting of 2749 patients treated with BCG or mitomycin C, a greater incidence of tumor recurrence was observed for patients treated with mitomycin C compared with BCG (46% versus 36%, respectively). This review also demonstrated improved results with BCG maintenance therapy for reducing the risk of tumor recurrence; however, this is currently controversial because of the potential increased toxicity with BCG maintenance. The incidence of associated cystitis related to intravesical therapy was more prevalent among BCG treated patients (54% versus 39%, respectively).¹³²

The standard of care for superficial bladder cancer with a high risk of recurrence has been established as intravesical BCG. Histologic persistence after 6 months indicates the need to change the intravesicular agent. Disease beyond 1 year indicates the need for a different treatment approach to prevent progression to muscle-invasive cancer. Investigators are testing the application of interferon α.^126,133 The role of external-beam radiation in the treatment of CIS is unclear. Especially among patients previously treated with multiple resections and intravesical therapy, the tolerance of external irradiation may be poor because of the pre-existing chronic irritation within the bladder mucosa. Van der Werf-Messing et al.¹³⁴ have reported excellent control rates (>90%) for limited T₁ tumors using interstitial therapy through an open cystotomy approach. In that report, interstitial radium-226 was used with an intended dose of 60 Gy delivered to the visualized bladder tumor. This approach, however, is generally not used in the United States. Patients with persistent CIS or refractory T_a/T₁ disease ultimately require cystectomy.

Surgical Treatment of Muscle-Invasive Bladder Cancer

The current standard of care for muscle-invasive bladder cancer in the United States is radical cystectomy with bilateral pelvic lymph node dissection. The use of neoadjuvant or adjuvant chemotherapy is still controversial; however, recently reported studies have demonstrated improved survival with neoadjuvant chemotherapy (see Neoadjuvant/Adjuvant Chemotherapy). In the United Kingdom and Europe the prevailing primary treatment is definitive external-beam radiation therapy with the option of salvage cystectomy.

Radical cystectomy in the male implies the en bloc removal of the bladder (with its peritoneal covering), perivesical adipose tissue, lower ureters, prostate gland and seminal vesicles, pelvic vas deferens, proximal urethra, and pelvic lymph nodes (Fig. 44-5). A total urethrectomy is performed for patients with CIS, multicentric tumors, and involvement of the bladder neck or prostatic urethra. A radical cystectomy in female patients is an anterior exenteration with sacrifice of the bladder (with its peritoneal covering), entire urethra, uterus, fallopian tubes, ovaries, anterior vaginal wall, and pelvic lymph nodes. A standard pelvic lymph node dissection generally includes removal of the bilateral external iliac, obturator, internal iliac (hypogastric), and common iliac lymph node chains. The dissection extends from the aortic bifurcation superiorly to the inguinal ligament and node of Cloquet inferiorly and the genitofemoral nerve laterally.^135,136 An extended dissection includes all nodes from the standard template plus paracaval, interaortocaval, para-aortic, and presacral lymph nodes. Herr reported on the prognostic effect of nodal status in 637 patients who underwent radical cystectomy and pelvic lymph node dissection.¹³⁷ For both node-negative and node-positive patients, improved survival and reduced local recurrence related to negative surgical margins and the number of lymph nodes removed at the time of the node dissection. The 5-year survival rates for patients who had 0 to 5, 6 to 10, 11 to 14, and more than 14 lymph nodes removed were 79%, 73%, 44%, and 33%, respectively. The corresponding local relapse rates in these aforementioned nodal groups were 4%, 7%, 8.5%, and 17%, respectively. A multivariate analysis demonstrated that the number of nodes removed was an independent predictor of survival outcome. This observation has been observed in other series and clinical trials.^102,103,138 One caveat, however, is that it is not clear if this improved outcome is related to therapeutic removal of occult nodal disease or if it represents a surrogate for the quality of surgery.

Figure 44-5 • Multifocal bladder tumor with papillary growth patterns. The cystectomy specimen includes en bloc removal of the bladder, lower ureters, prostate, seminal vesicles, proximal urethra, and pelvic lymph nodes.

Attempts to reduce the morbidity of radical cystectomy have focused on preserving potency in the male and developing urinary diversion alternatives that maximize continence and cosmesis. A nerve-sparing radical cystectomy requires preservation of the neurovascular bundles and the dorsal-venous complex surrounded by the lateral prostatic fascia located posterolateral to the prostatic capsule and membranous urethra.^136,139

Reconstructive techniques following radical cystectomy generally fall into two categories: incontinent or continent diversions. The standard urinary diversion, an incontinent diversion, is accomplished with a conduit derived from 15 cm of distal ileum to which the ureters are anastomosed (Fig. 44-6). An ostomy is created by attachment of the distal end of the ileum to the defect in the anterior abdominal wall.¹³⁶ More recently, a second type of urinary diversion, the creation of internal reservoirs, has become available to allow the patient to retain continence. Two classes of continent diversions of the urinary flow are (1) stomal reservoirs that require intermittent catheterization and (2) orthotopic neobladders anastomosed to the remaining distal urethra (Fig. 44-7). There are several techniques for creating a continent cutaneous diversion that can be catheterized including the cecoileostomy (Indiana pouch [Fig. 44-8]) and the ileostomy (Koch pouch). All involve constructing a detubularized pouch from intestine that becomes a spherical reservoir with a constructed sphincter to maintain continence. Because these reservoirs are composed of portions of bowel, metabolic and absorptive disturbances vary depending on the segments and lengths of bowel used.

Figure 44-6 • Ileal conduit reconstruction.

(From reference 140.)

Figure 44-7 • Orthotopic neobladder reconstruction.

(From reference 140.)

Figure 44-8 • Continent cutaneous diversion (Indiana pouch).

(From reference 140.)

Because of the longer operative times and complexity of these procedures, a higher risk of acute postoperative complications including urinary leak are noted as compared with ileal conduits. By 3 to 4 months the average catheterization interval is decreased to 6 to 8 hours. Nevertheless, the pouch is associated with a 2% mortality, a 5% incidence of serious complications, and a reoperative rate of 30%. Over the past decade, orthotopic neobladder reconstruction techniques have become the preferred method of urinary diversion in appropriately selected patients. Because this reconstruction leaves intact the external urinary sphincter, patients are able to void by performing the Valsalva maneuver. Patients must be carefully chosen for the orthotopic neobladders to avoid incontinence, particularly during sleep, because the guarding reflex is lost with the native bladder. Rates of nocturnal incontinence range from 15% to 40%.¹⁴⁰ Five-year survival rates for early stage T₂ tumors were often reported to be as low as 17% to 36% before 1980. Patients with stage T_3a and T_3b tumors had 5-year rates ranging from 18% to 31% and 20% to 25%, respectively. However, more recent studies of cystectomy using modern surgical techniques report survival rates as high as 50% to 88% for T₂, 36% to 69% for T_3a, and 11% to 47% for T_3b tumors.^135,141 A summary of recently reported outcomes in large surgical series is summarized in Table 44-2. The overall local control rates with radical cystectomy alone have been reported to be as high as 95%^142,143; however, in particular, patients with locally advanced or node-positive disease treated with radical cystectomy alone do not fare as well. Madersbacher et al.¹⁴⁴ reported in a recent series of patients treated with radical cystectomy alone isolated local recurrence rates of 9% to 20% for non–organ-confined (T_3b-T₄) node-negative disease and 13% for node-positive disease. Another surgery-only series reported by Hautmann et al.¹⁴⁵ demonstrated local recurrence rates of 16% and 20% in non–organ-confined and node-positive patients, respectively. Of note, this is likely an underestimate of the true local recurrence rates as patients with simultaneous local and distant recurrences are typically included in the metastatic cohort. The overall outcomes of these patients remain suboptimal even in series including neoaduvant and adjuvant therapies. Distant metastases occur in 20% to 50% of patients with locally advanced disease.^144,145 For patients with node-positive disease, long-term disease-free survival rates are poor at 25% to 35%.

Preoperative Radiation Therapy

In an effort to improve the outcome for patients with locally advanced bladder cancers treated with surgery, preoperative radiotherapy was commonly used from 1960 to 1980. The rationale for using this approach was to reduce the incidence of tumor bed failures and potential seeding of malignant cells from the operation. Although some retrospective studies have suggested that the addition of preoperative radiotherapy improves the overall outcome,^146,147 other series have found no benefit when results are compared with modern surgical series.^142,148,149 Several randomized prospective trials that compared preoperative radiotherapy plus radical cystectomy to radical cystectomy alone have demonstrated no significant differences between the treatment arms.^150–152 Nevertheless, it must be pointed out that these trials were performed before the improvement in surgical techniques and suffered as well from limited accrual, compromising their power to detect significant differences between the study arms. A review from the M.D. Anderson Hospital supports the notion that preoperative irradiation improves the outcome in a select cohort of patients.¹⁵³ In this report, 338 patients treated with preoperative radiotherapy between 1960 and 1983 were retrospectively compared with 232 patients treated at the same institution between 1985 and 1990. The majority of the latter group were treated with neoadjuvant or adjuvant chemotherapy. Despite this apparent advantage, there was a significant improvement in local control among patients with T_3b disease who received preoperative irradiation. Within this cohort of patients, there was a trend for improved disease-free survival and overall survival, although no significant differences were seen. The use of preoperative radiotherapy in the modern surgical era remains unclear, and at this time is not a standard approach in the United States.

Neoadjuvant/Adjuvant Chemotherapy

Survival in muscle-invasive bladder cancer is a function of stage. The stage-dependent reduction in survival is believed to be due to the presence of micrometastases at the time of cystectomy or radiation therapy. Thus long-term survival depends on prevention of distant metastasis. Several studies have demonstrated a survival advantage with neoadjvuant chemotherapy prior to definitive local therapy for muscle-invasive transitional cell carcinoma.

Bladder cancer is a chemotherapy-sensitive cancer. A wide variety of agents with varying mechanisms of action all have activity against bladder cancer. In metastatic disease, objective response proportions of 50% to 70%, and complete response rates of 20% to 30%, have been reported (see Chemotherapy for Metastatic Disease). This chemosensitivity provides the rationale for employing early systemic chemotherapy (adjuvant and neoadjuvant) in an effort to reduce the incidence of distant metastases to improve long-term survival. The success of adjuvant and neoadjuvant chemotherapy in other cancers to eradicate micrometastases has led to the application of these strategies in bladder cancer.

Neoadjuvant (pre-) treatment offers the advantage of in vivo drug sensitivity testing, which may provide prognostic information. In addition, preoperative chemotherapy has the advantage of shrinking and down-staging tumors, potentially allowing for easier surgery and raising the potential of a bladder-sparing approach. Furthermore, neoadjuvant chemotherapy delivers full-dose systemic chemotherapy, with the intent of allowing early treatment of micrometastases and reducing the incidence of subsequent metastatic disease.

Various neoadjuvant regimens have been evaluated in a number of prospective randomized trials. The Nordic Cystectomy I trial evaluated neoadjuvant chemotherapy in combination with low-dose irradiation and cystectomy.¹⁵⁴ Patients with T_1G3, T₂-T_4a N_XM₀ tumors were randomly assigned to either two cycles of chemotherapy with cisplatin and doxorubicin or no chemotherapy before cystectomy. In addition, all patients in this study received radiotherapy (4 Gy/day × 5 days) following the second cycle of chemotherapy. Initial results appeared to show a benefit for neoadjuvant chemotherapy. However, longer follow-up failed to demonstrate a statistically significant difference in overall survival between the two groups (59% versus 51%, P = 0.1).¹⁵⁵ Subgroup analysis also demonstrated no survival benefit for patients with T₁ and T₂ tumors, although patients with T₃ and T_4a tumors did show a 5-year survival benefit of 52% versus 37% (P = 0.03).

The largest neoadjuvant trial to date was a European study involving more than 900 patients. Between 1989 and 1995, patients with T_2G3, T₃, and T_4a, N₀ to N_X, M₀ tumors were randomized to receive either three cycles of cisplatin, methotrexate, and vinblastine (CMV) neoadjuvant chemotherapy (n = 491) or no chemotherapy (n = 485) before planned cystectomy or definitive radiation therapy (chosen at the discretion of the local treating institution).¹⁵⁶ Chemotherapy-associated mortality was 1% and cystectomy operative mortality was 3.7%. Approximately 42% of patients in each arm received radiotherapy as the primary therapy. Although at 4 years no significant difference in overall survival was seen, after longer follow-up (median 7.4 years), a statistically significant 6% absolute survival benefit was found in favor of the chemotherapy-treated patients.¹⁵⁷ In addition, the rate of pathologic complete response was found to be dramatically higher in the primary tumors of chemotherapy-treated patients (33% versus 13%).

A similar smaller randomized phase III trial of neoadjuvant methotrexate, vinblastine, doxorubicin (Adriamycin), and cisplatin (MVAC) plus cystectomy (n = 153) versus cystectomy alone (n = 154) was conducted by the U.S. Intergroup in patients with T₂ to T_4a N₀M₀ bladder cancer. After a median follow-up of 8.5 years, survival in the MVAC arm was marginally superior, with a hazard ratio of 0.74 (95% CI 0.55-0.99, P = 0.06). The estimated median survival of the MVAC arm was 77 months, while the no-MVAC arm was 46 months.¹⁵⁸

Two meta-analyses have been conducted to evaluate the effect of neoadjvuant chemotherapy on overall survival. One meta-analysis of individual patient data consisting of 98% of patients from known randomized trials using cisplatin combination chemotherapy has shown that neoadjuvant therapy provides a moderate but significant benefit in patients with T₂-T_4a disease. Patients who received neoadjuvant cisplatin combination chemotherapy experienced a 14% reduction in the risk of death (hazard ratio = .86 [CI = .70-.95]) and an absolute improvement in survival of 5% at 5 years.¹⁵⁹ Single-agent cisplatin chemotherapy was not shown to provide any survival benefit. A second meta-analysis was conducted that confirmed these results.¹⁶⁰ Neoadjuvant cisplatin combination chemotherapy prior to definitive local chemoradiotherapy has been tested in only one study, and did not appear to demonstrate a benefit in favor of chemotherapy. Several other neoadjuvant studies allowed either radiotherapy or cystectomy as local therapy following neoadjuvant chemotherapy. All of these studies have been included in the meta-analyses cited previously, and there was no significant association between the type of local therapy (cystectomy versus radiotherapy) and outcome.

Overall, data from neoadjuvant chemotherapy trials for bladder cancer suggest that the pT₀ rate is approximately 30% to 40% in patients receiving combination neoadjuvant cisplatin-based chemotherapy. Unfortunately, this pT₀ rate has not translated into a dramatic survival advantage for neoadjuvant chemotherapy. The fact that neoadjuvant therapy achieves high rates of pathologic complete responses out of proportion to the clinical benefit suggests that distant metastasis is an early event in locally advanced bladder cancer that is only modestly affected by chemotherapy. Older regimens for neoadjuvant treatment (e.g., single-agent cisplatin; cisplatin, cyclophosphamide, and doxorubicin [Adriamycin] [CISCA]) have not been proven to reduce the risk of death. However, modern regimens (MVAC and CMV) have shown a potential modest improvement in survival (6%-15% at 5 years). The incorporation of biologic agents with neoadjuvant chemotherapy may improve these outcomes.

Neoadjuvant chemotherapy is generally well tolerated, and does not appear to increase the rate of surgical complications. In metastatic disease, the use of gemcitabine/cisplatin (GC) rather than MVAC has become common because of its more favorable toxicity profile and similar activity to MVAC (see Chemotherapy for Metastatic Disease), although GC has not yet been formally tested in the neoadjuvant or adjuvant setting. The use of carboplatin-based regimens has increased in practice, although there is no data to support their use in this setting.

In contrast to neoadjuvant chemotherapy, adjuvant (postoperative or postradiation) chemotherapy allows for earlier local therapy (cystectomy or radiotherapy) without the delay of preoperative chemotherapy, resulting in immediate local therapy and tumor debulking. In addition, after cystectomy, risk stratification of patients based on pathologic examination of cystectomy specimens can guide subsequent adjuvant treatment. Therefore, only higher-risk patients (T₃-T_4a or node-positive) would be considered for chemotherapy.

There is no randomized data regarding the use of adjuvant chemotherapy following primary radiotherapy or chemoradiation for advanced bladder cancer. Several phase II studies of chemoradiation have been reported that include adjuvant combination chemotherapy. These studies have promising response rates and survival data. Radiation Therapy Oncology Group (RTOG) 97-06 tested three cycles of methotrexate, cisplatin, and vinblastine (MCV) chemotherapy in complete responders following induction and consolidation chemoradiation using concurrent cisplatin. Only 45% of patients completing chemoradiation were able to complete all three cyles of MCV adjuvant chemotherapy, although overall results of the strategy seemed promising.¹⁶¹ Another study evaluated two cycles of concurrent 5-fluorouracil (5-FU) and cisplatin during radiotherapy followed by two additional cycles of adjuvant 5-FU and cisplatin. Protocol therapy on this trial was completed by only 57% of patients enrolled.¹⁶² However, 5-year survival from both of these studies was 45% to 49% in patients who were cystectomy candidates, comparing favorably with large cystectomy series. Other phase II studies are underway to test the use of newer agents such as the taxanes and gemcitabine in this setting.

Several prospective randomized trials have attempted to address whether adjuvant chemotherapy improves survival after cystectomy. An early trial using single-agent cisplatin was unable to detect any advantage for adjuvant therapy.¹⁶³ Subsequently, a prospective randomized trial using combination chemotherapy was undertaken in patients with pT₃, pT₄, or node-positive transitional cell carcinoma of the bladder. Between 1980 and 1988, 91 patients were randomly assigned to adjuvant chemotherapy or observation. CISCA was the predominant chemotherapy administered on this trial, although some patients received other chemotherapy agents and regimens. This trial showed a significant increase in time to progression in favor of adjuvant chemotherapy (6.58 versus 1.92 years, P = 0.011), although overall survival was not significantly different at 5 years. In retrospective subgroup analyses, patients with a single positive lymph node appeared to derive the most benefit from adjuvant chemotherapy. Interestingly, patients with two or more positive lymph nodes did not appear to benefit from adjuvant chemotherapy.¹⁶⁴ This study has been criticized for the lack of standard treatment, as well as the failure to receive assigned therapy in a significant proportion of the patients.

In a study from Stanford University, 55 patients who underwent cystectomy and were found to have pT₃ and T₄ tumors with or without lymph node involvement were randomized to four cycles of CMV chemotherapy or observation. This study was halted early because of the observation of significant benefit in favor of patients receiving chemotherapy. With a median follow-up of 62 months, freedom from progression was superior in patients receiving adjuvant chemotherapy (37 versus 12 months, P = 0.01), although overall survival was not significantly different (63 versus 36 months, p = 0.32). Patients who developed progressive disease and who had not received adjuvant therapy subsequently received systemic therapy, perhaps accounting for the fact that overall survival was not significantly improved by the use of adjuvant therapy.¹⁶⁵

A third prospective randomized trial involved 49 patients with pT_3b, pT_4a, or pelvic lymph node involvement without any evidence of tumor remaining after cystectomy. These patients were randomized to three cycles of adjuvant MVAC or methotrexate, vinblastine, epirubicin, cisplatin (26 patients) or observation (23 patients). Only 18 of the 26 patients randomized to adjuvant chemotherapy actually went on to receive adjuvant chemotherapy. The study was designed to show a 35% improvement in disease-free survival at 5 years. However, an interim analysis showed a significant increase in disease-free survival among patients in the chemotherapy arm, resulting in early stopping of the study. In an intent-to-treat analysis, at 3.5 years, 63% of patients in the chemotherapy arm were disease free, compared with 13% in the observation arm. However, overall survival was not significantly improved in this study. The benefit of adjuvant therapy appeared strongest for those patients with lymph node involvement.¹⁶⁶

Because of the lack of statistical power and early closing of trials, it is difficult to reach conclusions regarding the true efficacy of adjuvant chemotherapy for advanced bladder cancer. A meta-analysis of data from six randomized controlled clinical trials of cisplatin-based adjuvant chemotherapy evaluated updated patient data for 491 patients. This analysis was limited in its power because of the relatively small sample size, but suggested that adjuvant chemotherapy was associated with a 25% reduction in the risk of death compared with controls.¹⁶⁷

Taken together, these data suggest that in the appropriate patients (T₃, T₄, or node-positive), adjuvant therapy improves disease-free survival. The question of whether adjuvant chemotherapy confers an overall survival benefit remains unanswered. Chemotherapy administered at the time of relapse with metastatic disease is quite effective, but rarely results in long-term remissions. This activity at the time of relapse may explain the apparent lack of efficacy of adjuvant strategies in improving overall survival. Studies evaluating early (adjuvant) versus delayed chemotherapy using MVAC, dose-dense MVAC, or GC have been closed because of poor accrual; the true benefit of adjuvant chemotherapy for advanced bladder cancer may remain unknown for some time.

Definitive Radiotherapy for Muscle-Invasive Bladder Cancer

Radiation therapy has been used in the management of bladder cancer for more than 50 years.¹⁶⁸ As a single modality, however, results were disappointing, with high local failure rates and poor survival.^169,170 In addition to suboptimal radiation techniques, poor patient selection and lack of systemic therapy to address micrometastatic disease contibuted to these poor results. Proper patient selection and the use of chemotherapy in combination with modern radiation therapy have improved these results considerably. There are, however, no reported modern prospective randomized studies directly comparing definitive chemoradiation with surgery. An ongoing randomized trial in the UK is designed to address this issue (Selective Bladder Preservation against Radical Excision trial) (public.ukcrn.org.uk). Patients offered definitive irradiation have typically been nonsurgical candidates with various comorbidities and more advanced disease, contributing to poorer prognoses than those selected for surgery. In addition, surgical series have the benefit of accurate pathologic staging, whereas radiation series rely on clinical staging with its intrinsic inaccuracies related to understaging. Comparison of cystectomy to radiotherapy series must be interpreted with the understanding that there are inherent selection-bias and stage-migration differences between these groups of patients as well as the differences in overall performance status. The primary goal of bladder preservation strategies remains disease control, with organ preservation a secondary albeit important treatment goal, and hence, depite these difficulties with data comparison, radical cystectomy outcomes remain the standard with which other therapies should be compared. In Great Britain, Canada, and Europe, external-beam radiotherapy is often first-line therapy, with salvage cystectomy reserved for treatment failures.

The 5-year survival rate for muscle-invasive bladder cancer treated with definitive external-beam radiotherapy alone is 20% to 40%. Most series report complete response rates of approximately 50%. Similar to surgical series, survival rates following definitive radiotherapy have been higher in patients with invasion limited to the bladder compared with those with extravesical (T_3b or T₄) disease.^142,171,172 Analysis of clinical characteristics has defined several prognostic factors that predict for local control, freedom from metastasis, and survival in patients treated with definitive irradiation. Most series show that depth of tumor invasion defined by stage is an important prognostic indicator.^172–175 Patients with stage T₄ disease have been shown to have 5-year survival rates often lower than 10%, indicating irradiation alone is likely to benefit only a minority of those with advanced disease. In addition to advanced T stage, other poor prognostic signs include tumors greater than 5 cm, residual disease after TURBT, and radiographic evidence of ureteral obstruction with hydronephrosis.¹⁷⁵ A multivariate analysis of 116 patients treated at Fox Chase Cancer Center showed that both local control and survival depended on the pretreatment clinical T stage and hemoglobin levels.¹⁷³ Grade was also shown to influence prognosis, with a 5-year survival of 43% in patients with grade I or II carcinoma versus 27% in those with grade III or IV histologic findings.¹⁷³ Patients with a papillary or mixed-tumor morphology have been shown to have higher local control rates and survival than those with solid lesions.¹⁶⁹ A report from M.D. Anderson Hospital of 135 patients, of whom the majority underwent an attempt at complete transurethral resection of tumor before definitive irradiation, indicated the importance of a cystoscopically verified complete response 2 to 6 months after the completion of therapy.¹⁷⁶ Multivariate analysis showed that a clinically complete response was the most important independent prognostic factor for survival and local pelvic control.^176,177 A study from Scotland of 333 patients with stage T₃ disease confirmed the correlation of a documented complete response to improved survival.¹⁷⁷ The presence or absence of histologic evidence of vascular invasion has not been shown to influence prognosis in the majority of series.^173–175

In an effort to improve local control rates with external-beam radiotherapy for patients with invasive bladder cancers, hyperfractionated treatment schedules have been used.^{146,171,178–180} The rationale behind this approach is derived from the recognition of tumor repopulation. Tumor repopulation by surviving clonogenic cells is believed to be one mechanism of treatment failure following fractionated radiotherapy and this phenomenon is increased with more protracted radiation schedules.¹⁸¹ Edsmyr and coworkers¹⁷⁹ reported a Swedish trial of 168 patients with T₂ to T₄ disease randomly assigned to receive either 1 Gy three times a day to a total of 84 Gy or 2 Gy once a day to a total of 64 Gy. Both regimens were given over 8 weeks, with an imposed mid-course 2-week treatment interruption. The 5-year survival for the group receiving 84 Gy was 37%, and for those receiving 64 Gy it was 16%.¹⁷¹ Patients with T₃ disease appeared to benefit most from the hyperfractionated schedule. An update with a follow-up period of at least 10 years shows that the benefit in survival for the entire group, and specifically for those with T₃ lesions, has persisted. To further evaluate the relationship of dose and overall treatment time, Majewski et al.¹⁸² conducted a retrospective study of 480 patients with T₂-T₃ bladder cancer treated with four different fractionation regimens: (1) conventional fractionation (1.8-2.5 Gy/day), (2) protracted radiation (1.6-1.7 Gy/day pelvic radiotherapy plus boost of 2.0 Gy/day), (3) accelerated hyperfractionation boost (2.0 Gy/day pelvic radiotherapy plus boost of 1.3-1.4 Gy/twice daily), and (4) accelerated hyperfractionation (1.2-1.5 Gy/twice daily pelvis plus boost radiotherapy). Although clinical prognostic factors including T stage were the most significant predictors of outcome, an increase in radiation dose was associated with improved local control (P = 0.011), although overall treatment time did not reach significance (P = 0.077). Horwich et al.¹⁸³ reported a prospective trial conducted at the Royal Marsden Hosptial comparing an accelerated hyperfractionated regimen to conventionally fractionated treatment in 229 patients with muscle-invasive bladder cancer (T₂-T₃) and minimal nodal disease. Patients were randomized to receive either 64 Gy in daily 2-Gy fractions to the whole bladder with margin or twice-daily radiation (1.8 Gy am dose and 2.0 Gy pm dose) to a total dose of 60.4 Gy over a 26-day period. Although the initial results of this study appeared promising, with longer follow-up there was no significant difference in local control or overall survival with the accelerated regimen. In addition, acute grade 2 and 3 bowel toxicity was higher in the accelerated arm, 40% versus 26% and 3% versus 0%, respectively. Acute grade 2 and 3 bladder toxicity appeared similar, at 16% versus 17% and 18% versus 19%, respectively. Treatment with accelerated and hyperfractionated schedules remains investigational and should be conducted in a protocol setting (see Multimodality Bladder-Preservation Therapy).

Multimodality Bladder-Preservation Therapy

In an effort to improve disease outcomes with radiation therapy, multimodality bladder-preserving approaches have been developed and typically include cystoscopic extirpation, radiation therapy, and concurrent chemotherapy. Several reports now substantiate the efficacy of combined chemotherapy and radiotherapy as an organ-preservation alternative for well-selected patients with localized transitional cell carcinoma of the bladder.^184–193 Current results suggest that this chemoradiation approach provides comparable survivorship to that expected for similar patients treated with radical cystectomy with a high rate of bladder preservation. The success of this treatment approach is predicated on the meticulous selection of patients with known favorable prognostic factors that are associated with improved response rates. This section summarizes the experience with combined modality therapy and discusses current trials that attempt to further explore this approach.

One of the earliest reports using combined modality therapy as a means for bladder preservation was conducted by the National Bladder Cancer Cooperative Group.¹⁹² From 1981 to 1986, 70 patients with T₂ to T₄ bladder cancer were enrolled in a prospective study, all with medical contraindications for radical cystectomy. Cisplatin was given at a dose of 70 mg/m² every 3 weeks concurrent with external-beam radiotherapy to a total dose of 64.8 Gy. Initially, 45 Gy was delivered to the entire bladder, perivesicular tissues, and adjacent lymph nodes using a four-field plan followed by a boost to the tumor within the bladder for an additional 19.8 Gy in 11 fractions delivered in 2 weeks. The overall complete response rate was 70%, and the 4-year survival among complete responders was 57%, compared with 11% among patients who did not achieve a complete response. Long-term survival rates were 64% for T₂ patients and 22% for T₃ patients. A subsequent randomized trial conducted by the National Cancer Institute of Canada between 1985 and 1989 comparing preoperative or definitive radiation therapy with or without concurrent cisplatin chemotherapy reported a significant improvement in local control in patients treated with cisplatin (67% versus 47%) without an effect on distant metastases or overall survival.¹⁹⁴ Of note, this trial is the only prospective, randomized trial to date comparing radiation alone with chemoradiation. Several single-institution studies have been performed demonstrating that a complete transurethral resection also affects local control prior to radiation therapy alone or concurrent with chemotherapy.^195,196

In Europe, similar studies were being conducted. Housset et al.¹⁸⁸ at the University of Paris reported results of a prospective study of 54 patients with operable T₂-T₄ bladder cancer treated with transurethral resection followed by 5-FU/cisplatin with concommitant twice-daily split-course radiotherapy to the whole pelvis (24 Gy in eight fractions over 17 days). A restaging cystoscopy with deep biopsies performed at 6 weeks postinduction demonstrated a 74% pathologic complete response rate. These complete responders were then treated with either radical cystectomy or definitive chemoradiation with an additional 20 Gy bladder boost (2.5 Gy twice-daily split course) concurrent with 5-FU/cisplatin. Among an initial 18 complete responders who underwent planned surgery after chemoradiation, no residual tumor was found in the cystectomy specimens in all cases. The 3-year disease-free survival rate was significantly better in responders compared with nonresponders (77% versus 23%). However, overall survival rates were similar (81% versus 77%), demonstrating that there was no negative survival impact on bladder-preservation attempts. Importantly, treatment was well tolerated. At the University of Erlangen, Dunst et al.¹⁹⁶ reported their experience using conservative surgery followed by radiation therapy with or without chemotherapy in a series of 245 patients treated bewtween 1982 through 1991 on a prospective protocol. Patients with high-risk T₁-T₃ bladder cancer were treated with a maximal transurethral resection followed by definitive radiotherapy consisting of 46 Gy to the pelvis and a whole-bladder boost to 54 Gy. After 1985, patients were treated with concurrent cisplatin or carboplatin routinely. A restaging cystoscopy and transurethral resection were performed 6 to 8 weeks following radiation, with salvage cystectomy reserved for persistent or recurrent disease. With a median follow-up of 5.9 years, the 5- and 10-year overall survival was 47% and 26%, respectively. The bladder preservation rate in 5-year survivors was 83%. This study also demonstrated the importance of the amount of residual tumor after transurethral resection with 5-year survival rates of 80% after R0, 53% after R1, and 31% after R2 resection (P < 0.01). Rodel and colleagues recently updated the University of Erlangen experience.¹⁸⁷ In that report, 289 patients with locally advanced tumors were treated with transurethral resection followed by concurrent platin-based chemotherapy and pelvic radiotherapy plus a tumor boost for a cumulative dose of 54 Gy. Chemotherapy was administered during the first and fifth week of radiotherapy and consisted of cisplatin 25 mg/m² daily for 5 days. Carboplatin was used for those patients with creatinine clearance less than 60 mL per minute or congestive heart disease. In 68% the full-dose chemotherapy was able to be administered, whereas in 32% the doses needed to be curtailed secondary to hematologic toxicity, gastrointestinal toxicity, or nephrotoxicity. The median follow-up in the study was 60 months. The complete response rates at the 6 week restaging TURBT for patients treated with radiotherapy in combination with cisplatin and 5-FU, cisplatin alone, and carboplatin was 87%, 82%, and 66%, respectively. These rates of response were significantly higher than the complete response rate (61%) observed among 126 patients who were treated with radiotherapy alone (P = 0.001). Among all patients who achieved a complete response, the 10-year local control rate was 64%. The overall likelihood of developing distant metastases and disease-specific survival was 35% and 42%, respectively, at 10 years. Of long-term survivors, 80% maintained functioning bladders, and cystectomy was only required in 2% of patients secondary to a contracted bladder after radiotherapy. In a multivariate analysis, early tumor stage and the performance of a complete TURBT before radiochemotherapy were predictors of a higher complete response rate and improved overall survival. The 5-year survival rates among patients who had a complete TURBT, microscopic residual disease, and gross residual disease after resection were 76%, 52%, and 34%, respectively (Fig. 44-9). The data also suggested that adjuvant radiotherapy is of benefit even among selected patients who had complete TURBT. In this latter group, the incidence of bladder preservation was 85%, compared with 75% reported by Herr¹⁹⁷ for highly selected patients with muscle-invasive tumors treated with TURBT alone.

Figure 44-9 • Data from the University of Erlangen demonstrating the impact of having a complete transurethral resection among patients treated with chemoradiation.

These highly encouraging results were the impetus for a phase II chemoradiation bladder-preservation study for patients with localized bladder cancer who were otherwise eligible for radical cystectomy conducted at Massachusetts General Hospital.¹⁹³ Fifty-three patients were enrolled in this study, of whom 15 had clinical T₂, 29 had T₃, and 9 had T₄ disease. Treatment consisted of a transurethral resection followed by two cycles of neoadjuvant MCV and then 45 Gy to the pelvis with two concomitant doses of cisplatin. The 36 patients with a complete response at 40 Gy (as demonstrated by repeat urologic examination including cystoscopy, examination under anesthesia, and urine cytology) were allowed to complete the pelvic radiation and proceed to a 24.8 Gy tumor boost with a final dose of cisplatin. Radical cystectomy was performed for 15 patients who could not tolerate the combined chemoradiation, had an incomplete response at the repeat cystoscopy, or required salvage after completion of the entire treatment plan. This study demonstrated that a multimodality approach resulted in a 5-year survival rate of 48%, comparable with the expected outcome after radical cystectomy based on historic controls. There were no major complications involving the bladder or rectum, with only two patients experiencing intermittent hematuria. Toxicity associated with chemotherapy included nausea and vomiting (73%), stomatitis (24%), and diarrhea (10%). A 58% rate of bladder conservation was achieved at median follow-up of 48 months without an apparent compromise in survival. Of 27 patients who were complete responders after neoadjuvant chemotherapy and 40 Gy of radiation, 81% maintained a normal, functioning bladder.¹⁹³ Shipley et al.¹⁸⁵ recently updated the results of 190 patients treated on various institutional protocols at Massachusetts General Hospital from 1986 to 1997 with a combined modality bladder-preservation approach. With a median follow-up time of 6.7 years, the 5- and 10-year overall survival rates were 54% and 36%, respectively. For patients with clinically staged T₂ disease, the 5- and 10-year survival rates were 62% and 41%, respectively. For patients with clinically staged T₃ to T₄ disease, the 5- and 10-year survival rates were 47% and 31%, respectively. Disease-free survival outcome with an intact bladder was also excellent in this group of selected patients. The 5- and 10-year disease-free survival rates with an intact bladder were 57% and 50%, respectively. For patients with clinically staged T₃ to T₄ disease, the 5- and 10-year disease-free survival with intact bladder rates were 35% and 34%, respectively. Bladder contractures requiring cystectomy were not observed in these patients.

Simulation and Treatment

During simulation and treatment, the patient is supine and the bladder is emptied to ensure reproducibility. CT simulation is the gold standard, as two-dimensional simulation using a cystogram has a risk of geographic miss. Visualization of the pelvic vessels and bladder tumor may be enhanced using intravenous contrast. At the time of simulation, a Foley catheter may be inserted and a small volume (25 to 30 ml) of radiopaque contrast material may be instilled with or without 10 to 15 ml of air for bladder visualization. One must be cautious to ensure that the bladder volume is not significantly increased such that daily reproducibility will be compromised. A rectal tube with or without barium contrast medium is placed to visualize the rectum. If using contrast, similar care regarding rectal volume reproducibility should be used. CT planning involves the delineation of normal tissue and target structures on CT images using treatment-planning software. Various approaches to defining the clinical target volume exist. The most common approaches are (1) a two-phase approach to encompass the whole pelvis (including pelvic nodes) followed by a boost to the bladder or involved bladder wall or (2) a single-phase approach encompassing the bladder alone to full dose. Alternatives may also include treatment to the involved bladder wall alone in an effort to spare uninvolved bladder and bowel. Proponents of bladder-only treatment cite no significant increase in pelvic recurrence in these patients and no negative effect on survival; however, direct comparisons are difficult.²¹³ A recently closed UK phase II trial, BC2001, is designed to address target volume and the results are anticipated.²¹⁴

The classic radiation fields include elective pelvic nodal irradiation in addition to the bladder and proximal urethra followed by a boost to the involved bladder (Fig. 44-10A). In general, a four-field box approach is used (anterior, posterior, and lateral fields) delivered through 15- to 25-MV x-rays. The initial fields encompass the entire bladder and the first-echelon draining lymph nodes. The superior border of the anterior-posterior–posterior-anterior fields is the L5 to S1 interspace. The inferior border of these fields is at the level of the bottom of the obturator foramina, and the lateral fields generally extend 1.5 to 2 cm beyond the widest portion of the bony pelvis. The anterior border of the lateral fields extends to the anterior bladder wall with a 1.5- to 2-cm margin. Often, the anterior border may extend several centimeters beyond the symphysis pubis. The posterior border of the lateral fields can be placed with a 2- to 3-cm margin posterior to the bladder, which will incorporate the presacral lymph nodes.

Figure 44-10 • A, Classic radiotherapy field using four-field box. B, Opposed laterals boost.

To properly plan the boost portion of therapy, which is typically limited to the involved area within the bladder, the radiation oncologist will need to know the location of the tumor based on the pretreatment cystoscopy with a detailed bladder map and findings from the examination under anesthesia or from pre-TURBT radiographic imaging and any extravesical disease should be included in the boost target volume. A treatment-planning CT scan can be obtained and target volumes delineated. Often, a full bladder is helpful during the boost simulation and treatment to minimize the irradiated bowel and uninvolved bladder mucosa in the high-dose volume. Care must be taken to ensure similar degrees of bladder filling during the planning process and during treatment, and ultrasound verification may be adviseable. Although opposed lateral fields (see Fig. 44-10B) have been used on prior RTOG protocols, a multifield technique may be used that minimizes the volume of normal bladder exposed to the prescription radiation dose. In some cases based on the location of the bladder tumor (i.e., limited to the dome or anterior-posterior wall), lateral fields alone may be sufficient for this phase of the therapy. More recently, the authors have incorporated 3-D treatment planning for planning the cone-down phase of therapy (Fig. 44-11). A multifield plan can be used in an effort to minimize the dose to uninvolved regions of the bladder as well as other normal tissue structures such as the rectum, bowel, and femoral heads.

Figure 44-11 • Three-dimensional conformal plan for the cone-down phase of therapy. Note the high-dose region confined to a small volume of bladder (site of known initial disease).

Treatment is initiated with four fields delivering daily 1.8 Gy fractions five times a week for a total dose of 45 Gy. At this dose, a 3-D plan is initiated for the cone-down approach and is directed to the tumor-containing normal bladder. The total dose is brought to 64.8-66.6 Gy using daily 1.8 Gy fractions.

Intensity-modulated radiation therapy has been implemented in many disease sites in an effort to improve treatment results and reduce toxicity. For bladder cancer treatment, this approach is particularly appealing, given the difficulty in delivering tumoricidal radiation doses in the pelvis adjacent to radiosensitive bowel and other organs. The steep dose gradients achievable with intensity-modulated radiation therapy, however, mean that organ-motion uncertainties may lead to significant tumor underdosage or normal tissue overdosage. Several studies show significant bladder motion during a course of standard radiotherapy.²¹⁵ Muren et al.²¹⁵ found bladder wall motion of greater than 15 mm in up to 40% of patients with maximal displacement of 29 to 36 mm as well as significant variations in rectal volumes and small bowel position. Pos et al.²¹⁶ measured organ motion from four weekly repeat CT scans and found that, despite treatment margins of 1.5 to 2 cm, the gross tumor volume was outside of the planning tumor volume at least once in 65% of patients treated with a full bladder. Most studies are in agreement that the predominant bladder motion is seen in the superior and anterior directions and the least motion noted in the inferior direction.

Image-guided radiation therapy may be the most promising step toward more precise radiation delivery. Using daily imaging, the radiation oncologist may be able to adapt treatment delivery to account for these large variations in target motion. This could potentially allow a reduction of the large treatment margins currently necessary, thus allowing safe dose escalation. Several different approaches are currently being explored, including fiducial marker placement and daily two-dimensional kV imaging²¹⁷ or 3-D cone-beam CT imaging,²¹⁸ which allows soft-tissue visualization immediately prior to treatment while the patient is on the treatment couch. One additional benefit of cystoscopically placed markers is the added ability to accurately define the tumor bed and reduce uncertainties in target definition.

Brachytherapy

Brachytherapy for bladder cancer was originally developed in the early twentieth century and typically consisted of permanent radon seed implantation or temporary radium needle insertion.²²¹ Because of crude techniques causing high complication rates and high radiation exposure to staff, brachytherapy for the treatment of bladder cancer fell out of favor. With the subsequent development of more precise techniques and afterloading capability, there has been renewed interest in this approach for bladder-preserving treatment. As with other bladder-sparing approaches, there is general agreement that small (<5 cm), solitary, organ-confined tumors are most amenable to brachytherapy-based treatment.

Although the experience with brachytherapy for localized bladder cancer has been limited in the United States, an extensive experience with radium needle implants has been reported by Van der Werf-Messing and colleagues from Rotterdam. The Rotterdam series established the importance of preoperative external-beam radiation prior to brachytherapy, as this significantly decreased scar metastases, which were a frequent component of treatment failure prior to its use. In the most recent Rotterdam report, 328 patients with T₂ disease and 63 patients with T₃ muscle-invasive bladder cancer were treated with preoperative radiation of three fractions each at a dose of 3.5 Gy followed by radium implantation.²²² Local recurrence at 5 years was 16% for T₂ tumors and 28% for T₃ tumors. Survival was similar to historic controls undergoing definitive external-beam radiotherapy or cystectomy, with 56% of patients with T₂ disease and 37% of those with T₃ disease alive at 5 years. In addition to stage, other poor prognostic factors were identified, including grade, vascular invasion, pathologic findings on intravenous pyelography, and more than one transurethral resection.

The Netherlands and France have the largest reported experiences establishing indications and techniques for bladder brachytherapy. Most series incorporate a suprapubic cystotomy with a unilateral or bilateral pelvic lymphadenectomy approximately 1 to 2 weeks following completion of preoperative radiation. Typically, a partial cystectomy is subsequently performed, removing the residual disease with a 1-cm margin. In some series, only a cystotomy is performed. A target area is identified and can be marked with radio-opaque markers to assist in treatment planning. In the afterloading technique, afterloading catheters are placed into the bladder wall target area in a parallel arrangement 1 to 1.5 cm apart (Fig. 44-12). The catheters are brought to the abdominal wall surface and secured. Treatment planning can currently be accomplished using 3-D CT-based dosimetry. Dose prescriptions vary considerably (30-60 Gy) with higher doses given in patients with T₁ and favorable T₂ tumors with shorter courses of external-beam radiation; however, there is currently no established consensus. Following treatment, catheters can be removed, generally with minimal anesthesia.

Figure 44-12 • Anterior-posterior (A) and lateral (B) radiograph demonstrating insertion of high-dose-rate catheters for bladder brachytherapy. Stranded dummy sources are shown and used to assist in treatment planning. Spacers are placed between strands (green arrow) to maintain parallelism.

Since the introduction of afterloading techniques, most current series use iridium.^223–226 Mazeron et al.²²⁵ treated 30 patients with T₂ disease and 5 with T₃ disease using a single fraction of 8.5 Gy, followed by partial cystectomy and iridium implantation, and attained excellent local control. Rozan and colleagues²²⁶ reported a multicenter French experience that included 98 patients with pT₁-stage disease, 66 with pT₂, 26 with pT_3a, 9 with pT_3b, and 1 with pT₄ treated by preoperative external-beam irradiation, partial cystectomy with lymphadenectomy, and placement of an iridium implant into the tumor bed. Preoperative radiation was targeted to the bladder in 44 patients and to the pelvis in 161 patients for a mean total dose of 11 Gy and a mean dose per fraction of 5.4 Gy. Of the 36 tumor recurrences, 9 were located at the original site and 19 occurred in a different site. They reported an overall 5-year survival of 67%. Survival was 77% for pT₁, 63% for pT₂, and 47% for pT_3a disease. Serious complications included 8 patients with hematuria, 17 patients with chronic cystitis, and 11 with fistulas. Blank et al.²²⁷ reported results of 122 patients treated with external-beam radiation followed by partial cystectomy (n = 37) or cystotomy (n = 85) and an iridium-192 implant. Brachytherapy consisted of either pulsed-dose-rate (n = 23) or continuous low-dose rate (n = 99) with doses varied from 20 to 50 Gy, depending on the extent of surgery and external-beam radiation dose. At 5 years, the local relapse-free survival was 76% and distant-metastasis-free survival was 83%. The overall survival rate was 49% at 10 years. Complications included two patients with ileus managed conservatively and three patients with crippled bladder function.

Salvage Therapy for Local Recurrence

The key to the success of bladder-preservation strategies with respect to outcomes is close cystoscopic surveillance following therapy. A significant concern is the development of recurrent disease. For muscle-invasive recurrence, prompt salvage cystectomy should be performed. As previously discussed, survival for these patients does not appear compromised compared with patients undergoing immediate cystectomy. The management of superficial recurrences is also important, because some of these lesions, even in a de novo setting, may exhibit aggressive behavior despite aggressive surgical management. Several reports have demonstrated that superficial recurrences in a retained bladder following multimodality bladder-preservation therapy may be managed as de novo disease.^228,229 Weiss et al.²²⁸ reported results on 531 consecutive patients treated with combined-modality bladder preservation and found that 17% (68/389) of patients achieving a complete response to therapy developed superficial recurrences (≤T₁). Of the 64 patients who underwent conservative management, consisting of a transurethral resection with or without intravesical therapy, 48% had no further recurrence, 33% experienced additional superficial recurrences, and 19% progressed to invasive disease. Compared with patients without bladder relapse, there was no difference in disease-specific survival (72% with superficial recurrence versus 79% without recurrence; P = 0.78). Zietman et al.²²⁹ reported results for 190 patients treated with chemoradiation for bladder cancer and found 57 separate superficial recurrences in 32 patients. Of the superficial recurrences, 67% occurred within the same subregion as the original invasive tumor. Of the 28 conservatively treated patients, 18 (64%) had no further recurrence during the period of observation. Seven had additional superficial recurrences, of which three were managed conservatively and four with cystectomy. Three of the 28 patients ultimately developed invasive tumors requiring cystectomy. Only 1 of the 32 patients with superficial relapse developed uncontrolled pelvic disease. The response of these superficial tumors to conservative therapy is similar to results in patients with primary superficial disease.

Chemotherapy for Metastatic Disease

Reported survival for patients with metastatic bladder cancer treated on chemotherapy trials varies widely. This wide variation can be explained by pretreatment disease and patient-related factors. In an effort to define the effect of pretreatment patient characteristics on clinical outcome, Bajorin and colleagues conducted a multivariate analysis evaluating 18 variables in 203 patients treated with MVAC.²³⁰ In this analysis, the only variables that were prognostic of survival were Karnofsky performance status (KPS) lower than 80% and the presence of visceral metastases. Patients with a KPS lower than 80% and visceral metastases had a median survival of 9.3 months, compared with patients who had no adverse prognostic factors and whose median survival was 33 months.²³⁰ Prognostic factors for survival were also evaluated in a 405-patient randomized trial of combination cisplatin-based chemotherapy. This analysis concluded that improved overall survival was associated with a KPS greater than 70%, absence of distant metastases, low or normal alkaline phosphatase, fewer than three sites of disease, and the absence of visceral metastases.²³¹ An understanding of prognostic features of patients enrolled in prospective clinical trials may allow more sophisticated and accurate interpretation of outcomes reported in phase 2 clinical trials, allowing for improved selection of treatments for phase III testing. Identification of additional pretreatment prognostic factors will allow more accurate identification of patients at high risk for death from their cancer.

Historically, MVAC had been considered the standard therapy for treating advanced bladder-cancer patients. The efficacy of MVAC was first reported in 1989 when Sternberg and colleagues treated 121 patients with advanced urothelial tract cancers and demonstrated a 72% response rate.²³² MVAC was subsequently compared with single-agent cisplatin and shown to be superior in terms of response rate and overall survival in patients with advanced bladder cancer.²³³ MVAC has also been compared with the previously used multiagent regimen CISCA: 110 patients were randomized to either MVAC or CISCA. MVAC was found to have both a higher objective response rate and a longer median survival.²³⁴ Despite superior outcome with MVAC, significant limitations include its severe toxicity (mucositis, infectious complications, and a 3% toxic death rate are consistently seen in randomized trials). In addition, despite high objective response rates, only a very small percentage of patients with metastatic bladder cancer have long-term disease-free survival (3.7% at 6 years).²³³

MVAC has been administered in a dose-intense fashion with the incorporation of prophylactic filgrastim to prevent excessive toxicity. A randomized phase III study was conducted comparing dose-intense MVAC administered every 2 weeks to standard MVAC in 263 patients with untreated metastatic transitional cell carcinoma. The dose-intense MVAC arm was associated with objective responses in 64% of patients compared with 50% with standard-dose MVAC. Long-term follow-up of this trial demonstrated that 24.6% of patients on the dose-intense arm were alive at 7.3 years, compared with 13.2% on the standard-dose MVAC arm, although the median survival was similar between the two regimens.²³⁵ The hazard ratio for survival was 0.76 (95% CI = 0.58-0.99). Toxicity with dose-intense MVAC treatment appears to be similar to or better than standard-dose MVAC treatment. Dose-intense MVAC remains a standard treatment option for patients with advanced transitional cell carcinoma.

The GC regimen has largely replaced MVAC as the standard of care based on a randomized trial comparing the two regimens.¹⁶⁹ This trial was designed to demonstrate superior efficacy for GC. Although the trial was not sufficiently powered to prove equivalence of GC to MVAC, this 405-patient trial showed similar anti-tumor efficacy (equal complete response rate and similar partial response rate) and a hazard ratio for survival approaching 1.0. An overall response proportion of approximately 50% was observed with both regimens. Long-term follow-up showed no difference in survival (13% in patients treated with GC, 15.3% in patients treated with MVAC). However, toxicity in GC-treated patients was substantially lower than observed in patients treated with MVAC, including a lower death rate, fewer infections, and less mucositis. This markedly superior therapeutic index has led to fairly wide acceptance of GC as the standard first-line therapy for metastatic bladder cancer.

Taxanes have significant activity in advanced transitional carcinoma. Paclitaxel has been tested in combination with gemcitabine and cisplatin and demonstrated significant activity in phase II testing.²³⁶ Phase III evaluation of the triplet regimen compared with GC alone has been reported in abstract form, and suggests that there is no significant improvement with the addition of a third agent to GC chemotherapy.²³⁷

Unfortunately, many patients with advanced transitional cell carcinoma have renal insufficiency and are unable to undergo therapy with cisplatin or other potentially nephrotoxic agents. Consequently, regimens with less nephrotoxic drugs have been developed. These have included single-agent treatments with drugs such as the taxanes, gemcitabine, and carboplatin.^238–247 The median survival reported for patients with renal insufficiency (in predominantly single-institution studies) ranges from 8 to 10 months, and is generally inferior to the results of the large multicenter trial (n = 405) comparing GC to MVAC in patients with nonimpaired renal function, which found median survivals of 13.8 and 14.8 months, respectively.²⁰² Whether this is attributable to a fundamentally worse prognosis in patients with renal insufficiency or to the use of less effective regimens is not known. Frequently-used combinations in metastatic transitional cell carcinoma patients with renal insufficiency include gemcitabine/carboplatin and carboplatin/paclitaxel. Optimal chemotherapy for patients with renal insufficiency remains undefined.

New drug targets have been identified, and novel agents against these targets are currently being tested. These targets include epidermal growth factor receptor, Her2, mTOR, fibroblast growth factor receptor 3, and vascular endothelial growth factor receptor, among others. Drugs that interfere with these targets have been designed, and clinical trials are ongoing. Although progress has made in improving outcomes with combination chemotherapy, there is still much room for progress.

Palliative Radiotherapy

Bladder cancer can metastasize to various visceral organs, causing symptoms that require palliation. Radiation therapy is an effective therapy for pain or hematuria associated with advanced disease. The treatment approach in advanced disease is dictated by a desire to deliver therapy for rapid relief of pain or hematuria with minimal inconvenience to the patient. This can be accomplished with a short course of radiation with a high dose per fraction. Srinivasan and associates²⁴⁸ compared in 41 patients a conventional regimen of 45 Gy in 12 fractions over 26 days to a hypofractionated treatment of 17 Gy in 2 fractions over 3 days.²⁴⁸ In patients receiving the hypofractionated regimen, 59% had a clearance of hematuria and 73% had an improvement in pain. This was superior to the conventional treatment that resulted in a 16% clearance of hematuria and a 37% improvement in pain. Median survival was higher at 14 months for those treated conventionally compared with 10 months for those receiving the hypofractionated treatment.

A palliative regimen of 30 Gy in six fractions of two fractions per week was reported by Salminen²⁴⁹ in 94 patients. Forty attained complete palliation and 27 had partial improvement, although many suffered acute side effects, including diarrhea in nearly 60% of patients. A regimen of 21 Gy in three fractions over 5 days resulted in palliation of 75 of 102 patients but at a cost of significant acute toxicity.²⁵⁰ External-beam irradiation has had mixed results in relieving ureteral obstruction.²⁵¹ The obstruction was relieved for 20%, unchanged in 24%, and worse in 56% after a 60-Gy course in 2-Gy fractions for 55 patients with 68 obstructed kidneys. This limited success was at a cost of vesicovaginal fistulas in 1 patient, contracted bladders in 22 patients, and diarrhea in 25 patients. Median survival was 11 months, with a median of 9.5 months spent at home. A randomized trial conducted by the Medical Research Council demonstrated that 21 Gy in 3 fractions were as efficacious as 35 Gy in 10 fractions for palliation of local bladder symptoms.²⁵² Symptomatic improvement was reported in 64% of patients in the short-course arm and 71% in the longer course but this did not reach statistical significance. Toxicity was modest and similar in both arms. Shorter courses of palliative radiation therapy are ideal for rapid palliation in poor-prognosis, ill patients whereas longer courses of therapy with or without chemotherapy may be appropriate for patients with a good performance status and a lower disease burden. When chemotherapy is used, lower daily doses (1.8-2.5 Gy) should be used with total doses of 45 to 50 Gy. The optimal regimen should be tailored to the individual patient.