Key points

Introduction

Bladder cancer is one of the most common cancer diagnoses in the United States, with an estimated 68,810 new cases in 2010, accounting for 7% of all cancers and 3% of cancer deaths. Approximately 70% of bladder cancers do not invade the underlying muscle at diagnosis ; this type of malignancy is broadly referred to as non-muscle invasive bladder cancer (NMIBC). Among NMIBCs, disorders can range from low-grade, superficial disease that often behaves in an indolent fashion to more aggressive high-grade lesions such as carcinoma in situ (CIS) and lesions invading the underlying lamina propria. These high-risk, high-grade lesions have the potential to progress to muscle-invasive disease that is typically not amenable to the bladder-sparing treatments often used for NMIBC.

Treatment of bladder cancer is costly, in both direct and indirect terms. Of all cancers, bladder cancer is the ninth most costly in the United States in terms of overall expenditures, and has the greatest lifetime treatment cost per patient diagnosed, surpassing colorectal, breast, prostate, and lung cancers. Among the 5 commonest cancers in elderly patients, review of Medicare data expenditures also found bladder cancer to be the costliest. Of patients diagnosed with low-grade NMIBC, approximately 50% recur and 5% progress to muscle-invasive disease. These recurrences necessitate multiple endoscopic resections and intravesical treatments combined with lifelong surveillance. Beyond the direct costs, NMIBC treatment also results in indirect costs such as lost productivity or the heightened psychological burden of a cancer diagnosis.

Introduction

Bladder cancer is one of the most common cancer diagnoses in the United States, with an estimated 68,810 new cases in 2010, accounting for 7% of all cancers and 3% of cancer deaths. Approximately 70% of bladder cancers do not invade the underlying muscle at diagnosis ; this type of malignancy is broadly referred to as non-muscle invasive bladder cancer (NMIBC). Among NMIBCs, disorders can range from low-grade, superficial disease that often behaves in an indolent fashion to more aggressive high-grade lesions such as carcinoma in situ (CIS) and lesions invading the underlying lamina propria. These high-risk, high-grade lesions have the potential to progress to muscle-invasive disease that is typically not amenable to the bladder-sparing treatments often used for NMIBC.

Treatment of bladder cancer is costly, in both direct and indirect terms. Of all cancers, bladder cancer is the ninth most costly in the United States in terms of overall expenditures, and has the greatest lifetime treatment cost per patient diagnosed, surpassing colorectal, breast, prostate, and lung cancers. Among the 5 commonest cancers in elderly patients, review of Medicare data expenditures also found bladder cancer to be the costliest. Of patients diagnosed with low-grade NMIBC, approximately 50% recur and 5% progress to muscle-invasive disease. These recurrences necessitate multiple endoscopic resections and intravesical treatments combined with lifelong surveillance. Beyond the direct costs, NMIBC treatment also results in indirect costs such as lost productivity or the heightened psychological burden of a cancer diagnosis.

Pathology/staging

NMIBC is defined by the stage classification of the cancer according to American Joint Committee on Cancer (AJCC) specifications. Cancer grade is an important component of risk stratification within NMIBC. The World Health Organization (WHO) recently adopted the term urothelial cell carcinoma (UCC) to replace the traditional term transitional cell carcinoma. UCC was formerly graded from 1 to 3, with grade 1 lesions representing well-differentiated UCC and grade 3 lesions representing poorly differentiated UCC. However, as per WHO recommendation, this has been replaced by dichotomous histologic grading with UCC lesions now categorized as either high grade or low grade. This grading decreases some of the histologic discrepancy and interpathologist variability that was often present with grade 2 tumors with the earlier classification. Papillary urothelial neoplasm of low malignant potential (PUNLMP) and papillomas are two additional noninvasive bladder lesions that are rarely aggressive. PUNLMPs progress in fewer than 3% of cases; however, because of the possibility of local recurrence and the low probability of progression, many advocate maintenance of surveillance. Papillomas do not progress and do not require long-term follow-up.

The bladder wall comprises 3 histologic layers. The inner urothelium is approximately 7 cell layers thick. Just deep to this, the lamina propria is a layer of blood vessels and lymphatics. The outermost muscularis propria predominantly functions to give the bladder its contracting ability. Stage classification Ta tumors are papillary tumors that do not extend deep to the urothelium. CIS (stage classification Tis) refers to a flat, high-grade lesion confined to the urothelium. Stage classification T1 tumors invade the lamina propria but do not extend into the underlying muscularis propria. This article focuses on the costs associated with management of lesions confined to the urothelium or lamina propria (ie, stage classification Ta, Tis, and T1), which are collectively referred to as NMIBC.

Risk factors and demographics

Tobacco smoking is the greatest known external risk factor for the development of bladder cancer. The relative risk among smokers for the development of bladder cancer is 2.8 in men and 2.7 in women compared with nonsmokers. This risk increases in proportion to a patient’s duration of tobacco use and with the amount of tobacco exposure over time. Other environmental exposures also increase the risk for the development of UCC. Certain occupations with frequent contact with petroleum products, stains, and dyes have an increased risk for bladder cancer. Such individuals include mechanics, leather workers, miners, and hairdressers.

Men outnumber women in bladder cancer diagnoses at a ratio of 3:1. Although bladder cancer is more common in men, women often present with late-stage UCC, likely as a result of misinterpretation of presentation with hematuria. The incidence of bladder cancer increases with age; most diagnoses occur during the eighth decade of life. UCC is more prevalent in the white population, with white men and women having a 3-fold increase in the risk of developing bladder cancer compared with African Americans.

Direct economic burden

The economic burden of diagnosis, treatment, and surveillance of NMIBC is substantial. Bladder cancer care in the United States in 2010 cost an estimated $3.98 billion. Lifetime per capita costs have been estimated between $96,000 and $187,000 in 2001 US dollars, or more than $230,000 in 2010 US dollars. Modeling patients whose course follows the best possible bladder cancer outcomes (patients with NMIBC with projected recurrences at the mean NMIBC rate), Avritcher and colleagues found that the lifetime cost of treatment of patients in a best-case scenario was $120,684. Management of recurrences accounted for approximately 60% of this lifetime cost. If all patients newly diagnosed with NMIBC in the last year received recommended surveillance care with no recurrences (a perfect case scenario), the cost to the health care system for the management of this cohort for 5 years would exceed $150 million.

Several factors contribute to the cost of NMIBC. Most patients present with either gross or microscopic hematuria, and initial evaluation for hematuria consists of cystoscopy and cross-sectional imaging. Urine cytology is often performed to evaluate for CIS that may be difficult to visualize cystoscopically or to better define the anticipated grade of identified bladder lesions. In instances of chronic kidney disease or contrast allergy, the expense of the initial evaluation is often compounded by the need to evaluate the upper urinary tract endoscopically, typically by performing retrograde pyelography or ureteroscopy under a general anesthetic.

If tumor or a suspicious lesion is visualized on initial cystoscopic examination, on radiographic examination, or if the urine cytology is suspicious for malignant cells, the initial tumor resection or biopsy of suspicious areas is usually done under general anesthesia. Intravesical instillation of mitomycin C is commonly indicated to reduce the risk of recurrence of bladder tumors, especially if the tumor seems to be low grade and noninvasive. For those diagnosed with NMIBC, the treatment and follow-up vary depending on the stage and grade of tumor. A repeat resection within 6 weeks is recommended for patients with high-grade T1 NMIBC disease because of a 64% risk of understaging when muscularis propria is absent from the initial resection and a 40% overall risk of understaging. Because management of upstaged NMIBC is different and cancer-specific outcomes suffer for mismanaged patients, repeat resection is not a candidate target for cost savings.

Following initial resection, among patients with high-grade NMIBC, intravesical bacille Calmette-Guérin (BCG) is typically indicated as first-line treatment. BCG is an attenuated mycobacterium that induces an immune response. Recommendations from the National Comprehensive Cancer Network (NCCN) advocate a 6-week induction course of BCG for those with high-grade Ta or T1 bladder cancer as well as those with CIS, followed by up to 3 years of maintenance therapy for multiple installations at varying time intervals, based on high-level evidence. Kilbridge and Kantoff used a cost model to show that use of BCG in a 65-year-old man with high-grade NMIBC costs approximately $4820 per year of life saved (2010 US dollars), a remarkable cost-effectiveness figure.

For patients with low-grade NMIBC, treatment and surveillance are less vigilant. Patients are often managed conservatively with surgical treatment of recurrences, without consideration of induction intravesical immunotherapy or chemotherapy. However, patients should receive a single instillation of intravesical chemotherapy concordant with resections, based on high-level evidence. Randomized trials and meta-analyses show that intravesical administration of mitomycin C given immediately after resection of a bladder tumor decreases tumor recurrence by approximately 35%.

For high-risk NMIBC refractory to treatment with BCG, salvage intravesical chemotherapy may help patients avoid major extirpative surgery, which is costly both in terms of expenditure and in the induced quality-of-life burden to the patient. Intravesical valrubicin has a single indication for its use as a salvage agent in high-risk NMIBC failing BCG. However, the cost-effectiveness of valrubicin for this indication may be marginally superior to that of radical cystectomy. Valrubicin showed durable benefit in 28% of patients with BCG-refractory NMIBC, and costs 6 times more than an induction course of BCG. Cost-effectiveness models that incorporate the disutility of cystectomy with urinary diversion may better account for the benefit of salvage valrubicin.

Prediction algorithms have attempted to better classify patients’ risks of recurrence and progression of bladder cancer to optimize surveillance strategies. The American Urological Association (AUA) categorizes patients into low-risk and high-risk groups based on grade and volume of cancer. Low-risk patients have low-volume, low-grade NMIBC. Intermediate-risk patients have higher volume low-grade NMIBC. The AUA categorizes all patients with high-grade NMIBC as high risk. The NCCN describes a similar risk stratification classification, but the NCCN categorizes all patients with low-grade NMIBC as low risk, independently of tumor volume. Intermediate-risk patients have high-grade stage classification Ta bladder cancer. High-risk patients with NMIBC have stage classification T1 and Tis tumors. The remainder of this article refers to the NCCN risk stratification and guidelines.

Patients with intermediate-risk or high-risk NMIBC require quarterly cystoscopies for the first 2 years after diagnosis with upper urinary tract evaluation every 1 to 2 years with cross-sectional imaging. Surveillance cystoscopy may be performed at increasing intervals if patients are recurrence free after 2 years. However, these patients are typically committed to a lifetime of upper tract surveillance. Although recommendations vary regarding surveillance for low-risk NMIBC, the presence of tumor at the initial 3-month cystoscopic evaluation is a prognostic marker for recurrent disease. NCCN guidelines recommend initial cystoscopy 3 months after resection, with subsequent cystoscopies at increasing intervals if the initial cystoscopy is negative for recurrence.

Long-term surveillance of NMIBC contributes substantially to the cost burden of the disease. Compared with muscle-invasive bladder cancer, in which costs are disproportionately driven by initial diagnosis and treatment, complications of these therapies and subsequent care, including end-of-life care, treatment complications account for only 4% of overall NMIBC costs. Because NMIBC surveillance mandates frequent invasive procedures and periodic cross-sectional imaging, these components of care dominate determinants of NMIBC costs. Changes in surveillance and treatment strategies have previously mirrored changes in physician reimbursement. In an effort to decrease the cost of hospital-based procedures, in 2005, Medicare adopted changes in reimbursements for office-based endoscopic procedures, which led to a large increase in office-based procedures and increased the cost of NMIBC surveillance and treatment.

To mitigate the costs associated with cystoscopic surveillance, to address concerns about patient noncompliance because of procedural discomfort, and to overcome the low sensitivity of cytology, other novel urinary markers have been examined that could obviate cystoscopy. Cytology has a median sensitivity of 69% for the detection of high-risk tumors within the bladder, 78% sensitivity for detection of CIS, but lower sensitivity for upper tract lesions within the ureter and renal pelvis. Thus, a negative cytology does not preclude cystoscopy, particularly in the setting of low-grade NMIBC or upper urinary tract tumors.

Commercial urinary markers for bladder tumor antigens (BTA stat and BTA trak) have superior sensitivity for low-grade tumors (48% vs 13% for cytology), but limited sensitivity overall, failing to detect 47% of tumors identified cystoscopically. In addition, these tests had low specificity and frequent false-positive results in the setting of urinary tract infections. ImmunoCyt tests for markers of exfoliated UCCs. Similar to BTA stat and BTA trak, ImmunoCyt has improved sensitivity for low-grade NMIBC compared with cytology, but has low specificity. ImmunoCyt seems to be particularly confounded in patients with cystitis or benign prostatic hyperplasia. Urinary testing for NMP22 has sensitivity of 59%, 90%, and 70% for low-grade NMIBC, high-risk NMIBC, and CIS, respectively. Reduced sensitivity in cases of infection, bladder calculi, and prior urinary tract instrumentation curtail its usefulness in the evaluation and surveillance of NMIBC.

Fluorescence in situ hybridization (FISH) detects aneuploidy in chromosomes 3, 7, and 17, as well as loss of the 9p21 locus of the P16 tumor suppressor gene. Sensitivity is limited in low-grade NMIBC, but increases in the detection of high-grade NMIBC. FISH urinary testing does have some usefulness in patients undergoing treatment with intravesical BCG, a setting in which cytology has limited usefulness, and may be of benefit in certain patients undergoing intravesical treatment who have ambiguous cytology results. Kamat and colleagues reported that high false-positive rates incurred with FISH tripled the cost associated with each additional cancer diagnosis compared with cystoscopy alone. To date, urinary markers have yet to be identified with adequate sensitivity or specificity to obviate cystoscopy and cytology and the potential cost reductions that could produce.

Novel endoscopic techniques have been developed to improve on detection of bladder cancer. Narrow-band imaging cystoscopy and fluorescence cystoscopy after intravesical instillation of 5-aminolevulinic acid or hexaminolevulinate have been shown to increase tumor detection rates compared with traditional white-light cystoscopy. Better detection was associated with decreased recurrence rates at 3-month surveillance, presumably because of complete treatment of all identified lesions at initial resection. Cauberg and colleagues showed significantly decreased recurrence rates from 30.5% with conventional cystoscopy to 15.0% with narrow-band imaging cystoscopy 3 months after initial resection. A large multicenter study identified a 9% absolute reduction in the recurrence of bladder tumors using fluorescence cystoscopy. However, these novel cystoscopic techniques are costly, and the proportion of lesions detected through fluorescence and narrow-band cystoscopy that are clinically important is unknown.

Management and surveillance of NMIBC remains costly. Unlike cancers that may be followed with serum laboratory tests or dedicated physical examinations, NMIBC requires expensive invasive procedures. Risk stratification can direct the intensity of surveillance required; patients with low-risk NMIBC can probably be followed less vigilantly than those with intermediate-risk or high-risk NMIBC. However, the investigation of other biomarkers for bladder cancer recurrence has not produced a test that has replaced the need for cystoscopy. Augmentation of conventional cystoscopy with new technologies has not shown convincing cost-effectiveness.