Treatment of non-muscle invasive bladder cancer (NMBIC) requires direct visual appreciation of the tumor. Transurethral resection that is dependent solely on white light cystoscopy (WLC) often fails to accurately stage or completely resect NMIBC. These deficiencies of WLC are significant contributors to the high rates of recurrence and eventual progression to muscle invasive disease. This article looks at technologies that are being used in adjunct to WLC to augment the urologist’s ability to identify, stage, and treat NMIBC.
Key points
- •
Non-muscle invasive bladder cancer (NMIBC) is confined to the epithelium and lamina propria.
- •
Complete eradication and bladder preservation is the goal of transurethral resection (TUR).
- •
Technological limitations contribute to failure of TUR.
- •
Failure results in recurrence and progression.
- •
New imaging modalities that increase the urologist’s ability to detect, stage, and treat NMIBC at TUR can have a significant impact in the treatment of NMIBC.
Fluorescence Cystoscopy
- •
Enhances tumor detection at initial resection for all stages and grades.
- •
Detection of carcinoma in situ is particularly enhanced with fluorescence cystoscopy (FC).
- •
Decreases rates of recurrence.
Narrow Band Imaging
- •
Shows similar efficacy in detection of tumors as FC.
- •
Does not require instillation of fluorophore in bladder before usage.
Optical Coherence Tomography
- •
Provides real-time tumor grading during TUR.
Confocal Laser Endomicroscopy
- •
Allows microscopic examination of urothelium in real-time for histopathologic analysis.
Bladder cancer remains one of the most difficult clinical conundrums for the practicing urologist. On the extremities of the disease, management of bladder cancer is simplified. The progressive nature of muscle-invasive disease dictates definitive management and the gold standard remains radical cystectomy. At the other extreme, the likelihood of a single, subcentimeter, superficial (Ta) low-grade tumor progressing to muscle invasion is less than 1%. Adequate management of such a lesion is eradication by transurethral resection (TUR) and a single instillation of intravesical chemotherapy at the time of resection. Recurrence and progression are monitored by quarterly cystoscopy for 1 year, decreasing in frequency thereafter if the patient remains free of tumor. Most patients, however, lie in between these 2 extremes. For these patients with non-muscle invasive bladder cancer (NMIBC), tumor recurrence and progression to muscle-invasive disease are a lifelong risk despite the multimodal application of chemotherapeutics, immune therapies, and repeated surgical resections.
Non-muscle invasive disease is confined to the epithelium (Ta) or the lamina propria (T1). A patient suspected of having bladder cancer undergoes evaluation with upper tract imaging, urine cytology, and cystoscopy. The standard for evaluation of the lower urinary tract is white light cystoscopy (WLC), followed by TUR of any tumor that is discovered. Accurate assessment of histopathologic grade and depth of tumor invasion is required for the implementation of proper treatment. This implementation is accomplished by complete and accurate endoscopic resection at the time of TUR. For TUR to be successful, the tumor must be removed in its entirety, along with a sufficient margin of surrounding mucosae and the muscle directly beneath the tumor.
WLC is, however, a technology fraught with deficiencies. Complete resection of papillary tumors is problematic and, as Brausi and colleagues demonstrated, often leads to early recurrence. As a result, a single TUR is often insufficient for the task. A repeat TUR, performed within 2 to 6 weeks of the initial TUR, shows residual disease a significant portion of the time depending on stage, grade, and tumor multiplicity. For pathologically confirmed disease confined to the epithelium (pTa), residual tumor is seen between 6% and 37% of the time. The rate is higher for disease that has been pathologically confirmed to invade into, but not through, the lamina propria (pT1) (33%–78%) and carcinoma in situ (CIS) (70%–80%) and incomplete resection is not the only limitation. High-grade Ta or T1 tumors are understaged 10% of the time even when muscle is present in the original resection specimen and will be found to be muscle invasive at re-resection. To what degree deficiencies of surgical technique contribute to this phenomenon is unknown. It is doubtless a factor, but there are also significant technological limitations that contribute to a surgeon’s inability to eradicate a bladder tumor completely at the time of resection.
Further evidence supporting this comes from the difficulty of finding and eradicating CIS as reported from trials that have recently explored the utility of random bladder biopsies. Despite the relative ease with which CIS is detected by bladder wash—fully 90% of patients with CIS are thought to have a positive cytology —standard WLC fails to detect a lesion 50% of the time. CIS is a flat, high-grade noninvasive lesion that, unlike Ta and T1 disease, has a high potential for progression when left untreated. Classically described as a red “velvety” lesion, CIS is often multifocal and may be macroscopically invisible. Occult CIS is common when high-grade pTa or pT1 lesions have been identified and is present upward of 40% of the time.
Of prognostic significance, CIS is, after tumor histopathologic grade, the second most important factor predicting progression to muscle invasive disease. Its detection is therefore paramount for accurate risk stratification and the implementation of treatment. Random bladder biopsies are often performed for just such a reason, as a random survey of the bladder mucosae for occult CIS. One recent study demonstrated random biopsies to be positive for CIS 8% of the time. Of clinical significance was the fact that, in a different study of random biopsying, random biopsies changed the indicated therapy 4.6% of the time.
Finally, the economic impact of bladder cancer is immense. From diagnosis to death, bladder cancer is the most expensive cancer in the United States. The inefficacy of cystoscopy for diagnosis and surveillance, in conjunction with the recurrent nature of the disease, is responsible for much of its financial burden. Technological advancements that can decrease recurrence have the potential to have a significant financial impact on health care costs if implemented properly.
Fluorescence cystoscopy
The topical application of the protoporphyrin 5-aminolevulinic acid (5-ALA) was first performed in the early 1990s by a group of Munich urologists. Before this, it had been shown that the intravenous injection of heme derivatives preferentially accrued in urothelial carcinomas. Heme derivatives are photoactive substances that absorb light at one wavelength and emit at another. Initially, the practical implementation of photodynamic tumor detection using heme derivatives, in particular porphyrin II, was attempted, but ultimately abandoned because of the systemic effects of skin photosensitization due to the porphyrins. Of practical importance, however, was the exposition that a photosensitizer could be used to enhance cystoscopy and tumor detection.
The Munich group, along with several other pioneering groups, showed that WLC could be effectively enhanced by the topical application of the protoporphyrin 5-ALA with minimal risk to the patient. Topical application managed to avoid the systemic complications associated with intravenous administration; however, the lipophilicity of 5-ALA still allowed for reasonably rapid local absorption over a period of several hours. These small proof-of-principle studies hinted, but were insufficiently powered to prove, that the instillation of 5-ALA before TUR could result in increased detection of all lesions and, in particular, 5-ALA had a distinct advantage in the enhanced detection of CIS. Subsequent randomized, prospective clinical trials would support these initial claims. 5-ALA has been consistently shown to enhance the detection of CIS greatly, with 33% to 36.7% more lesions detected. An advantage was also seen with the detection of pTa lesions and pT1 lesions, but it is not as pronounced (7%–11% and 3.6%–13%, respectively).
Enhanced detection at initial resection should lead to an increase in the rate of complete resection and a decrease of residual tumors seen at re-resection. Filbeck and colleagues demonstrated this in a randomized, prospective clinical trial. The WLC group had a 25.2% rate of residual tumor at 6 weeks, compared with 4.5% for the 5-ALA arm of the trial. On a stage-by-stage basis, the rate of residual tumor at 6 weeks for pTa was 17.8% and 3% for WLC and fluorescence cystoscopy (FC), respectively. pT1 had a greater rate of residual tumor for both, with the WLC arm having a rate of residual tumor of 36% and the FC arm having a rate of 11.8%. The difference was even greater when comparing residual CIS. In the WLC arm, CIS was found in 80% of patients at 6 weeks, whereas there was no residual CIS seen in the FC arm. This group also looked at the rate of recurrence in this trial and in a subsequent study with an 8-year follow-up. The rate of overall recurrence for the WLC arm was 44% versus 16% for the FC arm. Further analysis based on risk stratification was also performed by this group. At 8 years there was a significant difference in recurrence-free survival (RFS) when comparing the FC arm to the WLC arm for all risk groups. The WLC group performed remarkably similarly regardless of risk stratification at 8 years (42%–45% for low, intermediate, and high risk stratification). The benefit of FC was greatest for low-risk patients (patients with solitary low-grade pTa or pT1 tumors), with 81% having no recurrence at 8 years. As tumor multiplicity and grade increased, however, the advantage narrowed but was still significant. The intermediate-risk group had the same grade and stage as the low-risk group, but presented with multiple tumors initially, which resulted in an RFS of 70% in the FC arm at 8 years. High-grade disease, including CIS, conferred additional risk of recurrence at 8 years. Nonetheless, 60% of patients remained recurrence-free in the high-risk group.
Further refinements have resulted in hexaminolevulinate (HAL)—Hexvix in Europe (Photocure ASA, Oslo, Norway) and Cysview in North America (Photocure USA, Princeton, NJ)—largely supplanting 5-ALA as the protoporphyrin of choice. HAL is a hexyl ester derivative of 5-ALA. It is significantly more potent than 5-ALA due to its increased lipophilicity and tissue solubility. This increased lipophilicity and tissue solubility allow for decreased instillation time and surgeon convenience for practical usage during TUR without any increase in its complications. Like 5-ALA, HAL absorbs blue light (wavelength 380–480) and emits red light ( Fig. 1 ). Polarized light filters allow for visual enhancement of a tumor that appears red on a dark blue background.
Stenzl and colleagues have shown similar results with HAL in an international randomized, prospective trial that compared WLC-based and HAL-based FC. In this study there was a 32% increase in the detection of CIS in the fluorescence arm. Moreover, enhanced detection of tumors that were clinically relevant was also noted in this trial. The patients in the WLC arm underwent white light interrogation only, whereas the fluorescence arm group underwent an initial white light examination that was then followed by blue light interrogation. Tumors were mapped at both times and the results were recorded. In the fluorescence arm there was a 16% increase in the number of lesions detected. Of these lesions, 41% were high-grade Ta or T1 lesions that would not have been detected by WLC alone. This group also looked at rates of recurrence in a subsequent study. The time to recurrence was significantly improved in the fluorescence arm when compared with the white light–only arm (16.4 months vs 9.4 months, respectively).
A further advantage that FC has over standard WLC is its ability to more accurately demarcate tumor from normal mucosae and aid in the complete resection of a tumor that is visible with both modalities. A single-center prospective, randomized trial by Geavlete and colleagues found that 8.5% of tumors that had been resected under WLC had pathologically confirmed positive margins that could be visualized by FC ( Fig. 2 ). This group also showed in a subsequent study that at the time of initial resection FC was superior in detecting multiple tumors when compared with WLC (35.2% vs 14.1%, respectively). Furthermore, their analysis showed that HAL-based FC modified the risk stratification and the postoperative treatment regimen in several patients and aided a more accurate pathologic classification. Of particular note, patients who would had been declared “tumor free” by WLC, but the addition of HAL resulted in increased detection of bladder tumor. Stated another way, at initial TUR, there was a failure to find a tumor—the patient was “tumor-free”—7.7% of the time. In the FC arm, only 1.4% of patients were designated “tumor-free.”
Not all studies to date have shown a uniform advantage of FC over WLC. Schumacher and colleagues, in a randomized, multicenter, prospective trial, failed to show any advantage with respect to rates of recurrence for FC over WLC. An overall advantage for FC was seen in tumor detection and was consistent with other published trials—33% advantage in detecting CIS, 13% for pT1, and 6.7% for pTa—but no advantage was seen in rates of recurrence at 12 months in either low-risk or high-risk patients. One factor that could contribute to the difference seen in this trial is that it was a multicenter study. As a result, some of the surgeons who performed the TUR resections had limited experience with FC before their inclusion in the study.
Criticisms of HAL and 5-ALA are primarily twofold. One concern is its lack of specificity and high false-positive rate. The other concern is that neither influences disease progression and the number of individuals who progress to muscle-invasive disease is unchanged. Nonspecific uptake of heme derivatives does occur. This nonspecific uptake of heme derivatives is most notable in patients with a prior history of TUR and bacillus calmette-guerin (BCG) therapy. Inflammatory lesions, scars, and granulomas can all appear as red lesions on a blue background, the same as an FC-positive bladder tumor. Initial studies reported false-positive rates for FC as high as 51%. However, the inability to diagnose strictly based on appearance is not unique to FC and false-positive biopsying occurs in even in the most experienced hands under WLC. Most recent large prospective, randomized trials have similar rates for false-positive biopsying (10%–15%) regardless of the modality directing the biopsy, with a slight increase of 1% or 2% for blue light–directed biopsies.
With regard to progression, FC-based cystoscopy has not been conclusively shown to improve progression-free survival. Stenzl and colleagues found no difference between the 2 arms, whereas others have shown a modest improvement of 1% or 2%, but these were not statistically significant.
Several studies have looked into to the cost-effectiveness of fluorescence-based cystoscopy. Stenzl and colleagues first looked into the cost-effectiveness of 5-ALA and found that the savings could be substantial if the difference in the rate of recurrence was 20%. The savings were largely due to avoided TURs. This savings covered the added expense of the supplemental equipment needed for FC by a factor of 3. Daniltchenko and colleagues also found a significant cost-savings based primarily on the decrease in the total number of TURs that would need to be performed. Most convincingly, Zaak and colleagues performed a cost analysis on a per-patient basis and analyzed patients who had undergone WLC and FC. They found a savings of about $250 per patient per year over a 7-year follow-up.
Narrow band imaging
The propensity of bladder cancers to be hypervascular is exploited in narrow band imaging (NBI). White light emanating from the lamp source is filtered into 2 discrete wavelengths—415 (blue) and 540 (green)—that have only superficial tissue penetration. Correspondingly, these are the peak absorptive ranges for hemoglobin. When combined with high-resolution magnifying endoscopy, the vascular architecture of a bladder lesion is accentuated and the disparity between malignant and benign is heightened. The technology was initially used for the characterization of colonic lesions and was adapted to the cystoscope shortly thereafter. Vascular structures appear dark brown (capillaries) and green (veins) on a pink-white background (normal mucosae) ( Fig. 3 ). Unlike fluorescence cystoscopy, it does not require the instillation of a photoactive substance. Likewise, it avoids the added costs of the porphyrin and time associated with its instillation, and is more convenient to implement.
