Management of Low-grade Upper Tract Urothelial Carcinoma: An Unmet Need
Jay Raman, MD, FACS,1 Neal D. Shore, MD, FACS2
1Division of Urology, Penn State Health Milton S. Hershey Medical Center, Hershey, PA; 2Atlantic Urology Clinics and Carolina Urologic Research Center, Myrtle Beach, SC
Upper tract urothelial cancers (UTUC) are frequently managed by radical nephroureterectomy (RNU), a major operative procedure that may entail short-term morbidity and long-term decline in renal function. Kidney-sparing procedures offer a less invasive alternative to RNU for low-risk, low-grade UTUC (LG-UTUC). They are associated with similar disease-specific survival rates and better long-term renal function, albeit with a potentially increased risk of recurrence. Strategies to decrease LG-UTUC recurrence include improved risk stratification and enhanced endoscopic instrumentation. Chemoablation may represent an alternative, innovative kidney-sparing approach for LG-UTUC.
[Rev Urol. 2020;22(1):1–8]
© 2020 MedReviews®, LLC
Upper tract urothelial cancers (UTUC) may arise within the upper tract of the genitourinary system including the ureter and the pelvicalyceal system. They are a heterogeneous group of cancers whose prevalence appears to be increasing. They now affect 1 to 2/100,000 persons in the United States.1
Low-grade UTUC (LG-UTUC) account for approximately 30% of all UTUC, and are typically papillary tumors that are less aggressive than highgrade UTUC.2 As low-grade tumors develop, they may lead to symptoms such as urinary obstruction and/or hematuria, both of which need to be addressed. They are not, however, usually associated with high rates of cancer-specific mortality. An analysis of the Surveillance, Epidemiology, and End Results (SEER) database showed that high-grade disease, but not low-grade disease, was a significant predictor of cancer-specific mortality (Figure 1).3 Nonetheless, in a minority of cases, low-grade tumors can progress to high-grade disease, become locally invasive, and then metastasize.
As the risk of recurrence and progression for patients with LG-UTUC is not insignificant and as the complex anatomy of the upper tract renders endoscopic intervention challenging, low-grade disease has, oftentimes, been treated similarly to high-grade disease. Radical nephroureterectomy (RNU), which includes the resection of the kidney, ureter, and ipsilateral bladder cuff, has been a standard treatment for both low- and high-grade UTUC. In recent years, however, concerns have been raised that RNU can be associated with significant adverse surgical sequelae and can diminish long-term renal function, especially for the elderly. Kidney-sparing approaches, which are indicated in cases where RNU is contraindicated, can offer an alternative to RNU for patients with low-grade tumors who are at low risk of disease progression. This possibility has prompted the development of models that help identify which patients with low-grade tumors are candidates for kidney-sparing procedures. Current 2017 European guidelines define low-risk UTUC as low-grade, non-invasive, unifocal tumors that are <2 cm in diameter (Table 1).4
RNU is a major surgical intervention that can lead to both shortand long-term complications. Thirty-day complications were assessed in a study of 92 consecutive UTUC patients who underwent open or minimally invasive RNU.5 Thirty-eight percent of patients experienced a complication (49 complications in 35 patients). Overall, 75% of these complications were hematologic, gastrointestinal, or infectious. Gastrointestinal ileus and anemia requiring transfusion were the two most frequently reported complications; 28% of all complications were classified as major (Clavien III-V).5
Complete kidney removal also has significant impact on long-term kidney function even when a normal contralateral kidney is present preoperatively.6 The impact of RNU on estimated glomerular filtration rate (eGFR) was illustrated in a retrospective study of 374 patients who had a median preoperative eGFR of 72 mL/min/1.73m2.6 Six months after surgery, eGFR had declined by 32% to a median of 50 mL/min/1.73m2. Tumor grade had no effect on eGFR.6 In another retrospective analysis of 666 patients, the percentage of patients who had an eGFR ≥60 mL/min/1.73 m2 decreased from 37% of patients preoperatively to 16% of patients 3 to 6 months after RNU (P < 0.001); the percentage of patients with an eGFR ≥45 mL/min/1.73 m2 decreased from 72% to 52%.7 Univariate analyses showed that reduced kidney function (eGFR <60 mL/min/1.73 m2) correlated with reduced long-term overall survival after RNU in UTUC patients who had neither undergone adjuvant therapy nor experienced disease recurrence.7
The risk of developing significant chronic renal insufficiency disease post-RNU has also been shown to increase with age and with comorbidities, such as hypertension, diabetes mellitus, and obesity. A multivariate logistic regression analysis of data from 749 UTUC patients (2/3 RNU and 1/3 kidneysparing surgery) showed that RNU, as well as diabetes mellitus, hypertension, body mass index ≥30 kg/m2, age ≥60 years, and smoking, were risk factors for developing an eGFR <60 mL/min/1.73 m2.8 Patients with UTUC are typically elderly1 and have attendant comorbidities; 25% of patients have diabetes and 60% are hypertensive.8 As a result, many patients with UTUC are at especially high risk of developing clinically significant renal impairment post-RNU.
For LG-UTUC patients, choosing between RNU and a kidney-sparing procedure requires an assessment of the risk of surgical complications, of renal insufficiency with its attendant complications, and of recurrence and progression associated with a non-extirpative approach. A model has been developed to evaluate the risk of complications with RNU (Figure 2),9 but no prospective trials have compared outcomes after RNU and kidney-sparing techniques in patients with low-grade, low-risk tumors. Retrospective studies (Table 2) that compare outcomes can be used to support this decision-making process.10-16 These data, however, do not generally compare patients with similar characteristics and, therefore, inherently reflect physician understanding (and potential bias) about which patients are best suited for the respective procedures.
In retrospective studies of LG-UTUC, cancer-specific mortality is not generally higher when kidney-sparing procedures are compared with RNU.14,15 In a retrospective study of 202 patients with low-risk LG-UTUC, defined as a unifocal lesion without local invasion and a diameter <1 cm, there was no difference in overall survival between endoscopic treatment and RNU.15 In one retrospective study of 453 patients over the age of 65 years with low-grade disease, however, endoscopic surgery was a significant and independent long-term predictor of all-cause mortality (hazard ratio [95% confidence interval], 1.6 [1.3-2.2]) and cancer-specific mortality (hazard ratio [95% confidence interval], 2.1 [1.0-4.1]) compared with RNU.16 Survival curves became significantly different 24 months after intervention. This study is of particular note because baseline characteristics were matched using propensity scoring; therefore, the number of patient characteristics that could influence the surgical approach and outcome was lower than in most other retrospective studies.16 The matching algorithm did not, however, specifically include preoperative kidney function or disease characteristics (other than grade), both of which could have been contributors to overall risk in the endoscopic surgery group. Together, these results underscore the need for randomized prospective trials that compare the surgical interventions and assess the impact of selecting patients with low-grade tumors who are also low risk.
Although kidney-sparing approaches appear to have acceptable survival outcomes in patients with low-grade tumors, recurrence rates are not insignificant and are higher with kidney-sparing procedures than with RNU.10-12,14 A 2012 meta-analysis by Cutress and colleagues showed that patients with low-grade tumors who underwent ureteroscopy with ablation or percutaneous nephroscopic resection had a 3-year disease-specific survival rate of 96% to 100% and a 3-year rate of upper tract recurrence of 35% to 48% depending on the type of kidney-sparing procedure.11 In another study by Cutress and colleagues, in which 34 patients with grade 1 UTUC underwent endoscopy, 5- and 10-year diseasespecific survival rates were 100% and 80%, respectively, and the estimated 5- and 10-year rates of upper tract recurrence of grade 1 tumors was 36.6% and 78.7%, respectively. Similarly, the rate of bladder recurrence of grade 1 tumors increased from 11.5% to 48.2% after 5 and 10 years, respectively.12
Even in cases where disease recurs after a kidney-sparing procedure, not all recurrences necessitate RNU. The rate of kidney preservation after kidney-sparing procedures is high; it varies from approximately 60% to 80% depending on the study.12,13,17 In the 2012 Cutress and colleagues study, the rate of renal preservation in patients with grade 1 tumors was 96% after 5 years and 62% after 10 years.12 These numbers were significantly higher than those observed in patients with grade 3 tumors. These data suggest that in some cases, even though recurrence leads to removal of the kidney, for many low-grade tumors, a kidney-sparing approach can successfully preserve the kidney.
Lower surgical perioperative morbidity and preservation of long-term renal function are the expected benefits of kidney-sparing approaches. Complications, however, have not been systematically reported.17 In at least one retrospective analysis (N = 96), overall complications were significantly lower after endoscopic procedures than after RNU (9.3% vs 29.0%; P = 0.0002).13
Data suggest that using kidney-sparing procedures to resect LG-UTUC does not increase disease-specific or overall mortality and may reduce complications. Prospective data in well-defined patient populations that are both low grade and low risk are needed to confirm the data gathered from retrospective analyses.
Although kidney-sparing procedures provide an important alternative to RNU for low-risk LG-UTUC, challenges remain.
The well-known anatomical complexity of the calyceal system and the potential narrow ureteral lumen can result in access and visualization difficulties that can limit endoscopic tumor measurement, tumor biopsy technique, and subsequent tumor ablation. Endoscopic upper tract biopsies can require specialized biopsy forceps and baskets that may not be accessible in all surgical sites of service. As a result, histopathologic grading is susceptible to inter-observer variability and may be inaccurate.18 In a retrospective analysis of 77 nephroureterectomy specimens, 42% (8/19) of tumors categorized as low grade preoperatively using a combination of ureteroscopic biopsy and cytology were re-categorized as high grade after analysis of the surgical samples.19 This diagnostic uncertainty complicates the assessment of risk of disease progression, and errors in categorization increase the overall rate of recurrence in patients undergoing kidney-sparing procedures. Accurate and reliable preoperative diagnosing of grade remains an important technical challenge.
Poor access and visualization can lead to limited ablation and/or aborted procedures. In a small retrospective study of 41 patients with UTUC who underwent ureteroscopy within 60 days of laser tumor photoablation, UTUC tumors were detected in 51.2% of patients at the 2-month time point.20 Eighty-six percent of these tumors were in the same location as the original tumor. The authors suggested that these data may be indicative of incomplete ablation of the primary tumor during the initial procedure rather than tumor recurrence.
Adjuvant intracavitary therapy can be used post-surgically to reduce the rate of recurrence. A number of adjuvant therapies, such as bacillus Calmette-Guérin, mitomycin-C, and other immunomodulatory agents, have been incorporated into clinical practice, albeit with incomplete conclusions about treatment efficacy.21 In theory, antegrade instillation via a nephrostomy tube reliably exposes the urothelium to the topical agent, but the question of extravasation as well as tumor seeding have raised concerns regarding the technique’s optimal use. External-internal stents, such as open-ended ureteric stents ≥4F, have been utilized, but are somewhat time intensive both for the clinic and the patient and can occasionally lead to pyelovenous backflow during instillation and associated complications.21
The efficacy of adjuvant treatments is reduced further by the difficulties associated with maintaining contact between the selected drug treatment and the tumor in the presence of a continuous flow of urine and physiologic calycealureteral peristalsis. The most recent European guidelines include topical adjuvant therapies administered via antegrade instillation to reduce the risk of recurrence after kidney-sparing procedures but note that the level of evidence in support of this approach is low regardless of the instillation technique.4,21
Low-risk UTUC is a relatively new concept that has not been fully integrated into urologic practice. It builds from grade categorization to better identify which low-grade tumors are truly at low risk of progression. Educational efforts are needed to communicate the importance of distinguishing between low- and high-risk tumors.
The definition of low risk is still evolving. In 2015, the European Association of Urology guidelines set the size cut-off for low-risk tumors at <1 cm, but increased it to ≤2 cm in 2018.4,22 This size cut-off is likely to continue to change as better visualization techniques increase the accuracy of measurement. A recent retrospective analysis of patients diagnosed at baseline with low-grade, unifocal lesions without local invasion15 showed no significant difference in overall survival between endoscopic treatment and RNU in patients with lesions <1 cm (n = 202; 5-year overall survival of 69.2% vs 73.7%, respectively; P = 0.79; hazard ratio [confidence interval]: 1.08 [0.63-1.82]). By contrast, when a cut-off for size of <2 cm was used, overall survival was significantly lower in the endoscopy treatment group than in the RNU group (n = 850; 5-year overall survival of 69.3% vs 75.2%, respectively; P = 0.006; hazard ratio [confidence interval]: 1.43 [1.11-1.85]). The cut-off for no difference in overall survival for patients treated by endoscopy versus RNU was ≤1.5 cm.15 Although these data need to be validated with additional studies as the patients who received endoscopic treatment were older and healthier, they do illustrate the importance of the size cut-off in the definition of low risk.15
Similarly, the definition of low grade may have a significant impact on outcomes. European guidelines, which are based on the 2004 World Health Organization categorization, distinguish between low- and high-grade tumors.22 This classification was developed to reduce the ambiguity of the three-tiered grading system, in which roughly half of the grade 2 tumors met the criteria for low grade and the other half for high grade.
Clinical decision-making tools that combine patient comorbidity indices, histopathology, and tumor size should help identify which low-grade tumors are at low risk of progression and recurrence. Currently, such tools are focused on outcomes after RNU. Xylinas and colleagues developed a nomogram to predict intravesical recurrence after RNU.23 It included age, sex, history of bladder cancer, ureteral tumor location, laparoscopic surgical technique, endoscopic distal ureteral management, tumor stage, concomitant carcinoma in situ, and lymph node involvement.23 A different preoperative risk classification model included neutrophil-lymphocyte ratio and hydronephrosis to assess the risk of cancer-specific survival and recurrence-free survival after RNU.24 Similar tools could be developed to predict risk and outcomes after kidney-sparing procedures and to support the decision to proceed with a kidney-sparing procedure rather than an RNU.
Smaller, flexible ureteroscopes have increased access to the upper urinary tract and several image enhancement technologies have improved visualization.25 Digital ureteroscopes offer high-definition imaging, autofocus capabilities, and digital magnification. Narrowband imaging has been shown to improve the visualization of the upper tract, increase the number of detected tumors, and extend the limits of the tumors compared with white light.26 Similarly, photodynamic diagnosis, which uses photosensitizers and guided flexible ureteroscopes to identify malignant tissue, has been shown to enhance visualization and increase the number of tumors identified compared with white light ureteroscopy.27 A systematic review and metaanalysis of seven studies (194 patients) showed that the sensitivity and specificity of photodynamic diagnosis to detect UTUC was 95.8% and 96.6%, respectively.27 Better visualization of tumors should lead to more complete ablation and to better risk stratification as both the number of tumors and the size of tumors are key components in the evaluation of risk. Use of such techniques could help reduce the number of patients falsely categorized as low risk.
The accuracy of grading is still limited by the quality of the sample. Some ureteroscopic biopsy devices have been shown to improve the quality of the specimen, but not the accuracy of the histopathologic diagnosis. In a recent study that compared backloaded cup forceps, nitinol basket, and standard cup forceps specimens, higher quality urothelial samples were collected with the backloaded cup forceps and nitinol basket than with the standard cup forceps.28 The rate of grade concordance with surgical samples, however, was similar regardless of the biopsy collection device (75%-78%; P = 0.177).
In vivo high-resolution imaging, which provides real-time grading of urothelial lesions, may offer an alternative approach. A recent study of 35 patients with papillary UTUC showed that an attenuation coefficient of the optical coherence tomography signal of 4.0 mm21 could be used as a cut-off to discriminate between highand low-grade papillary UTUC.29 The sensitivity and specificity of this cut-off for high-grade papillary UTUC were 83% and 94%, respectively. In addition, a systematic review and meta-analysis (14 studies, 2031 patients) showed that fluorescence in situ hybridization, which detects genetic aberrations, had a better sensitivity for UTUC than cytology (84.0% vs 40.0%, respectively) and similar specificity (89.5% vs 95.9%, respectively).30 This study did not, however, specifically look at specificity and sensitivity for low-grade tumors.
Initiatives to increase dwell time, the time during which the therapeutic agent is in contact with the tumor, are ongoing. Thermo-reversible hydrogels are being developed as a chemotherapy delivery technique. Because they are liquids at chilled temperatures, they can be administered as a liquid via antegrade or retrograde approaches. They solidify into gels at body temperature, allowing the drug to remain in contact with the tumor for several hours before eventually being removed by urine flow. UGN-101 (UroGen Pharma, New York, NY) is a thermoreversible hydrogel formulation of mitomycin that has been shown in animal trials to remain in the upper tract and to release mitomycin for a period of 4 to 6 hours.31
This formulation has been studied as a chemoablative therapy for UTUC with the goal of eradicating tumors in situ. In a compassionate use trial, UGN-101 was administered to 22 patients with life-threatening or debilitating UTUC and no other therapeutic options. Complete response, partial response, and no response were reported in 36%, 23%, and 9% of patients, respectively. Among the patients with complete response, 63% experienced a recurrence by 30 months.32 In addition, UGN-101 is being evaluated in a prospective, open-label, single-arm trial named OLYMPUS. Seventy-one patients with LG-UTUC were given 6 weekly instillations of UGN-101 via a retrograde catheter. Four to 6 weeks after the last treatment, 41 patients (61%) showed complete response. After 9 months of follow-up, 3 of these patients had relapsed. Urinary tract infection and stricture were the most common adverse events.33 Analysis of the full data set when the study is completed will confirm whether such an approach could be used in low-risk patients with LG-UTUC in lieu of RNU and endoscopic kidney-sparing approaches.
Although advances in our understanding of tumor histopathology have demonstrated that there are significant differences between LG-UTUC and high-grade UTUC, the two disease states are often treated similarly. Consensus is growing that kidney-sparing approaches should be a primary consideration in the treatment of LG-UTUC. Endoscopic management of low-risk LG-UTUC currently offers an alternative to RNU that, in most cases, spares the kidney. The anatomical complexity of the upper tract, however, continues to make the use of conventional kidney-sparing techniques challenging; and better risk models, instrumentation, and tumor visualization are needed to improve outcomes. Chemoablation, which is a technique that improves contact of drug with the urothelium, could provide an alternative, kidney-sparing approach for primary tumor eradication. Chemoablation may provide the ideal combination of low recurrence rates, acceptable shortand long-term safety, and tumor ablation without kidney removal.
Dr. Shore receives research funding and/or consulting fees from Amgen, Astellas, Astra Zeneca, Bayer, BMS, Dendreon, Ferring, Janssen, Merck, Novant, Nymox, Pfizer, Sanofi, Tolmar, and UroGen. He has no stocks/salaries/patents to disclose. Dr. Raman receives research funding and/or consulting fees from Pacific Edge Biotechnologies, MDxHealth, and UroGen. He has an investment interest in American Kidney Stone Management.
The authors thank Phase to Phase Strategy, LLC, for their help in drafting this manuscript. Their work was paid for by UroGen Pharma.
Jelmyto™ (mitomycin) for pyelocalyceal solution (UGN-101) was approved by the FDA for the treatment of adult patients with LG-UTUC in April 2020.