Cystoscopy is an essential component of the evaluation for patients with suspected UTUC due to the risk of concurrent lower tract urothelial cancer in this population.

If there are no contraindications to its use, clinicians should perform a multiphase computed tomography (CT) scan with excretory phase imaging of the urothelium. A systematic review by Janisch et al. highlighted that CT urography was associated with a pooled sensitivity of 92% (95% CI: 85% to 96%), pooled specificity of 95% (95% CI: 88% to 98%), and a summary area under the ROC curve of 0.97 (95% CI: 0.96 to 0.98)¹¹ using histopathology or a negative clinical follow-up as referent standard. Two additional studies published after the systematic review reported results consistent with this review.^{12, 13}

In patients with contraindications to contrast-enhanced CT such as chronic kidney disease (CKD) or untreatable allergy to iodinated contrast medium, clinicians may utilize magnetic resonance (MR) urography. In a study of 88 patients, Takahashi et al. found MR urography was associated with a per-patient sensitivity of 63% to 74% and specificity of 96% to 97% for diagnoses of UTUC.¹⁴

For patients with contraindications to multiphasic CT and MR urography, clinicians may utilize retrograde pyelography in conjunction with non-contrast axial imaging to assess the upper urinary tracts. Renal ultrasound (US) may also have utility in providing additional diagnostic assessment. In a study of 151 imaged urinary tracts, retrograde pyelography was associated with a per-tract diagnostic sensitivity of 96%, specificity of 97%, positive predictive value (PPV) of 87%, and negative predictive value (NPV) of 97%¹⁵ with referent standard of histopathology (biopsy or final surgical pathology) and clinical follow-up for 3-5 years. While no reports have specifically evaluated the diagnostic accuracy of US for UTUC alone, one study of 575 patients evaluated the accuracy of US (followed by CT only if US was suspicious) for the diagnosis of upper urinary tract malignancies (renal cell carcinoma or UTUC) in a population with hematuria.¹³ In this cohort, US was associated with high sensitivity (100%) and specificity (97%) for upper tract malignancies, with a PPV of 18% and NPV of 100% with referent standard being histopathology from biopsy or radical nephroureterectomy (RNU). An important caveat is that renal cell carcinoma accounted for two-thirds of the cases in this series and that the study was not structured as a head-to-head comparison of US and CT.¹³ Recognizing these limitations, the Panel cannot recommend the routine use of renal sonography for evaluating patients with suspected UTUC though recognizes there may be clinical utility in the narrow indications where contrast-based CT and MRI studies cannot be performed.

After initial concerns of UTUC are discovered on imaging, endoscopy by antegrade or retrograde approach with tissue sampling and cytologic washing should be performed when diagnostic and prognostic details are needed. When performed, this diagnostic procedure should be performed as a standardized endoscopic examination including elements pertinent to clinical decision-making. At ureteroscopic evaluation, clinicians should document key descriptive features of UTUC including tumor size, number, location, focality, and appearance. These factors may guide further diagnostic testing and inform therapeutic interventions as well as provide points of comparison for subsequent ureteroscopic surveillance. An example checklist for standardized endoscopic diagnostic examination is provided in Table 3. The Panel recognizes and emphasizes a distinction between diagnostic and therapeutic endoscopic procedures for UTUC. Diagnostic procedures are intended as low-impact and typically brief interventions to provide clinical information required for risk-adapted patient care whereas therapeutic endoscopic procedures are often longer, more technically involved operations undertaken with curative intent and greater risk for surgical complications. Under either circumstance, it is recommended that standardized reporting of findings be documented. Diagnostic and therapeutic procedures may overlap under circumstances where discovered tumors are small and can be easily and completely treated at the time of endoscopy. It is further recognized that endoscopic procedures carry risks including perforation and tumor seeding, inside and outside the urinary tract, that may complicate future management. Data on the comparative risks of retrograde vs antegrade percutaneous approaches are insufficient to address the concern regarding potential risk of tumor seeding with percutaneous techniques.

Table 3: Standardized Upper Tract Endoscopy Suggested Reporting Elements

Different techniques exist for endoscopic approach including retrograde ureteroscopy versus antegrade percutaneous nephroscopy and/or ureteroscopy. Both approaches allow visualization of suspected lesions, and a variety of biopsy techniques can subsequently be employed which can successfully yield tissue adequate for diagnosis. Factors such as tumor location, configuration, size, and patient factors (e.g., prior cystectomy) may influence the chosen approach or technique. Data on comparative effectiveness across all clinical situations are lacking.     

Six studies evaluated the diagnostic accuracy of endoscopic (ureteroscopic) biopsy for UTUC, compared against a reference standard of surgical pathology (e.g., following nephroureterectomy [NU]) or surgical pathology plus clinical follow-up.^16-21 One study (n=93 patients with 118 biopsies) found ureteroscopic biopsy with forceps was associated with diagnostic sensitivity of 83% and specificity of 100%.¹⁸ Another study (n=45) reported fluoroscopically guided retrograde brush biopsy was associated with diagnostic sensitivity of 91% and specificity of 88% for UTUC.²¹ Additional details of biopsy and cytology sampling accuracy are provided in Appendix I. Although the sensitivity and specificity of biopsy by forceps or loop for yielding a diagnosis of UTUC may be higher than brush biopsy and/or fine-needle aspiration (FNA) and thus be preferred, patient and tumor characteristics will likely dictate the optimal biopsy technique. Mucosal abnormalities may be difficult to biopsy effectively and thus attempted tissue confirmation may be facilitated with the use of brush biopsies or percutaneous image-guided biopsy.

The Panel recognizes there are rare situations where endoscopic upper tract evaluation may not be necessary, when other diagnostic means clearly confirm the diagnosis of UTUC and thus histologic tissue confirmation is not clinically required. Such scenarios may include those patients with high-grade (HG) selective cytology or other source of tissue diagnosis, and clear and convincing radiographic findings of upper tract urothelial-based tumor(s) such as patients with an obvious enhancing, urothelial based soft-tissue filling defect on contrast-enhanced imaging with urography. Such situations may be particularly relevant in patients with a history of HG urothelial cancer. Other clinically justifiable scenarios for omitting diagnostic endoscopic evaluation may occur when findings would not influence decision-making, such as patients with severe co-morbidities who are ineligible for intervention or request expectant management. In such cases it is recommended that documentation of clinical rationale is provided.

Urine cytology can be helpful in identifying carcinoma in the upper tracts. Adjunctive cytologic barbotage washing with saline obtained from selective ipsilateral collection prior to use of any contrast is preferred to a voided urinary specimen due to improved cellular yield, to avoid potential contamination in case of concomitant bladder and/or prostatic urethral disease as well as theoretical dilution of the specimen from a normal contralateral unit, all of which further reduce sensitivity.²² Urine cytology classification has been standardized under The Paris System, which prioritizes the identification of HG cells while minimizing the ambiguity of non-HG findings. By this convention, urine cytology is reported according to seven categories: nondiagnostic, negative for HG urothelial carcinoma (NHGUC), atypical urothelial cells (AUC), suspicious for HG urothelial carcinoma (SHGUC), HGUC, low-grade (LG) urothelial neoplasm (LGUN), and other malignancies.²³ Adoption of The Paris System began in 2016, taking time to become more widely accepted, and thus may impact interpretation of studies with data obtained prior to use of this standard.  

Urine fluorescence in situ hybridization (FISH) testing may also be helpful in the diagnosis of UTUC. One previously described systematic review²⁴ including 14 studies (N=2,031) found that FISH was associated with high diagnostic accuracy for identifying UTUC with a pooled sensitivity of 84% (95% CI: 74% to 90%; range: 52% to 100%) and a pooled specificity of 90% (95% CI: 85% to 93%; range: 33% to 96%). The pooled positive and negative likelihood ratios were 7.96 (95% CI: 5.87 to 10.81) and 0.18 (95% CI: 0.11 to 0.29), respectively. Based on the head-to-head comparisons in the review, sensitivity of FISH was higher than for voided urine cytology, with similar specificity. However, use was not evaluated for selective, instrument-obtained samples from the suspected upper tract. Analogous to cytology, selective collection from the suspected renal and/or ureteral unit likely improves performance characteristics. Given the high sensitivity and low specificity of FISH compared to voided cytology, the Panel acknowledges the yet uncertain role of FISH and suggests such testing may be considered adjunctively to adjudicate atypical or suspicious cytology results.

A common clinical scenario when managing patients with UTUC, the finding of urothelial tumors in the lower tract (bladder or urethra) warrants appropriate independent guideline-directed management in the same surgical setting by biopsy, resection or ablation as clinically indicated. This feature of UTUC has been described clinically and further investigated through genomic studies, which show clonal similarity between upper and lower tract tumors, suggesting either downstream or upstream tumor implantation as a potential mechanism. The pathology findings from bladder tumor sampling often reflects that of upper tract tumors, though not reliably enough to be used as rationale for avoiding separate upper tract endoscopy and biopsy when feasible.^{2, 25}

Consensus on prioritization of procedure sequencing (managing bladder before or after same-setting ureteroscopy) is lacking and heavily scenario-dependent. Rationale for managing the bladder first include optimizing visualization within the bladder, avoiding back-pressure or back-washing into the upper tract in the case of post-ureteroscopy stenting, and permitting final confirmation of bladder hemostasis. Addressing the upper tract first may be preferred in cases of bulky bladder tumor involvement where complete resection is not possible or bulky upper tract disease in which risk assessment is the priority. Seeding of tumors from bladder to upper tract or from upper tract to the lower tract have been raised as legitimate concerns which some have addressed by advocating use of ureteral access sheaths in such circumstances, yet the benefits of this approach require further prospective study. 

Perforation or disruption of the urothelium in patients with UTUC can risk tumor seeding outside the urinary tract. Precautionary measures in cases of difficult ureteral access such as avoiding dilation or placing a stent without performing ureteroscopy and then returning one-two weeks later to repeat the procedure (pre-stenting) can decrease the risk of iatrogenic injury and provide opportunity for a safer and more successful procedure. Recognized perforation or injury events should be documented with immediate cessation of the procedure as soon as safely possible with additional steps to limit sequelae (e.g., stenting, bladder decompression with urethral catheter drainage to limit reflux, nephrostomy tube placement in cases of a completely obstructive ureteral tumor and evidence of contrast extravasation).

Findings from selective cytology and retrograde pyelography may provide useful, objective and sufficient information for risk stratification when endoscopic examination of the involved upper tract is not possible.²² Example scenarios may include washings taken at the time of percutaneous nephrostomy tube placement or during attempted retrograde ureteroscopy that is abandoned for safety concerns. Cytologic sampling from the upper urinary tract, either by barbotage (irrigation and aspiration) or by irrigation with passive collection (washings) can be used to improve cellular yield for cytologic evaluation and best performed prior to pyelography to avoid artifactual cellular changes from contrast solutions. The Panel recognizes that this approach is supported by evidence associated with Statement 2 above and felt that guidance on this scenario was warranted as a Conditional Recommendation to describe means for risk-directed patient care in the setting of limited data from endoscopy, biopsy, and imaging.  

Indications for ureteroscopy or percutaneous endoscopy of the upper urinary tract include such findings as lateralizing hematuria, suspicious selective cytology, and radiographic presence of a mass or urothelial thickening. Endoscopic procedures have risks for patient injury and the potential for tumor seeding in the presence of urothelial cancer. Performing upper tract endoscopy in the setting of a completely normal contralateral upper urinary tract without clinical indication or as a “screening” procedure is unnecessary, placing patients at undue risk and should not be performed.

The significant role of hereditary risk factors in numerous malignancies is well recognized and a topic that care providers must be familiar and comfortable discussing with patients.

LS is common among patients with UTUC, accounting for an estimated 7-20% of U.S. cases. However, LS is frequently unrecognized as a risk factor in this setting and warrants specific attention during clinical assessment. Routine evaluation should include a detailed personal and family history to ask about specific LS associated cancers to clinically identify at-risk patients and their family members.

LS is a familial, autosomal-dominant multi-organ cancer syndrome estimated to affect roughly 1 in 280 individuals in the U.S.²⁶ It is widely and strongly recommended (e.g., ASCO, National Comprehensive Cancer Network [NCCN], Centers for Disease Control and Prevention [CDC]) that patients with LS undergo routine screening due to increased life-long risk for developing associated malignancies, often occurring before 50 years of age, though not exclusively.²⁷ The most commonly encountered are colorectal (20-80%), urothelial (1-18%), and gastric cancers (1-13%) in both men and women; and endometrial (15-60%) and ovarian cancer (1-38%) in women. Practice guidelines by several organizations (e.g., US Multi-Society Task Force on Colorectal Cancer [MSTF], ASCO, European Society of Medical Oncology [ESMO], NCCN, American College of Gastroenterology [ACG], American College of Obstetrics and Gynecology [ACOG]) recommend routine clinical screening including the use of standardized genomic questionnaires for all patients with related gastrointestinal (colon, gastric) and gynecologic (endometrial, ovarian) cancers.

Competing LS-associated cancers or related suspicious findings can pose potential clinical challenges requiring involvement and coordination of multi-disciplinary care. In UTUC specifically, LS may increase the possibility of contralateral upper tract involvement, which is an important potential clinical consideration when developing a treatment plan. The Panel notes that developing data on systemic therapies focused on targeting LS-associated cancers are expected to impact future therapeutic options for these patients, further reinforcing the overlap between genetic risk factors and need for alignment with clinical guidelines for preventative and therapeutic management of UTUC in patients with LS.

For UTUC patients with familial risk factors, clinical suspicion, or interest in further testing for hereditary syndromes, clinicians can perform initial screening tests (described below), and should offer referral for genetic counseling and, if indicated, genetic testing.

Clinical screening criteria including standard Amsterdam II criteria and Bethesda guidelines (Table 4) are useful in providing background context yet are unreliable, difficult to implement, and fail to identify a significant proportion of patients with LS or sufficiently exclude patients from screening.²⁸ Routine tissue testing provides a more sensitive, first-line means to identify LS-associated features in tumor samples, thus providing clinically significant information for patient counseling and management as well as screening for family members. IHC testing for example, which is widely available, can preliminarily identify the altered proteins associated with LS, and thus help to identify patients who may have the syndrome, who then require confirmation with further genetic (germline) testing.

LS results from an inherited germline mutation in a group of DNA damage response genes responsible for biologic mechanisms of mismatch repair (MMR), specifically MLH1, MSH2, MSH6, PMS2, or EPCAM.²⁸ Alterations affecting the normal function of these genes results in an accumulation of DNA errors and increases the potential for cancer development. Tumor tissue testing by histologic studies such as IHC can indicate loss of these specific MMR proteins or evaluate for MSI status as a standard means to assess for the possibility of LS-association. Suspicious findings with these tests require further confirmatory testing, for which patients should be referred to a specialist for genetic counseling.

Recommendations and guidelines in other LS-related cancers strongly endorse universal MMR and MSI testing. A detailed analysis by the Evaluation of Genomic Applications in Practice and Prevention (EGAPP) Working Group from the CDC reported sufficiently strong evidence to recommend genetic testing routinely be offered to all patients with newly diagnosed colorectal cancers (CRC) due to the high rate (3%) of LS and the indication of significant, cost-effective clinical benefits for patients and family members.²⁸ Of the strategies investigated and endorsed, reflex IHC studies for MMR and MSI testing were highlighted for their high sensitivity and specificity. The  Panel acknowledges the EGAPP report did not evaluate or address testing in UTUC but similarly endorses an analogous strategy in that the Panel recommends genetic testing to all patients with UTUC due to the higher identified prevalence of LS association in UTUC relative to CRC.

The NCCN Guideline Genetic/Familial High-Risk Assessment: Colorectal version 1.2022 endorses universal MMR testing of all CRC and endometrial cancers and recommends considering universal testing of other LS-related malignancies regardless of the age of diagnosis including urothelial cancers. These recommendations are founded on findings from a large data set of over 15,000 LS-related cancers (including 551 urothelial cancers) screened with MSI, MMR and germline genomic testing which identified LS-association in over 16% of cases of patients with MSI-high signatures and specifically 37.5% in urothelial cancer cases.²⁷

From the guideline systematic review, four retrospective cohort studies were identified that evaluated factors associated with LS in patients diagnosed with UTUC.²⁹ Although the patient numbers in these series were modest, limiting the analyses, the use of MMR and MSI testing was significantly associated with identification of LS patients within studied cohorts when used adjunctively with standard clinical screening criteria. One study evaluated a universal molecular screening strategy against a genetic testing standard. Patients who screened positive by standard clinical criteria (Amsterdam II), had loss of one or more MMR proteins, or had high MSI, were considered to have “potential LS,” and were referred for germline testing. Of 115 patients screened, 13.9% had potential LS: 7.0% met Amsterdam II criteria; 11.3% had loss of at least one MMR protein; and 5.7% had high MSI. Of the 16 patients with potential LS, 9 completed germline testing, with LS confirmed in 6 patients (5.2% of the total 115 patients screened). These data are comparable with described results from large cohort screening studies in LS malignancies.²⁷

It must be acknowledged that MMR or MSI studies are screening tests and are neither genetic tests nor a gold standard for identifying LS. As such, these tests may miss 10% or more with the disease.³⁰ Germline genetic testing, also considered a molecular study, is a more definitive means of diagnosis requiring specific counseling and justified if available – yet may also fail to identify familial cases of Lynch-like syndromes caused by epigenetic phenomena. Universal molecular testing serves a key role along with clinical awareness when evaluating UTUC patients providing the opportunity for discussion of genetic risk factors with patients and sufficient indication for appropriate genetic counseling referral for any patient with UTUC. The Panel notes that identifying the presence of LS-associated and MSI-high cancers also has clinical implications related to therapeutic treatment options, including identified sensitivity of urothelial cancers with mutations in DNA damage repair genes to systemic agents such as immune checkpoint inhibitors and cis-platinum-based chemotherapy.^{31, 32}

Table 4: Clinical screening criteria for LS (also referred to as hereditary non-polyposis colorectal cancer [HNPCC])

Adapted from Revised Bethesda Guidelines for Hereditary Nonpolyposis Colorectal Cancer (Lynch Syndrome) and Microsatellite Instability and New clinical criteria for hereditary nonpolyposis colorectal cancer (HNPCC, Lynch syndrome) proposed by the International Collaborative group on HNPCC.^{33, 34}

Tumor features identifiable by endoscopic and radiologic assessment are strongly associated with disease risk and, therefore, necessary to properly inform risk stratification, treatment decision-making and assessment of treatment response.^35-41 Standard reporting of endoscopic findings is, therefore, critical to document and communicate objective clinical findings. At the time of examination by antegrade or retrograde approach, clinicians should document key features including the following:

• Sites of involvement (ureteral segment, renal pelvis, calyceal sites and lower tract)
• Number of tumors or presence of multifocality
• Tumor appearance (sessile, papillary, flat/villous)

It is also recommended that documentation include an estimate of the largest tumor size, if possible, by using a reference standard such as the scope tip, basket, laser fiber, biopsy forceps, or brush. Quality of visualization can impact the accuracy of endoscopic inspection (e.g., bleeding, difficulty in access, tumor location, artifacts from instrumentation) and should be documented in endoscopic reports.

Radiographic characterization of tumor features is also informative for clinical staging. As noted, retrograde urography should be performed concomitantly with upper tract endoscopic assessment and documentation of filling defects or evidence of urinary tract obstruction should be provided. Reporting from contrast-enhanced cross-sectional imaging should include details of tumor characteristics that suggest invasive features, obstruction of the urinary tract, and locoregional progression such as suspicious lymphadenopathy, and/or presence of metastatic disease.

Determining cancer-associated risk is critical to guide risk-adapted treatment selection and patient counseling. Tumor characteristics determined from the standardized process of clinical assessment described in this guideline allows categorization of tumors into high- and LR groups. The association of HG cancer (HG biopsy or cytology) with disease progression risk and pathologic stage T2 or greater disease defines the category of HR whereas LG cancer (LG biopsy and normal cytology) defines LR disease.

It is recognized that heterogeneity within these two categories exists, warranting further stratification by distinct clinically identifiable features. The factors below highlight these additional findings to aid sub-stratification within the HR and LR categories and to guide risk-adapted management strategies. An accounting of the data supporting these additional features is also included in Appendix tables II and III.

BIOPSY

The association of HG tumor on ureteroscopic biopsy with high-stage (HS) disease (≥pT2) on final pathology (13 studies, N=1,197) has a PPV of 60% (95% CI: 54% to 66%; range: 33% to 85%; I² = 57.0%) and a pooled NPV of 77% (95% CI: 73% to 82%; range: 67% to 100%; I² = 19.8%), indicating room for further refinement along the spectrum of favorable to unfavorable within the HR and LR groups.⁴² Sub-stratification features have been identified and reported in publications and nomogram formats and are recognized by the guideline Panel as being useful for further risk refinement.  

CYTOLOGY

Selective ipsilateral upper tract cytology provides supplemental histologic data to tumor biopsies and the finding of HG cytology in the setting of LG biopsy findings indicates the likely presence of higher-risk features (e.g., HG tumor) missed on biopsy sampling. Obtaining selective cytology after tumor biopsy can improve the yield of cells for cytologic analysis.⁴³

FISH

There are limited data on the independent value of FISH testing to identify advanced stage disease, and its routine use for this purpose cannot be supported now. The association of FISH with the presence of a HG tumor is recognized and may have value as an adjunct test in some scenarios where tissue sampling is challenging and cytology is indeterminate. Two studies (N=244) reported on the diagnostic accuracy of FISH for identifying HS disease in confirmed UTUC.^{44, 45} The studies used the reference standard of histologic confirmation on final surgical pathology report from NU or distal ureterectomy and defined HS as ≥pT2. Sensitivity was from 72% to 83% and specificity was 38% to 47%, for PPV of 58% to 60% and NPV of 63% to 67%. One of the studies found that FISH results were not significantly associated with increased likelihood of HS disease (p=0.12).⁴⁵

IMAGING

CT

The sensitivity and specificity of specific CT findings for identifying HG disease varies (Appendix II); the CT finding with the best combination of sensitivity and specificity was presence of heterogeneous texture (versus homogeneous; sensitivity: 70% and specificity: 100%) with a PPV of 100% and a NPV of 28%.^{46, 47} However, these predictive values should be interpreted with caution due to a low proportion of patients with LG UTUC in the study cohort. The presence of ipsilateral hydroureteronephrosis demonstrates limited diagnostic accuracy for the findings of HG UTUC, with a sensitivity of 40% to 43% and specificity of 60% to 66%, which was further supported by a study by Ng et al. in which hydronephrosis on CT was not significantly associated with HG UTUC (p=0.49).⁴⁷

Similarly, the sensitivity and specificity of specific CT findings for identifying HS UTUC, which is defined as >pT2, is variable. Five studies reported on the diagnostic accuracy of CT for identifying HS disease in confirmed UTUC.^46-50 Heterogenous texture on enhanced and even unenhanced CT imaging has been associated with invasive disease.⁵¹ In a study of 48 patients with UTUC, the presence of heterogeneous (versus homogenous) texture was associated with the best combination of sensitivity (91%) and specificity (58%) for advanced stage, with a PPV of 66% and NPV of 88%. In this series, identification of hydronephrosis on CT was associated with a 4-fold increase in the risk of HS UTUC (hazard ration [HR]: 4.0; 95% CI: 1.4 to 11.5, p=0.01). The sensitivity of multidetector CT identifying HS disease (≥pT3) ranges from 0.28 – 0.75, while the specificity ranges from 0.84-1.00.^{46, 48, 50} Specific features that have been proposed to predict HS UTUC include the presence of local invasion on CT and the presence of pathologically enlarged lymph nodes, both of which are associated with a relatively modest sensitivity of 0.49 and 0.22, respectively, with a higher specificity (0.85, and 0.98, respectively).⁴⁹

OTHER IMAGING

Retrograde pyelograms provide a roadmap for evaluation and possibly planning kidney-preserving strategies and should be considered at initial evaluation with images retained in the patient record. Modalities such as endoluminal US do not yet have a well-defined clinical role and warrant either prospective evaluation or, at a minimum, further testing under controlled clinical circumstances such as quality improvement studies. MRI can provide some soft tissue details in patients who cannot receive contrast, offering some advantages in such patients by identifying features of fat invasion with diffusion weighted imaging associated with very advanced, T3 disease.⁵² However, MRI can falsely over-estimate tumor stage due to surrounding tissue effects that may mimic tumor invasion such that establishing cutoffs for diffusion weighted imaging have not been well established.⁵³ At present, the Panel recommends such studies only as supplements to current standards of care.

TUMOR APPEARANCE, GROWTH CHARACTERISTICS AND LOWER TRACT INVOLVEMENT

Tumor features associated with more aggressive cancer risk include sessile versus papillary appearance, multi-focality, and, relatedly, pan-urothelial disease as indicated by history of prior cystectomy, concomitant or metachronous lower tract urothelial cancer or contralateral UTUC diagnosis.^{54, 55} Each of these features are considered unfavorable if present. Recognizing no single finding can reliably provide adequate staging or risk-stratification in isolation, several groups have tried to combine multiple variables to strengthen predictive ability. The Panel acknowledges the debate on issues including overlap in several of these features and aspects of additive risk when more than one feature is present, describing a spectrum of risk, which can be weighted and considered during shared decision-making and treatment election.

OTHER FEATURES

The Panel recognizes that other features have been identified that may have an association with disease risk (e.g.., tumor size, lymphovascular invasion (LVI) neutrophil-to-lymphocyte ratio, tp53 or FGFR3 mutation status). However, these variables may not be widely available, easily identified or measured, thereby limiting broad applicability. Specifically, tumor size has a described association with tumor grade and stage; however, measurement in the pre-surgical setting is not standardized and has not been shown to be independent of other more easily determined clinically identified features such as multifocality, invasion and obstruction. Importantly, data supporting tumor measurements from large retrospective databases have been derived from pathology reports after surgical resection and may not be applicable to pre-operative imaging or endoscopic tumor size estimates – therefore less clinically useful.⁵⁶

Table 5: Presurgical Clinical Risk Categories

Initial decisions regarding operative approach and administration of systemic therapy are based on patients’ baseline renal function and their estimated post-operative estimated glomerular filtration rate (eGFR). Patients undergoing NU have diminished postoperative renal function due to loss of a renal unit. A median decline in renal function after NU up to 32% has been reported which can induce or exacerbate a state of CKD, affect a patient’s candidacy to receive adjuvant chemotherapy ^{57, 58}

Patients with UTUC should, therefore, undergo an assessment of renal function and, for individuals who are scheduled to undergo NU and especially those who may require perioperative systemic treatment, an estimation of post-operative renal function should be made. Recommended tests include serum creatinine to calculate an eGFR and, at the clinician’s discretion for more refined evaluation, split function testing such as with differential renal scan or CT volumetric studies. Perioperative nephrology consultation can be considered as well, particularly in patients with pre-existing kidney disease. Attention should be paid in the settings of renal atrophy and hydronephrosis, which may alter clinical estimates of resulting post-operative renal function. Hydronephrosis caused by tumor obstruction may falsely under-estimate preoperative renal function and alter decision-making around the use of neoadjuvant chemotherapy (NAC). Thus, in settings of hydronephrosis, renal decompression either by indwelling ureteric stent or a percutaneous nephrostomy tube placed in an uninvolved renal calyx along with oral fluid hydration for 7-14 days before re-checking eGFR will help to establish a more accurate estimation of baseline renal function. Ureteric stenting is the preferred method of drainage given the known risk of tract seeding with percutaneous nephrostomy tubes in the setting of UTUC as well as quality of life considerations⁵⁹ Atrophy of the contralateral (unaffected) renal unit may lead to over-estimates of postoperative renal function in the setting of NU since the kidney with lower differential function will remain in situ^{60, 61} Results of renal function investigations can help with patient counseling, strategizing treatment sequence, and determination of downstream risks of CKD and potential dialysis. In patients with sufficiently poor CKD in which NU could precipitate ESRD, a post operative plan for dialysis in conjunction with nephrology colleagues should be in place preoperatively including plans for dialysis access. Referral to nephrology for detailed evaluation and recommendations for perioperative management is warranted in such cases.

In patients with pre-existing CKD or a solitary kidney, attempts to preserve renal function can be made, if oncologically feasible and appropriate, with segmental or endoscopic organ-sparing approaches which preferentially are associated with improved postoperative renal function.^62-64 Notably, in four studies reporting the impact of SU compared to NU on renal function, three found SU was associated with improved renal function versus NU with differences in eGFR ranging from 14 to 19 mL/m.^65-67

As stated in the Renal Mass and Localized Renal Cancer: Evaluation, Management, and Follow Up: AUA Guideline,^{68, 69} predictive factors for post-operative development of CKD or progression of pre-existing CKD include older age, diabetes mellitus, hypertension, as well as male sex, obesity, tobacco use, larger tumor size, and post-operative acute kidney injury.^70-76 Patients who present with eGFR less than 45 mL/ min/1.73m² or confirmed proteinuria are at particularly HR from a functional standpoint and should be considered for nephrology consultation. Patients who are expected to have an eGFR less than 30 mL/ min/1.73m² after intervention will also be at HR long-term, and a nephrologist should be involved in their care. Identifying modifiable risk factors including diabetes mellitus (DM), hypertension (HTN) and smoking is essential. Optimizing glycemic and blood pressure control, smoking cessation and minimizing risk of acute kidney injury (with avoidance of hypotension and nephrotoxic agents such as intravenous contrast or non-steroidal anti-inflammatory drugs) should reduce the degree of renal dysfunction in the perioperative period.⁷⁷ Of note, patients with DM are at even higher risk for acute kidney injury compared with those without DM, even among those with normal eGFR prior to nephrectomy.⁷⁰ With significant nephron mass loss, hyperfiltration can occur resulting in glomerular damage, exacerbation of proteinuria and progressive sclerosis with further decline in GFR. Therefore, repeat assessment of blood pressure, eGFR, and proteinuria should be performed soon after nephrectomy then again in three to six months to assess for development or progression of CKD. With any compromise in eGFR or presence of CKD complications, additional regular monitoring of kidney function should be performed and further management of CKD would be recommended with referral to nephrology. Careful management of DM and HTN and avoidance of substantial weight gain may slow or prevent CKD progression and should be prioritized on a long-term basis.

Providing patients with a detailed description of risks and benefits associated with treatment options is a mandatory requirement of care, and clinicians must be familiar with outcomes in upper tract management. Urothelial recurrences are common in the management of UTUC, regardless of approach, and mandate long-term surveillance for which patients must be prepared – including the potential need for additional treatments. Ablative options can provide local control including durable long-term kidney sparing outcomes but incur additional endoscopic surveillance requirements and associated risks such as stricture and infection.⁷⁸ Specifically, the use of chemoablative treatment with the reverse thermo-hydrogel preparation of mitomycin for pyelocaliceal instillation for LG tumors carries an FDA label warning for ureteral obstruction (>44%), bone marrow suppression, and embryo-fetal toxicity.⁷⁹ Systemic chemotherapy and immunotherapy treatments also have toxicities requiring specific counseling best provided as part of multi-disciplinary care.

LR UTUC is associated with low rates of metastatic progression and presents an opportunity for kidney preservation. Endoscopic management (by retrograde ureteroscopy, antegrade ureteroscopy, or percutaneous resection) is an established treatment option for urothelial cancer, including those involving the upper tract, and should be the first-line treatment for patients with LR favorable UTUC when technically feasible. Developing RCTs is a challenge in this setting since renal functional loss from NU represents a significant medical risk, leaving a small number of patients at individual centers with normal renal function who might be considered trial-eligible. There are 13 observational studies that have compared endoscopic management with NU showing similar cancer-specific survival (CSS) and improved renal functional outcomes for patients treated with endoscopic ablation.^{62-64, 80-89} As these are retrospective studies, characteristics of patients selected for endoscopic management versus those managed with NU are varied and results must be interpreted in the context of strong case-selection bias. A study by Grasso et al. (n=162) with a mean follow-up just over 3 years reported improved 5-year CSS for patients with LG UTUC who underwent ureteroscopic management versus those with any grade UTUC who underwent NU in which 10-year DSS were similar (5-year: 87% versus 64%; 10-year: 81% versus 78%).⁸² Further, studies by Rouprêt et al. (n=97) and Shenhar et al. (n=61) reported similar 5-year CSS for patients who underwent endoscopic management and NU (80% to 81% versus 84% and 89% versus 92%, p=0.96).^{64, 85}

Regarding renal functional outcomes, two studies ^{81, 87} reported similar renal function following endoscopic management or NU, and three studies reported better renal outcomes following endoscopic management.^62-64 Four studies reported similar rates of surgical complications or reported no statistically significant differences.^{64, 81, 84, 85}

In certain clinical scenarios of LR UTUC, complete endoscopic ablation may not be feasible. Such instances may be predicated on specific tumor (location and focality) and patient factors (age, comorbidities, baseline renal function, procedural risk). Chemoablation (in-situ tissue destruction) can be a treatment alternative in these situations. In an open label, single arm, phase III trial in patients with LG tumors measuring between 5 – 15 mm, Kleinmann et al. reported that a mitomycin containing reverse thermal gel yielded a 59% (42 of 71 renal units) complete response at primary disease evaluation, 1 month following a 6-week course of therapy.⁷⁹ A subsequent report from these investigators highlighted that 56% of evaluable complete responders remained disease free at 12 months post-therapy.⁷⁸ The observed benefit of mitomycin containing reverse thermal gel in these studies must be balanced against the risk of possible ureteral stricture. Importantly, chemoablation should not be used as a substitute for complete endoscopic ablation whenever feasible.

There is no high-quality evidence that specifically compares outcomes of endoscopic management versus NU for patients who meet specific criteria for LR unfavorable or HR favorable UTUC. Collectively, comparable cancer-specific survival and improved renal functional outcomes are reported for patients undergoing endoscopic management relative to NU (see discussion in the Guideline statement 13).^{62-64, 80-89} Some studies have included in their analyses patients with features of unfavorable LR disease (e.g., multifocal LG tumors).^{81, 82, 90} Further, Grasso et al. included patients with pan-urothelial disease and larger tumors but did note a higher rate of disease progression in these patients.⁸²

Tumors < 1.5 cm in size may be optimal for endoscopic ablation given a lower risk of invasive disease. Conversely, tumors ≥ 1.5 cm in size are associated with a > 80% risk of invasive disease, and tumors > 2.5 cm are associated with a lower disease-specific survival.⁹¹ Further, hazard ratio analyses of tumor size cutoffs of 1.5 and 2 cm have demonstrated significant relationships with stage ≥ pT2.⁹²

Endoscopic ablation has been reported in patients with imperative indications with tumors up to 6.0 cm. Scotland et al. described their institutional experience treating patients with tumors ≥ 2.0 cm in size and found 5-year recurrence-free survival (RFS), progression-free survival (PFS), and CSS rates of 10%, 65%, and 84%.⁹³ Therefore, larger tumors (≥ 1.5 cm) may be considered for ablation based on the provider’s experience and assessment of the need for kidney sparing surgery.

For patients with LR unfavorable disease who demonstrate progression in tumor size, focality, or grade, the Panel recommends against further endoscopic-assisted attempts and consideration of definitive resection via segmental ureterectomy (SU) or NU. In cases of HR favorable cancers managed endoscopically, clinicians must recognize the higher risks of disease progression and pivot early to definitive surgical resection when necessary.

Various approaches and techniques may be employed to successfully treat UTUC by ablation. Retrograde approaches including ureteroscopy with pyeloscopy is commonplace, while percutaneous techniques including antegrade pyeloscopy or ureteroscopy with ablation is typically reserved for larger tumors, those that are difficult to access in a retrograde fashion, or in patients who have undergone prior radical cystectomy or urinary diversion.

The energy source employed for ablation may vary based on availability of instrumentation and tumor characteristics. Thulium laser, holmium laser, Neodymium (Nd:YAG), and electrocautery devices (e.g., Bugbee) may all be deployed through an endoscope. Additionally, chemoablation may be employed either through retrograde ureteral catheter instillation or percutaneous access with fluoroscopic imaging guidance.

Optional use of a ureteral access sheath during the time of ureteroscopic ablation can provide some advantages for endoscopic assisted ablation when safely employed – allowing for repeated scope passage up and down the ureter for sampling and a means of fluid egress from the upper tract to avoid excess pelvicalyceal hydrostatic pressure from irrigation solutions. A study by Douglawi et al. demonstrated a lower rate of intravesical recurrence in patients who underwent ureteroscopy with an access sheath compared to those without a sheath prior to NU.⁹⁴ Prior to placement of any ureteral access sheath, the entire ureter should be directly visualized in order to avoid missing any luminal neoplasms, especially in the distal ureter.

Repeat endoscopic evaluation should take place within three months of the initial treatment due to the proclivity of UTUC to recur and for residual disease to remain after the first ablation. Optimal timing of follow-up endoscopic evaluation has not been well established noting that several factors may impact the indication and decision for short interval follow-up such as aspects of visualization. A study of 41 patients who underwent a second look ureteroscopy within 60 days of ablation showed a 51.2% cancer detection rate at the time of the second look.⁹⁵ A 30-day window on either side of this endpoint (i.e., 30 to 90 days) is justified to allow timely identification of recurrences and may be dictated by aspects such as tumor size, visualization, access, treatment efficacy, etc., as clinically indicated. Clinicians may wish to take a conservative approach with shorter interval endoscopic diagnostic and therapeutic endoscopic procedures for more challenging cases, particularly when incomplete treatment is a possibility. Repeat endoscopic assessment should occur within three-month intervals until no evidence of upper tract disease is identified.

There is ample evidence supporting the use of an immediate instillation of intravesical chemotherapy at the time of transurethral resection of a bladder tumor for urothelial carcinoma for the purpose of reducing the rate of intravesical tumor recurrence.^{96, 97} The principle of an immediate instillation of intravesical or pyelocaliceal (upper tract) chemotherapy at the time of endoscopic tumor ablation for UTUC is undertaken by extrapolation of the data supporting this practice in the management of urothelial carcinoma of the lower tract. At present, this is considered an optional part of routine practice. The available reported clinical experience reported in the upper tract is less compelling. A small, prospective, non-randomized single center cohort study by Gallioli et al., showed a strong trend in improving urothelial recurrence free survival (URFS) for patients treated with a single upper tract instillation of Mitomycin C after endoscopic ablation. Mean URFS was 29 months for the treated group compared to 19 months in patients who did not receive treatment (log-rank p = 0.067).⁹⁸ Though a small study including only 51 patients, there were controls for several potential confounding variables and low ROB was identified. A larger study (n=73) by Cutress et al. did not control for confounding variables and failed to identify a difference in RFS with adjuvant intraluminal chemotherapy.⁹⁹ In the Gallioli study, the majority of recurrences were observed in the bladder.⁹⁸ More recent work has explored the role of an adjuvant dose of upper tract mitomycin gel following endoscopic ablation with a report of 63% ipsilateral disease-free rate at 6.8 months following instillation, albeit with a 19% ureteral stenosis rate and no comparator group.¹⁰⁰ While acceptable, there are limited direct supporting data for this common practice in upper tract applications at this time.

Technical considerations

The optimal administration technique is not fully elucidated. Both ex vivo and in vivo porcine models suggest higher rates of topical therapy delivery to the pyelocaliceal system with retrograde administration by ureteral catheter.^101-104 However, the Panel considers each of the following delivery approaches to be acceptable: 1) antegrade perfusion by nephrostomy tube, 2) retrograde perfusion via ureteral catheter, and 3) bladder instillation by transurethral catheter with reflux via a double J ureteral stent. In the third scenario, it is recommended to perform a cystogram and demonstrate adequate reflux of contrast into the pyelocaliceal system. Finally, while bacillus Calmette-Guerin (BCG) is the mainstay of topical therapies for UTUC, the following agents have been described: mitomycin c, gemcitabine, docetaxel, epirubicin, adriamycin, thiotepa, and BCG with interfero^{105, 106}

Topical therapy may consist of a six-week induction course of BCG. Patients should be considered for topical therapy if imperative indications are present, including solitary kidney status, bilateral UTUC, or risk of progression to end-stage renal disease.

There is a dearth of literature specifically regarding the treatment of favorable HR UTUC with topical therapy (e.g., BCG). No randomized trials exist to compare outcomes of patients treated with topical therapy versus NU, and retrospective comparisons are limited by small cohorts.¹⁰¹ However, several small observational studies have evaluated the role of BCG as the primary treatment of upper tract CIS (UTCIS) and as an adjuvant treatment for Ta/T1 disease. These studies have been summarized in systematic reviews.^{102, 105} Regarding the treatment of CIS, generalizations of the literature are limited by the lack of a standard definition of UTCIS, variable methods of UTCIS detection (voided cytology, selective, cytology, ureteroscopic visualization, biopsy), and inconsistent measurements of successful treatment. These systematic reviews report rates of complete response ranging from 41% to 100%. Rates of recurrence, progression, and transition to radical NU likewise vary from 10% to 46%, 0% to 45%, and 45% to 100%, respectively. Further, AEs are reported in 0% to 92% of patients across studies and include cystitis, fever, sepsis, renal tuberculosis, ureteral stricture, and pericarditis.^{102, 105}

Regarding the treatment of UTUC Ta/T1 disease, Foerster et al. conducted a meta-analysis of 12 non-randomized observational studies including 212 patients who underwent adjuvant therapy following ablation of Ta/T1 disease.¹⁰⁶ Median follow-up was 31 months, during which time 39% of patients developed a recurrence. Nine studies (with ≥ 10 patients) were included for pooled survival estimates. CSS and overall survival (OS) were 94% (95% CI: 86 to 99%), and 71% (95% CI: 47 to 90%), respectively. One study included in the analysis included 22 renal units with Ta/T1 disease and reported progression in 41% of patients.¹⁰⁷ Although most studies have used BCG as the therapeutic agent, sub analyses showed no significant differences in outcomes when other agents were administered.

Failure of conservative strategies for kidney preservation includes the risk of cancer progression, potentially shifting from curable to an incurable form of UTUC. Clinical evidence of a change in tumor growth pattern toward a more aggressive subtype should prompt re-assessment of management strategy and consideration for more definitive treatment with extirpative surgical resection. This is especially true for LG cancers, which should not display evidence of aggressive biology including invasion, multifocal implantation, HG cytology, or change from non-obstructing to obstructing tumors. Such features should prompt a detailed discussion with patients about the observed findings, their clinical significance suggesting a shift in disease risk and consideration for change in strategy developed through shared decision-making. Data on outcomes comparing endoscopic management to extirpative surgery in different risk groups are limited but provide support for this transparent approach to counseling and managing patient expectations.

Thirteen retrospective studies compared endoscopic management versus RNU with baseline differences between treatment cohorts noted.^{62-64, 80-89} Eight studies compared RFS.^{62-64, 80, 81, 83-86, 89} Of three groups that reported adjusted risk estimates, one (n=120) found endoscopic management was associated with an increased risk of any (local, intravesical, or distant) recurrence (adjusted HR: 3.56; 95% CI: 1.73 to 7.35),⁸⁴ one study found endoscopic management associated with improved intravesical RFS (adjusted HR: 0.56; 95% CI: 0.25 to 1.25),⁸¹ and one found endoscopic management associated with increased risk of local recurrence (adjusted HR: 1.27; p=0.001) but no difference in risk of intravesical RFS (adjusted HR: 0.90; p=0.52).⁸⁶ Nine studies reported outcomes on CSS or all-cause mortality (ACM).^{62, 80-87} Three (n=453, 356, and 170) that controlled for confounding factors found endoscopic management was associated with worse CSS (propensity-matched HR: 2.1; 95% CI: 1.0 to 4.1; adjusted HR: 1.18; p=0.12; adjusted HR: 2.00; 95% CI: 0.33 to 12.50).^{81, 86, 89} Valid concerns for aspects such as accuracy of clinical staging, risks of undiagnosed HG cancers and disease-specific mortality associated with developing HR disease warrants vigilance in follow-up and recognizing clinical signs indicating thresholds for recommending altering care.

Some patients with UTUC have significant comorbid medical conditions that impose serious risks of severe, treatment-related complications from any form of intervention. Complication rates following RNU range from 15% to 50% including a 30-day mortality risk of 1%.¹⁰⁸ Such results do not reflect outcomes in non-operative cases where observation and palliative approaches are utilized. Discussion of treatment related risks including perioperative mortality may lead to a shared decision to proceed with active surveillance (whereby periodic assessments such as imaging or limited endoscopic assessment are performed) or watchful waiting/expectant management, where interventions are limited to palliation or awaiting symptomatic progression – especially in those with very limited life expectancy. In such cases, patients and family should be counseled and prepared for disease-related events such as bleeding, obstruction, infection, and pain with options for palliation that may be limited.

Two studies utilizing large databases compared non-surgical management versus surgery for UTUC. ^{109, 110} Both studies found non-surgical management was associated with worse OS versus surgical treatment though likely reflecting the compromised medical condition of these patients. Outcomes reported from the SEER database (n=8,304; 633 of whom did not undergo surgery) also observed that non-surgical management was associated with worse OS (median 1.9 versus 7.8 years; p<0.001) and 3-year CSS (74% versus 92%; p<0.001).¹¹⁰ Another study utilized the National Cancer Database (n=28,910; 3,157 of whom did not undergo surgery) and similarly found non-surgical management to be associated with worse OS (median 2.0 versus 5.6 years; p<0.0001).¹⁰⁹

RNU with complete bladder cuff excision (BCE) and lymphadenectomy is the standard of care for patients with HR UTUC. Principles of RNU include complete excision of ipsilateral upper tract urothelium, including the intramural portion of the ureter and ureteral orifice with negative margins, and avoidance of urinary spillage, such as by early low ligation of the ureter, to minimize the risk of seeding urothelial cancer outside the urinary tract.

The RNU specimen should be removed en bloc whenever technically feasible. Open, robotic, and laparoscopic approaches are suitable for RNU so long as the above oncologic and surgical principles are adhered to. The systematic literature review supporting these guidelines demonstrated equivalent oncologic outcomes for open and minimally invasive (laparoscopic, hand-assisted laparoscopic, robot-assisted laparoscopic) approaches to RNU.¹¹¹ Minimally invasive approaches were associated with favorable perioperative outcomes including shorter length of stay and fewer complications, and, therefore, are favored for most patients when principles of RNU can be maintained. Case selection criteria are difficult to assess in these analyses such that outcomes across the range of tumor staging could be a concern and used as rationale for preferentially offering open surgical approaches for large, bulky UTUC with clinical evidence for direct invasion to adjacent structures.¹¹²

Numerous studies demonstrate worse local and metastatic recurrence rates with associated decreased CSS and OS for patients who did not receive complete BCE.¹¹¹ BCE can be completed either extravesically or transvesically through a variety of approaches including open, minimally invasive or transurethral endoscopic techniques. Transurethral endoscopic approaches are associated with higher recurrence rates in the bladder and may limit the ability to utilize post-NU intravesical therapies if the bladder is not fully closed.¹¹³

Ureterectomy including SU with ureteroureterostomy and distal ureterectomy with ureteral reimplant are reasonable alternatives to RNU for well-selected patients. The literature review demonstrates equivalent oncologic outcomes for patients undergoing RNU and ureterectomy recognizing the inherent selection differences in the comparative cohorts.¹¹¹ The most favorable candidates for distal ureterectomy are patients who have ureteral tumors in the lower third of the ureter and a sufficiently mobile bladder with capacity to facilitate reimplantation with or without reconfiguration of the bladder to facilitate a tension-free anastomosis (i.e., Boari flap or psoas hitch maneuver). Patients most suitable for SU have small, unifocal tumors (typically 1 cm or smaller) tumors isolated to a short segment of the proximal or mid-ureter requiring resection of 2 cm or less of ureteral length to allow for primary ureteroureterostomy. Longer sections of ureteral involvement and resection may require more complex reconstruction techniques when kidney sparing is desired. Principles of ureterectomy in select cases include:

• Patient counseling to describe techniques, potential requirements for urinary reconstruction and associated complications including the potential impact on postoperative bladder function.
• Preoperative endoscopic assessment to evaluate sites of involvement and proximal extent of disease.
• Preoperative assessment of bladder capacity and function in cases where more extensive reconstruction such as a Boari flap are anticipated to permit a tension free ureterovesical anastomosis or the use of bowel segments.
• Intraoperative pathologic assessment (i.e., frozen sections) of proximal and distal margins to ensure complete resection with negative margins.
• Reasonable attempts to avoid of spillage of urine into the surgical field.
• Watertight, tension free closure to facilitate functional healing and avoid urine leak (of urine potentially contaminated with malignant cells).

Distal ureterectomy and reimplantation offers definitive curative management for tumors confined to the lower ureter while preserving kidney function. It is, therefore, the treatment of choice for patients with localized cancers in this location with an increased risk of disease recurrence and progression. Other approaches such as endoscopic assisted tumor ablation are considered alternative options to the gold-standard of extirpative resection and carry risk for upper tract tumor recurrence, with reported rates of 23% to 76%.⁹⁰ As such, these approaches may yield less optimal results and require multiple additional procedures. Of note, CIS limited to the region within the ureteral orifice. Topical therapies such as BCG along with refluxing ureteral stenting that has been used for in cases of CIS near the ureterovesical junction or transurethral resection of the transmural portion of the ureter for very distal tumors, as an extension of bladder resection procedures, when tumor is limited to the region inside the ureteral orifice and not beyond the bladder wall, thus anatomically managed as bladder cancer.

The management of the ureteral orifice during distal ureterectomy or RNU has been variably described. Traditionally, this aspect of UTUC surgery has been approached as a formal excision of the bladder cuff surrounding the ureteral orifice and entire ureteral tunnel in continuity with the ureter either via a transvesical (e.g., midline cystotomy) or extravesical approach. Depending on surgeon preference and expertise, this aspect of surgery can be approached via minimally invasive (e.g., laparoscopic, robotic-assisted laparoscopic) or open approaches. Others have advocated for a combined endoscopic deep incision surrounding the ureteral orifice or transurethral resection of the ureteral orifice with extravesical traction to complete the excision (the “pluck” technique). The resultant hiatus in the bladder in the location of the excised ureteral orifice with or without the bladder cuff can be closed formally in a watertight fashion in one or more layers; however, delayed closure by secondary intension in a decompressed bladder without formal bladder closure has also been described.

To date, no RCT has compared the different surgical techniques for managing the distal ureter and ureteral orifice during NU or distal ureterectomy for UTUC. This has been studied in retrospective observational studies ^114-120 with sample sizes ranging from 84 to 4,266 (total N=12,125), with low ROB in one study,¹¹⁴ moderate ROB in five studies,^{115-117, 119, 120} and high ROB in one study.¹¹⁸ Two retrospective studies^{114, 118} (N=420) demonstrated that formal BCE is associated with improved 5-year OS versus no BCE (71.5% versus 57.0%; p=0.001).¹¹⁸ In patients undergoing SU for UTUC of the distal ureter (N=84), BCE was independently associated with improved 5-year OS (92.3% versus 73.7%; adjusted HR: 0.31; 95% CI: 0.08 to 1.18).¹¹⁴

An association between formal BCE and CSS has been evaluated in seven studies (N=11,478).^{114-116, 118-120} BCE was associated with improved CSS versus no BCE in all studies except for two,^{115, 117} though some differences were small and/or not statistically significant (Appendix IV). However, in the two largest studies (n=4,266¹²⁰ and 4,210¹¹⁹), both of which evaluated patients who underwent NU, the adjusted HRs for cancer-specific mortality (CSM) with BCE versus no BCE were 0.88 (95% CI: 0.75 to 1.03)¹²⁰ and 0.76 (95% CI: 0.66 to 0.88).¹¹⁹ Of note, this association was also observed in patients with pT3 (adjusted HR: 0.8; p=0.04), pT4 (adjusted HR: 0.69; p=0.02), and N1-3 disease (adjusted HR: 0.72; p=0.04).¹¹⁹ Conversely, two studies did not demonstrate a statistically significant association between BCE and CSM.^115,117 There are insufficient data to date documenting associations between BCE and RFS or metastasis-free survival. Limited data demonstrate no significant difference in harms associated with BCE.

There are insufficient data to recommend one surgical approach to the bladder cuff and ureteral orifice over the other. However, to avoid the risk of incomplete resection of the distal ureter and transmural tunnel, the Panel recommends that a clinician should perform a formal BCE with watertight closure of the bladder cuff to avoid urinary extravasation from the bladder, facilitate more rapid catheter removal, and permit instillation of intravesical adjuvant chemotherapy in the perioperative setting.

Two prospective RCTs have demonstrated that a single instillation of intravesical chemotherapy around the time of NU reduces the risk of subsequent intravesical recurrence of urothelial carcinoma. A phase III trial by O’Brien et al. (ODMIT-C Trial) enrolled 284 patients with no prior history of bladder cancer who were undergoing NU for suspected UTUC to either a single post-operative intravesical dose of mitomycin-C (MMC) or standard management at the time of catheter removal.¹²¹ On the intention-to-treat analysis, 17% of the MMC arm developed a bladder recurrence in the first year compared to 27% in the standard treatment arm (p=0.055). By treatment as per protocol analysis, 17 of 105 patients (16%) in the MMC arm and 31 of 115 patients (27%) in the standard treatment arm developed a recurrence (p=0.03) with no reported serious AEs. A smaller phase II trial by Ito et al. randomized 77 patients to a single intravesical instillation of pirarubicin or standard care within 48 hours of RNU for UTUC with similar results.¹²² As such, the evidence strongly supports the use of single dose of intravesical chemotherapy around the time of RNU to reduce the risk of subsequent bladder recurrence. The exact timing of therapy has varied by study with the ODMIT-C trial instilling intravesical chemotherapy at the time of catheter removal, while other retrospective series reported instillation during surgery or up to 48 hours postoperatively.^121-123

Numerous agents have been used at the time of trans urethral resection of bladder tumor (TURBT) to reduce the risk of NMIBC recurrence but in the context of UTUC there is little data to support one intravesical chemotherapeutic over another. However, for many clinicians, the recent compelling data supporting the use of a single dose of intravesical gemcitabine at the time of TURBT for NMIBC to reduce the rate of intravesical recurrences combined with concerns about potential chemical peritonitis if there is extravesical extravasation of MMC has led many to convert their practice to the use of gemcitabine rather than MMC.^{123, 124} Nevertheless, in the absence of direct comparisons, ultimately the timing of therapy and choice of agent can be modified based on the agent availability and workflow suitable to the clinician.

Limited evidence exists to support a beneficial role for LND at time of NU or ureterectomy among patients with LR UTUC. To date, no RCT has compared LND versus no LND with respect to impact on oncologic outcomes.

Two recent systematic reviews¹²⁵ of observational studies compared LND versus no LND.¹²⁶ No statistically significant differences were noted with regard to LND in subsequent oncologic outcomes, including among patients with higher stage tumors. Therefore, the benefit of node dissection among patients with LR UTUC specifically remains unclear. In one study evaluating patients with cN0M0 UTUC (n=7,278) in the SEER database,¹²⁷ there was no statistically significant association between lymphadenectomy and OS or CSS in patients with T1 and T2 tumors. Given that most patients with LR UTUC have low-stage tumors on final pathology, while some may be upgraded or upstaged, LND may be considered at time of NU or ureterectomy at the discretion of the clinician according to clinically or radiographically suspicious regional lymphadenopathy or other intraoperative findings suggesting more advanced disease for which nodal staging may be warranted.

There have been no RCTs to evaluate the effect of LND on oncologic outcomes in patients undergoing NU or SU. Two recent systematic reviews (N=7,516 and N=22,665) of observational studies compared LND with no LND.¹²⁵ Findings of the reviews were consistent, with no statistically significant differences in oncologic outcomes, including among patients with higher stage tumors. While studies have attempted to adjust for confounders, the ROB in these studies is substantial, especially as it relates to selection bias: systematic differences in patient baseline characteristics, tumor grade, and tumor stage. Moreover, these studies are unable to confirm the extent or anatomic boundaries of the LND that was performed.

The Panel conducted a re-analysis of some of the studies described in the systematic review by Chan et al.¹²⁶ in which hazard ratios were all converted so the comparison was in the same direction (LND versus no LND), to pool data from all studies. In this meta-analysis, LND was associated with better RFS (four studies, HR: 0.58; 95% CI: 0.40 to 0.83) (Figure 1).

Figure 1: Reanalysis of recurrence-free survival from Chan 2020 systematic review; all hazard ratios converted to LND versus no LND.

• Notes:
• Only included ureteric arm patients with pT2 disease or above and N0M0
• Only included renal pelvic arm patients with pT2 disease or above and N0M0
• Only included patients with muscle invasive disease and locoregional recurrence
• Only included patients with locally advanced UTUC

Two additional recent cohort studies^{127, 128} published subsequently also compared LND with no dissection. One study of patients in the National Cancer Database (n=5,905) with clinically localized (≤cT4, N0M0) UTUC found LND was significantly associated with improved OS (adjusted HR: 0.87; 95% CI: 0.78 to 0.96), but worse 90-day mortality (adjusted OR: 1.53; 95% CI 1.03 to 2.27) compared to no LND.¹²⁸ The second study evaluated patients with cN0M0 UTUC (n=7,278) in the SEER database.¹²⁷ Compared to no LND, the study found performance of a LND was associated with slightly improved OS (adjusted HR: 0.87; 95% CI 0.80 to 0.95) and CSS (adjusted HR: 0.81; 95% CI: 0.73 to 0.95). When patients were stratified according to tumor stage, LND was significantly associated with improved OS in patients with T3 (adjusted HR: 0.88; 95% CI: 0.78 to 0.99) and T4 (adjusted HR: 0.74; 95% CI: 0.59 to 0.94) tumors; while estimates indicated no benefit or were not statistically significant in patients with T1 and T2 tumors. Findings were similar for CSS, with statistically significant benefits in patients with T3 (adjusted HR: 0.83; 95% CI: 0.73 to 0.98) and T4 (adjusted HR: 0.64; 95% CI: 0.47 to 0.88) tumors and non-statistically significant differences in patients with T1 and T2 tumors.

To date, no study has adequately assessed the distribution of lymph node metastases. As such, the appropriate template to yield maximal oncologic outcomes and prognostic information remains to be determined. However, based on anatomic principles, the Panel recommends that the following minimal templates may be considered in most settings of clinically non-metastatic HR disease (cN0M0).

• Tumors in the pyelocaliceal system: lymph nodes of the ipsilateral great vessel extending from the renal hilum to at least the inferior mesenteric artery.
• Tumors in the proximal 2/3 of the ureter: lymph nodes of the ipsilateral great vessel extending from the renal hilum to the aortic bifurcation.
• Tumors in the distal 1/3 of the ureter: ipsilateral pelvic LND to include at minimum the obturator and external iliac nodal packets. Internal and common iliac nodal packets may be removed in the appropriate clinical setting. Limited data suggest cranial migration of lymph node metastases to the ipsilateral great vessels such that higher dissection may be considered in the appropriate clinical setting and per clinician judgement.

Taken in sum, there is sufficient non-randomized evidence to suggest an oncologic benefit to LND at the time of NU for patients with “HR” stratification by guidelines, the Panel recommends LND at the time of NU or SU for patients with HR UTUC.

Survival outcomes in patients with HR UTUC after RNU and LND is poor owing to the aggressive nature of the disease and the resulting compromise to renal function that limits further therapeutic options. Optimizing perioperative systemic treatment to improve outcomes is commonly achieved in the neoadjuvant setting when dosing regimens may be better tolerated, allowing more courses to be completed, and permitting patients to proceed to appropriate surgical intervention. Several meta-analyses evaluating NAC for UTUC have identified evidence for improved pathologic outcomes, CSS, and OS with this approach.¹²⁹

Two recently completed NAC trials of cisplatin-based chemotherapy prior to RNU strongly support this position. The designs of these single-arm prospective trials were extrapolated from data in muscle-invasive bladder cancer that provide high level evidence for neoadjuvant cisplatin prior to cystectomy compared to surgery alone.^{130, 131} Both trials used selection criteria that predicted for existing muscle-invasive disease at baseline in over 65% of patients.¹³² In the ECOG 8141 study, four cycles of NAC with accelerated MVAC (aMVAC, methotrexate, vinblastine, adriamycin, and cisplatin with growth factor support every two weeks) were planned.¹³³ Eligible patients had HG UTUC of the renal pelvis or ureter, pathologically confirmed and correlated with cross-sectional imaging, were without evidence of locoregional or metastatic disease, and were planned for RNU. In 29/30 aMVAC treated patients, eligibility included ECOG performance status 0 or 1, and creatinine clearance >50 mL/min. The pathologic complete response rate (ypT0N0/Nx) following RNU in this group was 13.8% (90% CI: 4.9-28.8) and deemed worthy of further study. In addition, 62% of eligible treated patients had final pathologic stage of <ypT1N0/x, an encouraging endpoint given improved long-term outcomes for patients with non-muscle invasive cancer following NAC in UTUC in retrospective series. Accelerated MVAC was tolerated as expected with 80% of patients completing planned four cycles. No patients progressed prior to surgery, none died of chemotherapy- or surgery-related toxicity. At a median follow-up of 21.1 months, the median relapse-free survival for subjects treated with aMVAC and RNU was not reached.

Renal function outcomes were also evaluated in this trial. Baseline median creatinine clearance was 82 mL/min (53.7 -170) prior to aMVAC with two patients (6.7%) having creatinine clearance < 60 mL/min. Following chemotherapy, 20% had creatinine clearance <60 mL/min, all were still surgery eligible. As expected, the largest decline in renal function occurred post-RNU with 69% having calculated creatinine clearance <60 mL/min. Results from this study have informed the design of EA8192, a randomized phase II/III trial of neoadjuvant aMVAC +/- durvalumab chemotherapy for four neoadjuvant cycles prior to RNU, currently enrolling.

Coleman et al. presented data from a prospective Phase II open label trial in a similar population of patients with HG non-metastatic UTUC planned for RNU, with platinum eligible renal function based on calculated glomerular filtration rate >55 mL/ min/1.73m² by CKD-EPI.¹³⁴ In this fully accrued trial, 58 patients enrolled and were treated with a dosing schedule of split dose cisplatin (35 mg/m² on days 1 and 8) with gemcitabine (1,000 mg/m² on days 1 and 8) for 4 planned cycles prior to RNU. The study was powered at the 90% level to detect a significant reduction in pathologic stage <ypT2N0 in ≥60% of patients. This trial met its primary endpoint with 63% of patients achieving <ypT2N0 status following surgery, including 19% with complete pathologic response (ypT0N0). Median follow-up was 3.1 years and 2- and 5-year PFS was 78% and 65%, respectively. The 2- and 5-year OS were 93% and 79%, and for both PFS and OS, lower final pathologic stage correlated with better PFS and OS outcomes.

Similar to EA8141, there were no patients with cancer progression post NAC, which precluded surgery, and no treatment related deaths. Further, 89% of patients received at least 3 cycles of NAC, with 47% completing all 4. All patients proceeded to surgery. The strongly positive data from these phase II trials, the established high-level evidence seen in bladder cancer trials, the consistent findings from pooled meta-analytic data, and the compelling clinical challenges imposed by post-RNU renal function on cis-platinum eligibility support the standard use of NAC regimens for HR UTUC. Observed results seen with phase II trials also set an important benchmark for any future such studies in this disease.

Alternatives to cisplatin-based chemotherapy (i.e., immune checkpoint inhibitors, carboplatin, antibody drug conjugates, targeted FGFR therapies) are not recommended in the neoadjuvant setting (prior RNU or ureterectomy) outside of clinical trials.

Adjuvant platinum-based chemotherapy for select patients with UTUC post-RNU is a standard based on results from the randomized phase III POUT trial.¹³⁵ In this study, 261 chemotherapy-naïve patients were identified and enrolled post-RNU, with HR patients selected based on postoperative stage in non-metastatic patients of pT2–T4 pN0–N3 M0 or pTany N1–3. In this trial, patients were randomized to platinum chemotherapy day 1 based on eligibility (cisplatin, or carboplatin for glomerular filtration rate <50 mL/min) with gemcitabine days 1 and 8 for four planned adjuvant cycles to start within 90 days of RNU. The trial was designed to show improved disease-free survival (DFS) in the chemotherapy versus the observation arm, and after meeting an early efficacy point, accrual was halted. At a median follow-up of 30.3 months, subjects in the adjuvant chemotherapy arm had improved DFS (HR: 0.45; 95% CI: 0.30 to 0.68; p=0.0001) compared with those on observation. Subjects on the chemotherapy arm had a significantly lower risk of metastases or death compared to observation (HR: 0.48; 95% CI: 0.31 to 0.74; log-rank p=0.0007). Side effects of platinum chemotherapy were as expected with no grade 5 events. The completion rate of four adjuvant cisplatin cycles was low in this dataset at 58%, including 21% of patients who started with cisplatin but switched to carboplatin for post allocation decline in GFR.

A subgroup analysis demonstrated that outcomes for patients with lymph node involvement and those treated with carboplatin chemotherapy were worse than those without positive nodes or treated with cisplatin chemotherapy.¹³⁶ As the primary endpoint was powered based on the intent to treat population, speculation about these subgroups, the potential utilization of six versus four cycles for metastatic N+ disease or the impact of carboplatin in this setting are hypothesis generating discussions. Based on these data, carboplatin remains a reasonable choice for HR cisplatin-ineligible patients post-RNU if NAC was not given.

Two completed RCTs compared adjuvant checkpoint inhibitor therapy versus observation (IMvigor 010) or placebo (CheckMate 274) following surgery in patients with HR non-metastatic urothelial carcinoma (Appendix V).^{137, 138} Although the majority of patients in these studies underwent radical cystectomy for bladder primaries, 20% of patients in CheckMate 274 and 7% of IMvigor 010 patients underwent surgery for UTUC, with endpoints based on the intention to treat population. Inclusion criteria for both studies were patients with HR urothelial cancer defined as pT3, pT4a, or pN+ for patients who had not received neoadjuvant cisplatin-based chemotherapy and ypT2 to ypT4a or ypN+ for patients who had received neoadjuvant cisplatin.

In the IMvigor 010 trial, (n=406; 29 with UTUC) planned one year of adjuvant atezolizumab did not meet the primary endpoint of improved DFS compared to observation (19.4 months versus 16.6 months; HR: 0.89; 95% CI: 0.74 to 1.08).¹³⁹ Another study, the phase III randomized adjuvant study of pembrolizumab in muscle invasive and locally advanced urothelial carcinoma including UTUC patients (AMBASSADOR) versus observation trial, has completed accrual and is maturing with data yet to be presented.¹⁴⁰

The CheckMate 274 (n=709; 149 with UTUC) study of one year of planned adjuvant nivolumab did meet its co-primary endpoints, with improved DFS (definition per-protocol included within and outside of the urothelial tract) of 20.8 months (95% CI: 16.5 to 27.6) with nivolumab versus 10.8 months (95% CI: 8.3 to 13.9) with placebo in the intention to treat population.¹³⁸ The 6-month DFS benefit of 74.5% with nivolumab and 55.7% with placebo (HR: 0.55; 98.72% CI: 0.35 to 0.85; P<0.001) was even more striking in patients whose tumors expressed PD-L1 (>1%).

Additionally, non-urothelial tract RFS (77.0% versus 62.7%; HR: 0.72; 95% CI: 0.59 to 0.89), and distant metastasis free survival (MFS, 82.5% versus 69.8%; HR: 0.75; 95% CI: 0.59 to 0.94) were also improved. In a subgroup analysis of patients with UTUC, there was no difference in DFS for renal pelvic cancers (HR: 1.23; 95% CI: 0.67 to 2.23) or the ureter (HR: 1.56; 95% CI: 0.70 to 3.48) in either arm. The small sample size limits the statistical power to detect a difference, and thus the results from this subgroup analysis in UTUC are hypothesis generating only. Based on the strength of the overall evidence, adjuvant nivolumab was approved for UTUC and urothelial carcinoma of the bladder in patients with advanced disease identified from post-surgical pathology findings.

With respect to harms, nivolumab was well tolerated and similar to placebo with respect to overall AEs (98.9% versus 95.4%) and grade 3 or higher AEs (42.7% versus 36.8%).¹⁰⁷ However, nivolumab was associated with increased likelihood of treatment-related AEs (77.5% versus 55.5%) and grade 3 or higher treatment-related AEs (17.9% versus 7.2%). The most common toxicities in the nivolumab group were pruritus (23.1%), fatigue (17.4%), and diarrhea (16.8%); and the most common grade 3 or higher AEs were elevations in serum lipase (5.1%) and amylase (3.7%) levels, diarrhea (0.9%), colitis (0.9%), and pneumonitis (0.9%). Treatment-related death occurred in three patients treated with nivolumab (two due to pneumonitis and one due to bowel perforation). Toxicity-related treatment discontinuation was also higher from nivolumab compared to placebo (12.8% versus 2.0%); most frequently from pneumonitis (1.7%), rash (1.1%), colitis (0.9%), and increased alanine aminotransferase level (0.9%). These toxicities are similar to other checkpoint inhibitor studies with no new safety signals noted. No adjuvant studies have compared nivolumab to platinum-based chemotherapy regimens.

Based on the relative strengths of the available data, the Panel recommends the use of adjuvant platinum-chemotherapy over adjuvant nivolumab for eligible patients who did not receive NAC. Scenarios for use of adjuvant nivolumab include: 1) patients with contraindications to platinum-based chemotherapy (e.g., poor renal function, performance status, sensorineural hearing loss, neuropathy or congestive heart failure, allergy), 2) patients with HR pathology after NAC, 3) patients who refuse standard forms of adjuvant chemotherapy after appropriate counseling. 

No clear evidence supports upfront RNU without chemotherapy in the setting of known metastatic (M+) UTUC. Oncologic outcomes in the metastatic setting are strongly determined by response to systemic therapy, and surgical treatment has no demonstrable therapeutic efficacy for cytoreduction or as a single modality in this setting. Potential harms such as delay or inability to receive systemic therapy due to consequences of surgery can significantly and negatively impact oncologic outcomes and OS in this setting. Therefore, clinicians should favor systemic therapy and alternative approaches (i.e., radiotherapy with or without chemotherapy in selected cases) for inoperable or symptomatic patients with M+ UTUC. 

Retrospective studies suggesting clinical benefit from surgery to the primary site in patients with metastatic UTUC apply specifically to those who have already received first-line chemotherapy or from data sets where use of peri-operative chemotherapy is poorly documented, thus limiting interpretation and applicability due to strong selection biases and significant weaknesses in the data sets.^{141, 142}

The Panel emphasizes that, in the case of cN1-3 UTUC, the primary treatment is chemotherapy, and that surgery with curative intent be considered as a consolidation strategy after complete or, in select cases, partial response. Supporting data for this statement are derived mainly from studies of platinum-based chemotherapy regimens, and results should be interpreted contextually.

Patients with clinically suspicious lymph nodes are classified as HR unfavorable with likely established locally advanced or metastatic disease. These patients with clinically evident or biopsy proven regional nodal metastases who demonstrate suitable response to systemic therapy that converts their disease to a clinical state amenable to surgical resection should be offered surgical treatment if medically suitable. Pooled data from comparative outcomes utilizing NAC in patients with clinically node positive (cN+) disease supports this approach.

Three reviews included up to six individual studies were consistent in finding NAC associated with improved oncologic outcomes versus NU alone. The largest total sample (N=1,252) included a study of patients with cN+ M0 UTUC in the U.S. National Cancer Database (n=720).¹⁴³ Based on a pooled analysis of adjusted risk estimates, NAC was associated with significantly better OS (5 studies; adjusted HR: 0.53; 95% CI: 0.40 to 0.69) and CSS (2 studies; adjusted HR: 0.39; 95% CI: 0.23 to 0.67) compared to NU alone. Findings for OS were similar in the subgroup of patients with locally advanced tumors (≥cT3 or cN+; 4 studies; adjusted HR: 0.54; 95% CI: 0.41 to 0.72; p<0.001). A review including 5 studies common to all the reviews also found NAC associated with improved RFS versus NU (3 studies; HR 0.50; 95% CI: 0.37 to 0.66; p<0.0001).

Patients’ localized disease may be deemed unresectable or ineligible for extirpative surgical management due to significant medical comorbidities or other factors including refusal to accept surgical treatment (e.g., solitary kidney). Formulating alternative care options should be approached with multi-disciplinary input with a focus on realistic goals of care such as providing means of local control for functional preservation (e.g., renal function) and palliation (e.g., bleeding, infection). Appropriate patient counseling with an explanation of goals and expectations should be provided and documented. Clinical trials, where available, should be discussed with, sought out, and offered to eligible patients. Multi-modal approaches include combination of endoscopic management to maintain upper and lower tract function (e.g., stents, nephrostomies, ablation for bleeding and local control) in addition to systemic treatment options if available. Rarely, radiation, angioembolization, or percutaneous ablation for palliation of bleeding can be offered based on anecdotal case report data.^144-146

Surveillance regimen should be tailored to disease risk as well as treatment modalities taken.

Patients with LG (LR) UTUC managed with nephron-sparing approaches should, at a minimum, undergo cystoscopic surveillance three months after endoscopic treatment, then once again within the first year after treatment, then every six to nine months for two years. Upper tract imaging, preferably with CT urogram, should be done at least every six to nine months for the first several years but can then be done annually out to year five. Follow-up ureteroscopic evaluation should be performed at a regular interval within the first year but can then be performed with any symptoms or significant findings on upper tract imaging.

Risk of recurrence

Endoscopy and radiographic imaging can be utilized to evaluate the upper tracts for recurrence within the affected and contralateral system. Although risk of recurrence varies on disease characteristics as well as medical therapy (e.g., mitomycin gel), periodic evaluation will diagnose disease earlier in the disease progression. Of three studies that reported adjusted risk estimates, one study (n=120) found endoscopic management associated with increased risk of any (local, intravesical, or distant) recurrence (adjusted HR: 3.56; 95% CI: 1.73 to 7.35),⁸⁴ one study found endoscopic management associated with improved intravesical RFS (adjusted HR: 0.56; 95% CI: 0.25 to 1.25),⁸¹ and one study found endoscopic management associated with increased risk of local recurrence (adjusted HR: 1.27; p=0.001) but no difference in risk of intravesical RFS (adjusted HR: 0.90; p=0.52).⁸⁶ The follow-up evaluation schedule attempts to balance the morbidity and cost of follow-up with the risk of disease recurrence. Clinicians may elect to increase the intensity of surveillance above the minimum recommendations as listed in the guideline according to their assessment of an individual patient’s risk and shared decision-making.

Risk of metastasis

Another important reason for disease follow-up is to evaluate patients for metastatic disease; the likelihood of which for LG/LR disease is low. Three studies reported inconsistent results for metastasis: one study⁸² (n=162) reported higher MFS for patients with LG UTUC who underwent ureteroscopic management versus patients with any grade UTUC who underwent NU (5-year MFS 84% versus 60%; 10-year MFS 75% versus 54%), one study⁶⁴ (n=453) reported similar 5-year MFS for patients with LG UTUC who underwent endoscopic management versus NU (81% versus 84%, p=0.99), and one study⁸⁴ (n=120) that did not stratify results by UTUC grade reported similar likelihood of distant metastasis for endoscopic management versus NU (24% versus 27%).

In an asymptomatic patient with a history of LR UTUC, a clinician should perform baseline chest imaging, which can be done by chest X-ray or CT of the chest, but routine imaging thereafter is not required.

Surveillance regimen should be tailored to disease risk as well as treatment modalities received.

Some patients with HGT1 or less (HG Ta, T1, or CIS) may be managed with nephron-sparing treatments, such as endoscopic management, topical therapy, segmental ureteral resection (including distal ureterectomy) or other therapies (e.g., systemic therapy or surveillance). Patients with HG Ta, T1, or CIS urothelial carcinoma receiving nephron-sparing treatment should undergo cystoscopic surveillance starting within three months after treatment, continuing every 3-6 months for 3 years, and then every 6-12 months through year 5. Follow-up ureteroscopy should be performed at least once within three to six months of endoscopic therapy and subsequently at the discretion of the clinician. Ureteroscopy for patients who undergo definitive surgical management (e.g., SU or distal ureterectomy) should be performed at three to six months after resection, and then may be performed at the discretion of the clinician, who might prefer cross-sectional excretory imaging in lieu of ureteroscopy. For patients with HGT1 or less, managed by nephron-sparing treatment, upper tract imaging with CT urogram and basic metabolic panel (BMP) should be performed every 3-6 months for 3 years, then every 6-12 months for 2 years, and annually thereafter. Chest imaging with chest X-ray or CT of the chest is recommended every 6-12 months to evaluate for intrathoracic metastasis up to 5 years following last diagnosis/treatment. As discussed with respect to the initial characterization of UTUC, MR urography or retrograde pyelography combined with non-contrast axial imaging may be utilized in patients in whom the administration of iodinated contrast is contraindicated.

Risk of recurrence

The majority of studies evaluating risk of recurrence with nephron-sparing treatment are retrospective heterogeneous analyses that include both LG and HG tumors. A retrospective cohort study of 198 patients with pTa, pTis, or pT1 disease received either endoscopic treatment or RNU, of whom 15% and 25% had HG disease, respectively.⁴² Mean postoperative creatinine levels were slightly improved among patients who received nephron-sparing surgery (1.32 standard deviation 0.47] versus 1.64 [SD 0.79; p=0.048]). Among patients receiving endoscopic treatment, recurrence within the ipsilateral upper urinary tract was higher (25% versus 1.2%; p<0.001), but recurrence in the bladder was slightly lower (15% versus 36%; p=0.056). A smaller retrospective cohort of 43 patients undergoing either endoscopic treatment or RNU, of whom 20% and 91% had HG disease, respectively, showed a higher rate of bladder recurrence among patients undergoing nephron-sparing surgery (60% versus 18%; p=0.008).⁴⁰ Given the HR of recurrence in both the upper and lower urinary tract, risk-adapted surveillance suggests close monitoring to reflect a high recurrence risk within this patient population.

Risk of metastasis

In studies comparing endoscopic management versus NU, few patients undergoing endoscopic management had HG UTUC, and most patients received nephron-sparing treatment did so under palliative or emergent conditions (e.g., bleeding, renal failure).⁸² In one study, ureteroscopic management for HG UTUC in 14 patients was associated with worse 2-year OS (54% versus 77%), CSS (54% versus 78%), and MFS (34% versus 66%) versus NU in 80 patients (any grade; 71% with HG UTUC). At 5 years, survival was 0% in the HG ureteroscopy group, compared with 5-year OS of 58%, CSS of 64%, and MFS of 60% with surgery. One other study found percutaneous endoscopic management associated with worse OS (34.6 months versus 58.0 months) and CSS (27.8 months versus 56.7 months) in the subgroup patients with grade 3 UTUC (n=34).⁸³ In another retrospective cohort of 8,304 patients, 633 patients underwent nephron-sparing treatment of whom 39.7% had HG disease. Median OS and 3-year DSS were worse among those undergoing nephron-sparing treatment compared to those undergoing NU (1.9 years versus 7.8 years; p<0.001 and 73.7% versus 92.4%; p<0.001; respectively).⁵⁸ Given the comparatively worse OS, CSS, and MFS rates among patients with HG disease undergoing nephron-sparing surgery, a risk-adapted surveillance scheme should incorporate cross-sectional imaging of the abdomen and pelvis as well as chest imaging to evaluate sites of metastasis.

The development of a recurrence of urothelial carcinoma in the lower urinary tract or a positive cytology in the context of a patient with a history of UTUC should raise the possibility of recurrent disease in the upper tracts. Thus, patients who develop lower tract recurrence or a positive cytology without a clear etiology should undergo an evaluation of the upper tracts. Depending on the clinical scenario this may be via cross-sectional imaging or retrograde pyelography with or without selective upper tract cytology. If these modalities suggest the presence of upper tract involvement, then further evaluation, including endoscopy, is warranted.

Follow up after NU for patients with non-muscle invasive, node-negative UTUC should be largely focused on the risk of intravesical recurrence. Two systematic reviews of recurrence rates after NU found similar rates of intravesical recurrence after NU to be approximately 29% with a median time to recurrence of 6-12 months¹⁴⁷ or 22 months.¹⁴⁸ The study by Kapoor et al. also summarized the overall risk of recurrence to the contralateral upper tract at 2.2% (range 0% to 4.6%; mean follow-up 46.7 months).¹⁴⁷ Locke et al. led a large Canadian study of over a 1,000 patients that found similar results with a local recurrence rate (both bladder and upper tract) of 24% with a median time to recurrence of 7 months.¹⁴⁹ That study also noted that 91% of the local recurrences were identified within the first 2 years, while late recurrences did occur as late as 150 months after surgery. Therefore, in the first two years after NU, there should be regular attention paid to monitoring for intravesical recurrence through regular cystoscopic surveillance. After two years, the frequency can be significantly reduced though, as with non-muscle invasive bladder cancer, how long surveillance should be continued is not clear.¹⁵⁰ Periodic imaging of the upper tracts should be undertaken given the risk of recurrence to the contralateral upper tract, preferably with cross-sectional imaging such as CT urogram, though the rate is low enough that this can be done annually after NU.

The Locke et al. study broke down the risk of regional or distant recurrence by grade and stage when examining the risk of locoregional or distant metastases.¹⁴⁹ Patients with less than pT2 or pN+ disease were considered either LR (pTa-T1, pN0, LG, no LVI, and not multifocal) or intermediate (pTa-T1, pN0 and HG, LVI present, or multifocal). The 3-year estimated freedom from regional or distant metastases were 93% and 87% for these low- and intermediate-risk groups, respectively. The rate of intrabdominal recurrences in LR patients from this study was very low while in intermediate-risk patients it was 17%, with most occurring within the first 2 years. Thus, periodic imaging of the abdomen and pelvis is warranted, especially for those HG disease, LVI or tumor multifocality, particularly for the first two years. The risk of lung metastases for patients with less than pT2 or pN+ disease is overall low but can occur in those with HG disease so periodic chest imaging (see table 6) should be undertaken and can be done via chest x-ray or CT of the chest, though the former is likely sufficient, less costly, and associated with less radiation exposure.

Table 6: Surveillance after complete treatment. The following surveillance schedules are recommended in the setting of complete treatment where no residual or recurrent tumor is identified or clinically suspected. Earlier intervals of follow-up endoscopy may be used in cases of concern for incomplete treatment (e.g., larger tumors, more difficult access, poor visibility, disease biology). The Panel recognizes the limitations of the data on tumor recurrence and optimal intervals of follow-up which require further study. Any clinical findings of new or worsening disease should prompt re-evaluation.

Regional Recurrences (Bladder) following NU

Follow-up after NU for non-metastatic node-negative pT2 and higher disease requires surveillance for local and regional recurrence, intravesical recurrences, and distant metastases. A meta-analysis of 59 studies¹⁴⁷ evaluating recurrences following NU reported a 29% risk of intravesical recurrence within a median 6-12 months after RNU. A 2016 systematic review of 18 studies enumerated key risk factors for intravesical recurrence including male sex (HR: 1.37; p<0.001), previous bladder cancer (HR: 1.96; p<0.001), preoperative CKD (HR: 1.87; p=0.002), positive preoperative urinary cytology (HR: 1.56; p<0.001), ureteral tumor site (HR: 1.27, p<0.001), multifocality (HR: 1.61; p=0.002), invasive pathologic T-stage (HR: 1.38; p<0.001), presence of necrosis (HR: 2.17; p=0.02), laparoscopic approach (HR: 1.62; p=0.003), extravesical bladder cuff removal (HR: 1.22; p=0.02), and positive surgical margins (HR 1.90; p=0.004).¹⁴⁸ Additional independent risk factors for intravesical recurrence included having positive surgical margins (HR: 3.36; 95% CI: 1.36 to 8.33) and prior ureteroscopic biopsy (HR: 1.39; 95% CI: 0.88 to 2.19).¹⁵¹ A large Canadian study (n=1,029) similarly identified a 26% risk of urothelial (bladder or contralateral upper tract) recurrence with a mean time to recurrence of 7 months.¹⁴⁹ Post-NU bladder recurrences were observed in 21%, 26%, and 36% of patients with UTUC of the renal pelvis, ureter, or both, respectively. In patients with HR disease, defined as ≥pT2 or pN+ and any grade, with or without LVI or multifocality, bladder recurrences were observed in 53% of patients compared to 14% and 33% for low- and intermediate-risk disease; with 52% of recurrences in the bladder in the intermediate-risk categories occurring in the first year following NU. There are conflicting data regarding the risk of bladder recurrence according to the location of the primary tumor. Given the substantial risk of local (bladder) recurrences within the first years following NU, risk adapted surveillance with cystoscopy and urine cytology at routine intervals is indicated to facilitate prompt detection of bladder recurrences.

Locoregional recurrence, retroperitoneal nodal and distant metastases following NU

In the previously mentioned meta-analysis of 59 studies¹⁴⁷ evaluating recurrences following NU, in addition to the risk of intravesical recurrence, they reported recurrences of the retroperitoneum or pelvis occurred in 4.6% of patients within an average 32.7 months, and distant metastases occurred in 16.4% within an average of 46.8 months. Retroperitoneal lymph node metastasis occurred in 5.2% of patients within a mean of 46.8 months), while lung, liver, and bone metastases were observed in 4.8%, 4.1%, and 3.7% of patients, respectively. The median time to metastases was 13 months to 16 months (range 1 month to 50 months postoperatively). The large Canadian study (n=1,029) mentioned previously found 24% of patients experience locoregional (in the nephrectomy bed or retroperitoneal lymph nodes) or distant (lung, bone, liver, brain, or other) recurrence with a mean time to recurrence of 8 months.¹⁴⁹ In this analysis, 91% of local recurrences were diagnosed in the first 2 years, though late local recurrences (up to 150 months) were observed. Post-NU locoregional and distant recurrences were observed in 21%, 24%, and 31%, respectively. In patients with intermediate- and HR UTUC, while rare, the vast majority of recurrences to the nephrectomy bed, liver, and distant metastases to the lungs and bones were observed in patients with intermediate- and HR disease within 18-24 months following NU. Risk factors for recurrence following NU include risk factors associated with increased risk of recurrence were multifocality, stage T3-4, grade G3, and presence of lymph node metastasis; UTUC site in ureter versus renal pelvis was not an independent predictor.¹⁵² Given this risk of locoregional recurrence and metastasis in patients with > pT2 UTUC following NU, risk-adapted routine surveillance with contrast-enhanced cross-sectional imaging and urography is recommended, with decreasing intensity in years three to five, and subsequent follow-up surveillance recommended according to principles of informed/shared decision-making.

Of note, brain metastases are rare following NU, but have been observed only in patients with a prior history of HR UTUC, within an average of 18 months of NU.¹⁴⁷ Patients undergoing follow-up for HR UTUC following NU with acute neurological signs or symptoms should undergo prompt neurologic evaluation with cross-sectional imaging of the brain and/or spine by CT or MRI.

For patients undergoing follow-up for treated UTUC, additional site-specific imaging can be ordered as warranted according to clinical symptoms suggestive of local recurrence or metastatic spread. PET scans should not be obtained routinely but may be selectively considered for patients who are at risk for metastatic recurrence and are not able to have contrast enhanced CT and MRI. Finally, patients with findings suggestive of metastatic UTUC should be evaluated to define the extent of disease and referred to medical oncology for further management.

In addition to following patients for cancer recurrence or metastasis, clinicians should monitor patients for the sequelae of NU. Following NU for HR UTUC, the Panel recommends that patients should undergo periodic laboratory assessment including serum creatinine level, eGFR, and urinalysis. Other laboratory evaluations (e.g., CBC, LDH, liver function tests, and alkaline phosphatase) may be obtained at the discretion of the clinician or if advanced disease is suspected. In patients who develop progressive renal insufficiency or proteinuria should be referred to nephrology.

Referral to nephrology should be considered for patients with eGFR less than 45 mL/min/1.73m², confirmed proteinuria, diabetics with preexisting CKD, or whenever eGFR is expected to be less than 30 mL/min/1.73m² after intervention.

The long-term impact of renal dysfunction increases risks of osteoporosis, anemia, metabolic and cardiovascular disease, hospitalization and death. Effective treatment strategies are available to slow the progression of CKD and reduce cardiovascular risks, and therefore timely identification of progressive renal dysfunction and/or proteinuria can provide opportunity for medical intervention when indicated. The two formulas for monitoring eGFR commonly reported in the contemporary literature at this time are the Modification of Diet in Renal Disease and CKD – Epidemiology Collaboration (CKD-EPI) equations.

Risk factors such as smoking are associated with advanced disease stage, recurrence and worse CSM among patients with UTUC, with the highest risk among current smokers.¹⁵³ Therefore, clinicians should discuss and facilitate smoking cessation with patients at the time of diagnosis and treatment. UTUC is also associated with metabolic syndrome and obesity, with obesity adversely impacting disease-specific outcomes among patients undergoing RNU.^{154, 155} Clinicians should, therefore, encourage patients to adopt healthy lifestyle habits regarding exercise and a healthy diet to promote long-term health benefits and quality of life. Finally, clinicians should work with patients and their primary care providers to ensure that comorbidities are optimally managed throughout the course of care for UTUC and during surveillance to maximize quality of life during survivorship.