Predictive factors of recurrence and survival of upper tract urothelial carcinomas

Abstract

Objective

UTUCC is a rare tumor and most reports on prognostic factors come from small single centers series. The objective of this article was to provide an updated overview of current clinical, pathological and biological prognostic factors of UTUC.

Methods

PubMed was searched for records from 2002 to 2010 using the terms “prognostic factors”, “recurrence”, “survival”, and “upper tract urothelial carcinoma”. Among identified citations, papers were selected based on their clinical relevance.

Results

Classical clinical factors that influence UTUC prognosis include age, presence of symptoms, hydronephrosis and interval from diagnosis. Many biomarkers have shown promises to better appraise the natural course of UTUC although none is currently used in clinical practice. Stage, grade, lymph node metastases, lymphovascular invasion, tumor necrosis, and tumor architecture are strong pathological parameters. RNU is the standard treatment of localized UTUC. Both laparoscopic and open approaches seem to offer similar cancer control. Lymph node dissection increases staging accuracy and might confer a survival benefit.

Conclusion

RNU is the standard treatment for most patients with UTUC. Recent multicenter studies confirmed the prognostic value of classical prognostic parameters. Better survival prediction might be obtained with prognostic systems including clinical data and new biomarkers.

Introduction

Upper tract urothelial carcinoma (UTUC) is a rare tumor that represents 5% of all genitourinary malignancies.[1] It occurs more frequently in the renal pelvis than in the ureter with a ratio of 3:1.[2, 3] Classical risk factors for the development of UTUC include smoking, abuse of analgesics, chronic urinary tract infection, stone disease, and oncologic agents such as cyclophosphamide.[2] UTUC can develop after primary bladder cancer in up to 10% of cases. Secondary bladder cancer after primary UTUC is more common with a risk of 20 to 50%.[3–5]

Radical nephro-ureterectomy (RNU) with ipsilateral bladder cuff excision is the surgical standard of care for patients with non metastatic UTUC. Despite effective local therapy, disease recurrence and progression remain common. The most important prognostic factor of UTUC is disease stage. Five-year survival rates approach 90% for low stage tumors and decrease to < 30% in cases of regional nodal metastases and < 10% in case of distant metastases.[5] Endoscopic management can be an option for patients with small, unilateral low stage and low grade tumors [6, 7].

Because of the rarity of UTUC, most of the publications concerning UTUC were single center series, until recently. Although they largely contributed to the understanding of the disease, they were limited by small size and heterogeneous populations. To overcome this limitation and in an effort to better understand the natural history, a comprehensive database (UTUCC; the Upper Tract Urothelial Carcinoma Collaboration) incorporating the clinico-pathologic characteristics and outcomes of more than 1300 patients treated with RNU for UTUC at 13 academic centers worldwide was created in 2008.[8] Many publications focusing on prognostic factors came out of this collaborative effort. To externally validate the findings of these studies, a validation cohort based on over 700 RNU cases was created.

In the light of data, the objective of this review was to investigate prognostic factors of UTUC outcomes.

Clinical factors

Age

There is very limited data on the impact of age on clinical outcomes in UTUC. In a retrospective study from the UTUCC including 1453 patients, Shariat et al. reported that older age was an independent predictor of cancer-specific mortality (CSM).[9] This finding could be explained by a change in the biological potential tumor cells, a decrease in the host’s defense mechanisms, or differences in care patterns.[10] However, advanced age alone should not be an exclusion criterion for the aggressive treatment of potentially curable UTUC. A large proportion of elderly patients can be cured with RNU.[9] Therefore, chronological age alone is an inadequate indicator of outcomes in older UTUC patients and should not be used to deny a potentially curative intervention to elderly patients.[9]

Gender

No difference has been shown in histopathological features and outcomes between men and women treated with RNU for UTUC.[11, 12]

Symptoms

The presence of symptoms is a classical prognostic factor in renal cell carcinoma (RCC) patients.[13] The UTUCC studied 654 patients managed by RNU and found no difference in terms of recurrence-free survival between patients with incidental tumors and patients with local symptoms. However, patients with systemic symptoms had a significantly higher risk of disease recurrence (p<0.001).[14] Nevertheless, systemic symptoms could not predict recurrence or cancer death in multivariable analysis. Yet, symptom classification improved the accuracy of a predictive model comprising stage, grade and lymph node status. In a smaller series of 168 patients, Inman and colleagues built a preoperative model, where presence of constitutional symptoms (pain or weight loss) was a predictor of overall survival (p=0.007), and trended towards predicting cancer mortality (p=0.064).[15]

Hydronephrosis

In bladder cancer, hydronephrosis is a sign of advanced disease and a predictor of poor outcome.[16] In UTUC, two previous studies reported that non-visualization of the urinary tract, delayed excretion, or hydronephrosis were associated with invasive ureteral cancer.[17, 18]

Interval from diagnosis

As suggested in bladder cancer[19], a recent study showed that longer interval from diagnosis of UTUC to RNU was associated with aggressive features, such as more advanced stage and higher tumor grade, but not with disease recurrence or cancer-specific mortality. However, in the subgroup of patients with stage ≥ pT2, longer delay was associated with higher risk of disease recurrence and cancer-specific mortality.[20]

Biomarkers

Conventional serum markers

Recent research into the profiling of UTUC at the molecular level has begun to shed light on important mechanisms of pathogenesis, as well as providing a number of potential diagnostic and prognostic markers. Molecular markers have the potential to be used clinically to screen for, diagnose, or monitor the activity of diseases and to guide molecular targeted therapy or assess therapeutic response. The prognostic value of proteins implicated in the regulation of cell cycle (p53 and p27), apoptosis (bcl2, survivin), cell adhesion (E-Cadherin) and cell proliferation (Ki67) have been suggested in small single-center series.[3, 21, 22] E-Cadherin and Ki67 were shown to be independent prognostic factors of tumor recurrence in multivariable analysis.[21, 22] Microsatellite instability has also been identified as an independent prognostic parameter, mainly among patients with pT2-T3 N0 M0 UTUC.[3, 23] Although these markers are promising, none are used in clinical practice. Tissue microarray studies are currently ongoing as part of the UTUCC biomarker and validation project. Based on the lower tract urothelial carcinoma experience, a combination of complementary and yet independent molecular markers will likely better capture the biologic potential of each individual urothelial tumor, resulting in improved clinical decision-making.

Laparoscopic versus open RNU

There has been a controversy regarding the laparoscopic treatment of UTUC because of the fear of potential tumor seeding. Most of the studies are retrospective and it is unlikely that a large randomized analysis will be performed due to the rarity of UTUC. However, the retrospective series suggest that the laparoscopic approach is safe and effective in appropriately selected patients.

While some authors reported a higher risk of intravesical disease recurrence with laparoscopic approach [24], the studies were small, monocentric and multivariable analyses failed to characterize treatment modality as an independent risk factor for disease recurrence.[25, 26] A recent publication from the UTUCC including 1249 patients reported that laparoscopic RNU offered equivalent oncologic efficacy to open surgery in selected patients. There was no difference between the two approaches with respect to disease recurrence or cancer-specific mortality; although patients treated laparoscopically had significantly more favorable cancer characteristics which might have resulted in selection biases.[27] Other retrospective reports confirmed these findings.[27–29] There is only one prospective randomized study that included 80 patients treated with either laparoscopic or open RNU. With a follow-up of 44 months, the authors found comparable oncologic outcomes between the laparoscopic and open group. However, in patients with pT3 and high grade UTUC, there was a slight advantage in cancer specific survival in favor of the open RNU.[30] Results were comparable with the retroperitoneal approach, with low complications rate suggesting that it was also a safe procedure.[31, 32]

Endoscopic treatment

Endoscopic surgery can be considered in imperative situations (solitary kidney, bilateral disease, significant comorbidities) but also in selected patients with a small low grade tumor. Both retrograde and antegrade approaches can be used depending on tumor volume and location.[7] Retrograde uretersocopy has a low morbidity but requires smaller instruments that limit the size of the tumor that can be treated.[6] In addition, some portions of the upper urinary tract, such as the lower pole calyces, can sometimes not be reliably reached with working instruments. The antegrade approach allows the treatment of larger tumors but with the potential risk of tumor seeding. Only limited monocentric series report on the endoscopic treatment of UTUC with a recurrence risk of about 30%.[6, 7]

Adjuvant instillations

Because of the high local recurrence rate of patients managed endoscopically (around 30%), adjuvant topical chemotherapy has been utilized. The instillation can be accomplished via a nephrostomy tube or through a ureteral catheter. The same agents used to treat urothelial carcinoma of the bladder are used to treat tumors of the upper tracts (i.e., BCG, mitomycin and other chemotherapies). Although the cumulative experience appears encouraging, no long-term study has shown statistical improvement with relation to survival and recurrence rates.[33, 34] This may be due to insufficient numbers to show clinical significance, given the relative rarity of the disease, differential biology of upper urinary tract versus bladder tumors, or inadequate delivery systems that do not allow for uniform delivery and adequate dwell times to enable a clinical response.

Pathological factors

Tumor location

Historically, the prognostic impact of tumor location resulted in contradictory results. Tumors within the renal pelvis are more common than ureteral lesions, but small single-institution series have suggested that ureteral disease confers worse prognosis.[35] A European study, including 269 patients from three academic centers who underwent RNU at three academic centers, noted that tumor location was not independently associated with clinical outcomes.[36] A larger series from the UTUCC showed that in 1249 patients that there is no difference in terms of recurrence or cancer-specific mortality between ureteral and renal pelvic tumors (p=0.133).[37] This was confirmed in nine SEER registries relying on 2824 patients; Isbarn et al. failed to detect an association of tumor location with cancer-specific survival.[38] Finally, Favaretto et al. confirmed the lack of prognostic importance of tumor location when adjusted for the effects of standard pathologic variables in RNU patients.[28]

Tumor architecture

Tumor architecture (infiltrative or papillary) was found to be associated with disease recurrence in international multicenter analyses.[8, 39, 40] Five years after RNU, 40% of patients with papillary vs. 77% with infiltrative tumors had disease recurrence.[39] Langner at al. showed that infiltrative pattern was significantly associated with the development of metastatic disease and was an independent prognostic factor of survival.[41] Combining this parameter with histological grade and tumor location in 1453 patients, a preoperative prognostic model achieved 76.6% accuracy in predicting non-organ confined UTUC.[40]

Tumor stage

Tumor stage is currently the most important prognostic indicator of UTUC. Upper tract urothelial carcinomas can spread by direct invasion, mucosal seeding, hematologic and lymphatic routes. The metastatic potential, and therefore the prognosis, worsens with advancing tumor stage. All the largest series validated tumor stage as a prognostic indicator.[3, 5, 8] Patients with pTa-pT1 tumors have a 5-year cancer-specific survival > 90%, whereas patients with pT3-4 tumors have a 5-year cancer-specific survival of 40.5% and 19% respectively.[4, 36] Tumor stage is also associated with a higher risk of local recurrence in many studies.[8, 42–44] In the UTUC studies, advanced pathological stage was consistently associated with disease recurrence and cancer specific survival.[8, 45, 46]

Lymph node invasion

Up to 30% of patients with muscle-invasive UTUC present with lymph node (LN) metastasis at diagnosis, which is associated with poor prognosis.[2, 3] Lymph node invasion has been demonstrated to be one of the most important predictors of poor outcome in patients treated with RNU.[8, 36, 40, 43, 47] There are no definitive data supporting the survival benefit of lymph node dissection in UTUC patients. Indications for lymphadenectomy are extrapolated from bladder cancer data that advocate extensive pelvic lymph node dissection to improve both staging and survival in patients undergoing radical cystectomy.[48] In a study comprising 1130 patients from the UTUCC, 5 year specific survival was lower in patients with pN+ disease compared to those with pNx disease (35% vs. 69%, p<0.001).[47, 49, 50] Interestingly, patients with pNx disease had a worse prognosis than pN0 patients, particularly in patients with advanced T stage. Therefore, the authors recommend that patients suspected to have pT2-T4 disease should undergo lymphadenectomy to improve staging and eventually help guide decision-making regarding adjuvant chemotherapy. Based on patients from the same cohort, the authors further studied the importance of the extent of lymphadenectomy. [47, 49, 50] In the entire population, the number of LNs removed was not associated with cancer mortality. However, in the subgroup of pT0 patients, the extension of lymph node dissection seemed to be associated with better cancer-specific survival.[50] A minimum of 8 LNs seemed to be the most informative cut-off with 75% probability of detecting one or more positive LN, and accurate prediction of cancer-specific mortality.[49, 50] Finally, Bolenz et al. demonstrated that LN density (using a threshold of 30%) significantly affected cancer-specific mortality in UTUC patients. [51]

Lymphovascular invasion (LVI)

LVI has been shown to have an important prognostic role, and assessment of this feature may help identify patients who could benefit from multimodal therapy after RNU.[43, 52] Lymphatic vessels serve as the primary pathway for metastatic tumor cell spread in many types of cancer.[53] Early studies reported that LVI is an independent prognostic factor in UTUC.[43, 54, 55] More recently, the UTUCC confirmed that patients with LVI had worse cancer-specific survival than their counterparts without LVI.[56] LVI was an independent predictor of both disease recurrence and cancer-specific mortality, even among patients who had either negative lymph nodes or who did not undergo lymphadenectomy [56]. LVI was associated with established features of biologically aggressive UTUC, such as advanced stage, high tumor grade, metastasis to lymph nodes, infiltrative tumor architecture, tumor necrosis, and concomitant carcinoma in situ (CIS).[36, 57] In addition, lymphovascular invasion was an independent predictor of both disease recurrence and cancer-specific mortality. When evaluated in all patients, addition of LVI to standard pathologic features improved their predictive accuracy for both disease recurrence and cancer-specific mortality by a statistically significant but clinically small margin. This margin was significantly larger and clinically significant when the analyses were restricted to patients without lymph node metastasis and those who did not undergo a lymphadenectomy. Besides inclusion in predictive tools for selecting adjuvant therapy, LVI could be considered for inclusion in the TNM staging system such as in hepatic and testicular cancer. However, a limitation is the inherent difficulty in determining the presence of LVI at the morphological level, with significant differences between pathologists. Retraction artifacts of the surrounding stromal tissue can mimic vascular invasion. Therefore, experts have recommended reporting LVI only in unequivocal cases, and to make use of immunohistochemistry, if necessary. The use of immunohistochemical staining to identify the vessels remains controversial and not practical for everyday clinical use. It is of utmost importance that strict morphological criteria are established to standardize and make the diagnosis of LVI reproducible, and consequently allow its recommendation in daily clinical setting.[43]

Tumor grade

Tumor grade is currently divided into papillary urothelial neoplasia of low malignant potential, low-grade carcinomas and high-grade carcinomas. Until 2004, the most common classification used was the WHO classification of 1973, which distinguished three grades (G1, G2, G3).[58] High-grade tumors are more likely to invade the underlying connective tissue, muscle, and surrounding tissues and are also more likely to be associated with concomitant CIS.[16, 51] The prognostic role of pathological grade is controversial. Several authors did not find any prognostic role for tumor grade [36, 52, 59], whereas others reported an effect of grade on UTUC recurrence and/or progression.[42, 44] These discrepancies might be related to similarities between grade 1 and 2 tumors and to the well-known intra- and interobserver variability to assign grade. Most recent series, including the one from UTUCC, used a 2-tiered grading system and found that high tumor grade was a strong independent prognostic factor in patients treated with RNU.[8, 39, 40]

Tumor necrosis

Only a few reports have addressed the prognostic value of tumor necrosis in UTUC. Langner et al. showed that tumor necrosis was an independent predictor of clinical outcomes and could predict distant metastasis after RNU.[60] This was subsequently confirmed by Simone [61] and by two large UTUCC series. [62, 63] In multivariable analyses, tumor necrosis was an independent predictor of both disease recurrence and cancer-specific mortality.[62] Assessment of tumor necrosis may help to identify patients who could benefit from multimodal therapy. Therefore, evaluation of extensive tumor necrosis has been suggested to become part of standard pathologic reporting.[62] Further extended validation is needed before tumor necrosis can be used to guide clinical decision-making after RNU.

Carcinoma in situ (CIS)

Carcinoma in situ is a cytological lesion of the urothelium and the basal membrane, with pre-invasive and invasive potential. In bladder cancer, CIS is associated with an increased risk of disease recurrence and progression.[64] In a series including 79 patients, CIS was a significant independent parameter of subsequent bladder recurrence (p=0.005).[65] The UTUCC confirmed these findings in 1387 patients; the presence of concomitant CIS in patients with organ confined UTUC (≤ pT2 N0M0) was associated with worse outcomes and a significantly increased risk of both cancer recurrence and cancer-specific mortality.[66–68]

Conclusion

Radical nephroureterectomy is the standard treatment for most patients with UTUC. Both laparoscopic and open approaches offer similar cancer control in appropriately selected patients. Lymphadenectomy seems to offer both a staging and survival benefit. The recent worldwide collaborative efforts confirmed classical clinical and pathological prognostic parameters (i.e. stage, grade, lymph node metastases) and new ones (LVI, necrosis, architecture). To better appraise the course of UTUC, there is a need to identify new biomarkers that could serve as prognostic indicators to include patients into clinical trials or help guide the clinical decision making regarding the type of treatment, integration of multimodal treatment and, response to treatment.

Abbreviations

CIS

Carcinoma In Situ

CSM

Cancer Specific Mortality

LN

Lymph Node

LVI

Lymphovascular Invasion

RNU

Radical Nephroureterectomy

SEER

Surveillance Epidemiology and End Results

TNM

Tumor, Nodes, and Metastases

UTUC

Upper Tract Urothelial Carcinoma

UTUCC

Upper Tract Urothelial Carcinoma Collaboration