Urologic Oncology: Seminars and Original Investigations 33 (2015) 18.e21–18.e26

Original article

Feasibility of obtaining biomarker profiles from endoscopic biopsy specimens in upper tract urothelial carcinoma: Preliminary results Bishoy A. Gayed, M.D.a, Aditya Bagrodia, M.D.a, Mansi Gaitonde, M.D.a, Laura-Maria Krabbe, M.D.a, Matthew Meissner, M.D.a, Payal Kapur, M.D.b, Ramy F. Youssef, M.D.a, Arthur Sagalowsky, M.D.a, Yair Lotan, M.D.a, Vitaly Margulis, M.D.a,* a b

Department of Urology, University of Texas Southwestern Medical Center, Dallas, TX Department of Pathology, University of Texas Southwestern Medical Center, Dallas, TX

Received 14 July 2014; received in revised form 25 September 2014; accepted 26 September 2014

Abstract Objective: To prospectively evaluate the feasibility of obtaining a reliable histochemical assessment of cell cycle biomarkers from endoscopic biopsy specimens of patients with upper tract urothelial cancer. Methods: Overall, 17 patients were identified who had an available biopsy as well as those who underwent subsequent radical nephroureterectomy (RNU) or segmental ureterectomy (SU) for clinically localized high-grade upper tract urothelial cancer of the renal pelvis or ureter. Of those 17 patients, 15 (88%) had sufficient tissue to undergo immunohistochemical staining. Biopsies were obtained using various endoscopic techniques. Tumor characteristics were recorded and prospectively evaluated for immunohistochemical expression of 5 biomarkers: p21, p27, p53, cyclin E, and Ki67/pRb. Unfavorable prognostic score (PS) was defined as 42 altered markers. Results: The median age of the patients was 68 years (range: 53–82 y) with 87% being males. Of the 15 specimens, 9 (60%) tumors were organ confined (T r 2 and N0), and all were high grade. Of the 15 patients, 4 (27%), 7 (46.6%), 3 (20%), and 1 (6.7%) individuals had 1, 2, 3, and 5 markers altered on biopsy marker profiling, respectively, with Ki67 being the most frequent alteration (13/15; 87.7%). An overall concordance rate of 60% (9/15) was seen between biopsy and RNU/SU PS. Those patients with favorable biopsy biomarker PS were less likely to display adverse pathological features, with organ-confined disease in 7/11 (63.6%) patients and 9/11 (81.8%) being free of carcinoma in situ in the final specimen. Additionally, 10/11 (91%) had no evidence of necrosis and 7/11 (64%) had no evidence of lymphovascular invasion on final pathologic evaluation. Conclusions: Preliminary results suggest that obtaining interpretable biomarker profile of ureteroscopic biopsy specimens is feasible. Tumor heterogeneity and limited biopsy material may account for the discordance between biopsy and RNU/SU specimens. Meaningful biopsy biomarker profiling could serve as a powerful tool for individualizing treatment regimens and augmenting current predictive variables. Further studies are needed to evaluate clinical applicability. r 2014 Elsevier Inc. All rights reserved.

Keywords: Upper tract urothelial carcinoma; Cell cycle; Biomarkers

1. Introduction Upper tract urothelial cancer (UTUC) comprises roughly 5% of all urothelial cancers [1]. At presentation, 30% of patients will demonstrate invasive disease or locally advanced disease, or both, and up to 20% will harbor metastatic disease [1–3]. Nearly 45% to 60% of patients Corresponding author. Tel.: þ1-214-648-0567; fax: þ1-214-648-8786. E-mail address: [email protected] (V. Margulis). *

http://dx.doi.org/10.1016/j.urolonc.2014.09.017 1078-1439/r 2014 Elsevier Inc. All rights reserved.

with advanced disease will develop disease progression following surgery [2,4]. Despite available advancements in the treatment of UTUC, 5-year cancer-specific survival (CSS) rates for patients with advanced UTUC continue to be poor, ranging from 12.2% to 74.7% [2]. Current pathological prognostic factors such as stage, grade, tumor location, and lymphovascular invasion (LVI) provide important prognostic information but are not reliably ascertainable before extirpative surgery and fail to capture individual biological variability of tumors [5].

18.e22

B.A. Gayed et al. / Urologic Oncology: Seminars and Original Investigations 33 (2015) 18.e21–18.e26

Over the past few years, the importance of elucidating molecular markers that have the ability to probe the biological heterogeneity of tumors has gained tremendous momentum. Several studies by Shariat and Lotan have described the utility of a panel of cell cycle regulators (p53, p21, p27, and cyclin E) and proliferative molecular markers (Ki67) in accurately predicting disease-free survival and CSS in patients with urothelial bladder cancer [6–9]. Furthermore, a similar panel of tissue biomarkers obtained from transurethral resection specimens was shown to correlate with non–organ-confined disease at the time of radical cystectomy [10]. One of the main problems in appropriately counseling patients with UTUC is the inadequacy of clinical staging owing to small amounts of tissue obtained at diagnosis and the inability to completely resect the tumor endoscopically. As such, useful pathological features such as depth of penetration, concomitant carcinoma in situ (CIS), necrosis, and presence of LVI are difficult to assess before extirpative surgery. Consequently, appropriate selection of patients for multimodal therapy is difficult until after surgical extirpation. In this pilot study, we sought to evaluate the feasibility of obtaining an interpretable biomarker panel in the ureteroscopic biopsy specimens of patients with UTUC. In addition, correlation between pathological features and biopsy marker panel prognostic score (PS) was assessed. 2. Material and methods 2.1. Patient population In total, 17 patients were identified between January 2006 to February 2012 who had an available biopsy as well as who also underwent subsequent open or laparoscopic radical nephroureterectomy (RNU) or segmental ureterectomy (SU) with or without regional lymphadenectomy for clinically localized biopsy-proven high-grade UTUC of the renal pelvis or ureter. Of those 17 patients, 15 (88%) had sufficient tissue to undergo immunohistochemical staining. Recurrences in the bladder were coded as second primaries and not as local or distant recurrences. Biopsy specimens were obtained using endoscopic techniques. Patient and tumor characteristics were recorded into an institutional review board–approved database; primary tumors and biopsy specimens were prospectively evaluated for immunohistochemical expression of a panel of cell cycle and proliferative markers such as p21, p27, p53, cyclin E, and Ki67/pRb. In general, our pathologists require at least enough biopsy tissue to view tumor occupying at least 5 cells per high power field. 2.2. Immunohistochemistry and scoring Immunohistochemical staining from paraffin-embedded tumor blocks were performed for cyclin E1, Tp53, CDKN1A,

Fig. 1. Representative immunohistochemical assessment of endoscopic biopsies and surgical specimens from patients with UTUC utilizing cell cycle and proliferative biomarkers.

p27, and MKI67 employing the Dako Autostainer (Dako North America Inc, Carpinteria, CA, USA). The staining and scoring protocols for all antibodies were previously described in detail [9,11,12]. Briefly, specimens of normal urothelium from patients undergoing RNU/SU served as internal controls. Brightfield microscopy imaging along with an advanced color detection software (Automated Cellular Imaging System, Clarient, CA, USA) was used for semiquantitative scoring. Staining intensity and percentage of positive nuclei per area measurements were evaluated using 10 random spots within each specimen. Staining was done on tissue from the renal biopsy specimen performed as well as the RNU/SU specimens. Nuclear Tp53 immunoreactivity was altered when samples revealed 410% nuclear reactivity. CDKN1A immunoreactivity was altered when samples demonstrated o10% staining. Nuclear p27 and cyclin E were considered altered when samples demonstrated o30% nuclear reactivity. MKI67 immunoreactivity was altered when samples showed 420% nuclear reactivity (Fig. 1). Based on the number of altered biomarkers, we used a previously established PS such that altered expression of 0 to 2 biomarkers was considered favorable PS, whereas an alteration of at least 3 biomarkers signified an unfavorable PS [6,13].

3. Results Clinicopathological data are shown in Table 1. The median age of the patients was 68 years (range: 53–82 y) with 87% of the cohort being males. Median time to recurrence was 17 months (1–58), and median follow-up was 20 months (range: 3–58).

B.A. Gayed et al. / Urologic Oncology: Seminars and Original Investigations 33 (2015) 18.e21–18.e26

18.e23

patients, respectively. KI67 was concordant in both the biopsy and RNU/SU panel in 12/15 (80%) patients (Fig. 2).

Table 1 Clinicopathological features of the study cohort No. (%) Median age at surgery

68 (range: 53–82)

Sex Male Female

13/15 (87) 2/15 (13)

BMI 430 o30

5/15 (33.3) 10/15 (66.6)

pT category pTa pT1 PT2 pT3

6 2 1 6

pN category pNx pN0 pN1

6 (40) 7 (46.7) 2 (13.3)

(40.0) (13.3) (6.7) (40.0)

Necrosis No Yes

13 (86.7) 2 (13.3)

LVI No Yes

10 (66.7) 5 (33.3)

CIS No Yes

11 (73.3) 4 (26.7)

BMI ¼ body mass index.

Of the 15 specimens, 9 (60%) tumors were organ confined (T r 2 and N0) and all were high grade. At the time of analysis, 6/15 (40%) patients had recurred and 4/15 (26.6%) had died of their disease. 3.1. Correlation of individual markers between biopsy and RNU/SU Of the 15 patients, 4 (27%), 7 (46.6%), 3 (20%), and 1 (6.7%) had 1, 2, 3, and 5 markers altered on biopsy marker profiling, respectively. The mean number of altered markers in the biopsy group was 2.1. The most frequently altered marker in the biopsy tissue group was Ki67 with alterations in 13/15 (87.7%) patients. Of the 15 patients, 2 (13.3%), 7 (46.7%), 5 (33.3%), and 1 (6.7%) individuals had 1, 2, 3, and 4 markers altered on RNU/SU marker profiling, respectively. No patient had all 5 markers altered. The mean number of altered markers in the RNU/SU specimen was 2.3. The most frequently altered marker in the RNU/SU was also Ki67 (14/15; 93.3%). p21 was concordant in both the biopsy and RNU/SU panel in 11/15 (73%) patients. p27 was concordant in both the biopsy and RNU/SU specimen in only 5/15 (33%) patients. p53 and cyclin E were concordant in both the biopsy and RNU/SU in 11/15 (73%) and 9/15 (60%)

3.2. Correlation of PS between biopsy and RNU/SU An overall concordance rate of 60% (9/15) was seen between biopsy and RNU/SU PS. Overall, 11/15 (73%) of the cohort had favorable biopsy PS, whereas 9/15 (60%) of the cohort had favorable RNU/ SU PS. Of the 11 patients with a favorable biopsy PS, 7 (64%) also had a favorable RNU/SU PS. Moreover, 4/15 (26.7%) of the cohort had an unfavorable biopsy PS, whereas 6/15 (40%) of the cohort had an unfavorable RNU/SU PS. Of the 4 patients with an unfavorable biopsy PS, 2 (50%) also had an unfavorable RNU/SU PS. 3.3. Association of biopsy PS and pathological features Those patients with favorable biopsy biomarker PS were less likely to display adverse pathological features. Of 11 patients with favorable biopsy PS, 7 (63.6%) had organ confined disease, and 9 (81.8%) had no evidence of CIS in the final specimen. Additionally, 10/11 (91%) had no evidence of necrosis and 7/11 (64%) had no evidence of LVI on final pathological evaluation (Table 2). 4. Discussion In this study, we evaluated the feasibility of obtaining an interpretable biomarker panel consisting of cell cycle regulators and proliferative markers on biopsy specimens of patients with UTUC. Our pilot study shows that with current endoscopic tools, molecular marker profiling of ureteroscopic biopsy specimens of patients with UTUC is achievable. We were able to obtain an interpretable biomarker panel in 15/17 (88%) patients. Additionally, our results show a strong correlation of 4/5 individual markers between biopsy and RNU/SU. Altered p21, p53, cyclin E, and Ki67 all showed correlations of greater than 60% between biopsy and RNU specimen. In our cohort, 11/15 (73%) patients had favorable biopsy PS. Of the 11 patients with a favorable biopsy PS, 7 (64%) also had a favorable RNU/SU PS. Although evaluating oncological outcomes was not feasible owing to low numbers in this pilot study, it is interesting that of the 7 patients who had both a favorable biopsy PS and RNU/ SU PS, 6 (86%) were recurrence free at the time of analysis. We had a 60% concordance rate between our biopsy biomarker panel PS and our RNU/SU biomarker panel PS. Although 6 of the biopsy biomarker PS results did not correlate with the final RNU/SU biomarker PS, we believe this could be owing to technical limitations of evaluating small amounts of tissue or inherent tumor heterogeneity or both.

18.e24

B.A. Gayed et al. / Urologic Oncology: Seminars and Original Investigations 33 (2015) 18.e21–18.e26 Table 2 Relationship of biopsy biomarker profiles with final pathological characteristics of patients treated with extirpative surgery for UTUC Biopsy No. of favorable PS (%) Organ-confined disease Yes (rpT2, N0) No (ZT3, N1)

7 (63.6) 4 (36.4)

2 (50) 2 (50)

LVI No Yes

7 (63.6) 4 (36.4)

3 (75) 1 (25)

CIS No Yes

9 (81.8) 2 (18.2)

2 (50) 2 (50)

10 (90.9) 1 (9.1)

3 (75) 1 (25)

Necrosis No Yes

Fig. 2. Correlation of individual marker expression between endoscopic biopsy and corresponding final pathological specimen.

No. of unfavorable PS (%)

Our analysis of pathological features and biopsy PS revealed that those individuals with a favorable biopsy PS were less likely to harbor adverse pathological features. Individuals with favorable biopsy PS were more likely to have organ-confined disease than those with an unfavorable biopsy PS (63.6% vs. 50%). Furthermore, individuals were less likely to have evidence of necrosis, LVI, and CIS on final specimen (Table 2). Thus, an interpretable biopsy marker profile may be useful in selecting those patients who are more likely to harbor adverse pathological features on final specimen and devising individualized treatment regimens. The gold standard treatment for UTUC continues to be RNU with excision of bladder cuff. However, a substantial portion of these patients will develop systemic progression following RNU [2]. In a large multi-institutional study consisting of 1,363 individuals with UTUC, Margulis et al. [2] reported that nearly 30% of the cohort developed a recurrence at 3-year follow-up after RNU. Despite an accumulating body of evidence supporting application of platinum-based systemic chemotherapy in individuals with UTUC, only a minute portion of individuals with advanced UTUC receive systemic chemotherapy as an adjunct to local treatment. In a recent study evaluating the utilization of systemic therapy and oncologic outcomes after RNU for UTUC over the past 3 decades, Adibi et al. [14] showed that neoadjuvant and adjuvant systemic chemotherapy were used in 47 (3.2%) and 171 (11.7%) patients, respectively. Most concerning, however, was the lack of improvement in oncologic outcomes over the past 3 decades [14]. Neoadjuvant chemotherapy confers several theoretical benefits, notably, enhanced patient tolerability preceding RNU and treatment of micrometastatic disease [15]. However, several challenges, including patient selection and lack of final tissue specimen to help guide treatment

B.A. Gayed et al. / Urologic Oncology: Seminars and Original Investigations 33 (2015) 18.e21–18.e26

regimens, hamper thoughtful implementation of a neoadjuvant paradigm. Hence, establishment of molecular markers that may risk stratify patients for whom multimodality therapy may improve oncological outcomes is imperative to improving oncological outcomes. Although molecular markers have been studied extensively in patients with urothelial bladder cancer, the role of molecular markers in UTUC remains to be defined. Bagrodia et al. studied the use of a biomarker panel consisting of cell cycle regulators and a proliferative marker in individuals with UTUC treated with RNU. Individuals with an unfavorable tissue biomarker score (42 biomarkers altered) showed inferior CSS as well as a strong correlation with adverse pathological features [11]. Although the results are quite intriguing regarding the use of a panel molecular markers to predict oncological outcomes, improvements in oncological outcomes in individuals with UTUC center around identifying those individuals more likely at risk for progression or recurrence before extirpative surgery. Interestingly, our results revealed a strong correlation of 4 individual markers between biopsy and RNU/SU. Ki67 was the most altered marker (80%) on both biopsy and final RNU specimens. Ki67 has been implicated in several malignancies and has been shown to be prognostic marker of worse oncological outcomes as well aggressive pathological features [12,16]. Ki67 has also been shown to play critical role in urothelial bladder cancer [17]. In patients with organconfined bladder cancer, Ki67 overexpression was an independent predictor of disease-free survival and CSS [17]. Recently, Krabbe et al. published their results evaluating Ki67 in a cohort of patients with UTUC. Elevated expression of Ki67 was associated with worse recurrencefree survival as well as adverse pathological features [9]. This study has several important limitations, which center on the retrospective study design and potential associated bias. Additionally, the total number of patients was less, and larger studies may be needed to evaluate the statistical significance of this biopsy marker panel. We acknowledge that selecting tissue specimens for immunohistochemistry has a lot of variability; however, at our institution, our pathologists typically select tissue areas with the highest grade or most invasive areas. Additionally, driven by the technical difficulties arising from obtaining substantial tissue from ureteroscopic biopsies, we typically employ new-generation large-caliber biopsy forceps, baskets, and laser tissue excision to obtain maximal amount of tissue for pathological evaluation. Lastly, we believe the discordance between biopsy biomarkers and RNU biomarkers could have been owing to tumor heterogeneity and technical limitations of immunohistochemical assessment of biopsy tissue. As we continue to refine our techniques for immunohistochemical staining as well as gain further experience with handling and evaluation of biopsy tissue, we expect decreased discordance rates.

18.e25

5. Conclusion In this study, we present the first known analysis evaluating the feasibility of obtaining a molecular biopsy biomarker panel. With current endoscopic tools, obtaining an interpretable biomarker profile of ureteroscopic biopsy specimens is achievable and may provide significant prognostic information. With a concordance rate of 60% with RNU/SU biomarker panel, meaningful biopsy biomarker profiling could serve as a powerful tool for individualizing treatment regimens and augmenting current predictive variables. Further studies are needed to evaluate its clinical applicability.

References [1] Raman JD, Messer J, Sielatycki JA, Hollenbeak CS. Incidence and survival of patients with carcinoma of the ureter and renal pelvis in the USA, 1973–2005. BJU Int 2011;107:1059–64:[PubMed PMID: 20825397. Epub 2010/09/10. Eng]. [2] Margulis V, Shariat SF, Matin SF, Kamat AM, Zigeuner R, Kikuchi E, et al. Outcomes of radical nephroureterectomy: a series from the Upper Tract Urothelial Carcinoma Collaboration. Cancer 2009;115:1224–33:[PubMed PMID: 19156917. Epub 2009/01/22. Eng]. [3] Hall MC, Womack S, Sagalowsky AI, Carmody T, Erickstad MD, Roehrborn CG. Prognostic factors, recurrence, and survival in transitional cell carcinoma of the upper urinary tract: a 30-year experience in 252 patients. Urology 1998;52:594–601:[PubMed PMID: 9763077. Epub 1998/10/08. Eng]. [4] Ozsahin M, Zouhair A, Villa S, Storme G, Chauvet B, Taussky D, et al. Prognostic factors in urothelial renal pelvis and ureter tumours: a multicentre Rare Cancer Network study. Eur J Cancer 1999;35:738– 43:[PubMed PMID: 10505034. Epub 1999/10/03. Eng]. [5] Verhoest G, Shariat SF, Chromecki TF, Raman JD, Margulis V, Novara G, et al. Predictive factors of recurrence and survival of upper tract urothelial carcinomas. World J Urol 2011;29:495–501:[PubMed PMID: 21681525. Pubmed Central PMCID: 3526475. Epub 2011/06/ 18. Eng]. [6] Shariat SF, Karakiewicz PI, Ashfaq R, Lerner SP, Palapattu GS, Cote RJ, et al. Multiple biomarkers improve prediction of bladder cancer recurrence and mortality in patients undergoing cystectomy. Cancer 2008;112:315–25:[PubMed PMID: 18008359. Epub 2007/11/17. Eng]. [7] Shariat SF, Chade DC, Karakiewicz PI, Ashfaq R, Isbarn H, Fradet Y, et al. Combination of multiple molecular markers can improve prognostication in patients with locally advanced and lymph node positive bladder cancer. J Urol 2010;183:68–75:[PubMed PMID: 19913255. Epub 2009/11/17. Eng]. [8] Shariat SF, Ashfaq R, Sagalowsky AI, Lotan Y. Predictive value of cell cycle biomarkers in nonmuscle invasive bladder transitional cell carcinoma. J Urol 2007;177:481–7:[discussion 7. PubMed PMID: 17222615. Epub 2007/01/16. Eng]. [9] Krabbe LM, Bagrodia A, Lotan Y, Gayed BA, Darwish OM, Youssef RF, et al. Ki-67 is an independent predictor of oncological outcomes in patients with high-grade upper tract urothelial carcinoma: results of a prospective analysis and a review of the literature. J Urol 2013; 191(1):28–34:[PubMed PMID: 23871758. Epub 2013/07/23. Eng]. [10] Shariat SF, Passoni N, Bagrodia A, Rachakonda V, Xylinas E, Robinson B, et al. Prospective evaluation of a preoperative biomarker panel for prediction of upstaging at radical cystectomy. BJU Int 2014;113(1):70–6.

18.e26

B.A. Gayed et al. / Urologic Oncology: Seminars and Original Investigations 33 (2015) 18.e21–18.e26

[11] Bagrodia A, Youssef RF, Kapur P, Darwish OM, Cannon C, Belsante MJ, et al. Prospective evaluation of molecular markers for the staging and prognosis of upper tract urothelial carcinoma. Eur Urol 2012;62: e27–9:[PubMed PMID: 22552216. Epub 2012/05/04. Eng]. [12] Heslin MJ, Cordon-Cardo C, Lewis JJ, Woodruff JM, Brennan MF. Ki-67 detected by MIB-1 predicts distant metastasis and tumor mortality in primary, high grade extremity soft tissue sarcoma. Cancer 1998;83:490–7:[PubMed PMID: 9690542. Epub 1998/08/05. Eng]. [13] Lotan Y, Bagrodia A, Passoni N, et al. Prospective evaluation of a molecular marker panel for prediction of recurrence and cancer-specific survival after radical cystectomy. Eur Urol 2013; 6(3):465–71. [14] Adibi M, Youssef R, Shariat SF, Lotan Y, Wood CG, Sagalowsky AI, et al. Oncological outcomes after radical nephroureterectomy for upper

tract urothelial carcinoma: comparison over the three decades. Int J Urol 2012;19:1060–6:[PubMed PMID: 22882743. Epub 2012/08/14. Eng]. [15] Audenet F, Yates DR, Cussenot O, Roupret M. The role of chemotherapy in the treatment of urothelial cell carcinoma of the upper urinary tract (UUT-UCC). Urol Oncol 2010;31(4):407–13:[PubMed PMID: 20884249. Epub 2010/10/05. Eng]. [16] Jonat W, Arnold N. Is the Ki-67 labelling index ready for clinical use? Ann Oncol 2011;22:500–2:[PubMed PMID: 21343384. Epub 2011/02/24. Eng]. [17] Margulis V, Shariat SF, Ashfaq R, Sagalowsky AI, Lotan Y. Ki-67 is an independent predictor of bladder cancer outcome in patients treated with radical cystectomy for organ-confined disease. Clin Cancer Res 2006;12:7369–73:[PubMed PMID: 17189409. Epub 2006/12/26. Eng].