- Email: [email protected]
Evaluation of individual and cumulative sites of extrarenal tumor invasion in pT3a clear cell renal cell carcinoma
ARTICLE IN PRESS
Urologic Oncology: Seminars and Original Investigations 000 (2019) 1−6
Clinical-Kidney cancer
Evaluation of individual and cumulative sites of extrarenal tumor invasion in pT3a clear cell renal cell carcinoma Trevor A. Flood, M.D.a,*, Kevin Hogan, M.D.a, Luke T. Lavallee, M.D.b, Rodney H. Breau, M.D.b, Chris Morash, M.D.b, Eric C. Belanger, M.D.a, Nicola Schieda, M.D.c a
Department of Pathology and Laboratory Medicine, The Ottawa Hospital, Ottawa, Ontario, Canada b Division of Urology, Department of Surgery, The Ottawa Hospital, Ottawa, Canada c Department of Medical Imaging, The Ottawa Hospital, Ottawa, Canada Received 9 May 2019; received in revised form 19 August 2019; accepted 28 September 2019
Abstract Introduction: The Tumor-Node-Metastasis classification of renal cell carcinoma (RCC) for pT3a tumors includes sinus fat invasion (SFI), perinephric fat invasion (PFI), renal vein invasion (RVI), and/or pelvicaliceal system invasion (PSI). The purpose of this study was to determine the association between these patterns of invasion (assessed individually and cumulatively) with the development of metastases and cancer-specific mortality (CSM). Materials and methods: We identified 160 patients who underwent radical nephrectomy for pT3a clear cell RCC between 2011 and 2017. The association between individual patterns of invasion and metastases and cancer-specific survival were evaluated with multivariate logistic regression. Cox Hazard proportion ratios and Kaplan-Meier survival curves were generated for patterns of invasion (assessed individually and cumulatively). Results: The number of individual invasive patterns was as follows: 97/160 (61%) presented with RVI, 91/160 with SFI (57%), 62/160 with PFI (39%), and 24/160 (15%) with PSI. At multivariate analysis, both PFI and RVI were associated with metastases (P < 0.001 and 0.028, respectively). PFI (hazard ratio [HR] 4.12, 95% confidence interval [CI] 2.14−7.92; P < 0.001), RVI (HR 2.44, 95% CI 1.18−5.01; P = 0.015), SFI (HR 2.13, 95% CI 1.05−4.34; P = 0.036) had higher CSM, while PSI (HR 1.43, 95% CI 0.65-3.16; p = 0.38) did not show increased CSM. Furthermore, cumulative analysis showed that multiple invasive patterns resulted in worse CSM (p < 0.001). Conclusions: In our study, PFI was associated with the most aggressive behavior while PSI was the most indolent. Furthermore, the presence of more than one pattern of invasion was associated with worse CSM. These results indicate that reporting of the individual location and cumulative amount of pT3a patterns of invasion in clear cell RCC is clinically relevant. Ó 2019 Elsevier Inc. All rights reserved.
Keywords: Carcinoma; Kidney; Metastasis; Renal cancer; Clear cell renal cell carcinoma; Radical nephrectomy
1. Introduction The Tumor-Node-Metastasis (TNM) classification of renal cell carcinoma (RCC) is used to standardize disease extent among patients. The categories within the TNM system aim to group tumors together that share similar prognoses and behaviors. The pT3a subgroup of RCCs represents
Funding: This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. *Corresponding author. Tel.: +1 613-737-8899 ext 72179; fax: (613) 737-8712. E-mail address: [email protected] (T.A. Flood). https://doi.org/10.1016/j.urolonc.2019.09.030 1078-1439/Ó 2019 Elsevier Inc. All rights reserved.
a high-risk collection of tumors that have extended locally beyond the confines of the renal parenchyma. The previously used seventh edition of the American Joint Committee on Cancer (AJCC) staging system classified pT3a tumors as lesions that show perirenal fat invasion (PFI), sinus fat invasion (SFI), or renal vein invasion (RVI) [1]. The recently issued eighth edition of the AJCC TNM staging system added pelvicaliceal system invasion (PSI) as a fourth pattern of RCC invasion that defines the pT3a subcategory [2]. Despite several modifications to the TNM staging system for RCC, the significance of individual patterns of pT3a invasion (PFI, SFI, RVI, and more recently PSI)
ARTICLE IN PRESS 2
T.A. Flood et al. / Urologic Oncology: Seminars and Original Investigations 00 (2019) 1−6
remains uncertain. While some studies have reported that tumors displaying specific patterns of pT3a disease are not associated with an adverse prognosis, others have shown certain patterns to be associated with more aggressive behavior. Furthermore, even though PSI was added to the most recent TNM staging, conflicting results regarding the prognosis of patients with RCC and PSI have been reported [3−8]. Moreover, investigators have also shown that the presence of more than 1 pattern pT3a may impart a worse prognosis for the patient than when a single pattern is detected [9−11]. To our knowledge the incremental effect of PSI added into these cumulative predictive models has not been evaluated. The purpose of this study was to determine if the different patterns of pT3a invasion included in the AJCC eighth edition (PFI, SFI, RVI, and PSI), assessed both separately and cumulatively, were associated with metastases and cancer-specific mortality (CSM) in patients with pT3a clear cell RCC (CCRCC).
Fig. 1. Histologic sections demonstrating; (A) Example of renal vein invasion (RVI) demonstrating tumor thrombus (arrowheads) in muscular containing branch of the renal vein; (B) Sinus fat invasion (SFI) showing tumor cells (arrowheads) in direct contact with fat cells of the renal sinus; (C) Perirenal fat invasion (PFI) exhibited by malignant cells infiltrating into perirenal fat cells (arrowheads) beyond the renal capsule (arrow); (D) Pelvicaliceal system invasion (PSI) characterized by tumor cells involving the lumen of the collecting system (arrowheads) beyond the urothelium (arrow).
2. Methods 2.1. Patients With institutional review board approval, we conducted a consecutive search of our pathology database for patients with pT3a CCRCC who underwent radical nephrectomy between January 2011 and December 2017 at a single institution. We only included radical nephrectomies in our study so that we could confidently examine the pelvicaliceal system for PSI. Only CCRCCs were studied because this is the most common RCC and we did not want to confound our results by including other histologic subtypes that are associated with unique behaviors. The pathological stage was assigned according to the AJCC eighth edition TNM staging system by dedicated genitourinary pathologists at a single tertiary-care referral center for RCC [2]. Information recorded from patient electronic medical records included: PFI, SFI, RVI, PSI, patient age, tumor size, and World Health Organization/International Society of Urological Pathology (WHO/ISUP) nucleolar grade. 2.2. Specimen preparation and histopathologic analysis All radical nephrectomies performed by any surgical approach at the same tertiary care referral center were included. Specimens were fixed in 10% neutral buffered formalin for a minimum of 24 hours. Kidneys were bivalved using a probe inserted into the collecting system as a guide. The renal rein and its branches were opened and examined for foci of RVI. At least 1 section of neoplastic tissue was submitted for every centimeter of the largest dimension of tumor. Additional sections of the renal sinus were submitted when SFI was equivocal. Tissues were paraffin embedded and 4 mm thick sections were cut and stained with hematoxylin and eosin. Subspecialty trained
pathologists provided diagnoses for all radical nephrectomy specimens as part of routine clinical care. Fig. 1A shows an example of RVI, which was defined as identification of tumor into the renal vein or its segmental branches [12]. SFI was designated as tumor cells being in direct contact with fat or in loose connective tissue beyond the renal parenchyma, as well as involvement of endothelial lined spaces within the sinus (Fig. 1B) [12]. PFI was defined as tumor touching fat or extending as irregular tongues into the perinephric tissue, with or without desmoplasia (Fig. 1C) [12]. PSI was characterized by tumor involving the renal calyceal or pelvic submucosa/mucosa with or without luminal involvement (Fig. 1D) [3]. 2.3. Outcomes Outcomes in this study included: distant metastatic disease and CSM. Distant metastases were considered to be present when the patient developed a new lesion in a target organ on CT or MRI studies which were not present at baseline imaging or that grew greater than 20% compared to a baseline imaging study performed within 6 months [13] or when biopsy was performed with histological confirmation of metastatic disease. Metastases and death were determined by chart review and the maximum duration of follow-up was calculated from the date of surgery to the date of last follow-up or for the date of death from our institutional medical records. 2.4. Statistical analysis Skew tests were performed for quantitative data. For normally distributed variables, the mean § standard deviation was reported. For skewed data, the median and interquartile range (IQR) was reported. Multivariate logistic
ARTICLE IN PRESS T.A. Flood et al. / Urologic Oncology: Seminars and Original Investigations 00 (2019) 1−6
regression was performed comparing patient and tumor variables to the outcomes of distant metastases and CSM. Kaplan Meier curves were constructed for each pattern of pT3a invasion and for the cumulative number of pT3a patterns in the same patient. Cox hazard proportional ratios were calculated that compared each of the 4 patterns of pT3a invasion for both CSM and all cause mortality according to previously published guidelines [14]. In our Cox Hazard-Proportional model we studied 1 predictive variable for every 10 events [15]. Statistical analyses were performed using STATA version 13.0 (Statacorp, College Station, TX). A P value of <0.05 was considered statistically significant. 3. Results The search of our database identified 160 consecutive radical nephrectomy specimens with pT3a CCRCC. No patients received neoadjuvant therapy between diagnosis and surgery. Mean § standard deviation patient age was 64.66 § 10.36 years. The median tumor size was 7.5 (IQR 5.7−9.5) cm. The cohort had a median postoperative follow-up of 33.8 (IQR 20.6−55.4) months. Forty (40/160) patients died from CCRCC with a median of 25.4 (IQR 7.5 −40.0) months. Four patients died from causes other than CCRCC. Distant metastases were detected in seventy (70/ 160) patients based on positive imaging findings. These findings were confirmed in 31.4% (22/70) of patients by histologic sampling of the metastasis. Tables 1 and 2 summarize the results comparing the clinicopathologic features to outcomes. The number of individual invasive patterns was as follows: 97/160 (61%) presented with RVI, 91/160 with SFI (57%), 62/160 with PFI (39%), and 24/160 (15%) with PSI. After multivariate analysis, WHO/ISUP nucleolar grade (P < 0.001), tumor size (P = 0.008), RVI (P = 0.028), PFI (P < 0.001), and cumulative number of invasive patterns (P < 0.001) were significantly associated with distant metastases. WHO/ ISUP nucleolar grade (P < 0.001), PFI (P = 0.006), and
3
cumulative number of invasive patterns (P < 0.001) were associated with CSM. Kaplan-Meier curves for individual patterns of pT3a spread are provided in Fig. 2. Among the 4 patterns of pT3a spread, RVI (P = 0.023), SFI (P = 0.045), and PFI (P < 0.001); were associated with increased risk of CSM; whereas, PSI (P = 0.24) was not. Hazard ratios comparing CSM for the 160 patients and each individual invasive patterns were: 2.44 (95% CI, 1.18−5.01; P = 0.015) for RVI, 2.13 (95% CI, 1.05−4.34, P = 0.036) for SFI, 4.12 (95% CI, 2.14−7.92; P < 0.001) for PFI, and 1.43 (95% CI, 0.65 −3.16; P = 0.38) for PSI (Table 3). The hazard ratios for all-cause mortality for the 164 patients and the individual invasive patterns were: 2.25 (95% CI, 1.17-4.39; P = 0.017) for RVI, 2.43 (95% CI, 1.22−4.82; P = 0.011) for SFI, 3.02 (95% CI, 1.77−5.13; P < 0.001) for PFI, and 1.35 (95% CI, 0.62-2.94; P = 0.46) for PSI. The cumulative number of patterns of tumor invasion was as follows: 50.0% (80/160) showed 1 invasive pattern, 31.9% (51/160) had 2 patterns, 14.4% (23/160) had 3 patterns, and 3.8% (6/160) had all 4 invasive patterns. CSM was observed in: 10% (8/80) of patients with 1 invasive pattern, 27.4% (14/51) with 2 invasive patterns, 65.2% (15/23) with 3 patterns, and 50.0% (3/6) among patients who showed all 4 pT3a invasive patterns. Kaplan-Meier curves constructed for the cumulative number of patterns of pT3a invasion showed a worse CSM in patients whose tumor showed increasing number of invasive patterns (P < 0.001), Fig. 3. 4. Discussion This study evaluated individual patterns of pT3a CCRCC invasion and the cumulative effect of multiple invasive patterns with respect to metastatic disease and CSM. Of the 4 patterns of invasion included in the AJCC eighth editions pT3a subgroup, PFI was the most aggressive while PSI was the least aggressive. Our results also show that patients with an increasing number of extrarenal
Table 1 Clinicopathologic features of cohort of patients with pT3a CCRCC and their association with metastases and CSM after multivariate analysis. Variable
N (%)
Patients with metastases (%)
P value
Patients who died
P value
Age Gender Female Male WHO/ISUP nucleolar grade:a 2/4 3/4 4/4 Tumor size Sarcomatoid differentiation Necrosis Cumulative number of invasive patterns
−
−
0.028
−
0.066
43/160 (27) 117/160 (73)
20/70 (29) 50/70 (71)
0.58
13/40 (32) 27/40 (68)
0.69
29/160 (18) 88/160 (55) 43/160 (27) − 24/160 (15) 84/160 (53) −
3/70 (4) 36/70 (51) 31/70 (45) − 19/70 (27) 46/70 (66) −
<0.001
1/40 (2) 15/40 (38) 24/40 (60) − 17/40 (43) 30/40 (75) −
<0.001
a
No patients were graded 1/4.
0.008 0.65 0.42 <0.001
0.095 <0.069 0.76 <0.001
ARTICLE IN PRESS T.A. Flood et al. / Urologic Oncology: Seminars and Original Investigations 00 (2019) 1−6
4
Table 2 Individual invasive patterns and their association with metastases and CSM after multivariate analysis. RVI Metastatic disease Yes (N = 70) 73% (51/70) No (N = 90) 51% (46/90) RVI Cancer specific mortality (CSM) Yes (N = 40) 75% (30/40) No (N = 120) 56% (67/120)
P value
SFI
P value
PFI
P value
PSI
P value
0.009
61% (43/70) 53% (48/90)
0.22
61% (43/70) 21% (19/90)
<0.001
17% (12/70) 13% (12/90)
0.80
P value
SFI
P value
PFI
P value
PSI
P value
0.12
73% (29/40) 52% (62/120)
0.060
65% (26/40) 30% (36/120)
0.006
20% (8/40) 13% (16/120)
0.94
Fig. 2. Kaplan-Meier curves for individual patterns of pT3a spread. Solid lines represent presence of specific indicated pattern of invasion compared to tumors without indicated invasive pattern (dashed line): (A) Renal vein invasion (RVI); (B) Sinus fat invasion (SFI); (C) Perirenal fat invasion (PFI); and (D) Pelvicaliceal system invasion (PSI).
Table 3 Patterns of invasion and association with CSM after Cox regression analysis. pT3a Pattern RVI SFI PFI CSI a
Hazard ratioa 2.44 2.13 4.12 1.43
95% confidence interval (CI)
P value
1.18-5.01 1.05-4.34 2.14-7.92 0.65-3.16
0.015 0.036 <0.001 0.38
Cox proportion hazard ratios includes the four patterns of pT3a studied.
(pT3a) sites of spread have higher CSM. These findings indicate that identifying and reporting of each pattern of pT3a CCRCC invasion at time of pathologic analysis in radical nephrectomy specimens is important and may be of prognostic significance. To our knowledge this is the first study to examine the prognostic impact of the 4 different patterns of extrarenal invasion that define pT3a tumors in the newly accepted AJCC eighth edition. Our results show that a patient whose tumor exhibits additional sites of extrarenal invasion has a
ARTICLE IN PRESS T.A. Flood et al. / Urologic Oncology: Seminars and Original Investigations 00 (2019) 1−6
Fig. 3. Kaplan-Meier curves for the cumulative number of patterns of pT3a invasion detected in the same patient.
significantly worse CSM than when 1 pattern is present. These outcomes highlight the importance of identifying all patterns of invasion with an emphasis conveyed to surgeons regarding the significance associated with their cumulative effect. Both Kresowick et al. and Bedke et al. previously proposed similar recommendations after they independently showed that tumors demonstrating more than 1 invasive pattern (specifically, PFI and SFI in combination) had a worse prognosis than when either pattern was observed alone [9,11]. Similarly, Shah et al. showed that the presence of the 3 patterns of invasion defined by the AJCC seventh edition (PFI, SFI, and RVI) were associated with a higher risk of disease progression and cancer-related death after nephrectomy [10]. Our study is the first to examine the recently included PFI as a feature of pT3a tumors and has consistent findings with these previous reports. The reporting of the amount of patterns in nephrectomy specimens could allow for better prediction of tumor behavior and lead to improved planning of postoperative therapy, surveillance and determining candidacy for clinical trials. PSI was recently included as a feature of pT3a tumors in the new AJCC eighth edition TNM Staging System [2], despite the prognostic value of this feature being uncertain. Some authors have argued that PSI should not be included in the TNM Staging System [6,8,17,19] based on the results of several studies that have shown that this feature is associated with indolent behavior [6,8,17−19]. Our study shares similar findings in that we demonstrated that PSI was not significantly associated with metastatic disease or CSM. Interestingly, other author groups have shown that PSI is associated with a worse prognosis. The majority of these studies examined PSI as a variable across several different pT Stages and only 2 studies have examined the individual prognostic significance of PSI in pT3 tumors [4,16]. In the aforementioned studies by Anderson et al. and Chen et al., PSI was independently associated with worse cancer-specific survival. Unlike Anderson et al. and Chen et al., we restricted our analysis to include pT3a tumors and excluded pT3b and pT3c. Additionally, we limited our evaluation to
5
include only CCRCCs since it is the most common RCC and that other histologic subtypes are associated with unique prognoses. These differences in study design may account for the variation in prognostic significance of PSI observed between research groups. Additional analysis is required to validate the prognostic significance of PSI and to confirm if it should be included within the pT3a subcategory of the TNM Staging System. Aside from PSI, the pT3a group for RCCs includes any tumor demonstrating SFI, RVI, and/or PFI. The exact significance of these other individual patterns of invasion is also debated. The renal sinus is home to an abundance of lymphatics and veins which malignant cells can invade and use to disseminate widely [20]. Some studies have suggested that tumors associated with SFI behave more aggressively when compared to those with PFI [21−23]. Other author’s results demonstrate no significant difference between SFI and PFI when either feature is found in isolation [9,11,24,25]. Different prognoses may also exist between RVI and tumor invading into adipose tissue. Park et al. showed that patients with RVI were associated with a significant higher tumor recurrence and cancer death compared to those with PFI [26]. However, Novara et al. demonstrated that patients with RVI were actually associated with a better prognosis than tumors that invaded the fat [27]. Other groups have shown no difference between the patterns of invasion with regards to prognosis [10,25]. The Shah et al. study was the only manuscript to restrict their cohort to pT3a CCRCC and they found no prognostic significance between the individual patterns of invasion. In our study, PFI imparted the highest risk of CSM and previous studies have also shown that PFI may be a predictor of more aggressive behavior [28,29]. Indeed, our results indicate that PFI is associated with approximately a 4-times increased risk of CSM while the RVI and SFI may be associated with a roughly 2-fold increase. It is evident from these results among the current and previous studies that the individual patterns of invasion in pT3a RCCs are associated with diverse and variable prognoses although heterogeneity in study designs may account for some of these differences. The limitations of this study include the inherent biases associated with a retrospective single institution evaluation. Statistical analysis may have been influenced by small sample sizes, especially with regards to comparing the subgroups of tumor invasion, such as those demonstrating PSI. The smaller sample sizes are partly due to our strict inclusion criteria (limiting our cohort to CCRCC pT3a and radical nephrectomies). However, our choice to only examine radical nephrectomy specimens ensures adequate evaluation of the pelvicalyceal system to assess for PSI, a feature that we felt was fundamentally necessary to sufficiently complete the objectives of the study because of the potential for a false-negative evaluation of PSI at partial nephrectomy. By including only CCRCCs, we eliminated the confounding effect of different behaviors associated with
ARTICLE IN PRESS 6
T.A. Flood et al. / Urologic Oncology: Seminars and Original Investigations 00 (2019) 1−6
various RCC histologic subtypes, although at the cost of introducing bias and reducing our sample size. We feel that these strict inclusion criteria help strengthen our results through the creation of a uniform patient cohort and by ensuring thorough examination of the surgical specimens to assess for all pT3a patterns of invasion. Our low sample size restricted our ability to assess individual combinations of pT3a patterns in models, due to the risk of over-fitting our data, which may be a topic for future studies evaluating pT3a disease in RCC. 5. Conclusions Our results show that among individual pT3a patterns of extrarenal invasion that PFI is the most aggressive, while PSI is the most indolent. Moreover, the presence of a cumulative increase in the amount of pT3a invasive patterns in a patient’s tumor was associated with worse CSM. These results indicate that reporting of both the individual location and cumulative amount of pT3a patterns of CCRCC extrarenal invasion may allow for better risk stratification and follow-up after surgery. References [1] Edge S BD, Compton CC, Fritz AG, Greene FL, Trotti A. American Joint Committee on Cancer, American Cancer Society. AJCC Cancer Staging Manual. 7th ed New York: Springer-Verlag; 2010. [2] Amin MB ES, Greene FL, Byrd DR, Brookland RK, Washington MK, Gershenwald JE, et al. AJCC cancer staging manual. 8th ed Springer; 2017. [3] Bailey GC, Boorjian SA, Ziegelmann MJ, et al. Urinary collecting system invasion is associated with poor survival in patients with clear-cell renal cell carcinoma. BJU Int 2017;119(4):585–90. [4] Chen L, Ma X, Li H, et al. Invasion of the urinary collecting system is an independent prognostic factor in pT3 renal cell carcinoma. Urol Oncol 2016;34(7):293.e11-6. [5] Palapattu GS, Pantuck AJ, Dorey F, et al. Collecting system invasion in renal cell carcinoma: impact on prognosis and future staging strategies. J Urol 2003;170(3):768–72:discussion 72. [6] Schrader AJ, Rustemeier J, Varga Z, et al. Urinary collecting system invasion in renal cell carcinoma: incidence and long-term prognosis. Int J Urol 2009;16(9):718–22. [7] Verhoest G, Avakian R, Bensalah K, et al. Urinary collecting system invasion is an independent prognostic factor of organ confined renal cell carcinoma. J Urol 2009;182(3):854–9. [8] Waalkes S, Merseburger AS, Herrmann TR, et al. Urinary collecting system invasion is no independent prognostic factor in renal cell carcinoma. World J Urol 2010;28(3):283–8. [9] Bedke J, Buse S, Pritsch M, et al. Perinephric and renal sinus fat infiltration in pT3a renal cell carcinoma: possible prognostic differences. BJU Int 2009;103(10):1349–54. [10] Shah PH, Lyon TD, Lohse CM, et al. Prognostic evaluation of perinephric fat, renal sinus fat, and renal vein invasion for patients with pathological stage T3a clear-cell renal cell carcinoma. BJU Int 2019;123(2):270–6.
[11] Kresowik TP, Johnson MT, Joudi FN. Combined renal sinus fat and perinephric fat renal cell carcinoma invasion has a worse prognosis than either alone. J Urol 2010;184(1):48–52. [12] Trpkov K, Grignon DJ, Bonsib SM, et al. Handling and staging of renal cell carcinoma: the International Society of Urological Pathology Consensus (ISUP) conference recommendations. Am J Surg Pathol 2013;37(10):1505–17. [13] Nishino M, Jagannathan JP, Ramaiya NH, et al. Revised RECIST guideline version 1.1: What oncologists want to know and what radiologists need to know. AJR Am J Roentgenol 2010;195(2):281–9. [14] Abd ElHafeez S, Torino C, D’Arrigo G, et al. An overview on standard statistical methods for assessing exposure-outcome link in survival analysis (Part II): the Kaplan-Meier analysis and the Cox regression method. Aging Clin Exp Res 2012;24(3):203–6. [15] Harrell FE Jr., Lee KL, Califf RM, et al. Regression modelling strategies for improved prognostic prediction. Stat Med 1984;3(2):143–52. [16] Anderson CB, Clark PE, Morgan TM, et al. Urinary collecting system invasion is a predictor for overall and disease-specific survival in locally invasive renal cell carcinoma. Urology 2011;78(1):99–104. [17] Terrone C, Cracco C, Guercio S, et al. Prognostic value of the involvement of the urinary collecting system in renal cell carcinoma. Eur Urol 2004;46(4):472–6. [18] Mirza KM, Taxy JB, Antic T. Radical nephrectomy for renal cell carcinoma: its contemporary role related to histologic type, tumor zize, and nodal status: aA retrospective study. Am J Clin Pathol 2016;145 (6):837–42. [19] Uzzo RG, Cherullo E, Myles J, et al. Renal cell carcinoma invading the urinary collecting system: implications for staging. J Urol 2002;167(6):2392–6. [20] Bonsib SM, Gibson D, Mhoon M, et al. Renal sinus involvement in renal cell carcinomas. Am J Surg Pathol 2000;24(3):451–8. [21] Thompson RH, Leibovich BC, Cheville JC, et al. Is renal sinus fat invasion the same as perinephric fat invasion for pT3a renal cell carcinoma? J Urol 2005;174(4 Pt 1):1218–21. [22] Bonsib SM. The renal sinus is the principal invasive pathway: a prospective study of 100 renal cell carcinomas. Am J Surg Pathol 2004;28(12):1594–600. [23] Zhang Z, Yu C, Velet L, et al. The difference in prognosis between renal sinus fat and perinephric fat invasion for pT3a renal cell carcinoma: a meta-analysis. PLoS One 2016;11(2):e0149420. [24] Poon SA, Gonzalez JR, Benson MC, et al. Invasion of renal sinus fat is not an independent predictor of survival in pT3a renal cell carcinoma. BJU Int 2009;103(12):1622–5. [25] Margulis V, Tamboli P, Matin SF, et al. Location of extrarenal tumor extension does not impact survival of patients with pT3a renal cell carcinoma. J Urol 2007;178(5):1878–82. [26] Park M, Shim M, Kim M, et al. Prognostic heterogeneity in T3aN0M0 renal cell carcinoma according to the site of invasion. Urol Oncol 2017;35(7):458.e17-58.e22. [27] Novara G, Ficarra V, Antonelli A, et al. Validation of the 2009 TNM version in a large multi-institutional cohort of patients treated for renal cell carcinoma: are further improvements needed? Eur Urol 2010;58(4):588–95. [28] Brookman-May SD, May M, Wolff I, et al. Evaluation of the prognostic significance of perirenal fat invasion and tumor size in patients with pT1-pT3a localized renal cell carcinoma in a comprehensive multicenter study of the CORONA project. Can we improve prognostic discrimination for patients with stage pT3a tumors? Eur Urol 2015;67(5):943–51. [29] Yoo C, Song C, Hong JH, et al. Prognostic significance of perinephric fat infiltration and tumor size in renal cell carcinoma. J Urol 2008;180 (2):486–91:discussion 91.
Urologic Oncology: Seminars and Original Investigations 000 (2019) 1−6
Clinical-Kidney cancer
Evaluation of individual and cumulative sites of extrarenal tumor invasion in pT3a clear cell renal cell carcinoma Trevor A. Flood, M.D.a,*, Kevin Hogan, M.D.a, Luke T. Lavallee, M.D.b, Rodney H. Breau, M.D.b, Chris Morash, M.D.b, Eric C. Belanger, M.D.a, Nicola Schieda, M.D.c a
Department of Pathology and Laboratory Medicine, The Ottawa Hospital, Ottawa, Ontario, Canada b Division of Urology, Department of Surgery, The Ottawa Hospital, Ottawa, Canada c Department of Medical Imaging, The Ottawa Hospital, Ottawa, Canada Received 9 May 2019; received in revised form 19 August 2019; accepted 28 September 2019
Abstract Introduction: The Tumor-Node-Metastasis classification of renal cell carcinoma (RCC) for pT3a tumors includes sinus fat invasion (SFI), perinephric fat invasion (PFI), renal vein invasion (RVI), and/or pelvicaliceal system invasion (PSI). The purpose of this study was to determine the association between these patterns of invasion (assessed individually and cumulatively) with the development of metastases and cancer-specific mortality (CSM). Materials and methods: We identified 160 patients who underwent radical nephrectomy for pT3a clear cell RCC between 2011 and 2017. The association between individual patterns of invasion and metastases and cancer-specific survival were evaluated with multivariate logistic regression. Cox Hazard proportion ratios and Kaplan-Meier survival curves were generated for patterns of invasion (assessed individually and cumulatively). Results: The number of individual invasive patterns was as follows: 97/160 (61%) presented with RVI, 91/160 with SFI (57%), 62/160 with PFI (39%), and 24/160 (15%) with PSI. At multivariate analysis, both PFI and RVI were associated with metastases (P < 0.001 and 0.028, respectively). PFI (hazard ratio [HR] 4.12, 95% confidence interval [CI] 2.14−7.92; P < 0.001), RVI (HR 2.44, 95% CI 1.18−5.01; P = 0.015), SFI (HR 2.13, 95% CI 1.05−4.34; P = 0.036) had higher CSM, while PSI (HR 1.43, 95% CI 0.65-3.16; p = 0.38) did not show increased CSM. Furthermore, cumulative analysis showed that multiple invasive patterns resulted in worse CSM (p < 0.001). Conclusions: In our study, PFI was associated with the most aggressive behavior while PSI was the most indolent. Furthermore, the presence of more than one pattern of invasion was associated with worse CSM. These results indicate that reporting of the individual location and cumulative amount of pT3a patterns of invasion in clear cell RCC is clinically relevant. Ó 2019 Elsevier Inc. All rights reserved.
Keywords: Carcinoma; Kidney; Metastasis; Renal cancer; Clear cell renal cell carcinoma; Radical nephrectomy
1. Introduction The Tumor-Node-Metastasis (TNM) classification of renal cell carcinoma (RCC) is used to standardize disease extent among patients. The categories within the TNM system aim to group tumors together that share similar prognoses and behaviors. The pT3a subgroup of RCCs represents
Funding: This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. *Corresponding author. Tel.: +1 613-737-8899 ext 72179; fax: (613) 737-8712. E-mail address: [email protected] (T.A. Flood). https://doi.org/10.1016/j.urolonc.2019.09.030 1078-1439/Ó 2019 Elsevier Inc. All rights reserved.
a high-risk collection of tumors that have extended locally beyond the confines of the renal parenchyma. The previously used seventh edition of the American Joint Committee on Cancer (AJCC) staging system classified pT3a tumors as lesions that show perirenal fat invasion (PFI), sinus fat invasion (SFI), or renal vein invasion (RVI) [1]. The recently issued eighth edition of the AJCC TNM staging system added pelvicaliceal system invasion (PSI) as a fourth pattern of RCC invasion that defines the pT3a subcategory [2]. Despite several modifications to the TNM staging system for RCC, the significance of individual patterns of pT3a invasion (PFI, SFI, RVI, and more recently PSI)
ARTICLE IN PRESS 2
T.A. Flood et al. / Urologic Oncology: Seminars and Original Investigations 00 (2019) 1−6
remains uncertain. While some studies have reported that tumors displaying specific patterns of pT3a disease are not associated with an adverse prognosis, others have shown certain patterns to be associated with more aggressive behavior. Furthermore, even though PSI was added to the most recent TNM staging, conflicting results regarding the prognosis of patients with RCC and PSI have been reported [3−8]. Moreover, investigators have also shown that the presence of more than 1 pattern pT3a may impart a worse prognosis for the patient than when a single pattern is detected [9−11]. To our knowledge the incremental effect of PSI added into these cumulative predictive models has not been evaluated. The purpose of this study was to determine if the different patterns of pT3a invasion included in the AJCC eighth edition (PFI, SFI, RVI, and PSI), assessed both separately and cumulatively, were associated with metastases and cancer-specific mortality (CSM) in patients with pT3a clear cell RCC (CCRCC).
Fig. 1. Histologic sections demonstrating; (A) Example of renal vein invasion (RVI) demonstrating tumor thrombus (arrowheads) in muscular containing branch of the renal vein; (B) Sinus fat invasion (SFI) showing tumor cells (arrowheads) in direct contact with fat cells of the renal sinus; (C) Perirenal fat invasion (PFI) exhibited by malignant cells infiltrating into perirenal fat cells (arrowheads) beyond the renal capsule (arrow); (D) Pelvicaliceal system invasion (PSI) characterized by tumor cells involving the lumen of the collecting system (arrowheads) beyond the urothelium (arrow).
2. Methods 2.1. Patients With institutional review board approval, we conducted a consecutive search of our pathology database for patients with pT3a CCRCC who underwent radical nephrectomy between January 2011 and December 2017 at a single institution. We only included radical nephrectomies in our study so that we could confidently examine the pelvicaliceal system for PSI. Only CCRCCs were studied because this is the most common RCC and we did not want to confound our results by including other histologic subtypes that are associated with unique behaviors. The pathological stage was assigned according to the AJCC eighth edition TNM staging system by dedicated genitourinary pathologists at a single tertiary-care referral center for RCC [2]. Information recorded from patient electronic medical records included: PFI, SFI, RVI, PSI, patient age, tumor size, and World Health Organization/International Society of Urological Pathology (WHO/ISUP) nucleolar grade. 2.2. Specimen preparation and histopathologic analysis All radical nephrectomies performed by any surgical approach at the same tertiary care referral center were included. Specimens were fixed in 10% neutral buffered formalin for a minimum of 24 hours. Kidneys were bivalved using a probe inserted into the collecting system as a guide. The renal rein and its branches were opened and examined for foci of RVI. At least 1 section of neoplastic tissue was submitted for every centimeter of the largest dimension of tumor. Additional sections of the renal sinus were submitted when SFI was equivocal. Tissues were paraffin embedded and 4 mm thick sections were cut and stained with hematoxylin and eosin. Subspecialty trained
pathologists provided diagnoses for all radical nephrectomy specimens as part of routine clinical care. Fig. 1A shows an example of RVI, which was defined as identification of tumor into the renal vein or its segmental branches [12]. SFI was designated as tumor cells being in direct contact with fat or in loose connective tissue beyond the renal parenchyma, as well as involvement of endothelial lined spaces within the sinus (Fig. 1B) [12]. PFI was defined as tumor touching fat or extending as irregular tongues into the perinephric tissue, with or without desmoplasia (Fig. 1C) [12]. PSI was characterized by tumor involving the renal calyceal or pelvic submucosa/mucosa with or without luminal involvement (Fig. 1D) [3]. 2.3. Outcomes Outcomes in this study included: distant metastatic disease and CSM. Distant metastases were considered to be present when the patient developed a new lesion in a target organ on CT or MRI studies which were not present at baseline imaging or that grew greater than 20% compared to a baseline imaging study performed within 6 months [13] or when biopsy was performed with histological confirmation of metastatic disease. Metastases and death were determined by chart review and the maximum duration of follow-up was calculated from the date of surgery to the date of last follow-up or for the date of death from our institutional medical records. 2.4. Statistical analysis Skew tests were performed for quantitative data. For normally distributed variables, the mean § standard deviation was reported. For skewed data, the median and interquartile range (IQR) was reported. Multivariate logistic
ARTICLE IN PRESS T.A. Flood et al. / Urologic Oncology: Seminars and Original Investigations 00 (2019) 1−6
regression was performed comparing patient and tumor variables to the outcomes of distant metastases and CSM. Kaplan Meier curves were constructed for each pattern of pT3a invasion and for the cumulative number of pT3a patterns in the same patient. Cox hazard proportional ratios were calculated that compared each of the 4 patterns of pT3a invasion for both CSM and all cause mortality according to previously published guidelines [14]. In our Cox Hazard-Proportional model we studied 1 predictive variable for every 10 events [15]. Statistical analyses were performed using STATA version 13.0 (Statacorp, College Station, TX). A P value of <0.05 was considered statistically significant. 3. Results The search of our database identified 160 consecutive radical nephrectomy specimens with pT3a CCRCC. No patients received neoadjuvant therapy between diagnosis and surgery. Mean § standard deviation patient age was 64.66 § 10.36 years. The median tumor size was 7.5 (IQR 5.7−9.5) cm. The cohort had a median postoperative follow-up of 33.8 (IQR 20.6−55.4) months. Forty (40/160) patients died from CCRCC with a median of 25.4 (IQR 7.5 −40.0) months. Four patients died from causes other than CCRCC. Distant metastases were detected in seventy (70/ 160) patients based on positive imaging findings. These findings were confirmed in 31.4% (22/70) of patients by histologic sampling of the metastasis. Tables 1 and 2 summarize the results comparing the clinicopathologic features to outcomes. The number of individual invasive patterns was as follows: 97/160 (61%) presented with RVI, 91/160 with SFI (57%), 62/160 with PFI (39%), and 24/160 (15%) with PSI. After multivariate analysis, WHO/ISUP nucleolar grade (P < 0.001), tumor size (P = 0.008), RVI (P = 0.028), PFI (P < 0.001), and cumulative number of invasive patterns (P < 0.001) were significantly associated with distant metastases. WHO/ ISUP nucleolar grade (P < 0.001), PFI (P = 0.006), and
3
cumulative number of invasive patterns (P < 0.001) were associated with CSM. Kaplan-Meier curves for individual patterns of pT3a spread are provided in Fig. 2. Among the 4 patterns of pT3a spread, RVI (P = 0.023), SFI (P = 0.045), and PFI (P < 0.001); were associated with increased risk of CSM; whereas, PSI (P = 0.24) was not. Hazard ratios comparing CSM for the 160 patients and each individual invasive patterns were: 2.44 (95% CI, 1.18−5.01; P = 0.015) for RVI, 2.13 (95% CI, 1.05−4.34, P = 0.036) for SFI, 4.12 (95% CI, 2.14−7.92; P < 0.001) for PFI, and 1.43 (95% CI, 0.65 −3.16; P = 0.38) for PSI (Table 3). The hazard ratios for all-cause mortality for the 164 patients and the individual invasive patterns were: 2.25 (95% CI, 1.17-4.39; P = 0.017) for RVI, 2.43 (95% CI, 1.22−4.82; P = 0.011) for SFI, 3.02 (95% CI, 1.77−5.13; P < 0.001) for PFI, and 1.35 (95% CI, 0.62-2.94; P = 0.46) for PSI. The cumulative number of patterns of tumor invasion was as follows: 50.0% (80/160) showed 1 invasive pattern, 31.9% (51/160) had 2 patterns, 14.4% (23/160) had 3 patterns, and 3.8% (6/160) had all 4 invasive patterns. CSM was observed in: 10% (8/80) of patients with 1 invasive pattern, 27.4% (14/51) with 2 invasive patterns, 65.2% (15/23) with 3 patterns, and 50.0% (3/6) among patients who showed all 4 pT3a invasive patterns. Kaplan-Meier curves constructed for the cumulative number of patterns of pT3a invasion showed a worse CSM in patients whose tumor showed increasing number of invasive patterns (P < 0.001), Fig. 3. 4. Discussion This study evaluated individual patterns of pT3a CCRCC invasion and the cumulative effect of multiple invasive patterns with respect to metastatic disease and CSM. Of the 4 patterns of invasion included in the AJCC eighth editions pT3a subgroup, PFI was the most aggressive while PSI was the least aggressive. Our results also show that patients with an increasing number of extrarenal
Table 1 Clinicopathologic features of cohort of patients with pT3a CCRCC and their association with metastases and CSM after multivariate analysis. Variable
N (%)
Patients with metastases (%)
P value
Patients who died
P value
Age Gender Female Male WHO/ISUP nucleolar grade:a 2/4 3/4 4/4 Tumor size Sarcomatoid differentiation Necrosis Cumulative number of invasive patterns
−
−
0.028
−
0.066
43/160 (27) 117/160 (73)
20/70 (29) 50/70 (71)
0.58
13/40 (32) 27/40 (68)
0.69
29/160 (18) 88/160 (55) 43/160 (27) − 24/160 (15) 84/160 (53) −
3/70 (4) 36/70 (51) 31/70 (45) − 19/70 (27) 46/70 (66) −
<0.001
1/40 (2) 15/40 (38) 24/40 (60) − 17/40 (43) 30/40 (75) −
<0.001
a
No patients were graded 1/4.
0.008 0.65 0.42 <0.001
0.095 <0.069 0.76 <0.001
ARTICLE IN PRESS T.A. Flood et al. / Urologic Oncology: Seminars and Original Investigations 00 (2019) 1−6
4
Table 2 Individual invasive patterns and their association with metastases and CSM after multivariate analysis. RVI Metastatic disease Yes (N = 70) 73% (51/70) No (N = 90) 51% (46/90) RVI Cancer specific mortality (CSM) Yes (N = 40) 75% (30/40) No (N = 120) 56% (67/120)
P value
SFI
P value
PFI
P value
PSI
P value
0.009
61% (43/70) 53% (48/90)
0.22
61% (43/70) 21% (19/90)
<0.001
17% (12/70) 13% (12/90)
0.80
P value
SFI
P value
PFI
P value
PSI
P value
0.12
73% (29/40) 52% (62/120)
0.060
65% (26/40) 30% (36/120)
0.006
20% (8/40) 13% (16/120)
0.94
Fig. 2. Kaplan-Meier curves for individual patterns of pT3a spread. Solid lines represent presence of specific indicated pattern of invasion compared to tumors without indicated invasive pattern (dashed line): (A) Renal vein invasion (RVI); (B) Sinus fat invasion (SFI); (C) Perirenal fat invasion (PFI); and (D) Pelvicaliceal system invasion (PSI).
Table 3 Patterns of invasion and association with CSM after Cox regression analysis. pT3a Pattern RVI SFI PFI CSI a
Hazard ratioa 2.44 2.13 4.12 1.43
95% confidence interval (CI)
P value
1.18-5.01 1.05-4.34 2.14-7.92 0.65-3.16
0.015 0.036 <0.001 0.38
Cox proportion hazard ratios includes the four patterns of pT3a studied.
(pT3a) sites of spread have higher CSM. These findings indicate that identifying and reporting of each pattern of pT3a CCRCC invasion at time of pathologic analysis in radical nephrectomy specimens is important and may be of prognostic significance. To our knowledge this is the first study to examine the prognostic impact of the 4 different patterns of extrarenal invasion that define pT3a tumors in the newly accepted AJCC eighth edition. Our results show that a patient whose tumor exhibits additional sites of extrarenal invasion has a
ARTICLE IN PRESS T.A. Flood et al. / Urologic Oncology: Seminars and Original Investigations 00 (2019) 1−6
Fig. 3. Kaplan-Meier curves for the cumulative number of patterns of pT3a invasion detected in the same patient.
significantly worse CSM than when 1 pattern is present. These outcomes highlight the importance of identifying all patterns of invasion with an emphasis conveyed to surgeons regarding the significance associated with their cumulative effect. Both Kresowick et al. and Bedke et al. previously proposed similar recommendations after they independently showed that tumors demonstrating more than 1 invasive pattern (specifically, PFI and SFI in combination) had a worse prognosis than when either pattern was observed alone [9,11]. Similarly, Shah et al. showed that the presence of the 3 patterns of invasion defined by the AJCC seventh edition (PFI, SFI, and RVI) were associated with a higher risk of disease progression and cancer-related death after nephrectomy [10]. Our study is the first to examine the recently included PFI as a feature of pT3a tumors and has consistent findings with these previous reports. The reporting of the amount of patterns in nephrectomy specimens could allow for better prediction of tumor behavior and lead to improved planning of postoperative therapy, surveillance and determining candidacy for clinical trials. PSI was recently included as a feature of pT3a tumors in the new AJCC eighth edition TNM Staging System [2], despite the prognostic value of this feature being uncertain. Some authors have argued that PSI should not be included in the TNM Staging System [6,8,17,19] based on the results of several studies that have shown that this feature is associated with indolent behavior [6,8,17−19]. Our study shares similar findings in that we demonstrated that PSI was not significantly associated with metastatic disease or CSM. Interestingly, other author groups have shown that PSI is associated with a worse prognosis. The majority of these studies examined PSI as a variable across several different pT Stages and only 2 studies have examined the individual prognostic significance of PSI in pT3 tumors [4,16]. In the aforementioned studies by Anderson et al. and Chen et al., PSI was independently associated with worse cancer-specific survival. Unlike Anderson et al. and Chen et al., we restricted our analysis to include pT3a tumors and excluded pT3b and pT3c. Additionally, we limited our evaluation to
5
include only CCRCCs since it is the most common RCC and that other histologic subtypes are associated with unique prognoses. These differences in study design may account for the variation in prognostic significance of PSI observed between research groups. Additional analysis is required to validate the prognostic significance of PSI and to confirm if it should be included within the pT3a subcategory of the TNM Staging System. Aside from PSI, the pT3a group for RCCs includes any tumor demonstrating SFI, RVI, and/or PFI. The exact significance of these other individual patterns of invasion is also debated. The renal sinus is home to an abundance of lymphatics and veins which malignant cells can invade and use to disseminate widely [20]. Some studies have suggested that tumors associated with SFI behave more aggressively when compared to those with PFI [21−23]. Other author’s results demonstrate no significant difference between SFI and PFI when either feature is found in isolation [9,11,24,25]. Different prognoses may also exist between RVI and tumor invading into adipose tissue. Park et al. showed that patients with RVI were associated with a significant higher tumor recurrence and cancer death compared to those with PFI [26]. However, Novara et al. demonstrated that patients with RVI were actually associated with a better prognosis than tumors that invaded the fat [27]. Other groups have shown no difference between the patterns of invasion with regards to prognosis [10,25]. The Shah et al. study was the only manuscript to restrict their cohort to pT3a CCRCC and they found no prognostic significance between the individual patterns of invasion. In our study, PFI imparted the highest risk of CSM and previous studies have also shown that PFI may be a predictor of more aggressive behavior [28,29]. Indeed, our results indicate that PFI is associated with approximately a 4-times increased risk of CSM while the RVI and SFI may be associated with a roughly 2-fold increase. It is evident from these results among the current and previous studies that the individual patterns of invasion in pT3a RCCs are associated with diverse and variable prognoses although heterogeneity in study designs may account for some of these differences. The limitations of this study include the inherent biases associated with a retrospective single institution evaluation. Statistical analysis may have been influenced by small sample sizes, especially with regards to comparing the subgroups of tumor invasion, such as those demonstrating PSI. The smaller sample sizes are partly due to our strict inclusion criteria (limiting our cohort to CCRCC pT3a and radical nephrectomies). However, our choice to only examine radical nephrectomy specimens ensures adequate evaluation of the pelvicalyceal system to assess for PSI, a feature that we felt was fundamentally necessary to sufficiently complete the objectives of the study because of the potential for a false-negative evaluation of PSI at partial nephrectomy. By including only CCRCCs, we eliminated the confounding effect of different behaviors associated with
ARTICLE IN PRESS 6
T.A. Flood et al. / Urologic Oncology: Seminars and Original Investigations 00 (2019) 1−6
various RCC histologic subtypes, although at the cost of introducing bias and reducing our sample size. We feel that these strict inclusion criteria help strengthen our results through the creation of a uniform patient cohort and by ensuring thorough examination of the surgical specimens to assess for all pT3a patterns of invasion. Our low sample size restricted our ability to assess individual combinations of pT3a patterns in models, due to the risk of over-fitting our data, which may be a topic for future studies evaluating pT3a disease in RCC. 5. Conclusions Our results show that among individual pT3a patterns of extrarenal invasion that PFI is the most aggressive, while PSI is the most indolent. Moreover, the presence of a cumulative increase in the amount of pT3a invasive patterns in a patient’s tumor was associated with worse CSM. These results indicate that reporting of both the individual location and cumulative amount of pT3a patterns of CCRCC extrarenal invasion may allow for better risk stratification and follow-up after surgery. References [1] Edge S BD, Compton CC, Fritz AG, Greene FL, Trotti A. American Joint Committee on Cancer, American Cancer Society. AJCC Cancer Staging Manual. 7th ed New York: Springer-Verlag; 2010. [2] Amin MB ES, Greene FL, Byrd DR, Brookland RK, Washington MK, Gershenwald JE, et al. AJCC cancer staging manual. 8th ed Springer; 2017. [3] Bailey GC, Boorjian SA, Ziegelmann MJ, et al. Urinary collecting system invasion is associated with poor survival in patients with clear-cell renal cell carcinoma. BJU Int 2017;119(4):585–90. [4] Chen L, Ma X, Li H, et al. Invasion of the urinary collecting system is an independent prognostic factor in pT3 renal cell carcinoma. Urol Oncol 2016;34(7):293.e11-6. [5] Palapattu GS, Pantuck AJ, Dorey F, et al. Collecting system invasion in renal cell carcinoma: impact on prognosis and future staging strategies. J Urol 2003;170(3):768–72:discussion 72. [6] Schrader AJ, Rustemeier J, Varga Z, et al. Urinary collecting system invasion in renal cell carcinoma: incidence and long-term prognosis. Int J Urol 2009;16(9):718–22. [7] Verhoest G, Avakian R, Bensalah K, et al. Urinary collecting system invasion is an independent prognostic factor of organ confined renal cell carcinoma. J Urol 2009;182(3):854–9. [8] Waalkes S, Merseburger AS, Herrmann TR, et al. Urinary collecting system invasion is no independent prognostic factor in renal cell carcinoma. World J Urol 2010;28(3):283–8. [9] Bedke J, Buse S, Pritsch M, et al. Perinephric and renal sinus fat infiltration in pT3a renal cell carcinoma: possible prognostic differences. BJU Int 2009;103(10):1349–54. [10] Shah PH, Lyon TD, Lohse CM, et al. Prognostic evaluation of perinephric fat, renal sinus fat, and renal vein invasion for patients with pathological stage T3a clear-cell renal cell carcinoma. BJU Int 2019;123(2):270–6.
[11] Kresowik TP, Johnson MT, Joudi FN. Combined renal sinus fat and perinephric fat renal cell carcinoma invasion has a worse prognosis than either alone. J Urol 2010;184(1):48–52. [12] Trpkov K, Grignon DJ, Bonsib SM, et al. Handling and staging of renal cell carcinoma: the International Society of Urological Pathology Consensus (ISUP) conference recommendations. Am J Surg Pathol 2013;37(10):1505–17. [13] Nishino M, Jagannathan JP, Ramaiya NH, et al. Revised RECIST guideline version 1.1: What oncologists want to know and what radiologists need to know. AJR Am J Roentgenol 2010;195(2):281–9. [14] Abd ElHafeez S, Torino C, D’Arrigo G, et al. An overview on standard statistical methods for assessing exposure-outcome link in survival analysis (Part II): the Kaplan-Meier analysis and the Cox regression method. Aging Clin Exp Res 2012;24(3):203–6. [15] Harrell FE Jr., Lee KL, Califf RM, et al. Regression modelling strategies for improved prognostic prediction. Stat Med 1984;3(2):143–52. [16] Anderson CB, Clark PE, Morgan TM, et al. Urinary collecting system invasion is a predictor for overall and disease-specific survival in locally invasive renal cell carcinoma. Urology 2011;78(1):99–104. [17] Terrone C, Cracco C, Guercio S, et al. Prognostic value of the involvement of the urinary collecting system in renal cell carcinoma. Eur Urol 2004;46(4):472–6. [18] Mirza KM, Taxy JB, Antic T. Radical nephrectomy for renal cell carcinoma: its contemporary role related to histologic type, tumor zize, and nodal status: aA retrospective study. Am J Clin Pathol 2016;145 (6):837–42. [19] Uzzo RG, Cherullo E, Myles J, et al. Renal cell carcinoma invading the urinary collecting system: implications for staging. J Urol 2002;167(6):2392–6. [20] Bonsib SM, Gibson D, Mhoon M, et al. Renal sinus involvement in renal cell carcinomas. Am J Surg Pathol 2000;24(3):451–8. [21] Thompson RH, Leibovich BC, Cheville JC, et al. Is renal sinus fat invasion the same as perinephric fat invasion for pT3a renal cell carcinoma? J Urol 2005;174(4 Pt 1):1218–21. [22] Bonsib SM. The renal sinus is the principal invasive pathway: a prospective study of 100 renal cell carcinomas. Am J Surg Pathol 2004;28(12):1594–600. [23] Zhang Z, Yu C, Velet L, et al. The difference in prognosis between renal sinus fat and perinephric fat invasion for pT3a renal cell carcinoma: a meta-analysis. PLoS One 2016;11(2):e0149420. [24] Poon SA, Gonzalez JR, Benson MC, et al. Invasion of renal sinus fat is not an independent predictor of survival in pT3a renal cell carcinoma. BJU Int 2009;103(12):1622–5. [25] Margulis V, Tamboli P, Matin SF, et al. Location of extrarenal tumor extension does not impact survival of patients with pT3a renal cell carcinoma. J Urol 2007;178(5):1878–82. [26] Park M, Shim M, Kim M, et al. Prognostic heterogeneity in T3aN0M0 renal cell carcinoma according to the site of invasion. Urol Oncol 2017;35(7):458.e17-58.e22. [27] Novara G, Ficarra V, Antonelli A, et al. Validation of the 2009 TNM version in a large multi-institutional cohort of patients treated for renal cell carcinoma: are further improvements needed? Eur Urol 2010;58(4):588–95. [28] Brookman-May SD, May M, Wolff I, et al. Evaluation of the prognostic significance of perirenal fat invasion and tumor size in patients with pT1-pT3a localized renal cell carcinoma in a comprehensive multicenter study of the CORONA project. Can we improve prognostic discrimination for patients with stage pT3a tumors? Eur Urol 2015;67(5):943–51. [29] Yoo C, Song C, Hong JH, et al. Prognostic significance of perinephric fat infiltration and tumor size in renal cell carcinoma. J Urol 2008;180 (2):486–91:discussion 91.