Pathologic Characteristics of Solitary Small Renal Masses. Can They Be Predicted by Preoperative Clinical Parameters?

Transcription

1 Anatomic Pathology / Pathology of Small Renal Masses Pathologic Characteristics of Solitary Small Renal Masses Can They Be Predicted by Preoperative Clinical Parameters? Tom DeRoche, MD, 1 Esteban Walker, PhD, 2 Cristina Magi-Galluzzi, MD, PhD, 1 and Ming Zhou, MD, PhD 1 Key Words: Small renal mass; Renal cell carcinoma; Pathology; Clinical correlation Abstract We studied the pathologic features and whether clinical features could predict pathologic outcomes in small renal masses. The study included all adult patients with solitary, nonmetastatic renal masses 4 cm or smaller confirmed by nephrectomy or needle biopsy between 2004 and Tumor size, histologic type, Fuhrman nuclear grade, and stage were obtained from surgical pathology reports. Clinical variables included age, sex, tumor size, and symptomatology. The study included 290 men and 198 women (mean age, 59.3 years). Median tumor size was 2.6 cm (range, cm). Approximately 84% of masses were incidentally detected. Nonneoplastic entities, benign neoplasms, and low- and high-grade carcinoma accounted for 1.6%, 18.0%, 49.0%, and 31.4% of masses, respectively. Women were more likely to have a benign mass (27.3% vs 14.5% of men, P <.001). Age (P =.56), tumor size (mean, 2.63 vs 2.46 cm for benign; P =.08), and symptomatology (P =.46) were not associated with malignancy. Multivariate analyses using sex, age, tumor size, and symptomatology failed to produce a model useful to predict the pathology of individual tumors. This inability may argue for an increased role for needle biopsy in their evaluation. During the past few decades, there has been an increase in the incidence of small renal masses, owing in part to the increased use of abdominal cross-sectional imaging in a workup for an unrelated condition. 1-3 Although the majority of small renal masses represent renal cell carcinoma (RCC), a significant proportion is ultimately found to be benign or nonneoplastic on histopathologic examination. In a large 30-year retrospective study, Frank et al 4 found that 221 (23.3%) of 947 renal masses measuring less than 4 cm were benign. More recently, Snyder et al 5 found that 98 (18.5%) of 531 renal masses measuring 4 cm or less were benign. In addition, most small RCCs are low-grade. In the series by Frank et al, % of RCCs measuring less than 5 cm had low-grade histologic features. Several small, prospective studies examining outcomes of expectant management provide insight into the natural history of small renal masses. Kassouf et al 6 studied a cohort of 24 patients with small renal masses (mean, 3.3 cm). At a mean follow-up of 31 months, only 5 (21%) showed tumor growth, and there were no cases of metastasis. 6 Kouba et al, 7 in a study of 43 patients with a mean tumor size of 2.9 cm, showed tumor growth in 74% of patients (mean, 0.7 cm/y) but no cases of metastasis at a mean follow-up of 3 years. Therefore, there has been a paradigm shift in the management of these small masses, with partial nephrectomy and minimally invasive ablative therapies gaining increased acceptance. Active surveillance with serial imaging studies has been used as an alternative to surgery, mainly in patients with underlying comorbidities. Several studies have examined the relationship between tumor size and pathologic diagnosis, but few have examined the influence of additional clinical variables, including age, 560 Am J Clin Pathol 2008;130: Downloaded 560 from

2 Anatomic Pathology / Original Article sex, and symptomatology. In addition, there is little information regarding pathologic features of T1a tumors, which are more likely to be managed conservatively. Knowledge of the pathologic nature of these small renal masses will help clinicians and patients to choose the appropriate treatment modality. The objectives of this study were to describe the pathologic features of a contemporary series of solitary renal masses measuring 4 cm or less and to determine whether sex, age, tumor size, and/or symptomatic manifestations could predict the pathologic nature of such masses. Materials and Methods The Cleveland Clinic (Cleveland, OH) surgical pathology database was searched for nephrectomies (partial and radical) and renal mass biopsies during the 2-year period between July 2004 and July Cases of resections for upper tract urothelial neoplasms, retroperitoneal sarcomas, and end-stage renal disease were excluded. In addition, the data for patients with recurrent RCC or a documented history of von Hippel-Lindau syndrome were excluded. Only the data for adult patients with solitary, nonmetastatic renal masses 4 cm or smaller confirmed by nephrectomy or needle biopsy were included in the study. Pathologic data, including tumor size, histologic type, Fuhrman nuclear grade, and stage were obtained from surgical pathology reports. All diagnoses were made by genitourinary pathologists using the 2004 World Health Organization classification of renal neoplasms. In cases with a needle biopsy proven diagnosis followed by ablative therapy rather than resection, the tumor size was obtained from preoperative computed tomography scans. A low-grade neoplasm was defined as an organconfined tumor (2002 American Joint Committee on Cancer stage T1a) with a Fuhrman nuclear grade of 1 or 2 without lymphatic or vascular invasion, tumor necrosis, or sarcomatoid differentiation. Hybrid oncocytic tumors with features of renal oncocytoma and chromophobe RCC and carcinoid tumor were also classified as low-grade neoplasms. Tumors with any of the following characteristics were categorized as high-grade neoplasms: American Joint Committee on Cancer stage T3, lymph node involvement, microscopic or macroscopic vascular invasion, Fuhrman nuclear grade 3 or 4, or a sarcomatous component. Age, sex, and clinical manifestations were recorded. The clinical manifestations were categorized as symptomatic (flank pain, hematuria, or symptoms of metastatic disease) or asymptomatic. Descriptive statistics were obtained for continuous and categorical variables. Associations were analyzed depending on the nature of the variables involved (ie, t test, χ 2 test, logistic regression). Multivariate logistic regression was used to construct a model to predict the pathologic classification of individual tumors based on the clinical parameters. JMP 7.0 (2007, SAS Institute, Cary, NC) was used to analyze the data. The Cleveland Clinic Institutional Review Board approved the study. Results This study included 488 cases, including 465 resection and 23 needle biopsy specimens. The latter 23 cases represented biopsy specimens obtained before patients underwent cytoablative procedures and included 17 clear cell RCCs, 3 papillary RCCs, and 3 oncocytomas. The study population included 290 men and 198 women with a mean age of 59.3 years (range, years). Symptomatic manifestations, including gross hematuria and flank pain, were present in 78 patients (16.0%). There were no cases of synchronous metastasis. The median tumor size was 2.60 cm (range, cm). The pathologic diagnoses are listed in ztable 1z. Eight masses (1.6%) were nonneoplastic entities including benign complex cysts and xanthogranulomatous pyelonephritis. Benign neoplasms accounted for 18.0%, including oncocytoma, angiomyolipoma, metanephric adenoma, leiomyoma, and mixed epithelial and stromal tumor of the kidney. Clear cell, papillary, and chromophobe RCCs accounted for 53.9%, 15.2%, and 7.8%, respectively. There were 49.3% low-grade and 31.4% high-grade carcinomas. Among 392 malignant tumors, Fuhrman nuclear grades were 3 and 4 in 118 (30.1%) and 13 (3.3%), respectively. Invasion of perinephric or renal sinus fat was seen in 33 (8.4%). Lymphovascular invasion was identified in 18 (4.6%). Sarcomatoid differentiation and adrenal gland involvement were each seen in only 1 case. ztable 1z Pathologic Diagnoses of 488 Small Renal Masses Histologic Type No. (%) of Cases Renal cell carcinoma 392 (80.3) Clear cell 263 (53.9) Papillary 74 (15.2) Chromophobe 38 (7.8) Multilocular cystic 7 (1.4) Unclassified 6 (1.2) Hybrid oncocytic tumor 3 (0.6) Carcinoid tumor 1 (0.2) Benign tumors 88 (18.0) Oncocytoma 56 (11.5) Angiomyolipoma 24 (4.9) Mixed epithelial and stromal tumor/ 5 (1.0) cystic nephroma Metanephric adenoma 2 (0.4) Leiomyoma 1 (0.2) Nonneoplastic lesions 8 (1.6) Downloaded from Am J Clin Pathol 2008;130:

3 DeRoche et al / Pa t h o l o g y o f Sm a l l Renal Masses ztable 2z summarizes the pathologic classification of the tumors based on patient sex, age, tumor size, and symptomatology. Sex was associated with the pathologic classification, ie, women were less likely to have malignant lesions (72.7% vs 85.5% in men; P <.001). Age was not associated with the pathologic classification. Table 2 shows the age cutoff at 50 years. Other age cutoffs were explored, but no statistical significance was achieved. When used as a continuous variable, tumor size had no significant effect on pathologic characteristics, even after adjusting for sex. However, by using an optimal cutoff of 3.5 cm to dichotomize tumor size, 88.9% of tumors larger than 3.5 cm were malignant, compared with 77.9% for tumors 3.5 cm or smaller (P =.008; Table 2). High-grade carcinoma was seen in 50 (46.3%) of 108 masses larger than 3.5 cm compared with 103 (27.1%) of 380 masses measuring 3.5 cm or less (P <.001). These features were seen in 21.1% of masses smaller than 2 cm (vs 36% of masses 2 cm; P =.002). There was no significant difference in the proportion of high-grade carcinomas between the sexes (32.4% vs 29.8% for women; P =.54). Symptomatic manifestations did not predict the presence of malignancy, with 65 (83%) of 78 symptomatic and 327 (79.8%) of 410 incidentally found tumors proving malignant. Nor did it predict high-grade pathologic features in malignant tumors (47.7% [31/65] in symptomatic vs 37.4% [123/329] in incidentally found tumors; P =.12). One of the objectives of this study was to determine if clinical parameters, including age, sex, tumor size, and symptomatology could be used to predict tumors as being benign or malignant. Multivariate logistic regression was used to fit a model to predict the probability of the pathologic classification of any given tumor. As shown in ztable 3z, several such parameters, including sex, age combined with sex, age combined with tumor size, and sex combined with tumor size, were significantly associated with the pathologic classification. For example, 100.0% of men older than 68 years with a mass 3.5 cm or larger had malignant neoplasms. In contrast, 68.9% of women younger than 68 years with a mass smaller than 3.5 cm had malignant neoplasms (P =.0006) ztable 4z. However, when the model was used to predict the pathologic classification using the clinical parameters, the results were disappointing. ztable 5z shows that the model correctly predicted 390 of the 392 malignant tumors with a sensitivity of 99.5%. However, only 6 of the 96 benign lesions were predicted correctly with a specificity of 6%, or a false-positive rate of 93.8%. Discussion Although the majority of small renal masses are renal cell carcinomas, a significant proportion (14%-26%) are benign neoplasms, including oncocytomas and angiomyolipomas, or ztable 2z Pathologic Classification Stratified by Sex, Age, and Tumor Size in 488 Small Renal Masses No. (%) With No. of Patients Malignancy P Sex <.001 Male (85.5) Female (72.7) Age (y) (83.1) > (79.4) Tumor size (cm) (77.9) > (88.9) Symptoms.46 Yes (83) No (79.8) χ 2 test. ztable 3z Logistic Regression Model to Predict the Pathologic Classification by Patient Age, Sex, and Tumor Size Term Age.22 Sex.001 Tumor size.36 Age and sex.036 Age and tumor size.023 Sex and tumor size.034 The P value for the model is ztable 4z Pathologic Classification of Small Renal Masses Stratified by Age, Sex, and Tumor Size Tumor Size No. of No. (%) With Sex (cm) Age (y) Patients Malignancy Male <3.5 < (86.9) (75) 3.5 < (88) (100) Female <3.5 < (68.9) (74) 3.5 < (79) (92) P =.0006; χ 2 test. ztable 5z Prediction of Pathologic Classification of Small Renal Masses Using the Multivariate Logistic Regression Best-fit Model Final Pathologic Classification Prediction Benign Malignant Total Benign Malignant Total P 562 Am J Clin Pathol 2008;130: Downloaded 562 from

4 Anatomic Pathology / Original Article nonneoplastic lesions. 4,5,8,9 Given the increase in incidental detection of these masses, especially in older patients, there has been much interest in using clinical parameters to predict the pathologic nature of these lesions to guide therapeutic decisions. Patients who are likely to have a benign mass could potentially be treated in a more conservative manner. Few studies have examined the pathologic features of these small renal masses and the relationship between pathologic features in surgically resected specimens and clinical parameters, including tumor size, sex, age, and symptomatology, in a contemporary cohort. Among these parameters, tumor size seems to be the strongest predictor of malignant pathologic features. In the series by Frank et al, % of tumors smaller than 4 cm were benign compared with only 8% of tumors larger than 4 cm. Snyder et al 5 found that 18.5% of masses smaller than 4 cm were benign, compared with 12.7% of masses larger than 4 cm. Among small tumors, there is a disproportionate representation of relatively indolent papillary RCC, with the odds of clear cell RCC increasing with increased tumor size. 4 The present study yielded similar results: 19.6% of small renal masses were nonneoplastic or benign. The remaining 80.3% of solitary masses measuring 4 cm or smaller were malignant. It is important to note that 31.4% of masses showed highgrade pathologic features in our series, including Fuhrman grade 3 or 4, sarcomatoid differentiation, perinephric or renal sinus invasion, and vascular and adrenal invasion. Such highgrade histologic features were present in 27.1% of masses smaller than 3.5 cm and 21.1% of masses smaller than 2 cm. Therefore, even a small renal mass should be carefully examined for these adverse pathologic features. Sex has been shown to influence the classification of renal masses, with women nearly twice as likely as men to have a benign lesion. 5 In our series, female sex was significantly associated with benign lesions, with an odds ratio of 2.23 (95% confidence interval, ). In women, the increased incidence of benign lesions is related to the disproportionate representation of angiomyolipoma in this group (odds ratio, 18.1; 95% confidence interval, ). Therefore, a diagnosis of angiomyolipoma should be considered for a small renal mass in female patients, especially when such a diagnosis is also suspected radiologically. 10,11 The relationship between symptomatology and pathologic features of small renal masses was also explored in this study. The increasing use of cross-sectional imaging has resulted in an increase in incidentally detected renal masses. 2,3 Recent series have reported that 48% to 74% of all renal masses are detected incidentally. 8,12 Incidentally detected RCCs tend to be of lower grade and stage than RCCs found because of symptoms Among stage T1 masses, however, there is no significant difference in malignant or high-grade pathologic features relating to symptomatology. 12 In the present series, 84.0% of small renal masses were detected incidentally. Symptomatic manifestations were not significantly associated with malignancy, and, among malignant tumors, there was no association between symptomatology and the presence of high-grade pathologic features in renal masses smaller than 4 cm. Although sex and tumor size correlated with the final pathologic nature of the small renal masses, neither, by itself, could be reliably used to distinguish renal masses with different pathologic classifications. We therefore sought to examine clinical factors that, when combined, might predict pathologic features at resection, including sex, age, tumor size, and symptomatology. We did not include preoperative imaging results because a preliminary study did not find that radiologic findings contributed to such a prediction model. 18 We used logistic regression to obtain a best-fit model that used clinical data to predict the pathologic classification. Our model showed a statistically significant association with tumor classification and excellent sensitivity such that it correctly predicted 99.5% of malignant renal masses (390/392). However, it could only predict 6% (6/96) of benign renal masses correctly, and the false-positive rate was 93.8% (90/96). Because the majority of the small renal masses (~80%) were malignant tumors, it is expected that such a prediction model will have excellent sensitivity. The specificity of the model, ie, its ability to predict which patient is likely to have a benign or nonneoplastic mass, is, therefore, of more clinical significance because patients who are predicted to have benign masses can be managed conservatively. We, therefore, conclude that prediction of pathologic nature based on the clinical factors is unreliable for renal masses 4 cm or smaller. The inability to predict pathologic findings in small renal masses using clinical data may argue for a greater role for needle biopsy in the workup of small renal masses. Currently, this approach is mainly used in patients who are not operative candidates. The sensitivity of core needle biopsy in diagnosing malignancy in renal masses is 70% to 100%, and the specificity is 100% (for review, see Volpe et al 19 ). A potential downside to core needle biopsy in the assessment of small renal masses is intratumoral heterogeneity or the presence of an unsampled higher grade component within an RCC. This issue is most important in patients who are candidates for surveillance because the presence of a low-grade malignancy is sufficient to merit definitive surgical management in younger, healthy patients. There are conflicting data regarding assessment of nuclear grade in core needle biopsy specimens. Neuzillet et al 20 reported a 70% concordance between biopsy and resection nuclear grades, with no cases differing by more than 1 grade. More recently, Lebret et al 21 reported a concordance of 46%, with 4 nuclear grade 1 cases found to be grade 3 at resection. Intratumoral heterogeneity can also be problematic with regard to oncocytic neoplasms, with the potential for misinterpretation of a chromophobe RCC as oncocytoma Downloaded from Am J Clin Pathol 2008;130:

5 DeRoche et al / Pa t h o l o g y o f Sm a l l Renal Masses in a needle biopsy specimen. 22 If the sample is adequate, immunohistochemical analysis for cytokeratin 7 may be helpful in distinguishing these 2 entities. 23 Conclusions The majority (~80%) of solitary masses measuring 4 cm or less were malignant. Furthermore, high-grade histologic features were present in about one third of these cases. Sex is significantly associated with the pathologic nature of small renal masses, with benign neoplasms twice as common in women. Tumor size also correlated with the pathologic classification. However, age and symptomatic manifestations had no significant association with pathologic features. The pathologic nature of the small renal masses cannot be accurately predicted by the clinical factors, including sex, age, tumor size, and clinical manifestations. From the Departments of 1 Anatomic Pathology and 2 Quantitative Health Science, Cleveland Clinic, Cleveland, OH. Address reprint requests to Dr Zhou: Dept of Anatomic Pathology, Cleveland Clinic Foundation, 9500 Euclid Ave, L25, Cleveland, OH References 1. Hollingsworth JM, Miller DC, Daignault S, et al. Rising incidence of small renal masses: a need to reassess treatment effect. J Natl Cancer Inst. 2006;98: Jayson M, Sanders H. Increased incidence of serendipitously discovered renal cell carcinoma. Urology. 1998;51: Smith SJ, Bosniak MA, Megibow AJ, et al. Renal cell carcinoma: earlier discovery and increased detection. Radiology. 1989;170: Frank I, Blute ML, Cheville JC, et al. Solid renal tumors: an analysis of pathological features related to tumor size. J Urol. 2003;170: Snyder ME, Bach A, Kattan MW, et al. Incidence of benign lesions for clinically localized renal masses smaller than 7 cm in radiological diameter: influence of sex. J Urol. 2006;176: Kassouf W, Aprikian AG, Laplante M, et al. Natural history of renal masses followed expectantly. J Urol. 2004;171: Kouba E, Smith A, McRackan D, et al. Watchful waiting for solid renal masses: insight into the natural history and results of delayed intervention. J Urol. 2007;177: Duchene DA, Lotan Y, Cadeddu JA, et al. Histopathology of surgically managed renal tumors: analysis of a contemporary series. Urology. 2003;62: Schachter LR, Cookson MS, Chang SS, et al. Second prize: frequency of benign renal cortical tumors and histologic subtypes based on size in a contemporary series: what to tell our patients. J Endourol. 2007;21: Bosniak MA. Angiomyolipoma (hamartoma) of the kidney: a preoperative diagnosis is possible in virtually every case. Urol Radiol. 1981;3: Nelson CP, Sanda MG. Contemporary diagnosis and management of renal angiomyolipoma. J Urol. 2002;168: Schlomer B, Figenshau RS, Yan Y, et al. Pathological features of renal neoplasms classified by size and symptomatology. J Urol. 2006;176: Bretheau D, Lechevallier E, Eghazarian C, et al. Prognostic significance of incidental renal cell carcinoma. Eur Urol. 1995;27: Gudbjartsson T, Thoroddsen A, Petursdottir V, et al. Effect of incidental detection for survival of patients with renal cell carcinoma: results of population-based study of 701 patients. Urology. 2005;66: Konnak JW, Grossman HB. Renal cell carcinoma as an incidental finding. J Urol. 1985;134: Luciani LG, Cestari R, Tallarigo C. Incidental renal cell carcinoma: age and stage characterization and clinical implications: study of 1092 patients ( ). Urology. 2000;56: Porena M, Vespasiani G, Rosi P, et al. Incidentally detected renal cell carcinoma: role of ultrasonography. J Clin Ultrasound. 1992;20: Kutikov A, Fossett LK, Ramchandani P, et al. Incidence of benign pathologic findings at partial nephrectomy for solitary renal mass presumed to be renal cell carcinoma on preoperative imaging. Urology. 2006;68: Volpe A, Kachura JR, Geddie WR, et al. Techniques, safety and accuracy of sampling of renal tumors by fine needle aspiration and core biopsy. J Urol. 2007;178: Neuzillet Y, Lechevallier E, Andre M, et al. Accuracy and clinical role of fine needle percutaneous biopsy with computerized tomography guidance of small (less than 4.0 cm) renal masses. J Urol. 2004;171: Lebret T, Poulain JE, Molinie V, et al. Percutaneous core biopsy for renal masses: indications, accuracy and results. J Urol. 2007;178: Shah RB, Bakshi N, Hafez KS, et al. Image-guided biopsy in the evaluation of renal mass lesions in contemporary urological practice: indications, adequacy, clinical impact, and limitations of the pathological diagnosis. Hum Pathol. 2005;36: Leroy X, Moukassa D, Copin MC, et al. Utility of cytokeratin 7 for distinguishing chromophobe renal cell carcinoma from renal oncocytoma. Eur Urol. 2000;37: Am J Clin Pathol 2008;130: Downloaded 564 from