Genitourinary Imaging
Original Research
Malignancy Rate, Histologic Grade, and Progression of Bosniak Category III and IV Complex Renal Cystic Lesions
Audio Available | 
Share
OBJECTIVE. The primary purpose of this study is to determine the malignancy rate, histologic grade, and initial stage of surgically treated complex renal cysts classified as Bosniak category III or IV. For nonsurgical lesions, a secondary objective was to evaluate lesion progression on follow-up examinations.
MATERIALS AND METHODS. We searched our database for cystic lesions classified as Bosniak III or IV category on CT or MRI from January 2008 to April 2016. Surgically resected lesions, per category, were correlated with information on pathologic reports to obtain malignancy rates. For malignant lesions, histologic grade and initial stage were evaluated. Imaging follow-up of at least 2 years was used to evaluate progression of clinically followed lesions.
RESULTS. We included 86 lesions in 85 patients in the final analysis. Of the 60 surgically resected lesions (70%), 46 (77%) were malignant and 14 (23%) were benign. Malignancy rates were 72% for Bosniak category III lesions and 86% for Bosniak category IV lesions. Most malignant cysts were early-stage (pT1) cysts with low histologic grades (89% of Bosniak III lesions and 91% of Bosniak IV lesions). Follow-up studies of the surgically resected lesions did not show local recurrence, metastasis, or lymph node enlargement. Among patients with lesions managed by watchful waiting (n = 26), all lesions remained unchanged in terms of size and complexity after at least 2-years of follow-up.
CONCLUSION. Although high malignancy rates were observed for both Bosniak category III and IV lesions, our results suggest that such malignant cysts are usually early-stage tumors with a low histologic grade. Lesions that underwent follow-up remained unchanged on control examinations. These findings may indicate low aggressiveness of these lesions, supporting the idea that more conservative approaches may be used.
Keywords: Bosniak classification, CT, kidney cysts, MRI, renal cysts
Renal cystic lesions are extremely common, with a prevalence of at least 50% in populations older than 50 years [1], and they are often incidentally discovered on imaging examinations. The Bosniak classification for renal cystic lesions aims to predict the risk of malignancy by means of a standardized report based on CT findings [1–4], criteria that were later extended to MRI [3, 4], suggesting appropriate management regardless of lesion cause [5–7].
Bosniak categories I and II represent simple and minimally complicated cysts (small hemorrhagic cysts or cysts containing thin septa), respectively, and do not require monitoring or specific treatment [1–4]. Bosniak category IIF corresponds to indeterminate but probably benign findings that require follow-up, including an intrarenal hemorrhagic cyst larger than 3.0 cm and cysts containing thick calcifications or multiple thin septa [4, 8].
Bosniak category III and IV cysts are lesions with progressive malignancy rates, and surgical resection is often considered on the basis of urologic guidelines [9–11]. Although category III lesions are considered suspicious for malignancy given the presence of thick septa or septal enhancement, they include a broad spectrum of benign differential diagnoses, such as chronic abscess, chronically infected or hemorrhagic cyst, simple cyst after alcohol ablation, localized cystic disease, mixed epithelial and stromal tumor, and multilocular cystic nephroma [12, 13]. Category IV lesions are characterized by a solid enhancing component and are highly suspicious for malignancy, particularly the intracystic type of renal cell carcinoma (RCC), which is a less aggressive subtype than conventional RCC [14–17] and should not be confused with solid RCCs with a necrotic component (a more aggressive lesion) [18].
Several studies have aimed to evaluate the malignancy rates for Bosniak category III and IV lesions [14–17, 19–29]. However, most of these studies had small simple sizes [20–22], were conducted before the Bosniak classification update [20, 21], did not differentiate category IIF from category III [19, 22–24, 26], or had a short follow-up period (6 months) [29]. Some studies focused on pathologic findings (i.e., histologic grade) [14–17] but did not correlate with imaging findings or Bosniak classification.
Therefore, the aim of the present study is to evaluate the malignancy rate of complex renal cystic lesions classified as Bosniak categories III and IV, the histologic grade, stage at diagnosis, and progression rate on follow-up studies.
| Materials and Methods |   |
|---|
The institutional review board at Hospital Israelita Albert Einstein approved our retrospective study and provided a waiver of informed consent. We searched our database for CT and MRI reports from January 2008 to April 2016 that classified cystic complex renal lesions as Bosniak category III or IV (Figs. 1–4).
 View larger version (269K) | Fig. 1A —53-year-old woman with complex left renal cystic mass classified as Bosniak category IV on MRI. A, Coronal T2-weighted MR image shows cystic lesion with thick internal septation and solid parietal component. |
 View larger version (232K) | Fig. 1B —53-year-old woman with complex left renal cystic mass classified as Bosniak category IV on MRI. B, Coronal T1-weighted unenhanced (B) and contrast-enhanced (C) corticomedullary phase MR images show nodular heterogenous enhancement. Pathologic evaluation revealed benign anastomosing hemangioma. |
 View larger version (303K) | Fig. 1C —53-year-old woman with complex left renal cystic mass classified as Bosniak category IV on MRI. C, Coronal T1-weighted unenhanced (B) and contrast-enhanced (C) corticomedullary phase MR images show nodular heterogenous enhancement. Pathologic evaluation revealed benign anastomosing hemangioma. |
 View larger version (171K) | Fig. 2A —44-year-old man with complex right renal cystic mass classified as Bosniak category III on MRI. A, Axial fat-saturated T2-weighted (A), coronal T1-weighted unenhanced (B), and contrast-enhanced nephrographic phase (C) MR images show 1.2-cm cortical cyst with thick enhancing walls in upper third of right kidney. Final pathologic evaluation revealed cystic oncocytoma. |
 View larger version (243K) | Fig. 2B —44-year-old man with complex right renal cystic mass classified as Bosniak category III on MRI. B, Axial fat-saturated T2-weighted (A), coronal T1-weighted unenhanced (B), and contrast-enhanced nephrographic phase (C) MR images show 1.2-cm cortical cyst with thick enhancing walls in upper third of right kidney. Final pathologic evaluation revealed cystic oncocytoma. |
 View larger version (268K) | Fig. 2C —44-year-old man with complex right renal cystic mass classified as Bosniak category III on MRI. C, Axial fat-saturated T2-weighted (A), coronal T1-weighted unenhanced (B), and contrast-enhanced nephrographic phase (C) MR images show 1.2-cm cortical cyst with thick enhancing walls in upper third of right kidney. Final pathologic evaluation revealed cystic oncocytoma. |
 View larger version (283K) | Fig. 3A —53-year-old man with cystic renal mass in left kidney that was classified as Bosniak category IV on MRI. A, Coronal T2-weighted (A), unenhanced (B), and contrast-enhanced (C) MR images show right renal cyst with thick enhancing septa and mural nodules. Pathologic evaluation confirmed renal cell carcinoma of clear cell type with International Society of Urological Pathology histologic grade 2 and TNM stage pT1a. |
 View larger version (248K) | Fig. 3B —53-year-old man with cystic renal mass in left kidney that was classified as Bosniak category IV on MRI. B, Coronal T2-weighted (A), unenhanced (B), and contrast-enhanced (C) MR images show right renal cyst with thick enhancing septa and mural nodules. Pathologic evaluation confirmed renal cell carcinoma of clear cell type with International Society of Urological Pathology histologic grade 2 and TNM stage pT1a. |
 View larger version (243K) | Fig. 3C —53-year-old man with cystic renal mass in left kidney that was classified as Bosniak category IV on MRI. C, Coronal T2-weighted (A), unenhanced (B), and contrast-enhanced (C) MR images show right renal cyst with thick enhancing septa and mural nodules. Pathologic evaluation confirmed renal cell carcinoma of clear cell type with International Society of Urological Pathology histologic grade 2 and TNM stage pT1a. |
 View larger version (230K) | Fig. 4A —84-year-old man with complex right renal cystic mass classified as Bosniak category IV on MRI. A, Coronal fat-saturated T1-weighted contrast-enhanced nephrographic phase MR images obtained in 2012 (A) and 2016 (B) show exophytic cortical cyst with solid component. No significant changes between images were noted. Patient is undergoing active surveillance. |
 View larger version (207K) | Fig. 4B —84-year-old man with complex right renal cystic mass classified as Bosniak category IV on MRI. B, Coronal fat-saturated T1-weighted contrast-enhanced nephrographic phase MR images obtained in 2012 (A) and 2016 (B) show exophytic cortical cyst with solid component. No significant changes between images were noted. Patient is undergoing active surveillance. |
CT examinations were performed using several different 16- to 320-MDCT scanners, with the use of collimation up to 2 mm, including protocols for routine abdominal imaging, abdominal pain, renal masses, and CT urography, all of which involved, at a minimum, an unenhanced phase and a contrast-enhanced portal phase (i.e., a 70-second delay) or nephrographic phase (i.e., a 90-second delay). Iodine-based IV contrast medium (1–2 mL/kg of body weight) was delivered using a power injector at a rate of 2.0–3.0 mL/s.
MRI examinations were performed using several different 1.5- or 3-T scanners. Detailed protocols for routine abdominal MRI and renal mass MRI are shown in Appendix 1. At a minimum, axial and coronal fast spin-echo T2-weighted images, chemical shift–based images, DW images, and unenhanced and contrast-enhanced T1-weighted sequences were obtained in all cases (including corticomedullary, nephrographic, and excretory phases), for the area extending from the diaphragm to below the inferior renal pole (see Appendix 1). Gadolinium-based contrast medium (dose, 0.2 mL/kg of body weight) was injected at a rate of 2 mL/s.
Examinations were prospectively read in consensus by two board-certified abdominal radiologists with 3–20 years of experience, and all cases included in this study were reviewed by one of the authors. For image interpretation, cases were read on an integrated PACS radiology information system that included multiplanar reconstruction tools (Carestream Health).
The inclusion criterion was imaging reports of complex cystic renal lesions larger than 1.0 cm that had been classified as Bosniak category III or IV lesions on CT or MRI examinations. Exclusion criteria were examinations that had incomplete protocols or contained substantial artifacts, incomplete medical records (i.e., pathologic reports that were unavailable or were obtained from other institutions), and less than 2 years of follow up of unresected lesions. We primarily searched for abdominal CT or abdominal MRI as follow-up examinations. Any signs of metastatic disease noted on chest CT, head CT, or whole-body PET/CT were also evaluated.
All pathologic reports were related to surgical resection specimens (from partial or total nephrectomy), were interpreted by one of two uropathologists who had more than 10 years of experience, and included information on the Fuhrman nuclear grade or International Society of Urological Pathology grade as well as the TNM stage of the lesions.
With the use of statistical software (MedCalc, version 17.8, MedCalc software), an unpaired t test was used to evaluate differences in size among benign and malignant lesions.
| Results | |
|---|
A search of radiology reports resulted in the identification of 160 patients. Of those, 75 patients were excluded for the following reasons: 71 patients were not surgically treated and had less than 2 years of follow-up, three patients received percutaneous treatment and underwent no previous biopsies, and one patient underwent lesion resection at an outside institution and had unavailable pathologic findings.
The final study group included 85 patients with 86 lesions. No patient had a known hereditary syndrome associated with renal tumors. CT was the first examination performed for 44 patients (52%), and MRI was the first examination performed for 41 patients (48%). Of the 86 lesions, 55 were classified as Bosniak category III (64%) and 31 were classified as Bosniak category IV (36%).
Of the 55 Bosniak category III lesions, 39 (71%) were surgically resected, and malignancy was confirmed for 28 (72%), all of which were RCCs of the following subtypes: 19 clear cell carcinomas (68%), five cystic multilocular carcinomas (18%), two papillary tubular clear cell carcinomas (71%), one papillary carcinoma (4%), and one chromophobe carcinoma (4%). Benign status was confirmed for 11 resected lesions (28%): seven epithelial or fibrous cysts with nonneoplastic inflammatory changes (64%), two multilocular cystic nephromas (18%), and two oncocytomas (18%), a particularly rare cystic presentation [30].
Of the 31 Bosniak category IV lesions, 21 (68%) were surgically resected, and malignancy was confirmed for 18 of these 21 lesions (86%), all of which were RCCs of the following subtypes: 12 clear cell carcinomas (67%), five papillary carcinomas (28%), and one tubule papillary clear cell carcinoma (5%). Tubule papillary clear cell carcinoma is a subtype recently incorporated into the kidney tumor classification of the International Society of Urological Pathology and has peculiar morphologic findings, a typical immunohistochemical profile, indolent behavior, and low staging [31]. Benign status was confirmed for three resected lesions (14%): two epithelial or fibrous cysts with nonneoplastic inflammatory changes and one anastomosing hemangioma (Table 1). The anastomosing hemangioma is an unusual neoplasm of vascular origin with benign behavior [32].
Mean lesion size was 4.6 cm (range, 1.1–8.1 cm) for Bosniak category III lesions and 3.5 cm (range, 1.2–22.0 cm) for Bosniak category IV lesions. The mean size of malignant lesions was 4.1 cm, and that for benign lesions was 3.4 cm, with no significant statistical difference noted (p = 0.387).
Malignant lesions had a low Fuhrman grade (1 or 2) in 91% of cases (42/46). Only four lesions were classified as Fuhrman grade 3, and none were classified as Fuhrman grade 4. As for staging, 83% of malignant lesions (38/46) presented as stage pT1a, 7% as pT1b (3/46), and 11% (5/46) as pT2. Pathologic findings according to Bosniak classification are presented in Tables 2 and 3.
Of the 60 surgically treated lesions (39 Bosniak category III and 21 Bosniak category IV), 53 (88%) had imaging follow-up studies performed for at least 2 years after surgical resection (range, 24–84 months; mean, 42 months), and there was no evidence of local recurrence, metastases, or lymph node enlargement. For the group of surgically resected lesions with malignant status (46/60), MRI or CT of the abdomen was performed for 89% of cases (41/46), chest CT for 65% (30/46), head CT for 30% (14/46), and PET/CT for 11% (5/46).
All 26 lesions that were not resected were clinically monitored for at least 2 years (range, 24–96 months; mean, 53 months) and remained stable in size and complexity without signs of local invasion or distant metastasis. CT or MRI of the abdomen was performed for 100% of cases (26/26), whereas chest CT was performed for 50% (13/26), head CT for 50% (13/26), and PET/CT for 12% (3/26).
| Discussion | |
|---|
There is wide variability in the malignancy rates for complex renal cystic lesions reported in studies in the literature (ranging from 25% to 81% for Bosniak category III lesions and from 67% to 100% for Bosniak category IV lesions) [19–27], as is shown in Table 4. Technologic improvements and updates to the Bosniak classification may have played a role in these discrepancies because older studies showed lower malignancy rates (Bosniak category IIF lesions previously had been categorized as category III lesions before the classification update occurred). Variable sample sizes and the different criteria used as surrogates for benign and malignant lesions among studies may also have contributed to the discrepant results.
Most previous studies focused on malignancy rate only, whereas we sought to evaluate histologic grade, initial stage, and, also, progression on follow-up examinations, findings that we consider valuable for prognostic purposes. Bosniak category III and IV lesions were surgically managed in 69% of our cases. Although the overall malignancy rate for surgically resected complex renal cysts was high (77%), most lesions had initial staging and low histologic grades (89% were pathologic tumor stage pT1; 91% were Fuhrman nuclear grade or International Society of Urological Pathology grade 1 or 2), indicating a less aggressive behavior.
Benign lesions were the final diagnosis in 23% of resected lesions, with inflammatory changes in epithelial or fibrous cysts occurring most frequently (64%), most likely representing chronically infected or hemorrhagic epithelial cysts.
In the present study, we observed a high resection rate for Bosniak category III lesions (71%, which was higher than that for Bosniak category IV lesions) and a concomitant high malignancy rate but no evidence of local recurrence, metastases, or lymph node enlargement on imaging follow-up studies. Follow-up of the remaining lesions that were not resected revealed relatively long stability (mean, 53 months). Our findings suggest that a more conservative approach could be considered in certain clinical scenarios (for instance, for patients for whom surgical risks exist) [33], but they also suggest that many stable lesions presumably considered benign are likely malignant with indolent behavior and that there may not be a maximum follow-up time frame in which to consider benignity [34–36].
The present study has some limitations, most of which are related to its retrospective design. First, almost half of the initial cases not submitted for surgery were excluded because of a short follow-up. Second, a minimum of 2 years of follow-up does not necessarily indicate benign behavior; longer follow-up and a larger sample size should be addressed in future studies. Third, the criteria used to select patients for surgery were not evaluated because of the retrospective nature of the study. The experience of the urologist, comorbidities, and patient apprehension may have played a role in the decision and may have induced selection bias. Fourth, Bosniak classification was extracted from radiologic reports from several readers with different levels of expertise, and interobserver variability was not evaluated. However, we believe that the classification system is simple and straightforward, especially for category III and IV lesions.
| Conclusion | |
|---|
Although high malignancy rates were observed for both Bosniak category III and IV lesions, the results of this study suggest that those malignancies are often in an initial stage and have a low histologic grade. Nonresected lesions remained unchanged on follow-up examinations. These findings indicate low aggressiveness of these lesions, supporting the idea of more conservative approaches.
| References | |
|---|
