Deep learning based classification of solid lipid-poor contrast enhancing renal masses using contrast enhanced CT.

OBJECTIVE: Establish a workflow that utilizes convolutional neural nets (CNN) to classify solid, lipid-poor, contrast enhancing renal masses using multiphase contrast enhanced CT (CECT) images and to assess the performance of the resulting network. METHODS: In this institutional review board approved study of 143 patients with predominantly solid, lipid-poor, contrast enhancing renal lesions (46 benign and 97 malignant), patients with a pre-operative multiphase CECT of the abdomen and pelvis obtained between June 2009 and June 2015 were retrospectively queried. Benign renal masses included oncocytoma and lipid-poor angiomyolipoma and the malignant group included clear cell, papillary, and chromophobe carcinomas.Region of interests of whole tumor volumes were manually segmented, and CT phase images with the largest cross-section of the segmented tumor in the axial plane were used for assessment. Post-surgical pathological evaluation was used to establish diagnosis.The segmented images of renal masses were used as input to a CNN. The data were augmented and split into training (83.9%) and validation sets (16.1%) to determine the hyperparameters of the CNN. Thereafter. the performance of the resulting CNN was quantified using eightfold cross-validation. RESULTS: The CNN-based classifier demonstrated an overall accuracy of 78% (95% confidence interval: 76-80%), sensitivity of 70% (95% confidence interval: 66-74%), specificity of 81% (79-83%) and an area under the curve of 0.82. CONCLUSION: A CNN-based classifier to diagnose solid enhancing malignant renal masses based on multiphase CECT images was developed. ADVANCES IN KNOWLEDGE: It was established that a CNN-based classifier could be trained to accurately distinguish malignant renal lesions.

Renal angiomyolipoma with tuberous sclerosis complex: How it differs from sporadic angiomyolipoma in both management and care.

Renal angiomyolipoma (AML) is the most common benign tumor of the kidney. It consists of blood vessels, smooth muscle and fat components in varying proportions. AML is divided into the sporadic type and tuberous sclerosis complex (TSC)-associated type. TSC-associated AML develops at a younger age and tends to exhibit a much faster growth rate over time than sporadic AML. AMLs are classified as classic AML, fat-poor AML and epithelioid AML. Epithelioid AML, though rare, shows aggressive behavior leading to distant metastasis and mortality. TSC-associated AML is more likely to have an epithelioid component than sporadic AML. Active surveillance is the suggested management for small AML. Clinical intervention is mainly indicated when there is a substantial risk of rupture. Minimally invasive therapies, including partial nephrectomy, transcatheter arterial embolization, and mammalian target of rapamycin (mTOR) inhibitor treatment are employed for patients who require treatment. An updated algorithm for the management of AML is herein described. According to this algorithm, treatment intervention is recommended for TSC-associated AML >3 cm, even in asymptomatic cases. In cases with asymptomatic sporadic AML >4 cm in size or with an intra-tumoral aneurysm of >5 mm, treatment, including transcatheter arterial embolization or partial nephrectomy, is advised. The major complication of AML is intra-tumoral or retroperitoneal hemorrhage due to rupture that may be serious and life threatening. Thus, correct diagnosis, proper observation, and appropriate treatment are very important in the management of renal AML.

Radiomics of small renal masses on multiphasic CT: accuracy of machine learning-based classification models for the differentiation of renal cell carcinoma and angiomyolipoma without visible fat.

OBJECTIVE: To investigate the discriminative capabilities of different machine learning-based classification models on the differentiation of small (< 4 cm) renal angiomyolipoma without visible fat (AMLwvf) and renal cell carcinoma (RCC). METHODS: This study retrospectively collected 163 patients with pathologically proven small renal mass, including 118 RCC and 45 AMLwvf patients. Target region of interest (ROI) delineation, followed by texture feature extraction, was performed on a representative slice with the largest lesion area on each phase of the four-phase CT images. Fifteen concatenations of the four-phasic features were fed into 224 classification models (built with 8 classifiers and 28 feature selection methods), classification performances of the 3360 resultant discriminative models were compared, and the top-ranked features were analyzed. RESULTS: Image features extracted from the unenhanced phase (UP) CT image demonstrated dominant classification performances over features from other three phases. The two discriminative models "SVM + t_score" and "SVM + relief" achieved the highest classification AUC of 0.90. The 10 top-ranked features from UP included 1 shape feature, 5 first-order statistics features, and 4 texture features, where the shape feature and the first-order statistics features showed superior discriminative capabilities in differentiating RCC vs. AMLwvf through the t-SNE visualization. CONCLUSION: Image features extracted from UP are sufficient to generate accurate differentiation between AMLwvf and RCC using machine learning-based classification model. KEY POINTS: • Radiomics extracted from unenhanced CT are sufficient to accurately differentiate angiomyolipoma without visible fat and renal cell carcinoma using machine learning-based classification model. • The highest discriminative models achieved an AUC of 0.90 and were based on the analysis of unenhanced CT, alone or in association with images obtained at the nephrographic phase. • Features related to shape and to histogram analysis (first-order statistics) showed superior discrimination compared with gray-level distribution of the image (second-order statistics, commonly called texture features).

New radiologic classification of renal angiomyolipoma: frequently asked questions.

Recently, a new classification of renal angiomyolipoma (AML) has been introduced to clinical practice. This classification categorizes AMLs into three subtypes (fat-rich, fat-poor, and fat-invisible AMLs) based on quantitative CT and MRI criteria. Radiologists who get used to previous AML classification may have questions about how to apply a new classification. The purpose of this review is to answer the questions that are frequently asked about the new AML classification.

Clinical behavior and management of three types of renal angiomyolipomas.

OBJECTIVES: To analyze and present the demography, clinical behavior, especially the risk factors of tumor hemorrhage and management of sporadic angiomyolipoma (SAML), tuberous sclerosis complex associated angiomyolipoma (TSCAML) and epithelioid angiomyolipoma (EAML) in our institution. METHODS: A retrospective study of 587 patients who were diagnosed with renal angiomyolipoma in our institution between January 2000 and May 2015 was done. The AMLs were diagnosed by ultrasonography, CT, or MRI. EAML was confirmed by histopathology. Medical records and follow-up results were analyzed using the SPSS version 22 software. RESULTS: Out of 587 cases of renal AMLs, 87.4% were SAMLs, 8.7% were TSCAMLs and 3.9% were EAMLs. Most of the AML patients were asymptomatic. The most common presenting symptoms included flank pain and abdominal pain. The median tumor size of SAML, TSCAML, EAML were 4.7, 2.7, 10.5 cm respectively. Approximately half of SAMLs were conservatively treated, almost all TSCAMLs were treated conservatively, while all EAMLs were surgically treated. The median tumor size of hemorrhagic SAML cases was 8 cm versus non-hemorrhagic cases of 4.1 cm. The optimal cut-off point on the ROC curve for predicting SAML tumor hemorrhage was 7.35 cm. CONCLUSION: A larger tumor size, younger patient's age and higher BMI value correlated with a higher risk of tumor hemorrhage. For tumor sizes less than 7.35 cm, we recommend active surveillance or TAE for hemorrhage prevention. We also suggest that surgical management should be considered for patients with tumors larger than 7.35 cm, symptomatic and progressive AML, or suspicious EAML.

Renal Angiomyolipoma: Radiologic Classification and Imaging Features According to the Amount of Fat.

OBJECTIVE: The purposes of this article are to introduce the radiologic classifications of renal angiomyolipoma (AML) and the clinical implications, to show the imaging features of each type of AML, and to describe which types of AML should be biopsied. CONCLUSION: Renal AML can be classified according to amount of fat as fat rich, fat poor, or fat invisible. To detect fat, one needs to thoroughly evaluate the entire AML by controlling the size and shape of the ROI. Fat-invisible AML should be biopsied, and fat-poor AML requires further investigation to determine whether biopsy is necessary to differentiate it from renal cell carcinoma. If differentiation between AML and renal cell carcinoma is not clear with CT and MRI, percutaneous biopsy may be performed.

Lipomatous tumours in adrenal gland: WHO updates and clinical implications.

Adrenal lipomatous tumour is a group of adrenal tumours with a significant component of adipose tissue. According to the current World Health Organization (WHO) classification of tumours of endocrine organs, adrenal myelolipoma is the only entity amongst the group of tumours being described. In the literature, other more recently documented adrenal lipomatous tumours included 24 lipomas, 32 teratomas and 16 angiomyolipomas. Rare fatty tumours of the adrenal gland comprised liposarcoma, hibernoma, adrenocortical tumours with fat component and rare adrenal tumours with fat component. Myelolipoma comprises approximately 3% of primary adrenal tumour. It is noted more commonly in females and in the right adrenal gland. Approximately 40 bilateral myelolipomas were reported. The tumour is most frequently recorded in patients between fifth and seventh decades of life. Adrenal lipomas are often seen in males and in the right adrenal gland. They were commonly noted in patients in the sixth decade of life. The diagnosis could only be possible on examination of the surgically removed specimen. Adrenal teratomas were more common in females and with a bimodal age distribution. Slightly over 60% of the patients with adrenal teratoma are symptomatic. Adrenal angiomyolipomas were often symptomatic, more common in females and in the fifth decades of life. To conclude, adrenal lipomatous tumour is uncommon. They are often benign and non-functional. It is important to recognize the features of this group of lipomatous tumours in the adrenal gland as they are being detected on increasing incidence as a result of the wide-spread use of modern imaging modalities.

Renal angiomyolipomas: At least two diseases. A series of patients treated at two European institutions.

OBJECTIVE: The aim of this study is to analyze the outcome of renal angiomyolipomas (AML) at two European institutions. METHODS: The data were collected from patients with a primary AML who were treated at Gustave Roussy, Villejuif, France and Fondazione IRCCS Istituto Nazionale dei Tumori in Milan, Italy from 1998 to 2014. The specimens were classified as classic AML (C AML) or epithelioid AML (E AML) based on the percentage of epithelioid cells. RESULTS: There were 40 patients identified for the study (35 C AML, 5 E AML). One patient had an associated tuberous sclerosis complex. Six patients (15%) had bilateral AML. The imaging results were significantly different between C/E AML. E AML was associated with fewer bilateral lesions, more renal vein/vena cava extension, and more poor or non-fatty aspects. Surgery/active surveillance (AS)/chemo radiation were applied for 28/11/1 patients, respectively. The median tumor size was significantly smaller (3.75 cm) in patients receiving AS (median 15 cm when surgically resected). The median patient follow-up was 43 months. The three-year overall survival was significantly better for patients with C AML than E AML (100% versus 50%, p < 0.0001). The univariate analysis identified the OS prognostic factors were E AML histologic subtype (p < 0.001), poor/non fatty features (p = 0.002), and renal vein extension on imaging (p = 0.01). CONCLUSION: AML manifests as at least two different entities with significantly different outcomes. Epithelioid subtype, poor/non-fatty features, and renal vein involvement are all associated with worse survival.

Renal leiomyoma: a contemporary multi-institution study of an infrequent and frequently misclassified neoplasm.

Renal leiomyoma is an exceptionally rare benign mesenchymal tumor of the kidney predominantly arising in proximity of the renal capsule or pelvis. Its rarity and nonspecific clinical and imaging features may lead to radical or partial nephrectomy on the basis of preoperative suspicion of renal cell carcinoma. The diagnosis of renal leiomyoma is challenging because of the histologic overlap with lipid-poor angiomyolipoma (AML). We conducted a multi-institution study to characterize renal leiomyoma in greater detail. We collected and reviewed 24 cases diagnosed initially as renal leiomyoma in 10 institutions from North America, Canada, and Europe. Immunohistochemical expression of desmin, HMB-45, estrogen receptor (ER), progesterone receptor (PR), and cathepsin K was evaluated. Upon central review, 9 tumors were classified as renal leiomyoma, whereas the remaining were reclassified as AML (n=13), myolipoma (n=1), and medullary fibroma (n=1). All renal leiomyomas were solitary and occurred in female patients (mean age 63 y; range, 44 to 74 y). Tumor size ranged from 0.6 to 7.0 cm (mean 2.9 cm); 7 originated from the renal capsule or the subcapsular area and 1 from a large vessel in the renal sinus. All leiomyomas were diffusely positive for desmin and negative for HMB-45 and cathepsin K; 6/9 (67%) showed diffuse ER and PR expression, and 1 case showed focal ER positivity only. Renal leiomyoma should be included in the histologic differential diagnosis of solid renal masses, particularly in perimenopausal women. The main differential diagnosis is with lipid-poor AML, and cathepsin K plays a key role in distinguishing these 2 lesions.

Ten uncommon and unusual variants of renal angiomyolipoma (AML): radiologic-pathologic correlation.

Classic (triphasic) renal angiomyolipoma (AML) is currently classified as a neoplasm of perivascular epithelioid cells. For diagnosis of AML, the use of thin-section non-contrast enhanced CT (NECT) improves diagnostic accuracy; however, identifying gross fat within a very small AML is challenging and often better performed with chemical-shift MRI. Although the presence of gross intra-tumoural fat is essentially diagnostic of AML; co-existing intra-tumoural fat and calcification may represent renal cell carcinoma (RCC). Differentiating AML from retroperitoneal sarcoma can be difficult when AML is large; the feeding vessel and claw signs are suggestive imaging findings. AML can haemorrhage, with intra-tumoural aneurysm size >5 mm a more specific predictor of future haemorrhage than tumor size >4 cm. Diagnosis of AML in the setting of acute haemorrhage is complex; comparison studies or follow-up imaging may be required. Not all AML contain gross fat and imaging features of AML without visible fat overlap with RCC; however, homogeneity, hyperdensity at NECT, low T2-weighted signal intensity and, microscopic fat are suggestive features. Patients with tuberous sclerosis often demonstrate a combination of classic and minimal fat AML, but are also at a slightly increased risk for RCC and should be imaged cautiously. Several rare pathological variants of AML exist including AML with epithelial cysts and epithelioid AML, which have distinct imaging characteristics. Classic AML, although benign, can be locally invasive and the rare epithelioid AML can be frankly malignant. The purpose of this review is to highlight the imaging manifestations of 10 uncommon and unusual variants of AML using pathological correlation.

Time to abandon the term angiomyolipoma for non-PEComatous angiomyomatous (or angiomatous) oral tumors with adipocytes.

Angiomyolipoma (AML) is the most common benign mesenchymal neoplasm of the kidneys with well-established clinical and morphological features. The oral and maxillofacial pathology literature contains several examples that identify angiomyomatous proliferations of the oral mucosa that contain an adipocytic component as analogous to classic renal AMLs although they differ significantly in their immunohistochemical phenotype. Herein, through review of the pertinent oral pathology literature and the detailed description of 2 lesions, one an oral angiomyoma with an adipocytic component and the other an apparently hamartomatous angioleiomyomatous proliferation with adipocytes, we provide, in our opinion, a solid argument against the use of the term AML for non-PEComatous oral tumors.

Hepatic angiomyolipoma mimicking hepatic clear cell carcinoma.

Angiomyolipoma usually involves the kidney and rarely affects the liver. This study reports a case of angiomyolipoma of the liver in a 47-year-old Chinese woman. The patient did not present with abdominal pain and jaundice. Imaging showed a small mass in the right liver. The hepatic artery and portal vein were free from invasion. Partial hepatectomy was performed after a tentative diagnosis of clear cell carcinoma of the liver by needle biopsy. Histopathological examination of the resected specimen revealed angiomyolipoma originating in the liver. The post-operative clinical course was uneventful and, at the time of writing, the patient was well with no signs of recurrence 6 months after operation. To our knowledge this is the first documented case of an angiomyolipoma arising in the liver mimicking hepatic clear cell carcinoma.

Primary perivascular epithelioid cell tumor of the liver not related to hepatic ligaments: hepatic PEComa as an emerging entity.

Primary perivascular epithelioid cell tumor (PEComa) of the liver is a very rare example of an emerging family of hepatic PEC tumors. Only few cases have been described so far. We report the case of a large but benign hepatic PEComa in a 53-year-old man without signs of tuberous sclerosis. In contrast to recently described PEC-derived liver tumors in children and young adults, this neoplasm was not related to the hepatic ligaments but had developed deeply within the liver substance. The neoplastic cells displayed the complete phenotype typical for PEComas, i.e. reactivity for several melanoma markers and for smooth muscle actin. The unique relationship of myoid tumor cells to the adventitia of blood vessels prompted us, in comparison with published findings obtained with angiomyolipomas, to comment on the possible origin of the still enigmatic perivascular epithelioid cells.

Clinical correlates of renal angiomyolipoma subtypes in 209 patients: classic, fat poor, tuberous sclerosis associated and epithelioid.

PURPOSE: Angiomyolipomas classically present radiographically as fat containing lesions but some fail to demonstrate fat content. Histologically confirmed angiomyolipomas uniformly follow a benign course but rare epithelioid variants of angiomyolipoma can recur and metastasize. We investigated the clinical, radiographic and histological characteristics of each angiomyolipoma subtype. MATERIALS AND METHODS: Pertinent data were recorded for 209 patients surgically treated for angiomyolipoma in 219 kidneys from 1981 to 2007. Classic and fat poor angiomyolipomas were classified radiographically based on the presence or absence of fat and classified histologically based on the presence of triphasic, monophasic or epithelioid histology. RESULTS: Median radiographic size was 3.2, 4.9 and 10 cm in patients with a single angiomyolipoma (59% of patients), multiple angiomyolipomas and tuberous sclerosis (probable or definite), respectively. In these 3 groups 65%, 47% and 33% of lesions were not suspected radiographically (fat poor angiomyolipoma). Fat poor angiomyolipomas were more commonly single, smaller and in older patients. Triphasic histology was evident in 76% of angiomyolipomas with 16% demonstrating a predominance of 1 component and 8% containing epithelioid features. Despite potentially aggressive findings in 18% (eg presence within the perinephric fat, lymph node involvement) no angiomyolipoma recurred during a mean followup of 3.4 years (range 0 to 24). A total of 28 (13%) patients with angiomyolipoma had concomitant renal cell carcinoma. CONCLUSIONS: A surprisingly high number of resected angiomyolipomas was not suspected radiographically indicating the importance of precise radiographic characterization to minimize nephrectomy for fat poor angiomyolipoma, which should remain a research priority. In this sizeable single institution series no triphasic, monophasic or epithelioid angiomyolipoma recurred despite potentially aggressive findings in a substantial proportion of cases.

Pathological features of renal neoplasms classified by size and symptomatology.

PURPOSE: We examined the relationship between tumor size and pathological findings in a contemporary series of surgical renal lesions and we characterized the relationship of incidental and symptomatic tumors to pathological findings. MATERIALS AND METHODS: We retrospectively reviewed the records of patients treated surgically for renal lesions suspicious for malignancy between March 2000 and May 2005. Specimens were examined for a gross and microscopic description. Statistical analysis was used to determine the correlation of size and symptomatology. RESULTS: A total of 349 renal masses from 331 patients were identified. Of the 349 renal masses 56 (16.0%) were benign, 289 (82.8%) were renal cell carcinoma and 4 (1.1%) were other malignancies. The percent of malignant tumors increased from 72.1% for those less than 2 cm to 93.7% for those greater than 7 cm (OR 1.39, 95% CI 1.17 to 1.65). Of the 349 renal masses 258 (73.9%) were discovered incidentally and 91 (26.1%) were symptomatic. Mean size of incidental and symptomatic tumors was 3.7 and 6.2 cm, respectively (p < 0.001). When comparing T1 incidental and symptomatic tumors, there was no significant difference in the overall frequency of malignancy. When comparing T2 incidental and symptomatic tumors, the groups had similar malignancy rates (90.9% and 100%, respectively, p = 0.16). However, symptomatic lesions showed an increased incidence of high grade malignancy (78.4% vs 40.9%, p = 0.012). CONCLUSIONS: Smaller renal tumors are more likely to be benign or be a lower grade of malignancy. T1 renal tumors are more likely to be detected incidentally than T2 tumors. When T1 incidental and symptomatic tumors were compared, there was no difference between the malignancy rates. However, when T2 incidental and symptomatic tumors were compared, symptomatic tumors were more likely to be high grade malignancy.

Controversies related to epithelioid variant of renal angiomyolipoma: a review of the literature.

We performed a review of the epithelioid-type angiomas to determine their clinical course and propose an appropriate plan for follow-up. We present the case of a patient with possible adrenal carcinoma suggested by computed tomography for whom the histopathologic study revealed an epithelioid angiomyolipoma arising from the kidney. In the absence of consensus, we consider it worthwhile to register the very few cases diagnosed and record a detailed follow-up of the clinical course. We recommend surgical treatment and a follow-up regimen similar to that for renal carcinoma.

Cutaneous angiomyolipomas are HMB45 negative, not associated with tuberous sclerosis, and should be considered as angioleiomyomas with fat.

Angiomyolipomas are uncommon tumors in the skin. Neoplasms with the same name are more typically associated with a renal location, many in the context of tuberous sclerosis. Three cases of cutaneous angiomyolipoma are reported, which were all HMB45 negative and not associated with tuberous sclerosis. This is the first series of such tumors to be stained with HMB45. The presence of fat in histologically similar cutaneous angioleiomyomas has been previously documented and a review of 40 archival cases revealed fat in 1 case. The author proposes that cutaneous angiomyolipomas should be termed angioleiomyoma with fat to avoid confusion with unrelated non-cutaneous angiomyolipomas that are HMB45 negative and require investigation for tuberous sclerosis.

[Morphologic variants and immunohistochemical features of hepatic angiomyolipoma].

OBJECTIVE: To study the clinicopathological characteristics, immunohistochemical features and differential diagnosis of hepatic angiomyolipoma (AML). METHODS: The clinicopathological features of hepatic AML were systematically examined in 44 surgically resected tumor specimens, with additional immunohistochemistry study using 10 relevant antibodies. RESULTS: The tumors were composed of various amounts of three components, i.e. blood vessels, smooth muscle cells and adipose cells. According to the proportions of each of these tissue components, AML was subcategorized into the classical type (n = 13), myomatous type (n = 25), lipomatous type (n = 4), and angiomatous type (n = 2). Myoid cells displayed various morphology, including epithelioid, intermediate (ovoid or short spindle), spindle, oncocytic, and pleomorphic features. Hematopoietic elements were present as minor findings in eight tumors. Immunohistochemically, the tumor cells were strongly positive for HMB45 (44/44, 100%), SMA (38/38, 100%) and CD117 (30/38, 78.9%). CONCLUSIONS: A correct diagnosis of hepatic AML might be difficult due to its various growth patterns and cell types. HMB-45 positivity in the myoid cells is a key feature for hepatic AML. CD117 may be another useful ancillary marker for reaching a definite diagnosis.