Genitourinary Imaging Original Research

Transcription

1 Genitourinary Imaging Original Research Patel et al. CT of Synchronous Renal Masses Genitourinary Imaging Original Research Uday Patel 1 Navin Ramachandran James Halls Aneeta Parthipun Catherine Slide Patel U, Ramachandran N, Halls J, Parthipun A, Slide C Keywords: differentiating features, incidence metastases, renal masses DOI: /AJR Received January 18, 2011; accepted after revision May 30, All authors: Department of Radiology, St. George s Hospital and Medical School, Blackshaw Rd, London, SW17 0QT United Kingdom. Address correspondence to U. Patel (uday.patel@stgeorges.nhs.uk). WEB This is a Web exclusive article. AJR 2011; 197:W680 W X/11/1974 W680 American Roentgen Ray Society Synchronous Renal Masses in Patients With a Nonrenal Malignancy: Incidence of Metastasis to the Kidney Versus Primary Renal Neoplasia and Differentiating Features on CT OBJECTIVE. The purpose of this study was to establish the contemporaneous frequency of metastases within the kidney as opposed to primary renal tumors in patients with an active primary nonrenal malignancy and to identify the differentiating features. MATERIALS AND METHODS. We retrospectively identified all patients with an active primary nonrenal malignancy (group 1) who had also undergone at least 2 contrastenhanced abdominal CT examinations spaced 1 year apart. The radiologic and pathologic data of these cases were reviewed and the incidence of metastasis to the kidney versus primary renal tumors established. These data were compared with a separate group who presented with primary renal malignancy from the outset (group 2). RESULTS. In the study were 2340 patients with primary nonrenal malignancy (group 1) and 231 patients with a primary renal malignancy (group 2). For group 1, the mean age was 63 years and 51% were men; for group 2, the mean age was 59 years, and 58% were men. The differences were not statistically significant. Thirty-six patients in group 1 had a malignant renal mass; 21 were a result of kidney metastasis and the remaining 15 were a synchronous primary renal tumor (0.9% vs 0.6%). The kidney was the eighth most common site of metastatic spread. Metastases to the kidney were statistically more likely with higher tumor stage of the primary nonrenal malignancy (68% vs 46%, p = ) and in those with other sites of metastasis (p = 0.012, positive likelihood ratio [LR+] = 6.75). Compared with primary renal tumors, metastases to the kidney were more often solid (86% vs 53%, p = 0.019, LR+ = 3.7) and endophytic (76% vs 33%, p = 0.017, LR+ = 2.29). There were too few cases with calcification and bilateral tumors to reach a statistically significant conclusion. Tumor size, polar predominance, and enhancement pattern were similar in the two groups. The primary renal tumors seen in group 1 versus group 2 were similar regarding age and sex distribution, cell type, median size, and tumor stage. CONCLUSION. Metastases to the kidney are uncommon in modern radiologic practice (0.9%, 21/2340 in this study), and a renal mass seen in a patient with nonrenal malignancy is nearly as likely to be an incidental primary renal tumor. Metastasis is more likely in those with higher tumor stage or if other viscera are also affected and is usually an asymptomatic, small, endophytic, and solid mass. If a renal mass seen in a patient with primary nonrenal malignancy proves to be a synchronous primary renal tumor, its cell type and stage will be similar to sporadic primary renal tumors. A n incidental solid or semisolid renal mass in a patient without renalspecific symptoms can be reliably diagnosed on CT or MRI appearances alone [1]. Renal cell cancer will account for the majority. Angiomyolipoma can be excluded if there is macroscopic fat on an unenhanced study [2]. The remainder will be due to other cell types, such as oncocytoma, fat-poor angiomyolipoma, upper tract transitional cell cancers, or further rare renal tumor subtypes. Metastasis to the kidney is a mimic of primary renal tumor. However, it rarely presents as an incidental renal mass and in current practice radiologic appearances are sufficiently reliable to plan therapy for the incidentally detected renal mass, with biopsy reserved for selected cases [3]. Diagnostic evaluation can be challenging if a solid or complex cystic renal mass is seen in a patient with an active nonrenal malignancy [4]. The renal tumor may be either a metastatic focus or an incidental synchronous primary renal tumor. There are currently only W680 AJR:197, October 2011

2 CT of Synchronous Renal Masses sparse data on the imaging of kidney metastases. For example, their contemporaneous frequency (outside the autopsy room) is not known, and the distinguishing features have not been thoroughly described. Some investigators have argued that radiologic differentiation is impossible [5]. This distinction is clinically important because it influences treatment options. This study was carried out to establish how often this problem arises and whether these two entities can be radiologically differentiated. The primary aim of this study was to establish the contemporaneous frequency of synchronous malignant renal masses in patients under active cancer care for a nonrenal malignancy. The secondary aims were to establish the percentage from kidney metastasis versus a synchronous primary renal tumor and to define any discriminating features. In the rest of this article, the term kidney metastasis implies metastasis to the kidney from a malignancy of nonrenal origin. Materials and Methods Study Design A 2-year ( ) retrospective longitu dinal observational study was carried out in a 1000-bed institution, serving a population of approximately 1.3 million. A study waiver was obtained from the local institutional ethics committee because this was a retrospective study with no change in patient management. Study Population Adult patients (> 16 years) referred for CT staging of an active malignancy were identified by electronic interrogation of the radiology inform ation system. The presence of an active nonrenal malignancy was confirmed by manual review of clinical data. Discrepancies or data gaps were resolved by referral to clinical case notes. The inclusion cri teria were patients with pathologically proved non renal malignancy who were referred for cancer staging to the CT department. To ensure sufficient longitudinal follow-up data, only those who had undergone at least two IV contrast-enhanced ab dominal CT examinations (at least 1 year apart) that fully encompassed the kidneys were included. Patients who could not undergo a contrast-enhanced CT study (e.g., iodine allergy) were excluded. Tumor type and location were not considered. Further data were gathered by reference to the proceedings of a weekly urology-oncology multidisciplinary team meeting. This committee has been tasked to advise on the management of any suspected renal tumors identified in the hospital. All renal masses or abnormalities that are suspected to be malignant by any reporting radiologist in our department are highlighted and the report is forwarded to the committee. At the meeting, all radiologic and clinical data are reviewed and a management plan is formed, including the necessity for further imaging or biopsy. The results of these discussions are available for future review. One of the authors is a member of this team and maintains a contemporaneous record of the discussions and outcome of all suspected malignant renal masses seen. To further ensure capture of all relevant malignant renal masses, data were supplemented by electronic interrogation of the hospital pathology database, which is an institutional record of the histologic outcome of all patients with renal tumors who have undergone either surgical removal or renal biopsy. Identified renal masses, whether metastases or primary renal tumors, were split into two groups: patients who presented with a nonrenal malignancy and were found to have a malignant renal mass (group 1) and those who presented with a primary renal tumor from the beginning (group 2). CT Studies CT was performed on a 16- or 64-MDCT scanner (LightSpeed Ultra or LightSpeed VCT, respectively, GE Healthcare). Patients in group 1 had undergone abdominal CT after IV contrast injection of 100 ml of iohexol 300 mg/ml (Omnipaque, GE Healthcare), given at a rate of 3 ml/s. The scans were timed for the portal venous phase (65 seconds after injection) and reconstructed at a slice thickness of 2.5 mm. The patients who presented with a renal mass (group 2) had undergone a renal specific CT protocol: an unenhanced study followed by a contrastenhanced study at 90 seconds (similar contrast agent and volume as group 1); both studies were reconstructed at 2.5-mm collimation. To ensure uniform study comparison, only the contrastenhanced axial studies in either group were compared. Some patients had undergone additional studies, e.g., unenhanced or delayed studies of the abdomen, but these were ignored. Data Collection CT scans were reviewed on a PACS. The presence of a true enhancing solid or complex cystic (Bosniak type 3 or 4 cyst) renal mass was confirmed on double review, carried out by two uroradiologists with 3 and 15 years of experience. Both were blinded to the pathology of the renal masses and clinical details of the patient. To further reduce bias, only those images showing the kidneys were presented to the readers without details regarding scanning times. First, the initial scan was reviewed; if this showed a renal tumor, the 12-month scan was evaluated and the two compared. The diagnosis of kidney metastasis was made on histologic evidence (extirpative or biopsy) or the progression or improvement of the renal abnormality in concert with the patient s nonrenal malignancy. Patients with renal masses that did not change underwent imaging-guided biopsy after con firmation of the suspicious nature with further imaging, such as ultrasound, MRI, or a dedicated renal mass protocol CT. A primary renal tumor was diagnosed on histology alone. The pathologic type of the renal mass and the CT characteristics of the renal and nonrenal TABLE 1: Demographics of Patients With Primary Nonrenal Malignancy (Group 1) and Primary Renal Malignancy (Group 2) Parameter Group 1 Group 2 No. of patients Mean age (range) 63 (20 94) 59 (21 89) M:F ratio 1193: :95 Tumor types a Lymphoma 561 (24) Clear cell RCC 137 (59) Colorectal 328 (14) Papillary RCC 23 (10) Lung 304 (13) Chromophobe RCC 11 (5) Melanoma 307 (13) Eosinophilic RCC 3 (1.3) Breast 280 (12) Oncocytoma 21 (9) Female pelvic 210 (9) Angiomyolipoma 2 (1) Esophagus 94 (4) UTTCC 18 (8) Prostate 70 (3) Others 11 (5) Others 186 (8) Note Except where indicated otherwise, data in parentheses are percentages. RCC = renal cell carcinoma, UTTCC = upper tract transitional cell carcinoma. a Tumor type refers to the presenting malignancy (the nonrenal malignancy for group 1 and the primary renal tumor for group 2). AJR:197, October 2011 W681

3 Patel et al. malignancy were separately investigated by the other three authors. In masses with disseminated malignancy, the metastatic load was estimated by noting the number of organ systems involved. Each renal mass was evaluated for size (maximal axial dimension was recorded in accordance with our standard practice), tumor characteristics (whether solid or necrotic-cystic), global enhancement pattern (subjectively determined to be either homogeneous or heterogeneous), and presence of calcification. Because group 1 did not undergo unenhanced studies, only the contrast-enhanced studies were used to evaluate calcification. The mean density of the solid elements of the renal mass was measured in Hounsfield units and its position determined (left or right kidney; upper pole, interpolar, or lower pole; and whether the mass was exophytic or endophytic). An exophytic mass was one that protruded > 50% outside the kidney. Data Analysis Data were entered into a single database and the incidence of kidney metastasis was calculated. In the second part of the study, the demographics and imaging features of kidney metastasis were compared with primary renal cancers. Intergroup comparison was made using continuous and categoric statistics. A commercially-available statistics program was used (Medcalc, version , Medcalc Software). For all analysis, a p value < 0.05 was taken to imply statistical significance. Measurements were tested for normality of distribution before statistical testing, and the statistical tests used are specified. Results There were 2340 patients with a primary nonrenal malignancy on active therapy or surveillance (group 1) who fulfilled the entry criteria. A further 231 patients had presented with a primary renal malignancy (group 2). All patients had sufficient radiologic and clinical data, and there were no further exclusions. The demographic details of the two study groups are given in Table 1, and the groups are matched for age and sex distribution. The distribution of tumor types in group 1 reflects the referral patterns of our hospital, which is a center for the management of many tumors but particularly lymphoma and melanoma. Thus, these two tumor types may be over-represented. The histologic distribution of primary renal tumors (group 2) is similar to that described in the literature [6]. Ninety-six patients in group 1 had a renal abnormality in their initial staging CT and were discussed at the weekly meeting. After complete review of their radiologic history, 36 patients had masses that fitted the criteria for malignant renal masses. The rest had nonmalignant abnormalities, e.g., transient enhancement changes, infarction, hyperdense (Bosniak type 2) cyst, parenchymal calcification, and others. In all 36 cases, the renal mass was asymptomatic, and all patients had undergone at least two CT studies of the abdomen at least 1 year apart. In every patient, the mass was present on both studies. On review of the histologic records and follow-up studies, 21 of 2340 patients (0.9%) had a renal mass that fulfilled the criteria for kidney metastasis, whereas 15 of 2340 (0.6%) had a synchronous primary renal tumor. Metastasis was diagnosed on histology in 12 of 21 cases and on follow-up imaging (criteria used to diagnose metastasis as described earlier) in the remainder. The demographic and imaging features of these two subgroups are listed in Table 2. The kidney was the eighth most common site of metastasis. The common sites of metastasis TABLE 2: Outcome and Comparison of 36 Malignant Renal Masses Found in 2340 Patients With a Primary Nonrenal Malignancy Parameter Kidney Metastasis Incidental Primary Renal Tumor No. of masses Mean patient age (range) 61 (26 88) 56 (36 82) M:F ratio 11:10 9:6 Primary nonrenal malignancy (No.) Lymphoma (6) Colon(5) Lung (4) Bladder (3) Melanoma(4) Leukemia (2) Adenocarcinoma (2) Lung (2) Endometrial (1), colon (1) Myeloma (1) Thyroid (1) Ovarian (1) Gastric (1) Larynx (1) Stage of primary nonrenal malignancy T1 4, N0, M a T1 4, N1 2, M0 1 0 T1 4, N0, M T1 4, N1 2, M Mean no. of sites of metastasis (range) 2.4 (1 6) 1.1 (0 2) b Median renal tumor size (SD, 95% CI) 4 (5, ) 4.5 (3.3, ) 0.66 b Renal tumor characteristics Solid or necrotic:cystic 19:2 8: a Enhancement pattern c 10:11 6: a Calcification a Bilateral tumors 2 d a Median renal mass density (HU) (SD, 95% CI) 70 (21, 56 83) 62 (21, 57 77) 0.71 b Renal tumor location Left:right 13:8 7: a Upper pole a Interpolar 11 2 Lower pole 5 6 Exophytic:endophytic 5:16 10: b a Fisher s test or chi-square test as appropriate. b Unpaired Student t test or Mann-Whitney U test as appropriate. c Homogeneous versus heterogeneous enhancement pattern. d Both patients had multiple bilateral tumors. p W682 AJR:197, October 2011

4 CT of Synchronous Renal Masses were lymph nodes, lungs, liver, bones, adrenals, peritoneum, pleura, and kidney in that order. The three most common primary sites of metastasis to the kidney were lymphoma (6/21), lung (4/21), and melanoma (4/21). When the imaging and clinical characteristics of patients with primary nonrenal malignancy with kidney metastasis were compared with patients with primary nonrenal malignancy patients with a synchronous primary renal tumor, kidney metastases were found to be more common in those with a higher stage of nonrenal malignancy (68% in the kidney metastasis group had at least T1 4, N1 2, M1 2 stage disease vs 46% in the primary renal tumor group (p = ) (Table 2). Patients with kidney metastasis were also more likely to show other visceral metastasis (p = , positive likelihood ratio [LR+] = 6.75) and a heavier metastatic load as measured by the number of organ systems affected (mean, 2.4 systems involved vs 1.1, respectively, p = 0.008). Kidney metastases were also statistically more likely to be solid (86% vs 53%, respectively, p = 0.019, LR+ = 3.7). The enhancement pattern or the degree of enhancement was not significantly different between the two groups, and there were too few patients with calcification and bilaterality to reach a meaningful conclusion (Table 2). Regarding tumor location, the only differentiator was that metastases were more likely to be endophytic (76% vs 33%, p = 0.017, LR+ = 2.29). In those patients with a primary renal tumor in group 1, the most common cause of the presenting nonrenal malignancy was colon cancer (5/15), and 3 of 15 had bladder cancer. The clinical and CT characteristics of primary renal tumors in the two groups are compared in Table 3. There are no statistically significant differences in overall demographics or the tumor types, stages, and CT appearances. Discussion This study shows that the contemporaneous antemortem frequency of kidney metastasis is low (21/2340 patients, 0.9%). This figure is in contrast to (older) autopsy data that have reported frequencies of between 7% and 20% [7 10]. A further general finding is that in current practice a renal mass in a patient with a nonrenal malignancy is nearly as likely to be a primary renal tumor (21/2340 vs 15/2340, 0.9% vs 0.6% for kidney metastasis and primary renal tumor, respectively) in contrast to a previous study that found metastasis to be four times more frequent [5]. Figures 1 and 2 are two examples of synchronous primary renal tumors. It is important to know whether a renal mass identified in patient with an active nonrenal malignancy is a primary renal tumor or a metastasis because it directly influences the treatment of the renal tumor. Our study identifies three helpful differentiating signs on CT: the consistency of the renal mass, its location, and the overall burden of metastatic disease. A solid endophytic mass in a patient with multiple sites of metastasis is more likely to be a metastasis (Figs. 3 5) rather than a primary renal tumor (LR+, 6.75 vs 2.29, respectively). Other tumor characteristics were not helpful discriminators. Previous studies have observed either a lack of or poor enhancement in kidney metastasis [4, 5]. In our study, kidney metastasis had a median density of 70 HU after IV contrast administration, and this was not significantly different from primary renal tumors (median density, TABLE 3: Pathologic Outcome and Comparison of Primary Renal Tumors in Patients With and Without a Concurrent Nonrenal Malignancy Parameter With Primary Nonrenal Malignancy Without Primary Nonrenal Malignancy No. of patients Mean age (range) 56 (36 82) 59 (21 89) M: F ratio 9:6 136:95 Renal tumor types Clear cell RCC 10 (66%) 137 (59%) Papillary RCC 2 (13%) 23 (10%) Chromophobe RCC 0 11 (5%) Eosinophilic RCC 1 (7%) 3 (1.3%) Oncocytoma 1 (7%) 21 (9%) Angiomyolipoma 0 2 (1%) TCC 1 (7%) 18 (8%) Others 0 11 (5%) Renal tumor classification a T1 7 (47%) 117 (51%) 0.25 b T2 6 (40%) 79 (34%) T3a 2 (13%) 8 (3.4%) T3b 0 19 (8%) T3c T4 0 8 (3.4%) Median renal tumor size (SD, 95% CI) 4.5 (3.3, ) 5.3 (3.4, ) 0.7 c Tumor characteristics Solid or necrotic:cystic 8:7 132: b Enhancement pattern a 6:9 110: b Calcification b Median renal mass density (HU) (SD, 95% CI) 62 (21;57 77) 81 (27;74 87) 0.71 c` Renal tumor location Left:right 7:8 104:126 1 b Upper pole c Interpolar Lower pole 6 54 Exophytic:endophytic 10:5 158: b Note RCC = renal cell carcinoma, TCC = transitional cell carcinoma. a Tumor stage on histologic examination. b Fisher s test or chi-square test as appropriate. c Unpaired Student t test or Mann-Whitney U test as appropriate. p AJR:197, October 2011 W683

5 Patel et al. 62 HU). It is also worth noting that all cases of kidney metastasis were seen in patients with a known nonrenal malignancy, that the kidney was not the sole site of metastasis in any patient, and metastasis did not occur in group 2. We also found that synchronous primary renal tumors are similar to the general pattern of renal tumors seen in modern practice [6]. The majority are clear cell renal carcinoma, the minority being benign. Most are small (median axial dimension, 4.5 cm) early-category tumors (87% were category T2) and thus potentially curable [6]. There are no modern autopsy series devoted to kidney metastasis [8 11]. The most complete earlier autopsy study was by Abrams et al. [9] who reported that 12.6% of patients who died of disseminated carcinoma between the years 1943 and 1947 had metastasis to the kidney. The kidney was the tenth most common target of metastatic spread and thus nearly similar to its antemortem rank in our study. Tumors reported to more frequently metastasize to the kidney are lymphoma, lung, breast, melanoma, stomach, cervix, colon, pancreas, prostate, and the contralateral kidney [7]. This was largely confirmed in another autopsy study of 81 cases with kidney metastasis in which the three most common primary tumors were lung, stomach, and breast [10]. As noted earlier, the reported incidence of kidney metastasis is higher in the postmortem series. It is possible that this reflects smaller lesions identified on autopsy that are beyond the resolution of CT; however, we have not been able to confirm (or refute) this because these autopsy studies do not provide any size data. Regarding radiologic series, previous studies were selected case series with advanced clinical stage and may not accurately reflect modern experience. For example, 7 of 12 patients with adequate follow-up in one study died within 3 months [9]. We were able to identify only one previous study [12] concerning incidence of kidney metastasis in radiologic practice, which found only two cases in 1000 patients with malignancy. Sánchez- Ortiz and coworkers [4] reported on 100 patients with a nonrenal malignancy and a renal mass. This was a selected study and not designed to assess the incidence of kidney metastasis, but similar to our study, they found that the most common causes were lymphoma and lung metastasis and that the likelihood of kidney metastasis in a patient with a clinically localized nonrenal malignancy was low. Bilateral kidney metastases (Fig. 6) have Fig year-old woman who was diagnosed with chronic lymphocytic lymphoma. Staging CT image shows irregular mass lesion lying centrally in left kidney. CT-guided percutaneous biopsy confirmed grade 2 renal cell carcinoma, which is currently under surveillance. Fig year-old man who was diagnosed with rectal carcinoma. Initial staging CT image shows partially exophytic enhancing cystic lesion within interpolar region of right kidney (arrow). Rectal carcinoma was locally confined on staging MRI (not shown), and renal mass was thought to be synchronous primary renal tumor. Anterior resection of rectal cancer confirmed localized disease. Patient underwent subsequent partial nephrectomy, and histology confirmed clear cell renal cell carcinoma. Fig year old woman with biopsy-confirmed non small cell lung cancer. Staging CT image identifies metastatic deposit within hepatic segment VI (black arrow) and ill-defined endophytic enhancing lesion in left kidney (white arrow). Metastases to kidney were found to be more commonly endophytic in our study. Patient underwent chemotherapy for non small cell lung cancer with reduction in size of both lesions. W684 AJR:197, October 2011

6 CT of Synchronous Renal Masses commonly featured in previous studies (42% [5], 50% [11], 80% [13], 100% [14], and 40% [15]). Bilaterality was not a statistically significant finding in our study (2/21, 9.5%), but the sample size was small and the disease less advanced (Fig. 6). It is possible that we have underestimated the diagnostic value of bilaterality as an indicator of kidney metastasis because our figure (9.5%) is still higher than the reported frequency of synchronous primary renal tumors (1 5%) [5]. A study such as ours raises the possibility of an association between renal cell cancer and other malignancies, and this has been previously speculated. A relationship between A Fig year-old man diagnosed with malignant melanoma. A, Initial staging CT image shows probable metastatic disease anterior to inferior vena cava (short arrow) and small low-density lesion within left kidney (long arrow). B, Follow-up CT image obtained 3 months later shows increase in size of both retroperitoneal disease and renal mass. Unlike in this example, metastasis to kidney was more commonly endophytic in our study (Fig. 3). renal cell cancer and lymphoma (particularly non-hodgkin lymphoma) is suspected [16, 17] as are more tentative links with melanoma [18] and rectal cancer [19]. Conversely, other data suggest that renal carcinoma itself confers a higher risk of a second malignancy. In a study of 1425 patients with renal cell cancer followed for 7 years [20], 257 developed or had a history of a second malignancy: prostate, bladder, lung, breast, colorectal, melanoma, and lymphoma the most common and 53 of 257 were synchronous presentations. Other studies have also alluded to this finding. Rabbani et al. [21] reported that prostate, breast, colon, bladder, and Fig year-old man with metastatic melanoma. Coronal CT image shows pulmonary metastases (white arrow), hepatic metastases (black arrow), and small lesion in upper pole of left kidney (curved arrow). All lesions progressed in congruity on subsequent imaging. non-hodgkin lymphoma were the five most common cancer types seen with renal cell cancer, either synchronously or later. In our study, colon and bladder cancer were most commonly seen with a synchronous primary renal neoplasia, but our study does not necessarily support an association between renal and other cancers because the incidence of primary renal tumors in group 1 (0.6%) is not dissimilar to the estimated prevalence of asymptomatic renal carcinoma (estimated range, %) [22]. There are some limitations to our study. First, we did not have pathologic confirmation of metastasis to the kidney in all our cases; instead, we used the less-robust indicator of radiologic progression or regression in concert with the primary lesion. To reduce selection bias, we recruited all patients presenting with malignancy over the study period, but some sources of bias could not be controlled. Because we are a tertiary referral center, some tumor types (and possibly overall tumor stage) may be overrepresented in our hospital population, e.g., melanoma. Conversely, those with advanced metastasis beyond therapy may be underrepresented because they may not be referred to our institute or may not have lived long enough for a second CT examination. The study design will also have underestimated the frequency of kidney metastasis in that we excluded those patients who did not undergo a follow-up CT study at 1 year because the nature of their renal mass was not determinable. Evaluation of certain tumor characteristics was also not B AJR:197, October 2011 W685

7 Patel et al. Fig year-old man with non-hodgkin lymphoma. Staging CT image shows multiple bilateral renal masses that improved with lymphomaspecific chemotherapy. Bilateral disease was not found to be useful discriminator between metastatic or primary renal masses in our study. ideally performed. Calcification was assessed only on contrast-enhanced scans because these studies were consistently available. It is possible that subtle calcification may have been masked by contrast enhancement, and we may have underestimated the value of calcification as a diagnostic sign. Enhancement was not a useful discriminator, but we only measured the contrast-enhanced renal mass density because we did not have unenhanced studies in group 1 and also we compared 65-second contrast-enhanced scans in group 1 to scans obtained at 90 seconds in group 2. Thus, we were not able to study early enhancement, a feature more commonly seen with clear cell carcinoma. This may have helped with diagnostic differentiation. Even having recruited a large study population, the number of kidney metastasis in our study is small, resulting in possible measurement error. We may have overestimated the given likelihood ratios or conversely failed to detect small but real differences with other potentially useful diagnostic signs, e.g., bilaterality. Finally, these findings may be particular to our patient population and may not be generally applicable. Further data from other centers would be helpful. To summarize, in patients with an active nonrenal malignancy, the kidney was found to be the eighth most common site of metastatic spread, but the incidence of kidney metastasis was low. When seen on routine contrast-enhanced CT, metastasis to the kidney typically presents as an asymptomatic, small, solid, and homogeneously-enhancing endophytic renal mass. The frequency of primary renal tumors was not dissimilar to that expected from the background prevalence of incidental renal tumors. However, if the prognosis of the presenting nonrenal malignancy is good, biopsy correlation is advised because those with a primary renal tumor will prove to have an early-stage and treatable primary renal malignancy. References 1. Israel GM, Bosniak MA. How I do it: evaluating renal masses. Radiology 2005; 236: Simpson E, Patel U. Diagnosis of angiomyolipoma using computed tomography-region of interest < or = 10 HU or 4 adjacent pixels < or = 10 HU are recommended as the diagnostic thresholds. Clin Radiol 2006; 61: Silverman SG, Gan YU, Mortele KJ, Tuncali K, Cibas ES. Renal masses in the adult patient: the role of percutaneous biopsy. Radiology 2006; 240: Sánchez-Ortiz RF, Madsen LT, Bermejo CE, et al. A renal mass in the setting of a nonrenal malignancy: when is a renal tumor biopsy appropriate? Cancer 2004; 101: Choyke PL, White EM, Zeman RK, et al. Renal metastases: clinicopathologic and radiologic correlation. Radiology 1987; 162: Russo P. Partial nephrectomy for renal cancer: part 1. BJU Int 2010; 105: Meilstrup JW, Mosher TJ, Dhadha RS, Hartman DS. Other renal tumors. Semin Roentgenol 1995; 30: Bracken RB, Chica G, Johncon DE, Luna M. Secondary renal neoplasms: an autopsy study. South Med J 1979; 72: Abrams H, Spira R, Goldstein M. Metastases in carcinoma: analysis of 1000 autopsied cases. Cancer 1950; 3: Klinger ME. Secondary tumors of the genitourinary tract. J Urol 1951; 65: Wagle DG, Moore RH, Murphy GP. Secondary carcinomas of the kidney. J Urol 1975; 114: Pagani JJ. Solid renal mass in the cancer patient: second primary renal cell carcinoma versus renal metastasis. J Comput Assist Tomogr 1983; 7: Nishitani H, Onitsuka H, Kawahira K, et al. Computed tomography of renal metastases. J Comput Assist Tomogr 1984; 8: Mitnick JS, Bosniak MA, Rothberg M, et al. Metastatic neoplasm to the kidney studied by computed tomography and sonography. J Comput Assist Tomogr 1985; 9: Bhatt GH, Bernardino ME, Graham SD Jr. CT diagnosis of renal metastasis. J Comput Assist Tomogr 1983; 7: Tihan T, Filippa DA. Coexistence of renal cell carcinoma and malignant lymphoma: a casual relationship or coincidental occurrence. Cancer 1996; 77: Ohsawa M, Hashimoto M, Yasunaga Y, et al. Characteristics of non-hodgkin s lymphoma complicated by renal cell malignancies. Oncology 1998; 55: Schmid-Wendtner MH, Baumert J, Wendtner CM, et al. Risk of second primary malignancies in patients with cutaneous melanoma. Br J Dermatol 2001; 145: Cullinane CA, Jarrahy R, Wilson TG, Ellenhorn JD. Asymptomatic renal neoplasms in the rectal cancer patient. Am Surg 2001; 67: Beisland C, Talleraas O, Bakke A, Norstein J. Multiple primary malignancies in patients with renal cell carcinoma: a national population-based cohort study. BJU Int 2006; 97: Rabbani F, Grimaldi G, Russo P. Multiple primary malignancies in renal cell carcinoma. J Urol 1998; 160: Fenton JJ, Weiss NS. Screening computed tomography: will it result in overdiagnosis of renal carcinoma. Cancer 2004; 100: W686 AJR:197, October 2011