William W. Mayo-Smith

Incidental Adrenal Lesions: Principles, Techniques, and Algorithms for Imaging Characterization

Incidental adrenal lesions are commonly detected at computed tomography, and lesion characterization is critical, particularly in the oncologic patient. Imaging tests have been developed that can accurately differentiate these lesions by using a variety of principles and techniques, and each is discussed in turn. An imaging algorithm is provided to guide radiologists toward the appropriate test to make the correct diagnosis.

Microwave Ablation of Lung Malignancies: Effectiveness, CT Findings, and Safety in 50 Patients

PURPOSE To retrospectively evaluate effectiveness, follow-up imaging features, and safety of microwave ablation in 50 patients with intraparenchymal pulmonary malignancies. MATERIALS AND METHODS This HIPAA-compliant study was approved by the institutional review board; informed consent was waived. From November 10, 2003, to August 28, 2006, 82 masses (mean, 1.42 per patient) in 50 patients (28 men, 22 women; mean age, 70 years) were percutaneously treated in 66 microwave ablation sessions. Each tumor was ablated with computed tomographic (CT) guidance. Follow-up contrast material-enhanced CT and positron emission tomographic (PET) scans were reviewed. Mixed linear modeling and logistic regression were performed. Time-event data were analyzed (Kaplan-Meier survival estimates and log-rank statistic). All event times were the time to a patients first event (alpha level = .05, all analyses). RESULTS At follow-up (mean, 10 months), 26% (13 of 50) of patients had residual disease at the ablation site, predicted by using index size of larger than 3 cm (P = .01). Another 22% (11 of 50) of patients had recurrent disease resulting in a 1-year local control rate of 67%, with mean time to first recurrence of 16.2 months. Kaplan-Meier analysis yielded an actuarial survival of 65% at 1 year, 55% at 2 years, and 45% at 3 years from ablation. Cancer-specific mortality yielded a 1-year survival of 83%, a 2-year survival of 73%, and a 3-year survival of 61%; these values were not significantly affected by index size of larger than 3 cm or 3 cm or smaller or presence of residual disease. Cavitation (43% [35 of 82] of treated tumors) was associated with reduced cancer-specific mortality (P = .02). Immediate complications included pneumothorax (Common Terminology Criteria for Adverse Events [CTCAE] grades 1 [18 of 66 patients] and 2 [eight of 66 patients]), hemoptysis (four of 66 patients), and skin burns (CTCAE grades 2 [one of 66 patients] and 3 [one of 66 patients]). CONCLUSION Microwave ablation is effective and may be safely applied to lung tumors. (c) RSNA, 2008.

Adnexal masses in pregnancy: Surgery compared with observation

OBJECTIVE: To estimate whether the delay of surgery impacts the risk of adverse maternal and fetal outcomes in patients diagnosed with an adnexal mass during pregnancy. METHODS: A review was performed of pregnant patients diagnosed with an adnexal mass 5 cm or greater in diameter. Data collected included age, gravity/parity, gestational age at diagnosis, and presenting symptoms. Ultrasound examinations were evaluated for mass size and complexity. Pregnancy outcome, complications, and surgical pathology were reviewed. RESULTS: Between 1990 and 2003, 127,177 deliveries were performed at our institution. An adnexal mass 5 cm in diameter or greater was diagnosed in 63 (0.05%) patients. Pathologic diagnosis was available for 59 (94%) patients. The remaining 4 patients were lost to follow-up and excluded from the analysis. Antepartum surgery was performed in 17 patients (29%): 13 because of ultrasound findings that suggested malignancy and 4 secondary to ovarian torsion. The remaining patients were observed, with surgery performed in the postpartum period or at time of cesarean delivery. The majority of masses were dermoid cysts (42%). Four patients were diagnosed with ovarian cancer (6.8% of masses, 0.0032% of deliveries), and one patient (1.7%) had a tumor of low malignant potential. Antepartum surgery due to ultrasound findings that caused concern was performed on all 5 women diagnosed with a malignancy or borderline tumor, compared with 12 (22%) of the patients with benign tumors (P < .01). CONCLUSION: In select cases, close observation is a reasonable alternative to antepartum surgery in patients with an adnexal mass during pregnancy. LEVEL OF EVIDENCE: II-3

Percutaneous Image-guided Thermal Ablation and Radiation Therapy: Outcomes of Combined Treatment for 41 Patients with Inoperable Stage I/II Non–Small-Cell Lung Cancer

PURPOSE To evaluate the clinical outcomes in patients with early-stage non-small-cell lung cancer (NSCLC) after combined treatment with thermal ablation and radiation therapy (RT). MATERIALS AND METHODS Forty-one patients with inoperable stage I/II NSCLC tumors underwent thermal ablation and RT at our institution between 1998 and 2005. Thirty-seven radiofrequency (RF) ablation procedures and four microwave ablation procedures were performed. Ablations were followed by standard-fraction external-beam RT within 90 days (n = 27) or postprocedural brachytherapy (n = 14). Survival and local recurrence were the primary endpoints evaluated by Kaplan-Meier analysis. RESULTS The median follow-up was 19.5 months. The overall survival rates were 97.6% at 6 months, 86.8% at 1 year, 70.4% at 2 years, and 57.1% at 3 years. Patients with tumors smaller than 3 cm (n = 17) had an average survival time of 44.4 +/- 5.4 months (SE). Patients with tumors 3 cm or larger (n = 24) had an average survival time of 34.6 +/- 7.0 months (P = .08). Local recurrence occurred in 11.8% of tumors smaller than 3 cm after an average of 45.6 +/- 4.1 months and in 33.3% of the larger tumors after an average of 34.0 +/- 7.8 months (P = .03). Outcomes in the brachytherapy and RT groups did not differ significantly. Nine of 15 pneumothoraces required chest tube drainage (22.0%). CONCLUSIONS Thermal ablation followed by RT for inoperable stage I/II NSCLC has a relatively low rate of complications that are easily managed. Combined therapy may result in an improved survival compared with either modality alone.

Diagnostic yield of 58 consecutive imaging-guided biopsies of solid renal masses: should we biopsy all that are indeterminate?

OBJECTIVE The purpose of our study was to report the diagnostic yield of 58 consecutive imaging-guided biopsies of solid renal masses. MATERIALS AND METHODS We retrospectively reviewed all percutaneous renal biopsies of solid masses performed at our institution over 83 consecutive months from May 1998 to March 2005 through a query of our radiology department procedure database. Fifty-five CT and three sonographic biopsies were performed at our institution during this time. A solid renal mass was documented prior to biopsy by contrast-enhanced CT (n = 48), gadolinium-enhanced MRI (n = 6), or sonography (solid noncystic masses, n = 4). The average maximal mass diameter was 3.1 cm (range, 1.0-11.0 cm). Forty-seven (81%) of the 58 biopsies were performed immediately before percutaneous ablation. Forty-four (76%) of the biopsies were performed using a coaxial technique with side-cutting automated biopsy needles (16-20 gauge), and 14 (24%) were fineneedle aspirations with a Franseen needle (20 gauge) using a tandem technique. In 19 cases, immunohistochemistry or histochemistry (Hale colloidal iron stain) was used to establish or confirm the diagnosis. Medical records and radiology and pathology reports were reviewed for all patients. RESULTS An adequate sample size was obtained in 55 (95%) of 58 renal masses and led to a definitive diagnosis in 52 (90%) of the 58. Renal cell carcinoma accounted for 36 (69%) of 52 diagnostic biopsies. The diagnosis of a benign lesion was made in 14 (27%) of 52 biopsies. Lymphoma (1/58) and metastatic disease (1/58) accounted for the remaining two diagnostic biopsies. Three biopsy samples obtained inadequate sample volumes, and an additional three samples were thought to have adequate sample volume but were not diagnostic. A single false-negative biopsy result was identified after growth was seen on follow-up imaging and subsequent nephrectomy revealed renal cell carcinoma. CONCLUSION Imaging-guided biopsy of a solid enhancing renal mass was diagnostic in 52 (90%) of 58 consecutive biopsies. The diagnosis of a benign lesion was made in 27% of diagnostic biopsies. Because of the advances in biopsy and histology techniques, the role of imaging-guided biopsy should be reconsidered.

Cytologic Diagnosis of Pancreatic Cystic Lesions

PURPOSE: To prospectively evaluate fine needle aspiration biopsy (FNAB) of pancreatic cystic lesions. STUDY DESIGN: We performed a blind, prospective study on percutaneous aspirates from 28 radiographically identified cysts, including 6 inflammatory cysts (5 pseudocysts and 1 abscess), 4 serous cystadenomas, 1 cystic islet cell tumor, 5 mucinous cystic neoplasms, 6 mucinous cystadenocarcinomas and 6 nonpancreatic cysts. RESULTS: Four of six (67%) cystadenocarcinomas were identified as malignant, and the other two, which lacked sufficient morphologic criteria for malignancy, as consistent with mucinous cystic neoplasm. Two of five mucinous cystic neoplasms were correctly classified. One, which contained atypical cells, did not appear to be mucinous on the ThinPrep, and one, which lacked an epithelial component, was suggested because of the presence of mucin in the background. The fifth one contained inflammatory cells only. One of four serous cystadenomas produced a diagnostic specimen. FNAB of the cystic islet cell tumor was nondiagnostic. Five of six inflammatory cysts (83%) were correctly diagnosed, whereas one case produced an acellular, nondiagnostic specimen. Six of 28 (23%) cases were nonpancreatic cysts, aspirated under the presumption that they were pancreatic cysts based on radiologic studies: only one case, a papillary cystadenocarcinoma of the stomach, was correctly diagnosed; the other five cases were nondiagnostic, and in two of these the assumption that the cysts were pancreatic in origin precluded an accurate classification. CONCLUSION: FNAB of pancreatic cystic lesions can differentiate mucinous from nonmucinous pancreatic cysts and provide definitive evidence of malignancy. In some cases, serous cystadenomas can be diagnosed. Pseudocysts can be suspected on the basis of an inflammatory smear lacking both epithelial cells and background mucin, but this finding is not specific. Nonpancreatic lesions constitute a significant percentage of cases aspirated as pancreatic cysts and present a major pitfall in cytologic interpretation.

MR differentiation of phaeochromocytoma from other adrenal lesions based on qualitative analysis of T2 relaxation times

AIM To determine the diagnostic accuracy of MR imaging in differentiating phaeochromocytoma from other adrenal lesions. MATERIALS AND METHODS Sixty-seven adrenal masses (65 patients) including 17 phaeochromocytomas were imaged using T2-weighted pulse sequences on 0.6 T and 1.5 T GE MR units. The adrenal lesions were qualitatively assessed by three observers and divided into three categories (benign adenomas, malignant lesions and phaeochromocytomas) based on lesion signal intensity relative to liver and cerebrospinal fluid. RESULTS Eleven phaeochromocytomas (65%) were correctly identified while the remaining six (35%) were misclassified, five as malignant lesions and one as a benign adenoma, because of atypical low signal intensity on T2-weighted MR images. Conversely, six non-phaeochromocytomas (three benign adenomas, two adrenal carcinomas and one metastasis) were wrongly classified as phaeochromocytomas because of very high lesion signal intensity. The sensitivity of MR imaging for diagnosing phaeochromocytoma was 64.7%, specificity 88.0%, positive predictive value 64.7% and negative predictive value 88.0%. CONCLUSION There is considerable overlap between the MR appearance of phaeochromocytoma and other adrenal lesions. A phaeochromocytoma cannot be excluded on the basis of a lack of high signal intensity on T2-weighted MR imaging.

FDG PET characterization of renal masses : Preliminary experience

OBJECTIVE This study was undertaken to evaluate the potential efficacy of fluorine-18 2-fluoro-2-deoxy-D-glucose positron emission tomography (FDG PET) to detect renal tumours and to characterize indeterminate renal cysts. SUBJECTS AND METHODS Twenty-six PET scans were performed in 21 patients (14 PET scans in 10 patients with malignant renal tumours and 12 PET scans in 11 patients with Bosniak type 3 indeterminate renal cysts). Pathological proof was obtained in 18 of 21 patients (10 with solid neoplasms, eight with indeterminate cysts). Imaging was performed 1 h after injection of 5-10 mCi of FDG with IV administration of Lasix (10 mg) 20 mins after injection. Two consecutive 9.7-cm image segments were scanned to cover the entire renal areas. RESULTS PET accurately depicted solid neoplasms as areas of increased uptake in nine of 10 patients. Bilateral renal cell carcinomas were missed in one diabetic patient. All but one indeterminate renal cysts were correctly classified as benign (photopenic areas), but an indeterminate cyst with a 4-mm papillary neoplasm was wrongly classified as benign. There were no false positive PET interpretations. The mean tumour-to-kidney ratio was 3.0 for malignant lesions. CONCLUSION We conclude that FDG PET scanning shows promise in the evaluation of indeterminate renal cysts. A positive PET scan in the appropriate clinical setting obviates the need for cyst aspiration. A negative PET scan in conjunction with a negative cyst aspiration offers confirmatory evidence of benignity. Our preliminary results are encouraging and further work is ongoing.

Sensitivity and specificity of eight CT signs in the preoperative diagnosis of internal mesenteric hernia following Roux-en-Y gastric bypass surgery

AIM To evaluate the sensitivity and specificity of eight previously reported computed tomography (CT) signs in diagnosing internal mesenteric hernia following Roux-en-Y gastric bypass surgery. MATERIALS AND METHODS Preoperative CT images of nine patients with surgically proven internal mesenteric hernia as a complication of gastric bypass surgery and 10 matched control patients were reviewed in a blinded fashion by three radiologists. The presence of eight previously reported signs of internal mesenteric hernia was assessed: mesenteric swirl sign, hurricane eye sign, mushroom sign, small bowel obstruction, clustered small bowel loops, small bowel other than duodenum located behind the superior mesenteric artery (SMA), presence of the jejunal anastomosis to the right of the midline, and engorged mesenteric lymph nodes. The sensitivity and specificity were calculated for each sign, as well as inter-observer reliability in recognizing these signs. RESULTS Mesenteric swirl was the most predictive sign of internal hernia (sensitivity 78-100%, specificity 80-90%). Other CT signs showed good specificity (70-100%), but sensitivities were low (0-44%). The presence of a small-bowel obstruction and engorged mesenteric nodes was found to be 100% specific in predicting the presence of an underlying hernia. There was substantial inter-observer agreement in detecting mesenteric swirl sign (kappa=0.48-0.79), but agreement was relatively poor for all other signs. CONCLUSION Mesenteric swirl is an easily recognized CT sign, and is the best indicator of internal hernia following Roux-en-Y gastric bypass surgery. Other reported CT signs are diagnostically insensitive. The presence of small-bowel obstruction with engorged mesenteric nodes is highly specific in diagnosing internal mesenteric hernia.

The CT small bowel faeces sign: Description and clinical significance

OBJECTIVE To describe a new CT sign of the GI tract: the small bowel faeces sign, and discuss its significance. METHODS The small bowel faeces sign consists of gas bubbles mixed with particulate matter in dilated segments of small bowel. This was a retrospective study of 22 patients demonstrating this sign from 1989 to 1993. Final diagnosis was established by surgical, medical or laboratory findings. To determine the prevalence of the sign, the CT examination of 120 consecutive separate control patients were evaluated. RESULTS All 22 of the patients demonstrating this sign required hospitalization with surgical or medical intervention. Eighteen of 22 had mechanical small bowel obstruction. The remaining four patients had other abnormalities of small bowel to account for the finding on CT. Twelve of the 22 patients were treated with surgery and the remaining 10 patients were treated with nasogastric tubes (n = 6) or other medical therapy (n = 4). None of the 120 control patients demonstrated the sign. CONCLUSION The presence of gas and particulate material resembling faeces in a dilated segment of small bowel on CT is abnormal. Most (18/22; 82%) patients with this sign had small bowel obstruction.