Imaging of spindle cell lipoma.

AIM: To review the evaluation, diagnosis, and treatment of spindle cell lipoma (SCL) with emphasis on the location of these tumours and the spectrum of magnetic resonance imaging (MRI) and computed tomography (CT) appearances. MATERIALS AND METHODS: The MRI and CT findings of 27 histopathologically proven SCLs were evaluated retrospectively. Imaging features evaluated included margins, percentage visible fat, MRI signal characteristics, oedema, and contrast enhancement patterns. RESULTS: Patient ages ranged from 18 to 80 years with an average age of 56.5 years. Men were affected twice as frequently as women (M=18, F=9). SCLs ranged in size from 2 to 10 cm, with an average greatest dimension of 5.5 cm. Five lesions (19%) contained no visible fat on CT or MRI, and the leading differential diagnosis of high-grade soft-tissue sarcoma diagnosis was suggested by referring surgeons. Five lesions (19%) had <50% fatty areas, nine lesions (52%) demonstrated >50% but <90% fat at MRI or CT. Only three of 25 lesions (12%) had an appearance of a typical lipoma on unenhanced MRI sequences. All SCLs that were imaged with contrast medium (n = 18) demonstrated some degree of enhancement, with eight (44%) showing marked enhancement, four (22%) showing moderate, and six (33%) minimal enhancement. CONCLUSION: SCLs have considerably variable imaging appearances and may have minimal or no visible fat at MRI or CT. Imaging features may make it difficult to distinguish this benign tumour from a potentially higher-grade malignant tumour.

CT in the selection of patients with abdominal or pelvic sarcoma for reoperative surgery.

BACKGROUND: Retroperitoneal or visceral sarcoma may recur with disease limited to the abdomen and pelvis. In this clinical situation, further surgical treatments in an attempt to control the disease may be appropriate. CT is used to help select patients for additional surgical interventions. STUDY DESIGN: Preoperative abdominal and pelvic CT scans of 33 patients with recurrent sarcoma were reviewed retrospectively. All patients underwent reoperative surgery and, when appropriate, perioperative intraperitoneal chemotherapy. Patients were divided into two groups according to survival and disease status: alive with no evidence of disease (n = 7) and alive with disease or dead of disease (n = 26). Twenty-two CT indices were studied retrospectively for each patient and evaluated statistically. RESULTS: The presence of large (greater than 5 cm) tumor volume in 3 of the 13 abdominopelvic regions resulted in a significant difference in the prognosis between the groups of patients. These findings included tumor in the left lower quadrant (p = 0.032), tumor in the pelvis (p = 0.008), and tumor in the distal jejunum (p = 0.032). Two other CT indices that showed a significant difference in survival between the groups were involvement of five abdominopelvic regions or fewer (p = 0.008) and a peritoneal cancer index of 15 or less (p = 0.03). A statistical approach using a tree-structured diagram showed that patients with tumor diameter greater than 5 cm in the pelvis accompanied by tumor involvement of more than one segment of small bowel had a 0% probability of postoperative disease-free survival. In contrast, patients with tumor diameter less than 5 cm in the pelvis on CT had an 86% probability of disease-free survival. CONCLUSIONS: For patients with recurrent sarcoma, selection criteria were generated by a preoperative CT of the abdomen and pelvis. In this disease, CT was a reliable diagnostic test for predicting benefit from further surgical interventions and should be used in the future to help select patients for an aggressive versus a palliative approach.

Biopsy of musculoskeletal tumors. Current concepts.

Biopsy is a key step in the diagnosis of bone and soft tissue tumors. An inadequately performed biopsy may fail to allow proper diagnosis, have a negative impact on survival, and ultimately necessitate an amputation to accomplish adequate margins of resection. Poorly performed biopsy remains a common finding in patients with musculoskeletal tumors who are referred to orthopaedic oncology centers. The principles by which an adequate and safe biopsy of musculoskeletal tumors should be planned and performed are reviewed, and the surgical approach to different anatomic locations is emphasized.

Muscle infarction in patients with diabetes mellitus: MR imaging findings.

PURPOSE: To describe the magnetic resonance (MR) imaging findings in diabetic patients with muscle infarction and to describe commonly associated clinical features. MATERIALS AND METHODS: The MR imaging studies of 21 patients with diabetic muscle infarction were reviewed retrospectively. Of the 21 patients, 12 were women, and nine were men; the mean age was 48 years (range, 30-77 years). RESULTS: Eight patients had bilateral lower-extremity involvement; six had involvement confined to the right lower extremity and seven to the left. The thigh was involved in 17 patients (81%). One or more of the musculi vastus, the most frequently affected muscle group, were affected in 16 patients (76%). Four patients (19%) had isolated calf involvement. MR imaging studies showed diffuse enlargement of involved muscle groups and partial loss of normal fatty intermuscular septa. MR imaging also allowed identification of areas of subfascial fluid in 16 patients (76%) and subcutaneous edema in 19 patients (90%). MR imaging showed involved muscle groups best with T2-weighted, inversion-recovery, and gadolinium-enhanced sequences, where the infarcted muscles appeared diffusely hyperintense compared with adjacent muscles. Comparison of T2-weighted and gadolinium-enhanced MR images of nine patients showed enlarged, enhancing muscles in all patients and small, focal, rim-enhancing fluid collections in six of nine patients (66%). CONCLUSION: Diabetic muscle infarction is suggested in diabetic patients with sudden onset of severe pain in the thigh or calf muscles who have MR imaging findings of diffuse edema and swelling of multiple thigh and calf muscles (often in more than one compartment).

Begnign bone tumors of childhood.

The diagnosis of a bone tumor in a child can be a source of great anxiety for the patient, the parents, and the treating physician. Fortunately, most bone tumors in children are benign. Although there are a variety of benign bone tumors that affect skeletally immature patients, most have such characteristic clinical and radiographic presentations that the diagnosis can be made with reasonable accuracy without a biopsy. However, some benign bone tumors can simulate a malignant process and may be best handled by referral to a person trained in orthopaedic oncology for additional evaluation. Treatment alternatives are in part related to the Musculoskeletal Tumor Society stage of the lesion. Recurrences of certain lesions, such as aneurysmal bone cysts and osteoblastomas, can be problematic. By becoming familiar with the presentation of the more common benign bone tumors in children, physicians will be able to alleviate fears, establish a diagnosis, and make treatment recommendations in the most effective manner.

Sclerotic lesions of the cervical spine in sarcoidosis.

Sarcoidosis is a common multisystem disorder characterized by noncaseating epithelial granulomata, with osseous involvement typically seen in 5% of patients. While the lace-like or cystic pattern frequently seen in radiographs of the phalanges is well appreciated, sclerotic lesions of the spine are uncommon. We review a case of sarcoidosis of the cervical spine with sclerotic changes that mimicked blastic metastatic disease.

Parosteal osteosarcoma: value of MR imaging and CT in the prediction of histologic grade.

PURPOSE: To evaluate the use of magnetic resonance (MR) imaging and computed tomography (CT) for predicting the histologic grade of parosteal osteosarcomas. MATERIALS AND METHODS: Sixty parosteal osteosarcomas were analyzed for tumor size and location, presence of a cleavage plane, intramedullary extension, soft-tissue mass (distinct from ossified mass), and the presence and pattern of ossification. Axial and longitudinal views were evaluated for specific osseous sites within the bone. Tumors were classified as low grade (grade 1) or high grade (grades 2-3). RESULTS: There were 32 low-grade lesions and 28 high-grade lesions. Average maximal lengths of low- and high-grade tumors were 7.7 and 15.0 cm, respectively. A cleavage plane was present in 20 (62%) low-grade and 19 (68%) high-grade lesions. On cross-sectional images, intramedullary extension was present in 13 (41%) low-grade and 14 (50%) high-grade lesions. A focal soft-tissue mass distinct from the ossific matrix was identified in 25 (89%) high-grade lesions and in only two (6%) low-grade lesions. All 17 high-grade lesions evaluated with MR imaging were of predominantly high signal intensity on T2-weighted images. CONCLUSION: A poorly defined soft-tissue component distinct from the ossific matrix is the most distinctive feature of high-grade parosteal osteosarcoma and may be an optimal site for biopsy.

Percutaneous transpedicular biopsy of vertebral body lesions.

STUDY DESIGN: This prospective study evaluates the use of transpedicular biopsy in obtaining diagnostic tissue from vertebral body lesions. OBJECTIVE: To report the authors' experience of all (N = 32) percutaneous transpedicular biopsies performed between 1990-1994. SUMMARY OF BACKGROUND DATA: Previous articles have discussed the value of open biopsy of the vertebral body using a Craig needle. A large series of closed percutaneous transpedicular biopsies have not been reported. METHODS: The authors evaluated 32 patients (26 outpatients, six inpatients) who underwent transpedicular biopsy for T1-L4 lesions of the vertebral bodies. None of the tumors had an extraosseous component. Biopsy specimens were obtained from 25 lesions using C-arm fluoroscopy; seven were guided by computed tomography. All biopsies were performed with a 14- to 17-gauge bone biopsy needle. RESULTS: The needle passed through the pedicle into the site of disease in all patients, as confirmed by C-arm fluoroscopy or computed tomography. There were 22 malignancies; four isolated compression fractures, two at T6, one at T7, one at T8; four cases of infection or inflammation; and one case each of Paget's disease and myelofibrosis. Two patients required a second biopsy because the tissue sample was suspicious for lymphoma but not diagnostic. All 26 outpatients were discharged after a 2-hour observation period. There were no complications. CONCLUSION: Transpedicular biopsy of deep vertebral body lesions using a bone biopsy needle under computed tomography or fluoroscopy guidance can be performed safely and efficaciously as an outpatient procedure.

Radiology of peritoneal carcinomatosis.

In summary, the radologic appearance of peritoneal carcinomatosis and sarcomatosis is best understood with a thorough knowledge of the natural history of the disease. Computed tomography (CT) examinations of the abdomen and pelvis are insensitive to small volumes (< 5 cm) of diffuse peritoneal seeding. However, when larger volumes of peritoneal tumor are present, CT can be helpful defining the patients who are good candidates for complete cytoreduction (no small bowel disease) from those who are less likely to have a complete cytoreduction (extensive small bowel disease with clumping and obstruction). Thus CT examination plays a critical role in the identification of patients with mucinous tumors who are operative candidates. Further research is needed to improve sensitivity and in the monitoring of recurrence in patients with gastrointestinal cancer.

Abdominal computed tomographic scan in the selection of patients with mucinous peritoneal carcinomatosis for cytoreductive surgery.

BACKGROUND: Cytoreductive surgery and intraperitoneal chemotherapy have been used to treat peritoneal carcinomatosis. A complete surgical resection is required for optimal results to be achieved. This study evaluated the preoperative computed tomographic (CT) findings in patients with mucinous peritoneal carcinomatosis in order to predict the probability of a complete resection. STUDY DESIGN: Computed tomographic scans of the abdomen and pelvis were reviewed retrospectively in 45 patients with a diagnosis of mucinous peritoneal carcinomatosis who were treated with surgery and intraperitoneal chemotherapy. According to the completeness of cytoreduction, patients were divided into two groups. Patients in the first group (n = 25) had complete cytoreduction (CR) with no tumor deposits 2.5 mm in diameter or larger left behind. The surgical resection of tumor was incomplete in the second group of patients (n = 20). Sixteen CT parameters were initially examined in each group of patients and statistically evaluated according to the completeness of the cytoreductive surgical procedure. RESULTS: The incidences of six CT findings were significantly different in the two groups of patients. These findings were: tumor volume in small bowel mesentery (p < 0.001), tumor volume in proximal jejunum (p = 0.003), tumor volume in distal jejunum (p = 0.002), tumor volume in proximal ileum (p = 0.003), mesentery configuration (p < 0.001), and obstruction of bowel segments by tumor (p < 0.001). A statistical approach using a tree-structured diagram showed that patients with both obstruction of bowel segments by tumor and tumor diameter greater than 0.5 cm on small bowel surfaces exclusive of distal ileum on preoperative CT scan, had an 88 percent probability of incomplete resection. Patients without these two CT findings had a 92 percent probability of complete resection. CONCLUSIONS: This study shows that selection criteria for patients with mucinous peritoneal carcinomatosis are available on a preoperative CT scan of the abdomen and pelvis. Patients whose scans show obstruction of bowel segments by tumor and tumor diameter greater than 0.5 cm on small bowel surfaces exclusive of distal ileum are unlikely to be candidates for cytoreductive surgery for the treatment of peritoneal carcinomatosis.

Treatment of large subchondral tumors of the knee with cryosurgery and composite reconstruction.

The radiographic and clinical outcomes of 9 patients with large benign aggressive or low-grade malignant periarticular tumors of the knee who were treated with cryosurgery and composite reconstruction (cementation, bone graft, and internal fixation) in lieu of primary resection were analyzed. The minimum followup was 2 years (range, 24-103 months). There were 6 giant cell tumors and a single case each of chondroblastoma, chondrosarcoma, and fibrosarcoma in the study group. Six lesions involved the distal femur and 3 the proximal tibia. Functional outcomes were graded according to the Enneking Modified System for the Functional Evaluation of Tumor and The Knee Rating Scale of The Hospital for Special Surgery. All lesions extended to within 2 mm of the articular surface. Local tumor control was achieved in 8 patients (89%). The single local recurrence was successfully managed with repeat curettage, cryosurgery, cementation, and internal fixation. All 9 patients had excellent functional outcomes according to both evaluation scales.

MR imaging of soft-tissue masses: role of gadopentetate dimeglumine.

To assess the effectiveness of gadopentetate dimeglumine in the magnetic resonance (MR) imaging evaluation of soft-tissue masses without osseous involvement, 30 patients underwent MR imaging before and after administration of contrast material (0.1 mmol/kg). Of the 30 lesions, 22 were benign and eight were malignant; histologic confirmation was available in all lesions except one benign lesion. Overall, enhancement was detected in 26 (87%) of 30 lesions: 18 (82%) of the 22 benign lesions and eight (100%) of eight malignant lesions. Enhancement was characterized as homogeneous (two [11%] benign lesions, two [25%] malignant lesions), inhomogeneous (11 [61%] benign lesions, six [75%] malignant lesions), or peripheral (five [28%] benign lesions, no malignant lesions). Of the 19 lesions assessed for a change in enhancement over time, seven (37%) showed an increase and two (11%) showed a decrease in signal intensity. The authors conclude that benign and malignant soft-tissue lesions could not be differentiated solely on the basis of enhancement (pattern, degree, or time course).

Giant cell tumor of the tendon sheath: MR findings in nine cases.

OBJECTIVE: The purpose of this study was to determine the MR findings in patients with giant cell tumors of the tendon sheath. MATERIALS AND METHODS: MR imaging findings in nine surgically proved cases of giant cell tumors of the tendon sheath were evaluated on both T1- and T2-weighted images. Of nine lesions, three involved the foot, two involved the thumb, two involved the knee, one involved the proximal part of the tibia, and one involved the proximal part of the femur. RESULTS: All nine lesions were hypointense on T1-weighted images; the signal intensity of most of the tumors was approximately equal to that of skeletal muscle. On the T2-weighted images, three lesions were hypointense relative to skeletal muscle, two lesions were approximately isointense relative to skeletal muscle, and two lesions were slightly hyperintense relative to skeletal muscle but hypointense relative to fat. The remaining two lesions had a more heterogeneous appearance on T2-weighted images. CONCLUSION: On both T1- and T2-weighted images, giant cell tumor of the tendon sheath has a signal intensity similar to that of its pathologic counterpart, pigmented villonodular synovitis. The decreased signal intensity on both T1- and T2-weighted images is an uncommon appearance of extraarticular soft-tissue masses, in particular when they occur in the hands or feet, and this may suggest the diagnosis of giant cell tumor of the tendon sheath.

Fluoroscopically guided peritoneal catheter placement for intraperitoneal chemotherapy.

PURPOSE: Surgical placement of intraperitoneal catheters for intraperitoneal chemotherapy is associated with bowel perforation, peritonitis, and catheter occlusion. The authors evaluated the safety and efficacy of fluoroscopically guided placement of temporary intraperitoneal catheters for chemotherapy. PATIENTS AND METHODS: Two hundred one intraperitoneal catheter placements were attempted in 88 patients with peritoneal carcinomatosis or sarcomatosis. The peritoneum was punctured with 22-gauge needles and exchange was made with use of Seldinger technique and liberal injections of contrast material at each step for 8.3- or 8.5-F multiple-side-hole catheters. Placement sites included all four quadrants and the midline. Computed tomographic (CT) peritoneography was performed prior to chemotherapy. RESULTS: One hundred ninety (94.5%) of 201 attempted catheter insertions were technically successful. Results of CT peritoneography were available in 175 cases and showed free distribution of peritoneal contrast material in 39% (n = 69), partial loculation in 38% (n = 67), and extensive loculation in 22% (n = 39). Catheters remained in place for a median of 5 days (range, 2-6 days). Significant complications occurred in 11 procedures (5.5%). There were seven unintended bowel intubations; all were treated conservatively except one that required surgical repair. One other patient developed necrotizing fasciitis requiring surgical débridement. Three other patients (1.5%) developed mild peritonitis responsive to antibiotics. Technical success, complications, and peritoneal distribution of contrast material did not correlate with the site of catheter placement. CONCLUSION: Percutaneous catheter placement with use of small-gauge needles for initial puncture is safe and efficacious in patients requiring short-term peritoneal access for chemotherapy.

Protein S deficiency: imaging findings.

Protein S deficiency is an autosomal-dominant inherited disorder of coagulation associated with spontaneous and recurrent venous thrombosis. Approximately 5% of patients with deep venous thrombosis of the lower extremities who are less than 45 years old have a deficiency of protein S. Patients frequently have spontaneous and recurrent deep venous thrombosis and pulmonary embolism during young adulthood. Thrombosis occurs less commonly in the splanchnic and cerebral veins. Arterial thrombosis is rare but is seen in the cerebral circulation. Radiologists should include protein S deficiency in the differential diagnosis of unexplained thrombosis in young patients. This pictorial essay illustrates the range of imaging findings encountered in patients with this disorder.

Evaluation of computed tomography in patients with peritoneal carcinomatosis.

BACKGROUND: Radiologic tests to determine the size and location of cancer on peritoneal surfaces are needed for patient management and for clinical research activity. METHODS: Computed tomography (CT) of the abdomen and pelvis were obtained on 45 patients with a diagnosis of peritoneal carcinomatosis before implementation of a complete exploratory laparotomy with biopsy or resection of all tissues suspicious of malignancy. CT findings and surgical findings were recorded and the data analyzed by dividing the abdomen into 9 regions or into 15 anatomic sites. RESULTS: In an overall patient analysis, the sensitivity was 79%. In the nine different abdominal regions, sensitivity was greater than 80% in the right flank, left flank, and right lower and left lower abdominal regions. Sensitivity was lowest in the pelvis (60%). Negative predictive value in the pelvis was 20%. The volume of tumor present within an abdominal region or anatomic site greatly influenced the sensitivity. A sensitivity of only 28% was recorded when tumor nodules were less than 0.5 cm in thickness. This increased to 90% when nodule thickness was greater than 5 cm. CONCLUSIONS: CT is not a reliable diagnostic test for low volume tumor on peritoneal surfaces, and the greatest inaccuracies were recorded in the pelvis. Its diagnostic value increased as tumor volume increased.

Dedifferentiated liposarcoma of the extremities: imaging findings in four patients.

OBJECTIVE: The term "dedifferentiated" liposarcoma is used to describe a distinctive lesion in which a well-differentiated liposarcoma is juxtaposed with a high-grade nonlipogenic sarcoma. Dedifferentiated liposarcoma is probably the most common of all the dedifferentiated sarcomas, occurring almost exclusively in the mediastinum, the retroperitoneum, and the inguinal/paratesticular regions. We report the imaging findings in four cases of dedifferentiated liposarcoma of the lower extremities. MATERIALS AND METHODS: The radiologic images and clinical histories of four patients with histologically verified dedifferentiated liposarcoma were retrospectively studied. The mean age of the patients was 61 years (range, 33-79 years). All lesions occurred in the thigh. Spin-echo MR images were available for review in two cases and CT scans in the two remaining cases. Plain radiographs were available in all cases. RESULTS: All lesions were large, with a mean maximum size of 24 cm (range, 12-30 cm), and had a significant fatty component radiologically. Both MR and CT showed a well-delineated fatty component of the lesions and a closely apposed nonfatty region. Radiographs showed well-defined bone within one lesion, bone and amorphous calcification in another, and a single punctate calcification in a third. Two patients had a history of excision of a lipomatous lesion in the region of the mass. One patient had pulmonary metastases. CONCLUSION: The potential for deep well-differentiated fatty tumors of the extremities to dedifferentiate is not generally recognized. Although the different types of liposarcoma cannot be reliably distinguished with imaging studies, a well-defined nonlipomatous mass juxtaposed with a predominantly fatty tumor is suggestive of a dedifferentiated liposarcoma.

Imaging of soft tissue tumors.

In summary, MR imaging is the preferred modality for the evaluation of a soft tissue mass after plain films have been taken. The radiologic appearance of certain soft tissue tumors or tumorlike processes such as myositis ossificans, benign fatty tumors, intramuscular hemangiomas, pigmented villonodular synovitis, and certain hematomas may be sufficiently unique to allow a strong presumptive radiologic diagnosis. It must be emphasized that MR cannot reliably distinguish between benign and malignant lesions, and when radiologic evaluation is nonspecific, one is ill advised to suggest a lesion is benign or malignant solely on the basis of its MR appearance. CT may be useful in specific instances for the identification of subtle soft tissue mineralization in those patients in whom lesions are not adequately evaluated by radiographs. Ultrasonography may be useful in the assessing of recurrent disease as well as in establishing tumor vascularity.

Liposarcoma of the extremities: MR and CT findings in the histologic subtypes.

The findings on images of liposarcomas of the extremities in 48 patients (26 men and 22 women aged 20-85 years) were reviewed to correlate the histologic subtypes with radiologic appearance. Computed tomographic scans were obtained in 36 patients; magnetic resonance (MR) images, in 27 patients. The study group had 19 myxoid, 12 well-differentiated, nine round cell, and eight pleomorphic liposarcomas. Well-differentiated liposarcomas were predominantly composed of fat, typically with thick septa, which were hyperintense on T2-weighted spin-echo MR images. The heterogeneity of tumor helped differentiate tumor subtypes. Eleven of 19 myxoid tumors were mildly heterogeneous. Round cell and pleomorphic subtypes, which tend to be aggressive, were generally heterogeneous. Well-differentiated liposarcomas may be confidently differentiated from other subtypes of liposarcoma. Except for the well-differentiated subtypes, liposarcomas often contain no fat. Moderate to marked heterogeneity is common in high-grade liposarcomas; myxoid liposarcomas tend to be homogeneous and may mimic cysts.

CT and MR of neurofibromatosis of the bladder.

Neurofibromatosis of the genitourinary system is rare. We present the CT and MR findings of neurofibromas of the bladder in three patients with von Recklinghausen disease (neurofibromatosis, Type 1). In one case, genitourinary involvement was the primary presentation of the disease. Both CT and MR imaging revealed diffuse and nodular bladder wall involvement, along with pelvic sidewall and adjacent soft tissue abnormalities. The CT attenuation coefficients measured near soft tissue density. On T1-weighted spin echo MR images the tumors revealed signal intensity slightly greater than that of skeletal muscle. Neurofibromas showed markedly increased signal intensity on T2-weighted images relative to the surrounding soft tissues, with marked enhancement in two cases imaged following Gd-DTPA administration. Obstructive hydronephrosis was present in all cases, presumably due to neurofibromas involving the trigonal region. Pelvic sidewall tumors were visualized as rounded, nodular masses extending into the obturator foramina. In the evaluation of patients with von Recklinghausen disease, MR imaging, compared with CT, more clearly defined tumor extent within the bladder, pelvic sidewalls, and surrounding soft tissues.