A 51-year-old woman presented with moderate knee pain refractory to conservative measures. Radiographs revealed a well-defined, but irregularly-shaped ovoid soft tissue density at the posterior superior aspect of the infrapatellar fat pad of Hoffa. On magnetic resonance imaging, the mass was mildly heterogeneous and had intermediate signal in the proton density series and mixed signal intensity on the T2-weighted images. A low-signal rim could be seen around a portion of the lesion. Arthroscopic resection was performed and a vascular stalk was encountered. The differential diagnosis included: ganglion cyst, meniscal cyst, intra-articular lipoma, villous proliferation of the synovial membrane, Hoffa disease, and intracapsular chondroma. Histologic examination revealed a fibrous capsule partially surrounding a benign lipomatous neoplasm containing an abundance of thin- and thick-walled blood vessels with periadventitial myxoid stroma.
To our knowledge, this is the first reported case of an intra-articular angiomyxolipoma in the literature. At 8-month follow-up, the patient was asymptomatic with no sign of recurrence. We postulate a low recurrence rate based on the insidious growth rate and benign histological appearance of such lesions. To our knowledge, no other intra-articular lipomatous lesion of the knee has been removed solely with arthroscopic techniques. Further follow-up information is needed to better understand the natural course of these lesions.
Intra-articular knee masses are relatively rare clinical entities with a known differential diagnosis. The most common lesions seen clinically and in radiologic evaluation of the knee include meniscal cysts, villous proliferation of the synovial membrane, Hoffa disease, and ganglion cysts. This article presents a novel case of angiomyxolipoma in the knee. This is the first reported case of an intra-articular angiomyxolipoma and also the first reported case of a lipomatous lesion removed entirely with arthroscopic techniques.
A 51-year-old woman presented with right knee pain. She had anterior knee pain for a number of years and had been treated with nonsteroidal anti-inflammatory medications and physical therapy by her primary care physician. This provided only temporary relief of her symptoms.
Over the past 6 months, the patient had increasing pain in her anterior knee that worsened with hyperflexion activities. She also reported a deep aching at night if active during the day. However, she reported no mechanical symptoms, specifically locking, catching, or instability of the knee. Based on her symptoms, she underwent magnetic resonance imaging (MRI) scan of her right knee.
On physical examination, the patient had no effusion of her knee. Range of motion of the right knee was from 0° to 125° with pain at the extreme of flexion. There was full extension and flexion to 135° in the contralateral left knee. Palpation of the joint line elicited no tenderness or palpable masses. There was no varus or valgus instability of the right knee. Anterior and posterior drawer tests were negative. The Lachmans test was also negative. Neurologic examination of the lower extremity revealed no deficits and the dorsalis pedis and posterior tibialis pulses were symmetric in both lower extremities.
Radiographs obtained to evaluate the patients knee pain showed an abnormality on the lateral view only (Figure 1). A well-defined, but slightly irregularly shaped ovoid soft-tissue density could be visualized at the posterior/superior aspect of the infrapatellar fat pad of Hoffa. An MRI scan of the knee confirmed a soft tissue mass in this region. On the proton density image (Figure 2A), it was a mildly heterogeneous and mostly intermediate signal. At the inferior aspect of the lesion, an area was brighter and isointense to the marrow and subcutaneous fat. No significant amount of low signal was seen on the proton-weighted images to suggest hemosiderin deposition. The mass was of mixed signal intensity on the T2-weighted images (Figure 2B) with areas of bright, intermediate, and low signal. A low-signal rim could be seen around a portion of the lesion. As on the proton density images, the inferior portion of the lesion was isointense to fat.
| || || |
|Figure 1: Lateral radiograph of the knee demonstrating an ovoid soft tissue density at the posterior aspect of the infrapatellar fat pad of Hoffa. Figure 2: A sagittal proton density spin echo (TR/TE 1200/25) MRI scan of the knee revealing a well-defined intermediate signal mass at the posterior/superior aspect of Hoffa’s fat pad that corresponds to the soft tissue density seen on radiograph (A). A sagittal T2 fast spin echo (TR/TE 4500/90) MRI scan of the knee showing the mass in the infrapatellar fat pad. It is of mixed signal intensity on the T2-weighted images with areas of bright, intermediate, and low signal (B). |
Arthroscopy was performed. The lesion was located posterior to the superior aspect of the anterior fat pad as seen in this arthroscopic image (Figure 3). An excisional biopsy was performed with the use of electrocautery as bleeding was encountered during its removal. The gross specimen was removed from the knee and sent to pathology for permanent section (Figure 4). The histologic patterns are revealed in Figure 5.
| || |
|Figure 3: Arthroscopic image showing the mass engaging the femoral trochlea during knee flexion. Figure 4: Arthroscopic image taken during excisional biopsy as the entire lesion was removed from the anterolateral portal. |
| || || |
| ||Figure 5: A slide showing a lipomatous tumor containing a prominent vascular component with periadventitial myxoid stroma. A delicate capsule can also be seen at the periphery (hematoxylin-eosin, original magnification ×100) (A). A hematoxylin-eosin stain revealing benign features including: absence of cytologic atypia, mitoses, necrosis, and lipoblasts seen in this field. Additionally, the patchy perivascular lymphoplasmacytic infiltrate is apparent (hematoxylin-eosin, original magnification ×300) (B). Myxoid stroma stains for Alcian Blue (pH 2.5) (Alcian Blue (pH 2.5), original magnification ×100) (C). The cellular component of the myxoid stroma (as well as the endothelial cells lining the vascular lumens) is immunoreactive for CD34 (CD34, original magnification ×200) (D). The cellular component of the myxoid stroma (as well as the endothelial cells and adipocytes) is immunoreactive for Vimentin (Vimentin, original magnification ×200) (E). |
Grossly, the mass was soft to rubbery, well-demarcated, partially encapsulated, and composed of yellow fatty lobules surrounded by a gelatinous stroma with focal central hemorrhage. Routine histologic sections showed a variably defined fibrous capsule partially surrounding a benign lipomatous neoplasm containing an abundance of thin- and thick-walled blood vessels with periadventitial myxoid stroma (Figure 5A). There were no cytologic atypia, mitotic figures, areas of necrosis, or lipoblasts. A patchy perivascular lymphoplasmacytic infiltrate was also seen (Figure 5B). The myxoid stroma stained positively for Alcian Blue (pH 2.5) (Figure 5C) and the cellular component was immunoreactive for CD34 (Figure 5D) and vimentin (Figure 5E), but not for smooth muscle actin, muscle specific actin, or S-100, supporting our diagnosis.1-3
Based on the clinical and pathologic review, the diagnosis of angiomyxolipoma was made. The differential diagnosis of an intra-articular knee mass includes, but is not limited to: ganglion cysts, meniscal cysts, intra-articular lipoma, villous proliferation of the synovial membrane, Hoffa disease, and intracapsular chondroma.
Intra-articular lipomas are rare but present with pain and no history of trauma or injury. Physical findings are not helpful in the diagnosis as they may present with or without a palpable mass. In our patient, history and physical examination findings were not helpful in making the diagnosis. These patients posed a severe diagnostic challenge before the advent of advanced imaging studies. However, even after review of the MRI results in this case, the diagnosis remained unclear. Only after the specimen was reviewed by pathology was the diagnosis proposed. The imaging characteristics excluded a simple lipoma. A simple lipoma is uniformly isointense to fat on all MRI sequences and this lesion only had fatty signal in its inferior pole.
Intra-articular ganglion cysts are uncommon. A 1.3% incidence was reported in a retrospective review of all knee MRIs at one institution.4 Most cysts were located in the intercondylar notch with only 3 located in the fat pad. All were isointense to fluid with low TI-weighted signal intensity and high T2-weighted signal intensity. True lipomas and ganglion cysts both present as painful knees and can only be differentiated by their MRI findings. Only 22% of the patients found to have a cyst on MRI eventually underwent aspiration or excision.
In a series of 1571 surgically excised menisci, 7.1% were found to have >1 cyst. They were found in association with horizontal tears, providing a means for synovial fluid to communicate with the lesion.5 They were more commonly found in the lateral meniscus, but may occur anywhere in the joint. In our described case, no meniscal pathology was identified on imaging or during arthroscopic visualization, suggesting an intra-articular ganglion is not a diagnosis. The MRI findings also helped exclude a ganglion or meniscal cyst. A cyst is isointense to fluid on MRI and therefore should be uniformly bright on a T2-weighted image and intermediate in signal on proton density images. The intraoperative appearance of a solid mass and postoperative pathology conclusively ruled out a cyst.
Similar to true lipomas, lipoma arborescens may present with slow enlargement of a painless mass accompanied by intermittent effusions.6 These lesions, however, present as villous projections of fatty synovial tissue and have been found after trauma, in inflammatory arthropathies, or may arise de novo.7 Magnetic resonance imaging findings are used to distinguish between true lipomas and lipoma arborescens. Hill et al7 described pathognomonic findings for lipoma arborescens including: a synovial mass with a frondlike architecture, fat-signal intensity on all pulse sequences, suppression of signal with fat-selective presaturation, associated joint effusion, potential chemical-shift artifact, and absence of magnetic susceptibility effects from hemosiderin. Histologically, no villous synovial proliferation was noted, making it unlikely it was the correct diagnosis.
Hoffa disease is a syndrome of infrapatellar fat pad impingement.8 It may begin with trauma to the fat pad, causing hemorrhage and hypertrophy of the fat pad. This begins a cycle of further impingement and enlargement as it is crushed between the femur and tibia. Chondroid metaplasia can occur and these lesions are then known as synovial chondromas. Magnetic resonance imaging findings are characterized by increased signal intensity on T2-weighted MRIs. In the subacute and chronic phases, fibrin and hemosiderin have low-signal intensity on both T1- and T2-weighted MRIs. Fibrous tissue turns into fibrocartilaginous tissue that may ossify. Thus, the lesion may be visible on plain radiographs.8 This patient had no history of trauma and no ossification could be seen on radiographs. Bowing of the patellar tendon from mass effect and areas of low signal of all sequences due to fibrin and hemosiderin are features frequently seen in Hoffas disease that were not visible on this patients MRI.
Both localized nodular synovitis and villous nodular synovitis are unlikely given these MRI findings. On MRI, areas of low-signal intensity on all pulse sequences are characteristic as the lesions have a tendency to bleed, resulting in hemosiderin deposition. The presence of fat in this mass and lack of significant low signal on proton density images are radiologic features that make this diagnosis unlikely. Localized nodular synovitis clinically can present as a solitary mass with anterior knee pain or mechanical symptoms. These lesions do not have to have hemosiderin deposition. However, pathology often shows giant cell proliferation not seen in this mass.9,10 The diffuse form of the disease is clinically different from the solitary mass seen in this case with no joint changes or signs of bleeding.11
The closest alternative diagnosis in this case is an intra-articular lipoma. True intra-articular lipomas are rare with only a few definitive cases in the literature.6,7,12 A few other reports of lipomas in the hip13 and in the spine14,15 have been published, but the exact incidence of these lesions is not clear.6 True intra-articular lipomas are typically characterized by a small, solitary, round mass composed of adipose tissue enclosed by a thin fibrous capsule.6 Similar to the case described by Matsumoto et al,6 the mass in our case was enclosed by synovium and had a vascular stalk. However, unlike the case described by Matsumoto et al6 that was found in the suprapatellar fat pad, the mass was felt to have arisen from the infrapatellar fat pad.
Similarly, no villous synovial proliferations were found, ruling out the diagnosis of lipoma arborescens. Although MRI findings of lipoma arborescens have been well described, MRI findings of true intra-articular lipoma have not.6 In our patient, the lesion had MRI findings, suggesting a fatty component with signal changes in a portion of the lesion isointense to fat. The heterogeneity of the mass on MRI is consistent with the vascular and cellular components seen histologically. Although the histologic features of angiomyxolipomas are characteristic, as originally described by Mai et al2 in 1996, a differential diagnosis for lipomatous tumors with a prominent vascular component includes angiomyxolipoma,3 angiomyxoma,3 myxoid spindle cell lipoma,1,3 low-grade myxofibrosarcoma,3 and liposarcoma.2,16 The conspicuous absence of smooth muscle elements by both histologic and immunohistochemical examination precludes angiomyolipoma, while the presence of adipose tissue excludes angiomyxoma. The prominent vascular component combined with the relative lack of organized, palisading bundles of spindle cells suggests a diagnosis other than myxoid spindle cell lipoma. The absence of organized spindle cells together with an absence of cytologic atypic also excludes low-grade myxofibrosarcoma. Finally, the lack of cytologic atypia, mitotic figures, and lipoblasts conclusively excludes a diagnosis of liposarcoma and its variants.17
This is the first case report, to our knowledge, of an intra-articular angiomyxolipoma. As a comparison, little is known about the natural history of true intra-articular lipomas. No case report elaborates on long-term follow-up of these lesions.6,7,12 All of the reported cases were diagnosed after the lesion was excised.6,7,12 We postulate a low recurrence rate based on the insidious growth rate and benign histological appearance of such lesions. To our knowledge, no other intra-articular lipomatous lesion of the knee has been removed solely with arthroscopic techniques. This patient went home the same day as surgery and at 1-month follow-up, has improvement in her pain although full return of flexion had yet to happen. At 8 month follow-up, the patient was completely asymptomatic with full ROM. Further follow-up information is needed to better understand the natural course of these lesions.
- Lee HW, Lee DK, Lee MW, Choi JH, Moon KC, Koh JK. Two cases of angiomyxolipoma (vascular myxolipoma) of subcutaneous tissue. J Cutan Pathol. 2005; 32(5):379-382.
- Mai KT, Yazdi HM, Collins JP. Vascular myxolipoma (angiomyxolipoma) of the spermatic cord. Am J Surg Pathol. 1996; 20(9):1145-1148.
- Tardio JC, Martin-Fragueiro LM. Angiomyxolipoma (vascular myxolipoma) of subcutaneous tissue. Am J Dermatopathol. 2004; 26(3):222-224.
- Bui-Mansfield LT, Youngberg RA. Intraarticular ganglia of the knee: prevalence, presentation, etiology, and management. AJR Am J Roentgenol. 1997; 168(1):123-127.
- Barrie HJ. The pathogenesis and significance of menisceal cysts. J Bone Joint Surg Br. 1979; 61(2):184-189.
- Matsumoto K, Okabe H, Ishizawa M, Hiraoka S. Intra-articular lipoma of the knee joint. A case report. J Bone Joint Surg Am. 2001; 83(1):101-105.
- Hill JA, Martin WR III, Milgram JW. Unusual arthroscopic knee lesions: case report of an intra-articular lipoma. J Natl Med Assoc. 1993; 85(9):697-699.
- Jacobson, JA, Lenchik L, Ruhoy MK, Schweitzer ME, Resnick D. MR imaging of the infrapatellar fat pad of Hoffa. Radiographics. 1997; 17(3):675-691.
- Nau T, Chiari C, Seitz H, Weixler G, Krenn M. Giant-cell tumor of the synovial membrane: localized nodular synovitis in the knee joint. Arthroscopy. 2000; 16(8):E22.
- Yoo JH, Yang BK, Park JM. Localized nodular synovitis of the knee presenting as anterior knee pain: a case report [published online ahead of print July 26, 2007]. Knee. 2007; 14(5):398-401.
- Tyler WK, Vidal AF, Williams RJ, Healey JH. Pigmented villonodular synovitis. J Am Acad Orthop Surg. 2006; 14(6):376-385.
- Pudlowski RM, Gilula LA, Kyriakos M. Intraarticular lipoma with osseous metaplasia: radiographic-pathologic correlation. AJR Am J Roentgenol. 1979; 132(3):471-473.
- Margheritini F, Villar RN, Rees D. Intra-articular lipoma of the hip. A case report. Int Orthop. 1998; 22(5):328-329.
- Dietemann JL, Bonneville JF, Runge M, Jeung MY, Weintraub A, Wackenheim A. Computed tomography of lumbar apophyseal joint lipoma: report of three cases. Neuroradiology. 1989; 31(1):60-62.
- Husson JL, Chales G, Lancien G, Pawlotsky Y, Masse A. True intra-articular lipoma of the lumbar spine. Spine (Phila PA 1976). 1987; 12(8):820-822.
- Sánchez Sambucety P, Alonso TA, Agapito PG, Moran AG, Rodríguez Prieto MA. Subungual angiomyxolipoma. Dermatol Surg. 2007; 33(4):508-509.
- Weiss S, Goldblum J. Soft Tissue Tumors. Philadelphia, PA: Mosby; 2008.
Drs Bergin, Milchteim, and Faulks are from the Department of Orthopedic Surgery, Drs Beaulieu and Schwartz are from the Department of Pathology, and Dr Brindle is from the Department of Radiology at The George Washington University Hospital, Washington, DC.
Drs Bergin, Milchteim, Beaulieu, Brindle, Schwartz, and Faulks have no relevant financial relationships to disclose.
Correspondence should be addressed to: Patrick F. Bergin, MD, Department of Orthopedic Surgery, George Washington University, 2150 Pennsylvania Ave NW, Washington, DC 20037 (firstname.lastname@example.org).