Jonathan Tresley, MD

Take-Home Points

• Rare and histologically indistinguishable from osteosarcoma of bone.
• Most common presentation is an enlarging mass in the thigh or buttock.
• Secondary extraosseous osteosarcoma usually arises in the field of prior external beam radiation or brachytherapy.
• Radiographic pattern of mineralization is central amorphous or cloudlike.
• On cross sectional imaging, the soft-tissue mass is separate from the underlying bone and periosteum.

Aside from multiple myeloma, osteosarcoma is the most common primary malignancy of bone, but extraosseous osteosarcoma is rare and accounts for only 1% of soft-tissue sarcomas and only 4% of all osteosarcomas.^1-3 Benign mesenchymal tumors, such as lipomas, are common, and they are estimated to outnumber their malignant counterparts by more than a factor of 100. However, the true ratio is unknown, as many clinically benign lipomas are not biopsied.⁴ Conventional lipoma is the most common lipoma and is biologically indolent. Conventional lipoma generally does not transform biologically into a more aggressive type of neoplasm—unlike atypical lipomatous tumors, which may demonstrate this type of evolution with multiple local recurrences.

This article is the first report of a case of radiation-associated extraosseous osteosarcoma that developed within a benign conventional lipoma. The patient provided written informed consent for print and electronic publication of this case report.

Case Report

In March 2013, a 72-year-old woman presented to a general surgeon with a right thigh mass of several weeks’ duration. The patient, who had a remote history of thyroid carcinoma, underwent thyroidectomy in 1991, excision of melanoma of the chest in 1998, and resection and adjuvant external beam radiotherapy (30 fractions) for Merkel cell carcinoma of the right proximal lateral leg (malignancy images unavailable) at an outside institution in 2003. Regional lymph node dissection at the time was negative. The patient remained disease-free the next 10 years. On presentation, magnetic resonance imaging (MRI) showed a 2.2-cm mass encircled by a tumor of lipomatous tissue within the vastus intermedius muscle, adjacent to but separate from the right distal femur (Figures 1A-1C).

Clinical examination findings suggested the sarcoma had arisen at the margins of the radiation field, but more than 10 years had passed since initial treatment, and records were unavailable for confirmation. Results of a computed tomography (CT)-guided biopsy performed at an outside institution revealed a high-grade malignancy, either an extraosseous osteosarcoma or a dedifferentiated liposarcoma. After the biopsy, the patient developed a severe medial compartment hematoma that required angiography and embolization. She was then referred to the division of orthopedic surgical oncology at our institution.

Physical examination revealed marked ecchymosis of the left groin at the access site for embolization as well as massive ecchymosis and swelling along the right distal thigh, medial knee, and medial lower leg. The neurovascular structures were intact with full motor function and sensation distally, as well as normal distal pulses. No inguinal adenopathy was identified. The proximal portion of the prior radiation tattoo was at the inferior extent of the lesion on MRI.

The patient was treated with doxorubicin and ifosfamide (2 cycles) while waiting for the hematoma to shrink. Contrast-enhanced MRI showed a 2.2-cm enhancing mass with isointense T1 signal and heterogeneously hyperintense STIR (short tau inversion recovery) signal surrounded by a circumscribed nonenhancing lipomatous tumor within the vastus intermedius muscle, adjacent to the distal femoral cortex. There was no invasion of the bone, and a fat plane between the enhancing mass and the femoral cortex was identified (Figures 2A-2E).

Fluorine 18 (¹⁸F) fluorodeoxyglucose (FDG) positron emission tomography (PET)/CT (FDG-PET/CT) showed marked hypermetabolic activity within the soft-tissue mass (maximum standardized uptake value, 7.0), surrounded by metabolically nonactive fat. No hypermetabolic lung, bone, or soft-tissue metastases were seen. CT and plain radiographs showed the nonfat portion of the tumor had soft-tissue density and contained a central and peripheral curvilinear pattern of mineralization (Figures 3A-3C, 4A-4B).

The primary differential diagnosis included myositis ossificans, chondrosseous lipoma, parosteal lipoma (ossifying variant), liposarcoma with metaplastic bone, and dedifferentiated liposarcoma with osteosarcoma or chondrosarcoma component (see Discussion section).

After 3 cycles of neoadjuvant chemotherapy with doxorubicin and ifosfamide, MRI showed a marked reduction in hematoma size, to 2.4 cm × 0.7 cm × 3.2 cm (estimated volume, ~3 mL), from 10 cm × 3.4 cm × 7.3 cm (estimated volume, ~130 mL), so the decision was made to proceed with surgery, excising the hematoma and sarcoma separately. First, wide resection of the hematoma yielded a 7-cm × 4-cm resection specimen with negative margins on frozen section. Subsequently, definitive radical resection of the tumor with wide margins yielded a 13-cm × 9-cm × 4-cm specimen. The resection specimen contained an intramuscular, mobile, encapsulated 2.0-cm × 1.5-cm × 1.0-cm mass with 2 components. The first was a tan-white solid mass containing thin deposits of calcified matrix, and the second, which surrounded the first, was composed of well-circumscribed soft yellow lobulated adipose tissue (Figure 5).

Microscopic evaluation revealed that the tan-white mass consisted of a hypercellular proliferation of malignant spindle and polyhedral cells that exhibited marked pleomorphism and hyperchromasia and produced extracellular coarse lace-like neoplastic bone characteristic of a high-grade extraskeletal osteosarcoma (Figures 6A-6D).

The sarcoma was sharply demarcated from the surrounding fatty component, which consisted of lobules of mature white adipocytes with no cytologic atypia, characteristic of a lipoma. An estimated 60% of the neoplasm was the lipoma, and the other 40% was the osteosarcoma. Immunohistochemistry revealed the tumor cells from both components to be negative for desmin, myogenin, CDK4, and MDM2. P16 showed cytoplasmic staining of the malignant cells, and these results helped exclude the possibility of dedifferentiated liposarcoma. All resection margins were negative, including the deep margin of the femoral periosteum. In addition, the resected hematoma did not contain malignant cells.

After surgery, the patient’s dermatologist performed a shave biopsy of a lentiginous lesion anterior to the knee. Subsequently, the patient began having increasing knee pain and developed, on the lower extremity, small areas of erythema that were attributed to mild cellulitis. Four months after surgery, emergent contrast-enhanced MRI showed enhancement of thickened synovium of the knee joint (Figure 7).

The patient underwent arthroscopic lavage and synovial biopsy for septic arthritis after knee aspiration yielded 51,000 white blood cells with a negative bacterial culture. The biopsy yielded acute and chronic inflammatory cells compatible with infection. No malignant cells were identified, and the bacterial culture was negative.

Since the lavage, the patient remained in good condition. There was no evidence of local recurrence on contrast-enhanced MRI (Figure 8), or metastases the first year, and she remained clinically free of disease the first 22 months of follow-up.

 

Discussion

Extraosseous osteosarcoma, typically a high-grade malignant neoplasm of the soft tissues that produces osteoid or cartilaginous matrix, is histologically indistinguishable from osteosarcoma of bone.

It usually occurs in the sixth decade of life, and there is a slight male predominance.^1,3,5,6 The most common presentation is an enlarging mass that may be painful. This mass often originates within the deep soft tissues of the lower extremities, especially the thigh and buttock, and less frequently in the upper extremity, retroperitoneum, and torso.⁶ Secondary extraosseous osteosarcoma accounts for 4% to 13% of extraosseous osteosarcoma and usually arises in the field of prior external beam radiation or brachytherapy.^1-3

Conventional lipoma, the most common subtype of lipoma, is a benign mesenchymal tumor. Other subtypes are hibernoma, fibrolipoma, angiolipoma, myelolipoma, spindle-cell lipoma, pleomorphic lipoma, and atypical lipomatous tumor.⁷ Atypical lipomatous tumor and well-differentiated liposarcoma are distinguished from each other by location: The World Health Organization recommends the term atypical lipomatous tumor for tumors that arise in the extremities and trunk lesions and well-differentiated liposarcoma for neoplasms that develop in the retroperitoneum, peritoneum, mediastinum, spermatic cord, and thoracic cavity.⁸ On PET, hypermetabolic activity is nonspecific and can be seen in malignant tumors and some benign reactive processes, such as evolving heterotopic ossification. However, simple lipomas, including those with mature ossification or dystrophic calcification, do not manifest increased FDG avidity.⁹

We are not aware of any published cases of extraosseous osteosarcoma arising within a conventional lipoma. A limited number of cases of coexisting conventional lipoma and spindle-cell lipoma or liposarcoma have been reported.^10-13 Retroperitoneal liposarcoma with areas of dedifferentiation into osteosarcoma has also been described.¹⁴ Development of malignant fibrous histiocytoma and liposarcoma have also been reported within intraosseous lipomas.¹⁵ One theory is based on premalignancy as a biological concept as opposed to a morphologic one. In other words, lesions that may be considered morphologically benign may already have the biological phenotype for malignancy that is not yet reflected morphologically.¹⁶ However, it has been suggested that such findings may instead result from initial sampling error or histologic misdiagnosis.^17,18There is a spectrum of findings on imaging studies of extraosseous osteosarcoma. Plain radiographs show a soft-tissue density with variable degrees of central calcification that reflects mineralization of deposited neoplastic bone. The pattern of calcification is characteristically amorphous or cloudlike, as opposed to the ring-and-arc observed in cartilage matrix. On CT, the soft-tissue mass of extraosseous osteosarcoma is separate from the underlying bone and periosteum—a defining characteristic that distinguishes it from conventional intramedullary and juxtacortical osteosarcoma.⁶ The central pattern of amorphous calcification helps to differentiate extraosseous osteosarcoma from heterotopic ossification, which characteristically demonstrates zonation, with trabecular architecture and mature cortical bone peripherally.¹ Enhancement of extraskeletal osteosarcoma tends to be heterogeneous and depends on the quantity of necrosis. Extraskeletal osteosarcoma tends to be isointense on T1-weighted MRI and mildly hyperintense on T2-weighted MRI.^1,6 Areas of very low signal intensity on both T1- and T2-weighted MRI may reflect mineralization.¹⁹ If intratumoral hemorrhage has occurred, there may be signal intensity of blood products of various ages.^1,3 Tumors with abundant hemorrhage can be mistaken for hematoma. FDG-PET radiotracer accumulation tends to be intense peripherally with variable central activity depending on quantity of necrosis and hemorrhage.¹The radiologic differential diagnosis includes myositis ossificans, chondrosseous lipoma, parosteal lipoma (ossifying variant), liposarcoma with metaplastic bone, dedifferentiated liposarcoma with osteosarcoma or chondrosarcoma component, and malignant mesenchymoma. Other common soft-tissue sarcomas, such as fibrosarcoma, leiomyosarcoma, and pleomorphic undifferentiated sarcoma, are excluded by the presence of fat within the tumor. The radiographic pattern of osteoid matrix produced by the tumor in our patient may be seen in heterotopic ossification, but the absence of mature ossification with zonation was evidence against heterotopic ossification, and microscopically it was neoplastic rather than reactive osteoid. In addition, it is possible that, because of the small size of the soft-tissue component, it was difficult to appreciate the less mature osteoid matrix peripherally. The lack of characteristic rings and arcs helps exclude benign and malignant cartilage containing neoplasms. Malignant mesenchymoma is a diagnosis of exclusion, and such tumors are usually better classified as sarcomas that have undergone heterologous differentiation.

The histologic diagnosis of extraosseous osteosarcoma requires identification of malignant mesenchymal cells that secrete neoplastic osteoid that may or may not mineralize. It is important to exclude the possibility that the malignant bone-forming tumor is part of a different type of sarcoma, the most common being dedifferentiated liposarcoma. Immunohistochemistry can be helpful in this situation, as dedifferentiated liposarcomas demonstrate nuclear expression of MDM2, CDK4, and p16, a constellation of findings rare in conventional and extraosseous osteosarcoma.^20-23 Osteosarcoma has not previously been reported as arising in a lipoma; in our patient’s case, we excluded the possibility that the fatty component represented an underlying atypical lipomatous tumor/well-differentiated or dedifferentiated liposarcoma on the basis of morphology and lack of expression of MDM2, CDK4, and p16.

Although histologically identical to osteosarcoma of bone, extraosseous osteosarcoma is treated differently because of its relatively decreased chemosensitivity and radiosensitivity. Treatment tends to be focused on limb-sparing wide local excision, and local recurrence complicates about 50% of cases.¹ Neoadjuvant or adjuvant treatment with radiation or chemotherapy is often provided.⁶ Platinum and doxorubicin chemotherapeutic agents, which are first-line treatments for osteosarcoma of bone, tend to be less effective in extraosseous osteosarcoma, and ifosfamide is more often used instead.⁵

Primary extraosseous osteosarcoma classically has a poor prognosis, with 2- to 3-year mortality of 50%, and prognosis tends to be worse for secondary radiation-induced sarcomas than for primary sarcomas.^2,6 However, with there being improved treatment protocols involving surgery and chemoradiation, more recent 5-year survival rates without metastatic disease are between 60% and 80%, though there is no definite consensus regarding the optimal systemic therapy regimen.^1,24 In a 2014 review of 53 patients who presented with localized disease, Choi and colleagues²⁵ identified a 3-year cumulative 39% incidence of death caused by disease, and in 2016 Sio and colleagues²⁶ reported that 55% of patients, most of whom had stage 3 disease, were alive at median follow-up of 45 months. Similar to osteosarcoma of bone, metastases may develop up to 10 years after primary treatment and are most commonly to the lung (80%-88%). Because extraosseous osteosarcoma is rare, no definite prognostic factors have been determined, but metastases at presentation and large tumor size (>5 cm) likely portend a worse prognosis.^2,3,27 Fibroblastic and chondroblastic subtypes may have a slightly better prognosis.^6,28

Conclusion

Extraosseous osteosarcoma is a rare malignancy that should be considered in the appropriate clinical and imaging scenario. This article is the first report of a case of a radiation-associated extraosseous osteosarcoma that developed within a lipoma with preoperative and postoperative multimodality imaging.

Am J Orthop. 2017;46(3):E200-E206. Copyright Frontline Medical Communications Inc. 2017. All rights reserved.

Take-Home Points

• Rare and histologically indistinguishable from osteosarcoma of bone.
• Most common presentation is an enlarging mass in the thigh or buttock.
• Secondary extraosseous osteosarcoma usually arises in the field of prior external beam radiation or brachytherapy.
• Radiographic pattern of mineralization is central amorphous or cloudlike.
• On cross sectional imaging, the soft-tissue mass is separate from the underlying bone and periosteum.

Aside from multiple myeloma, osteosarcoma is the most common primary malignancy of bone, but extraosseous osteosarcoma is rare and accounts for only 1% of soft-tissue sarcomas and only 4% of all osteosarcomas.^1-3 Benign mesenchymal tumors, such as lipomas, are common, and they are estimated to outnumber their malignant counterparts by more than a factor of 100. However, the true ratio is unknown, as many clinically benign lipomas are not biopsied.⁴ Conventional lipoma is the most common lipoma and is biologically indolent. Conventional lipoma generally does not transform biologically into a more aggressive type of neoplasm—unlike atypical lipomatous tumors, which may demonstrate this type of evolution with multiple local recurrences.

This article is the first report of a case of radiation-associated extraosseous osteosarcoma that developed within a benign conventional lipoma. The patient provided written informed consent for print and electronic publication of this case report.

Case Report

In March 2013, a 72-year-old woman presented to a general surgeon with a right thigh mass of several weeks’ duration. The patient, who had a remote history of thyroid carcinoma, underwent thyroidectomy in 1991, excision of melanoma of the chest in 1998, and resection and adjuvant external beam radiotherapy (30 fractions) for Merkel cell carcinoma of the right proximal lateral leg (malignancy images unavailable) at an outside institution in 2003. Regional lymph node dissection at the time was negative. The patient remained disease-free the next 10 years. On presentation, magnetic resonance imaging (MRI) showed a 2.2-cm mass encircled by a tumor of lipomatous tissue within the vastus intermedius muscle, adjacent to but separate from the right distal femur (Figures 1A-1C).

Clinical examination findings suggested the sarcoma had arisen at the margins of the radiation field, but more than 10 years had passed since initial treatment, and records were unavailable for confirmation. Results of a computed tomography (CT)-guided biopsy performed at an outside institution revealed a high-grade malignancy, either an extraosseous osteosarcoma or a dedifferentiated liposarcoma. After the biopsy, the patient developed a severe medial compartment hematoma that required angiography and embolization. She was then referred to the division of orthopedic surgical oncology at our institution.

Physical examination revealed marked ecchymosis of the left groin at the access site for embolization as well as massive ecchymosis and swelling along the right distal thigh, medial knee, and medial lower leg. The neurovascular structures were intact with full motor function and sensation distally, as well as normal distal pulses. No inguinal adenopathy was identified. The proximal portion of the prior radiation tattoo was at the inferior extent of the lesion on MRI.

The patient was treated with doxorubicin and ifosfamide (2 cycles) while waiting for the hematoma to shrink. Contrast-enhanced MRI showed a 2.2-cm enhancing mass with isointense T1 signal and heterogeneously hyperintense STIR (short tau inversion recovery) signal surrounded by a circumscribed nonenhancing lipomatous tumor within the vastus intermedius muscle, adjacent to the distal femoral cortex. There was no invasion of the bone, and a fat plane between the enhancing mass and the femoral cortex was identified (Figures 2A-2E).

Fluorine 18 (¹⁸F) fluorodeoxyglucose (FDG) positron emission tomography (PET)/CT (FDG-PET/CT) showed marked hypermetabolic activity within the soft-tissue mass (maximum standardized uptake value, 7.0), surrounded by metabolically nonactive fat. No hypermetabolic lung, bone, or soft-tissue metastases were seen. CT and plain radiographs showed the nonfat portion of the tumor had soft-tissue density and contained a central and peripheral curvilinear pattern of mineralization (Figures 3A-3C, 4A-4B).

The primary differential diagnosis included myositis ossificans, chondrosseous lipoma, parosteal lipoma (ossifying variant), liposarcoma with metaplastic bone, and dedifferentiated liposarcoma with osteosarcoma or chondrosarcoma component (see Discussion section).

After 3 cycles of neoadjuvant chemotherapy with doxorubicin and ifosfamide, MRI showed a marked reduction in hematoma size, to 2.4 cm × 0.7 cm × 3.2 cm (estimated volume, ~3 mL), from 10 cm × 3.4 cm × 7.3 cm (estimated volume, ~130 mL), so the decision was made to proceed with surgery, excising the hematoma and sarcoma separately. First, wide resection of the hematoma yielded a 7-cm × 4-cm resection specimen with negative margins on frozen section. Subsequently, definitive radical resection of the tumor with wide margins yielded a 13-cm × 9-cm × 4-cm specimen. The resection specimen contained an intramuscular, mobile, encapsulated 2.0-cm × 1.5-cm × 1.0-cm mass with 2 components. The first was a tan-white solid mass containing thin deposits of calcified matrix, and the second, which surrounded the first, was composed of well-circumscribed soft yellow lobulated adipose tissue (Figure 5).

Microscopic evaluation revealed that the tan-white mass consisted of a hypercellular proliferation of malignant spindle and polyhedral cells that exhibited marked pleomorphism and hyperchromasia and produced extracellular coarse lace-like neoplastic bone characteristic of a high-grade extraskeletal osteosarcoma (Figures 6A-6D).

The sarcoma was sharply demarcated from the surrounding fatty component, which consisted of lobules of mature white adipocytes with no cytologic atypia, characteristic of a lipoma. An estimated 60% of the neoplasm was the lipoma, and the other 40% was the osteosarcoma. Immunohistochemistry revealed the tumor cells from both components to be negative for desmin, myogenin, CDK4, and MDM2. P16 showed cytoplasmic staining of the malignant cells, and these results helped exclude the possibility of dedifferentiated liposarcoma. All resection margins were negative, including the deep margin of the femoral periosteum. In addition, the resected hematoma did not contain malignant cells.

After surgery, the patient’s dermatologist performed a shave biopsy of a lentiginous lesion anterior to the knee. Subsequently, the patient began having increasing knee pain and developed, on the lower extremity, small areas of erythema that were attributed to mild cellulitis. Four months after surgery, emergent contrast-enhanced MRI showed enhancement of thickened synovium of the knee joint (Figure 7).

The patient underwent arthroscopic lavage and synovial biopsy for septic arthritis after knee aspiration yielded 51,000 white blood cells with a negative bacterial culture. The biopsy yielded acute and chronic inflammatory cells compatible with infection. No malignant cells were identified, and the bacterial culture was negative.

Since the lavage, the patient remained in good condition. There was no evidence of local recurrence on contrast-enhanced MRI (Figure 8), or metastases the first year, and she remained clinically free of disease the first 22 months of follow-up.

 

Discussion

Extraosseous osteosarcoma, typically a high-grade malignant neoplasm of the soft tissues that produces osteoid or cartilaginous matrix, is histologically indistinguishable from osteosarcoma of bone.

It usually occurs in the sixth decade of life, and there is a slight male predominance.^1,3,5,6 The most common presentation is an enlarging mass that may be painful. This mass often originates within the deep soft tissues of the lower extremities, especially the thigh and buttock, and less frequently in the upper extremity, retroperitoneum, and torso.⁶ Secondary extraosseous osteosarcoma accounts for 4% to 13% of extraosseous osteosarcoma and usually arises in the field of prior external beam radiation or brachytherapy.^1-3

Conventional lipoma, the most common subtype of lipoma, is a benign mesenchymal tumor. Other subtypes are hibernoma, fibrolipoma, angiolipoma, myelolipoma, spindle-cell lipoma, pleomorphic lipoma, and atypical lipomatous tumor.⁷ Atypical lipomatous tumor and well-differentiated liposarcoma are distinguished from each other by location: The World Health Organization recommends the term atypical lipomatous tumor for tumors that arise in the extremities and trunk lesions and well-differentiated liposarcoma for neoplasms that develop in the retroperitoneum, peritoneum, mediastinum, spermatic cord, and thoracic cavity.⁸ On PET, hypermetabolic activity is nonspecific and can be seen in malignant tumors and some benign reactive processes, such as evolving heterotopic ossification. However, simple lipomas, including those with mature ossification or dystrophic calcification, do not manifest increased FDG avidity.⁹

We are not aware of any published cases of extraosseous osteosarcoma arising within a conventional lipoma. A limited number of cases of coexisting conventional lipoma and spindle-cell lipoma or liposarcoma have been reported.^10-13 Retroperitoneal liposarcoma with areas of dedifferentiation into osteosarcoma has also been described.¹⁴ Development of malignant fibrous histiocytoma and liposarcoma have also been reported within intraosseous lipomas.¹⁵ One theory is based on premalignancy as a biological concept as opposed to a morphologic one. In other words, lesions that may be considered morphologically benign may already have the biological phenotype for malignancy that is not yet reflected morphologically.¹⁶ However, it has been suggested that such findings may instead result from initial sampling error or histologic misdiagnosis.^17,18There is a spectrum of findings on imaging studies of extraosseous osteosarcoma. Plain radiographs show a soft-tissue density with variable degrees of central calcification that reflects mineralization of deposited neoplastic bone. The pattern of calcification is characteristically amorphous or cloudlike, as opposed to the ring-and-arc observed in cartilage matrix. On CT, the soft-tissue mass of extraosseous osteosarcoma is separate from the underlying bone and periosteum—a defining characteristic that distinguishes it from conventional intramedullary and juxtacortical osteosarcoma.⁶ The central pattern of amorphous calcification helps to differentiate extraosseous osteosarcoma from heterotopic ossification, which characteristically demonstrates zonation, with trabecular architecture and mature cortical bone peripherally.¹ Enhancement of extraskeletal osteosarcoma tends to be heterogeneous and depends on the quantity of necrosis. Extraskeletal osteosarcoma tends to be isointense on T1-weighted MRI and mildly hyperintense on T2-weighted MRI.^1,6 Areas of very low signal intensity on both T1- and T2-weighted MRI may reflect mineralization.¹⁹ If intratumoral hemorrhage has occurred, there may be signal intensity of blood products of various ages.^1,3 Tumors with abundant hemorrhage can be mistaken for hematoma. FDG-PET radiotracer accumulation tends to be intense peripherally with variable central activity depending on quantity of necrosis and hemorrhage.¹The radiologic differential diagnosis includes myositis ossificans, chondrosseous lipoma, parosteal lipoma (ossifying variant), liposarcoma with metaplastic bone, dedifferentiated liposarcoma with osteosarcoma or chondrosarcoma component, and malignant mesenchymoma. Other common soft-tissue sarcomas, such as fibrosarcoma, leiomyosarcoma, and pleomorphic undifferentiated sarcoma, are excluded by the presence of fat within the tumor. The radiographic pattern of osteoid matrix produced by the tumor in our patient may be seen in heterotopic ossification, but the absence of mature ossification with zonation was evidence against heterotopic ossification, and microscopically it was neoplastic rather than reactive osteoid. In addition, it is possible that, because of the small size of the soft-tissue component, it was difficult to appreciate the less mature osteoid matrix peripherally. The lack of characteristic rings and arcs helps exclude benign and malignant cartilage containing neoplasms. Malignant mesenchymoma is a diagnosis of exclusion, and such tumors are usually better classified as sarcomas that have undergone heterologous differentiation.

The histologic diagnosis of extraosseous osteosarcoma requires identification of malignant mesenchymal cells that secrete neoplastic osteoid that may or may not mineralize. It is important to exclude the possibility that the malignant bone-forming tumor is part of a different type of sarcoma, the most common being dedifferentiated liposarcoma. Immunohistochemistry can be helpful in this situation, as dedifferentiated liposarcomas demonstrate nuclear expression of MDM2, CDK4, and p16, a constellation of findings rare in conventional and extraosseous osteosarcoma.^20-23 Osteosarcoma has not previously been reported as arising in a lipoma; in our patient’s case, we excluded the possibility that the fatty component represented an underlying atypical lipomatous tumor/well-differentiated or dedifferentiated liposarcoma on the basis of morphology and lack of expression of MDM2, CDK4, and p16.

Although histologically identical to osteosarcoma of bone, extraosseous osteosarcoma is treated differently because of its relatively decreased chemosensitivity and radiosensitivity. Treatment tends to be focused on limb-sparing wide local excision, and local recurrence complicates about 50% of cases.¹ Neoadjuvant or adjuvant treatment with radiation or chemotherapy is often provided.⁶ Platinum and doxorubicin chemotherapeutic agents, which are first-line treatments for osteosarcoma of bone, tend to be less effective in extraosseous osteosarcoma, and ifosfamide is more often used instead.⁵

Primary extraosseous osteosarcoma classically has a poor prognosis, with 2- to 3-year mortality of 50%, and prognosis tends to be worse for secondary radiation-induced sarcomas than for primary sarcomas.^2,6 However, with there being improved treatment protocols involving surgery and chemoradiation, more recent 5-year survival rates without metastatic disease are between 60% and 80%, though there is no definite consensus regarding the optimal systemic therapy regimen.^1,24 In a 2014 review of 53 patients who presented with localized disease, Choi and colleagues²⁵ identified a 3-year cumulative 39% incidence of death caused by disease, and in 2016 Sio and colleagues²⁶ reported that 55% of patients, most of whom had stage 3 disease, were alive at median follow-up of 45 months. Similar to osteosarcoma of bone, metastases may develop up to 10 years after primary treatment and are most commonly to the lung (80%-88%). Because extraosseous osteosarcoma is rare, no definite prognostic factors have been determined, but metastases at presentation and large tumor size (>5 cm) likely portend a worse prognosis.^2,3,27 Fibroblastic and chondroblastic subtypes may have a slightly better prognosis.^6,28

Conclusion

Extraosseous osteosarcoma is a rare malignancy that should be considered in the appropriate clinical and imaging scenario. This article is the first report of a case of a radiation-associated extraosseous osteosarcoma that developed within a lipoma with preoperative and postoperative multimodality imaging.

Am J Orthop. 2017;46(3):E200-E206. Copyright Frontline Medical Communications Inc. 2017. All rights reserved.

Take-Home Points

• Rare and histologically indistinguishable from osteosarcoma of bone.
• Most common presentation is an enlarging mass in the thigh or buttock.
• Secondary extraosseous osteosarcoma usually arises in the field of prior external beam radiation or brachytherapy.
• Radiographic pattern of mineralization is central amorphous or cloudlike.
• On cross sectional imaging, the soft-tissue mass is separate from the underlying bone and periosteum.

Aside from multiple myeloma, osteosarcoma is the most common primary malignancy of bone, but extraosseous osteosarcoma is rare and accounts for only 1% of soft-tissue sarcomas and only 4% of all osteosarcomas.^1-3 Benign mesenchymal tumors, such as lipomas, are common, and they are estimated to outnumber their malignant counterparts by more than a factor of 100. However, the true ratio is unknown, as many clinically benign lipomas are not biopsied.⁴ Conventional lipoma is the most common lipoma and is biologically indolent. Conventional lipoma generally does not transform biologically into a more aggressive type of neoplasm—unlike atypical lipomatous tumors, which may demonstrate this type of evolution with multiple local recurrences.

This article is the first report of a case of radiation-associated extraosseous osteosarcoma that developed within a benign conventional lipoma. The patient provided written informed consent for print and electronic publication of this case report.

Case Report

In March 2013, a 72-year-old woman presented to a general surgeon with a right thigh mass of several weeks’ duration. The patient, who had a remote history of thyroid carcinoma, underwent thyroidectomy in 1991, excision of melanoma of the chest in 1998, and resection and adjuvant external beam radiotherapy (30 fractions) for Merkel cell carcinoma of the right proximal lateral leg (malignancy images unavailable) at an outside institution in 2003. Regional lymph node dissection at the time was negative. The patient remained disease-free the next 10 years. On presentation, magnetic resonance imaging (MRI) showed a 2.2-cm mass encircled by a tumor of lipomatous tissue within the vastus intermedius muscle, adjacent to but separate from the right distal femur (Figures 1A-1C).

Clinical examination findings suggested the sarcoma had arisen at the margins of the radiation field, but more than 10 years had passed since initial treatment, and records were unavailable for confirmation. Results of a computed tomography (CT)-guided biopsy performed at an outside institution revealed a high-grade malignancy, either an extraosseous osteosarcoma or a dedifferentiated liposarcoma. After the biopsy, the patient developed a severe medial compartment hematoma that required angiography and embolization. She was then referred to the division of orthopedic surgical oncology at our institution.

Physical examination revealed marked ecchymosis of the left groin at the access site for embolization as well as massive ecchymosis and swelling along the right distal thigh, medial knee, and medial lower leg. The neurovascular structures were intact with full motor function and sensation distally, as well as normal distal pulses. No inguinal adenopathy was identified. The proximal portion of the prior radiation tattoo was at the inferior extent of the lesion on MRI.

The patient was treated with doxorubicin and ifosfamide (2 cycles) while waiting for the hematoma to shrink. Contrast-enhanced MRI showed a 2.2-cm enhancing mass with isointense T1 signal and heterogeneously hyperintense STIR (short tau inversion recovery) signal surrounded by a circumscribed nonenhancing lipomatous tumor within the vastus intermedius muscle, adjacent to the distal femoral cortex. There was no invasion of the bone, and a fat plane between the enhancing mass and the femoral cortex was identified (Figures 2A-2E).

Fluorine 18 (¹⁸F) fluorodeoxyglucose (FDG) positron emission tomography (PET)/CT (FDG-PET/CT) showed marked hypermetabolic activity within the soft-tissue mass (maximum standardized uptake value, 7.0), surrounded by metabolically nonactive fat. No hypermetabolic lung, bone, or soft-tissue metastases were seen. CT and plain radiographs showed the nonfat portion of the tumor had soft-tissue density and contained a central and peripheral curvilinear pattern of mineralization (Figures 3A-3C, 4A-4B).

The primary differential diagnosis included myositis ossificans, chondrosseous lipoma, parosteal lipoma (ossifying variant), liposarcoma with metaplastic bone, and dedifferentiated liposarcoma with osteosarcoma or chondrosarcoma component (see Discussion section).

After 3 cycles of neoadjuvant chemotherapy with doxorubicin and ifosfamide, MRI showed a marked reduction in hematoma size, to 2.4 cm × 0.7 cm × 3.2 cm (estimated volume, ~3 mL), from 10 cm × 3.4 cm × 7.3 cm (estimated volume, ~130 mL), so the decision was made to proceed with surgery, excising the hematoma and sarcoma separately. First, wide resection of the hematoma yielded a 7-cm × 4-cm resection specimen with negative margins on frozen section. Subsequently, definitive radical resection of the tumor with wide margins yielded a 13-cm × 9-cm × 4-cm specimen. The resection specimen contained an intramuscular, mobile, encapsulated 2.0-cm × 1.5-cm × 1.0-cm mass with 2 components. The first was a tan-white solid mass containing thin deposits of calcified matrix, and the second, which surrounded the first, was composed of well-circumscribed soft yellow lobulated adipose tissue (Figure 5).

Microscopic evaluation revealed that the tan-white mass consisted of a hypercellular proliferation of malignant spindle and polyhedral cells that exhibited marked pleomorphism and hyperchromasia and produced extracellular coarse lace-like neoplastic bone characteristic of a high-grade extraskeletal osteosarcoma (Figures 6A-6D).

The sarcoma was sharply demarcated from the surrounding fatty component, which consisted of lobules of mature white adipocytes with no cytologic atypia, characteristic of a lipoma. An estimated 60% of the neoplasm was the lipoma, and the other 40% was the osteosarcoma. Immunohistochemistry revealed the tumor cells from both components to be negative for desmin, myogenin, CDK4, and MDM2. P16 showed cytoplasmic staining of the malignant cells, and these results helped exclude the possibility of dedifferentiated liposarcoma. All resection margins were negative, including the deep margin of the femoral periosteum. In addition, the resected hematoma did not contain malignant cells.

After surgery, the patient’s dermatologist performed a shave biopsy of a lentiginous lesion anterior to the knee. Subsequently, the patient began having increasing knee pain and developed, on the lower extremity, small areas of erythema that were attributed to mild cellulitis. Four months after surgery, emergent contrast-enhanced MRI showed enhancement of thickened synovium of the knee joint (Figure 7).

The patient underwent arthroscopic lavage and synovial biopsy for septic arthritis after knee aspiration yielded 51,000 white blood cells with a negative bacterial culture. The biopsy yielded acute and chronic inflammatory cells compatible with infection. No malignant cells were identified, and the bacterial culture was negative.

Since the lavage, the patient remained in good condition. There was no evidence of local recurrence on contrast-enhanced MRI (Figure 8), or metastases the first year, and she remained clinically free of disease the first 22 months of follow-up.

 

Discussion

Extraosseous osteosarcoma, typically a high-grade malignant neoplasm of the soft tissues that produces osteoid or cartilaginous matrix, is histologically indistinguishable from osteosarcoma of bone.

It usually occurs in the sixth decade of life, and there is a slight male predominance.^1,3,5,6 The most common presentation is an enlarging mass that may be painful. This mass often originates within the deep soft tissues of the lower extremities, especially the thigh and buttock, and less frequently in the upper extremity, retroperitoneum, and torso.⁶ Secondary extraosseous osteosarcoma accounts for 4% to 13% of extraosseous osteosarcoma and usually arises in the field of prior external beam radiation or brachytherapy.^1-3

Conventional lipoma, the most common subtype of lipoma, is a benign mesenchymal tumor. Other subtypes are hibernoma, fibrolipoma, angiolipoma, myelolipoma, spindle-cell lipoma, pleomorphic lipoma, and atypical lipomatous tumor.⁷ Atypical lipomatous tumor and well-differentiated liposarcoma are distinguished from each other by location: The World Health Organization recommends the term atypical lipomatous tumor for tumors that arise in the extremities and trunk lesions and well-differentiated liposarcoma for neoplasms that develop in the retroperitoneum, peritoneum, mediastinum, spermatic cord, and thoracic cavity.⁸ On PET, hypermetabolic activity is nonspecific and can be seen in malignant tumors and some benign reactive processes, such as evolving heterotopic ossification. However, simple lipomas, including those with mature ossification or dystrophic calcification, do not manifest increased FDG avidity.⁹

We are not aware of any published cases of extraosseous osteosarcoma arising within a conventional lipoma. A limited number of cases of coexisting conventional lipoma and spindle-cell lipoma or liposarcoma have been reported.^10-13 Retroperitoneal liposarcoma with areas of dedifferentiation into osteosarcoma has also been described.¹⁴ Development of malignant fibrous histiocytoma and liposarcoma have also been reported within intraosseous lipomas.¹⁵ One theory is based on premalignancy as a biological concept as opposed to a morphologic one. In other words, lesions that may be considered morphologically benign may already have the biological phenotype for malignancy that is not yet reflected morphologically.¹⁶ However, it has been suggested that such findings may instead result from initial sampling error or histologic misdiagnosis.^17,18There is a spectrum of findings on imaging studies of extraosseous osteosarcoma. Plain radiographs show a soft-tissue density with variable degrees of central calcification that reflects mineralization of deposited neoplastic bone. The pattern of calcification is characteristically amorphous or cloudlike, as opposed to the ring-and-arc observed in cartilage matrix. On CT, the soft-tissue mass of extraosseous osteosarcoma is separate from the underlying bone and periosteum—a defining characteristic that distinguishes it from conventional intramedullary and juxtacortical osteosarcoma.⁶ The central pattern of amorphous calcification helps to differentiate extraosseous osteosarcoma from heterotopic ossification, which characteristically demonstrates zonation, with trabecular architecture and mature cortical bone peripherally.¹ Enhancement of extraskeletal osteosarcoma tends to be heterogeneous and depends on the quantity of necrosis. Extraskeletal osteosarcoma tends to be isointense on T1-weighted MRI and mildly hyperintense on T2-weighted MRI.^1,6 Areas of very low signal intensity on both T1- and T2-weighted MRI may reflect mineralization.¹⁹ If intratumoral hemorrhage has occurred, there may be signal intensity of blood products of various ages.^1,3 Tumors with abundant hemorrhage can be mistaken for hematoma. FDG-PET radiotracer accumulation tends to be intense peripherally with variable central activity depending on quantity of necrosis and hemorrhage.¹The radiologic differential diagnosis includes myositis ossificans, chondrosseous lipoma, parosteal lipoma (ossifying variant), liposarcoma with metaplastic bone, dedifferentiated liposarcoma with osteosarcoma or chondrosarcoma component, and malignant mesenchymoma. Other common soft-tissue sarcomas, such as fibrosarcoma, leiomyosarcoma, and pleomorphic undifferentiated sarcoma, are excluded by the presence of fat within the tumor. The radiographic pattern of osteoid matrix produced by the tumor in our patient may be seen in heterotopic ossification, but the absence of mature ossification with zonation was evidence against heterotopic ossification, and microscopically it was neoplastic rather than reactive osteoid. In addition, it is possible that, because of the small size of the soft-tissue component, it was difficult to appreciate the less mature osteoid matrix peripherally. The lack of characteristic rings and arcs helps exclude benign and malignant cartilage containing neoplasms. Malignant mesenchymoma is a diagnosis of exclusion, and such tumors are usually better classified as sarcomas that have undergone heterologous differentiation.

The histologic diagnosis of extraosseous osteosarcoma requires identification of malignant mesenchymal cells that secrete neoplastic osteoid that may or may not mineralize. It is important to exclude the possibility that the malignant bone-forming tumor is part of a different type of sarcoma, the most common being dedifferentiated liposarcoma. Immunohistochemistry can be helpful in this situation, as dedifferentiated liposarcomas demonstrate nuclear expression of MDM2, CDK4, and p16, a constellation of findings rare in conventional and extraosseous osteosarcoma.^20-23 Osteosarcoma has not previously been reported as arising in a lipoma; in our patient’s case, we excluded the possibility that the fatty component represented an underlying atypical lipomatous tumor/well-differentiated or dedifferentiated liposarcoma on the basis of morphology and lack of expression of MDM2, CDK4, and p16.

Although histologically identical to osteosarcoma of bone, extraosseous osteosarcoma is treated differently because of its relatively decreased chemosensitivity and radiosensitivity. Treatment tends to be focused on limb-sparing wide local excision, and local recurrence complicates about 50% of cases.¹ Neoadjuvant or adjuvant treatment with radiation or chemotherapy is often provided.⁶ Platinum and doxorubicin chemotherapeutic agents, which are first-line treatments for osteosarcoma of bone, tend to be less effective in extraosseous osteosarcoma, and ifosfamide is more often used instead.⁵

Primary extraosseous osteosarcoma classically has a poor prognosis, with 2- to 3-year mortality of 50%, and prognosis tends to be worse for secondary radiation-induced sarcomas than for primary sarcomas.^2,6 However, with there being improved treatment protocols involving surgery and chemoradiation, more recent 5-year survival rates without metastatic disease are between 60% and 80%, though there is no definite consensus regarding the optimal systemic therapy regimen.^1,24 In a 2014 review of 53 patients who presented with localized disease, Choi and colleagues²⁵ identified a 3-year cumulative 39% incidence of death caused by disease, and in 2016 Sio and colleagues²⁶ reported that 55% of patients, most of whom had stage 3 disease, were alive at median follow-up of 45 months. Similar to osteosarcoma of bone, metastases may develop up to 10 years after primary treatment and are most commonly to the lung (80%-88%). Because extraosseous osteosarcoma is rare, no definite prognostic factors have been determined, but metastases at presentation and large tumor size (>5 cm) likely portend a worse prognosis.^2,3,27 Fibroblastic and chondroblastic subtypes may have a slightly better prognosis.^6,28

Conclusion

Extraosseous osteosarcoma is a rare malignancy that should be considered in the appropriate clinical and imaging scenario. This article is the first report of a case of a radiation-associated extraosseous osteosarcoma that developed within a lipoma with preoperative and postoperative multimodality imaging.

Am J Orthop. 2017;46(3):E200-E206. Copyright Frontline Medical Communications Inc. 2017. All rights reserved.

Take-Home Points

• Torn, displaced, and entrapped UCL is a Stener lesion.
• Hyperabduction injury with pain and joint laxity on examination.
• MRI and ultrasound are useful in evaluating UCL tears.
• Ultrasound offers dynamic evaluation.
• Must be treated appropriately to avoid pain, instability, and osteoarthritis.

In the literature, hyperabduction injuries to the thumb metacarpophalangeal (MCP) joint have been referred to interchangeably as gamekeeper’s thumb and skier’s thumb. Historically, though, gamekeeper’s thumb was initially described in hunters with chronic injury to the ulnar collateral ligament (UCL),¹ and skier’s thumb typically has been described as an acute hyperabduction injury of the UCL.^2-5 The proximal portion of a torn UCL may retract with further abduction and displace dorsally, becoming entrapped by the adductor pollicis aponeurosis insertion, known as a Stener lesion.⁶

The first MCP joint is stabilized by static and dynamic structures that contribute in varying degrees in flexion and extension of the joint. The static stabilizers include the proper and accessory radial and UCLs, the palmar plate, and the dorsal capsule. The UCL originates at the dorsal ulnar aspect of the first metacarpal head at the metacarpal tubercle about 5 mm proximal to the articular surface. The UCL courses distally in the palmar direction to insert volar and proximal to the medial tubercle of the proximal phalanx about 3 mm distal to the articular surface.⁷ In flexion, the proper collateral ligament is taut and is the primary static stabilizer. In extension, the accessory collateral ligament, which inserts on the palmar plate, is taut and is the primary static stabilizer.^8-11

The dynamic stabilizers include the extrinsic muscles (flexor pollicis longus, extensor pollicis longus and brevis) and the intrinsic muscles (abductor pollicis brevis, adductor pollicis, flexor pollicis brevis) inserting on the thumb at the distal phalanx and proximal phalanx and at the base of the first metacarpal.^8-10

The adductor pollicis originates from the volar third metacarpal, capitate, and hamate and has a dual insertion on the thumb.¹² There is a direct insertion onto the palmar proximal phalanx at the medial tubercle, distal and dorsal to the phalangeal insertion of the UCL.

There is also a broad aponeurosis that inserts onto the extensor hood expansion, dorsal to the insertion of the UCL (Figures 1A-1C and 2A, 2B).^7,8,13

We report the case of an acute hyperabduction injury of the thumb MCP joint with radiographic, ultrasound, and magnetic resonance imaging (MRI) findings consistent with a Stener lesion and subsequently confirmed with intraoperative photographs. The patient provided written informed consent for print and electronic publication of this case report.

Clinical Findings

A 33-year-old healthy man had persistent left hand pain and grip weakness after performing a handstand. He presented to the orthopedic hand clinic 20 days after injury, having failed nonoperative management (use of nonsteroidal anti-inflammatory drugs and soft thumb spica splint). Physical examination revealed soft-tissue swelling and focal tenderness to palpation at the ulnar aspect of the thumb MCP joint. Despite bilateral first MCP joint laxity on varus and valgus stress without identification of a firm endpoint, pain was elicited only on valgus stress of the left first MCP joint. Given the laxity and the left thumb soft-tissue swelling with pain, plain radiographs, ultrasound, and MRI were used to evaluate for severity of presumed left thumb UCL injury.

Imaging Findings

Plain radiographs showed normal bony anatomy without fracture, normal joint space, and mild soft-tissue swelling at the left thumb MCP level (Figures 3A, 3B).

Ultrasound confirmed a complete tear of the UCL, which was flipped in a proximal direction and projected dorsally in relation to the direct insertion of the adductor tendon (Figure 2B). MRI showed focal disruption of the UCL at the level of the left thumb MCP joint with associated MCP joint effusion (Figures 4A-4F).

Low T1 signal intensity over the adductor aponeurosis at the level of the metacarpal head corresponded with the torn and proximally retracted UCL. There was associated bone marrow edema at the radial and volar aspects of the thumb metacarpal head and low-grade strain of the abductor pollicis brevis. The thumb flexor and extensor tendons appeared normal. Although possibly secondary to patient positioning, mild volar subluxation of the proximal phalanx in relation to the metacarpal head was queried.

 

Surgical Findings

Given laxity with pain at the UCL on stress testing, MRI and ultrasound findings, and continued pain and instability of the thumb with pinching and grasping during activities of daily living, the patient and orthopedic hand surgeon proceeded with surgical intervention. Preoperative examination under anesthesia confirmed significant laxity on valgus stress without a palpable endpoint (Figures 5A, 5B).

During surgery, retraction of the extensor hood revealed the completely torn and displaced UCL, entrapped dorsally and proximally to the adductor aponeurosis, characteristic of a Stener lesion. After the primary repair of the UCL, the extensor hood was seen partially retracted in a normal location superficial to the normal deep position of the repaired UCL (Figures 6A, 6B).

Discussion

Hyperabduction injuries to the thumb may rupture the UCL of the MCP joint of the thumb or cause a bony avulsion of the base of the proximal phalanx. Injury to the UCL, most often at its distal portion,^4,14,15 may result in a sprain or full-thickness tear of the ligament.

Subsequently, the ligament may remain in situ, or the proximal segment may retract proximal to the adductor aponeurosis with continued abduction of the thumb. On release of the abduction force, the proximal UCL segment is displaced dorsally and proximally by the inferior aspect of the adductor aponeurosis. The UCL becomes entrapped by the adductor aponeurosis and cannot reduce spontaneously.¹⁵ This displacement was initially described by Stener⁶ in 1962 and is referred to as a Stener lesion (Figures 1A-1C).

It is vital for the radiologist to identify a Stener lesion because a nondisplaced tear of the UCL is often treated nonsurgically, but UCL tears displaced more than 3 mm and Stener lesions usually must be operated on to avoid chronic instability, pain, and osteoarthritis.^2-5,8,12-23 Sensitivity and specificity of MRI in evaluating UCL injuries are reported to be almost 100%, with resolution of 1 mm using current surface coils.²³ There are various UCL injury patterns, including partial tears, displaced and nondisplaced complete tears, and even complex injuries, such as an incomplete tear with the torn portion retracted as a Stener lesion.²² MRI is needed to establish the extent of injury, as 90% of complete tears that are displaced at least 3 mm, and all tears with retraction proximal and superficial to the aponeurosis (true Stener lesions), failed immobilization and required surgical treatment.²³Although they vary in the literature, mean sensitivity and specificity of ultrasound in detecting UCL tears in level I studies have been reported as 76% and 81%, respectively.²⁴ When Melville and colleagues²¹ applied their ultrasound criteria—including absence of normal UCL fibers traversing the first MCP joint as well as heterogeneous masslike tissue at least partially proximal to the apex of the metacarpal lateral tubercle—they were able to distinguish displaced full-thickness tears from nondisplaced full-thickness tears with 100% accuracy. Hergan and colleagues²⁵ found that the diagnostic accuracy of MRI was superior to that of ultrasound; while MRI accuracy was perfect, 12% of patients were incorrectly diagnosed with ultrasound, with false-positive or false-negative tendon-edge displacement. In our experience, ultrasound is uniquely useful in its ability to characterize the real-time dynamic interaction of the UCL with the adductor aponeurosis. It has been observed that passive flexion of the first interphalangeal joint moves the adductor aponeurosis in isolation, allowing differentiation from the subjacent UCL.²¹ Had a partial tear been in the differential diagnosis of our patient’s Stener lesion, such a maneuver under ultrasound visualization would have solved the dilemma. In addition, ultrasound allows for comparison with the contralateral ligament at the time of examination should a diagnostic dilemma arise.

As many have reported both bony avulsion of the base of the proximal phalanx and concomitant injury to the UCL, identification of a bony avulsion does not exclude a ligamentous injury and the possibility of a Stener lesion (Figure 7).^16,19

In one study, 14% of patients with injury to the UCL sustained a concomitant bony avulsion of the UCL insertion.²³ However, presence of the avulsion fragment did not alter management, and only those fragments involving more than 20% of the articular surface were considered true fractures and treated as such.

Conclusion

A Stener lesion—retraction of a completely torn UCL becoming entrapped dorsally and proximally to the adductor insertion—can cause pain, instability, and ultimately osteoarthritis if not treated appropriately. The orthopedic surgeon should have a high index of suspicion for a Stener lesion in the appropriate clinical scenario and consider all imaging modalities for diagnosis. Likewise, it is of utmost importance for the radiologist to identify imaging findings of a Stener lesion, as physical examination alone may be limited in its ability to characterize injury severity. Both MRI and ultrasound are useful in evaluating UCL tears, and ultrasound provides the additional benefit of dynamic visualization and comparison with the contralateral side.

Am J Orthop. 2017;46(3):E195-E199. Copyright Frontline Medical Communications Inc. 2017. All rights reserved.

Take-Home Points

• Torn, displaced, and entrapped UCL is a Stener lesion.
• Hyperabduction injury with pain and joint laxity on examination.
• MRI and ultrasound are useful in evaluating UCL tears.
• Ultrasound offers dynamic evaluation.
• Must be treated appropriately to avoid pain, instability, and osteoarthritis.

In the literature, hyperabduction injuries to the thumb metacarpophalangeal (MCP) joint have been referred to interchangeably as gamekeeper’s thumb and skier’s thumb. Historically, though, gamekeeper’s thumb was initially described in hunters with chronic injury to the ulnar collateral ligament (UCL),¹ and skier’s thumb typically has been described as an acute hyperabduction injury of the UCL.^2-5 The proximal portion of a torn UCL may retract with further abduction and displace dorsally, becoming entrapped by the adductor pollicis aponeurosis insertion, known as a Stener lesion.⁶

The first MCP joint is stabilized by static and dynamic structures that contribute in varying degrees in flexion and extension of the joint. The static stabilizers include the proper and accessory radial and UCLs, the palmar plate, and the dorsal capsule. The UCL originates at the dorsal ulnar aspect of the first metacarpal head at the metacarpal tubercle about 5 mm proximal to the articular surface. The UCL courses distally in the palmar direction to insert volar and proximal to the medial tubercle of the proximal phalanx about 3 mm distal to the articular surface.⁷ In flexion, the proper collateral ligament is taut and is the primary static stabilizer. In extension, the accessory collateral ligament, which inserts on the palmar plate, is taut and is the primary static stabilizer.^8-11

The dynamic stabilizers include the extrinsic muscles (flexor pollicis longus, extensor pollicis longus and brevis) and the intrinsic muscles (abductor pollicis brevis, adductor pollicis, flexor pollicis brevis) inserting on the thumb at the distal phalanx and proximal phalanx and at the base of the first metacarpal.^8-10

The adductor pollicis originates from the volar third metacarpal, capitate, and hamate and has a dual insertion on the thumb.¹² There is a direct insertion onto the palmar proximal phalanx at the medial tubercle, distal and dorsal to the phalangeal insertion of the UCL.

There is also a broad aponeurosis that inserts onto the extensor hood expansion, dorsal to the insertion of the UCL (Figures 1A-1C and 2A, 2B).^7,8,13

We report the case of an acute hyperabduction injury of the thumb MCP joint with radiographic, ultrasound, and magnetic resonance imaging (MRI) findings consistent with a Stener lesion and subsequently confirmed with intraoperative photographs. The patient provided written informed consent for print and electronic publication of this case report.

Clinical Findings

A 33-year-old healthy man had persistent left hand pain and grip weakness after performing a handstand. He presented to the orthopedic hand clinic 20 days after injury, having failed nonoperative management (use of nonsteroidal anti-inflammatory drugs and soft thumb spica splint). Physical examination revealed soft-tissue swelling and focal tenderness to palpation at the ulnar aspect of the thumb MCP joint. Despite bilateral first MCP joint laxity on varus and valgus stress without identification of a firm endpoint, pain was elicited only on valgus stress of the left first MCP joint. Given the laxity and the left thumb soft-tissue swelling with pain, plain radiographs, ultrasound, and MRI were used to evaluate for severity of presumed left thumb UCL injury.

Imaging Findings

Plain radiographs showed normal bony anatomy without fracture, normal joint space, and mild soft-tissue swelling at the left thumb MCP level (Figures 3A, 3B).

Ultrasound confirmed a complete tear of the UCL, which was flipped in a proximal direction and projected dorsally in relation to the direct insertion of the adductor tendon (Figure 2B). MRI showed focal disruption of the UCL at the level of the left thumb MCP joint with associated MCP joint effusion (Figures 4A-4F).

Low T1 signal intensity over the adductor aponeurosis at the level of the metacarpal head corresponded with the torn and proximally retracted UCL. There was associated bone marrow edema at the radial and volar aspects of the thumb metacarpal head and low-grade strain of the abductor pollicis brevis. The thumb flexor and extensor tendons appeared normal. Although possibly secondary to patient positioning, mild volar subluxation of the proximal phalanx in relation to the metacarpal head was queried.

 

Surgical Findings

Given laxity with pain at the UCL on stress testing, MRI and ultrasound findings, and continued pain and instability of the thumb with pinching and grasping during activities of daily living, the patient and orthopedic hand surgeon proceeded with surgical intervention. Preoperative examination under anesthesia confirmed significant laxity on valgus stress without a palpable endpoint (Figures 5A, 5B).

During surgery, retraction of the extensor hood revealed the completely torn and displaced UCL, entrapped dorsally and proximally to the adductor aponeurosis, characteristic of a Stener lesion. After the primary repair of the UCL, the extensor hood was seen partially retracted in a normal location superficial to the normal deep position of the repaired UCL (Figures 6A, 6B).

Discussion

Hyperabduction injuries to the thumb may rupture the UCL of the MCP joint of the thumb or cause a bony avulsion of the base of the proximal phalanx. Injury to the UCL, most often at its distal portion,^4,14,15 may result in a sprain or full-thickness tear of the ligament.

Subsequently, the ligament may remain in situ, or the proximal segment may retract proximal to the adductor aponeurosis with continued abduction of the thumb. On release of the abduction force, the proximal UCL segment is displaced dorsally and proximally by the inferior aspect of the adductor aponeurosis. The UCL becomes entrapped by the adductor aponeurosis and cannot reduce spontaneously.¹⁵ This displacement was initially described by Stener⁶ in 1962 and is referred to as a Stener lesion (Figures 1A-1C).

It is vital for the radiologist to identify a Stener lesion because a nondisplaced tear of the UCL is often treated nonsurgically, but UCL tears displaced more than 3 mm and Stener lesions usually must be operated on to avoid chronic instability, pain, and osteoarthritis.^2-5,8,12-23 Sensitivity and specificity of MRI in evaluating UCL injuries are reported to be almost 100%, with resolution of 1 mm using current surface coils.²³ There are various UCL injury patterns, including partial tears, displaced and nondisplaced complete tears, and even complex injuries, such as an incomplete tear with the torn portion retracted as a Stener lesion.²² MRI is needed to establish the extent of injury, as 90% of complete tears that are displaced at least 3 mm, and all tears with retraction proximal and superficial to the aponeurosis (true Stener lesions), failed immobilization and required surgical treatment.²³Although they vary in the literature, mean sensitivity and specificity of ultrasound in detecting UCL tears in level I studies have been reported as 76% and 81%, respectively.²⁴ When Melville and colleagues²¹ applied their ultrasound criteria—including absence of normal UCL fibers traversing the first MCP joint as well as heterogeneous masslike tissue at least partially proximal to the apex of the metacarpal lateral tubercle—they were able to distinguish displaced full-thickness tears from nondisplaced full-thickness tears with 100% accuracy. Hergan and colleagues²⁵ found that the diagnostic accuracy of MRI was superior to that of ultrasound; while MRI accuracy was perfect, 12% of patients were incorrectly diagnosed with ultrasound, with false-positive or false-negative tendon-edge displacement. In our experience, ultrasound is uniquely useful in its ability to characterize the real-time dynamic interaction of the UCL with the adductor aponeurosis. It has been observed that passive flexion of the first interphalangeal joint moves the adductor aponeurosis in isolation, allowing differentiation from the subjacent UCL.²¹ Had a partial tear been in the differential diagnosis of our patient’s Stener lesion, such a maneuver under ultrasound visualization would have solved the dilemma. In addition, ultrasound allows for comparison with the contralateral ligament at the time of examination should a diagnostic dilemma arise.

As many have reported both bony avulsion of the base of the proximal phalanx and concomitant injury to the UCL, identification of a bony avulsion does not exclude a ligamentous injury and the possibility of a Stener lesion (Figure 7).^16,19

In one study, 14% of patients with injury to the UCL sustained a concomitant bony avulsion of the UCL insertion.²³ However, presence of the avulsion fragment did not alter management, and only those fragments involving more than 20% of the articular surface were considered true fractures and treated as such.

Conclusion

A Stener lesion—retraction of a completely torn UCL becoming entrapped dorsally and proximally to the adductor insertion—can cause pain, instability, and ultimately osteoarthritis if not treated appropriately. The orthopedic surgeon should have a high index of suspicion for a Stener lesion in the appropriate clinical scenario and consider all imaging modalities for diagnosis. Likewise, it is of utmost importance for the radiologist to identify imaging findings of a Stener lesion, as physical examination alone may be limited in its ability to characterize injury severity. Both MRI and ultrasound are useful in evaluating UCL tears, and ultrasound provides the additional benefit of dynamic visualization and comparison with the contralateral side.

Am J Orthop. 2017;46(3):E195-E199. Copyright Frontline Medical Communications Inc. 2017. All rights reserved.

Take-Home Points

• Torn, displaced, and entrapped UCL is a Stener lesion.
• Hyperabduction injury with pain and joint laxity on examination.
• MRI and ultrasound are useful in evaluating UCL tears.
• Ultrasound offers dynamic evaluation.
• Must be treated appropriately to avoid pain, instability, and osteoarthritis.

In the literature, hyperabduction injuries to the thumb metacarpophalangeal (MCP) joint have been referred to interchangeably as gamekeeper’s thumb and skier’s thumb. Historically, though, gamekeeper’s thumb was initially described in hunters with chronic injury to the ulnar collateral ligament (UCL),¹ and skier’s thumb typically has been described as an acute hyperabduction injury of the UCL.^2-5 The proximal portion of a torn UCL may retract with further abduction and displace dorsally, becoming entrapped by the adductor pollicis aponeurosis insertion, known as a Stener lesion.⁶

The first MCP joint is stabilized by static and dynamic structures that contribute in varying degrees in flexion and extension of the joint. The static stabilizers include the proper and accessory radial and UCLs, the palmar plate, and the dorsal capsule. The UCL originates at the dorsal ulnar aspect of the first metacarpal head at the metacarpal tubercle about 5 mm proximal to the articular surface. The UCL courses distally in the palmar direction to insert volar and proximal to the medial tubercle of the proximal phalanx about 3 mm distal to the articular surface.⁷ In flexion, the proper collateral ligament is taut and is the primary static stabilizer. In extension, the accessory collateral ligament, which inserts on the palmar plate, is taut and is the primary static stabilizer.^8-11

The dynamic stabilizers include the extrinsic muscles (flexor pollicis longus, extensor pollicis longus and brevis) and the intrinsic muscles (abductor pollicis brevis, adductor pollicis, flexor pollicis brevis) inserting on the thumb at the distal phalanx and proximal phalanx and at the base of the first metacarpal.^8-10

The adductor pollicis originates from the volar third metacarpal, capitate, and hamate and has a dual insertion on the thumb.¹² There is a direct insertion onto the palmar proximal phalanx at the medial tubercle, distal and dorsal to the phalangeal insertion of the UCL.

There is also a broad aponeurosis that inserts onto the extensor hood expansion, dorsal to the insertion of the UCL (Figures 1A-1C and 2A, 2B).^7,8,13

We report the case of an acute hyperabduction injury of the thumb MCP joint with radiographic, ultrasound, and magnetic resonance imaging (MRI) findings consistent with a Stener lesion and subsequently confirmed with intraoperative photographs. The patient provided written informed consent for print and electronic publication of this case report.

Clinical Findings

A 33-year-old healthy man had persistent left hand pain and grip weakness after performing a handstand. He presented to the orthopedic hand clinic 20 days after injury, having failed nonoperative management (use of nonsteroidal anti-inflammatory drugs and soft thumb spica splint). Physical examination revealed soft-tissue swelling and focal tenderness to palpation at the ulnar aspect of the thumb MCP joint. Despite bilateral first MCP joint laxity on varus and valgus stress without identification of a firm endpoint, pain was elicited only on valgus stress of the left first MCP joint. Given the laxity and the left thumb soft-tissue swelling with pain, plain radiographs, ultrasound, and MRI were used to evaluate for severity of presumed left thumb UCL injury.

Imaging Findings

Plain radiographs showed normal bony anatomy without fracture, normal joint space, and mild soft-tissue swelling at the left thumb MCP level (Figures 3A, 3B).

Ultrasound confirmed a complete tear of the UCL, which was flipped in a proximal direction and projected dorsally in relation to the direct insertion of the adductor tendon (Figure 2B). MRI showed focal disruption of the UCL at the level of the left thumb MCP joint with associated MCP joint effusion (Figures 4A-4F).

Low T1 signal intensity over the adductor aponeurosis at the level of the metacarpal head corresponded with the torn and proximally retracted UCL. There was associated bone marrow edema at the radial and volar aspects of the thumb metacarpal head and low-grade strain of the abductor pollicis brevis. The thumb flexor and extensor tendons appeared normal. Although possibly secondary to patient positioning, mild volar subluxation of the proximal phalanx in relation to the metacarpal head was queried.

 

Surgical Findings

Given laxity with pain at the UCL on stress testing, MRI and ultrasound findings, and continued pain and instability of the thumb with pinching and grasping during activities of daily living, the patient and orthopedic hand surgeon proceeded with surgical intervention. Preoperative examination under anesthesia confirmed significant laxity on valgus stress without a palpable endpoint (Figures 5A, 5B).

During surgery, retraction of the extensor hood revealed the completely torn and displaced UCL, entrapped dorsally and proximally to the adductor aponeurosis, characteristic of a Stener lesion. After the primary repair of the UCL, the extensor hood was seen partially retracted in a normal location superficial to the normal deep position of the repaired UCL (Figures 6A, 6B).

Discussion

Hyperabduction injuries to the thumb may rupture the UCL of the MCP joint of the thumb or cause a bony avulsion of the base of the proximal phalanx. Injury to the UCL, most often at its distal portion,^4,14,15 may result in a sprain or full-thickness tear of the ligament.

Subsequently, the ligament may remain in situ, or the proximal segment may retract proximal to the adductor aponeurosis with continued abduction of the thumb. On release of the abduction force, the proximal UCL segment is displaced dorsally and proximally by the inferior aspect of the adductor aponeurosis. The UCL becomes entrapped by the adductor aponeurosis and cannot reduce spontaneously.¹⁵ This displacement was initially described by Stener⁶ in 1962 and is referred to as a Stener lesion (Figures 1A-1C).

It is vital for the radiologist to identify a Stener lesion because a nondisplaced tear of the UCL is often treated nonsurgically, but UCL tears displaced more than 3 mm and Stener lesions usually must be operated on to avoid chronic instability, pain, and osteoarthritis.^2-5,8,12-23 Sensitivity and specificity of MRI in evaluating UCL injuries are reported to be almost 100%, with resolution of 1 mm using current surface coils.²³ There are various UCL injury patterns, including partial tears, displaced and nondisplaced complete tears, and even complex injuries, such as an incomplete tear with the torn portion retracted as a Stener lesion.²² MRI is needed to establish the extent of injury, as 90% of complete tears that are displaced at least 3 mm, and all tears with retraction proximal and superficial to the aponeurosis (true Stener lesions), failed immobilization and required surgical treatment.²³Although they vary in the literature, mean sensitivity and specificity of ultrasound in detecting UCL tears in level I studies have been reported as 76% and 81%, respectively.²⁴ When Melville and colleagues²¹ applied their ultrasound criteria—including absence of normal UCL fibers traversing the first MCP joint as well as heterogeneous masslike tissue at least partially proximal to the apex of the metacarpal lateral tubercle—they were able to distinguish displaced full-thickness tears from nondisplaced full-thickness tears with 100% accuracy. Hergan and colleagues²⁵ found that the diagnostic accuracy of MRI was superior to that of ultrasound; while MRI accuracy was perfect, 12% of patients were incorrectly diagnosed with ultrasound, with false-positive or false-negative tendon-edge displacement. In our experience, ultrasound is uniquely useful in its ability to characterize the real-time dynamic interaction of the UCL with the adductor aponeurosis. It has been observed that passive flexion of the first interphalangeal joint moves the adductor aponeurosis in isolation, allowing differentiation from the subjacent UCL.²¹ Had a partial tear been in the differential diagnosis of our patient’s Stener lesion, such a maneuver under ultrasound visualization would have solved the dilemma. In addition, ultrasound allows for comparison with the contralateral ligament at the time of examination should a diagnostic dilemma arise.

As many have reported both bony avulsion of the base of the proximal phalanx and concomitant injury to the UCL, identification of a bony avulsion does not exclude a ligamentous injury and the possibility of a Stener lesion (Figure 7).^16,19

In one study, 14% of patients with injury to the UCL sustained a concomitant bony avulsion of the UCL insertion.²³ However, presence of the avulsion fragment did not alter management, and only those fragments involving more than 20% of the articular surface were considered true fractures and treated as such.

Conclusion

A Stener lesion—retraction of a completely torn UCL becoming entrapped dorsally and proximally to the adductor insertion—can cause pain, instability, and ultimately osteoarthritis if not treated appropriately. The orthopedic surgeon should have a high index of suspicion for a Stener lesion in the appropriate clinical scenario and consider all imaging modalities for diagnosis. Likewise, it is of utmost importance for the radiologist to identify imaging findings of a Stener lesion, as physical examination alone may be limited in its ability to characterize injury severity. Both MRI and ultrasound are useful in evaluating UCL tears, and ultrasound provides the additional benefit of dynamic visualization and comparison with the contralateral side.

Am J Orthop. 2017;46(3):E195-E199. Copyright Frontline Medical Communications Inc. 2017. All rights reserved.