A thirty-eight-year-old man presented to us with a painful enlarging mass on the medial aspect of his left knee. The pain started after a fall while he was walking approximately two months previously. The pain occurred mostly on weight-bearing, and the patient reported that he had not had any pain at night or at rest initially. However, he fell again about two weeks prior to presentation and subsequently noticed a mass. The mass had been enlarging since then along with worsening of the pain; recently, he had also noticed the onset of pain at night and at rest. He reported that he had not had any constitutional symptoms or any numbness or weakness in the limb. On examination, the patient was a healthy-appearing man who was in mild discomfort. There was tender, diffuse swelling over the anteromedial side of his left knee. There was increased warmth but no erythema or overlying skin changes. The mass was noncompressible and did not change with limb elevation. Quadriceps atrophy was visible. The knee was painful during active and passive range of motion, and grossly it appeared stable. There was no inguinal lymphadenopathy, and the distal neurovascular examination had normal findings.
Radiographs showed an expansile, eccentric lytic lesion in the medial femoral condyle (Fig. 1). The computed tomographic (Fig. 2) and magnetic resonance imaging scans showed a 6.1 × 5.4 × 4.2-cm lesion in the medial femoral condyle. No fluid levels were seen on the magnetic resonance imaging scan. The cortices were markedly thinned and expanded, and there was a soft-tissue component. No matrix formation was evident. The lesion involved the subchondral area and crossed the midline to involve a part of the lateral condyle. A computed tomography scan of the chest and a bone scan revealed no other focus of disease. Serum calcium, phosphate, and alkaline phosphatase levels were normal. A core needle biopsy showed numerous multinucleated giant cells of an osteoclastic type with large groups of small spindle-shaped cells, most of which were consistent with a giant-cell tumor. After the available management options were discussed with the patient, a decision was made to perform a distal femoral replacement.
During preoperative medical clearance, the patient was diagnosed as having essential hypertension (160/115 mm Hg). He underwent a cardiac workup, which included a normal stress test and an ejection fraction of 66% on echocardiogram. He underwent a wide local excision of the tumor, and a distal femoral replacement was inserted with cement one week later, once the blood pressure was stabilized. A fixed-hinge bearing nonmodular implant (BIAM New Materials [Jing Hang], Beijing, China) was used. The perioperative period was uneventful, and the patient left the hospital on the fourteenth postoperative day.
The gross specimen measured 13.0 × 8.0 × 6.0 cm, and the histological findings were consistent with a benign giant-cell tumor, consisting of uniformly distributed mononuclear cells and multinucleated giant cells resembling osteoclasts. The mononuclear cells had round, ovoid, or elongated nuclei often with prominent nucleoli. Hemorrhage, necrosis, and small foci of reactive osteoid formation were found in some areas. Mitotic figures were rare and ranged from zero to ten per ten high-power fields. Neither atypical mitoses nor cytological atypia were present. The margins were negative, and no vascular involvement was seen. The findings on microbiological examination were negative. A diagnosis of a benign giant-cell tumor of the distal end of the femur was made.
Forty months later, the patient presented to the emergency room with progressive chest tightness, shortness of breath, and swelling in the lower limbs. A computed tomography scan of the chest (Fig. 3) and a transthoracic echocardiogram showed an enlargement of the right ventricle with a mass in the ventricular cavity obstructing the tricuspid valve and extending to the right atrium, creating concerns of a possible acute obstruction26,27. A computed tomography scan of the abdomen and pelvis did not reveal any other focus of disease. Serum calcium, phosphate, and alkaline phosphatase levels were normal. Cardiac exploration showed a 5.5 × 4.8 × 4.4-cm mass that was soft, smooth, gray, and deep red in the right ventricle and had invaded the anterior part of the ventricular wall, including the epicardium, the tricuspid valve, and the right atrium. There was a large amount of old thrombus in the right atrium, which was cleared. An intralesional resection was attempted; however, complete resection was not possible because of the extent of the lesion. A pericardial patch was used to reconstruct the defect in the ventricular wall. No adjuvant treatment was given.
The histopathological and immunohistochemical findings were consistent with a benign giant-cell lesion, similar to the previously excised lesion in the distal end of the left femur (Fig. 4). The mononuclear stromal cells expressed Ki67 (a proliferation marker) in about 5% to 10% of the cells; P53 and P63 (tumor-associated markers) in 30% to 50% and 20% of the cells, respectively; smooth muscle antigen in <5% of the cells; and BCL2 (or apoptosis marker) in <10% of the cells. Because of this unusual metastasis and the question of malignancy, this case was reviewed again by several experienced pathologists, who see about 100 cases of giant-cell tumor per year. All agreed that this was a so-called benign metastasis. Although the patient showed some clinical improvement in the beginning, he redeveloped progressive cardiac failure and subsequently renal and liver failure. He refused any further medical intervention and died nine months later because of multiorgan failure.
Giant-cell tumor of bone is regarded as an aggressive neoplasm of the skeletal system, which can metastasize despite its benign appearance1-13. Ever since the first description of the metastatic potential of a benign giant-cell tumor by Finch and Gleave in 19261, increased awareness and improved methods of detection have led to the increased recognition of metastatic lesions6. Although the lungs have been the most common site of metastases, involvement of other visceral organs has also been reported occasionally1-25. However, to our knowledge, a metastasis to the heart from a benign giant-cell tumor of the bone has never been reported.
The data on metastatic giant-cell tumor is mostly derived from reports of lung metastases. Most metastases are hematogenous, although lymphatic spread has been reported2,23. Postulated mechanisms of metastases include active, so-called biologic predeterminism24 of the giant cells to invade the interstitium, destroy vessel walls, and intravasate by means of enzymes with subsequent extravasation, lodgment, and growth in target organs. Since the histological characteristics of the nodules found in the lung are identical to those of the so-called benign tumor of the primary site, others consider these as implants derived from tumor emboli formed in areas of hemorrhage and thrombus formation within the primary tumor and not as true metastases5-11. Although few studies have correlated the biologic activity of the giant cell tumor with the mitotic index, and overexpression of Ki67 and p53 proteins, there is currently no definitive agreed-on method to predict distant metastasis in giant-cell tumor on the basis of histological analysis, immunohistochemistry, proliferation index, vascular density, or flow cytometry14,25,28,29.
Most metastases are diagnosed within the first few years after the diagnosis of the primary tumor, although isolated cases of longer intervals, as much as twenty-four years, have been reported2-10. Reported potential risk factors for metastasis in giant-cell tumor include the primary site, local recurrences, and the method of treatment of the primary tumor2-10,24,25. Common primary sites with a predilection to metastases include the distal end of the radius, the distal end of the femur, the proximal part of the tibia, and the sacrum. Patient age or sex has not been identified as a risk factor for metastatic disease. The role of surgical manipulation in promoting metastases is controversial. The growth rate (as measured by doubling time) of metastases is extremely low, supporting the concept that they probably were present at the time or even before the diagnosis of the primary lesion13. However, it is estimated that there is a sixfold increase in the risk for development of lung metastases after a local recurrence of giant-cell tumor5-10. In larger series, the likelihood of recurrence and metastasis was also found to be increased in tumors that were radiographically aggressive, especially those with soft-tissue extension2-10.
The natural history of these benign metastatic lesions is also unpredictable. Complete excision of metastases has been very successful with good long-term survival, but visceral metastases have been reported to be fatal in up to 43% of cases2-10. This wide range may be explained by variable interpretation of what constitutes a histologically benign or malignant giant-cell tumor, the relatively small number of patients in each series, the different time frames of investigation, and the variable lengths of follow-up in various series. However, the mortality rate has been declining and, in a recently published study, none of the twenty-four patients with visceral metastases died because of the disease2. This may be due to earlier diagnosis (smaller and fewer lesions) by the use of advanced imaging modalities (such as computed tomography compared with chest radiographs in the older studies) and subsequent aggressive surgical excision of metastatic lesions. These suggest that the prognosis of giant-cell tumor with visceral metastases may be more favorable than was originally thought2-10. Although most authors favor excision whenever feasible, there are several reports where the metastases have completely regressed spontaneously or, after a biopsy, have remained static for years2-10. Long-term survival has been possible even with residual and/or multiple pulmonary lesions2-10. Thus, if the lesions appear stable on serial imaging studies without pulmonary compromise, then surgery can be delayed or even avoided. However, we believe that an aggressive approach may be necessary for a cardiac metastasis because of the potential of mechanical obstruction and/or catastrophic embolization26,27,30.
Although radiation and/or traditional chemotherapy have occasionally been used for metastatic disease, their role remains unclear, and these modalities are not recommended in most cases of benign giant-cell tumor2-10. On the other hand, several other systemic therapies have been described. Corticosteroids have been successfully used in the control of unresectable metastases23. Bisphosphonates have been shown to inhibit the growth of giant-cell-tumor cell lines with a reduction in the local recurrence rate30,31. Recently, Thomas et al.32 reported on the use of denosumab (human monoclonal antibody that specifically inhibits RANKL [receptor activator of nuclear factor-?B ligand] and, thus, the osteoclast-mediated bone destruction) in recurrent or unresectable giant-cell tumor. However, more studies are warranted to document the role of these novel agents in the treatment of metastatic giant-cell tumor.
Tumors of the heart are rare, occurring in 0.001% to 0.02% in pooled autopsy series26,27. Metastatic heart tumors occur more frequently, at a rate that is at least 100 times higher than that of primary tumors, and typically in patients with disseminated disease26,27. Malignant tumors from all tissues except the central nervous system have been reported to metastasize to the heart26. Most of these are melanomas; pulmonary, breast, or esophageal carcinomas; or lymphomas and leukemias. A solitary cardiac metastasis from an extremity sarcoma has been described33. Metastases may disseminate to the heart by lymphatic or hematogenous pathways, local spread, or by a transvenous extension26,27. Most of them are subclinical, and less than one-third are diagnosed before death. Suspicion should be raised and diagnostic studies employed as soon as symptoms of heart failure, chest pain, new-onset heart murmurs, embolism, or rhythm disturbances develop or as soon as there is radiographic evidence of cardiomegaly. The location in the heart, not the histologic subtype, determines the symptoms, which can mimic any form of heart disease. Patients typically present with the triad of intracardiac obstruction, distant embolization, and systemic or constitutional symptoms26,27,33. The management of cardiac metastatic disease is difficult and is dictated by the nature of the primary tumor, the previous therapy, the extent of metastatic spread, signs and symptoms, and the feasibility of resection26,27,33-36. A combination of echocardiography, computed tomography, and cardiac magnetic resonance imaging provides excellent anatomical definition for the purpose of preoperative planning36.
Although distant metastasis from a so-called benign giant-cell tumor is rare, it is important to recognize this and differentiate it from malignant giant-cell tumor as well as from the usual malignant metastases. The histological characteristics of these "benign" metastases are usually the same as those of the benign primary lesion with no tendency to dedifferentiate. On the other hand, a primary malignant giant-cell tumor is a lesion in which there are areas of synchronous high-grade sarcoma growth next to areas of benign giant-cell tumor37. A secondary malignant giant-cell tumor is a metachronous high-grade sarcomatous growth superimposed on a previous, biopsy-verified, benign giant-cell tumor that has been treated by surgery and/or radiation therapy37. This differentiation is important because the outcomes associated with malignant giant-cell tumors are usually poor, in contrast to the benign metastases. In a study from the Rizzoli Institute37, eleven of seventeen patients with malignant giant-cell tumor died from visceral metastases.
In conclusion, the sudden onset of cardiovascular symptoms in a patient with a well-controlled primary tumor should evoke suspicion of cardiac metastasis. Increased awareness of the possibility of latent cardiac metastasis after complete resection of a primary noncardiac tumor may aid in the diagnosis of the cardiac mass and may help to direct further appropriate management.
Note: The authors thank Martin M. Malawer, MD, for his evaluation and critical comments made during the preparation of this manuscript.