Lytic Lesions of Bone: A Histopathological and Radiological Correlative Study

Academic Medical Journal of India 83 Volume III - Issue 3 www.medicaljournal.in July - September 2015 ORIGINAL RESEARCH Lytic Lesions of Bone: A Histopathological and Radiological Correlative Study Lilarani Vijayaraghavan, a Lailaraji N, a Gopakumar TS, b Roy N, c Neelima Radhakrishnan d a. Department of Pathology, ; b. Department of Orthopedics, Government Medical College, Thiruvananthapuram, ; c. Department of Radiodiagnosis, ; d. Department of Pathology, ESIC Medical College, Paripally, Kollam, * Corresponding Author: Lilarani Vijayaraghavan, Department of Pathology, Government Medical College Trivandrum, Medical College Post, Thiruvananthapuram,. Email: neelimajayadevan@gmail.com Abstract Published on 28th August, 2015 In this retrospective study, the importance of cooperation between the surgeon, the radiologist, and the pathologist in diagnosing a lesion of bone was highlighted. (1) To classify lytic lesions of bone according to the histopathological diagnosis. (2) To study the clinical as well as the radiological findings. (3) To correlate the radiological diagnosis with the histopathological diagnosis. (4) To measure the agreement between the radiological and histopathological diagnosis on specific lesions.the radiological characteristics of lytic lesions were assessed based on the Lodwick system and the type of periosteal reaction present. These lytic lesions of bone were histopathologically assessed, and the radiological diagnosis was correlated with the histopathological diagnosis. The agreement between the radiological and histopathological diagnosis on specific lesions was measured statistically and compared with similar other studies conducted by different authors. Of the 212 cases studied, 80% cases showed radiological and pathological correlation. The radiologic-pathologic correlation in our study was low compared to other studies in the cases of tumor like conditions and primary bone tumors. In all the other cases correlation percentage remained the same. Key Words: Lytic Lesions, Bone, Radiology, Histopathology, Lodwick Classification, World Health Organization Classification Cite this article as: Vijayaraghavan L, Lailaraji N, Gopakumar TS, Roy N, Radhakrishnan N. Lytic Lesions of Bone: A Histopathological and Radiological Correlative Study. Academic Medical Journal of India. 2015 Aug 28;3(3):83-87. Introduction A lytic lesion of bone is defined as the destruction of an area of bone, due to a disease process. It can be due to benign tumors, malignant tumors, infections, metastasis to bone, and numerous other tumor like conditions. Many of the bone lesions look similar and are nearly impossible to differentiate by plain film alone. A concerted multidisciplinary approach including clinical, radiological and histopathology in needed in the early diagnosis of lytic bone lesions. 1 Conventional X-rays are always the initial method of detection and localization of a tumor. 2 A definite pathologic interpretation should never be rendered if radiological information is not available. Furthermore, after a biopsy the radiologic features may get altered. Hence, imaging should be done before attempting a bone biopsy. 2 This study was undertaken with the objectives to classify the lytic bone lesions according to the histopathological diagnosis and to correlate the radiological diagnosis with the histological diagnosis. Objectives 1. To classify lytic lesions of bone according to the histopathological diagnosis 2. To study the clinical as well as the radiological findings 3. To correlate the radiological diagnosis with the histopathological diagnosis 4. To measure the agreement between the radiological and histopathological diagnosis on specific lesions. Methods Materials for this study were obtained from Department of Orthopaedics, Medical College, Thiruvananthapuram,. This retrospective study was conducted for a period of 2-year starting from May 2005 to December 2007. The study was started after obtaining the clearance from the Ethical Committee of our institute. Bone biopsies, amputation specimens with the relevant clinical findings, X-ray findings, computed tomography, and magnetic resonance imaging (MRI) findings, clinical and radiological differential diagnoses were recorded from the patient request form received in our Pathology Department. Further details if needed any were collected from the orthopedicians. X-rays were collected and discussed with consultants of radiodiagnosis. Radiological parameters, such as site of involvement, pattern of destruction, margins, expansion of cortical shell, sclerotic margin, periosteal reaction, and soft tissue extension, were studied to obtain a definite impression. The radiological features were assessed based on the Lodwicks method of classification. Both amputation specimens as well as image-guided biopsy samples were studied. The specimens received were fixed in 10% formalin, grossed, processed and sections taken from paraffin embedded tissues. The sections were stained with routine hematoxylin and eosin stains. Immunohistochemical stains *See End Note for complete author details

Academic Medical Journal of India 84 Volume III - Issue 3 were done if indicated. For bone tumors and tumor like lesions classification and histopathological findings proposed by the World Health Organization (WHO) was adopted. Tumors were divided into benign and malignant according to WHO classification. Our final histopathological diagnoses were correlated with the radiological differential diagnoses. The measurement of agreement was calculated based on Cohens Kappa value. Results A total of 212 cases were studied with histopathological follow-up. Lytic lesions of bone were classified into five broad categories. Among the 212 cases, 64 were inflammatory conditions, 21 were tumor-like lesions, 44 were benign tumors, 43 were primary bone malignancies, and 40 were secondary bone malignancies (Figure 1). The most frequent age group affected by lytic lesions of bone in our study was 10-19 years (Figure 2). The most common lesion in this age group was osteosarcoma. 124 cases were male and 88 were female (Figure 3). Except fibrous dysplasia, chondrosarcoma, solitary bone cyst, and multiple myeloma all other lesions showed male predominance. The most frequent presenting complaints were pain and swelling (194 and 87 cases), respectively. All the cases of metastases presented with a clinical history of pathological fracture. Most of the inflammatory conditions presented with fever. In our study, the most frequent radiological appearance encountered was lytic lesion with well-defined margin, which was detected in 113 patients. 95 cases showed lytic lesion with ill-defined margins. Matrix calcification was seen in only one out of five chondrosarcomas. 20 out of 33 cases of giant cell tumor showed typical soap bubble appearance in radiology. Tumors with soft tissue extension were clearly delineated in MRI. A few representative lesions with their imaging findings and corresponding histopathological features are shown in Figures 4 and 5. The most frequent aspirated site was a femur. Among the 212 cases of lytic lesions of bone, 171 showed pathologic radiologic correlation (80.66%). In 5 cases, diagnosis was inconclusive due to inadequate material. Negative correlation Figure 3: Male female ratio of lytic lesions of bone in this study Figure 1: Five broad categories of lytic bone lesions in the present study a b c d Figure 2: Age groups affected by lytic bone lesions e f g Figure 4: (a) Well-defined lytic lesion bone; (b) diaphyseal lesion with multiple lucencies; (c) diaphyseal lytic lesion with multiple lucencies in the left tibia; (d) lytic lesion femur with sharply defined margins; (e) lytic lesion tibial condyle with soap bubble appearance; (f) lytic lesion with soft tissue expansion and periosteal elevation; (g) magnetic resonance imaging second metatarsal bone - Intraosseous mass with cortical erosion and soft tissue component; (h) pathological fracture trochanter h

Academic Medical Journal of India 85 Volume III - Issue 3 with radiological findings was seen in 38 cases. The percentage of lytic lesions showing histopathological and radiological correlation is given in Tables 1 and 2. Discussion Lytic lesions of bone constitute a major proportion of orthopedic specimens in our department. The clinical manifestations of bone lesions are often nonspecific, and they are often mistaken for osteomyelitis or trauma. The appropriateness criteria established by the American College of Radiology, dictate that for the initial evaluation of a bone lesion radiographs should be the first line of investigation. If the radiograph shows normal or indeterminate findings, additional imaging studies are frequently required. In 1958, Jaffe and Hydson pointed out the importance of cooperation between the surgeon, the radiologist and the a e b f Figure 5: (a) Bone marrow granuloma with Langhans giant cell and epithelioid cells; (b) adamantinoma with epithelial and spindle cell elements; (c) spindle cell adamantinoma with sheets of spindle cells; (d) eosinophilic granuloma with eosinophils, magrophages with grooved nuclei and lymphocytes; (e) giant cell tumor with osteoclast like giant cells and plump stromal spindle cells; (f) osteosarcoma with pleomorphic malignant spindle cells and intervening malignant osteoid; (g) Ewings sarcoma with small round cells in sheets; (h) renal cell carcinoma with sheets of clear cells and highly vascular stroma c g d h pathologist in diagnosing a lesion of bone. 1 Precise radiological differential diagnose of bone sarcomas is uncertain beyond indicating the presence of a tumor and biopsy remains the ultimate diagnostic method of choice. 3 Unlike in lesions of other organ systems, the bone pathologist must have a basic idea regarding the appearances of all bone lesions in imaging. Hence, a proper correlation between imaging and histopathological findings is often the key to appropriate diagnosis. Our study as well as similar studies by other authors pointed out that results of a percutaneous biopsy can be extremely effective and accurate as total excision of the tumor. 4 Unusual histologic features as well as rare entities always cause diagnostic difficulty whether it is biopsy or complete excision. They will show variable features in imaging also based on the different and unusual tissue elements present. In such cases, the pathologist must know whether the radiological pattern is also variable throughout. In our study, most of the tumor like conditions, metastases, lymphomas, myelomas and rare tumors like angiosarcomas were difficult to correlate between histopathology and radiology as the imaging features were overlapping. Yamaguchi et al. in 1996 studied 52 cases of metastatic bone tumors are usually classified pointed out that into osteolytic, osteoblastic mixed, and intertrabecular corresponding to their radiological patterns Radiographs and bone scans often fail to show metastatic lesions especially those with the intertrabecular pattern. 5 Murphey et al. pointed out that it is not possible to differentiate between hemangioendothelioma, hemangiopericytoma, and angiosarcoma radiologically. These lesions are predominantly lytic and may reveal a honeycomb or hole within hole appearance similar to that of hemangioma. 6 Many of these tumors are difficult to diagnose by routine histopathology alone. Further immunohistochemical markers, hematological Table 1: Histopathological Radiological correlation Diagnosis Inflammatory condition Total number of cases Positive correlation Number (%) Inadequate sample Chronic osteomyelitis 44 40 (90.91) 4 Tuberculosis 20 17 (85) 3 Tumor like conditions Aneurysmal bone cyst 7 5 (71.43) 2 2 Nonossifying fibroma 1 1 (100) Fibrous dysplasia 5 5 (100) Intraosseous ganglion 1 0 (0) 1 Intraosseous epidermal cyst 1 0 (0) 1 Eosinophilic granuloma 1 0 (0) 1 Solitary bone cyst 4 3 (75) 1 Brown tumor of hyperparathyroidism 1 0 (0) 1 Negative correlation

Academic Medical Journal of India 86 Volume III - Issue 3 Table 2: Histopathological Radiological correlation Diagnosis Benign neoplasms Total number of cases Positive correlation Number (%) Inadequate sample Giant cell tumor 33 29 (87.87) 2 4 Desmoplastic fibroma 6 3 (50) 3 Adamantinoma 2 2 (100) Chondromyxoid fibroma 1 1 (100) Chondroblastoma 2 2 (100) Primary malignant neoplasms Osteosarcoma 18 17 (94.44) Chondrosarcoma 5 3 (60) 2 Ewing s sarcoma 9 7 (77.77) 2 Multiple myeloma 8 4 (50) 4 Epithelioid angiosarcoma 1 0 (0) 1 Lymphoma 2 0 (0) 2 Metastases to bone 40 33 (82.50) 7 Negative correlation workup, and molecular genetic studies are needed to arrive at a final diagnosis. Whereas, despite unusual histological forms of adamantinoma and different types of osteosarcoma positive correlation were possible as imaging findings and location in such cases were typical. The unusual spindle cell variant of adamantinoma case in our study was sent to two other higher centers for further confirmation despite the imaging features being classical of adamantinoma. The spindle cell variant of adamantinoma can be misdiagnosed as osteofibrous dysplasia, synovial sarcoma, Fibroblastic osteosarcoma, fibrosarcoma. The histological variants of adamantinoma can be better delineated by immunohistochemistry for cytokeratin. 7 Conclusions The results of our study were compared with many other studies similar conducted by different authors. The radiologicpathologic correlation in our study was low compared to other studies in cases of tumor like conditions such as intraosseous ganglion, eosinophilic granuloma, epidermoid cyst and primary malignancies like Ewings sarcoma, chondrosarcoma, and multiple myeloma. 8-13 In all the other lytic lesions, correlation reports were the same. The reasons for low positive correlation in these lesions may be due to the less number of cases obtained, short period of follow up and because of the nonspecific radiological findings obtained in these lesions. Limitations Few cases in this series were reported inconclusive due to inadequate sampling. A few cases in our study showed low radiologic-pathologic correlation. Yet another limitation in our study was the lack of availability of certain immunomarkers as well as molecular genetic tests to confirm the diagnosis of a few tumors. For these cases we had to depend on other higher centres within and outside India to confirm our diagnosis. Abbreviations WHO: World Health Organization CT: Computed tomographic imaging MRI: Magnetic resonance imaging End Note Author Information 1. Lilarani Vijayaraghavan, Professor,Department of Pathology, Government Medical College, Th iruvananthapuram, 2. Lailaraji N, Professor, Department of Pathology, 3. Gopakumar TS, Professor, Department of Orthopedics, 4. Roy N, Professor, Department of Radiodiagnosis, 5. Neelima Radhakrishnan, Associate Professor,Department of Pathology, ESIC Medical College, Paripally, Kollam, Conflict of Interest None declared.

Academic Medical Journal of India 87 Volume III - Issue 3 Acknowledgment We sincerely thank our Head of Department Professor S. Sankar for his critical review and comments. We also thank the Ethical Committee of our institute for giving us the permission to conduct this study. References 1. Jaffee HL, Hydson TM. Tumors and tumorous conditions of the bones and joints. J Clin Orthop 1958;17:27-39. 2. Watt I. Radiology in the diagnosis and management of bone tumours. J Bone Joint Surg Br 1985;67:520-9. 3. Dollahite HA, Tatum L, Moinuddin SM, Carnesale PG. Aspiration biopsy of primary neoplasms of bone. J Bone Joint Surg Am 1989;71:1166-9. 4. Jelinek JS, Murphey MD, Welker JA, Henshaw RM, Kransdorf MJ, Shmookler BM, et al. Diagnosis of primary bone tumors with imageguided percutaneous biopsy: Experience with 110 tumors. Radiology 2002;223:731-7. 5. Yamaguchi T, Tamai K, Yamato M, Honma K, Ueda Y, Saotome K. Intertrabecular pattern of tumors metastatic to bone. Cancer 1996;78:1388-94. 6. Murphey MD, Fairbairn KJ, Parman LM, Baxter KG, Parsa MB, Smith WS. From the archives of the AFIP. Musculoskeletal angiomatous lesions: Radiologic-pathologic correlation. Radiographics 1995;15:893-917. 7. Hazelbag HM, Taminiau AH, Fleuren GJ, Hogendoorn PC. Adamantinoma of the long bones. A clinicopathological study of thirty-two patients with emphasis on histological subtype, precursor lesion, and biological behavior. J Bone Joint Surg Am 1994;76:1482-99. 8. Schajowicz F, Clavel Sainz M, Slullitel JA. Juxta-articular bone cysts (intra-osseous ganglia): A clinicopathological study of eighty-eight cases. J Bone Joint Surg Br 1979;61:107-16. 9. Kilpatrick SE, Wenger DE, Gilchrist GS, Shives TC, Wollan PC, Unni KK. Langerhans cell histiocytosis (histiocytosis X) of bone. A clinicopathologic analysis of 263 pediatric and adult cases. Cancer 1995;76:2471-84. 10. Fisher ER, Gruhn J, Skerrett P. Epidermal cyst in bone. Cancer 1958;11:643-8. 11. Reinus WR, Gilula LA, Shirley SK, Askin FB, Siegal GP. Radiographic appearance of Ewing sarcoma of the hands and feet: Report from the Intergroup Ewing Sarcoma Study. AJR Am J Roentgenol 1985;144:331-6. 12. Björnsson J, McLeod RA, Unni KK, Ilstrup DM, Pritchard DJ. Primary chondrosarcoma of long bones and limb girdles. Cancer 1998;83:2105-19. 13. Gompels BM, Votaw ML, Martel W. Correlation of radiological manifestations of multiple myeloma with immunoglobulin abnormalities and prognosis. Radiology 1972;104:509-14.