Image guided core needle biopsy of musculoskeletal lesions: are nondiagnostic results clinically useful?

Abstract

Background: The clinical utility of nondiagnostic core needle biopsies is not fully understood. Understanding the clinical and radiologic factors associated with nondiagnostic core needle biopsies may help determine the utility of these nondiagnostic biopsies and guide clinical decision making.

Questions/purposes: We asked (1) whether benign or malignant bone and soft tissue lesions have a higher rate of nondiagnostic core needle biopsy results, and which diagnoses have the lowest diagnostic yield; (2) how often nondiagnostic results affected clinical decision-making; and (3) what clinical factors are associated with nondiagnostic but useful core needle biopsies.

Methods: A retrospective study was performed of 778 consecutive image-guided core needle biopsies of bone and soft tissue lesions referred to the musculoskeletal radiology department at a single institution. The reference standard was (1) the final diagnosis at surgery or (2) clinical followup. Diagnostic yield was calculated for the most common diagnoses. Clinical and imaging features related to each nondiagnostic core needle biopsy were assessed for their association with clinical usefulness. Useful nondiagnostic biopsies were defined as those that help guide treatment. Each lesion was assessed before biopsy by the orthopaedic oncologist as (1) "likely to be benign" or (2) "suspicious for malignancy." The overall diagnostic yield was 74%.

Results: Malignant lesions had higher diagnostic yield than benign lesions: 94% (323 of 345) versus 58% (252 of 433), yielding a relative risk (RR) of 1.61 and 95% CI of 1.48 to 1.75. Soft tissue lesions had a higher diagnostic yield than bone lesions: 82% (291 of 355) versus 67% (284 of 423); RR, 1.22; 95% CI, 1.22 (1.12-1.33). Ganglion cyst (36%, four of 11), myositis ossificans (40%, two of five), Langerhans cell histiocytosis (0%, 0 of four), and simple bone cyst 0%, 0 of six) had the lowest diagnostic yield. Of the nondiagnostic biopsies assessed for clinical usefulness by the orthopaedic oncologist, 60% (85 of 142) of the biopsies were useful in guiding clinical decision making. Useful nondiagnostic core needle biopsy results occurred more often in painless, nonaggressive lesions, assessed as "likely to be benign" before biopsy.

Conclusions: Nondiagnostic core needle biopsy results in musculoskeletal lesions are not entirely useless. At times, they can be supportive of benign processes and can help avert unnecessary surgical procedures.

Fig. 1A–C

A 56-year-old man had a melanoma soft tissue metastasis, which was diagnostic at core needle biopsy. Malignant soft tissue lesions had the highest diagnostic yield (94%) in our study. Sagittal (A) T1-weighted and (B) fat-saturated T1-weighted postcontrast MR images show a soft tissue anterior knee mass with solid enhancement (white arrow). (C) A CT-guided core needle biopsy image shows the biopsy needle in the lesion.

Fig. 2A–D

A 23-year-old man had a left proximal fibular simple (unicameral) bone cyst, which was nondiagnostic at core needle biopsy. Benign bone lesions had the lowest diagnostic yield (32%) in our study. (A) An AP radiograph shows a well-circumscribed lytic lesion in the proximal fibula (black arrow). A healing fracture is seen at the inferior margin of the lesion. The lesion (arrows) is low signal on the (B) axial T1-weighted and high signal on the (C) axial T2-weighted MR images. (D) A CT-guided core needle biopsy image shows a biopsy needle in the lesion.

Fig. 3A–C

A 54-year-old woman has a focus of red marrow in the distal femur, which was nondiagnostic at core needle biopsy. This lesion was painless and seen incidentally on routine knee MR images. Given the lack of pain and the MRI features suggestive of red marrow, no additional intervention was performed. Followup MRI studies over 2 years showed no change. (A) The sagittal T1-weighted MR image shows a well-circumscribed lesion with intermediate T1 signal intensity (white arrow) that is higher than muscle (asterisk), suggesting red marrow. On the (B) sagittal T2-weighted MR image, the lesion is high signal (white arrow) relative to muscle (white asterisk). (C) A CT-guided core needle biopsy image shows the biopsy needle in the lesion.

Fig. 4A–C

A 67-year-old woman had a history of breast cancer and a painful left humerus lesion, which was nondiagnostic at core needle biopsy. Given the pain, history of malignancy, and aggressive radiographic features, surgical biopsy was performed revealing a plasmacytoma. (A) An AP radiograph shows a lytic lesion with a wide transition zone and surrounding periostitis (black arrow) in the midhumeral shaft. The (B) bone scan shows focal radiotracer uptake in the lesion (black arrow). (C) A CT-guided core needle biopsy image shows the biopsy needle in the lesion.