Retrospective Review of Radiographic Imaging of Tibial Bony Stress Injuries in Adolescent Athletes With Positive MRI Findings: A Comparative Study

Abstract

Background: It is difficult to diagnose and grade bony stress injury (BSI) in the athletic adolescent population without advanced imaging. Radiographs are recommended as a first imaging modality, but have limited sensitivity and, even when findings are present, advanced imaging is often recommended.

Hypothesis: It was hypothesized that the significance of radiographs is underestimated for BSI in the adolescent with positive clinical examination and history findings.

Study design: Case series.

Level of evidence: Level 4.

Methods: A total of 80 adolescent athletes with a history of shin pain underwent clinical examination by an orthopaedic surgeon. On the day of clinical examination, full-length bilateral tibial radiographs and magnetic resonance imaging (MRI) scans were obtained. MRI scans were reviewed using Fredericson grading for BSI. At the completion of the study, radiographic images were re-evaluated by 2 musculoskeletal (MSK) radiologists, blinded to MRI and clinical examination results, who reviewed the radiographs for evidence of BSI. Radiographic results were compared with clinical examination and MRI findings. Sensitivity, specificity, negative predictive value, and positive predictive value were calculated based on comparison with MRI.

Results: All radiographs were originally read as normal. Of the tibia studied, 80% (127 of 160) showed evidence of BSI on MRI. None of the original radiographs demonstrated a fracture line on initial review by the orthopaedic surgeons. Retrospective review by 2 MSK radiologists identified 27% of radiographs (34 of 127) with evidence of abnormality, which correlated with clinical examination and significant findings on MRI. Review of radiographs found evidence of new bone on 0 of 28 Fredericson grade 0, 0 of 19 Fredericson grade I, 11 of 80 (13.7%) Fredericson grade II, 18 of 28 (64%) Fredericson grade III, and 5 of 5 (100%) Fredericson grade IV. Sensitivity of radiographs showed evidence of new bone on 27% (34 of 127) of initial radiographs, with presence more common with greater degree of BSI, as 23 of 33 (70%) were higher-grade injuries (III of IV) of BSI. Specificity and positive predictive value were 100%, while negative predictive value was 17%.

Conclusion: These findings highlight the importance of initial radiographs in identifying high-grade BSI. As radiographs are readily available in most office settings of sports medicine physicians, this information can influence the management of adolescent athletic BSI without the need to delay treatment to obtain an MRI.

Clinical relevance: Adolescent athletes with radiographic evidence of BSI should be treated in a timely and more conservative manner, given the likelihood of higher-grade BSI. In addition, clinicians knowledgeable of the radiographic findings of high-grade BSI should feel more confident that a negative initial radiograph is not likely to be a high-grade BSI and can modify their treatment plans accordingly.

Keywords: MRI; adolescent athlete; bony stress injury; new bone; radiograph; radiography.

Figure 1.

Enhanced view of subtle periosteal new bone with use of magnification enhancement.

Figure 2.

(a) Anterior posterior radiograph of the left tibula/fibula shows PNB pathology. (b) Arrow demarcating PNB formation at the left posterior mid to distal cortex seen on lateral radiograph of same patient. (c) Axial T2 shows corresponding moderate periosteal and endosteal bone marrow edema pattern. PNB, periosteal new bone.

Figure 3.

Examples of ENB (different patients). (a) Lateral radiograph with ENB. Arrow demarcates ENB formation at the left posteromedial proximal to midcortex. (b) STIR image. AP/lateral radiograph and axial MRI indicating endosteal sclerosis. (c, d) Axial views of lateral mid tibia with ENB; 3 had both evidence of PNB and ENB. A total of 27 tibia demonstrated that bony changes were seen on lateral radiographs, 17 were seen on AP; 7 tibias demonstrated findings on both AP and lateral views. Overall, 34 of 124 (27%) tibia showed evidence of PNB or ENB formation that coincided with evidence of BSI on MRI and noted location of pain on clinical examination. Four tibias demonstrated discordant PNB, which was not associated with a site of pain but was associated with patients who noted a previous BSI. The average thickness of PNB ranged from 1.5 mm to 5 mm (average, 2.5 mm) and all findings (27 of 146) on lateral view films were located on the posterior cortex. PNB was found in the proximal third of the tibia on 9 radiographs, in the middle third on 19, and in 6 on the distal tibia. Bone sclerosis was seen on 3 radiographs, of which 2 were seen in the distal third of the tibia. A nonossifying fibroma was found on 4 of 160 (2.5%) of tibias imaged. Using evidence of PNB or ENB on radiograph yielded a sensitivity of 27%, specificity of 100%, PPV of 100%, and NPV of 17%. AP, anterior posterior; BSI, bony stress injury; ENB, endosteal new bone; MRI, magnetic resonance imaging; NPV, negative predictive value; PNB, periosteal new bone; PPV, positive predictive value; STIR, short TI inversion recovery.