Scintigraphy has its advantage in the early detection of perfusion or metabolism defects before radiologic abnormalities would appear. Scintigraphy can reflect perfusion defects within 24 hours if the test is available and also diagnose and predict ONFH about 14 months earlier than simple radiography [19, 26]. Meyers et al. reported a 95% diagnostic accuracy of scintigraphy in predicting avascular necrosis in his prospective study on FNF patients . Turner et al. also conducted a study with 30 FNF cases using Technetium-99m scintigraphy, and approximated the accuracy of the test at 93% .
The diagnostic accuracy for the qualitative evaluation of ONFH using SPECT in our study was 83.3% and was suboptimal compared to other studies [27, 28]. Meyer et al., however, included variable indications besides FNF, such as traumatic hip dislocation or idiopathic ischemic necrosis of the femoral head, in assessing diagnostic accuracy . In the results of Turner et al., there were two patients with absent vascular activity in the bilateral femoral head . Depending on the interpretation of these two patients, the diagnostic accuracy may vary from 93–86%. In our study, there was one female patient (13-years-old at diagnosis) whose preoperative perfusion status was viable despite the displacement (Garden type 3 at initial diagnosis). Serial clinical and radiological follow-up results showed favorable outcomes but the patient was later diagnosed with ONFH at a postoperative eight-month follow-up. Considering variable clinical factors related to osteonecrosis, such as the patient's age, the degree of displacement, the effect of surgical reduction, and the revascularization potential, the preoperative evaluation of perfusion status should be more complicated. Furthermore, there is still no clear guidelines available for displaced FNF patients, and subjective decisions or the surgeon's preference are used instead. Therefore, a rationale to increase the diagnostic accuracy and positive predictive value of preoperative SPECT/CT in assessing femoral head perfusion status is needed.
Various attempts have been made to predict ONFH by relative quantification. Stromqvist et al. reported that an uptake ratio of the lesion to the unaffected side below 0.9 preoperatively, or 1.0 postoperatively, indicated a high risk of ONFH in displaced FNF patients . Holmberg and Thorngren reported that bone union was expected when the uptake ratio to the unaffected side was over 0.90, whereas complicated outcomes were found under 0.90 . Despite the virtues, conventional planar bone scintigraphy produces poor quality images that cannot provide precise anatomical information, resulting in low specificity. Planar bone scintigraphy also produces a summated 2-dimensional image that can lead to false-negative results by overlying and adjacent bony structures. In this study, therefore, we tried to reveal the predictive value of SPECT/CT based on quantitative analysis of the perfusion status of the femoral head using an ROI.
Quantitative analysis using the head-to-head ratio in SPECT/CT showed a high diagnostic accuracy of 100% in predicting osteonecrosis, and the sensitivity and specificity were both 100%. The results were similar to the head-to-acetabulum ratio, with a diagnostic accuracy of 96.7%, and sensitivity and specificity of 85.7% and 100%, respectively. The concept of ROI in quantitative analysis helps to localize the optimal anatomical area for diagnosis and overcome the distortion of the perfusion status. Using the ipsilateral acetabular dome as a reference point for ROI is useful in certain patients whose perfusion status of the contralateral hip is not comparable due to other medication conditions.
There were several limitations to our study. First, the relatively small number of enrolled patients could diminish the reliability of the clinical application of the results. However, our quantitative evaluation results were satisfactory, showing high diagnostic accuracy, as well as sensitivity and specificity. The follow-up duration was also sufficient enough to detect most of the complications related to osteonecrosis. Second, postoperative follow-up with SPECT/CT was not conducted. Several studies reported the optimal timing of the test at two weeks after surgery, as revascularization, particularly after fracture reduction, is maximized at that time [31–33]. We tried to take SPECT/CT scans two weeks and six months postoperatively, but overcorrection related to implants in the attenuation correction process made the analysis impossible. If further studies identify a solution to the problem of attenuation correction, revascularization and changes in perfusion, with or without reduction, as well as their effects on the occurrence of avascular necrosis, could be revealed. Finally, the prediction of ONFH was solely based on the preoperative perfusion state. Various factors affecting osteonecrosis, including patient age, the degree of fracture displacement, and the quality of reduction, should be analyzed. Compared to a study by Turner et al., in which half of the patients were Garden type 3 (50%) fractures, type 1 fractures (56.7%) outnumbered other types of fractures in our study . Since the displacement is minimal, it is hard to predict the perfusion status both quantitatively and qualitatively.