Diagnostic Accuracy of Dual-energy CT for Bone Marrow Edema in Patients with Acute Knee Injuries: A Systematic Review and Meta-analysis


 Background: The prevalence of knee injury is high and early diagnosis is significant to guide clinical treatment. MRI is recognized as the gold standard for detecting bone marrow edema (BME) in patients with acute knee injury, but limitations still exist. Dual-energy CT (DECT) is investigated as a promising alternative.Methods: We systematically retrieved studies from EMBASE, Scopus, PUBMED, and the Cochrane Library and collected gray literatures. According to PRISMA-DTA guidelines, a systematic review was performed from inception to July 31, 2021, assessing the diagnostic accuracy of DECT for detecting BME in at least 10 adult patients with acute knee injuries and with an MRI reference standard. Study details were independently extracted by two reviewers. Meta-analysis was performed using a bivariate mixed-effects regression model with subgroup analysis performed to evaluate for sources of variability. Results: Nine studies evaluating 290 patients between the ages of 23–53 with acute knee injuries undergoing DECT and MRI were included in analysis. Summary sensitivity, specificity, and AUC values for BME were 85% (95% confidence interval (CI) 77–90%), 96% (95% CI 93–97%), and 0.97 (95% CI 0.95–0.98), respectively. There were no statistically significant differences in specificity and sensitivity amongst comparative subgroups to account for presumed variability amongst studies.Conclusion: DECT is accurate for detecting BME in patients with acute knee injuries and can be used as an alternative to MRI, particularly when MRI is contraindicated or unavailable.


Introduction
The prevalence of knee injury is high and early diagnosis is signi cant to guide clinical treatment. Early diagnosis and early treatment can prevent the further development of injury, especially for traumatic bone marrow injury without obvious fracture line [1]. Bone marrow edema (BME) is a universally acknowledged biomarker of subtle bone injury such as microfracture and hemorrhage concentrated in the trabecular bone and can be detected by MRI accurately [2]. MRI, therefore, is recognized as the gold standard for detecting occult fractures in patients with acute knee injury [1]. However, MRI requires a long examination time and is limited by certain medical devices such as pacemakers. Furthermore, the patient must remain still during the procedure, and this is occasionally impossible for elderly or trauma patients [2,3].
Dual-energy CT (DECT) is investigated as a promising alternative to MRI with advanced three-material decomposition technique which can remove the materials with relevant photoelectric effects to BME such as calcium and iodine [4][5][6]. The clinical value of DECT in assessing BME in adult patients with acute knee injuries is still under debate and exploration. The accuracy of DECT needs to be proven systematically prior to implementation as a substitute to MRI. Therefore, the aim of this systematic review and meta-analysis was to synthesize and evaluate published data on the accuracy of DECT for detecting BME in adult patients with acute knee injuries and to determine whether MRI is still indispensable for these patients.

Methods
This systematic review and meta-analysis was reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analysis -Diagnostic Test Accuracy (PRISMA-DTA) guidelines [7]. The approaches to evidence searching and data processing described in this article were based on the Cochrane Collaboration's diagnostic test accuracy method [8]. No ethical approval or informed consent was required for this article because all data were obtained from published literatures. Searching for studies, identifying eligibility, extracting data, and assessing quality were performed by two investigators independently. Any disagreement was resolved through discussion until they reach a consensus.

Search strategy
Four electronic databases, PubMed, EMBASE, Scopus and Cochrane Library, were searched for entries recorded from the time of database inception to July 31, 2021. Vocabulary and syntax were speci cally adapted according to the database. The initial search included variations of title/abstract/keywords and medical subject heading terms including "dual energy" AND ("computed tomography" OR "CT") AND "bone" AND "edema" and were modi ed as needed by individual database. No language restrictions were applied [9]. To nd any additional possible records, a gray literature search was performed by one author with evaluation of recent annual meeting for the American Roentgen Ray Society (ARRS), as well as recent years of conference proceedings from the Radiological Society of North America (RSNA) and the European Congress of Radiology (ECR). Conference abstracts that met the inclusion criteria and were not yet available in full text were included in analysis.
Literatures from each database and other sources were combined into a list from which duplicates were removed and were then screened initially with title and abstract for relevance. Subsequently, full-text review of the potentially relevant studies was conducted.

Inclusion criteria
The following criteria needed to be met for the studies included in this systematic review: (1) BME was evaluated around knee joints, (2) BME was the target nding and DECT was the index test, (3) MRI was used as the reference standard, (4) at least 10 patients were included, (5) su cient data could be extracted to construct a 2 × 2 contingency table. Non-original researches (including case reports, commentaries, consensus statements, guidelines, and narrative reviews) were excluded. If more than one study provided overlapping data, only the most comprehensive or latest study was included.

Data extraction
Requisite data extracted and recorded in the standardized Excel les (version 16.54, Microsoft). Study characteristics including surname of the rst author, year and country of publication, prospective versus retrospective study design, time between DECT and MRI, number of readers, presence of consensus reading, and reader experience were recorded. Patient characteristics including total number and inclusion interval of patients, mean age (Range), number of male and female, and number of knees (regions) with BME after acute knee injury were recorded. Additionally, speci c site of pathology, true positive (TP), false positive (FP), false negative (FN), true negative (TN), and threshold value was recorded for each assessment.

Risk of bias assessment
The methodological quality of the included studies was appraised according to the Quality Assessment of Diagnostic Accuracy Studies (QUADAS)-2 tool, which based on four separate domains: (1) patient selection, (2) index test(s), (3) reference standard, and (4) ow and timing [10]. Risk of bias (ROB) was assessed in each domain, and concerns about applicability were assessed in the rst three domains with signaling questions. These questions were answered with "yes" for a low risk of bias/concern, "no" for a high risk of bias/concern, or "unclear" when relevant information was not clearly provided. Studies considered high risk for any single signaling question within a domain were considered to have a high ROB for that respective domain.

Statistical analyses
Generally, a 2 × 2 contingency table was developed for each included study. If both qualitative and quantitative analysis were reported in a study, two separate contingency tables were created. Pooled sensitivity, speci city, positive likelihood ratio (PLR), negative likelihood ratio (NLR), and diagnostic odds ratio (DOR) were calculated using the bivariate meta-analysis framework (bivariate mixed effects regression model). In addition, summarized receiver operating characteristic (sROC) curves were constructed, with the area under the curve (AUC) depicting the accuracy of the tests.
To examine potential sources of variability, extensive subgroup analysis was performed on the studies. And univariate analysis was performed when bivariate regression model did not converge due to limited number of studies (< 4) and zero values in the 2 × 2 contingency table. A two-sided p value < 0.05 was considered statistically signi cant in all statistical tests. Simultaneously, statistical evaluation for publication bias was performed. Stata version 16 (StataCorp, College Station, TX) was used to analyze data from the included studies, and Review Manager Software version 5.4 (Cochrane Collaboration, Oxford, UK) was used to assess the methodological quality of the included studies.

Literature search
The PRISMA ow diagram for the literature search is depicted in Figure 1. A total of 237 articles were identi ed by searching databases and removing duplicates. After initial screening of titles and abstracts, 64 articles were further assessed by scrutinizing the full texts against the predesigned criteria, and 9 articles [11][12][13][14][15][16][17][18][19] were eventually included in analysis.

Study characteristics
The main characteristics of the included studies are summarized in Table 1. Nine studies involved a total of 290 patients with mean age ranging between 23 and 53 years and evaluated 2809 bone regions around the knee for BME. Four studies were performed in European institutions, four studies were performed in Asian institutions and one study was performed in a North American institution. Seven study was prospective, two studies were retrospective. Time interval between DECT and MRI was variable by study, but generally performed within 1 week. All studies involved at least two readers with ve studies undergoing consensus review.

Diagnostic performance
Integrated data for individual studies by site of pathology is summarized in Table 2. Region-based characterization is used for all studies to ensure the consistency of data. As shown in Figure 2, the pooled sensitivity and speci city for DECT detecting BME in patients with acute knee injuries are 85% (95% con dence interval (CI) 77-90%) and 96% (95% CI 93-97%), respectively. As shown in Figure 3  Subgroup analysis is demonstrated in Table 3. All studies were divided into two groups according to each of six signi cant factors. Studies with a retrospective study design were assessed with univariate analysis due to limited number of studies (< 4) in subgroup. Studies using a qualitative assessment method had a trend toward lower sensitivity compared with quantitative studies (80% (68-88%) versus 91% (85-95%), NS), but no difference in speci city. Studies from North American and European institutions had a trend toward lower speci city compared with Asian institutions (93% (88-95%) versus 98% (95-99%), NS), but no difference in sensitivity. No other signi cant differences were found between the comparison subgroups to account for presumed variability amongst studies.

Risk of bias assessment
Results for the QUADAS-2 tool for evaluating individual studies for risk of bias and applicability is included in Figure 6. Studies were predominantly low risk for both risk of bias and applicability. One study was rated high risk of bias for ow and timing domain for inappropriate time interval (more than 2 weeks) between DECT and MRI [11]. Two study assessed both quantitative and qualitative evaluation and only the quantitative evaluation was deemed high risk as the threshold is identi ed in advance [13,15]. One additional study was considered high risk of bias for index test for applying a retrospective threshold as well [18].

Publication bias
As shown in Figure 7, Deek's funnel plot asymmetry test indicated the existence of publication bias (p = 0.13).

Discussion
Knee injury is a common health problem at all ages and early diagnosis and early treatment can prevent the further development of injury. Bone marrow edema, as a universally acknowledged biomarker of subtle bone injury, can be detected by MRI accurately. However, a long examination time, a stationary state, incompatibility of certain medical devices such as pacemakers suggest that limitations still exist. At present, DECT is investigated as a promising alternative to MRI with advanced three-material decomposition technique which can remove the materials with relevant photoelectric effects to BME such as calcium and iodine. The clinical value of DECT in assessing BME in adult patients with acute knee injuries is still under debate and the accuracy of DECT needs to be proven. Therefore, the aim of this systematic review and meta-analysis was to synthesize and evaluate published data on the accuracy of DECT for detecting BME in patients with acute knee injuries and to determine whether MRI is still indispensable for these patients.
This meta-analysis demonstrates that DECT is highly speci c and accurate for detecting BME in patients with acute knee injuries when using MRI as the reference standard, with excellent pooled speci city (96% (95% CI 93-97%)) and AUC (0.97 (95% CI 0.95-0.98)). These results indicate that DECT can be used as alternative to MRI in patients with occult but suspected bone fracture after acute knee injury, particularly when MRI is contraindicated or unavailable. A slightly lower sensitivity of DECT for detecting BME when using an MRI reference standard (85% (95% CI 77-90%)) suggests that patients with negative results on DECT but persistent clinical manifestations may still require an MRI to detect more concealed bone injuries.
An extensive subgroup analysis to examine for potential sources of presumed variability and increase the con dence of the generalizability of these results was a highlight. There were no statistically signi cant differences in speci city and sensitivity between these comparative subgroups. Notably, studies using a qualitative assessment method had a trend toward lower sensitivity compared with quantitative studies and studies from North American and European institutions had a trend toward lower speci city compared with Asian institutions. This study provides several opportunities for future research to further evaluate particular subgroups, including DECT protocols not yet involved and determine the optimized applicability and interpretation of DECT dependent on patient and imaging speci c details in patients with acute knee injuries. Further, although most studies were low risk for bias and applicability, a few studies used retrospective thresholds and one study used prolonged intervals between DECT and MRI which increase the risk of bias in both index test and ow and timing domains, respectively.
In addition, some intrinsic limitations of DECT maintain MRI as the preferred imaging modality at present, including (1) the non-speci city of BME for fracture (e.g., bone infarction can result in BME); (2) accompanied bone injuries when soft tissues such as meniscal and ligament are frequently injured; (3) increased radiation dose; (4) risk for non-diagnostic virtual non-contrast reconstructions; (5) and limited eld of view in virtual non-contrast reconstructions [13,14,18]. Be that as it may, this systematic review and meta-analysis reveals that DECT can be applied as an alternative to MRI for patients with acute knee injuries under appropriate circumstances. Patients with a negative DECT but high suspicion for bone injuries may still require an MRI to evaluate for occult fractures.

Conclusions
Based on the results of this meta-analysis, it could be concluded that DECT is accurate for detecting BME in patients with acute knee injuries with high with high sensitivity and speci city, indicating it can be used as an alternative to MRI, particularly when MRI is contraindicated or unavailable. Availability of data and materials The datasets used and/or analyzed during the present study are available from the corresponding author on reasonable request. No., number; DECT, dual-energy computed tomography; MRI, Magnetic resonance imaging; NR, not reported. Table 2 Integrated data for individual studies by site of pathology Figure 2 Forest plots of the sensitivity (left) and speci city (right) of dual-energy CT for detecting bone marrow edema in patients with acute knee injuries.

Figure 3
Forest plots of the PLR and NLR of dual-energy CT for detecting bone marrow edema in patients with acute knee injuries.

Figure 4
Forest plots of the diagnostic score and DOR of dual-energy CT for detecting bone marrow edema in patients with acute knee injuries.

Figure 5
Summarized receiver operating characteristic curve (sROC) of dual-energy CT for detecting bone marrow edema in patients with acute knee injuries.

Figure 6
Results of QUADAS-2 tool evaluation of individual studies for risk of bias and applicability. Red in gure indicates high risk, yellow represents unclear risk and green means low risk.