Patients undergoing isolated primary ACL reconstruction by the senior author (PG) over a 12-month period were prospectively recruited to the study. Paediatric patients (aged less than 18 years) and patients undergoing revision surgery were excluded. Ethics approval was obtained from our local research ethics committee, and all patients provided informed consent prior to inclusion in the study.
A single bundle ACL reconstruction was performed with a four-strand short graft, using the Tape Lock Screw technique (TLS®, FH Orthopaedics, Heimsbrunn, France) [10]. This technique utilises the ipsilateral semitendinosus tendon, which is fashioned into a four-strand closed loop (50-60mm in length, depending on height and gender) and pre-tensioned to 500N for one minute. Femoral and tibial tunnels are created independently in a retrograde manner to match the diameter of each end of the graft. The tunnels are centred over the midpoint of the anatomic footprints of the native ACL. The graft is introduced using an all-inside technique and secured with polyethylene terephthalate tapes passed through each end of the closed tendon loop and suspended in the bone tunnels with interference screws.
All patients underwent a standardised LDCT six weeks post operatively. A Siemens Definition AS+ CT scanner (Siemens Somatom Definition AS+; Siemens Healthcare, Forchheim, Germany) was used. The data was then processed through the clinical software application iterative Metal Artifact Reduction (iMAR) algorithm for metal artifact reduction. Data was imported into Philips 2nd edition software (Philips Medical Systems North America, Shelton, Conn) to create a complete 3D volume reconstruction of each femur (to 10cm above the femoral condyle) and tibia (to 1cm below the tibial tubercle). The femoral 3D reconstruction had at least 90% overlap of femoral condyles and then virtual subtraction of the medial femoral condyle at the highest point of the intercondylar notch, leaving the most medial sagittal aspect of lateral femoral condyle with tunnel position on view (Figure 1). The tibial 3D reconstruction was an axial view, adjusted to view the most superior aspect of the proximal tibia with the femur and patella removed (Figure 2). The 3D reconstructions were imported into IntelliSpace Patient Archiving and Communication System (IntelliSpace PACS Enterprise; Philips) for measurement.
Radiation Dose and Associated Risk
The LDCT protocol used the following technique factors; 80kVp, 80mAs (effective), 128 x 0.6mm collimation, and helical scan with 1mm slice reconstructions. The scan coverage was approximately 10cm from above the femoral condyles to 1cm below the tibial plateau. The organ and effective dose applicable to research participants was estimated using three methods.
The first method was estimated using the International Commission on Radiological Protection 103 methodology (Table 2) [12]. This method used the assumption that the radiosensitive organs within the irradiation area (red bone marrow, bone surface and skin mass) constitutes less than 10% of the total mass of each of these organs in the body. This assumption was based on a conservatively estimated irradiation bone mass of less than 500g (i.e. 300cm3 x 1.65g/cm3) and a total body bone mass greater than 5kg [23]. The percentage of irradiated skin compared to total skin would be significantly less than 10%. The organ dose for radiosensitive organs was estimated using the central axis CT Dose Index (CTDI), where CTDI free-in-air would offer a maximum delivered dose to any point within the irradiated region. It was assumed that, given the relatively small cross-sectional area of the irradiated anatomy, organ doses would be approximately equal to, or slightly less than this value. The CTDI free-in-air value was taken from measurements of the Siemens Definition AS model outputs for technique factors matched as closely as possible to those proposed for this study [24]. This was calculated to be 7.8mGy/100mAs (Table 3).
The second method was estimated using the Impact CT dose program which bases its data on the Monte Carlo modelling of the National Radiological Protection Board (NRPB) [25]. This calculation used the deposition of energy at the proximal femur to estimate organ and effective doses for a scan distal to the knee. Based on this model and taking into account the estimates used in method one, the organ and effective doses (with the assumption an average sized patient is 70kg) for the LDCT scan, calculations can be seen in Figure 3 and are summarised in Table 4.
The third method used recently published data on Monte Carlo modelling of radiation transport from current generation CT scanners through voxel-based digital phantoms has provided a conversion factor (k-factor) to convert the Dose-Length Production (DLP) obtained from a given CT scanner to an effective dose value [26]. The k-factor derived for an 80kVp CT scan of an adult male or female was reported to be 0.0004. With a conservatively estimated DLP for this scan of 50mGy, the effective dose estimated using method would be <0.1mSv. Based on the above estimations and methods, the total effective dose for a single procedure is <0.5mSv.
Femoral Tunnel Measurement
The quadrant method described by Bernard was used to assess femoral tunnel position (Figure 1) [11, 12]. Point A is the centre of the femoral tunnel on the medial aspect of the lateral femoral condyle. A rectangular reference frame is superimposed. Distance B is the total sagittal diameter of the condyle, measured along the intercondylar notch roof, limited by the shallowest and deepest contours of the condyle. Distance C is the height of the intercondylar space, measured as the perpendicular distance between the notch roof and a parallel line tangential to the lowest point on the femoral condyle. Distance B and Distance C define the intercondylar space and create the axes for the quadrant system. Distance D is measured between Point A and the deep contour of the condyle, parallel to the notch roof. Distance E is the perpendicular distance between Point A and the notch roof. Femoral tunnel position in the sagittal plane is the defined by calculating ratios D/B and E/C.
Tibial Tunnel Measurement
The tibial tunnel position was assessed using a rectangular reference frame, as has been previously described (Figure 2) [12, 13]. The posterior border is drawn tangential to the posterior margins of the medial and lateral articular surfaces. The anterior border is drawn parallel to the posterior border, tangential to the anterior margin of the medial articular surface. The medial and lateral borders are drawn tangential to the most medial and lateral articular margins, respectively, perpendicular to the posterior border. Distance F is the mediolateral (ML) dimension of the tibial plateau. Distance G is the AP dimension of the tibial plateau. Point H is the centre of the tibial tunnel. Distance I and Distance J are measured perpendicularly from the medial and anterior borders of the reference frame, respectively. Tibial tunnel position is then defined in the axial plane by calculating ratios I/F and J/G.
Statistics
Femoral and tibial tunnel position was measured for each case independently by three observers (VV, SKe, SKa) at two time intervals, four weeks apart, in a blinded fashion. Images were de-identified and presented in random order, and the sequence was changed for the second measurement. Statistical analysis was performed using SPSS version 23.0 software (IBM Corp. IBM SPSS Statistics for Windows, Armonk, New York). Intra- and inter-rater reliability was calculated using a 2-way random absolute agreement model ICC and standard error of measurement (SEM). Single measures ICC was used to determine intra-rater reliability and average measures ICC was used to determine inter-rater reliability by comparing the means of 2 measurements of each variable.