External Adhesive Marker Technique Shows Excellent Reproducibility for Measuring Patellar Tendon‐to‐Lateral Trochlear Ridge Distance on Magnetic Resonance Imaging

Abstract

Purpose

To evaluate the agreement between a simple magnetic resonance imaging‐based measurement method for determining the patellar tendon‐to‐lateral trochlear ridge (PT‐LTR) distance using an external adhesive marker and a previously validated conventional measurement technique.

Methods

Patients diagnosed with patellofemoral instability before undergoing primary patellar stabilization surgery at a single institution between 2020 and 2024 were eligible for inclusion. PT‐LTR distances were measured in a blinded fashion by 2 musculoskeletal‐trained radiologists using the conventional technique and by an orthopaedic group using the simple technique. Interrater reliability between the musculoskeletal‐trained radiologists and between the orthopaedic groups was calculated using intraclass correlation coefficients (ICCs) and concordance correlation coefficients (CCCs) and subsequently intermethod reliability between the 2 groups. The measurements were plotted on concordance correlation graphs.

Results

One hundred knees (38 right, 62 left) from 97 patients (37 men, 60 women; mean age 22 ± 9 years) were included. The average distance for PT‐LTR measured by the radiologists using the conventional technique was 10.9 ± 7.5 mm and by the orthopaedic group using the simple technique was 9.3 ± 6.8 mm. Excellent interobserver agreement was noted for the radiologist group (ICC 0.941, 95% confidence interval [CI], 0.914‐0.960; CCC 0.941, 95% CI, 0.915‐0.959) and orthopaedic group (ICC 0.976, 95% CI, 0.964‐0.985; CCC 0.976, 95% CI, 0.965‐0.984). Excellent intermethod agreement was noted between the 2 techniques (ICC 0.903, 95% CI, 0.775‐0.942; CCC 0.902, 95% CI, 0.861‐0.932).

Conclusions

There was excellent interobserver reliability for measuring PT‐LTR distance using both the conventional radiologist‐based technique and the external adhesive marker technique. Excellent intermethod agreement was also shown between the 2 techniques, with comparable mean measurements between groups; thus, the simple technique can be used as a reproducible alternative for PT‐LTR distance.

Level of Evidence

Level III, retrospective comparative study.

Patellofemoral instability (PFI) is a disorder of the knee extensor mechanism that can occur with a wide range of severity. Risk factors for recurrent PFI include anatomic abnormalities such as increased tibial tubercle‐trochlear groove (TT‐TG) distance, increased patellar tendon‐to‐lateral trochlear ridge (PT‐LTR) distance on magnetic resonance imaging (MRI), trochlear dysplasia, patella alta, younger age, and skeletal immaturity. ¹ , ²

PT‐LTR distance serves as a reliable metric to assess a lateralized force vector of the patellar tendon (PT) as a risk factor for recurrent lateral PFI. ³ The sensitivity and specificity for PT‐LTR distance measured on MRI to detect patellar instability in a pediatric cohort were shown to be 73% sensitive and 89% specific at a cutoff of 5.6 mm. ⁴ Traditionally, the measurement of PT‐LTR distance requires the use of advanced radiographic software, where axial MRI images are superimposed onto one another with transposed lines and measurements. ⁵ , ⁶ By this method, PT‐LTR distance has been shown to be reliable, specific, and reproducible. However, access to the necessary advanced software on MRI (or computed tomography [CT]) may not be available universally, especially to the orthopaedist in clinic or office.

A simple technique has been previously validated as reliable and accurate in measuring TT‐TG distance in patients with PFI without using advanced radiographic imaging software. ⁷ The purpose of this study was to evaluate the agreement between a simple MRI‐based measurement method for determining the PT‐LTR distance using an external adhesive marker and a previously validated conventional measurement technique. We hypothesized that a strong positive correlation between measurements obtained using the simple and conventional methods.

METHODS

Patient Selection

Consecutive patients with a primary diagnosis of PFI who were managed operatively at a single institution between 2020 and 2024 were retrospectively reviewed. A primary diagnosis of PFI was determined utilizing an internal institutional research informatics tool to identify patients who presented for patellar instability or dislocation and received subsequent operative management. Institutional Review Board approval was obtained before initiation of this study (IRB: 15‐000601). Patients were included if they had available preoperative MRI of their affected knee before undergoing primary patellofemoral stabilization surgery. Patients were excluded if they did not have available MRI or did not undergo primary patellofemoral stabilization surgery. Additional patients were excluded if the MRI was not sufficient to allow for measurement of the PT‐LTR distance; for example, the study included incomplete axial coverage, excessive motion artifact that obscured visualization of the PT, the posterior condylar axis, or the lateral trochlear ridge (LTR). There were no restrictions on age or sex.

Magnetic Resonance Imaging

MRI studies of the knee were obtained in the standard, supine position, with the knee fully extended within a dedicated knee coil. Axial T2‐weighted, fat‐suppressed images were acquired on either a 1.5‐T or 3.0‐T GE MRI scanner (GE Healthcare, Waukesha, WI), with the following parameters at 1.5‐T: fast spin echo, 3000 to 5000 ms repetition time, 50 ms echo time, 3 to 4 mm slice thickness with 0.5 mm skip between slices, field of view of 16, and an echo train length of 8 to 10; and at 3‐T: fast spin echo, 4000 to 6000 ms repetition time, 45 ms echo time, 3 mm slice thickness with 0 mm skip between slices, field of view of 16, and an echo train length of 10 to 12.

Conventional Technique of PT‐LTR Distance Measurement

PT‐LTR distances were measured for all patients on MRI T2‐weighted axial images by 2 experienced musculoskeletal radiologists (N.G.R., M.D.R.). All measurements were made separately, in a blinded fashion. The 2 radiologists made their measurements on a Visage viewer platform (Richmond, VIC, Australia). These measurements were made by drawing a line perpendicular to the posterior condylar reference line of the femur, with propagation of that line on all images in the axial image stack. Next, the most anterior aspect of the LTR was identified, and that point was marked. A line (the LTR line) was drawn through this point perpendicular to the tangential line of the posterior condyles (Figure 1a). This line was propagated on all image slices. Next, the most proximal image showing the PT without the distal patellar bone was selected. On this slice, the most lateral aspect of the PT was marked, and an axial line in plane with the PT was drawn from this point to the point on the LTR line intersecting with the PT (Figure 1b). This measurement was termed the PT‐LTR distance in accordance with prior studies. ⁴ If the lateral aspect of the PT insertion was located medial to the LTR line, it was reported as 0 mm.

FIGURE 1.

Axial magnetic resonance images of the (a,b) conventional technique and the (c,d) simple technique of PT‐LTR distance of a left knee. Conventional technique: (a) a line is drawn along the posterior condylar axis and deemed the PCRL. A point is measured at the level of (a) the most anterior aspect of the LTR (marked by white arrow). A line, the LTR line (marked by yellow arrow), is drawn through this LTR point perpendicular to the tangential line of the previously drawn PCRL line. On the most proximal image showing the PT without the distal patellar bone was selected (b), the most lateral aspect of the PT was marked (arrow), and a line in plane with the PT was drawn from this point to the LTR line (red line), representing the PT‐LTR distance. Simple technique: (c) a line is drawn along the posterior condylar axis and deemed PCRL line. A point is measured at the most anterior aspect of the lateral trochlea ridge (marked by white arrow). The LTR line (marked by yellow arrow) is established similar to the conventional method, and the lateral border of a reusable adhesive external alignment aid was placed perpendicular to the PCRL. This was left in place while the most proximal image showing the PT without the distal patellar bone was selected. On this slice (d), the most lateral aspect of the PT was marked, and a line in plane with the PT was drawn from this point to the LTR line marked by the adhesive aid. (LTR, lateral trochlear ridge; PCRL, posterior condylar reference line; PT, patellar tendon; PT‐LTR, patellar tendon‐to‐lateral trochlear ridge).

Simple Technique of PT‐LTR Distance Measurement

PT‐LTR distances were measured for all patients on MRI axial images by a trained orthopaedic research assistant and a sports medicine fellowship‐trained orthopaedic surgeon (R.R.A., A.J.T.). All measurements were made separately, in a blinded fashion. The orthopaedic team made their measurements on InfinityView (Version r1 build 2025.09.25.2), which does not have the ability to propagate measurements on all images in an axial image stack. For the simple technique, a line was drawn perpendicular to the posterior condylar reference line (Figure 2a). Next, the most anterior aspect of the LTR was identified (Figure 2b), and the lateral border of a reusable, adhesive alignment aid was then placed on the computer monitor (the LTR line) perpendicular to the tangential line of the posterior condyles (Figures 1c and 2c). This was left in place on all image slices. Next, the most proximal image showing the PT without the distal patellar bone was selected. On this slice, the most lateral aspect of the PT was marked, and an axial line in plane with the PT was drawn from this point to the point on the LTR line marked by the adhesive aid (Figures 1d and 2d). This measurement was termed the PT‐LTR distance. If the lateral aspect of the PT insertion was located medial to the LTR line, it was recorded as 0 mm.

FIGURE 2.

Visual representation of steps required to measure PT‐LTR using a schematic illustration of a left knee MRI in the axial plane. First, the PCRL line is established (marked by yellow line) (a). Next, a line perpendicular to the PCRL (marked by red dotted line) is propagated through the most anterior aspect of the LTR (b). The lateral edge of a reusable adhesive aid is next placed along this perpendicular line to mark this point (c), so that when the images are advanced distally below the level of the patella, (d) the distance from the LTR to the lateral aspect of the PT can be measured (PT‐LTR). (LTR, lateral trochlear ridge; MRI, magnetic resonance imaging; PCRL, posterior condylar reference line; PT, patellar tendon; PT‐LTR, patellar tendon‐to‐lateral trochlear ridge.)

Statistical Analyses

Interrater reliability in the radiologist group and in the orthopaedic group was calculated using intraclass correlation coefficients (ICCs) and concordance correlation coefficients (CCCs) with 95% confidence intervals (CIs). Intermethod reliability between the 2 groups were calculated using ICC and CCC. ICC and CCC values greater than 0.75 were considered excellent agreement, 0.75 to 0.40 fair agreement, and less than 0.40 poor agreement. ⁸ The measurements made by each of the 4 readers were plotted on concordance correlation graphs in relation to their calculated CCC values. Intrarater reliability of the simple technique was assessed after a washout period. Finally, a post hoc assessment of precision in ICC estimates was performed to determine 95% CIs and half‐widths. Statistical analyses were performed using RStudio, version 4.4.1 (RStudio, PBC, Boston, MA).

RESULTS

One hundred knees (38 right, 62 left) in 97 patients (37 men, 60 women) with a mean age of 18 ± 5 were included. The average PT‐LTR distances measured by the 2 musculoskeletal‐trained radiologists using the conventional technique of the same cohort were 10.8 ± 7.2 mm and 11.0 ± 7.9 mm. The average of their combined measurements was 10.9 ± 7.5 mm. The average PT‐LTR distances measured by each member of the orthopaedic group using the simple technique of the same cohort were 9.5 ± 6.8 mm and 9.1 ± 6.8 mm. The average of their combined measurements was 9.3 ± 6.8 mm.

Analysis of the interrater reliability for the conventional technique performed by the radiologist group revealed an ICC of 0.941 (95% CI, 0.914‐0.960) and CCC of 0.941 (95% CI, 0.915‐0.959), indicating excellent interobserver agreement and reliability (Table 1; Figure 3a). Analysis of the interrater reliability for the simple technique performed by the orthopaedic group revealed an ICC of 0.976 (95% CI, 0.964‐0.985) and CCC of 0.976 (95% CI, 0.965‐0.984), indicating excellent interobserver agreement and reliability (Table 1; Figure 3b). After a washout period, intrarater reliability for the simple technique additionally showed an ICC of 0.947 (95% CI, 0.922‐0.964) and CCC of 0.946 (95% CI, 0.924‐0.962).

TABLE 1.

Intraclass Reliability for Overall Measurement of PT‐LTR for the Simple Technique, Conventional Technique, and Simple Versus Conventional Techniques

Intraclass Reliability	ICC (95% CI)	CCC (95% CI)
Intrarater reliability
Simple technique	0.947 (0.922‐0.964)	0.946 (0.924‐0.962)
Interrater reliability*
Simple technique	0.976 (0.964‐0.985)	0.976 (0.965‐0.984)
Conventional technique	0.941 (0.914‐0.960)	0.941 (0.915‐0.959)
Intermethod reliability†
Simple versus conventional techniques	0.903 (0.775‐0.950)	0.902 (0.861‐0.937)

CCC, concordance correlation coefficient; CI, confidence interval; ICC, intraclass correlation coefficient; PT‐LTR, patellar tendon‐to‐lateral trochlear ridge.

^*Agreement between rater 1 and 2 for the simple (orthopaedic) and conventional (radiologists) techniques.

^†Agreement between simple and conventional techniques.

FIGURE 3.

Concordance correlation graphs depicting (a) the agreement between radiologists 1 and 2 using the conventional PT‐LTR distance measurement technique and (b) the agreement between members 1 and 2 of the orthopaedic group using the simple PT‐LTR distance measurement technique. (PT‐LTR, patellar tendon‐to‐lateral trochlear ridge.)

When comparing the mean PT‐LTR distances for the simple technique method to the conventional technique method, the mean discrepancy in the PT‐LTR distance between the 2 techniques was 1.6 mm. Analysis of the intermethod reliability of the 2 techniques revealed an ICC of 0.903 (95% CI, 0.775‐0.950) and CCC of 0.902 (95% CI, 0.861‐0.932), indicating excellent interobserver agreement and reliability (Table 1; Figure 4).

FIGURE 4.

Concordance correlation graph depicting the agreement between the conventional and simple PT‐LTR measurement techniques. (PT‐LTR, patellar tendon‐to‐lateral trochlear ridge.)

A subanalysis of patients with trochlear dysplasia was additionally performed to assess the reproducibility of the simple technique in measuring PT‐LTR within this population. The ICC and CCC were calculated for PT‐LTR measurements obtained using both techniques among these patients. Intermethod reliability between the simple and conventional techniques showed strong agreement, with an ICC of 0.867 (95% CI, 0.677‐0.936) and a CCC of 0.865 (95% CI, 0.787‐0.916) (Table 2).

TABLE 2.

Intraclass Reliability for Measurement of PT‐LTR in Patients With Trochlear Dysplasia for the Simple Technique, Conventional Technique, and Simple Versus Conventional Techniques

Intraclass Reliability	ICC (95% CI)	CCC (95% CI)
Interrater reliability*
Simple technique	0.967 (0.942‐0.981)	0.966 (0.943‐0.980)
Conventional technique	0.931 (0.885‐0.959)	0.929 (0.885‐0.957)
Intermethod reliability†
Simple versus conventional techniques	0.867 (0.677‐0.936)	0.865 (0.787‐0.916)

CCC, concordance correlation coefficient; CI, confidence interval; ICC, intraclass correlation coefficient; PT‐LTR, patellar tendon‐to‐lateral trochlear ridge.

^*Agreement between rater 1 and 2 for the simple (orthopaedic) and conventional (radiologists) techniques.

^†Agreement between simple and conventional techniques.

Further analysis was performed to identify potential sources of variation between the conventional and simple techniques. The mean PT‐LTR measured for each patient in the orthopaedic group was compared with the individual measures recorded by each radiologist. If a difference of greater than the smallest standard deviation (6.8 mm) was identified, the MRI was reviewed for potential anatomic variations that may have been responsible for a wide PT‐LTR disagreement. Three specific patients were identified, with differences between the mean PT‐LTR measured with the simple technique and PT‐LTR measured with the conventional technique of 12.9, 8.6, and 7.1 mm. Potential sources of variation identified in these patients included trochlear dysplasia, leading to a flattened LTR obscuring the anterior‐most point for reliably measuring PT‐LTR.

A post hoc assessment of precision was performed to determine if the sample size of 100 included was sufficient to determine intermethod reliability. The calculated intermethod ICC of 0.903 with 100 measures and 2 measurement methods showed a 95% CI for ICC of 0.865 to 0.941 and a half‐width of ±0.038. This indicates the study sample size was sufficient to determine ICC with high precision. An increase of the sample size to 150 only narrowed the 95% CI half‐width to ±0.031, suggesting only minimal gains in precision with an increase in sample size.

DISCUSSION

This study showed that PT‐LTR distance can be measured reliably using a simple technique employing an external alignment aid and that this technique has excellent intermethod reliability when compared with a conventional technique requiring advanced imaging software. Among 100 knees with PFI, the mean PT‐LTR distance with the conventional radiologist method was 10.9 ± 7.5 mm, whereas the simple technique in the orthopaedic group measured a mean of 9.3 ± 6.8 mm. Both techniques had excellent interrater reliability, with ICC and CCC both greater than 0.94 for the radiologist group and 0.97 for the orthopaedic group. Reproducibility remained strong in a subgroup analysis of patients with trochlear dysplasia, with a CCC of 0.865. These findings show that the simple technique described can be used as a reproducible alternative for measuring PT‐LTR distance.

TT‐TG distance has been used most commonly when assessing lateralization of the tibial tubercle to determine a potential need for TTO and predicting a risk of recurrent PFI. ⁷ , ⁹ , ¹⁰ Though the simple technique outlined in our study has been validated for measuring TT‐TG distance, PT‐LTR distance may provide additional important information regarding anatomic risk factors for instability. ¹¹ PT‐LTR distance may better delineate the role of both the bony and soft‐tissue elements contributing to patellar instability, such as the trochlear containment and an elevated lateral vector of pull across the extensor mechanism relative to the trochlea that may not be factored into measurements like TT‐TG distance. ⁴ , ¹² PT‐LTR distance, when greater than 5.6 mm, has been shown to have similar sensitivity yet higher specificity for a history of patellar dislocation compared with TT‐TG. ⁴ Therefore, PT‐LTR distance, especially when measured with this simple technique, can offer a reproducible and practical complement to TT‐TG distance and shows greater specificity for instability events.

The simple technique outlined in our study was based on the methodology described by Camp et al., ⁷ where the technique for measuring TT‐TG distance on MRI with this method was validated. The simple technique, using the external alignment guide, has since been used for measuring TT‐TG distance on MRI in 2 studies with cohorts of patients diagnosed with patellar dislocation of 59 and 87 knees. ¹³ , ¹⁴ Although CT scans can be used to measure TT‐TG distance, MRI is more commonly utilized because of no radiation exposure to patients while also allowing simultaneous assessment of soft tissues and articular cartilage. ⁷ , ¹⁵ , ¹⁶ Camp et al. ¹⁷ also revealed that the values of measured TT‐TG distance are not interchangeable between CT and MRI, with TT‐TG values significantly higher on CT compared with the same knee measured on MRI. This lack of equivalence has yet to be shown with PT‐LTR measures and is a potential area of future study. Additionally, although external alignment aids have previously been validated for TT‐TG, validation of 1 stacking‐based technique does not necessarily extend to other measurements such as PT‐LTR. Differences in distinct anatomic landmarks and line constructions introduce measurement variability, including lateral‐trochlear ridge morphology in trochlear dysplasia and PT identification. Because of possibilities of method‐specific errors, independent validation is required.

PT‐LTR distance is a relatively novel measure that may be used as a surrogate of the complex interplay of biomechanical components in PFI. Reliable measurement of PT‐LTR distance allows surgeons to better quantify lateralization of the PT and translate this information into surgical decision‐making. For instance, a distance of 13 mm or greater has also been shown to identify knees at risk of a persistent postoperative J‐sign after medial patellofemoral ligament reconstruction with TTO, adding to the value of PT‐LTR distance in preoperative risk assessment and surgical planning. ¹⁸ The simplicity and reproducibility of the simple method described are clinically relevant in instances where imaging software does not allow for superimposed measures and marks to superimpose on axial MRI sequences in review of imaging. The use of an external alignment aid shown in the present study is easy to implement and reproducible. Consequently, having a fast, efficient, and reproducible method to perform PT‐LTR measurements can help further expedite patient care.

Limitations

There were limitations associated with the present study. First, the retrospective design introduces potential for selection bias. Second, this study was conducted at a single tertiary referral institution, which may limit the generalizability of the findings to other centers with different imaging protocols, software, or patient populations. Third, although the simple technique is straightforward, a potential learning curve exists for accurate placement and stabilization of the external alignment aid. Fourth, because no established external reference standard exists for PT‐LTR measurement, this study did not evaluate diagnostic accuracy or predictive validity. Finally, the sample size did not allow for subgroup analyses by sex, laterality, or additional anatomic abnormalities like trochlear dysplasia, which may influence PT‐LTR distance.

CONCLUSIONS

There was excellent interobserver reliability for measuring PT‐LTR distance using both the conventional radiologist‐based technique and the external adhesive marker technique. Excellent intermethod agreement was also shown between the 2 techniques, with comparable mean measurements between groups; thus, the simple technique can be used as a reliable and reproducible alternative for measuring PT‐LTR distance.

DISCLOSURES

The authors (A.J.T., C.L.C., M.H., A.J.K., D.B.F.S.) declare the following financial interests/personal relationships which may be considered as potential competing interests: A.J.T. reports a relationship with the American Orthopaedic Society for Sports Medicine: Early Career Member Committee that includes: board membership; and a relationship with Bodycad that includes: consulting. C.L.C. reports other professional activities with Arthrex, Incorporation; and other intellectual property. M.H. reports relationships with DJO, Limited Liability Company, Moximed, and Stryker that include: consulting, the Journal of Cartilage and Joint Preservation that includes: board membership, and Elsevier that includes: intellectual property royalties. A.J.K. reports relationships with American Journal of Sports Medicine, Arthroscopy Association of North America, International Cartilage Regeneration & Joint Preservation Society, and Knee Surgery, Sports Traumatology, Arthroscopy that include: board membership; Arthrex, Incorporation that includes: consulting; and other intellectual property. D.B.F.S. reports relationships with Smith & Nephew, Incorporation and Relive Biotechnologies that include: consulting; and Orthoregeneration Network Foundation, International Cartilage Regeneration & Joint Preservation Soceity, American Academy of Orthopaedic Surgeons Committee on Devices, Biologics, and Technology, and Cartilage that include: board membership. The other authors (R.R.A., G.H.W., J.D.M., N.G.R., M.D.R.) declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.