Study Design
This study was a pilot and feasibility randomized clinical trial with participants randomized into two arms with parallel assignment and balanced allocation. The study was approved by the University of Delaware Institutional Review Board, prospectively registered (ClinicalTrials.gov ID: NCT03694730), and conducted according to CONSORT guidelines.(17,18) To be included, participants had to have a clinical diagnosis of patellar tendinopathy and be between the ages of 16 and 40 years old. The diagnostic criteria for patellar tendinopathy was 1) pain and stiffness isolated to the patellar tendon and 2) load-dependent symptoms, which increased as the demands on the patellar tendon increased.(2,19) Potential participants were excluded if they had an injury, other than patellar tendinopathy, that limited their ability to participate in treatment and/or testing. Additionally, individuals were excluded if they had knee surgery or an injection, tenotomy, or shockwave to the patellar tendon within the last 6 months. All participants received written and oral information about the study prior to participation and provided informed consent.
Participants were randomly assigned into one of two treatment groups: 1) a Pain-Guided Activity (PGA) group and 2) a Pain-Free Activity (PFA) group. All participants completed patellar tendon loading exercises three times a week for 12 weeks using a standardized treatment program.(11,12) During the first six weeks of treatment, participants were asked to limit their physical activity outside of treatment, in accordance with their treatment strategy. Participants in the PGA group were allowed to continue their usual recreational activities, using the pain-monitoring model to guide activity intensity (Figure 1).(15) The PFA group were not allowed to perform running, jumping, or activities that provoke their patellar tendon pain outside of treatment. However, they were allowed to perform activities that were not pain-provoking, such as swimming or the elliptical. Additionally, participants completed evaluations at baseline, 6- and 12-weeks to assess symptom severity, psychological factors, tendon morphology and mechanical properties, lower extremity function, and quadriceps muscle performance. All supervised treatments and evaluation sessions were completed at the University of Delaware Health Sciences Complex in Newark, DE. Enrollment, randomization, evaluations and treatment were all completed by the same physical therapist precluding blinding.
Recruitment and Randomization
Participants were recruited from the University of Delaware community and the surrounding region between January 2019 and February 2020. A recruitment strategy was developed that incorporated multiple media platforms and recruitment locations to ensure a representative sample of individuals with patellar tendinopathy (Supplementary material). Interested individuals were directed to an online screening questionnaire for preliminary eligibility assessment. Following completion of this questionnaire, interested individuals were contacted by a member of the research team to provide additional information about the study, clarify responses on the eligibility questionnaire, and schedule their in-person screening and informed consent. Access to potential participants, results of eligibility screening, and willingness to be randomized were recorded using Research Electronic Data Capture (REDCap) (Vanderbilt University, Nashville, TN).(20) Additionally, recruitment sources and reasons for ineligibility or declined participation were recorded.
The randomization scheme was generated by a biostatistician and stratified for sex without blocking. The randomization scheme was stored as a spreadsheet on a password protected computer. Participants were randomized after being deemed eligible and committing to partake. The participants were notified of their group assignment at the first treatment session by the treating physical therapist.
Treatment Protocol
All participants completed a modified version of the Heavy-Slow Resistance protocol three times a week for 12-weeks.(11,12) The original protocol consists of three exercises, the squat, leg press, and hack squat. In the modified version, the hack squat was replaced with the knee extension due to equipment availability. Participants complete four sets of each exercise per session using a 6-second count per repetition (3 second eccentric, 3 second concentric phase). An auditory metronome was used to pace each repetition. Over the course of the protocol, the load is progressively increased and repetitions are decreased (Figure 2). Resistance levels for each phase of the treatment protocol are dosed based on the participants 5-repetition max (5RM) for each exercise, which were performed at the initial treatment and approximately every two weeks after. All supervised treatments were completed by the same physical therapist and participants were required to attend at least one supervised session every two weeks to complete 5RM testing. Since patellar tendinopathy typically occurs in athletic individuals who are familiar with strength training, participants were given the option to complete other treatments at their personal gym or team facility. At 6- and 12-weeks, participants were asked to rate their level of satisfaction on a 10-point Likert scale (0 = Not satisfied, 10 = Very Satisfied).
Compliance, Retention and Safety
Compliance with exercises and activity modifications were tracked using paper training diaries. For each day, participants were asked to record the treatment exercises completed, whether they performed running and/or jumping activity, any other physical activity performed, and their pain levels upon waking. Furthermore, if participants performed physical activity other than treatment exercises, they were asked to rate their pain levels prior to, during, and after the activity. The training diaries were collected weekly and reviewed by the research team. For each week, the number of times treatment exercises were completed, days the training diary was completed, days of running or jumping activity, and days in compliance with activity restrictions (weeks 1-6) was recorded. Participants were considered compliant with treatment if they performed at least two of the three prescribed exercise sessions per week. Additionally, the number of missed follow-up evaluations, drop-outs, and adverse events were recorded.
Clinical Outcomes
Symptom Severity
Symptom severity was assessed using the Victorian Institute of Sports Assessment – Patellar Tendon (VISA-P) questionnaire. The VISA-P is an 8-item questionnaire designed to assess patellar tendinopathy symptom severity and the impact on physical function.(21) Scores range from 0 to 100 with lower scores indicating greater disability. This instrument has a minimally clinically important difference (MCID) of 13 points.(22)
Psychological Factors
Participants fear of movement and re-injury, or kinesiophobia, were captured using the Tampa Scale of Kinesiophobia (TSK-17).(23,24) Higher levels of kinesiophobia have been associated with worse recovery of lower extremity function in Achilles tendinopathy.(25) Additionally, in pilot studies, we found that the majority of patients with patellar tendinopathy have clinically meaningful levels of kinesiophobia.(26) Scores range from 17 to 68 points, with higher scores indicating a greater fear of movement and re-injury.
The presence and severity of negative emotional states was measured using the Depression, Anxiety, and Stress Scale (DASS-21).(27) The DASS-21 has previously been used to evaluate the mental health of athletes at a variety of competition levels.(28,29) The DASS-21 consists of 21 questions that can be divided into three subscales, 1) Depression, 2) Anxiety, and 3) Stress. Scores range from 0 to 63 points, with higher scores indicating a greater degree of negative emotional states.
Tendon Morphology
B-mode ultrasound imaging was performed at the patellar tendon using a LOGIC e Ultrasound (GE Healthcare, Chicago, IL) system with a wide-band linear array probe (5.0 – 13.0 MHz) at 10 MHz and 2.5 cm depth to assess tendon morphology. Participants were positioned in supine with the knee flexed to 30° and supported by a bolster.(30) Three extended field of view long-axis images were completed on each limb from the tibial tuberosity to the inferior pole of the patellar to obtain maximal tendon thickness. Additionally, three short-axis images were taken at 1 cm distal to the inferior pole of the patella on each limb to obtain cross-sectional area (CSA). A custom MatLab code was used to identify the maximal tendon thickness and Osirix MD imaging software (Pixmeo, Geneva, Switzerland) was used to measure CSA. The average of three images was used for analysis.
Tendon Mechanical Properties
Continuous shear wave elastography (cSWE) was used to evaluate patellar tendon mechanical properties with a Sonix MDP Q+ (Ultrasonix, Vancouver, BC, Canada) ultrasound scanner with a L14-5/38 probe.(31,32) For this technique, participants were seated on an adjustable plinth with their legs stabilized at 90° of hip and knee flexion. The inferior pole of the patella and the tibial tuberosity were identified and a mark was placed 1 cm distal to the inferior pole of the patella, along the imaginary line connecting the two bony landmarks. The ultrasound probe was centered over this mark, in line with the long axis of the tendon. A Minishaker Type 4810 (Bruel and Kjaer, Norcross, GA, USA) was placed on the quadriceps tendon and used to produce shear waves at 11 different frequencies (322, 339, 358, 379, 402, 429, 460, 495, 536, 585, and 643 Hz). As each frequency propagated through the patellar tendon, the ultrasound probe captured raw radiofrequency data at 6438 frames/sec. A custom MATLAB code was used for post-processing to provide estimates of static shear modulus and viscosity, as described by Cortes et al and Corrigan et al.(31,32) Three trials were performed per limb and the average of three trials was used for analysis.
Lower Extremity Function
The single-leg counter-movement jump (CMJ) and single-leg drop CMJ were used to evaluate lower extremity function.(33) These tests have high reliability and have previously been used to assess function in lower extremity tendinopathies.(15,33,34) For the CMJ, participants began by standing on a single leg on flat ground with their hands behind their back. They were instructed to jump as high as they can, landing on the same leg with which they took off from the ground.(33) The drop CMJ was performed similarly except that participants assumed the starting position on a 20 cm high box. They were instructed to “drop” off of the box and then jump as high as they can once they contacted the ground.(33) For both tests, an infrared light mat (MuscleLab®, Ergotest Innovations, Stathelle, Norway) was used to record flight time, which was then used to estimate jump height. The average of three trials was used for analysis.
Quadriceps Muscle Performance
A knee extension maximal voluntary isometric contraction (MVIC) with the burst-superimposition method was used to evaluate knee extension strength and quadriceps muscle activation.(35) This technique has demonstrated reliability and has been utilized in a variety of chronic knee injuries.(35–42) Participants were seated on a KinCom dynamometer (Model 50 H, Isokinetic International, Chattanooga, TN, USA) at 90° of hip flexion and 60° of knee flexion for the tested limb. Self-adhesive electrodes were placed over the distal vastus medialis and proximal vastus lateralis muscle bellies. After familiarization with procedures and a standardized warm-up, participants were instructed to perform a 5-second MVIC. During the MVIC, a supramaximal, 10-pulse (600 μs, 130 V, 100 pulses per second) train of electrical stimulation was applied to the muscle using an electrical stimulator (Grass Technologies, Champaign, IL). Verbal encouragement was provided throughout each trial. If the participant was unable to activate quadriceps fully or they did not reach and maintain their peak MVIC prior to delivery of the burst, testing was repeated up to 4 times, with 3 minutes rest between trials. The MVIC force and force attributable to the electrical stimulation was recorded. The best trial, based on force production and visual inspection of the force production graph, was selected to calculate quadriceps central activation ratios (CAR = [MVIC force/burst augmented force] x 100%). The CAR is a measure of quadriceps inhibition, where lower values indicate a greater degree of quadriceps inhibition.
Alterations to Study Protocol After Initiation
Due to the COVID-19 pandemic, in-person human subjects research was halted on March 17th, 2020 by the Institutional Research office at the University of Delaware. At that time, there were seven active participants in the study. One had completed treatment and was scheduled for their 12-week follow-up, the remaining participants had completed their 6-week evaluations and were in the final phase of the treatment protocol. For those participants still in treatment, a modified version of the treatment protocol was created utilizing resistance bands so participants could continue treatment without access to fitness facilities. Bands of varying resistance (Rogue Monster Bands, Rogue Fitness, Columbus, OH) were mailed to the participants to ensure that they could replicate the resistance of isotonic exercises as closely as possible. If possible, remaining follow-up evaluations were completed remotely with the participants completing questionnaires online. Therefore, follow-up measures of tendon morphology and mechanical properties, lower extremity function and quadriceps muscle performance were not collected for these participants.
Statistical Analysis
Statistical analysis was performed using R version 3.6.3 and IBM SPSS version 26 (Chicago, IL) statistics software.(43,44) The target sample size was determined a priori based on a minimally clinically important improvement of 13 points in the VISA-P from baseline to 12-weeks using values obtained from a prior study.(11,22) It was determined that 10 participants would be required per group with 80% power and alpha set at 0.05. To account for drop-out, the target recruitment was set at 15 participants per group (30 total).
Descriptive statistics were calculated for demographic information at baseline and outcome measures at each timepoint for both groups. Demographic characteristics between groups were compared using Student’s t-tests or Mann-Whitney U tests when the assumption of normality was not met. Additionally, descriptive statistics were calculated for results of recruitment, randomization, compliance, retention, and safety. A priori criteria were not established to determine if a full clinical trial should be conducted.
A 2x3 Generalized Linear Mixed Model (GLMM) was used to test the change over time for both groups for the primary outcomes: symptom severity, psychological factors, tendon morphology and mechanical properties, lower extremity function and quadriceps muscle performance using the intention to treat principle.(45–48) For morphology, mechanical properties, lower extremity function and quadriceps muscle performance the most symptomatic limb was used for the analysis. If participants had bilateral symptoms, they were asked to identify their most symptomatic limb in their baseline questionnaires. Group (PGA or PFA) and time point (baseline, 6-, and 12-weeks) were included as fixed effects. A compound symmetric covariance matrix was used to model the correlation among residuals. GLMM models are able to garner accurate estimates in the presence of missing data without excluding entire cases. Allowing for anyone with an observation at a time point to be included, assuming that data is at least missing at random.(48).
To test the assumption of normality and to look for outliers, residuals were tested using Shapiro-Wilks tests, and screened for outliers. If time was significant, all pair-wise comparisons were tested post-hoc. Mean differences among timepoints were compared to the smallest detectable change (SDC) and minimally clinically important difference (MCID) to assess the magnitude of effect for all outcomes (Supplementary material). Alpha was set at 0.05 for all tests.