Longitudinal quantitative T2 mapping and SPECT-CT assessment of unloading therapy within a randomized controlled trial for medial knee osteoarthritis

Introduction

Osteoarthritis (OA) is characterized by progressive degeneration of articular cartilage, accompanied by changes in subchondral bone, synovial inflammation, and osteophyte formation (1). These processes lead to joint space narrowing, pain, stiffness, and loss of function. Although conventional radiography is still considered the gold standard for diagnosing and monitoring OA, it is limited to detecting features of advanced OA. There is a growing interest in advanced quantitative imaging of OA processes with the aim to diagnose and monitor OA in a more sensitive way that ideally also correlates well with clinical symptoms (2). Quantitative imaging can support the development of treatments for early OA, such as disease-modifying OA drugs, by enabling earlier detection of OA and facilitating monitoring of treatment effectiveness (3). Transverse relaxation time mapping (T2 mapping), expressed in T2 relaxation times, is a widely applied quantitative magnetic resonance imaging (MRI) technique in OA research that is able to assess the collagen deterioration of articular cartilage (4,5). As OA progresses, collagen fiber orientation and network integrity degrade, reducing water binding within the collagen matrix and leading to increased T2 relaxation times in affected cartilage regions. Single-photon emission computed tomography-computed tomography (SPECT-CT) visualizes subchondral bone remodeling with a nuclear tracer bound to a bisphosphonate that is absorbed in regions of active bone turnover and thus accumulates in osteoarthritic joints (6,7). Unlike various other imaging methods for OA, SPECT-CT captures the current metabolic activity of the disease rather than focusing on its structural damage. With the availability of advanced iterative reconstruction techniques in recent years, SPECT-CT can now be analyzed quantitatively using standard uptake values (SUVs) (8). We know that these quantitative imaging techniques are able to detect OA in an earlier stage as both articular cartilage deterioration and remodeling of the subchondral bone occur well before thinning of the cartilage or subchondral bone changes are visible on conventional radiography (9-11). However, their use for the assessment of the effects of OA treatment have only been sparsely reported (12-16). In this study, we aim to explore whether T2 mapping and quantitative SPECT-CT are able to detect changes in knee articular cartilage and subchondral bone after unloading therapy with an unloader brace and a high tibial osteotomy (HTO) in patients with medial knee OA and a varus knee malalignment (17,18). Patients with medial knee OA that receive an unloading treatment for the medial compartment are a particularly interesting group for quantitative imaging research as both the unloaded affected knee compartment and the healthy compartment, that becomes more heavily loaded by the therapy, can be evaluated. For T2 mapping, this could lead to an increase of T2 values in the lateral compartment, while T2 values in the medial compartment stabilize. For quantitative SPECT-CT, this may lead to decreased SUV in the medial compartment, with concurrent increase in the lateral compartment. In this explorative study, we assess differences in quantitative imaging outcomes before and after initiation of the unloading treatment with an interval of 1 year and compare the medial and lateral knee compartments at both time points. We assess whether there is a correlation between the two imaging techniques in the observed changes over time. Finally, we examine the relationships between the longitudinal quantitative imaging results and clinical outcomes. We present this article in accordance with the CONSORT reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-aw-2039/rc).

Methods

Subjects and treatment

Patients with symptomatic knee OA were included in a multicenter randomized controlled trial (RCT) on the efficacy of an unloading brace versus an HTO treatment, provided they were considered suitable for both treatments by their attending orthopedic surgeon. The trial, named ‘Should a patient with medial knee OA be operated?’, was conducted in nine hospitals in the Netherlands (Erasmus MC University Medical Center Rotterdam, Sint Fransciscus Gasthuis Rotterdam, Martini Hospital Groningen, University Medical Center Groningen, Meander Medical Center Amersfoort, Sint Maartenkliniek Woerden, Sint Jans Gasthuis Weert, Amphia Hospital Breda, Flevoziekenhuis Almere), with patient recruitment between August 2014 and February 2019. The trial was registered in the Dutch Trial Register on January 24^th, 2014 (NTR NL4200, https://onderzoekmetmensen.nl/en/node/27146/pdf). The inclusion criteria were: age 18–65 years, confirmed radiographic medial knee OA Kellgren & Lawrence (K&L) grade I–III and a varus knee malalignment of 0–14 degrees. Exclusion criteria were: lateral knee OA K&L grade ≥ II, knee flexion <100°, previous lateral meniscectomy and rheumatoid arthritis. These criteria are in concordance with the Dutch National Guidelines for the treatment of unilateral knee OA (19). When patients had bilateral knee complaints, the most symptomatic knee was included in the study. The primary outcome of the RCT was knee pain after 1 year of follow-up assessed with the pain subscale of the Knee Injury and Osteoarthritis Outcome Score (KOOS), which is presented in the manuscript on the study’s clinical outcomes (20). In this manuscript, a comprehensive account of the research methodology is available, including trial design regarding the primary outcome and information on adverse events. In the current paper, we discuss secondary outcomes, namely the results of the quantitative MRI and SPEC-CT imaging performed in the study. Patients received an MRI scan with quantitative T2 mapping at time of inclusion and 1 year after the date of inclusion. SPECT-CT was only performed in patients that agreed to undergo this additional examination. In these patients, a scan was made on the same day as the MRI scan. All patients visited the Erasmus MC University Medical Center Rotterdam for the quantitative imaging, so both the MRI and SPECT-CT for both time points were made using the same imaging equipment. Patients filled in patient reported outcome measures (PROMs) at time of inclusion and 1 year later. For this study, we used the KOOS (21,22). The KOOS questionnaire consists of 5 subscales: ‘symptoms’, ‘pain’, ‘activities of daily living’, ‘sport and recreation’ and ‘quality of life’. After the baseline scans and filling in the baseline PROMs, the patients were randomized to either a treatment with an unloader brace or an HTO. Patients were referred back to their attending orthopedic surgeon for initiation of the treatment. The unloader brace (Unloader One, Össur hf., Reykjavík, Iceland) that was used is this study is an off-the-shelf brace available in different sizes that was individually fitted by a certified orthotist. The brace applies a valgus stress to the knee. It aims at shifting the load from the medial to the lateral knee compartment and thus unloading the osteoarthritic medial compartment. The HTO was performed according to the preferred surgical technique in the participating hospitals. This could either be a medial opening wedge or a lateral closing wedge osteotomy using a titanium plate and screws (TomoFix, DePuy Synthes, PA, USA), or a lateral closing wedge osteotomy using two cobalt-chrome staples (Stepped High Tibial Osteotomy Staples, Stryker, MI, USA). In both techniques, the varus malalignment of the knee was surgically adjusted to a 3 to 4 degrees valgus overcorrection.

Image acquisition

T2 mapping was performed on a 3 Tesla MRI scanner (Discovery MR750, GE Healthcare, Milwaukee, WI, USA) with a dedicated eight-channel transmit and receive knee coil (Invivo, Gainesville, FL, USA). The T2 mapping sequence was a 3D fast spin echo sequence with 5 echo times (3, 13, 27, 40, 68 ms); an in-plane resolution of 0.5 mm × 0.8 mm; and a 3 mm slice thickness (23,24). The scan time was 9:40 minutes. A 3D high spatial resolution fat-saturated fast spoiled gradient-echo (FSPGR) sequence was performed at baseline for cartilage segmentation as it provided a better contrast between the cartilage and the surrounding tissue than the T2 mapping scan. The SPECT-CT scan using two gamma cameras (Symbia T series; Siemens Healthcare, Erlangen, Germany) was made 3 hours after the administration of approximately 550 megabecquerel (MBq) technetium-99m hydroxydiphosphonate (99mTc-HDP). A low-dose CT-scan of the knee was made directly after the SPECT acquisition. By registering the SPECT image to the CT scan, the disease activity was visualized at the correct anatomical location.

Image analysis

The T2 mapping scans were analyzed with an in-house developed MATLAB (R2021a; The MathWorks, Natick, MA, USA) software tool that uses Elastix to register the different images (25,26). Full-thickness femoral and tibial cartilage masks were segmented on the sagittal slices of the FSPGR scan. Segmentation was conducted manually on five central slices of the medial and five central slices of the lateral compartment. The T2 mapping scans of both the baseline and the follow-up time points were registered to the FSPGR scan using rigid registration. Subsequently, T2 relaxation times in the segmented masks were calculated voxelwise. Femur and tibia were registered separate from each other to account for differences in knee flexion between the two scans (25). A region of interest (ROI) analyses was defined by dividing the masks into a femoral weight-bearing, tibial weight-bearing and femoral posterior sub-region. The outer perimeters of the menisci delineated the weightbearing ROIs of the femur and tibia. The femoral cartilage behind the posterior border of the menisci was considered the posterior femoral ROI. Weighted mean T2 relaxation times were calculated for each ROI, using the reciprocal square root of the Cramér-Rao lower bound as weight factor (25). The SPECT-CT scans were quantitatively reconstructed using Hermes Hybrid Recon (Version 1.1.2, Hermes Medical Solutions AB, Stockholm, Sweden). Attenuation and Monte Carlo-based scatter correction were applied by means of a low-dose CT. Images were iteratively reconstructed using Ordered Subset Expectation Maximization with 5 iterations and 15 subsets. A 0.5 cm Gaussian post reconstruction filter was used. The reconstruction uses a predetermined calibration factor to enable quantification of the activity concentration in becquerel/milliliter (Bq/mL). SUVs were obtained by normalizing the activity concentration for net injected activity and patient weight. For the segmentation of the SPECT-CT scans, two 5 cm wide cubes were drawn. One in the medial knee compartment and one in the lateral. We chose not to define subregions for the femur and tibia and neither for the weight-bearing and posterior femoral compartment. This limitation arose because the cubical shape did not allow complete separation of the femur and tibia. Care was taken to exclude activity from the osteotomy, the patellofemoral joint, the tibiofibular joint and the tibial tuberosity. In each region, the maximum SUV (SUVmax) was measured, determined by the highest voxel value in the region. SUVmax was used for quantitative analysis as it is less dependent on exact ROI delineation than the mean SUV (SUVmean). Given the heterogeneous and focal nature of tracer uptake in subchondral bone, SUVmax was considered the most robust and reproducible parameter.

Randomization, blinding, and treatment allocation

After informed consent and baseline assessment, patients were randomized 1:1 to either treatment group. Randomization was centrally performed using computer-generated numbers with stratified balanced block randomization (block sizes 2–6), stratified by surgeon experience with HTO (> or <20 procedures per year) and sex. Microsoft Access was used for the randomization algorithm. Personnel who enrolled had no access to the random allocation sequence. Due to the nature of the interventions, neither patients nor orthopedic surgeons were blinded to treatment allocation.

Sample size

As this study presents the secondary outcomes of an RCT, it was not powered to detect differences in T2 mapping or SPECT-CT outcomes. Power calculation for the RCT was based on the KOOS and is described in detail in the manuscript on the study’s clinical outcomes (20). The initial target sample size was 124 patients (62 per group). Due to slow recruitment, the sample size was recalculated in consultation with the grant provider and the Dutch Orthopedic Association to 64 patients based on the observed baseline standard deviation (SD) from our study cohort, which was considerably smaller than the initially assumed SD.

Statistical analysis

A paired t-test was used to compare the differences in quantitative imaging outcomes between the baseline and follow-up scans of both knee compartments. The differences in quantitative imaging outcomes between both knee compartments at both time points were also assessed with a paired t-test. We used linear regression to examine the correlation between the change in T2 relaxation times and the change in SUVmax over time. We also used linear regression to examine the correlation between the changes in the quantitative imaging outcomes and the changes in the different KOOS subscale outcomes. For this analysis, we used the delta of the KOOS subscales as a dependent variable and delta of the T2 relaxation times or the SUVmax as an independent variable. The above-mentioned analyses were performed for the whole group, but also separate for both treatment arms (unloader brace and HTO). We performed an independent t-test to assess whether there was a difference between the brace and HTO patients in T2 values or SUVmax change over time (a between group comparison). In case of crossovers in treatment, we analyzed patients in the group of the treatment they received by the time of the follow-up measurements (an as treated analyses). A two-sided P value of less than 0.05 was considered statistically significant. Statistical analysis was performed using SPSS (IBM SPSS Statistics for Windows, Version 28.0.1.0 Armonk, NY, USA).

Ethical consideration

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the institutional review board of the Erasmus MC University Medical Center Rotterdam (No. MEC-2013-492). All participating hospitals were informed of and agreed to the study. Informed consent was taken from all individual participants.

Results

Population

Population characteristics are described in Table 1. Fifty-one patients were included in the RCT. At baseline, all patients received a T2 mapping scan, with 23 additionally undergoing SPECT-CT. The average time between baseline measurements and start of treatment was 24 days for the brace group and 74 days for the HTO group. At 12 months’ follow-up, T2 mapping was performed in 39 patients and SPECT-CT in 20 patients. Figure 1 shows a detailed flowchart of the available scans per treatment arm at both time points. All patients filled out the baseline PROMs. One patient was allocated to the brace treatment arm, but received an HTO after 7 months because of unsatisfactory results of the brace treatment. All patients that received the follow-up MRI and SPECT-CT scans underwent the same imaging at baseline and also completed the clinical questionnaires. The analyses of the follow-up T2 mapping scan in patients that underwent an HTO resulted in some problems because of the implanted material (Figure 2). In multiple patients, it caused visual distortion of the cartilage in certain ROIs or resulted in registration errors. In 13 patients, at least one cartilage ROI was not available for analyses due to visual distortion by the metal induced artifacts. In 16 patients, one or more cartilage ROIs had to be segmented manually due to registration errors. The above-mentioned problems mainly occurred in the tibial ROIs. In four patients, the implanted osteotomy material was removed before the follow-up scan. There were no issues with the MRI follow-up analyses of the patients that received brace treatment.

Figure 1 Flowchart of the available quantitative imaging scans per treatment arm at both time points. Twelve patients were not available for the follow-up quantitative MR imaging. Three patients received a unilateral or total knee arthroplasty before the follow-up measurements. One patient was allocated to the HTO treatment, but was excluded for this treatment by the attending surgeon because of significant patellofemoral OA on the SPECT-CT scan. This patient was excluded from the follow-up analysis because no unloading treatment was given. Two other patients refrained from the HTO therapy after randomization and did not respond to the invitation for follow-up measurements. Four patients started their treatment but did not respond to the invitation for follow-up measurements or did not want to travel to the hospital for these measurements. FU, follow-up; HTO, high tibial osteotomy; MR, magnetic resonance; SPECT-CT, single-photon emission computed tomography-computed tomography; TKA, total knee arthroplasty.

Figure 2 Multimodal imaging of the knee compartments. (A) Sagittal T2 map of medial femoral condyle. Higher T2 relaxation times (in milliseconds) represent a more deteriorated condition of the cartilage. (B) Coronal SPECT-CT image showing high activity in the medial compartment (left side of the image). (C) Sagittal T2 image of the medial compartment after an HTO showing artefacts caused by the implanted material. HTO, high tibial osteotomy; SPECT-CT, single-photon emission computed tomography-computed tomography.

T2 mapping

In Table 2, the average T2 relaxation times are shown for all patients, as well as separated per treatment arm. In general, there was an increase in the T2 values at the follow-up measurement compared to the baseline scan. The whole group showed a statistically significant increase in T2 values in all weight-bearing regions except for the medial tibial plateau [T2 relaxation times of baseline vs. follow-up in ms: medial weight bearing femoral condyle (wbFC) 44.2 vs. 44.3; lateral wbFC 38.3 vs. 39.9; medial weight bearing tibial plateau (wbTP) 40.9 vs. 40.4; lateral wbTP 34.8 vs. 37.3]. When the results were separated per treatment arm, the statistically significant increase of the lateral weight-bearing regions was only observed in the HTO group. In this group, the T2 values of the medial compartment did not statistically significantly change over time (baseline vs. follow-up: medial wbFC 45.6 vs. 44.4; lateral wbFC 38.4 vs. 39.9; medial wbTP 41.6 vs. 42.1; lateral wbTP 35.2 vs. 39.2). On the other hand, the brace group showed a statistically significant increase of the medial weight-bearing femoral condyle, while the cartilage of the lateral compartment did not change significantly (baseline vs. follow-up: medial wbFC 42.5 vs. 44.1; lateral wbFC 38.2 vs. 38.9; medial wbTP 40.2 vs. 39.5; lateral wbTP 34.4 vs. 35.5). When comparing the medial and lateral ROIs of the baseline scan, we observed statistically significant higher medial weight-bearing T2 values compared to the lateral ROIs of both the femur and the tibia for both the group as a whole and separated per treatment (medial vs. lateral of all patients: wbFC 44.2 vs. 38.3; wbTP 40.9 vs. 34.8; brace group: wbFC 42.5 vs. 38.2; wbTP 40.2 vs. 34.4; HTO group: wbFC 45.6 vs. 38.4; wbTP 41.6 vs. 35.2). The follow-up scans also showed higher T2 values in the medial weight-bearing ROIs, but there was only a statistically significant difference in the total group of patients and in the tibial plateaus of the brace group (medial vs. lateral of all patients: wbFC 44.3 vs. 39.9; wbTP 40.5 vs. 37.3; brace group: wbFC 44.1 vs. 38.9; wbTP 39.5 vs. 35.5; HTO group: wbFC 44.4 vs. 40.4; wbTP 42.1 vs. 39.2). The posterior femoral condyle cartilage did not show any statistically significant differences, neither between baseline and follow-up, nor between medial and lateral. Comparison of the change in T2 relaxation times between the brace and HTO group did not show any statistically significant differences in any of the cartilage ROIs.

SPECT-CT

In Table 3, the average SUVmax values are shown for all patients, as well as separated per treatment arm. In the comparison between the baseline and follow-up scans of all patients combined, we observed a statistically significant decrease in SUVmax in the medial compartment (SUVmax of baseline vs. follow-up: medial 13.6 vs. 12.0; lateral 7.1 vs. 7.2). When comparing the two treatment arms, this statistically significant decrease in the medial compartment was only seen in the HTO group (medial 13.4 vs. 11.0; lateral 7.4 vs. 7.6). No statistically significant changes in SPECT-CT activity of the lateral compartment were seen between baseline and follow-up in all groups. The SUVmax values of the medial compartment were statistically significant higher than the lateral compartment for both the group as a whole and separated per treatment at baseline (medial vs. lateral of all patients 13.6 vs. 7.1; brace group 13.8 vs. 6.6; HTO group 13.4 vs. 7.4) and follow-up (medial vs. lateral of all patients 12.0 vs. 7.2; brace group 14.4 vs. 6.4; HTO group 11.0 vs. 7.6). Comparison of the change in SUVmax between the brace and HTO group did not show any statistically significant differences for both the medial and the lateral compartment.

Correlation of change in T2 and SUV

We did not observe a correlation between both imaging modalities in any of the ROIs. This applied for both the group as a whole and separated per treatment.

Correlation of change in quantitative imaging and clinical symptoms

In Table S1, the average KOOS values are shown for the brace patients and the HTO patients. Major improvements of the KOOS subscale scores were seen for the HTO patients while little of no improvement was seen for the brace patients. We did not observe a correlation between the quantitative imaging outcomes of the T2 mapping or SPECT-CT and the clinical outcomes as reported in the KOOS questionnaire for all ROIs and all KOOS subscales for both the group as a whole and separated per treatment.

Discussion

The results of this explorative study using quantitative T2 mapping and SPECT-CT in a multicenter RCT on the effects of unloading therapy for medial knee OA, show that both techniques are clearly able to differentiate between the medial and lateral compartments. The statistically significant different T2 values of the medial and lateral weight-bearing ROIs indicate that, besides cartilage loss, the remaining cartilage also has a different, deteriorated, composition, which is in in accordance with previous research (3,5,27). The SPECT-CT results support the concept that OA causes changes to the subchondral bone as a separate phenomenon from the cartilage changes in the OA process (7,28). The changes in T2 relaxation times and SUVmax over time show clear differences between both unloading therapies. The results suggest that HTO effectively transfers load from medial to lateral compartment, as evidenced by increased lateral T2 values and decreased medial SUVmax at follow-up, unlike the brace group. These results are in line with the results of the clinical outcomes of this RCT showing that patients receiving an HTO improve clinically based on the PROMs after 1 year of follow-up while the brace group shows almost no improvement (20). Only limited literature has been published on longitudinal quantitative imaging assessment of OA unloading therapies. A study using T1rho and T2 mapping to evaluate the effect of a medial opening wedge HTO using temporarily fixation with an external fixator (hemicallotasis) on the medial weight-bearing femoral and tibial cartilage showed a significant decrease in T2 values 1 year after surgery, corresponding to an improvement of the cartilage. No significant differences were observed in the T1rho results (12). This study did, however, not examine the effects of the HTO on the lateral cartilage. Another study using delayed gadolinium enhanced MRI of cartilage (dGEMRIC) in the assessment of the effects of HTO and knee distraction did not show statistically significant changes in cartilage quality 2 years after knee distraction (16). On the other hand, the dGEMRIC values 2 years after an HTO showed lower proteoglycan concentration in medial cartilage (associated with poorer quality) and higher concentration in lateral cartilage. These contradictory findings were attributed to the idea that the proteoglycan concentration also depends on the pressure on the cartilage, which is changed after HTO, resulting in increased lateral compression and a slight medial decompression. A previous study in which SPECT-CT was performed before and after a medial opening wedge HTO showed a reduction of bone tracer uptake in the medial compartment (29). This study, however, did not perform a quantitative assessment of these findings.

We did not observe any correlation between the change in T2 values and the change in SUVmax in any of the ROIs. A possible explanation is that, although both techniques show the consequences of load transfer in the HTO group, they depict it in a different way. T2 mapping shows increased deterioration of the lateral compartment, while quantitative SPECT-CT shows a decrease in disease activity of the medial compartment after HTO treatment. There was, however, also no correlation between the change of the T2 values of the lateral weight-bearing ROIs and the change in SUVmax of the medial compartment in HTO group. We did not observe a correlation between the change of the quantitative imaging outcomes and the change in clinical outcomes, as reported in the KOOS questionnaire. Although somewhat disappointing, this was not unexpected, as to date no imaging technique has displayed a distinct correlation with the severity of the clinical OA symptoms. As illustrated in this paper, T2 mapping and quantitative SPECT-CT assess different aspects of the OA process. T2 mapping provides a measure for early structural damage, while SPECT-CT offers compelling insights into OA by revealing the current metabolic activity of the disease. This makes both techniques complementary in OA research. When implementing these techniques in clinical research of clinical practice, they have their advantages and disadvantages. The advantage of T2 mapping over SPECT is that it does not require an intravenous agent, there is no waiting time between injection and scanning, and the spatial resolution is higher. Furthermore, when a SPECT-CT scan is made shortly after performing the osteotomy, there will be substantial activity resulting from the osteotomy that might interfere with the measurement of the OA process. On the other hand, SPECT-CT is not affected by artifacts caused by the implanted metal in an HTO procedure. Our study also showed relatively greater quantitative differences between the affected and unaffected compartment with SPECT-CT. This might suggest that SPECT-CT is more sensitive than T2 mapping for capturing subtle changes, but this is not supported by our longitudinal data. A strength of this unique study is the comprehensive assessment of OA interventions using of multimodal quantitative imaging techniques. The study also has limitations. First, we studied a small sample size with a skewed distribution between the patients that received the brace and patients that received the HTO treatment. The skewed distribution was primarily due to our choice for block randomization. Initially, 124 people were to participate in the study, which was taken into account in the block randomization. Additionally, stratification was performed in the randomization for all the nine centers that participated in the RCT. Due to slow inclusion, an interim analysis was executed which justified a reduced number of inclusions based on the primary clinical outcomes. The study was concluded after 51 patients. This resulted in the small sample size and the skewed distribution. There was an even larger mismatch between the treatment arms in the patients that received an SPECT-CT scan, because this scan was not performed in every patient. Because of the limited sample size, we did not perform sub-analyses for the different osteotomy techniques. Another limitation is that in nine patients, we did not obtain a follow-up MRI and SPECT-CT scan. The fact that the imaging was done at one location, while patients were included in different centers across the country, meant that some of these patients did not want to travel for the follow-up scans. We chose to perform all scans at one location as quantitative imaging results can vary between different scanning protocols and vendors and therefore cannot be pooled (30). This emphasizes the need for standardization in using quantitative imaging in multicenter studies (9). An additional limitation was the time lag between the moment of inclusion and the start of therapy, especially in HTO. As a result, the time between intervention and follow-up scan was relatively short. However, we believe that the outcome of the difference between brace and HTO would not be different if the therapy had been performed sooner after randomization. The disparities discovered might have been even more significant had the therapies been given more time to take effect. We had one protocol violation in the study as a patient was excluded from unloading therapy because of significant patellofemoral OA on the SPECT-CT scan. This scan was not part of the regular work-up for medial knee OA, so the attending surgeon normally does not have this information. However, the scan could not be anonymized for safety reasons. This allowed the surgeon to view the results of the scan and decided to forego the unloading therapy. There was one treatment crossover during the course of the study. One patient that was randomized to the brace group but received an HTO 7 months after inclusion. This was allowed according to the study protocol. Since we assumed that the influence of the treatment on the structural properties of cartilage and bone has no subjective component, we decided to analyze this patient in the HTO group (as treated analyzes). A final, important limitation is that the RCT was not powered to detect differences in T2 mapping or SPECT-CT outcomes. Negative findings, especially the absence of correlations between imaging modalities and between imaging and clinical outcomes, should be interpreted with caution given the small sample size, imbalance between treatment groups, and incomplete follow-up imaging. Furthermore, a considerable number of cartilage ROIs in the HTO group required exclusion or manual correction due to metal-induced artifacts and registration errors which may have influenced longitudinal comparisons and between-group analyses. On the other hand, we previous showed that reliable T2 values can be obtained from cartilage ROI in the vicinity of HTO material as long as the cartilage is not visually distorted (31).

Conclusions

In conclusion, both T2 mapping and quantitative SPECT-CT show clear differences between the medial and lateral compartment of patients with medial knee OA and are able to detect changes of an unloading therapy after a relative short period of time of only 1 year. Our results suggest that HTO accomplishes a load transfer from the medial to the lateral compartment, while the unloader brace does not. Both techniques depict the OA processes in a different way and are therefore useful and complementary for monitoring OA therapies.

Acknowledgments

We would like to thank Reinoud W. Brouwer, Robert D. A. Gaasbeek, Sorin G. Blendea, Rutger C. I. van Geenen, Ronald J. van Heerwaarden, Frank Th. G. Rahusen, Sjoerd K. Bulstra, and Mark L. M. Falke for their efforts in including patients and initiating the treatment. This work partially contains Joost Verschueren’s thesis of the PhD degree in 2024, completed at Erasmus University Rotterdam in the Netherlands.

Footnote

Reporting Checklist: The authors have completed the CONSORT reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-aw-2039/rc

Trial Protocol: Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-aw-2039/tp

Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-aw-2039/dss

Funding: This research was supported by the Dutch Arthritis Foundation (No. NR BP12-03-401). The funder was not involved in the design, conduct, analysis and reporting of the trial, except for the approval of lowering the sample size.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-aw-2039/coif). E.H.G.O. serves as an unpaid Associate Editor of Quantitative Imaging in Medicine and Surgery. All authors declared that this research was supported by the Dutch Arthritis Foundation (No. NR BP12-03-401). D.H.J.P. receives research funding from GE Healthcare for other projects. S.M.A.B.Z. receives independent research grants from the EU, ZonMW, and ReumaNL and serves as a paid deputy editor of OAC. E.H.G.O. receives research support from General Electric Healthcare. The authors have no other conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the institutional review board of the Erasmus MC University Medical Center Rotterdam (No. MEC-2013-492). All participating hospitals were informed of and agreed to the study. Informed consent was taken from all individual participants.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Hunter DJ, Bierma-Zeinstra S. Osteoarthritis. Lancet 2019;393:1745-59. [Crossref] [PubMed]
2. Guermazi A, Hayashi D, Jarraya M, Roemer FW. The role of imaging in disentangling the enigma of osteoarthritis. Skeletal Radiol 2023;52:2005-6. [Crossref] [PubMed]
3. Karsdal MA, Tambiah J, Hochberg MC, Ladel C, Bay-Jensen AC, Arendt-Nielsen L, Mobasheri A, Kraus VB. Reflections from the 2021 OARSI clinical trial symposium: Considerations for understanding biomarker assessments in osteoarthritis drug development - Should future studies focus on disease activity, rather than status? Osteoarthr Cartil Open 2022;4:100262. [Crossref] [PubMed]
4. Oei EH, van Tiel J, Robinson WH, Gold GE. Quantitative radiologic imaging techniques for articular cartilage composition: toward early diagnosis and development of disease-modifying therapeutics for osteoarthritis. Arthritis Care Res (Hoboken) 2014;66:1129-41. [Crossref] [PubMed]
5. Mosher TJ, Dardzinski BJ. Cartilage MRI T2 relaxation time mapping: overview and applications. Semin Musculoskelet Radiol 2004;8:355-68. [Crossref] [PubMed]
6. Dickson JC, Armstrong IS, Gabiña PM, Denis-Bacelar AM, Krizsan AK, Gear JM, Van den Wyngaert T, de Geus-Oei LF, Herrmann K. EANM practice guideline for quantitative SPECT-CT. Eur J Nucl Med Mol Imaging 2023;50:980-95. [Crossref] [PubMed]
7. Kim J, Lee HH, Kang Y, Kim TK, Lee SW, So Y, Lee WW. Maximum standardised uptake value of quantitative bone SPECT/CT in patients with medial compartment osteoarthritis of the knee. Clin Radiol 2017;72:580-9. [Crossref] [PubMed]
8. Ljungberg M. Absolute Quantitation of SPECT Studies. Semin Nucl Med 2018;48:348-58. [Crossref] [PubMed]
9. Chalian M, Li X, Guermazi A, Obuchowski NA, Carrino JA, Oei EH, Link TM. The QIBA Profile for MRI-based Compositional Imaging of Knee Cartilage. Radiology 2021;301:423-32. [Crossref] [PubMed]
10. Burstein D, Gray M, Mosher T, Dardzinski B. Measures of molecular composition and structure in osteoarthritis. Radiol Clin North Am 2009;47:675-86. [Crossref] [PubMed]
11. Crema MD, Roemer FW, Marra MD, Burstein D, Gold GE, Eckstein F, Baum T, Mosher TJ, Carrino JA, Guermazi A. Articular cartilage in the knee: current MR imaging techniques and applications in clinical practice and research. Radiographics 2011;31:37-61. [Crossref] [PubMed]
12. Nishioka H, Nakamura E, Hirose J, Okamoto N, Yamabe S, Mizuta H. MRI T1ρ and T2 mapping for the assessment of articular cartilage changes in patients with medial knee osteoarthritis after hemicallotasis osteotomy. Bone Joint Res 2016;5:294-300. [Crossref] [PubMed]
13. Theologis AA, Schairer WW, Carballido-Gamio J, Majumdar S, Li X, Ma CB. Longitudinal analysis of T1ρ and T2 quantitative MRI of knee cartilage laminar organization following microfracture surgery. Knee 2012;19:652-7. [Crossref] [PubMed]
14. Holtzman DJ, Theologis AA, Carballido-Gamio J, Majumdar S, Li X, Benjamin C T. (1ρ) and T(2) quantitative magnetic resonance imaging analysis of cartilage regeneration following microfracture and mosaicplasty cartilage resurfacing procedures. J Magn Reson Imaging 2010;32:914-23. [Crossref] [PubMed]
15. Welsch GH, Mamisch TC, Domayer SE, Dorotka R, Kutscha-Lissberg F, Marlovits S, White LM, Trattnig S. Cartilage T2 assessment at 3-T MR imaging: in vivo differentiation of normal hyaline cartilage from reparative tissue after two cartilage repair procedures--initial experience. Radiology 2008;247:154-61.
16. Besselink NJ, Vincken KL, Bartels LW, van Heerwaarden RJ, Concepcion AN, Marijnissen ACA, Spruijt S, Custers RJH, van der Woude JAD, Wiegant K, Welsing PMJ, Mastbergen SC, Lafeber FPJG. Cartilage Quality (dGEMRIC Index) Following Knee Joint Distraction or High Tibial Osteotomy. Cartilage 2020;11:19-31.
17. Brouwer RW, Huizinga MR, Duivenvoorden T, van Raaij TM, Verhagen AP, Bierma-Zeinstra SM, Verhaar JA. Osteotomy for treating knee osteoarthritis. Cochrane Database Syst Rev 2014;2014:CD004019.
18. Duivenvoorden T, Brouwer RW, van Raaij TM, Verhagen AP, Verhaar JA, Bierma-Zeinstra SM. Braces and orthoses for treating osteoarthritis of the knee. Cochrane Database Syst Rev 2015;2015:CD004020. [Crossref] [PubMed]
19. Indicatiestelling bij geïsoleerde mediale en laterale artrose van de knie. Date of access: March 18, 2026. Available online: https://richtlijnendatabase.nl/richtlijn/geisoleerde_mediale_en_laterale_artrose_van_de_knie/indicatiestelling_bij_geisoleerde_mediale_en_laterale_artrose_van_de_knie.html
20. Stam M, Verschueren J, Van Outeren MV, Brouwer RW, Gaasbeek RDA, Blendea SG, Van Es EM, Reijman M, Bierma-Zeinstra SMA. BvO-study group. Unloader brace or high tibial osteotomy in the treatment of the young patient with medial knee osteoarthritis: a randomized controlled trial. Acta Orthop 2025;96:102-9.
21. de Groot IB, Favejee MM, Reijman M, Verhaar JA, Terwee CB. The Dutch version of the Knee Injury and Osteoarthritis Outcome Score: a validation study. Health Qual Life Outcomes 2008;6:16. [Crossref] [PubMed]
22. Roos EM, Roos HP, Lohmander LS, Ekdahl C, Beynnon BD. Knee Injury and Osteoarthritis Outcome Score (KOOS)--development of a self-administered outcome measure. J Orthop Sports Phys Ther 1998;28:88-96. [Crossref] [PubMed]
23. Verschueren J, van Tiel J, Reijman M, Bron EE, Klein S, Verhaar JAN, Bierma-Zeinstra SMA, Krestin GP, Wielopolski PA, Oei EHG. Influence of delayed gadolinium enhanced MRI of cartilage (dGEMRIC) protocol on T2-mapping: is it possible to comprehensively assess knee cartilage composition in one post-contrast MR examination at 3 Tesla? Osteoarthritis Cartilage 2017;25:1484-7.
24. Chen W, Takahashi A, Han ET. 3D Quantitative Imaging of T1rho & T2 (Abstract). Proc Annu Meet ISMRM 2011;19:231.
25. Bron EE, van Tiel J, Smit H, Poot DH, Niessen WJ, Krestin GP, Weinans H, Oei EH, Kotek G, Klein S. Image registration improves human knee cartilage T1 mapping with delayed gadolinium-enhanced MRI of cartilage (dGEMRIC). Eur Radiol 2013;23:246-52. [Crossref] [PubMed]
26. Klein S, Staring M, Murphy K, Viergever MA, Pluim JP. elastix: a toolbox for intensity-based medical image registration. IEEE Trans Med Imaging 2010;29:196-205. [Crossref] [PubMed]
27. Dunn TC, Lu Y, Jin H, Ries MD, Majumdar S. T2 relaxation time of cartilage at MR imaging: comparison with severity of knee osteoarthritis. Radiology 2004;232:592-8. [Crossref] [PubMed]
28. Maas O, Joseph GB, Sommer G, Wild D, Kretzschmar M. Association between cartilage degeneration and subchondral bone remodeling in patients with knee osteoarthritis comparing MRI and (99m)Tc-DPD-SPECT/CT. Osteoarthritis Cartilage 2015;23:1713-20. [Crossref] [PubMed]
29. Mucha A, Dordevic M, Testa EA, Rasch H, Hirschmann MT. Assessment of the loading history of patients after high tibial osteotomy using SPECT/CT--a new diagnostic tool and algorithm. J Orthop Surg Res 2013;8:46. [Crossref] [PubMed]
30. Verschueren J, Eijgenraam SM, Klein S, Poot DHJ, Bierma-Zeinstra SMA, Hernandez Tamames JA, Wielopolski PA, Reijman M, Oei EHG T. (2) mapping of healthy knee cartilage: multicenter multivendor reproducibility. Quant Imaging Med Surg 2021;11:1247-55. [Crossref] [PubMed]
31. Verschueren J, Meuffels DE, Bron EE, Klein S, Kleinrensink GJ, Verhaar JAN, Bierma-Zeinstra SMA, Krestin GP, Wielopolski PA, Reijman M, Oei EHG. Possibility of quantitative T2-mapping MRI of cartilage near metal in high tibial osteotomy: A human cadaver study. J Orthop Res 2018;36:1206-12. [Crossref] [PubMed]