Intra-articular Hyaluronic Acid in Mild to Moderate Knee Osteoarthritis

Introduction:

The medical management of knee osteoarthritis (OA performing intra-articular treatment with hyaluronic acid (HA) has recently become more widely accepted. The effectiveness of HA regarding the improvement of pain and function - also compared to placebo - has been shown by in vitro and clinical studies. However, most in vivo studies are based on patient well being, clinical scores and plain radiographs, which have however been shown neither to be an efficient way of monitoring OA progression nor to analyze the effect of HA on cartilage changes (1, 2).

Recently MRI has gained increasing importance as imaging modality for assessing cartilage changes in general but especially regarding degenerative changes in case of OA. Quantitative MRI (qMRI) studies were used to assess cartilage volume and thickness in patients with OA. In a longitudinal study Eckstein et al. (3) found an approximated annual loss of cartilage between 4 to 6%. Moreover, the effect of treatment on morphologic parameters can be evaluated objectively in-vivo which has been the major limitation of the validation process of drug therapy in OA so far. However, until now only three studies in the recent literature have used MRI to evaluate the effect of intraarticular HA on the articular cartilage of the knee joint. While Ozturk et al. (4) and Cubukçu et al. (5) evaluated the articular cartilage using a semiquantitative grading only Anandacoomarasamy et al. (6) have utilized validated techniques to assess cartilage defects and cartilage volume quantitatively for a follow-up period of six months. The so far published analyses have focused on entire cartilage plates, although it is likely that certain subregions of these plates are stronger affected by cartilage loss than others. Wirth et al. (7) demonstrated that cartilage loss can be measured in subregions (i.e., central, internal, external) of the femoro-tibial cartilage plates which allows for additional insight into the spatial distribution of tissue loss throughout the cartilage plates in OA (7, 8). Assuming that cartilage is not homogeneously lost throughout the plates, this approach may be used to identify subregions with a higher rate of and sensitivity to change, which may, in turn, permit a reduction of sample size and follow up periods in clinical studies to be able to demonstrate structure modifying effects of pharmaceutical compounds on disease progression (7-9).

The T2 relaxation time, another MRI based quantitative parameter, on the other hand is sensitive to tissue hydration and organizational properties of the collagen fiber matrix. It is thus considered to reflect compositional and architectural aspects as a surrogate parameter for cartilage matrix quality in cartilage degeneration (10). In mild OA of the knee, the mean cartilage T2 values, the standard deviation and the entropy were increased, indicating that in mild OA the T2 values are not only elevated but also are more heterogeneous as compared to healthy cartilage (11). It is generally assumed that changes of cartilage volume and thickness appear later and may also occur at a slower rate than T2 changes of cartilage in OA. Hence, the T2 relaxation properties of cartilage may be useful as non invasive parameters to analyse the impact of HA, especially with respect to the - in most cases - relatively short follow up periods available.

Therefore the aim of this study was to assess the impact of intraarticular hyaluronic acid on clinical outcome and on volumetric (including subregions) and T2 relaxation based changes of articular cartilage in patients with moderate knee OA.

Material and methods Patients 34 patients with age rang of 40 - 64 years (mean 48 ± 9 yrs) presenting with a clinical history of mild to moderate OA, unilateral osteoarthritis of the knee at the time of inclusion were enrolled. All patients were referred from visiting orthopaedic surgeons and presented with OA grade II of the knee joint radiologically confirmed according to the Kellgren-Lawrence grading system (12).

Exclusion criteria were as follows: a) intraarticular injections in the affected knee, b) oral application of glucosamine and chondroitin sulphate during the last 6 months prior to the beginning of the study, c) clinically significant knee joint effusion or d) if knee joint infection was known or suspected and/or trauma or other specific conditions such as neoplasm, diabetes mellitus, osteonecrosis were present, which might potentially interfere with the completion of the trial. Another exclusion criterion was known rheumatoid arthritis or any other inflammatory arthritis diagnosed by American College of Rheumatology criteria.

Study design The study was conducted in a prospective, randomized way with a 6-months follow-up period. The patients were randomly assigned either to the group receiving treatment with 2ml Na-Hyaluronate (HA; MW 1.2 x 106; Ostenil, TRB Chemedica) or the group without treatment. The treatment group (n=17) were weekly treated with an intraarticular injection of HA during the first five weeks on day 0, 7, 14, 21 and 28 (see Table 1). Any pretreatment with non-steroidal anti inflammatory drugs (NSAID) had to be discontinued 7 days before the study started whereas concomitant physiotherapy was allowed. Patients were withdrawn from the study if severe reactions to the injections occurred or if there was evidence of an active infection of the injected joint at any time during the study period.

Clinical Assessments To determine clinical effectiveness of HA therapy, all patients were clinically evaluated prior to first injection (baseline) and after 6, 12 and 24 weeks (Table 1). Clinical assessment comprised self-reported visual analog pain scale (VAS) and the Western Ontario and McMaster University Osteoarthritis Index (WOMAC) assessing the three dimensions of pain, function, and joint stiffness in knee OA. Additionally, the physical function and the patient as well as the physician global score were measured. Also the body mass index (BMI) was assessed.

Quantitative MRI: Volumetric Parameters A high resolution 3D T1-weighted FLASH Waterexcitation (WE) sequence with a thickness of 1.5 mm and an in-plane resolution of 0.31 mm x 0.31 mm was obtained for each patient at baseline and 24 weeks after on a 1.5 Tesla MR-scanner using a dedicated circularly polarized knee coil (Siemens Medical Solutions, Erlangen, Germany). This protocol has been validated for quantitative assessment of cartilage morphology (3, 13-15). Coronal images were acquired for the femoro-tibial compartiment and transverse images to assess the patellar cartilage.

The MR- data were sent to the image analysis center, quality controlled, and converted to a proprietary format (Chondrometrics GmbH, Ainring, Germany). Images were read in pairs in a blinded to acquisition order fashion. The cartilage plates of the patella (P), central (weight-bearing) medial femur (cMF), central lateral femur (cLF), medial tibia (MT) and the lateral tibia (LT) were quantified as previously described using a 3D digital post-processing algorythm using the dedicated Chondrometrics Works software (16). On the coronal images, cMF and cLF started anteriorly at the first partition with interruption by the subchondral bone of diverging trochlea into the femoral condyles. Posteriorly, the last partition showing the circular structure of the posterior femoral condyles (central bone with surrounding cartilage) was identified. The partition located at 60%(2/3) between the anterior and posterior landmark was the most posterior one to be included in the cMF and cLF. Quality control of all segmentations of each dataset was performed by a single person (FE). Proprietary software was used to determine cartilage volume (VC), total area of the subchondral bone (tAB) and part of tAB covered with cartilage (cAB), mean cartilage thickness (ThCcAB), and mean thickness when counting all denuded areas as 0 mm cartilage thickness (ThCtAB). Changes were computed for the medial and lateral femorotibial compartments (MFTC/LFTC) by summing up the values of the medial tibia and femur, and the lateral tibia and femur, respectively, at baseline and follow up (17, 18).

Additionally, five subregions (central, internal, external, anterior, posterior) were determined based on the subchondral bone area (tAB) in the tibia, with the central subregion occupying 20% of the total subchondral bone area (7). The central tibial region was defined by a perpendicular cylinder around the center of gravity of the tibial subchondral bone area with the diameters being adapted to its individual shape (7). Since the weight-bearing femoral condyles are limited in anterior-posterior extension (femoral trochlea anteriorly and posterior femoral condyle posteriorly), they were divided into a central, internal and external strip-like region of interest, respectively, each occupying 33.3% of the subchondral bone area (7). Cartilage thickness (ThCtAB) was determined in all subregions.

Quantitative MRI: T2-Mapping For T2 relaxation time calculation a fat-saturated Multislice-Multiecho Turbo-Spin-Echo-sequence (MSME) (TR 3000ms / TE 13.2ms / 8 echoes, echo spacing 13.2ms / Bandwidth (BW) 130Hz / interleaved acquisition) and a 3D T1-w fast low angle shot (FLASH) sequence with selective water excitation (TR 14.2ms / TE 7.2ms / FA 15° / BW 130Hz) with identical position and spatial resolution were acquired (19) in coronal plane at baseline, and 6, 12 and 24 weeks after. Both sequences were directed to the center of the knee joint with a resolution of 0.6² x 3 mm³ (256² matrix interpolated to 512²) and a field of view of 16 cm. The FLASH sequence is used for cartilage segmentation (19-21) then being superimposed on the T2 map to calculate the cartilage T2 values. Acquisition time for the FLASH-sequence was 2 min 55 sec, for the MSME-sequence 12 min 48 sec. Interactive segmentation and 3D reconstruction (15) of the tibial cartilage plates were performed consecutively.