Predicting intra-articular residence time of hyaluronan-based viscosupplements. a comparison between human and artificial synovial fluid as degradation media

Abstract

Purpose: Hyaluronan(HA)-based viscosupplements, although resorbable, present different in vivo residence times, which are mainly related to structural modifications of the native polymer. The residence time of unmodified HA generally does not exceed 5 days, while chemically crosslinked HA can be detected in the joint up to one month. Artificial synovial fluid (ASF) have been described in the literature to mimic rheology of or drug release in human SF (SF). These models suffer from a limitation from the biological point of view: the absence of hyaluronidases, HA degrading enzymes expressed by the chondrocytes in contact with SF, well described in literature. A novel in vitro preparation of ASF with hyaluronidase is hereby reported and tested with three marketed viscosupplements: Sinovial® HL (SHL, IBSA Farmaceutici), which contains native HA of both low and high MW, Hyalubrix® (HBX, Fidia Farmaceutici S.p.A.), which contains unmodified high MW HA and Hymovis® (HYM, Fidia Farmaceutici S.p.A.), composed of HA modified with hexadecyl alkyl chains via amidic bonds (MW comprised between 500 and 700 kDa). The effect of this ASF towards HA molecular weight and rheological properties was reported and compared with the degradation profile in centrifuged human SF. Methods: ASF was formulated with the following composition: NaCl 8 g/L, KCl 0.2 g/L, Na2HPO4 1.7 g/L, Bovine Serum Albumin (BSA) 19 g/L, Bovine Testes Hyaluronidase (BTH) 6 U/mL, Gentamicin 0.01 mg/mL (pH 7.4). The salt and BSA composition and concentration in SF are well reported in the literature. At the best of our knowledge, the quantification of hyaluronidase activity in human SF is not reported. Considering that Hyaluronidase (Hyal) activity titrated in serum is at least 6 U/mL in BTH equivalent, and that Hyal (MW 55-60 kDa) can permeate synovial membrane, the same concentration of BTH was added to artificial SF. Centrifuged (cell free) human SF collected from OA patients was used as “degradation media”. Sinovial® HL, Hyalubrix® and Hymovis® were separately incubated either in ASF or SF at 37 °C for 14 days at a w/w ratio of approximately 1:1, under mild shaking. At different time points (0, 1, 7 and 14 days) the viscoelastic moduli and dynamic viscosity of the mixtures were measured with an Anton Paar MCR 92 rheometer. Moreover, at each time point the mixtures were diluted in water and filtered through Merck Vivaspin® 2 centrifugal concentrators (MWCO = 100 kDa, comparable to the synovial membrane cut-off) in order to quantitate the permeated and retained HA fractions. PBS+SF at a w/w ratio of 1:1 was used as blank for HA measurement. The percent recovery of the HA fraction with MW ≤ 100 kDa was determined by SEC-TDA (Omnisec Resolve/Reveal, Malvern instruments). Results: The degradation profile of HYM, HBX and SHL in SF or ASF is shown in the graph below: after 14 days the viscoelasticity (expressed as storage modulus: G’) of HYM/SF, HBX/SF and SHL/SF is slightly decreased; however, the elasticity of all viscosupplements in ASF is completely impaired after only few hours. After 14 days, the percent recovery of high MW HA (>100 kDa) confirms the rheological data: in SF, HYM, HBX and SHL were not depolymerized to low MW polymers while, in ASF, the percent recovery in higher MW HA was only about 45%, 10% and 13% for HYM, SHL and HBX respectively. The strong depolymerizing activity of ASF was confirmed for all the viscosupplements tested; nonetheless, HYM showed higher resistance to degradation in ASF compared to SHL. Conclusions: In this work, two different in vitro and ex vivo models were compared in order to predict the in vivo residence time of three HA-based viscosupplements in the intra-articular environment. Based on the literature, after a single intra-articular administration of high MW HA (2000 kDa), the measured residence time was as short as about 5 days. The ex vivo model with centrifuged human SF bears the limitation of the absence of hyaluronidases (a cell surface glycosylphosphatidylinositol-anchored protein, removed with the chondrocytes during the centrifugation process), while the assay in ASF added with BTH turned out to be more reliable and can therefore be employed during viscosupplement development to predict the intra-articular residence time of HA-based products