Article Text


Analysis of lapine cartilage matrix after radiosynovectomy with holmium-166 ferric hydroxide macroaggregate
  1. O T Mäkelä1,
  2. M J Lammi2,
  3. H Uusitalo3,
  4. M M Hyttinen2,
  5. E Vuorio3,
  6. H J Helminen2,
  7. R-M Tulamo1
  1. 1Faculty of Veterinary Medicine, University of Helsinki, Finland
  2. 2Department of Anatomy, University of Kuopio, Finland
  3. 3Department of Medical Biochemistry and Molecular Biology, University of Turku, Finland
  1. Correspondence to:
    Dr O T Mäkelä, DVM, Department of Clinical Veterinary Sciences; Faculty of Veterinary Medicine, PO Box 57, FIN-00014 University of Helsinki, Finland;


Objective: To study the short and long term effects of radiosynovectomy on articular cartilage in growing and mature rabbits.

Methods: The articular cartilage of the distal femurs of rabbits was examined four days, two months, and one year after radiosynovectomy with holmium-166 ferric hydroxide macroaggregate ([166Ho]FHMA). Arthritic changes were evaluated from histological sections by conventional and polarised light microscopy, and glycosaminoglycan measurements using safranin O staining, digital densitometry, and uronic acid determination. Proteoglycan synthesis was studied by metabolic [35]sulphate labelling followed by autoradiography, and electrophoretic analysis of extracted proteoglycans. Northern analyses were performed to determine the mRNA levels of type II collagen, aggrecan, and Sox9 in cartilage samples.

Results: Radiosynovectomy had no major effect on the histological appearance of articular cartilage in mature rabbits, whereas more fibrillation was seen in [166Ho]FHMA radiosynovectomised knee joints of growing rabbits two months after treatment, but not after one year. Radiosynovectomy did not cause changes in the glycosaminoglycan content of cartilage or in the synthesis or chemical structure of proteoglycans. No radiosynovectomy related changes were seen in the mRNA levels of type II collagen, whereas a transient down regulation of aggrecan and Sox9 mRNA levels was seen in young rabbits two months after [166Ho]FHMA radiosynovectomy.

Conclusions: [166Ho]FHMA radiosynovectomy caused no obvious chondrocyte damage or osteoarthritic changes in mature rabbits, but in growing rabbits some transient radiation induced effects were seen—for example, mild cartilage fibrillation and down regulation of cartilage-specific genes.

  • radiosynovectomy
  • rabbits
  • articular cartilage
  • collagen
  • proteoglycans
  • GAGs, glycosaminoglycans
  • [166Ho] FHMA, holmium-166 ferric hydroxide macroaggregate
  • IM, intramuscularly
  • MEM, minimum essential medium
  • SDS, sodium dodecyl sulphate

Statistics from

Synovectomy has sometimes been used to treat refractory joint inflammation instead of more conventional treatments such as intra-articular cortisone treatment. Radiosynovectomy provides an attractive alternative to surgical synovectomy and has been in clinical use for 50 years.1 The objective of radiosynovectomy is to destroy the pannus and inflamed synovium by direct radiation, with the expectation that the regenerated synovium will be free of disease. A radiopharmaceutical agent with an appropriately strong β emission can be administered directly into the joint, where it affects the synovial lining, leaving the adjacent cartilage apparently unaffected.2 Radiosynovectomy has been shown to give results at least as good as those of arthroscopic synovectomy.3,4 The benefits to patients include increased joint movement and reduced swelling, pain, and effusion.5

Although mature cartilage has been considered to be resistant to radioactivity,2,6,7 minor injury to articular cartilage remains a concern of radiosynovectomy. In some studies β radiation has been associated with chondrocyte injury, and radiosynovectomy using phosphorus-32 (32P, βmax 1.7 MeV) with injuries of both articular cartilage and the growth plate.8,9 A report on the use of gold-198 ( 198Au, βmax 0.96 MeV) indicated that rat articular cartilage is extremely resistant to radiation.6

In addition to direct radiation, synovial damage, subsequent joint inflammation, and fibrosis may also contribute to articular cartilage damage after radiosynovectomy. Extensive fibrosis of the subsynovial tissues has been reported after radiosynovectomy with yttrium-90 (90Y, βmax 2.3 MeV), dysprosium-165 ferric hydroxide macroaggregate ([165Dy]FHMA, βmax 1.3 MeV), or 32P.10–12 Fibrosis may hamper filtration and resorption of synovial fluid, and lead indirectly to articular cartilage damage because articular cartilage relies entirely on synovial fluid for its nutrition.13 Despite fibrosis, less cartilage destruction was reported in knees treated with 90Y than in untreated ones.10 In other investigations on allergic arthritis, there was no evident sparing of articular cartilage after treatment with radiosynovectomy.11,12

Holmium-166 (166Ho) has excellent physical properties for radiosynovectomy. A reasonably short half life (27 h) decreases the risks associated with eventual leakage of the radioisotope, and strong β radiation energy (βmax 1.8 MeV) ensures efficient radiation of the synovium. Its ionising γ radiation (6% 81 keV) increases the total radiation to the body only slightly. Extra-articular leakage of the radioisotope is minimised with the use of 166Ho conjugated with ferric hydroxide macroaggregate particles ([166Ho]FHMA).14,15

Despite these advantages, the effects of radiosynovectomy with [166Ho]FHMA on articular cartilage have not been previously investigated. This study aimed at filling this gap in our knowledge and studying the effects of [166Ho]FHMA radiosynovectomy on rabbit articular cartilage using morphometric, biochemical, and molecular biological analyses.



This study was performed on 51 New Zealand white rabbits. At the onset of the study 29 animals were growing (mean age 3.7 months, range 106–122 days, weight 2.4 kg, referred to as young rabbits) and 22 were mature (mean age 1 year 3 months, range 446–519 days, weight 5 kg, referred to as old rabbits). The rabbits were housed in individual stainless steel cages with a floor area of 0.6×0.6 m, and height of 0.45 m. The study protocol was approved by the Animal Care and Use Committee of the University of Helsinki.

Protocols for [166Ho]FHMA injection and sample preparation

[166Ho]FHMA was injected into the knee joints through the suprapatellar ligament of 26 young and 20 old rabbits anaesthetised with medetomidine (0.3–0.5 mg/kg body wt intramuscularly (IM); Domitor, Orion Co, Espoo, Finland) and ketamine (20–25 mg/kg body wt IM; Ketalar, Warner Lambert/Parke Davis, Solna, Sweden). One knee of each rabbit was injected with 0.4 ml of radioactive [166Ho]FHMA (referred to as the [166Ho]FHMA treated knee), and the contralateral knee with 0.4 ml of non-radioactive [165Ho]FHMA (referred to as the control knee). [166Ho]FHMA was prepared under good manufacturing practice conditions (MAP Medical Technologies OY, Tikkakoski, Finland). Another five negative control animals received no injections. Old rabbits were treated with a higher dose of [166Ho]FHMA owing to their larger body size (on average 121 MBq in old rabbits v 75 MBq in young animals). The animals were killed at periods of four days, two months, and one year after [166Ho]FHMA injections (6–10 animals in each group).

The hind limbs were removed, skinned, and the femur was fixed onto a ball joint specimen holder. Three cartilage slices were then cut perpendicular to the surface with a high speed machine tool (Dremel, Emerson Electric Co, St Louis, MO, USA) equipped with two cutting discs and a 2 mm spacer. Each sample was continuously cooled with ice cold 0.9% sodium chloride during the harvesting procedure. The first slice was taken from the patellar surface just above the femoral condyles for histology, quantitative autoradiography, and densitometric analyses of glycosaminoglycans (GAGs). The second slice was cut approximately 2 mm proximal to the first one for total sulphate incorporation and uronic acid analyses. The third slice was cut from the superior part of the patellar surface of the femur for electrophoretic analysis of proteoglycans. All the slices were stored in minimum essential medium supplemented with Earle’s salts (MEM, Life Technologies, Paisley, Scotland) including 100 U/ml penicillin and 100 μg/ml streptomycin at 4°C until the next day, when they were labelled with [35]sulphate (50–100 μCi/ml) in Dulbecco’s MEM at 37°C in a 5% CO2 atmosphere. After six hours the labelling medium was replaced by ice cold phosphate buffered saline.

Cartilage histology and assessment of the superficial collagen network

Specimens for histology were fixed in 4% formaldehyde at 4°C for 20 hours followed by decalcification in buffered 10% EDTA and 4% formaldehyde for 12 days at room temperature, dehydration with alcohol, and embedding in Paraplast Plus wax (Lancer Division of Sherwood Medical, Kildare, Ireland).16 Sections 3 μm thick were cut for safranin O staining, and 5 μm thick sections for haematoxylin and eosin staining, and subsequent histological assessment. The superficial collagen network was assessed by quantitative polarised light microscopy.17

Digital densitometry of GAGs and thickness measurement of uncalcified cartilage

Densitometric determination of GAG concentration in safranin O stained sections was carried out in the middle of the femoral groove as described earlier.16,17 The thickness of the uncalcified articular cartilage was measured and divided into 12 fractions. The two most superficial fractions represented the superficial, the next five fractions the intermediate, and the deepest five fractions the deep zone. The results were calculated as the mean area integrated optical density proportional to one square micrometre in each zone (OD/μm2).

Uronic acid content

The GAG content of cartilage was calculated as micrograms of uronic acid measured with the m-hydroxydiphenyl method per milligram of wet weight.18

[35]Sulphate incorporation and autoradiography

For the analysis of total [35]sulphate incorporation, cartilage specimens were digested overnight at 60°C with 0.05% proteinase K (Roche Boehringer Mannheim, Mannheim, Germany) and 10 mM EDTA in 0.1 M phosphate buffer (pH 7.4). Unincorporated sulphate precursors were removed by PD-10 gel filtration columns (Amersham Pharmacia-Biotech, Uppsala, Sweden), and the radioactivity was measured by liquid scintillation counting (Winspectral, Wallac, Turku, Finland) from PD-10 eluates, and from medium samples using a water soluble scintillation cocktail (Optiphase Hisafe III, Wallac).

For autoradiography, tissue blocks containing articular cartilage and the underlying bone were prepared as described earlier,19,20 with the exception that no background staining was used. Microscopic analysis of the grain area fraction was performed in the middle of the femoral groove using image analysis. Each microscopic view was divided into 10 fields from the cartilage surface to the tidemark. The two most superficial fractions represented the superficial zone, the four consecutive fields the middle zone, and the deepest four the deep zone. The results are presented as the average grain fraction percentage of the total tissue volume.

Electrophoretic analysis of proteoglycan monomers

Cartilage was dissected from the underlying bone, weighed, and extracted with 500 μl of 4 M GuCl in 50 mM sodium acetate (pH 5.8) containing inhibitors of bacterial growth and proteases.20 Extraction was continued at 4°C for 24 hours. Proteoglycans were precipitated in 75% ethanol at 4°C and washed twice in ice cold 70% ethanol. The samples were dissolved in sodium dodecyl sulphate as previously described21 for electrophoresis in submerged horizontal 5 mm 1.2% agarose gels. The gels were scanned for densitometry, and exposed to autoradiography films for two weeks. The scanned images of films and gels were analysed by image analysis software (IP-Lab, Scananalytics, Fairfax, VA, USA).20

Northern hybridisations

For extraction of total RNA, both femoral condyles were cut off with a bone cutter, transferred into liquid nitrogen, and kept at −80°C. Before extraction of total RNA using a modification of the guanidinium isothiocyanate method, the samples were pulverised under liquid nitrogen.22 For northern analyses, 10 μg aliquots of total RNA were denatured by glyoxal and dimethyl sulphoxide. After electrophoresis on 0.75% agarose gels the RNAs were transferred overnight onto Pall Biodyne membranes as recommended by the supplier (Pall Europe, Portsmouth, UK). The filters were prehybridised and hybridised with cDNA clones for rabbit type II collagen,23 human aggrecan core protein,24 and Sox9,25 and for the 28S rRNA26 according to a protocol recommended by the supplier. All inserts were labelled with [32P]deoxycytidine triphosphate using the random priming method (Boehinger Mannheim, Germany). After high stringency washes, the bound radioactivity was detected and quantified on a molecular imager phosphorimager (Bio-Rad, Hercules, CA, USA), and the mRNA signals were corrected for variations in the 28S rRNA levels determined by hybridisation.

Statistical analysis

Differences in parameters between treated and control knees were tested with the non-parametric Wilcoxon matched pairs signed rank test because the data did not fulfil the criteria for parametric testing.


Table 1 summarises the results of the histological analyses of rabbit knee joints at four days, two months, and one year after radiosynovectomy. In young rabbits superficial articular cartilage fibrillation was seen more frequently in [166Ho]FHMA treated joints than in control joints. Vertical clefts disrupting the articular cartilage surface were only seen in old rabbits: in three animals processed four days after radiosynovectomy, and in two animals processed two months after the treatment. These changes occurred equally in [166Ho]FHMA treated and control knees.

Table 1

Number of knees with articular cartilage fibrillation‡

The average thickness of uncalcified cartilage in young rabbits was 388 μm, and in old rabbits 505 μm. No differences in cartilage thickness were found between [166Ho]FHMA treated knees and control knees.

The intensity of safranin O staining increased from the superficial zone towards deep articular cartilage (data not shown). Synovectomy with [166Ho]FHMA did not alter safranin O staining in any part of the articular cartilage. Neither were any differences observed in GAG concentration of the articular cartilage determined by uronic acid analyses for the [166Ho]FHMA treated and control knees (table 2). Metabolic labelling with [35]sulphate did not reveal any significant effect of radiosynovectomy on proteoglycan synthesis in the incorporation analysis of tissue samples (table 2) or in autoradiographic analysis of the cartilage zones (table 3, fig 1).

Table 2

Concentration of uronic acid (μg/mg articular cartilage) and total amount of [35]sulphate incorporated (pmol SO4/mg cartilage wet wt/h)

Table 3

Autoradiography slides; total grain volume fraction (% of the total tissue volume occupied by grains) in superficial, middle, and deep cartilage layers

Figure 1

Sections of articular cartilage at the patellar surface of the femur prepared for autoradiography. In an old rabbit four days after [166Ho]FHMA treatment an equal level of [35]sulphate incorporation is seen in all cartilage zones of (A) control and (B) [166Ho]FHMA treated samples. S, superficial zone; I, intermediate zone; D, deep zone. Bar =200 μm.

Analysis of the size distribution of proteoglycan monomers by SDS agarose gel electrophoresis showed that the proteoglycans were more homogeneous in size in young rabbits than in the old ones (fig 2). No obvious changes induced by [166Ho]FHMA radiosynovectomy were seen in the size distribution.

Figure 2

SDS agarose gel electrophoresis of proteoglycans from lapine cartilage. Proteoglycans were extracted from cartilage samples from the patellar surface of the femur two months and one year after [166Ho]FHMA treatment of young rabbits. Proteoglycans were separated in 1.2% agarose gels and stained with toluidine blue. Afterwards the gels were exposed to autoradiography films. Densitometric analysis of films (A and C) and stained gels (B and D) is shown, as well as representative samples from gels and films. 2M, samples taken two months after [166Ho]FHMA treatment; 1Y, samples taken one year after [166Ho]FHMA treatment.

When calculated for a constant amount of 28S rRNA, the levels for proα1(II) collagen mRNA were considerably higher in growing rabbits than in rabbits aged over one year. The proα1(II) collagen mRNAs remained unchanged after radiosynovectomy, but reduced levels of aggrecan and Sox9 mRNA were seen in the [166Ho]FHMA treated knees of young rabbits two months after treatment in comparison with the contralateral control knees. At the one year follow up, no differences were seen in the mRNA levels between the [166Ho]FHMA treated and control knees (fig 3).

Figure 3

A summary of changes in the mRNA levels of proα1(II) collagen, aggrecan, and Sox9 after [166Ho]FHMA treatment. The data are compiled from quantitative analyses of northern hybridisations of young and old rabbits at times shown above the diagrams. *Significance in northern analyses (p< 0.05).


Histological analysis and assessment of fibrillation indicated that [166Ho]FHMA radiosynovectomy caused no obvious adverse effect on articular cartilage in old rabbits. In young rabbits, however, mild cartilage fibrillation was seen more frequently in the radiosynovectomised knees, but only transiently two months after the treatment. Five old rabbits had vertical clefts in both the radiosynovectomised and control knees, probably representing a naturally occurring phenomenon. A benefit of using young animals for long term experiments is that they are less likely to have active degenerative processes at the time of experiment. On the other hand, young animals with their greater reparative abilities might make the experimental results less applicable to adult joints with signs of cartilage degeneration.27 The results of this study suggest that the cartilage of young rabbits is more sensitive to radioactivity induced metabolic alterations. The results also show that active matrix production in growing rabbits helps to prevent permanent cartilage damage.

None of the biochemical analyses gave any information which might explain the increased frequency of fibrillation in the radiosynovectomised knees of young rabbits. Cartilage thickness remained unchanged after radiosynovectomy, and there was no decrease in articular cartilage GAG concentration as has been found after surgical synovectomy.28–30 Similarly to the present study, no decrease in GAG concentration was detected after 90Y radiosynovectomy in dogs.31 Biochemical analyses also showed that [166Ho]FHMA radiosynovectomy had no effect on the average size of extracted proteoglycans. In old rabbits, proteoglycan heterogeneity can be explained by the fact that the normal aging process causes partial core protein cleavage of proteoglycans in tissues, whereas the newly synthesised proteoglycans produced by the same tissue exhibit remarkable homogeneity.32

Chondrocyte defects in the superficial zone have been reported after 90Y radiosynovectomy.8 Because the superficial chondrocytes would most likely be the first targets of damage caused by the synovial radioactivity, autoradiographic analysis of [35]sulphate labelled histological sections was performed. No decrease in autoradiographic grain density could be seen in any cartilage zone after [166Ho]FHMA radiosynovectomy. Similar results have been reported after surgical synovectomy,28 whereas increased synthesis of proteoglycans has been reported after 90Y radiosynovectomy.31 Higher rates of proteoglycan synthesis and turnover have been associated with incipient cartilage damage.33 No such increase was evident after [166Ho]FHMA radiosynovectomy. On the other hand, decreased levels of aggrecan mRNA were seen in the knee cartilage of young rabbits two months after radiosynovectomy. The significance of this observation remains to be determined, as no corresponding changes were seen in the proteoglycan or GAG content of the tissue at this time.

It has been proposed that synovectomy results in some degree of enzymatic degradation of cartilage, which continues until synovial regeneration is complete.28 As discussed above, increased matrix synthesis has been associated with attempted articular cartilage repair and with early osteoarthritis.34–36 In our study no up regulation of mRNAs for Sox9, an activator of the chondrocyte phenotype, nor proα1(II) (for type II collagen) and aggrecan, two key components of cartilage matrix, was detected. Careful tissue preparation and analysis of 28S rRNA to confirm equal loading of samples minimised the effect of RNA derived from subchondral bone on the levels of cartilage-specific mRNAs. A lack of a repair process in the present model after [166Ho]FHMA radiosynovectomy indirectly suggests that no severe radiation induced damage had occurred in articular cartilage. In contrast, radiosynovectomy in young rabbits resulted in a transient decrease (at two months) in mRNA levels of Sox9 and aggrecan. Earlier studies have shown that Sox9 is a transcriptional activator of both the Col2a1 gene37,38 and the aggrecan gene.39 The reason why reduced Sox9 mRNA levels did not result in down regulation of proα1(II) mRNA remains unexplained. As the proα1(II) mRNAs levels remained unchanged, it is unlikely that the reduction in the aggrecan and Sox9 mRNAs reflected irreversible chondrocyte damage. The importance of the decreased levels of aggrecan and Sox9 mRNAs remains open because no chondrocyte damage was apparent in the autoradiographic analysis of cartilage samples. Neither were any adverse effects detected in the chemical composition of articular cartilage.

Our results therefore indicate that intra-articular [166Ho]FHMA treatment does not cause cartilage damage in adult lapine joints despite the reported radiation damage on synovial tissue.40 However, studies showed transient harmful effects of [166Ho]FHMA treatment on immature rabbit cartilage. A higher rate of cartilage matrix synthesis in immature animals compared with adult animals41 might make the chondrocytes in young cartilage more susceptible to transient irradiation induced derangement of matrix production. Further studies are needed to determine whether cartilage in rheumatoid or osteoarthritic joints resembles young cartilage and might therefore be more susceptible to degenerative changes. Although some activation has been seen in cartilage matrix production in a transgenic mouse model for osteoarthritis,41 the mRNA levels of matrix components were much lower than in growing mice. Although [166Ho]FHMA has already been used in some clinical patients, further studies are needed to define the effects of [166Ho]FHMA treatment in the osteoarthritic joint of man.


We are grateful for the expert help of Eija Rahunen, Elma Sorsa (Department of Anatomy, Kuopio, Finland), and Merja Lakkisto (Department of Medical Biochemistry and Molecular Biology, University of Turku, Finland). The authors thank MS Pirkko Penttilä (MAP Medical Technologies OY, Tikkakoski, Finland) for preparing the radiopharmaceutical agent [166Ho]FHMA.

This study was supported by the MAP Medical Technologies OY and the Academy of Finland (project 52940).


View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.