Background Methotrexate (MTX) is a well established drug for the treatment of rheumatoid arthritis (RA). Pharmacological studies have shown, that most of the drug is eliminated by the kidneys. This results in a short plasma half-life and low concentrations of MTX in inflamed joints. In studies with tumour-bearing rats we showed, that due to the pathophysiological conditions in solid tumours and the high demand of tumour cells for amino acids and energy, albumin accumulates in tumours (up to 20% of the initially injected amount) and only a relatively small proportion is found in the liver (about 6%). Covalent coupling of MTX, fluorescent dyes and radioactive labels in a 1:1 molar ratio to albumin leads to conjugates which behave in-vivo like native albumin. In phase I clinical trials with tumour patients MTX-HSA showed promising results and relatively high dosages were well tolerated over several years. Like tumour cells, synovial cells of patients with RA show a high rate of proliferation and are in an active metabolic state. The plasma protein albumin might therefore be an appropriate carrier to increase drug concentrations in inflamed joints.

Objectives To proof if albumin is a suitable drug carrier to target MTX to inflamed joints in an RA animal model and isolated human RA synoviocytes.

Methods Laseroptic, scintigraphic and radiopharmacokinetic experiments with fluorescent and radioactive labelled albumin were performed to study the distribution of albumin in mice with and without CIA (n = 5 each point of the time course). The uptake of fluorescence labelled albumin by human RA synoviocytes was investigated by laser scanning microscopy in-vitro. The pharmacokinetics of MTX in arthritic mice was studied using 3H-labelled MTX. The therapeutic efficiency of MTX-HSA, MTX and a combination of both drugs (injected in different dosages 2×/week) was investigated by its ability to inhibit the development of arthritis (n = 3–35/group).

Results About 3% of the initially injected radioactive albumin was detected in each inflamed paw within a few hours after application, while in healthy paws only 0.3% of the tracer was found (accumulation by a factor of 6 in respect to the weight of the paws). Scintigraphic scans confirmed this results. Only 0.03% of 3H-MTX was measured in the inflamed paws. In the liver only 7% of the radiolabeled albumin but 27% of 3H-MTX was found. Illumination by laser light after injection of fluorescent labelled albumin showed strong fluorescence in arthritic but not in healthy joints of paws affected by CIA. The examination of human synoviocytes in-vitro showed catabolism of albumin in lysosomes. In equivalent dosages (7.5 mg MTX/kg twice a week) MTX-HSA was significantly (p = 0.017) superior to MTX in inhibiting the development of CIA (control: 68% arthritis incidence, MTX: 57%, MTX-HSA 28%). To achieve a comparable result as MTX-HSA almost a five times higher dosage of MTX (35 mg/kg) was necessary (37% arthr. inc.). Both drugs act synergistically (3.75 mg/kg MTX plus 0.8 mg/kg MTX-HSA: 29% arthr. inc.) in respect that 1.6 mg/kg MTX-HSA had no significant effect (57%).

Conclusion Using albumin as drug carrier allows to target drugs such as MTX specifically to inflamed joints. This might increase the therapeutic efficacy of MTX in human RA. Clinical trials to proof this concept are in preparation.