We investigated the therapeutic efficacy of recombinant human interleukin-2 (rhIL-2)-loaded, in situ gelling, physically crosslinked dextran hydrogels, locally applied to SL2 lymphoma in mice. The physical crosslinking was established by stereocomplex formation between d-lactic acid oligomers and l-lactic acid oligomers grafted separately to dextrans. The stereocomplex hydrogel as described in our manuscript has several favourable characteristics, which enables its use as system for the controlled release of pharmaceutically active proteins. Firstly, the hydrogel system is a physically crosslinked system. In physically crosslinked gels, the use of chemical crosslinking agents is avoided. Such agents can potentially inactivate the protein and can covalently link the protein to the hydrogel network. Secondly, the hydrogel formation takes place at room temperature and physiological pH, and, importantly, in an all-aqueous environment. All factors are important to preserve the three-dimensional structure, and thus the biological activity, of the protein to be entrapped and released from the gels. Thirdly, the gel formation does not occur instantaneously. This means that a liquid formulation can be injected which solidifies after injection (in situ gel formation is possible). Fourthly, no pH drop during degradation is expected during degradation. As a control, free rhIL-2 was administered locally in either a single injection or at five consecutive days. All mice received the same total dose of rhIL-2. The rhIL-2-loaded hydrogels released most IL-2 over a period of about 5 days. The biocompatibility and biodegradability of the gels were excellent, as there were no acute or chronic inflammatory reaction and as the gels were replaced completely by fibroblasts after 15 days. The therapeutic efficacy of rhIL-2-loaded in situ gelled hydrogels is very good, as was demonstrated in DBA/2 mice bearing SL2. The therapeutic effect of a single application of gels loaded with 1 x 10(6) IU rhIL-2 is at least comparable to the therapeutic effect of injection of an equal dose of free rhIL-2. All mice cured with rhIL-2-loaded hydrogels survived a subsequent challenge, rejecting 10(6) intraperitoneal (i.p.) injected SL2 cells. In conclusion, this study demonstrates that in situ gelling, physically crosslinked dextran hydrogels slowly release encapsulated rhIL-2 in such a way that it is intact and biologically and therapeutically active. These hydrogels may greatly enhance the clinical applicability of rhIL-2 immunotherapy as only a single treatment is required and as these hydrogels are completely biodegradable.