Topical delivery of PD-1 inhibitors with laser-assisted passive diffusion and active intradermal injection: Investigation of cutaneous pharmacokinetics and biodistribution patterns

Background and Objectives Current cancer immunotherapeutic treatment with PD-1 inhibitors is administered systemically. However, a local treatment strategy may be advantageous as it could provide targeted drug delivery as well as attenuate side effects seen with systemic treatments. For keratinocyte cancers, where surgical excision is not always applicable, an alternate local treatment approach would be beneficial. This study aims to examine cutaneous pharmacokinetics and biodistribution of the PD-1 inhibitor nivolumab, locally delivered either by ablative fractional laser (AFL)-assisted passive diffusion or active intradermal injection, in vivo. Materials and Methods In vivo pig skin was either exposed to CO2 AFL (80 mJ/mb by two stacked pulses of 40 mJ/mb) at 5% or 15% density followed by topical application of nivolumab (1 mg/ml, 100 mu l/10 x 10 mm) or intradermally injected with nivolumab (1 mg/ml, 100 mu l). Cutaneous nivolumab delivery was evaluated at different timepoints (0, 1, 2, 4 hours and 2 days) at two tissue depths (100-800 and 900-1600 mu m) by ELISA. Visualization of cutaneous biodistribution was shown in vertical tissue sections using HiLyte Fluor(TM) 488 SE labeled nivolumab for fluorescence microscopy whereas nivolumab was DOTA-tagged with Dysprosium before the laser ablation-inductively coupled plasma-mass spectrometry analysis (LA-ICP-MS). Results Our in vivo study revealed different pharmacokinetic and biodistribution patterns for the AFL- and injection techniques. A superficial horizontal band-like uptake of nivolumab was provided with AFL-assisted passive diffusion whereas a deep focal deposition was seen with active intradermal injection, compared with controls showing remnant deposition on the skin surface. AFL-assisted nivolumab uptake in upper dermis peaked after 4 hours (p < 0.01). The cutaneous concentration of nivolumab achieved by intradermal injection was markedly higher than with AFL, the highest deposition with intradermal injection was detected at time 0 hours in both upper and deep dermis (p < 0.01) and decreased throughout the study period, although the concentration remained higher compared with saline control injections at all time points (0 hours -2 d) (p < 0.01). Conclusion Local cutaneous delivery of nivolumab with either AFL or intradermal injection revealed two different pharmacokinetic and biodistribution patterns. Passive AFL-assisted diffusion of nivolumab resulted in enhanced uptake after 4 hours, while intradermal actively injected nivolumab showed immediate enhanced cutaneous deposition with retention up to 2 days after injection. The two local delivery techniques show potential for development of individualized treatment strategies depending on the clinical tumor appearance.