Abstract Poor oral bioavailability limits the use of many chemopreventives in the prevention and treatment of cancer. We previously reported a novel concept in which polycaprolactone (PCL) implants embedded with test agents provide sustained delivery for long duration (months to >1year), reduce effective dose substantially and increase bioavailability. While tested successfully for various agents, this formulation results in an initial burst release and does not apply to heat-labile compounds. Furthermore, it takes 2 or more years for the polymeric implant to biodegrade. To overcome these limitations, we hypothesized that formulation of low mol wt polymeric implants will provide more sustained and higher release compared with higher mol. wt polymers; the low mol. wt polymers are also expected to biodegrade in weeks to months. To test part of this hypothesis, we used three different mol wt PCL polymers and two co-polymers, poly(D, L-lactide-co-glycolide) (PLGA). The method involves i) preparation of blank PCL implants (1.4 mm dia), and ii) coating of 30-40 layers by dipping blank implants, with intermittent drying, in 10-20% of different mol wt PCL solutions in dichloromethane containing 10% curcumin in tetrahydrofuran. The curcumin implants when tested for in vitro release showed that i) the burst release effect and the amount released varied with mol wt of the polymer. Implants prepared with PCL of 3,600 mol wt (PCL-3.6K) did not show any burst release phenomenon. Instead, these implants took nearly 7 days to render maximum daily release and thereafter the rate of release declined gradually over a period of two weeks. Implants prepared from higher mol wt polymers (PCL-31K and PCL-112K) were accompanied with some initial burst release and the rate of release declined gradually. Cumulative release measured over two weeks was found to be inversely proportional to the mol wt of PCL: PCL-3.6K (30%) > PCL-31K (24%) > PCL-112K (21%). On the other hand, the co-polymers, PLGA (50:50) and PLGA (75:25) showed only 1% and 0.5% total release, respectively. To determine if the implant delivery reduced the effective dose in a tumor model, nude mice were inoculated with human lung cancer (A549) cells and then treated with coated implants of withaferin A, a potent triterpenoid isolated from the traditional Ayurvedic herb “Ashwagandha”. Withaferin A given via the coated implants significantly inhibited (>50%) human lung cancer (A549) xenograft, while it was ineffective when the same dose was administered i.p. Together, our data suggest that coated implants of low mol wt PCL can accommodate heat-labile compounds, enhance bioavailability, furnish more sustained and higher cumulative release, and elicits anti-tumor activity. Our ultimate goal is to use low mol polymeric implants that will biodegrade in few months and provide continuous release of the drug. (Supported from CA-118114, KLCRP and Duggan Endowment). Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 2883. doi:1538-7445.AM2012-2883