In vivo evaluation of temperature-responsive antimicrobial-loaded PNIPAAm hydrogels for prevention of surgical site infection

Abstract

Surgical site infections (SSIs) are a persistent clinical challenge. Local antimicrobial delivery may reduce the risk of SSI by increasing drug concentrations and distribution in vulnerable surgical sites compared to what is achieved using systemic antimicrobial prophylaxis alone. In this work, we describe a comprehensive in vivo evaluation of the safety and efficacy of poly(N-isopropylacrylamide-co-dimethylbutyrolactone acrylamide-co-Jeffamine M-1000 acrylamide) [PNDJ], an injectable temperature-responsive hydrogel carrier for antimicrobial delivery in surgical sites. Biodistribution data indicate that PNDJ is primarily cleared via the liver and kidneys following drug delivery. Antimicrobial-loaded PNDJ was generally well-tolerated locally and systemically when applied in bone, muscle, articulating joints, and intraperitoneal space, although mild renal toxicity consistent with the released antimicrobials was identified at high doses in rats. Dosing of PNDJ at bone-implant interfaces did not affect normal tissue healing and function of orthopedic implants in a transcortical plug model in rabbits and in canine total hip arthroplasty. Finally, PNDJ was effective at preventing recurrence of implant-associated MSSA and MRSA osteomyelitis in rabbits, showing a trend toward outperforming commercially available antimicrobial-loaded bone cement and systemic antimicrobial administration. These studies indicate that antimicrobial-loaded PNDJ hydrogels are well-tolerated and could reduce incidence of SSI in a variety of surgical procedures.

Keywords: NIPAAm polymers; hydrogel biocompatibility; local antimicrobial delivery; surgical site infection; sustained release.

Figure 1.

(a) Biodistribution of ¹⁴C-labeled PNDJ 70 days following intramuscular injection in rats. Differences between high and low molecular weight PNDJ were negligible. Data are shown as mean percent injected dose (%ID) + s.d. (HM_W: n = 5; LM_W: n = 6). (b) Systemic markers of nephrotoxicity (BUN) and hepatotoxicity (ALT) determined in a dose escalation study following subcutaneous injection of PNDJ with tobramycin and vancomycin in rats (serum chemistry data for creatinine and ALP in Supplementary Figure 4). Evidence of mild nephrotoxicity and hepatotoxicity was observed, particularly in the highest dose group. Data are shown as mean + s.d.; normal range values are provided for each marker;⁽²⁴⁾ significant differences (*) are reported where adjusted p < 0.05 (Groups 1–4: n = 6; Group 5: n = 4). (c) Pharmacokinetics of tobramycin and vancomycin measured for Groups 1–5 in the dose escalation study described in (b). Data are shown as mean ± s.d. (Groups 1–4: n = 6; Group 5: n = 4). (d) Representative histological images of kidneys from moderate and high dose groups in the dose escalation study. The central region of the sample from the moderate dose group (Group 2) is characterized by tubular degeneration, inflammatory infiltrates, and tubular regeneration. Tubules with vacuolated lining epithelium (arrows) surround the previously described area. In the high dose group (Group 5), the predominant feature is extensive tubular epithelial vacuolation (arrows). Small areas also feature more pronounced tubular degeneration (circled). Kidney sections stained with hematoxylin and eosin; scale bar = 100 μm. (e) Systemic markers of nephrotoxicity (BUN) and hepatotoxicity (AST) toxicity in canines (n = 3) receiving PNDJ with tobramycin and vancomycin (0.44 – 0.52 g/kg) during total hip arthroplasty (serum chemistry data for creatinine and ALT in Supplementary Figure 5). Serum concentrations of nephrotoxicity markers BUN and creatinine were not different or declined from pre-dose values. Hepatotoxicity marker AST showed a transient increase at 3 days and resolved to baseline by 7 days, while ALT declined from pre-dose levels. Data are shown as mean + s.d.; significant differences (*) are reported where adjusted p < 0.05 (n = 6). (f) Pharmacokinetics of tobramycin and vancomycin released from PNDJ dosed via intramuscular injection in the maximum feasible dose study described in (e). Data are shown as mean ± s.d. (n = 6).

Figure 2.

Representative histological sections of sites dosed with saline control (left column) or PNDJ gel (right column) with tobramycin and vancomycin in a surgical dead space site (SDS) in the quadriceps after 28 days (a, b) and 56 days (c, d), and in an intra-articular injection (IA) in the knee after 28 days (e, f). (a, b) Quadriceps skeletal muscle exhibits regions replaced by dense cellular granulation tissue (asterisks) with similar appearance for saline and PNDJ. A void indicating remaining PNDJ gel is visible in (b) in the bottom right of the image, along with a focal area of mineralization (arrow) which may contain a remnant of gel. (c, d) At 8 wk post injection, there is a decrease in granulation tissue (asterisks) within the muscle. In the site receiving gel, there was no remaining mineralized material. (e, f) Cartilage appears histologically normal at 28 days post injection for both saline and PNDJ groups. Skeletal muscle was stained with hematoxylin and eosin; articular cartilage was stained with Safranin-O and fast green counterstain; scale bars = 200 μm.

Figure 3.

PNDJ compatibility with bone ingrowth on titanium transcortical plugs. (a) The implantation site was prepared by drilling a pilot hole (3.5 mm diameter, 8 mm depth) in either the proximal tibia or distal femur of rabbits (1, 2). The pilot hole was filled with PNDJ containing gentamicin (3, 4). Titanium transcortical plugs (5) were then coated with PNDJ containing gentamicin so that the entire implant was covered, including the grooved ingrowth surfaces (6) prior to press-fitting the implant into the hole until flush with the bone (7, explanted distal femur). (b) Macroscopic sections of tibia and femur, demonstrating normal bone healing and apposition onto implant surfaces in areas in contact with cortical bone for samples treated with PNDJ gel with high concentration gentamicin (3.1 wt%), PNDJ gel with low concentration gentamicin (0.3 wt%), as well as a PBS negative control 8 weeks after implantation. Tissue was stained with toluidine blue; images were obtained at normal magnification. High-resolution composite micrographs of each sample are provided in Supplementary Figure 7.

Figure 4.

(a) PNDJ hydrogels loaded with antimicrobials dosed during total hip arthroplasty (THA) in dogs. Following acetabular and femoral reaming, the cavities were filled with PNDJ gel containing tobramycin and vancomycin. Implants were then coated with gel and press-fit in place using standard technique. At wound closure, gel was injected through a catheter placed within the joint capsule to fill the wound space. (b) Representative radiographs in one animal at 2, 6, and 12 weeks after implantation demonstrating stable press fit implants without subsidence or osteolysis adjacent to the implants. Plastic embedded histological sections of the tissue ingrowth surfaces on the (c) acetabular cup and the (d) femoral stem demonstrate dense fibrous tissue (blue) and bone (pink) throughout the ingrowth surfaces. Bone is present where expected at points of cortical contact. (e) The interface between bone (B) and implant (I) demonstrates complete infiltration of collagen producing fibroblasts (single arrows). Focal calcification also occurred within this zone. Cartilaginous metaplasia (C) was noted in one area in the gap between bone and implant. Where the implant was in direct contact with cortical bone, few or no fibroblasts were seen. Formation of new bone with osteoblastic rimming was noted (double arrows). None of the sections presented any evidence of infiltration of acute inflammatory cells (neutrophils) or mononuclear giant cells commonly associated with foreign body reaction. Histological sections were stained with Stevenel’s blue and van Gieson stain; images are shown at 40X magnification. Full slide images of three sections from each animal are shown in Supplementary Figure 8.

Figure 5.

Efficacy of locally applied PNDJ with tobramycin against S. aureus infections. (a) Osteomyelitis biofilm-based infection prior to debridement. (b) Implantation of low-dose antimicrobial-loaded bone cement with tobramycin and a metal implant. (c) Injection of PNDJ with tobramycin alongside a metal implant after debridement of the SSI. (d) PNDJ with tobramycin eliminated all MSSA (6/6) and MRSA (6/6) osteomyelitis infections. Antimicrobial loaded bone cement (ALBC) with tobramycin and systemic tobramycin were only evaluated against MSSA infections (n = 7 each group).