Objective—To evaluate the use of serum concentrations
of biochemical markers of bone metabolism
(osteocalcin [OC], bone-specific alkaline phosphatase
[BS-ALP], and deoxypyridinoline [DPYR]) to compare
healing in infected versus noninfected fractures and in
fractures with normal repair versus delayed (nonunion)
repair in rabbits.
Animals—32 female 9- to 10-month-old New Zealand
Procedure—A femoral fracture defect was made in
each rabbit. Rabbits were assigned to the following
groups: the bone morphogenetic-2 gene treatment
group with either noninfected nonunion or infected (ie,
inoculation of defects with Staphylococcus aureus)
nonunion fractures or the luciferase (control) gene
treatment group with either noninfected nonunion or
infected nonunion fractures. Serum samples were
obtained before surgery (time 0) and 4, 8, 12, and 16
weeks after surgery. Callus formation and lysis grades
were evaluated radiographically at 16 weeks.
Results—Serum OC and BS-ALP concentrations
decreased from time 0 at 4 weeks, peaked at 8
weeks, and then decreased. Serum DPYR concentration
peaked at 4 weeks and then decreased, independent
of gene treatment group or fracture infection
status. Compared with rabbits with noninfected fractures,
those with infected fractures had lower serum
OC and BS-ALP concentrations at 4 weeks, higher
serum OC concentrations at 16 weeks, and higher
serum DPYR concentrations at 4, 8, and 16 weeks.
Combined serum OC, BS-ALP, and DPYR concentrations
provided an accuracy of 96% for prediction of
fracture infection status at 4 weeks.
Conclusions and Clinical Relevance—Measurement
of multiple serum biochemical markers of bone
metabolism could be useful for clinical evaluation of
fracture healing and early diagnosis of osteomyelitis.
( J Am Vet Med Assoc 2003;64:727–735)