The structure of the pars interarticularis of the lower lumbar vertebrae and its relation to the etiology of spondylolysis, with a report of a healing fracture in the neural arch of a fourth lumbar vertebra.

Defects in the pars interarticularis are rarely present at birth, and seldom appear clinically before the age at which a child can stand upright. The incidence then increases up to the third or fourth decade (Rowe and Roche 1953, Stewart 1953). In radiological surveys of normal industrial and military populations, spondylolysis has been reported in up to 76 per cent (Friedman, Fischer and van Demark 1946 ; Moreton 1969 ; Hector 1972), though the incidence may be as much as 45 per cent in certain races and families (Hasebe 1913, Bakke 1931, Willis 1931, Friberg 1939, Roche and Rowe 1951, Stewart 1953, Baker and McHoffick 1956, Nathan 1959, Wiltse 1962, Berquet 1965, Francillon and Schreiber 1966, Lester and Shapiro 1968, Kettelkamp and Wright 1971). There are therefore genetic factors which may predispose an individual to the defect. But spondylolysis is not a congenital lesion, nor are developmental errors considered relevant to the etiology (Batts 1939, Mutch and Walmesley 1956). The precipitating factor is now believed to be a mechanical one, fatigue or impact failure leading to fracture of the pars interarticularis on one or both sides (Roberts 1947, Unander-Scharin 1950, Anderson 1956, Nathan 1959, Newman 1959, de Palma and Marone 1959, Newman 1963, Murray and Coiwill 1968, Arct 1971). Despite the growing body ofevidence in support of the theory that spondylolysis follows a fracture, there are notably few accounts of the normal anatomy of the region affected. The pathological anatomy of spondylolysis has been fully described (Willis 1931 ; Junghanns 1933; Raney 1945 ; Roche 1949 ; Gill, Manning and White 1955 ; Bosworth, Fielding, Demarest and Bonaquist 1955 ; Nathan 1959 ; Sullivan and Bickell 1960 ; Wiltse 1962). The pars at the site of the defect “may be of deep triangular shape or wide and flat” (Willis 1931); and Wiltse (1962) referred to cortical bone in the region of spondylolysis but did not define its extent. Gallois and Japiot (1925) radiographed sections ofwhole vertebrae and described the trabecular structure in each region of the spine, though in the lumbar neural arch they showed little of the pars, only reporting bony fasciculi at the superior margins of the laminae and between the transverse processes. SchlUter (1960) referred to cortical bone on the inferior and superior margins ofthe pedicles but did not describe its extent into the pars. Lethe (1962) also described cortical bone in the neural arch and demonstrated the direction of osteons histologically. It was therefore decided to study the bony structure of the neural arch and to consider the forces it withstands in the course of daily activities and occasional exertions. Accordingly, specimens of the fourth and fifth lumbar vertebrae were obtained at necropsy from seven subjects aged seventeen to sixty-seven (see Table I) in order to study the normal structure.