ORTHOPEDICS 2009; 32:893
December 2009, LINK

Anatomic Variance of Interfacet Distance and Its Relationship to Facet Arthrosis and Disk Degeneration in the Lumbar Spine

By Jason David Eubanks, MD; Jason O. Toy, BA; Patrick Messerschmitt, MD; Daniel R. Cooperman, MD; Nicholas U. Ahn, MD. Copyright ® 2009 SLACK Incorporated. All rights reserved.

Abstract

Inadequate separation of the L4-S1 facets predisposes individuals to spondylolytic defects. We proposed that decreased interfacet separation is a risk factor for increased lumbar degenerative disease. This study examined the correlation between interfacet distance and degenerative disease of the lumbar spine.

Four hundred forty-four cadaveric lumbar spines were examined for evidence of lumbar facet arthrosis and disk degeneration. Arthrosis at each level was graded from 0 to 4 on a continuum from no arthritis to complete ankylosis. These results were then examined in relation to interfacet spread. Interfacet distances were measured at each level (L1-S1). The difference in interfacet spread (L4-S1) was then correlated to facet arthrosis and disk degeneration. In individuals younger than 50 years (n=251), increased interfacet spread (L4-S1) was associated with less facet arthrosis at the L5/Sl level (P<.05). Similarly, in individuals younger than 40 years (n=149), increased interfacet spread (L4-S1) was associated with less disk degeneration at the L5/S1 level (P<.03).

Insufficient increase in interfacet distances (L4-S1) correlates with a greater risk of developing and maintaining spondylolytic defects. Similarly, this study suggests that increased interfacet spread (L4-S1) protects against early degenerative changes at the L5/Sl level. The more pyramidal the L4-S1 facet cascade, the lower the arthrosis. This effect appears early in the degenerative process before facet arthrosis and disk degeneration have become ubiquitous. In individuals younger than 50 years, increased interfacet distance correlates with less L5/Sl facet arthrosis.

Degenerative disease of the spinal elements is a ubiquitous finding in the aging lumbar spine, with 97% of individuals older than 49 years showing macroscopic evidence of disk degeneration1 and 100% of individuals older than 60 years showing evidence of facet arthrosis.2 While the reasons behind lumbar deterioration are many, it is clear that functional changes within the spinal unit’s 3-joint complex are integral in the degenerative process.3,4

It is also clear that the components of this functional spinal unit are uniquely oriented in the human species, with defining anatomical characteristics that distinguish humans from all other mammals.5 Some of these differences have been described in conjunction with spondylolysis6 and have been shown to correspond to the development and the maintenance of spondylolytic defects.5 More specifically, individuals with increasing interfacet spread from L4-S1 seem to be protected against spondylolysis, while those with smaller interfacet distances appear to be predisposed to the development and maintenance of spondylolytic defects.5

With this in mind, we examined the relationship between this uniquely human anatomy and the ubiquitous human problem of lumbar spine degenerative disease in an effort to help explain why some individuals may be predisposed to the early development of degenerative changes in the lumbar spine. It has been shown that facet orientation in the lumbar spine corresponds to arthrosis7; however, to our knowledge, the relationship between interfacet spread and degenerative changes in the lumbar spine has not been explored.

Understanding that the greatest compressive forces as well as most disk herniations are seen at the L5-S1 level1 and that interfacet distance is greatest from L4-S1,5 we explored the relationship between interfacet distance and degenerative disease of the lumbar spine. We hypothesized that just as increasing interfacet distance (L4-S1) is protective against developing and maintaining spondylolytic defects,5 so may increasing interfacet distance be protective in regard to degenerative changes, decreasing the incidence of vertebral rim osteophytosis, and facet osteophytosis at the lumbosacral junction.

Materials and Methods

Four hundred forty-four cadaveric lumbar spines from the Hamann-Todd Osteological Collection in Cleveland, Ohio, were randomly chosen for examination. The specimens in the collection are dried, disarticulated spines from >3100 individuals who died in Cleveland between the years of 1893 and 1938. The present study included 342 male and 102 female cadavers (Table), a ratio consistent with the collection’s overall gender ratio of approximately 4:1. In the present study, 193 samples were African-American with the remaining specimens white. All specimens were randomly chosen based on ease of access in the collection.

The gross specimens were examined morphologically and the interfacet distances were evaluated from L1-S1 using digital calipers to measure the distance from the center of the superior articular facets at each level, as described by Ward and Latimer.5 In keeping with their methodology,5 we similarly measured the vertebral body distances, measuring the maximum transverse breadth, to serve as a standardized control. Intraobserver error using this measuring method and these points of measurement has been previously reported to be <2%.5 We chose 10 specimens and performed similar intraobserver and interobserver calculations, with each specimen subjected to 3 measurements taken at the lowest 3 lumbar levels on separate occasions by 2 different examiners (J.D.E., P.M.). We found intraobserver, as well as interobserver, error to be <1%.

These interfacet distances were then correlated to the degree of vertebral rim osteophytosis and facet osteophytosis at the L5-S1 level. This information concerning degenerative disease was taken from previous work involving the same specimens.2,8 In this work,2,8 the degenerative disease was graded from no arthrosis to complete ankylosis on a scale from grade 0 to grade IV, as suggested by Kettler and Wilke9 using previously defined markers of degenerative disease.1,6 The grading scale was taken from previously published work2 that describes the osteophytic reaction in the facets as well as the peripheral rim osteophytes in the vertebral endplates. The scale follows:
  • Grade 0: normal facet joints or vertebral endplates.
  • Grade I: mild arthrosis, with evidence of osteophytic reaction involving up to 50% of the facet joint or vertebral endplates.
  • Grade II: moderate arthrosis, with evidence of osteophytic reaction involving 50% to 100% of facet joint or vertebral endplates.
  • Grade III: severe arthrosis, with evidence of osteophytic reaction involving 100% of the facet joints or vertebral endplates. Osteophytes are hypertrophic and bridging the joint space.
  • Grade IV: complete ankylosis.
Using a Spearman’s test, we correlated the difference in interfacet distances (L4-S1) with the maximum level of arthrosis at the L5-S1 level. This analysis was done for both vertebral rim osteophytosis and facet osteophytosis independently. A previous work8 established that the fifth decade appears to be a transition point in the degenerative process, where the changes in the disk begin to outpace those in the facet joints. For this reason, we focused our comparison for significance on individuals younger than 50 years (n=259) in the facet osteophytosis group and younger than 40 years (n=159) in the vertebral rim osteophytosis group.

Results

Data for the measured interfacet distances and vertebral body distances can be seen in Figure 1. It is evident in Figure 1A that the interfacet distances increase gradually from L1-S1. We have termed this incremental increase in interfacet distances a “cascading effect” to help illustrate this relationship. This cascading effect for interfacet distances results from the statistically significant (P<.001) differences in interfacet spread between successive lumbar levels, except L1-L2. For vertebral body distances, a statistically significant (P<.001) difference existed between all levels except L4 and Sl and L5 and S1.



Figure 1: Average interfacet distance L1-S1 (A). Average vertebral body distance L1-S1 (B).

While measurements were taken from L1-S1, the greatest difference in spread among multiple levels existed between L4-S1 (P<.05), as opposed to the L4-L5 or L5-S1 taken separately. This is in keeping with the evidence presented by Ward and Latimer,5 who found greater increase in interfacet distances from L4-S1 in normal spines than those with lytic defects. Therefore, we concentrated on these levels, the base of the pyramid, to calculate the difference in interfacet distances from L4 to S1 in all the specimens. These results can be seen graphically in Figure 2. This figure highlights the broad range of interfacet distances between specimens, extending from 1.06 to 21.9 mm, with the majority of specimens falling into the range of 5 mm to 1.5 cm, with a mean of 11.46 mm. After calculating the difference in interfacet spread from L4-S1, we then correlated these distances with previous findings2,8 concerning vertebral rim osteophytosis and facet osteophytosis.


Figure 2: Distribution of interfacet distances between L4-S1. Note the broad range from 1.6 to 22 mm.

We found that in individuals younger than 50 years (n=251), increased interfacet distance (L4-S1) was associated with less facet arthrosis at the L5/Sl level (P<.05). Similarly, in persons younger than 40 years (n=149), increased interfacet distance (L4-S1) correlated with less vertebral rim osteophytosis at the L5/S1 level (P<.03). Age was corrected for in all analyses performed. These results have been graphically displayed for facet osteophytosis in Figure 3. These differences did not hold true for the L4-L5 level, in which there was no significant difference in rates of degenerative disease (P=.146).



Figure 3: L5-S1 facet arthrosis vs interfacet distance in individuals younger than 50 years. Note the increased concentration of high-grade arthrosis at lower interfacet distances compared to Figure 3B, as well as the greater slope in the line of best fit (A). L5-S1 facet arthrosis vs interfacet distance in individuals younger than 50 years. There is a broad distribution of data points without the clear concentration seen in Figure 3A (B).

In both analyses, we corrected for the size of the vertebral bodies at L4 and L5 as well as for the height of the subject (available at time of death).

Furthermore, pilot analyses demonstrated that there was no significant relationship between interfacet distance and vertebral body size at the corresponding levels or subject height. Thus, we ensured that the significant association with facet osteophytosis and vertebral rim osteophytosis was not a result of the specimen or subject size but rather a direct result of the interfacet distance.

Pilot analyses also demonstrated that race and sex were not significantly associated with any of the measured parameters, including interfacet distances, vertebral rim osteophytosis, and facet osteophytosis, even when age was corrected for. Thus, these variables also did not appear to have a confounding effect on our results.

When comparing Figure 3B with Figure 3A, it appears that in individuals older than 50 years, there is a broader spread of data points, indicating the ubiquitous nature of arthrosis in this age category, irrespective of L4-S1 interfacet difference. In both Figures 3A and B, almost without exception, the highest grades of arthrosis are seen in individuals with smaller interfacet spreads (ie, <1.5 cm). Again, using a Spearman’s test, we found that smaller L4-S1 interfacet distances were correlated with decreased facet osteophytosis in individuals younger than 50 years (P<.05) at the L5-S1 level. This difference appears graphically in Figure 3A, in which a larger concentration of high grades of arthrosis at lower interfacet distances is seen.

Similar findings were encountered for vertebral rim osteophytosis. The preponderance of L5-S1 arthrosis appeared in smaller interfacet differences (<1.5 cm) and this was statistically significant in individuals younger than 40 years (P<.03). When attempting to quantify what specific interfacet spread predisposes an individual to early arthrosis, there was not a statistically significant difference between 1-mm increments to suggest that an interfacet distance of 1.3 cm is protective whereas an interfacet distance of 1.1 cm is not. However, in the case of facet osteophytosis, we noted a trend for the largest difference in maximum arthrosis when the cut-off for interfacet distances is <1.2 cm. Similarly, when vertebral rim osteophytosis was considered, an interfacet distance <1.5 cm appeared to display the largest difference. Clearly, with the average interfacet distances equal to 11.46 mm, individuals who have interfacet distances slightly greater than the mean (12-15 mm) appear to have a protective advantage in regard to the development of lumbar degenerative disease.

We found no positive association between interfacet distances from L4-S1 and facet osteophytosis or vertebral rim osteophytosis at L4/5 (P>.05). That a positive association was only identified at L5/S1 is likely a function of the fact that L5/S1 is subject to the greatest amount of stress and thus would likely display the most degenerative changes at an earlier age. It should also be noted that we found no significant association between interfacet distances at any level and corresponding facet osteophytosis or vertebral rim osteophytosis (P>.05). Rather, the association found was only with interfacet spread and degenerative changes at L5/S1.

Discussion

In their cadaveric study, Ward and Latimer5 postulated that perhaps interfacet spread might be related to spondylolysis. Examining 30 cadaveric specimens with spondylolysis, they were able to show a positive correlation between interfacet distance and spondylolysis—namely, increasing interfacet spread from L4-S1 appeared to be protective against the development and maintenance of spondylolytic defects.5

Using this paradigm, we explored the relationship between interfacet distances in this cascading configuration and the incidence of L5-S1 degenerative disease. While we believe the anatomic findings presented in this study are important to our understanding of lumbar degenerative disease, we recognize the limitations of this study. As earlier work detailed, this is a cadaveric study, performed on dried, disarticulated specimens. As such, it neglects symptoms the subjects may have experienced in addition to even earlier degenerative changes that may be evident with magnetic resonance imaging (MRI). However, while MRI may detect even earlier degenerative changes in the disks, it should not significantly affect the bony interfacet distance; therefore, in the absence of significant L5-S1 facet remodeling at young ages, our findings may be substantiated further in younger ages with more sophisticated imaging studies showing early soft tissue degeneration in those with small interfacet distances.

In his cadaveric study of spondylolysis, Nathan6 described spondylolytic defects as the resultant of a “pincer type” effect, in which the pars articularis of L5 was caught between the inferior articular process of the L4 facet and the superior articular process of the of the S1 facet. Over time, these superior and inferior facets cause thinning of the L5 isthmus or actual depressions (“prespondylolysis”) and eventual gaps (true spondylolytic defects) in the pars.6 What Nathan6 postulated as a “creeping fracture” we now believe to be a stress or fatigue fracture resulting from repetitive trauma on the pars.10

Ward and Latimer5 were able to show a positive association between this spondylolysis phenomenon and interfacet distance.5 Fujiwara et al7 has shown that facet orientation correlates with arthrosis in the lumbar elements. More specifically, increased sagittal orientation was observed in the L3-L4, L4-L5, and L5-S1 facets of nondegenerative spondylolisthesis individuals with higher grades of facet arthrosis. In individuals with spondylolisthesis, the same trend was seen at the L4-L5 and L5-S1 levels. Synthesizing the ideas presented by these previous studies, we postulated that increased interfacet distance may also affect facet osteophytosis, such that individuals with greater interfacet distances in the caudal portion of the lumbar spine may have lower rates of significant arthrosis.

Examining specimens from the same Hamann-Todd collection, previous reports demonstrated that facet osteophytosis and vertebral rim osteophytosis become ubiquitous by the seventh decade of life.8 Further, while the bony changes of facet arthrosis predominate in the early decades, the bony evidence of disk degeneration (vertebral rim osteophytosis) begins to outpace facet osteophytosis in the fifth decade.8 With this information, we focused our examination on individuals younger than the fifth decade of life to find that in individuals younger than 50 years, increased interfacet distance (L4-S1) was associated with less facet osteophytosis at the L5/Sl level. Similarly, when looking at vertebral rim osteophytosis, in individuals younger than 40 years, increased interfacet distance (L4-S1) was also associated with less vertebral rim osteophytosis at L5/S1. While a previous work2 showed L4/L5 to have the highest prevalence and degree of arthrosis, our study found no statistically significant (P>.146) association between interfacet distance and degenerative disease at the L4/L5 level, only the L5/S1 level (P<.05).

These results appear to suggest that individuals who have a smaller spread in the interfacet distance from L4-S1 are at an anatomical disadvantage for the development of early degenerative changes in the lumbar spine at the L5-S1 junction. Just as increasing interfacet distance (L4-S1) appears to spread the facets sufficiently to prevent the development and maintenance of spondylolytic defects,5 it also appears that increasing interfacet difference from L4-S1 appears to be associated with less early degenerative changes at the L5-S1 level. Perhaps the more pyramidal the configuration of the most caudal lumbar facets, the more efficient the dissipation of forces across the L5-S1 functional spinal unit and the lower the incidence of arthrosis. In other words, the greater the pyramidal configuration of the lumbar facet spread, the better the transfer of mechanical forces through the L5-S1 level into the sacrum and the less resultant junctional arthrosis. While finite element techniques have demonstrated that the facets of the lowest 2 lumbar levels experience and transmit significant loads,11,12 as of now we are unaware of any biomechanical studies specifically addressing interfacet distances and load transmission. Further biomechanical studies on this subject may be warranted.

Conclusion

Just as insufficient increase in interfacet distance (L4-S1) has been shown to correlate with a greater risk of developing and maintaining spondylolytic defects,5 so also does this study suggest that increased interfacet distance (L4-S1) correlates with degenerative changes at the L5/Sl level. The more pyramidal the L4-S1 facet configuration, the lower the arthrosis at the L5-S1 junction. This associated effect appears early in the degenerative process before facet osteophytosis and vertebral rim osteophytosis have become ubiquitous. In individuals younger than 50 years, increased interfacet distances is associated with less L5/Sl facet arthrosis. This effect appears to translate to the disk space as well, where individuals younger than 40 years with a larger interfacet spread display decreased vertebral rim osteophytosis at the L5/S1 level.

References
  1. Miller JA, Schmatz C, Schultz AB. Lumbar disc degeneration: correlation with age, sex, and spine level in 600 autopsy specimens. Spine (Phila Pa 1976). 1988; 13(2):173-178.
  2. Eubanks JD, Lee MJ, Cassinelli E, Ahn NU. Prevalence of lumbar facet arthrosis and its relationship to age, sex and race: an anatomic study of cadaveric specimens. Spine (Phila Pa 1976). 2007; 32(19):2058-2062.
  3. Borenstein D. Does osteoarthritis of the lumbar spine cause chronic low back pain? Curr Rheumatol Rep. 2004; 6(1):14-19.
  4. Kirkaldy-Willis WH, Farfan HF. Instability of the lumbar spine. Clin Orthop Relat Res. 1982; (165):110-123.
  5. Ward CV, Latimer B. Human evolution and the development of spondylolysis. Spine (Phila Pa 1976). 2005; 30(16):1808-1814.
  6. Nathan H. Spondylolysis: its anatomy and mechanism of development. J Bone Joint Surg Am. 1959; 41(2):303-320.
  7. Fujiwara A, Tamai K, An HS, et al. Orientation and osteoarthritis of the lumbar facet joint. Clin Orthop Relat Res. 2001; (385):88-94.
  8. Eubanks JD, Lee MJ, Cassinelli E, Ahn NU. Does lumbar facet arthrosis precede disc degeneration? A postmortem study. Clin Orthop Relat Res. 2007; (464):184-189.
  9. Kettler A, Wilke HJ. Review of existing grading systems for cervical or lumbar disc and facet joint degeneration. Eur Spine J. 2006; 15(6):705-718.
  10. Wiltse LL, Widell EH Jr, Jackson DW. Fatigue fracture: the basic lesion is inthmic spondylolisthesis. J Bone Joint Surg Am. 1975; 57(1):17-22.
  11. Goel VK, Kong W, Han JS, Weinstein JN, Gilbertson LG. A combined finite element and optimization investigation of lumbar spine mechanics with and without muscles. Spine (Phila Pa 1976). 1993; 18(11):1531-1541.
  12. Kong WZ, Goel VK, Gilbertson LG. Prediction of biomechanical parameters in the lumbar spine during static sagittal plane lifting. J Biomech Eng. 1998; 120(2):273-280.
Authors

Dr Eubanks is from the Department of Orthopedics, Case Western Reserve University School of Medicine, and Mr Toy and Drs Messerschmitt, Cooperman, and Ahn are from the Department of Orthopedics, Case Western Reserve University, Cleveland, Ohio.

Drs Eubanks, Messerschmitt, Cooperman, and Ahn and Mr Toy have no relevant financial relationships to disclose.

Correspondence should be addressed to: Jason David Eubanks, Department of Orthopedic Surgery, Case Western Reserve University School of Medicine, University Hospitals Case Medical Center, 11100 Euclid Ave, Cleveland, OH 44106 (jason.eubanks [at] uhhospitals [dot] org).