Osteophyte Formation in the Lumber Spine and Relevance to Low Back Pain (Pathogenesis) Part 2

Osteophyte and gene polymorphism

There were no significant differences in the disc score, osteophyte score or the ratio of group B in all polymorphisms, though Arg/Arg polymorphism in ADH2 tended to be less frequent. (p=0.051) (Table 3) Results of the logistic regression model to select gene polymorphism factors associated with the presence of osteophyte formation without disc height narrowing were shown in Table 4. In the polymorphism of alcohol dehydrogenase

	Genotypes			P value
	His/His (n=212)	Arg/His (n=124)	Arg/Arg (n=19)
Disc score	1.33 ± 1.56	1.42 ± 1.47	1.30 ± 1.83	0.812
Osteophyte score	8.12 ± 3.12	7.91 ± 3.05	6.75 ± 2.90	0.160
No. of Group A+C/B	147/65	98/26	17/2	0.051

Group A=the cases with osteophyte formation with disc height narrowing Group B=the cases with osteophyte formation without disc height narrowing Group C=the cases with no osteophyte formation

Table 3. Disc score and Osteophyte score for ADH2 polymorphysm

(ADH2; Arg47His), the prevalence of osteophyte formation without disc height narrowing (group B) was less in His/Arg (OR=0.57, 95%CI=0.33-0.97,p=0.041) and Arg/Arg (OR=0.41, 95%CI=0.1-1.5,p=0.18) than His/His. In the other polymorphisms, there were no significant differences in osteophyte formation without disc height narrowing.

	GroupA+C (n=262)	Group B (n=93)	Genotypes (Arg/Arg, Arg/His, His/His)
	Alelle Frequency (Arg, His)		Arg/Arg	Arg/His versus His/His	Arg/Arg + Arg/His
ADH2 Arg His	0.25 0.75	0.16 0.84
P value			0.183	0.035	0.027
Odds ratio (95% CI)			0.41 (0.10-1.50)	0.57 (0.33-0.97)	0.55 (0.32-0.93)

Allele frequencies were estimated by the gene counting method. P values were adjusted for age, gender, BMI, body fat ratio, bone stiffness, smoking habit and alcohol consumption.

Table 4. Multivariate logoistic regression analysis of the effect of ADH2polymorphism on the prevalence of osteophyte without disc height narrowing

LBP and osteophyte

The prevalence of LBP was 40.4% (156 cases) with average VAS scale of 34.9+28.5 (10-100). Back muscle strength was significantly lower in the LBP group than in the non-LBP group. (p<0.05) Disc score was significantly higher in the LBP group than in the non-LBP group (p<0.01), whereas there was no significant difference in the osteophyte score between the two groups. (Table 5) Characteristics of the groups A, B, and C were shown in Table 6. In group C, male subjects, Brinkman index and drinkers were significantly fewer than in group A and the group B.

	LBP group (n=156)	non-LBP group (n=224)
Age (years)	68.7 + 6.8	67.7 + 5.9
Gender (male/female)	59/97	93/131
Height (cm)	154.0 + 8.3	155.5 + 8.0
Body weight (kg)	58.0 +9.0	58.6 + 9.5
BMI (kg/m2)	24.3 + 3.0	24.2 +3.0
Body fat ratio (%)	28.2 +7.2	27.4 +6.9
Brinkman index	277.3 + 475.4	226.6 +425.3
Drinker (no. (%))	41(26.2)	53 (23.7)
VAS	34.9 + 28.5*	3.5 + 8.0*
Back muscle strength (kg)	63.6 + 24.2**	72.6 + 28.4**
Bone stiffness	97.0 + 18.6	96.7 + 21.0
Disc score	1.72 + 1.71*	1.09 + 1.33*
Osteophyte score	8.28 + 3.10	7.71 + 3.11

BMI = body mass index, VAS = visual analogue scale, * p<0.01, **p<0.05 Table 5. Characteristics of the LBP group and the non-LBP group

Although vacuum phenomenon was more frequent in the group A (p<0.01), the presence of vertebral fracture and degenerative spondylolisthesis were equivalent to the group B and C. Disc score was significantly higher in the group A than in the group C. (p<0.01) Osteophyet score was significantly higher in the group A than in the group B. (p<0.05) (Fig.4) Both VAS scale and the prevalence of LBP were significantly greater in group A than group B and group C. In group B, VAS scale and numbers of LBP were equivalent to those in group B, but significantly less than those in group A. (Fig. 5)

Fig 4. Disc score and Osteophyte score in the group A, B and C.

	Group A (n=217)	Group B (n=99)	Group C (n=71)
Age (years)	68.9+ 6.2	67.8+ 6.2	65.6+6.6
Gender (male/female)	98/119	48/51	7/64 **
Height (cm)	155.0 + 8.2	156.5+ 8.8	151.9+6.1
Body weight (kg)	58.5 + 9.4	60.2 +9.9	55.2+7.2
BMI (kg/m2)	24.3 + 3.1	24.5 + 3.0	23.9+2.8
Body fat ratio (%)	27.0 + 7.4	27.3 + 6.2	30.7+5.8
Brinkman index	232.2 + 427.1	384.1 + 532.8	95.6+276.0 "***
Drinker (no. (%))	52 (30.0)	35 (35.3)	8 (11.3) *"
Back muscle strength (kg)	70.5 + 28.6	71.6 + 28.5	60.5 + 17.3
Bone stiffness	98.5 + 19.6	95.8 + 20.5	93.4 + 20.1
Vertebral fracture (%)	3.26	3.06	2.85
Vacuum phenomenon (%)	45.53 **	7.37	4.35
Degenerative spondylolisthesis	21.96	18.37	28.17

Group A=the cases with osteophyte formation with disc height narrowing Group B=the cases with osteophyte formation without disc height narrowing, Group C=the cases with no osteophyte formation * p<0.01 vs Group A; ** p<0.01 vs Group B; *** p<0.05 vs Group A

Table 6. Characteristics of the groups A, B, and C

Fig. 5. Visual analogue scale and low back pain in the group A, B and C.

Discussion

It is commonly recognized that the degenerative changes that occur in the intervertebral discs are the point of departure of osteophyte formation. During the degeneration process, the discs undergo progressive structural changes in the form of dehydration of the nucleus and disintegration of the annulus fibrosus resulting in decreased disc height (Buckwalter et al., 1995), and lead to an increase in the compression stiffness and reduction in disc fiber strain (Kim et al., 1991). Biomechanically, the nucleus has lost some of its proteoglycan and water contents and increased its collagen content (Andersson, 1998). With progressive matrix alterations of the nucleus, changes in disc morphology such as a reduction in disc height become visible in plane radiographs. Degenerative changes within may result in an alteration of its mechanical properties, increased flexibility and decreased disc height, which in turn contribute to changes in the local stress within the disc (An et al., 2004). There is a general agreement that changes induced by aging lead to alternations in the thickness of the disc, but some differences are seen in the account of the effect of aging on the thickness of the lumbar disc. Vernon-Roberts et al. stressed that reduction of the disc height with age is inevitable (Vernon-Roberts et al., 1977), however, an increase in disc height with age has been reported. (Twomery et al., 1987, Amonoo-Kuofi, 1991, Roberts et al., 1997) Shao et al. demonstrated that the vertebral endplates became more concave with age, resulting the lumbar disc height increase (Shao et al., 2002). The effect of aging on the disc height has not been well understood, and the term of disc degeneration is imprecisely defined.

On the other hands, osteophytes form as a specific tissue reaction to these stresses and strains (Bick, 1995), and are attributed to higher stress more frequently anteriorly than posteriorly. Schmorl et al. formulated a pathogenic hypothesis that as a result of tears in the attachment of the annulus fibrosus into the margibnal ring of the vertebral body, the nucleus protrudes forward against the anterior longitudinal ligament. The increased strain causes the formation of spurs in the area of its attachment to the periosteum covering the cortex of the vertebral bodies (Schmorl et al., 1932). (Schmorl’s rim lesion theory) Colins formulated a theory of osteophyte formation that associates degeneration of the entire intervertebral disc (collapsed disc) and the resultant anterior protrusion with subsequent osteophyte formation. The protrusion of the disc lifts the periosteum lateral to anterior longitudinal ligament and stimulates new subperiosteal bone (Collons, 1949). (Collins’ bulging disc theory) According to Macnab’s theory, osteophytes form as a result of instability between adjacent vertebral bodies (Macnab, 1971)(Macnab’s instability theory). Traction spur, which projects horizontally and never curve toward the disc, differentiated from claw osteophytes (Nathan et al., 1962). Nathan concluded that osteophytes form as a natural physiologic response to compressive loads, serving to stabilize the spine (Nathan et al., 1962). In any case, there is wide agreement about the close association of disc degeneration with osteophyte formation and precedence in disc degeneration over vertebral deformities (Nathan et al., 1962, Vernon-Roberts et al., 1977, Lipson et al, 1980, Milgram, 1982). Our study date results provide further evidence substantiating that osteophyte formation and disc height narrowing are not always closely correlated, as identified by the prevalence of osteophytes without disc height narrowing in about 30% and the lack of correlation between disc height reduction and osteophytes. This finding stresses that these two features of spinal "degenerative" changes represent different factors affecting lumbar spine and the potential for osteophyte formation caused by factors other than spinal degeneration. Oishi et al. showed that intervertebral disc degeneration and osteophyte formation of the vertebral bodies represented different factors affecting the lumbar spine in postmenopausal Japanese women; however, the difference of osteophytes with or without disc degeneration was not mentioned. There are no detailed studies concerning osteophytes not accompanied with disc degeneration. We considered it informative to investigate the features of such osteophytes that are often observed clinically (Oishi et al., 2003).

There are few epidemiological data about osteophyte formation on the lumbar spine compared to the number of studies about osteoarthritis of the knee and hip. Nathan reported with regard to the frequency and degree of development of anterior osteophytes that the prevalence of osteophytes was greater in whites of both sexes than in Negroes with no statistically significant differences, with the frequency being much higher in the males in both races (Nathan et al., 1962). Our results revealed a significant influence of gender, smoking and alcohol consumption on osteophyte formation irrespective of the presence of disc height narrowing, however, showed no epidemiological differences between osteophyte formation with disc height narrowing and without narrowing, namely the differences in osteophytes depending on intervertebral disc degeneration. The present study illustrated that the prevalence of LBP in group B was significantly lower than in group A, and this suggests that lumbar disc narrowing may have a propensity for LBP, indicating osteophytes may prevent the clinical manifestation of pain. While data from many studies suggest an association between LBP and osteophyte formation (Frymoyer et al., 1984, Biering-Sorensen et al., 1985, Symmons et al., 1991, Pye et al., 2004), several studies indicate that osteophyte formation do not have an independent association with LBP (van Tulder et al., 1997, O’Neill et al., 1999, Schepper et al. 2004). Whether the stabilization of osteophytes or low frequency of disc degeneration decreased the prevalence of LBP is not clear. However, when osteophyte formation occurs before disc degeneration advances as a physiologic response to stabilize the spine, LBP may be evitable. While many studies have focused on LBP in relation to lumbar disc degeneration (Parkkola et al., 1993, Paajanen et al., 1997, Luoma et al., 2000, Jarvik et al., 2001, Videman et al., 2003), there are no reports regarding the association of LBP with osteophytes without lumbar disc degeneration.

Discal degeneration is generally considered as the primary source of LBP. In addition to nociceptive nerve fibers in the annulus and nucleus that can be sensitized by the cytokines and neuropeptides present in the degenerated disc, other sources of notiception can be found in the spinal unit including muscles, ligaments and facet joints (Freemont et al., 1997, Benoist, 2003). Nociception coming from these various tissues makes it difficult to distinguish from osteophytes in spinal pain. Thus, the cause of the decreased LBP should not be determined to be osteophyte formation before disc degeneration, although, it would be intriguing to investigate the genetic predisposition in cases with osteophytes without disc degeneration.

Several studies on factors associated with genetic susceptibility to spinal osteophyte formation, such as VDR (Videman et al., 2001, Jordan et al., 2005) and TGFB1 (Yamada et al., 2000) referred to osteophytes with spinal degeneration. Our results did not show any relationship between these polymorphisms and osteophyte formation without disc degeneration. Alcohol dehydrogenase P subunit is an enzyme that converts ethanol to acetaldehyde, whose gene, ADH2 located in 4q22, has a functional polymorphism Arg47His (Matsuo et al., 1989). Both ADH2 and ALDH2 (aldehyde dehydrogenase 2) are polymorphic, and genetic polymorphisms have been shown to functionally affect alcohol detoxification. The enzyme activity is higher in the 47His allele (ADH2*2) than in the 47Arg allele (ADH2*1) (Yin et al., 1984), and the former leads to a higher rate of oxidation of ethanol, resulting in an arginine/histidine exchange in the protein. In particular, an association of the 47His allele with flushing has been reported (Takeshita et al., 1996), and results of a number of studies seem to indicate that the 47His allele protects against alcohol abuse and alcoholism in Asians (Muramatsu et al., 1995, Shen et al., 1997) and Caucasians (Whitfield et al., 1998, Borras et al., 2000). In Japanese, the incidence of 47Arg allele is low, different from Caucasians (Sherman et al., DI, 1993, Higuchi et al., 1994). On the other hand, most alcoholics exhibit radiographic evidence of osteopenia (Bilke et al., 1985), leading to a hypothesis that reduced osteoblast activity resulting in underfilling of resorptive lacunae is primary responsible for alcohol-induced bone loss (Turner et al., 2001). Ethanol has been shown to increase bone resorption (Callaci et al., 2004) and to decrease trabecular bone volume (Rico et al., 1987). Additionally, administration of ethanol to healthy volunteers results in an acute decrease in serum osteocalcin levels (Rico et al., 1987, Nielsen et al., 1990). The present study demonstrated that carriers of 47Arg allele might suppress osteophyte formation unaffected by intervertebral disc degeneration, and this could be supported by these studies showing that ethanol contributes to decreased bone formation. Further research will be required to investigate the osteophyte development and molecular characterization, however, our study would encourage further studies on the mechanisms underlying osteophyte formation.

Conclusion

Osteophyte formation of the lumbar spine without disc degeneration was investigated, and estimated the implications of osteophytes from the viewpoint of LBP and gene polymorphism. The 47His polymorphism in the ADH2 may act to suppress osteophyte formation unaffected by disc degeneration. The subjects with osteophyte development preceding intervertebral disc degeneration had a lower risk of LBP compared with those without osteophytes.