SEARCH

SEARCH BY CITATION

Keywords:

  • ankylosing spondylitis;
  • osteitis condensans ilii;
  • sacroiliac joint

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. REFERENCES

Aim:  To determine whether subjects with radiological evidence of osteitis condensans ilii exhibit symptoms and signs in common with sacroiliitis when compared with an age-matched control group and to examine demographic features.

Methods:  The Dunedin Hospital radiology database was searched for all subjects with changes of osteitis condensans ilii over a 10-year period. An age-matched control group with plain X-rays of the pelvis was recruited from the same database. All subjects were sent a questionnaire enquiring about back pain and details of previous pregnancies. Those who responded to the questionnaire were invited for clinical assessment.

Results:  Thirty-five individuals with osteitis condensans ilii were identified over the 10-year period. All were female and reported prior pregnancy supporting an association between osteitis condensans ilii and pregnancy. Stress testing of the sacroiliac joints was associated with greater tenderness in the osteitis condensans ilii group with a mean of 1.8 positive tests out of a possible 4, compared to 0.8 in the control group (Wilcoxon rank-sum test P = 0.02). Comparison between the two groups showed no difference in number of pregnancies, newborn weight, presence of back pain, back pain assessed by the Oswestry Low Back Pain Questionnaire or loss of function using the Bath Ankylosing Spondylitis Functional Index (BASFI).

Conclusion:  Osteitis condensans ilii is associated with tenderness during sacroiliac joint compression tests and should be considered in the differential diagnosis when sacroiliac joint tenderness is elicited.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. REFERENCES

Osteitis condensans ilii (OCI) is an uncommon incidental finding on plain X-ray, characterized by sclerosis of predominantly the iliac bone adjacent to an otherwise normal sacroiliac joint.1 First described in 1926, OCI is seen almost exclusively in women. It is usually bilateral and symmetrical. Early reports suggested pregnancy and trauma as causative mechanisms.2 It is generally regarded as a self-limiting phenomenon and is mostly asymptomatic.3

Radiologists recognize the importance of differentiating OCI from sacroiliitis, but little is known about the clinical features that might be associated with such a dramatic X-ray appearance (Figs 1,2). In particular, to our knowledge, there have been no clinical studies comparing symptoms in individuals with OCI and a control population.

image

Figure 1. X-rays from a patient with bilateral osteitis condensans ilii.

Download figure to PowerPoint

image

Figure 2. X-rays from a patient with ankylosing spondylitis.

Download figure to PowerPoint

The aim of our study was to investigate clinical, radiological and demographic features in a cohort of individuals identified with OCI and compare these with an age-matched control group who had undergone X-rays of the pelvis for a variety of reasons and were identified as having normal sacroiliac joints.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. REFERENCES

Ethical approval was obtained from the Lower South Regional Ethics Committee. The study conforms to the provisions of the World Medical Association's Declaration of Helsinki.

The Dunedin Hospital radiology database was searched for all reports containing ‘osteitis condensans ilii’ over a 10-year period from 1997–2007. The X-rays were reviewed by a single musculoskeletal radiologist to confirm the presence of OCI. The radiology database was then searched to find age-matched controls with normal sacroiliac joints using a 2 : 1 ratio of controls to cases. With 20 cases, proportions in this group could be estimated to within 0.20 (worst case) using 95% confidence intervals and differences of 0.4 between 20 cases and 40 controls could be detected using a significance level of 0.05 with 80% power. Cases and controls were sent a two-page questionnaire enquiring about back pain, stiffness, treatment and details of previous pregnancies. Given previous suggestions that OCI may relate to pregnancy and labour we enquired after number of pregnancies, gestation, birth weight, presentation, instrumentation and complications of labour.

All those who responded to the questionnaire were invited for clinical assessment. The Modified Oswestry Low Back Pain Disability Questionnaire was chosen as a well validated instrument to asses low back pain.4 Subjects who denied any back pain were assigned a score of ‘0’ and did not complete the questionnaire. The Bath Ankylosing Spondylitis Functional Index (BASFI) was used as a measure of functional impairment.5 This was chosen in view of the similarity of radiological changes in both OCI and ankylosing spondylitis. We felt the BASFI had face validity in assessing function in individuals with OCI. Although specifically designed for assessing function in individuals with ankylosing spondylitis, activities relating to sacroiliac and lower back disability are predominant, including ability to put on socks, stand unsupported, climb stairs and undertake physically demanding activities. A 100-mm visual analogue pain scale (VAS) with the descriptor ‘how has your back pain been in the last week?’ and the anchors ‘very good’ and ‘very bad’ at either end was also administered.

A blinded assessor examined both cases and controls for sacroiliac joint (SIJ) tenderness on stress testing using four widely accepted pain provocation techniques. These included direct pressure on the anterior superior iliac spines (distraction test), direct pressure over the lower sacrum with the subject prone (sacral thrust test), compression of the pelvis with the patient lying on one side (compression) and forced flexion of one hip maximally towards the opposite shoulder with hyperextension of the contralateral hip joint.3 Schober's test, height and weight were also recorded.

Clinical assessment for SIJ tenderness, although commonly performed in clinical practice, has been shown to be of low reliability.6,7 In order to demonstrate the presence or absence of SIJ tenderness we employed a combination of provocation tests. This multi-test regimen has been found to improve clinical accuracy in the demonstration of SIJ pain in a recently published study using magnetic resonance imaging (MRI) evidence of sacroiliitis as a gold standard.8 A systematic review of the literature recommended five techniques which had been found to have sensitivity and specificity of greater than 60% in at least one study. These included the distraction test, compression, sacral thrust, resisted hip abduction and thigh thrust/posterior shear.7

All X-rays were reviewed by the radiologist to confirm the presence of OCI, and to assess symmetry of X-ray findings.

As the number of controls recruited was less than anticipated, a second cohort of controls were approached in the same way. Due to the poor response rate among controls, matching was broken and groups were compared using the Wilcoxon rank-sum (Mann-Whitney two-sample) test except for the proportion with at least one positive SIJ stress test, where an exact test for proportions was used.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. REFERENCES

Over the 10-year period selected, we identified 35 individuals with OCI. Fifteen of the 35 with OCI and 17 of 90 controls contacted responded to the questionnaire (Table 1). Thirteen from each group agreed to a further clinical assessment. All subjects with OCI were female and recorded at least one previous pregnancy. Stress testing of the SIJs was associated with greater tenderness in the OCI group with a mean of 1.8 positive tests compared to 0.8 in the control group (Wilcoxon rank-sum test P = 0.02). SIJ tenderness was present in 12/13 subjects with OCI, compared with 6/13 controls. There was a significant association between OCI and any SIJ tenderness using Fisher's exact test (P = 0.03).

Table 1.  Clinical and demographic characteristics of subjects
Characteristic:OCI Mean or percentage (SD)Controls Mean or percentage (SD)P-value
  1. BMI, body mass index; VAS, visual analogue scale; BASFI, Bath Ankylosing Spondylitis Functional Index; SIJ, sacroiliac joint.

Age (years)40.8 (11.5)44.3 (16.3)0.71
n = 15n = 17 
Female100%100% 
BMI31.2 (6.3)27.0 (4.1)0.05
n = 13n = 13 
Number of pregnancies 2.4 (0.9) 2.5 (1.3)0.86
n = 15n = 17 
Oswestry Low Back Pain Disability Questionnaire18.3 (15.1)15.1 (13.8)0.47
n = 13n = 13 
VAS pain scale (cm) 4.0 (2.8) 3.5 (2.8)0.55
n = 13n = 13 
BASFI 2.7 (1.8) 3.5 (2.0)0.41
n = 13n = 13 
Schober's text (cm) 4.5 (1.1) 5.2 (3.8)0.29
n = 13n = 13 
SIJ stress testing (number positive stress tests per subject out of 4) 1.8 (1.0) 0.8 (1.0)0.02
n = 13n = 13 
At least one positive SIJ stress test12 60.03
n = 13n = 13 

There was a tendency to higher BMI in the cases compared to the controls (mean 31.2 cf. 27.0 P = 0.05). However, there was no evidence that BMI was associated with SIJ stress testing scores among the 26 participants (Spearman's rho = 0.17, P = 0.41), so this does not explain the group differences.

There was no difference between the two groups with respect to number of pregnancies, birth weight, presence of back pain, back pain assessed by the Oswestry Low Back Pain Questionnaire or disability using the BASFI (Table 1). An analysis of age demographics between responders and non-responders showed no statistically significant difference, with responders 4.8 years older on average (95% CI – 0.5–10.1, P = 0.08). Radiologic review showed that 26/34 (76.5%) had bilateral changes of osteitis condensans ilii (OCI) with 14/26 having symmetrical changes; 8/34 (23.5%) had only unilateral evidence of OCI.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. REFERENCES

We have demonstrated that subjects with OCI are more likely to have SIJ tenderness than controls. Chronic back pain and disability were not associated with OCI. All patients with OCI reported prior pregnancy; however there was no difference in recorded pregnancy-associated variables between the cases and controls.

The small number of individuals identified with definite radiographic changes consistent with OCI was surprising in view of the estimated prevalence of 1.6–3.0%.9,10 There was a low response rate in both cases and controls. Although this low level of response could introduce responder bias, there was no significant difference between responders and non-responders in terms of age (P = 0.08). A slightly higher age for responders was noted (4.8 years) which is a common finding in postal surveys. The similar age demographics in responders and non-responders reduces the likelihood that response was a significant source of bias when comparing the cases and controls. Recall bias in relation to pregnancy-related questions is possible, but again this is unlikely to affect comparisons between groups or responders versus non-responders.

Previous studies of OCI have found a high prevalence of back pain among subjects with OCI, but have made no comparison with a control group. Hare and Haggart published the first clinical series of OCI in 1945 and described 23 cases. All were female, but not all had a history of pregnancy.11 All subjects reported back pain with 8/23 recalling the onset of symptoms as just prior to or following delivery. They reported a positive ‘sacroiliac test’ in all cases, but the method for this examination was not described.

A further report published in 1950 reviewed 100 women with OCI, of whom 22 were nulliparous.12 Chronic low-back pain was commonly reported, with symptoms exacerbated by activity and relieved by rest in all cases. Forty-eight patients identified pregnancy with the onset of their symptoms and others reported onset of symptoms following rapid weight gain or excessive physical fatigue. Interestingly, unlike our cohort, tests for SIJ tenderness, including stress testing, were negative. They found 64 cases to be overweight. Histology of the lesion, from those who had undergone surgery, was also described in this paper. They described focal areas of fibrosis within the bone marrow and evidence of increased osteoblastic activity. They also reported an increase in the density of trabeculations. Similar histological findings were reported in three patients by Rojko et al. in 1959.13

In 1971 Numaguchi reported a series of 48 cases of OCI.9 All were female and 44 of 48 (91.6%) recorded prior pregnancy. Bilateral involvement was seen in the majority (87.5%), as with our group (76.5%). Low back pain was present in 26 subjects, 17 of whom had alternative pathology seen on X-ray which may have explained their symptoms. Tenderness on palpation of the sacroiliac region was present in the minority of cases (8/48: 16.7%). For 29 subjects, follow-up X-rays were available, and partial or complete resolution of OCI changes was documented in 22/29 (75.9%).

We hypothesize that ligamentous laxity of the sacroiliac joints during pregnancy may lead to instability of the joint which persists in some individuals, leading to sclerosis and tenderness on stressing of the joint. This may persist for many years after pregnancy, but is not associated with chronic pain or disability. It is therefore important to appreciate that OCI should be included in the differential diagnosis when SIJ tenderness is elicited.

Further research in this area would be valuable. Prospective studies in OCI are limited by the large number of participants needed and the risk of X-ray exposure to the mother and fetus. A larger-scale longitudinal study would help to reveal the natural history of OCI. Performing MRI on OCI subjects would be of interest to determine whether bone edema or soft tissue involvement is a feature, and would limit radiation exposure to the pelvic organs.

ACKNOWLEDGMENTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. REFERENCES

We would like to thank Debra McNamara for her help with contacting the participants and administering the questionnaires.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. REFERENCES
  • 1
    Sicard JA, Gally L, Haguenau J (1926) Osteites condensantes, a etiologie inconnue. J Radiol D’electrol 10, 5037.
  • 2
    Berent Fritz (1933) Beitrage zur pathologie der kreuzdarmbeinfugen. Arch F Orthop U Unfall-Chir 32, 6426.
  • 3
    Muhammad Asim Khan (2003) Clinical features of ankylosing spondylitis. In: HochbergMC, SilmanAJ, SmolenSJ, WeinblattME, WeismanMH, (eds) Rheumatology, 3rd edn, pp. 1177. Mosby.
  • 4
    Fritz JM, Irrgang JJ (2001) A comparison of a modified Oswestry low back pain disability questionnaire and the Quebec back pain disability scale. Phys Ther 81, 7768.
  • 5
    Calin A, Garrett S, Whitelock H, et al . (1994) A new approach to defining functional ability in ankylosing spondylitis: the development of the bath ankylosing spondylitis functional index. J Rheumatol 21, 22815.
  • 6
    Van der Wurff P, Meyne W, Hagmeijer RHM (2000) Clinical tests of the sacroiliac joint: a systematic methodological review. Part 2: validity. Manual Ther 5, 8996.
  • 7
    Stuber KJ (2007) Specificity, sensitivity, and predictive values of clinical tests of the sacroiliac joint: a systematic review of the literature. J Can Chiropr Assoc 51, 3041.
  • 8
    Ozgocmen S, Bozgeyik Z, Kalcik M, Yildirim A (2008) The value of sacroiliac pain provocation tests in early active sacroiliitis. Clin Rheumatol 27, 127582.
  • 9
    Numaguchi Y (1971) Osteitis condensans ilii, including its resolution. Radiology 98, 18.
  • 10
    Dihlmann W (1962) Seltaene Beckenveranderungen bei der Spondylitis ankylopoetica (Morbus Bechterew). Fortschr Rontgenstr 97, 10911.
  • 11
    Hare HF, Haggart GE (1945) Ostseitis condensans ilii. JAMA 128, 7238.
  • 12
    Shipp FL, Haggart GE (1950) Further experience in the management of osteitis condensans ilii. J Bone Joint Surg 32, 8417.
  • 13
    Rojko A, Farkas K (1959) Osteitis condensans ossis ilii. Acta Orthopaedica 29, 10820.