Ann Nucl Med (2014) 28:227–231 DOI 10.1007/s12149-013-0799-0
ORIGINAL ARTICLE
Quantitative sacroiliac scintigraphy for pediatric patients: comparison of two methods Murat Fani Bozkurt • Pinar Kiratli
Received: 15 May 2013 / Accepted: 14 December 2013 / Published online: 25 December 2013 Ó The Japanese Society of Nuclear Medicine 2013
Abstract Objective Quantitative bone scintigraphy is a useful method to diagnose sacroiliitis. However, there is significant overlap between healthy and pathological sacroiliac index (SI) values for adult patients, while there are no such sufficient data for children. This study was aimed mainly to assess normal SI values in different age groups of pediatric patients using 2 different quantitative methods. Materials and methods Normally reported bone scans of 79 children were retrospectively reviewed. Two different methods were used for quantitation. For the first method, sacrum was used as a background site while L5 vertebra was used instead for the second method. Right/left SI values of both methods were compared with each other in relation with gender and different age groups, as group 1 (1–5 years), group 2 (6–10 years), group 3 (11–15 years) and group 4 (16–18 years). Additional comparison was made with a group of young-adult population of 21–30 years old as group 5 to assess the effect of age. Results Gender-based comparison yielded significantly higher SI for females for the first method, while no significant difference existed for the second one. Significant increase in SI with both methods was found as age increased. Significantly lower SI was found from the second method, when similar age groups like group 1–2 or group 2–3 were compared with each other, while no such difference existed for the first method. For each individual patient from any age group, method-based comparison resulted in a significantly different SI with both methods.
M. F. Bozkurt (&) P. Kiratli Department of Nuclear Medicine, Hacettepe University Faculty of Medicine, 06100 Ankara, Turkey e-mail: [email protected]; [email protected]
Conclusions In pediatric population, SI tends to increase as age increases. Quantitation method using sacrum as background yields significantly higher SI for female gender. Alternative use of L5 as background site for quantitation performs well in children. Since two methods resulted in significantly different SI, individualized cut-off values for each age group for any method are practically warranted. Keywords Gender
Children Sacroiliitis Scintigraphy Age
Introduction Sacroiliitis is a common manifestation of spondyloarthropathies which are mostly accepted as a spectrum of different disease types originating from undifferentiated spondyloarthropathy to well-established ankylosing spondylitis [1]. Although the most common onset age of ankylosing spondylitis is midtwenties, a variable percentage of patients start their symptoms at childhood and adolescent ages [1, 2]. Therefore, the early diagnosis of this disease is very important to begin appropriate treatment. The clinical diagnosis of sacroiliitis may be difficult to confirm especially in the early stages of disease. Sacroiliac needle biopsy seems to be a very sensitive and specific diagnostic tool to diagnose early stages of sacroiliitis [3]. However, it is not widely preferred in routine practice since it is an invasive procedure. Direct radiography of sacroiliac joints is mostly normal at early disease stages. Along with magnetic resonance imaging, sacroiliac scintigraphy serves as a useful non-invasive diagnostic tool to diagnose active sacroiliitis, especially at early stages of disease, in which direct radiography is usually normal [2, 4, 5].
123
228
Sacroiliac scintigraphy, as being a functional imaging modality, can provide a metabolic information for the disease process. Scintigraphy with bone radiopharmaceuticals shows the osteoblastic activity within the sacroiliac joints and the degree of osteoblastic activity uptake correlates well with the disease activity, even if there is no or subtle morphological changes assessed by radiological methods. In the scintigraphic evaluation of sacroiliac joints, quantitative analysis is commonly preferred along with visual analysis. Quantitative analysis of scintigraphic uptake consists of comparison of uptake ratios of both sacroiliac joints to any selected adjacent background structure, mostly sacrum among other sites used so far [6– 10]. The advantage of adding quantitative analysis is that it provides a more objective perspective to visual interpretation. However, significant overlap between healthy and pathological sacroiliac index values (SI) has been reported for adult patients [6, 8, 11–13]. Especially young adults have higher values for SI even if they do not suffer from sacroiliitis, mostly due to their growing pattern [13]. Therefore, standard SI values which are obtained from the population without any sacroiliac disease are absolutely needed to set a threshold to diagnose diseased individuals. Such studies are present for adult population in the literature [6, 8–13]. However, there is no sufficient data on this particular subject for pediatric patients to the best of our knowledge, in spite of some published studies on sacroiliitis of mostly pyogenic origin in pediatric patients [14–16]. This study was aimed to assess normal SI in different age groups of pediatric patients using 2 different quantitative methods.
Ann Nucl Med (2014) 28:227–231
adjusted dose for children according to EANM dosage card [17]. The dose of MDP for young-adult patients was 20 mCi. None of the patients had any findings or symptoms such as back pain or joint stiffness suggesting sacroiliitis, besides none had any systemic disease like SLE or DM. All the scans were reported as normal and all patients’ data were analyzed with two different methods for quantitative analysis as follows: First method Rectangular regions-of-interests (ROIs) were drawn over the right and left sacroiliac joint as well as the sacrum which represented the background activity (Fig. 1). The sacrum ROIs were placed in the center of the sacrum, to keep the standard among patients. Sacroiliac activity index (SI) was calculated as the ratio between the counts from the right and left sacroiliac joints divided by sacrum for each. All ROIs were drawn and semiquantitative analysis was done by the same nuclear medicine physician (MFB). Second method The second method included ROIs over L5 vertebra instead of the sacrum as the background activity (Fig. 1). Vertebral ROIs were placed centrally. SI was calculated as the ratio between the counts from the right/left joints divided by the counts from L5 vertebra. All ROIs were drawn and semiquantitative analysis was done by the same nuclear medicine physician (MFB).
Materials and methods A total of 79 bone scans of pediatric patients (30 females, 49 males, mean age 11 ± 6 years, age range 3–18 years) who were referred for various entities other than sacroiliitis in a period of 25 months were retrospectively reviewed to obtain semiquantitative analysis of sacroiliac joints. In addition, a total of 20 normally reported bone scintigraphy of young-adult patients (11 females, 9 males, mean age 25 ± 3 years, age range 21–30 years) without any referral for sacroiliitis were retrospectively reviewed for comparison with pediatric patients. The scintigraphic imaging included spot pelvis anterior and posterior planar images (10 % energy window set at a peak of 140 keV of Tc-99m, 256 9 256 9 256 matrix, 500 kcount preset count) obtained with a dual-headed gamma camera equipped with low-energy high-resolution collimators (Siemens E-cam, Erlangen Germany) at 3–4 h following intravenous injection of Tc-99m-methylene diphosphonate (MDP) with an
123
Fig. 1 Placement of the ROIs over right (ROI0) and left sacroiliac joints (ROI1) as well as sacrum (ROI2) and L5 vertebra (ROI3) on posterior planar image. ROI2 is used for the first quantitation method where ROI3 is used for the second one
Ann Nucl Med (2014) 28:227–231
229
Comparative analysis of the first and second method The right and left SI ratios with background sacrum counts (Rs and Ls) calculated according to the first method were compared with those calculated from L5 as background counts (Rv and Lv) from the second method in relation with gender. Comparative analysis of the two methods was also done for different age groups of children to assess the influence of age. For this analysis, the patients were grouped according to their ages as years; group 1 (n = 19, age 1–5 years), group 2 (n = 21, age 6–10 years), group 3 (n = 21, age 11–15 years) and group 4 (n = 18, age 16–18 years). To assess the influence of age from a different perspective, comparative analysis of the two methods was also performed for young-adult group of patients as group 5 (n = 20, age 21–30 years). Statistical analysis Descriptive statistics are provided as means and standard deviations or as medians and ranges. To compare two groups with each other, Student’s t test was used. When variables could not be assumed to be normally distributed Mann–Whitney rank sum test was used instead. A P \ 0.05 was considered to be significant.
Results Side-wise comparison of quantitative sacroiliac scintigraphy of a total of 79 children (30 girls, mean age 11 ± 6 years, age range 3–18 years) showed no significant difference between mean values of right SI ratios for the first (Rs) and the second method (Rv) along with no significant difference between left SI ratios for the first (Ls) and second method (Lv) (Table 1). However, when the same side mean values of SI ratios from two different
methods were compared to each other, there was statistically significant difference (Table 2). Gender-based comparison of the two methods yielded significantly higher SI ratios for females for the first method (Rs 1.33 ± 0.26 vs 1.50 ± 0.32, P = 0.01; Ls 1.33 ± 0.28 vs 1.49 ± 0.33, P = 0.02) while no such significant difference existed for the second method (Table 3). Age-based comparative analysis of the two methods revealed a significant increase in absolute values for SI ratios with both methods as age increased, on comparison with distant age groups (Table 4). Age-based analysis depending on the specific age groups showed that there were significantly lower absolute SI values with the second method, when similar consecutive age groups like group 1–2 or group 2–3 were compared with each other, while no such difference with the first method was found (Table 4). Age-based analysis of the two methods showed insignificant results when SI values of older children (group 4) were compared with those of young-adult population (group 5), although there was significant difference between SI ratios of younger children (group 1) and those of young-adult (group 5) (Table 4). For each individual patient from any of the age groups, method-based comparison resulted in a significantly different SI with both methods when mutually one-to-one comparison of each absolute SI value was performed (P \ 0.01).
Discussion Bone scintigraphy serves as a useful diagnostic imaging method to detect sacroiliitis. Quantitative sacroiliac scintigraphy can increase the diagnostic capability of the method and provide a more objective perspective to interpretation of bone scans [6, 8–12]. However, despite
Table 1 Side-wise comparison of mean SI values for all patients for the first and second methods SI
First method
P value
Rs
Ls
1.23 ± 0.26
1.25 ± 0.25
0.66
Second method
P value
Rv
Lv
1.14 ± 0.21
1.13 ± 0.20
0.89
Table 2 Comparison of same side SI values according to two different methods SI
Right-side values
P value
Rs
Rv
1.23 ± 0.26
1.14 ± 0.21
0.02
Left-side values
P value
Ls
Lv
1.25 ± 0.25
1.13 ± 0.20
0.01
123
230
Ann Nucl Med (2014) 28:227–231
Table 3 Comparison of mean SI values for each gender and for each method SI
Male
Female
P*
Rs
1.33 ± 0.26
1.50 ± 0.32
0.01
Rv
1.22 ± 0.21
1.26 ± 0.25
0.48
Ls
1.33 ± 0.28
1.49 ± 0.33
0.02
Lv
1.23 ± 0.23
1.24 ± 0.28
0.64
* Difference between genders
quantitative analysis, there is still significant discrepancy on the fact that bone scintigraphy can accurately differentiate active sacroiliitis from normal sacroiliac joints. The major drawback of the method is that SI ratios may show a wide range and are highly variable, especially in young adulthood. A study by Lin and Wang [13] concluded that SI ratios were higher in males younger than 30 years old and that those ratios tend to decrease by age at both genders. To the best of our knowledge, there is yet no such study on pediatric patients in which the influence of age and gender has been assessed. Most of the published data on sacroiliitis in pediatric patients are related to pyogenic sacroiliitis and spondyloarthropathies [14–16]. Therefore, this study was firstly aimed to assess whether there is any effect of age to SI ratios in a group of pediatric patients without any symptoms and findings suggesting sacroiliitis and, secondly, the use of an alternative quantitation method was assessed in comparison with the conventional SI quantitation method. This alternative method included L5 vertebral ROI instead of sacrum, which is the most common background site for quantitative sacroiliac scintigraphy. Although there is no published data related to disadvantages of sacrum as a background site in the literature, in routine practice, prominent sacral promontorium may cause underestimation of SI ratios because of relatively increased sacral activity at least for some patients. Thus, L5 vertebra was chosen as an alternative background site for the second quantitation method. Lumbal vertebrae are among the most common bony structures in which osteodegenerative changes are frequently seen. This issue
may theoretically present as a disadvantage of this method for adult patients, taking into account that osteodegenerative changes are not expected to be seen at pediatric age. Age-based comparison of the methods showed in contrast to previous study by Lin and Wang [13] that SI ratios were significantly higher for girls compared to boys, for the first method while no effect of gender was found for the second method. Another interesting result from the present study is that SI ratios by two methods were totally different from each other when mutual one-to-one correspondence was analyzed. This means that setting a particular threshold for each quantitative method needs to be done separately and that the same method—whether the first or the second— has to be used at the follow-up studies. Although the absolute values from two methods were totally different from each other, SI ratios tended to increase as age increases independently from the method applied. It can be hypothesized that SI values continue to increase as the children grow up and get mature, presumably until the age of 30 taking into account the previous studies although related data are limited [11, 13]. The results of the current study showed that age-based analysis depending on both methods came up with significantly higher SI values at young-adult population (group 5) on comparison with younger children (group 1). This result may also support the hypothesis that SI ratios tend to increase as age increases and it is independent from the method used. Different age groups of pediatric patients yielded different absolute SI values. However, analysis between similar (consecutive) age groups showed that SI ratios differed significantly when L5 was chosen as background site, in contrast with sacrum of the first method where no such difference was observed. Depending on the current study, as SI ratios tend to increase by age in parallel with growing and maturation, this finding may contribute to the fact that the second method may represent the normal osteoblastic metabolism slightly more precisely compared to the first method. However, its clinical significance needs to be verified in further studies.
Table 4 Comparison of mean SI values for all age groups (group 1, 2, 3, 4 and 5) in relation with 2 quantitation methods SI
Group 1
Group 2
Group 3
Group 4
Group 5
P*
P**
P***
Rs
1.16 ± 0.26
1.42 ± 0.26
1.41 ± 0.23
1.51 ± 0.23
1.50 ± 0.29
0.94
\0.01
0.87
Rv
1.04 ± 0.11
1.11 ± 0.16
1.25 ± 0.14
1.23 ± 0.14
1.29 ± 0.36
0.04
\0.01
0.22
Ls
1.15 ± 0.25
1.43 ± 0.25
1.39 ± 0.26
1.55 ± 0.26
1.52 ± 0.32
0.67
\0.01
0.79
Lv
1.03 ± 0.10
1.12 ± 0.10
1.22 ± 0.16
1.26 ± 0.16
1.28 ± 0.33
0.04
\0.01
0.89
* Difference between group 2 and 3 ** Difference between group 1 and 5 *** Difference between group 4 and 5
123
Ann Nucl Med (2014) 28:227–231
Another interesting result from the current study is that SI ratios of females were higher than those of males for the first method even though there was no such significant difference between genders for the second method. In our opinion, the fact that SI ratios for females were higher than those for males for the first method while no such significant difference was present for the second method may possibly be attributed to difference of anatomy between two genders. Although this is only a speculative explanation, we think that the relatively more prominent gluteal muscular and/or fat tissue of female gender may possibly cause more attenuation of counts from the sacrum. When L5 vertebra is used as a reference center instead of sacrum, such attenuation effect from the gluteal region of female gender might be expected to be lower compared to that of sacrum, depending on the anatomic localizaton and relation of sacrum and L5 vertebra with regard to gluteal soft tissue. However, more studies are surely warranted to analyze the possible causative factors between two genders.
2. 3.
4.
5. 6.
7.
8.
9.
10.
Conclusion The choice of L5 vertebra as an alternative to sacrum as background ROI site performs well for pediatric patients and can be preferred in case there is prominent sacral promontorium activity which may possibly hamper quantitation resulting in false interpretation. Regardless of the method used, SI ratios of children tend to increase as their ages increase. Thus, it may not seem possible to obtain a single SI ratio threshold value to differentiate sacroiliitis from normal metabolic activity in pediatric patients. It would be reasonable for each center to generate their own threshold values for different age groups of children. Conflict of interest
231
11.
12.
13. 14.
15.
None declared. 16.
References 1. Burgos-Vargas R. The assessment of the spondyloarthiritis international society concept and criteria for the classification of axial spondyloarthritis and peripheral spondyloarthritis:A critical appraisal for the pediatric rheumatologist. Pediatric Rheumatol
17.
Online J. 2012;31:10(1):14. http://www.ped-rheum.com/content/ 10/1/14. Braun J, van der Heijde D. Imaging and scoring in sacroiliitis. Best Pract Res Clin Rheumatol. 2002;16(4):573–604. Gong Y, Zheng N, Chen SB, Xiao ZY, Wu MY, Liu Y, et al. Ten years’ experience with needle biopsy in the early diagnosis of sacroiliitis. Arthritis Rheum. 2012;64(5):1399–406. doi:10.1002/ art.33453. Inanc N, Atagu¨ndu¨z P, Sen F, Biren T, Turog˘lu HT, Direskeneli H. The investigation of sacroiliitis with different imaging techniques in spondyloarthropathies. Rheumatol Int. 2005;25(8):591–4 (Epub 2004). Grigoryan M, Roemer FW, Mohr A, Genant HK. Imaging in sacroiliitis. Curr Rheumatol Rep. 2004;6(2):102–9. Bozkurt MF, Ug˘ur O, Ertenli I, Caner B. Combined use of bone and bone marrow scintigraphies for the diagnosis of active sacroiliitis. Ann Nucl Med. 2001;15(2):117–21. Fogelman I, Maisey MN, Clarke SEM. Sacroiliitis. In: Fogelman I, Maisey MN, Clarke SEM, editors. An atlas of clinical nuclear medicine. 2nd ed. Singapore: Martin Dunitz Ltd; 1994. p. 84. Ryan PJ, Fogelman I. Sacroiliac quantitative bone scintigraphy in ankylosing spondylitis:any clinical relevance? Nucl Med Commun. 1993;14:19–20. Davis MC, Turner MA, Charters JR, Golden HE, Ali A, Fordham EW. Quantitative sacroiliac scintigraphy:the effect of method of selection of region of interest. Clin Nucl Med. 1984;9:334–40. Huanga JY, Tsaib MF, Kaoc PF, Chena YS. Automatic computer-aided sacroiliac joint index analysis for bone scintigraphy. Comput Methods Programs Biomed. 2010;98(1):15–26. doi:10. 1016/j.cmpb.2009.07.010 (Epub 20). Ayres J, Hilson AJ, Maisey MR, Laurent R, Panayi GS, Saunders AJ. An improved method for sacroiliac joint imaging:a study of normal subjects, patients with sacroiliitis and patients with lowback pain. Clin Radiol. 1981;32:441–5. Prakash S, Gopinath PG, Bhargava S, Mehra NK, Malaviya AN. Evaluation of quantitative sacroiliac scintigraphy for the early detection of sacroiliitis. Eur J Nucl Med. 1983;8:531–4. Lin WY, Wang SJ. Influence of age and gender on quantitative sacroiliac joint scintigraphy. J Nucl Med. 1998;39:1269–72. Wu MS, Chang SS, Lee SH, Lee CC. Pyogenic sacroiliitis—a comparison between paediatric and adult patients. Rheumatology. 2007;46:1684–7. Taylor ZW, Ryan DD, Ross LA. Increased incidence of sacroiliac joint infection at a children’s hospital. J Pediatr Orthop. 2010;30(8):893–8. Stoll ML, Bhore R, Dempsey-Robertson M, Punaro M. Spondyloarthritis in pediatric population: risk factors for sacroiliitis. J Rheumatol. 2010;37(11):2402–8. doi:10.3899/jrheum.100014 (Epub 2010 Aug 3). Lassman M, Biassoni L, Monsieurs M, Franzius C, Jacobs F. EANM dosimetry and paediatrics committees. The new EANM paediatric dosage card. Eur J Nucl Med Mol Imaging. 2009;36(3):540–1.
123
ORIGINAL ARTICLE
Quantitative sacroiliac scintigraphy for pediatric patients: comparison of two methods Murat Fani Bozkurt • Pinar Kiratli
Received: 15 May 2013 / Accepted: 14 December 2013 / Published online: 25 December 2013 Ó The Japanese Society of Nuclear Medicine 2013
Abstract Objective Quantitative bone scintigraphy is a useful method to diagnose sacroiliitis. However, there is significant overlap between healthy and pathological sacroiliac index (SI) values for adult patients, while there are no such sufficient data for children. This study was aimed mainly to assess normal SI values in different age groups of pediatric patients using 2 different quantitative methods. Materials and methods Normally reported bone scans of 79 children were retrospectively reviewed. Two different methods were used for quantitation. For the first method, sacrum was used as a background site while L5 vertebra was used instead for the second method. Right/left SI values of both methods were compared with each other in relation with gender and different age groups, as group 1 (1–5 years), group 2 (6–10 years), group 3 (11–15 years) and group 4 (16–18 years). Additional comparison was made with a group of young-adult population of 21–30 years old as group 5 to assess the effect of age. Results Gender-based comparison yielded significantly higher SI for females for the first method, while no significant difference existed for the second one. Significant increase in SI with both methods was found as age increased. Significantly lower SI was found from the second method, when similar age groups like group 1–2 or group 2–3 were compared with each other, while no such difference existed for the first method. For each individual patient from any age group, method-based comparison resulted in a significantly different SI with both methods.
M. F. Bozkurt (&) P. Kiratli Department of Nuclear Medicine, Hacettepe University Faculty of Medicine, 06100 Ankara, Turkey e-mail: [email protected]; [email protected]
Conclusions In pediatric population, SI tends to increase as age increases. Quantitation method using sacrum as background yields significantly higher SI for female gender. Alternative use of L5 as background site for quantitation performs well in children. Since two methods resulted in significantly different SI, individualized cut-off values for each age group for any method are practically warranted. Keywords Gender
Children Sacroiliitis Scintigraphy Age
Introduction Sacroiliitis is a common manifestation of spondyloarthropathies which are mostly accepted as a spectrum of different disease types originating from undifferentiated spondyloarthropathy to well-established ankylosing spondylitis [1]. Although the most common onset age of ankylosing spondylitis is midtwenties, a variable percentage of patients start their symptoms at childhood and adolescent ages [1, 2]. Therefore, the early diagnosis of this disease is very important to begin appropriate treatment. The clinical diagnosis of sacroiliitis may be difficult to confirm especially in the early stages of disease. Sacroiliac needle biopsy seems to be a very sensitive and specific diagnostic tool to diagnose early stages of sacroiliitis [3]. However, it is not widely preferred in routine practice since it is an invasive procedure. Direct radiography of sacroiliac joints is mostly normal at early disease stages. Along with magnetic resonance imaging, sacroiliac scintigraphy serves as a useful non-invasive diagnostic tool to diagnose active sacroiliitis, especially at early stages of disease, in which direct radiography is usually normal [2, 4, 5].
123
228
Sacroiliac scintigraphy, as being a functional imaging modality, can provide a metabolic information for the disease process. Scintigraphy with bone radiopharmaceuticals shows the osteoblastic activity within the sacroiliac joints and the degree of osteoblastic activity uptake correlates well with the disease activity, even if there is no or subtle morphological changes assessed by radiological methods. In the scintigraphic evaluation of sacroiliac joints, quantitative analysis is commonly preferred along with visual analysis. Quantitative analysis of scintigraphic uptake consists of comparison of uptake ratios of both sacroiliac joints to any selected adjacent background structure, mostly sacrum among other sites used so far [6– 10]. The advantage of adding quantitative analysis is that it provides a more objective perspective to visual interpretation. However, significant overlap between healthy and pathological sacroiliac index values (SI) has been reported for adult patients [6, 8, 11–13]. Especially young adults have higher values for SI even if they do not suffer from sacroiliitis, mostly due to their growing pattern [13]. Therefore, standard SI values which are obtained from the population without any sacroiliac disease are absolutely needed to set a threshold to diagnose diseased individuals. Such studies are present for adult population in the literature [6, 8–13]. However, there is no sufficient data on this particular subject for pediatric patients to the best of our knowledge, in spite of some published studies on sacroiliitis of mostly pyogenic origin in pediatric patients [14–16]. This study was aimed to assess normal SI in different age groups of pediatric patients using 2 different quantitative methods.
Ann Nucl Med (2014) 28:227–231
adjusted dose for children according to EANM dosage card [17]. The dose of MDP for young-adult patients was 20 mCi. None of the patients had any findings or symptoms such as back pain or joint stiffness suggesting sacroiliitis, besides none had any systemic disease like SLE or DM. All the scans were reported as normal and all patients’ data were analyzed with two different methods for quantitative analysis as follows: First method Rectangular regions-of-interests (ROIs) were drawn over the right and left sacroiliac joint as well as the sacrum which represented the background activity (Fig. 1). The sacrum ROIs were placed in the center of the sacrum, to keep the standard among patients. Sacroiliac activity index (SI) was calculated as the ratio between the counts from the right and left sacroiliac joints divided by sacrum for each. All ROIs were drawn and semiquantitative analysis was done by the same nuclear medicine physician (MFB). Second method The second method included ROIs over L5 vertebra instead of the sacrum as the background activity (Fig. 1). Vertebral ROIs were placed centrally. SI was calculated as the ratio between the counts from the right/left joints divided by the counts from L5 vertebra. All ROIs were drawn and semiquantitative analysis was done by the same nuclear medicine physician (MFB).
Materials and methods A total of 79 bone scans of pediatric patients (30 females, 49 males, mean age 11 ± 6 years, age range 3–18 years) who were referred for various entities other than sacroiliitis in a period of 25 months were retrospectively reviewed to obtain semiquantitative analysis of sacroiliac joints. In addition, a total of 20 normally reported bone scintigraphy of young-adult patients (11 females, 9 males, mean age 25 ± 3 years, age range 21–30 years) without any referral for sacroiliitis were retrospectively reviewed for comparison with pediatric patients. The scintigraphic imaging included spot pelvis anterior and posterior planar images (10 % energy window set at a peak of 140 keV of Tc-99m, 256 9 256 9 256 matrix, 500 kcount preset count) obtained with a dual-headed gamma camera equipped with low-energy high-resolution collimators (Siemens E-cam, Erlangen Germany) at 3–4 h following intravenous injection of Tc-99m-methylene diphosphonate (MDP) with an
123
Fig. 1 Placement of the ROIs over right (ROI0) and left sacroiliac joints (ROI1) as well as sacrum (ROI2) and L5 vertebra (ROI3) on posterior planar image. ROI2 is used for the first quantitation method where ROI3 is used for the second one
Ann Nucl Med (2014) 28:227–231
229
Comparative analysis of the first and second method The right and left SI ratios with background sacrum counts (Rs and Ls) calculated according to the first method were compared with those calculated from L5 as background counts (Rv and Lv) from the second method in relation with gender. Comparative analysis of the two methods was also done for different age groups of children to assess the influence of age. For this analysis, the patients were grouped according to their ages as years; group 1 (n = 19, age 1–5 years), group 2 (n = 21, age 6–10 years), group 3 (n = 21, age 11–15 years) and group 4 (n = 18, age 16–18 years). To assess the influence of age from a different perspective, comparative analysis of the two methods was also performed for young-adult group of patients as group 5 (n = 20, age 21–30 years). Statistical analysis Descriptive statistics are provided as means and standard deviations or as medians and ranges. To compare two groups with each other, Student’s t test was used. When variables could not be assumed to be normally distributed Mann–Whitney rank sum test was used instead. A P \ 0.05 was considered to be significant.
Results Side-wise comparison of quantitative sacroiliac scintigraphy of a total of 79 children (30 girls, mean age 11 ± 6 years, age range 3–18 years) showed no significant difference between mean values of right SI ratios for the first (Rs) and the second method (Rv) along with no significant difference between left SI ratios for the first (Ls) and second method (Lv) (Table 1). However, when the same side mean values of SI ratios from two different
methods were compared to each other, there was statistically significant difference (Table 2). Gender-based comparison of the two methods yielded significantly higher SI ratios for females for the first method (Rs 1.33 ± 0.26 vs 1.50 ± 0.32, P = 0.01; Ls 1.33 ± 0.28 vs 1.49 ± 0.33, P = 0.02) while no such significant difference existed for the second method (Table 3). Age-based comparative analysis of the two methods revealed a significant increase in absolute values for SI ratios with both methods as age increased, on comparison with distant age groups (Table 4). Age-based analysis depending on the specific age groups showed that there were significantly lower absolute SI values with the second method, when similar consecutive age groups like group 1–2 or group 2–3 were compared with each other, while no such difference with the first method was found (Table 4). Age-based analysis of the two methods showed insignificant results when SI values of older children (group 4) were compared with those of young-adult population (group 5), although there was significant difference between SI ratios of younger children (group 1) and those of young-adult (group 5) (Table 4). For each individual patient from any of the age groups, method-based comparison resulted in a significantly different SI with both methods when mutually one-to-one comparison of each absolute SI value was performed (P \ 0.01).
Discussion Bone scintigraphy serves as a useful diagnostic imaging method to detect sacroiliitis. Quantitative sacroiliac scintigraphy can increase the diagnostic capability of the method and provide a more objective perspective to interpretation of bone scans [6, 8–12]. However, despite
Table 1 Side-wise comparison of mean SI values for all patients for the first and second methods SI
First method
P value
Rs
Ls
1.23 ± 0.26
1.25 ± 0.25
0.66
Second method
P value
Rv
Lv
1.14 ± 0.21
1.13 ± 0.20
0.89
Table 2 Comparison of same side SI values according to two different methods SI
Right-side values
P value
Rs
Rv
1.23 ± 0.26
1.14 ± 0.21
0.02
Left-side values
P value
Ls
Lv
1.25 ± 0.25
1.13 ± 0.20
0.01
123
230
Ann Nucl Med (2014) 28:227–231
Table 3 Comparison of mean SI values for each gender and for each method SI
Male
Female
P*
Rs
1.33 ± 0.26
1.50 ± 0.32
0.01
Rv
1.22 ± 0.21
1.26 ± 0.25
0.48
Ls
1.33 ± 0.28
1.49 ± 0.33
0.02
Lv
1.23 ± 0.23
1.24 ± 0.28
0.64
* Difference between genders
quantitative analysis, there is still significant discrepancy on the fact that bone scintigraphy can accurately differentiate active sacroiliitis from normal sacroiliac joints. The major drawback of the method is that SI ratios may show a wide range and are highly variable, especially in young adulthood. A study by Lin and Wang [13] concluded that SI ratios were higher in males younger than 30 years old and that those ratios tend to decrease by age at both genders. To the best of our knowledge, there is yet no such study on pediatric patients in which the influence of age and gender has been assessed. Most of the published data on sacroiliitis in pediatric patients are related to pyogenic sacroiliitis and spondyloarthropathies [14–16]. Therefore, this study was firstly aimed to assess whether there is any effect of age to SI ratios in a group of pediatric patients without any symptoms and findings suggesting sacroiliitis and, secondly, the use of an alternative quantitation method was assessed in comparison with the conventional SI quantitation method. This alternative method included L5 vertebral ROI instead of sacrum, which is the most common background site for quantitative sacroiliac scintigraphy. Although there is no published data related to disadvantages of sacrum as a background site in the literature, in routine practice, prominent sacral promontorium may cause underestimation of SI ratios because of relatively increased sacral activity at least for some patients. Thus, L5 vertebra was chosen as an alternative background site for the second quantitation method. Lumbal vertebrae are among the most common bony structures in which osteodegenerative changes are frequently seen. This issue
may theoretically present as a disadvantage of this method for adult patients, taking into account that osteodegenerative changes are not expected to be seen at pediatric age. Age-based comparison of the methods showed in contrast to previous study by Lin and Wang [13] that SI ratios were significantly higher for girls compared to boys, for the first method while no effect of gender was found for the second method. Another interesting result from the present study is that SI ratios by two methods were totally different from each other when mutual one-to-one correspondence was analyzed. This means that setting a particular threshold for each quantitative method needs to be done separately and that the same method—whether the first or the second— has to be used at the follow-up studies. Although the absolute values from two methods were totally different from each other, SI ratios tended to increase as age increases independently from the method applied. It can be hypothesized that SI values continue to increase as the children grow up and get mature, presumably until the age of 30 taking into account the previous studies although related data are limited [11, 13]. The results of the current study showed that age-based analysis depending on both methods came up with significantly higher SI values at young-adult population (group 5) on comparison with younger children (group 1). This result may also support the hypothesis that SI ratios tend to increase as age increases and it is independent from the method used. Different age groups of pediatric patients yielded different absolute SI values. However, analysis between similar (consecutive) age groups showed that SI ratios differed significantly when L5 was chosen as background site, in contrast with sacrum of the first method where no such difference was observed. Depending on the current study, as SI ratios tend to increase by age in parallel with growing and maturation, this finding may contribute to the fact that the second method may represent the normal osteoblastic metabolism slightly more precisely compared to the first method. However, its clinical significance needs to be verified in further studies.
Table 4 Comparison of mean SI values for all age groups (group 1, 2, 3, 4 and 5) in relation with 2 quantitation methods SI
Group 1
Group 2
Group 3
Group 4
Group 5
P*
P**
P***
Rs
1.16 ± 0.26
1.42 ± 0.26
1.41 ± 0.23
1.51 ± 0.23
1.50 ± 0.29
0.94
\0.01
0.87
Rv
1.04 ± 0.11
1.11 ± 0.16
1.25 ± 0.14
1.23 ± 0.14
1.29 ± 0.36
0.04
\0.01
0.22
Ls
1.15 ± 0.25
1.43 ± 0.25
1.39 ± 0.26
1.55 ± 0.26
1.52 ± 0.32
0.67
\0.01
0.79
Lv
1.03 ± 0.10
1.12 ± 0.10
1.22 ± 0.16
1.26 ± 0.16
1.28 ± 0.33
0.04
\0.01
0.89
* Difference between group 2 and 3 ** Difference between group 1 and 5 *** Difference between group 4 and 5
123
Ann Nucl Med (2014) 28:227–231
Another interesting result from the current study is that SI ratios of females were higher than those of males for the first method even though there was no such significant difference between genders for the second method. In our opinion, the fact that SI ratios for females were higher than those for males for the first method while no such significant difference was present for the second method may possibly be attributed to difference of anatomy between two genders. Although this is only a speculative explanation, we think that the relatively more prominent gluteal muscular and/or fat tissue of female gender may possibly cause more attenuation of counts from the sacrum. When L5 vertebra is used as a reference center instead of sacrum, such attenuation effect from the gluteal region of female gender might be expected to be lower compared to that of sacrum, depending on the anatomic localizaton and relation of sacrum and L5 vertebra with regard to gluteal soft tissue. However, more studies are surely warranted to analyze the possible causative factors between two genders.
2. 3.
4.
5. 6.
7.
8.
9.
10.
Conclusion The choice of L5 vertebra as an alternative to sacrum as background ROI site performs well for pediatric patients and can be preferred in case there is prominent sacral promontorium activity which may possibly hamper quantitation resulting in false interpretation. Regardless of the method used, SI ratios of children tend to increase as their ages increase. Thus, it may not seem possible to obtain a single SI ratio threshold value to differentiate sacroiliitis from normal metabolic activity in pediatric patients. It would be reasonable for each center to generate their own threshold values for different age groups of children. Conflict of interest
231
11.
12.
13. 14.
15.
None declared. 16.
References 1. Burgos-Vargas R. The assessment of the spondyloarthiritis international society concept and criteria for the classification of axial spondyloarthritis and peripheral spondyloarthritis:A critical appraisal for the pediatric rheumatologist. Pediatric Rheumatol
17.
Online J. 2012;31:10(1):14. http://www.ped-rheum.com/content/ 10/1/14. Braun J, van der Heijde D. Imaging and scoring in sacroiliitis. Best Pract Res Clin Rheumatol. 2002;16(4):573–604. Gong Y, Zheng N, Chen SB, Xiao ZY, Wu MY, Liu Y, et al. Ten years’ experience with needle biopsy in the early diagnosis of sacroiliitis. Arthritis Rheum. 2012;64(5):1399–406. doi:10.1002/ art.33453. Inanc N, Atagu¨ndu¨z P, Sen F, Biren T, Turog˘lu HT, Direskeneli H. The investigation of sacroiliitis with different imaging techniques in spondyloarthropathies. Rheumatol Int. 2005;25(8):591–4 (Epub 2004). Grigoryan M, Roemer FW, Mohr A, Genant HK. Imaging in sacroiliitis. Curr Rheumatol Rep. 2004;6(2):102–9. Bozkurt MF, Ug˘ur O, Ertenli I, Caner B. Combined use of bone and bone marrow scintigraphies for the diagnosis of active sacroiliitis. Ann Nucl Med. 2001;15(2):117–21. Fogelman I, Maisey MN, Clarke SEM. Sacroiliitis. In: Fogelman I, Maisey MN, Clarke SEM, editors. An atlas of clinical nuclear medicine. 2nd ed. Singapore: Martin Dunitz Ltd; 1994. p. 84. Ryan PJ, Fogelman I. Sacroiliac quantitative bone scintigraphy in ankylosing spondylitis:any clinical relevance? Nucl Med Commun. 1993;14:19–20. Davis MC, Turner MA, Charters JR, Golden HE, Ali A, Fordham EW. Quantitative sacroiliac scintigraphy:the effect of method of selection of region of interest. Clin Nucl Med. 1984;9:334–40. Huanga JY, Tsaib MF, Kaoc PF, Chena YS. Automatic computer-aided sacroiliac joint index analysis for bone scintigraphy. Comput Methods Programs Biomed. 2010;98(1):15–26. doi:10. 1016/j.cmpb.2009.07.010 (Epub 20). Ayres J, Hilson AJ, Maisey MR, Laurent R, Panayi GS, Saunders AJ. An improved method for sacroiliac joint imaging:a study of normal subjects, patients with sacroiliitis and patients with lowback pain. Clin Radiol. 1981;32:441–5. Prakash S, Gopinath PG, Bhargava S, Mehra NK, Malaviya AN. Evaluation of quantitative sacroiliac scintigraphy for the early detection of sacroiliitis. Eur J Nucl Med. 1983;8:531–4. Lin WY, Wang SJ. Influence of age and gender on quantitative sacroiliac joint scintigraphy. J Nucl Med. 1998;39:1269–72. Wu MS, Chang SS, Lee SH, Lee CC. Pyogenic sacroiliitis—a comparison between paediatric and adult patients. Rheumatology. 2007;46:1684–7. Taylor ZW, Ryan DD, Ross LA. Increased incidence of sacroiliac joint infection at a children’s hospital. J Pediatr Orthop. 2010;30(8):893–8. Stoll ML, Bhore R, Dempsey-Robertson M, Punaro M. Spondyloarthritis in pediatric population: risk factors for sacroiliitis. J Rheumatol. 2010;37(11):2402–8. doi:10.3899/jrheum.100014 (Epub 2010 Aug 3). Lassman M, Biassoni L, Monsieurs M, Franzius C, Jacobs F. EANM dosimetry and paediatrics committees. The new EANM paediatric dosage card. Eur J Nucl Med Mol Imaging. 2009;36(3):540–1.
123
