Purpose: To analyze the efficacy of oral bisphosphonate therapy (alendronate and risedronate) in patients with β-Thalassaemia evaluating incidence of vertebral fractures (VFs) and bone mineral density (BMD) values.
Methods and Materials: 60 patients (31 M, 29 F) with β-thalassaemia major of average age 35.02 ± 6.32 afferent at the center of Ferrara were selected retrospectively with age> 25 years, at least 1 between reduced BMD (Z-score<-2 premenopausal women and men aged <50), osteoporosis (T-score <-2.5), ≥ 1 fragility fracture. We analyzed BMD values, thoracolumbar radiographs for VFs evaluation (according Genant classification) and biochemical data at 2, 3 and 5 years.
Results: The mean duration of therapy was 6.23 ± 1.93 years. We observed an improvement from baseline of lumbar spine BMD of 4% to 2 yr of treatment, 2% to 3 yr and 3% to 5 yr. BMD is increased by 1% in the femur (F) and 2% at the neck of the femur (FN) at 2 yr of treatment, at 3 yr + 1% (F) and 0% (FN), at 5 aa -3% (F) and -2% (FN). Incidence of new vertebral fractures was of 27%.
Conclusion: Therapy was useful in rising BMD at the lumbar level with an improvement of Z and T-score and in stabilizing BMD at the femoral level (F and FN) in the first 3 yr of therapy with a slight reduction at 5 yr, but VFs in these well-treated patients were still common and morphometry measurement remains a useful test in surveillance of these patients.
Purpose: Dual X-ray absorptiometry (DXA) is the most common mode of bone mineral density (BMD) evaluation in young adults with b-thalassaemia major (TM). In addition to DXA that determines an areal BMD, Quantitative Computerised Tomography (QCT) also detects volumetric BMD which is independent of bone size. Our aim is to evaluate concordance of BMD values in thalassaemic patients obtained by QCT and DXA.
Methods and Materials: We enrolled 15 patients with TM from the Hospital of Ferrara who underwent both a DXA scan of lumbar spine in regular follow-up and an unenhanced CT, performed for other causes, within 4 months. CT images of lumbar spine were evaluated with a quantitative dedicated post-processing software calculating volumetric BMD for L1-L4 vertebrae. BMD values of both methods were expressed as Z-scores and T-scores and the results were correlated.
Results: Of 10 patients (6 males < 50 aged, 4 premenopausal woman), the overall prevalence of severely low bone density (Z-score ≤-2) was 9/10 (90%) by DXA and 3/10 (30%) by QCT with concordance in 4/10 patients (40%). In all 10 patients of this group QCT Z-score was higher than DXA Z-score with significant difference(p-value=0.002). 5 patients(>50 years old) showed T-score ≤-2,5 correspondent to osteoporosis both on DXA and QCT with 100% of concordance.
Conclusion: Our data show a discrepancy in lumbar BMD Z-scores by DXA and QCT in thalassaemic patients suggesting that DXA BMD could be falsely low in young subjects with short bones and QCT could be more accurate to assess it.
Purpose: To evaluate the applicability of in vivo hydroxyapatite (HA)-specific bone mineral density (BMD) measurements based on non-contrast-enhanced phantomless dual-layer spectral CT (DLCT).
Methods and Materials: BMD were obtained using spectral information from DLCT as well as with quantitative CT (QCT) examining a spine phantom with three artificial vertebral bodies with known HA densities, while simulating different patient positions and grades of obesity. Furthermore, HA-specific BMD values were determined using non-contrast routine DLCT in 174 vertebrae (33 patients; 66±18 years; 33% women; 45% with prevalent fractures) and the results were compared with QCT-based BMD. The discriminative power of HA-specific BMD for differentiating subjects with versus without osteoporotic fractures was analysed using ROC analyses, also for an extended cohort of 79 patients (66±18 years; 52% women; 38% with prevalent fractures).
Results: HA-specific BMD measurements in the phantom were more accurate than QCT, particularly when simulating obese patients. In vivo, strong correlations were shown between DLCT and QCT (r=0.987, p<0.001) and a high agreement in a Bland-Altman plot was observed. ROC areas under the curve (AUC) were 0.889 for QCT- and 0.878 for DLCT-based BMD in 33 patients. The AUC for the extended cohort was 0.858 for HA-specific BMD measurements, with an optimal cut-off at 81 mg/ml.
Conclusion: In vivo, DLCT-based HA-specific BMD measurements showed comparable results to QCT measurements as well as for distinguishing patients with vertebral fractures and those without. This suggests that HA-specific BMD measurements obtained from phantomless DLCT may be used for opportunistic osteoporosis screening.
Purpose: To evaluate the consistency of bone mineral density (BMD) measurements between a phantomless dual-energy CT scanner and conventional quantitative CT (QCT) in vitro.
Methods and Materials: QCT and dual-energy CT scans of 23 fresh sheep vertebrae were performed on the same CT scanner which generated dual-energy images from rapid kilovoltage switches (80kv-140kv). A fixed tube voltage of 120kV was applied for QCT scans with the bone equivalent phantom placed below the samples. Trabecular BMD analyses of QCT images were performed using QCTpro software. The dual-energy CT images were processed to generate calcium hydroxyapatite (HAP) density images based on HAP-water material pairs. HAP densities of the same regions were measured. Pearson correlation, paired t test, and intraclass correlation coefficient (ICC) were applied for consistency assessment.
Results: BMD measurements of QCT ranged from136 mg/cm3 to 486 mg/cm3, and the BMD measurements of DECT ranged from 128mg/cm3 to 419mg/cm3. BMD derived from DECT images were significantly lower than that of QCT (p<0.001), while they displayed a strongly linear correlation (r=0.989, p<0.001). The ICC between the two measurements was 0.974 (p<0.001).
Conclusion: In vitro, the HAP densities acquired by dual-energy CT displays a highly linear correlation with BMD measured by QCT of the same specimens. Although the results from dual-energy CT were not completely same as QCT, the phantomless dual-energy CT has the potential to be applied for BMD assessment clinically.
Purpose: To assess how bone morphology and mineralization changes with age in women.
Methods and Materials: The non-dominant hand was analyzed with digital X-ray radiogrammetry (DXR) in 1,429 women aged 40-74 years undergoing mammography screening twice with 18 to 24 months interval, depending on age. Yearly change in bone parameters was calculated by averaging the individual changes.
Results: Both the inner and outer diameter of the metacarpals increased with age while DXR bone mineral density (DXR-BMD) decreased. There was a continuous increase in outer diameter in adult age while there was a five-fold increase in yearly inner diameter growth and DXR-BMD loss during 48 to 53 years of age. In the age-interval 40-47 the change in W and ID was on average 4.65um and 8.8 um /year respectively, compared to 1.9 um and 49 um /year after 53 years of age.
Conclusion: The faster decrease in DXR-BMD observed during and after menopause is caused by resorption of the inner cortical surface, increasing the inner diameter that no longer is matched by the increase in outer diameter. To our knowledge it has not previously been shown how individual bone diameter changes with age in adults. The automatized analysis and high reproducibility of DXR-BMD made that analysis possible in a great number of subjects despite a short time-interval. We speculate that most bones in the human body grow in the same pattern as observed in the metacarpals, partly explaining the decreasing BMD at older age.
Purpose: To evaluate the prevalence of sarcopenia in patients with and without diagnosis of malignant tumour. To assess the relationship between sarcopenia and osteoporosis in patients with malignancy.
Methods and Materials: 70 patients age ≥65 (mean age 71.5), 33 male and 37 female, that underwent abdominal CT were included in the study. Half of patients (N=35) had diagnosis of malignant tumour, majority colorectal cancer (37%). Trabecular bone mineral density at level of L1 vertebral body, in upper vertebral body part, using single slice technique, was measured for detecting osteoporosis. Density and cross-section area of psoas muscles at L3 vertebral body mid level were measured for diagnosing sarcopenia. Cut-off values for CT detection of sarcopenia and osteoporosis established by previous researches were used. Chi square test, Pearson correlation, and linear regression analysis were performed.
Results: 40% of patients with diagnosis of malignant tumour had muscle density diagnostic for sarcopenia, compared to 11.4% in patients without cancer. Differences for sarcopenia, assessed by psoas muscle density values, between patients with malignancy and non-malignant group were statistically significant (χ2=6.06, p=0.016). Osteoporosis values weren’t significantly different between groups (χ2=0.06, p=0.811).There was moderate correlation between bone density and psoas muscle density in patients with malignancy (r=0.441, p<0.05). Regression analysis showed sarcopenia was positively correlated with age (odds ratio=1.169, p=0.036) and malignancy (odds ratio=0.117, p=0.08).
Conclusion: Sarcopenia is significantly higher in patients with diagnosis of cancer, compared to patients without malignancy diagnosis. Sarcopenia is associated with osteoporosis in malignancy and should be mentioned in abdominal CT report.
Purpose: Bone marrow fat (BMF) is an ectopic fat storage depot connecting metabolism and bone health. Evidence on changes in BMF with weight loss is limited. Therefore we aimed to analyze changes in BMF content following dietary induced weight loss.
Methods and Materials: Magnetic resonance imaging data of the HELENA-trial, an RCT among 143 non-smoking, obese participants (BMI at baseline between 25 and 40 kg/m2, 50% female), were used to quantify BMF content before and after a 12 week dietary intervention phase. Regions of interest (2cm²) were evaluated manually on a post-processing software (OsiriX, Pixmeo SARL, Bernex, Switzerland) using the proton density fat fraction map, based on mean counts from L1 and L2. The study cohort was divided into weight loss quartiles based on changes in body weight between baseline and week 12.
Results: Average BMI at baseline was highly similar across the weight loss quartiles (Q1:32.1±4.1kg/m2; Q2: 31.1±3.7kg/m2; Q3: 30.9±3.4kg/m2; Q4: 31.5±3.7kg/m2). Diet induced weight loss varied between 0.0±0.2% for Q1 to -11.3±0.6% for Q4. Relative changes in BMF were 0.7±2.4%, -3.5±2.0%, -1.4±2.4%, and -12.7±3.3% for Q1 to Q4. Across all four weight loss quartiles and for the two group comparison Q1 vs. Q4 there was a significant difference (p<0.01) for changes in BMF. For the two group comparisons Q1 vs. Q2 and Q1 vs. Q3 there were no significant differences (p>0.05) observed.
Conclusion: Our data suggest that BMF levels decrease with weight loss. Further research will unravel the association between the amount of weight loss and BMF.
Purpose: We investigated the in-vivo effect of increasing BMI and waist circumference on TBS precision error.
Methods and Materials: A population of postmenopausal Caucasian women was distributed in three different BMI (normal, overweight, class I obesity), plus two further groups based on waist circumference (WC) diameter (≤88 cm and >88 cm respectively). In-vivo precision error was calculated on 30 consecutive subjects that were scanned two times, with patient repositioning. Coefficient of Variation, percent least significant change (LSC%) and reproducibility were calculated according to the International Society for Clinical Densitometry guidelines.
Results: Ninety-five women aged 66 ± 10 (mean ± standard deviation) were included. No significant differences were found for BMD and TBS precision errors, respectively, when comparing BMI groups and WC groups. BMD reproducibility ranged from 95.9% (BMI >30 kg/m2) to 97.5% (BMI <25 kg/m2). TBS reproducibility ranged between 95.8% (BMI= 25-29.9 kg/m2, WC >88 cm) and 96.6% (BMI <25 kg/m2). With the exception of obese group, a significant difference was found between BMD and TBS reproducibility, being that of TBS slightly lower than BMD. A significant decrease of TBS was found between normal and obese subjects, as well as between WC groups; BMD variations between groups were not statistically significant.
Conclusion: TBS precision error is not affected by BMI and WC differences. TBS reproducibility showed to be slightly lower than that of BMD, but this difference was mitigated in obese patients. A negative association was found between the amount of fat mass and TBS mean values.
Purpose: To determine the dual-energy computed tomography (DECT) attenuation properties of meniscal calcifications in calcium pyrophosphate deposition (CPPD) in vivo, and assess whether DECT was able to discriminate meniscal CPP deposits from calcium hydroxyapatite (HA) in subchondral and trabecular bone.
Methods and Materials: Patients with clinical suspicion of crystal-related arthropathy and knee DECT scans were retrospectively assigned to CPPD (n=19) or control (n=21) groups depending on the presence/absence of chondrocalcinosis on DECT. Regions of interest were drawn in meniscal calcifications, non-calcified menisci, as well as subchondral and trabecular bone. Five DECT parameters were obtained: CT numbers (HU) at 80 and 140 kV, dual-energy index (DEI), electron density (ρe), and effective atomic number (Zeff). The four different knee structures were compared using mixed linear models.
Results: Meniscal calcifications (n=89) in CPPD patients had mean±SD CT numbers at 80 and 140 kV of 257±64 and 201±48 HU, respectively; with a DEI of 0.023±0.007, and ρe and Zeff of 140±35 and 8.8±0.3, respectively. Meniscal CPP deposits were readily distinguished from calcium HA in subchondral and trabecular bone (p≤0.01), except at 80 kV separately (p=0.28). Zeff and ρe both significantly differed between CPP deposits and calcium HA in subchondral and trabecular bone (p<0.0001).
Conclusion: This proof-of-concept study shows that DECT can discriminate meniscal CPP deposits from calcium HA in subchondral and trabecular bone in vivo, paving the way for the biochemical signature assessment of intra- and juxta-articular calcium crystal deposits.
Purpose: To evaluate the accuracy and which DECT findings are characteristic of diagnosing foot and ankle gout in patients without uric acid crystal deposition in first metatarsophalangeal joint (MTPJ).
Methods and Materials: Among 97 foot and ankle DECT obtained from January 2016 to September 2017, 27 foot and ankles without first MTPJ green pixelation were enrolled. Two blinded musculoskeletal radiologists evaluated the images and classified the examination findings as positive or negative for the presence of uric acid crystals. Readers also assessed following results of green pixelation: volume, site, size, shape, vanishment after increase minimal HU from 130 to 150. Reference standard was the results of joint aspiration and over one year clinical follow up. Logistic regression analyses were conducted to identify which findings were significantly associated with gout.
Results: The sensitivity and specificity of DECT without first MTPJ green pixelation was 0.96 and 0.67. The mean volume of green pixelation was 1.7cm3. All false negative patients had submillimeter artifact. The DECT findings showed significant differences between artifact and gout with respect to shape, vanishment after increase minimal HU from 130 to 150, site (p<0.05). Multivariate logistic regression showed correlations between artifact and shape (non-oval, sharp, flat, odd ratio 27.4), and vanishment after increase minimal HU (odd ratio 19.2).
Conclusion: The specificity of DECT of foot and ankle without green pixelation in first MTPJ is unsatisfactory due submillimeter artifact. Green pixelation with non-oval, sharp, flat shape and which vanishes after increased minimal HU in post-processing may associated with artifact.
Purpose: T2*mapping at 7T was used to investigate alterations in tendon composition and their relation to demographics of patients with DM1.
Methods and Materials: Eighteen DM1 patients (11f/7m, 28.25±7.14y, BMI 22.3±2.28) were enrolled together with eighteen sex, age and BMI matched healthy controls (11f/7m, 27.35±5.93y, BMI 21.61±2.67). MRI was performed on a 7T whole body scanner with a dedicated multichannel receive proton knee coil. FS-PDw images were acquired for morphological evaluation of patellar tendon. A variable TE sequence (10 TEs from 1.21-24.48ms) for mono-exponential T2*mapping. Differences in T2*values between diabetics and controls were compared. Sensitivity/specificity was evaluated. T2*values were correlated with BMI and age.
Results: On PDw images, tendons appeared unremarkable in all participants. Mean T2*values in patients and controls were 1.29±.41ms and 0.73±.14ms, respectively (p<0.001). AUC was 0.92 (95%CI, 0.82-1), a cut-off set to 0.89ms showed a sensitivity of 88.9% and specificity of 99.44%. BMI and T2* values correlated strong/statistically significant in diabetics (r=0.62, p<0.01) and weak in controls (r=0.26, p=0.3). Age and T2*values correlated weak in diabetics (r=0.255, p=0.31) and strong/statistically significant in controls (r=0.6, p<0.01).
Conclusion: Although inconspicuous in routine MRI, tendons of DM1 patients exhibited altered T2*values with a high diagnostic sensitivity/specificity. Furthermore, T2*values of DM1 tendons appear to be influenced by BMI. Age did not influence T2*values in DM1 patients - however, it did in controls. This may reflect a stronger impact of higher body weight to the deteriorated tendon tissue and a partial abolition of age related tendon degeneration in DM1 patients.