Exploring Vertebral Height Deficits in Patients with Thalassemia and Sickle Cell Disease,

Data for this analysis were collected from two sources: the Thalassemia Clinical Research Network Pain Survey study and the CHRCO Clinical Bone Density database. The TCRN Pain study used a validated pain survey to collect data on 45 patients from CHRCO at multiple time points. Pain data was then analyzed along with bone mineral density (BMD) scans if performed within 6 months. BMD Z-scores were abstracted from the CHRCO database along with patient demographics. Full lateral spine scans conducted at the time of BMD scan were re-analyzed by one observer using the Vertebral Fracture Analysis (VFA) software (Hologic v12.6.1.), and scored according to Genant for vertebral abnormalities (crush, wedge, biconcavity, Grades 1 to 3).

A total of 254 VFA scans were re-analyzed from 93 patients with Thal (21.5 ± 10.7 yrs, 55% female), 45 SCD (38.1 ± 12.7 yrs, 61% Female) and 11 healthy control subjects (29.9 ± 10.2 yrs, 64% female). The patients with Thal were primarily Asian (78%), with 68.8% B-thal, 24%E-Bthal and 12.1% Hb H; while 98% of the SCD patients were Black, all HbSS. Patients with Thal were more likely to have low bone mass (any BMD Z-score ≧ −2.0) compared to those with SCD (73% vs. 33%, p<0.01). Of the patients with Thal who had VFA scans, 25 (26.9%) had at least one vertebral abnormality, compared to 16 (34.8 %) of the 45 patients with SCD (p=NS). SCD patients had significantly more vertebral abnormalities compared to healthy controls (p=0.04). However, a similar percentage of Thal patients had substantial vertebral height deficits (Grade 2 or 3; 15.1%) compared to those with SCD (13.3%). The average number of vertebrae with abnormalities within an individal scan was lower in Thal (1.6±0.2) compared to SCD (2.8 ± 0.4; p=0.003). Thal were more likely to have wedge-type deformities in the lumbar region, while SCD had biconcavity-type deformed vertebrae in the thoracic region. Of the patients with multiple VFA scans, 12 progressed toward vertebral abnormality during a 7 year period, all with Thal. The presence of a vertebral abnormality was positively related to age in Thal but not SCD (24.2 ± 1.2 vs. 20.6 ± 0.8 yrs, p=0.029) with a trend towards a relationship to low bone mass (p=0.12). Abnormal VFA was unrelated to gender and vitamin D level. In a sub-set of the Thal subjects who participated in the TCRN pain study, two-thirds reported significant pain within the past month, 16% severe and 27% reported pain localized in the lower back. However in the patients for which both VFA scans and pain studies were performed (n=32), bodily pain was not related to vertebral abnormalities.

A surprisingly high percentage of both SCD and Thal patients had significant vertebral abnormalities. In many cases the etiology of the observed abnormal morphology unclear but appears to be more developmental in SCD whereas in Thal a consequence of collapse secondary to inherent bone fragility and/or trauma. It is also yet to be determined if this phenomenon results from skeletal effects of the underlying disease or the various treatments (i.e., chelation toxicity). The prevalence of vertebral abnormality observed in this cohort of Thal patients (27%) is twice that previously reported from the TCRN (12%). The prevalence of vertebral abnormality by DXA scan has never before been reported in SCD. It is conceivable that the high rate of lower back pain in Thal is related to the observed abnormal vertebrae, however we were unable to explore this thoroughly given the small sample size. This study was limited by the absence of multiple VFA scans for all subjects and the discordance of a temporal relationship between the pain studies and VFA scans. Further research should focus on the clinical relevance of vertebral abnormalities and its etiology in these at risk patient populations.

No relevant conflicts of interest to declare.