Radiomics: a new tool to differentiate adrenocortical adenoma from carcinoma

Background

The main challenge in the management of indeterminate incidentally discovered adrenal tumours is to differentiate benign from malignant lesions. In the absence of clear signs of invasion or metastases, imaging techniques do not always precisely define the nature of the mass. The present pilot study aimed to determine whether radiomics may predict malignancy in adrenocortical tumours.

Methods

CT images in unenhanced, arterial, and venous phases from 19 patients who had undergone resection of adrenocortical tumours and a cohort who had undergone surveillance for at least 5 years for incidentalomas were reviewed. A volume of interest was drawn for each lesion using dedicated software, and, for each phase, first-order (histogram) and second-order (grey-level colour matrix and run-length matrix) radiological features were extracted. Data were revised by an unsupervised machine learning approach using the K-means clustering technique.

Results

Of operated patients, nine had non-functional adenoma and 10 carcinoma. There were 11 patients in the surveillance group. Two first-order features in unenhanced CT and one in arterial CT, and 14 second-order parameters in unenhanced and venous CT and 10 second-order features in arterial CT, were able to differentiate adrenocortical carcinoma from adenoma (P < 0.050). After excluding two malignant outliers, the unsupervised machine learning approach correctly predicted malignancy in seven of eight adrenocortical carcinomas in all phases.

Conclusion

Radiomics with CT texture analysis was able to discriminate malignant from benign adrenocortical tumours, even by an unsupervised machine learning approach, in nearly all patients.

Biomarkers, imaging and disease activity indices in patients with early axial spondyloarthritis: the Italian arm of the SpondyloArthritis-Caught-Early (SPACE) Study

The study aimed to evaluate biomarkers facilitating early diagnosis of axial spondyloarthritis (axSpA) and correlations between them and disease activity parameters and imaging indexes. Patients with low back pain (LBP) (≥3 months, ≤2 years, onset ≤45 years) participating in the Italian arm of the SpondyloArthritis-Caught-Early SPACE study underwent a physical examination, questionnaires, laboratory tests, X-rays and MRI of the spine and sacroiliac joints (SIJ). An expert rheumatologist formulated axSpA diagnosis in accordance with Assessment of SpondyloArthritis International Society (ASAS) criteria. Disease activity and physical functioning were assessed using imaging, clinical and serological indices. Spine and SIJ MRI and X-rays were scored independently by 2 readers using the SPARCC, mSASSS and NY-criteria. Patients were classified as: subjects with signs of radiographic sacroiliitis (r-axSpA), subjects with signs of sacroiliitis on SIJ-MRI but not on X-rays (nr-axSpA MRI SIJ+) or subjects with no signs of sacroiliitis on MRI/X-rays but with >2 SpA features and signs of bone oedema on MRI spine (nr-axSpA MRI SIJ-/undifferentiated SpA). Significant differences were found in the prevalence of radiographic sacroiliitis, active sacroiliitis on MRI and SPARCC SIJ scores. Biomarker levels were not significantly increased in any of the patient groups. The correlations between IL-17 and IL-23 and other indices were not significant; correlations were found between IL-22 and BASFI, BASG1, HAQ, VAS pain, between mSASSS and MMP3, and between the latter and hsCRP. Although not significantly higher in any of the three groups, IL-22, MMP3 and hsCRP values were correlated with some disease activity indexes and with mSASSS. Large observational studies are required to confirm these preliminary findings.

Spine and sacroiliac joints on magnetic resonance imaging in patients with early axial spondyloarthritis: prevalence of lesions and association with clinical and disease activity indices from the Italian group of the SPACE study

Our aim was to determine the prevalence of spine and sacroiliac joint (SIJ) lesions on magnetic resonance imaging (MRI) in patients with early axial spondyloarthritis (axSpA) and their correlation with disease activity indices. Sixty patients with low back pain (LBP) (≥3 months, ≤2 years, onset ≤45 years), attending the SpA-clinic of the Unità Operativa Complessa Reumatologia of Padova [SpondyloArthritis-Caught-Early (SPACE) study], were studied following a protocol including physical examination, questionnaires, laboratory tests, X-rays and spine and SIJ MRI. Positive spine and SIJ MRI and X-rays images were scored independently by 2 readers using the SPARCC method, modified Stoke ankylosing spondylitis spine score and New York criteria. The axial pain and localization of MRI-lesions were referred to 4 sites: cervical/thoracic/lumbar spine and SIJ. All patients were classified into three groups: patients with signs of radiographic sacroiliitis (r-axSpA), patients without signs of r-axSpA but with signs of sacroiliitis on MRI (nr-axSpA MRI SIJ+), patients without signs of sacroiliitis on MRI and X-rays (nr-axSpA MRI SIJ-). The median age at LBP onset was 29.05±8.38 years; 51.6% of patients showed bone marrow edema (BME) in spine-MRI and 56.7% of patients in SIJ-MRI. Signs of enthesitis were found in 55% of patients in the thoracic district. Of the 55% of patients with BME on spine-MRI, 15% presented presented a negative SIJMRI. There was a significant difference between these cohorts with regard to the prevalence of radiographic sacroiliitis, active sacroiliitis on MRI and SPARCC SIJ score. The site of pain correlated statistically with BME lesions in thoracic and buttock districts. Since positive spine-MRI images were observed in absence of sacroiliitis, we can hypothesize that this finding could have a diagnostic significance in axSpA suspected axSpA.

[The T1 time dependence of water measured in normal and neoplastic lung tissues]

The authors determined the time dependence of the average T1 relaxation time for the constitutive water of normal and neoplastic lung tissues. In healthy tissues, the T1 dependence appears strictly as a single decaying exponential function down to a threshold value which is invariant over time. The time change of T1 in neoplastic tissues is represented by monoexponential functions, in the short and long run, with a sudden drop 2 weeks after beginning measurements. Moreover, the time dependence is independent of the examined variant of epidermoid carcinoma. We report the kinetic constants of the in vitro aging of the tissues. The presence of a cooperative process, involving neoplastic lung water, to explain the features of aging changes, is also suggested.

[The proton relaxation times in normal and neoplastic thyroid tissue. An in vitro study]

Aim of the work is the in vitro evaluation of normal and tumoral thyroid tissue relaxation times, in order to obtain useful information for MR Imaging in vivo. Water proton high-resolution MR spectra, together with T1 and T2 relaxation times, have been measured in normal and tumoral thyroid tissues. The longitudinal relaxation time T1 was particularly sensitive to pathologic tissue: significant differences (mean: 243 ms) were observed between normal and tumoral tissues, which is a useful finding for the successful application of MR techniques to the detection of tumors in the thyroid gland. The average T1 relaxation time, for each class of examined tumors, correlates qualitatively with the degree of malignancy. On the contrary, the transverse relaxation time T2 was insensitive to pathologic changes in the tissues. The analysis of water proton spectra showed the overall Free Induction Decay constant T2 to allow an easier discrimination than T2 between normal and pathologic thyroid tissue. Thus, T1-weighted sequences probably provide better discrimination than T2-weighted scans in the evaluation of normal and tumoral thyroid tissues.