Work

Circadian Deregulation in Multiple Myeloma: BMAL1 / CLOCK Expression Patterns and Diagnostic Performance

Year: 2026

Type: article

Abstract: BACKGROUND: Circadian clock disruption has emerged as a relevant axis in cancer; however, the expression patterns and diagnostic relevance of BMAL1 and CLOCK in multiple myeloma (MM) remain insufficiently defined. METHODS: BMAL1 and CLOCK mRNA expression was quantified by RT-qPCR in bone marrow samples from 46 newly diagnosed MM patients and 13 healthy controls. Group comparisons, ROC analyses, multivariate logistic regression, partial correlation, and machine-learning models (logistic regression, random forest, and support-vector machine) with stratified 5-fold cross-validation (including class-weighted sensitivity analysis) were performed. RESULTS: BMAL1 expression was significantly higher in MM than in controls (1.12 ± 0.33 vs. 0.85 ± 0.33, p = 0.016; Cohen's d = 0.83), whereas CLOCK showed no significant univariate difference (p > 0.2). BMAL1 demonstrated moderate diagnostic performance (AUC = 0.729), while CLOCK alone showed limited discrimination (AUC = 0.617). Combined analysis modestly improved performance (AUC = 0.756). In multivariate logistic regression, BMAL1 remained a strong independent predictor of MM (OR = 3343.4, p = 0.0046), whereas CLOCK showed an inverse adjusted association (OR = 0.0037, p = 0.019). BMAL1 and CLOCK expression were strongly correlated in MM (ρ ≈ 0.80), persisting after adjustment for age, ISS stage, and β2-microglobulin (partial r = 0.847, p = 2.2 × 10-13). Among classifiers, the random forest achieved the best performance (AUC = 0.832; AUPRC = 0.952). Specificity was limited across classifiers, indicating that these performance estimates are exploratory and require validation in larger cohorts. CONCLUSION: MM is characterized by BMAL1 upregulation within a preserved but imbalanced BMAL1/CLOCK axis rather than uniform shifts across circadian genes. Although diagnostic performance is moderate, consistent signals across statistical and machine-learning approaches support an MM-associated circadian expression pattern warranting further validation.

Source: International Journal of Laboratory Hematology

DOI: 10.1111/ijlh.70154