MGUS and Chip: Two Faces, but Not of the Same Medal

Clonal, pre-clinical expansions of hematopoietic cells are increasingly recognized. Clonal hematopoiesis of indeterminate potential (CHIP) has been recently described. It is characterized by the presence of mutations usually involved in myeloid neoplasia in people without any sign of hematological malignancy. Well before CHIP, another pre-clinical clonal expansion has been described for plasma cells (PC) based on the presence of a serum monoclonal protein, namely monoclonal gammopathy of undetermined significance (MGUS). Despite both being frequent, the association between these two bone marrow (BM) clonal entities has not been determined yet. Since the incidence of MGUS and CHIP rises with age reaching >10% of people >70y, we analyzed a unique oldest-old population dataset to define a possible association between these two different BM clonal disorders.

We analyzed a cohort of 777 patients with a median age of 91 years (range, 81 - 104), significantly higher than previous studies on MGUS. 579 (74.5%) were females and 198 (25.5%) males. CHIP was assessed in all but 45 through sequencing of a myeloid specific gene panel. Serum protein electrophoresis was available in all of them to assess the prevalence of MGUS. The prevalence of CHIP and MGUS were 17.5% (128/732) and 9.5% (74/777), respectively. Importantly, CHIP and MGUS did not associate in our cohort but rather showed a non-significant trend towards anti-correlation (Fisher's Exact test, p-val = 0.09). We then tested associations with different clinical and laboratory features, finding that, as expected, MGUS associated with higher concentration of gamma-globulins (Wilcoxon test, p-val = 0.01414), but also with absolute lower levels of albumin (Wilcoxon test, p-val = 0.01398). No significant association was found with the mean corpuscular value (MCV), hemoglobin levels or age. These results were confirmed by a logistic univariate model, in which also the male gender resulted associated with the presence of a monoclonal component with borderline significance. Then, to further corroborate our results we performed univariate linear regression analyses. Given that CHIP and MGUS are mainly considered two aging conditions, we investigated in linear regression models the impact of age, as a continuous variable, on other clinical features. In particular, we observed that the increase of age was significantly correlated with lower albumin levels (F-stat: 1.921e+04, Adj R-squared: 0.9288, p-val: < 2.2e-16), decreasing hemoglobin concentration (F-stat: 1.434e+05, Adj R-squared: 0.9898, p-val < 2.2e-16), increasing levels of gamma-globulins (F-stat: 9.21e+04, Adj R-squared: 0.9843, p-val < 2.2e-16) and higher MCV (F-stat: 18.04, Adj R-squared: 0.01144, p-val 2.3e-05). Then, a significant anti-correlation between gamma-globulin and serum albumin levels was also confirmed (F-stat: 2.807e+04, Adj R-squared: 0.9501, p-val < 2.2e-16). Finally, we implemented two different multivariate logistic models using as independent variables the presence or absence of MGUS or CHIP, respectively. In these models, the presence of a monoclonal component was only positively associated with a higher level of gamma-globulins (Est: 0.16238, p-val = 0.00079) and not with increasing age (Est: 0.02680, p-val = 0.31819). Interestingly, the presence of an MGUS confirmed a tendency to an anti-correlation with the presence of CHIP (Est: -0.78959, p-val = 0.06468). On the contrary, CHIP as independent variable resulted significantly correlated with increasing age (Est: 0.07396, p-val = 0.00021). The spectrum of mutations in CHIP cases with or without MGUS was not significantly different, with DNMT3A and TET2 being the most frequently mutated genes.

Our study shows that, in a large cohort of oldest-old patients, CHIP and MGUS are not correlated but follow two seemingly independent patterns, showing a tendency to a mutual exclusivity and associating with different clinical and laboratory values. One limitation of our study is the skewing towards female sex, where MGUS is less prevalent, but this is explained by the cohort's median age knowing that females have a longer expected life. Based on this, our findings that CHIP but not MGUS increased with age in our cohort suggest different selective pressures in this extreme age range. These results warrant further investigation as to whether there could be age-specific drivers for MGUS, and their clinical relevance.