On the relevance of immunodeficiency evaluation in haematological cancer

To the editor, B cell lymphoproliferative disorders (B‐CLPD) are unique models to explore tumor‐induced immunodeficiency, since immune cells are both the agents and the targets of immune defects. Secondary immunodeficiency (SID) appears early as a consequence of tumor growth and immune escape mechanisms, and becomes more severe in parallel with clonal expansion. The immune deficiency also worsens with tumor treatment, rendering the patient vulnerable to serious, and often fatal, infections. An immunosuppressive microenvironment, orchestrated by malignant B cells and bone marrow stroma cells, assists in the survival, growth and dispersion of the monoclonal tumor, while simultaneously impairing normal B cell maturation, and therefore protective antibody production, as well as polyclonal CD4 and CD8 T lymphocyte responses. Innate immunity is also altered, with hypocomplementemia, skewed immature dendritic cells, defective natural killer cytotoxicity and expansion of myeloid‐derived suppressor cells and monocyte‐derived nurse‐like cells, among other changes. This raises the question: could we, through earlier diagnosis of SID, predict not only the risk of serious (and often fatal) infections as well as haematological cancer progression? The evolving SID results in serious infections that impact quality of life and prognosis and is the second leading contributor to mortality in B‐CLPD patients, as well as a major driver of health care costs. Unfortunately, the new therapies that target the malignant immune cells have infectious side effects, worsening the SID. Importantly, infections are an independent prognostic factor of clonal progression in chronic lymphocytic leukemia (CLL) and multiple myeloma (MM); however, there are no proven or approved biomarkers of SID in the current prognostic guidelines for B‐CLPD nor consensus on the evaluation of immune status, unless infections have occurred. Immune markers are not currently considered in deciding risk‐adapted therapies. So far, only significantly reduced serum immune globulin isotypes distinct from that the monoclonal protein (immunoparesis) are included in the prognostic score of smoldering myeloma, suggested by the Spanish group. The first detectable immune event in CLL, B cell lymphomas and MM, is the suppression of production of immunoglobulin heavy and light chains in isotypes contrary to those of the clonal B cells, called isotype‐matched hypogammaglobulinemia or immunosuppression (IMI). For instance, in an IgA‐kappa clonal expansion, decreased polyclonal IgA‐lambda production by the normal plasma cells can precede that of increased clonal M protein. Therefore, this specific immunodeficiency biomarker could be detected earlier than clonal progression which provides an early marker of the SID. Severe IMI, defined as a >50% reduction below the lower limit of normal of the respective immunoglobulins, has been shown to be an independent biomarker of both infectious risk and cancer progression in MGUS, MM, and CLL, although there is still scarce data in NHL. Conversely, recovery from IMI may indicate humoral reconstitution during remission after treatment. This may also provide an early and accurate biomarker of cancer progression which would be of great interest in premalignant stages, such as monoclonal gammopathy of undetermined significance (MGUS). Here, earlier identification of patients with risk of progression to MM (only 1% of total MGUS per year) could prompt treatment leading to a positive impact and better survival. Accurate SID biomarkers could aid in deciding the best timelines for aggressive diagnostic techniques, such as bone marrow aspirates or biopsy, for treatment initiation and adjustment, and for individualization in high‐risk patients with all lymphoid malignancies (Figure 1). In particular, the relevance of early diagnosis of CLL progression during the “watch and wait” phase, and of early treatment as well as the risk of serious infections, remains to be clarified. Previous studies showed only that prominent hypogammaglobulinemia (IgG <4 g/L) is associated with a risk of pneumonia in B‐ CLPD and low IgA and IgG have been related to significantly reduced survival in CLL. While immunoglobulin replacement therapy has long been used in B‐CLPD for SID and autoimmune complications, hypogammaglobulinemia has been questioned as a reliable biomarker of infectious risk in B‐CLPD. A recent study, using a machine learning approach, in CLL to predict high risk of infection by using 188 parameters, including serum IgG/IgA/IgM (but not immunisation responses) concluded that low levels did not have a major impact and so these measurements did not enter into their high infectious risk algorithm. However, functional assays, such as antibody responses to immunization, as used in primary immune deficiencies, are a better surrogate biomarker of infectious risk. The recent EMA indication to initiate immunoglobulin treatment in such SID highlights the