First, we briefly review the current evidence for the immunological approach to the prevention of infectious complications associated with defective antibody responses in CLL and MM. We then survey the current challenges derived from the improvements in diagnosis and therapy of hematological malignancies and discuss some practical issues. Finally, we consider the potential indices for better selection of patients most likely to benefit from early intervention with IVIg or SCIg.

Not all patients with hypogammaglobulinemia present with infectious complications, so a poor response to pneumococcus was suggested as a good predictor of infections (13, 31); furthermore, serious infections can occur in the absence of hypogammaglobulinemia due to other reasons such as neutropenia or T cell suppression. Low baseline levels of specific exposure antibodies against various bacterial, viral, fungal, and protozoan pathogens have been described in these disorders (44). B-cell antimicrobial dysfunction increases progressively from monoclonal gammopathy of undetermined significance (MGUS) to Waldenstrom macroglobulinemia to MM (31, 44). In therapy-naive CLL patients, seroconversion from positive to negative IgG values was noted for EBV- (3.6%) and, most frequently, for VZV-specific IgG (18%), while IgG specific for CMV was preserved (45). Significantly, lower responses to vaccination in CLL patients with respect to healthy subjects have been reported against diverse antigens. Specific antibody production to polysaccharide antigens (T-cell-independent response), such as classical 23-valent pneumococcal vaccine (PPV23), is markedly impaired in these patients (13, 31, 46). Defective response to protein antigens, such as tetanus toxoid and influenza virus, is also apparent (47, 48). Interpretation of vaccine responses is complex and requires diverse clinical considerations (49). For the evaluation of primary responses, an increase greater than threefold after 4 weeks with respect to prevaccination levels is considered normal (49). The use of Salmonella typhi Vi vaccine (50) with pure polysaccharide extract may add clinical value in this population.

To evaluate the role of immunological deficiencies and to monitor patients with hematological malignancy, a complete clinical history of infections is recommended at diagnosis and during follow-up, as well as quantification of serum immunoglobulins (51) and circulating lymphocyte subsets, including CD4 and CD8 T cells as well as B cells (provided the B cell count in CLL is not excessively high) (Table 2). Neutrophil counts should be also regularly monitored.

A recent review by Dhalla et al. (9) has highlighted the relevant role of routine immunological evaluation for secondary specific antibody deficiency to protein and polysaccharide immunizations in CLL as a method for predicting patients prone to infections. These responses should be monitored every 6–12 months and after significant bacterial infections or immunosuppressive therapy, and this approach could be extended to other hematological malignancies.

IgG subclass evaluation could be useful. In a large series of CLL patients, subclass deficiency (particularly IgG3 and IgG1 subclass deficiency) better correlated with recurrent or significant infections than hypogammaglobulinemia itself (100% of IgG subclass deficiency versus 50% of hypogammaglobulinemia, respectively) (52). In another study, decreased concentrations of IgG4 and IgG2 were associated with increased susceptibility to infection (17). However, other studies have not shown association between IgG subclass deficiency and infection in CLL (53).

A recent study showed more serious infections in secondary than in primary antibody deficiency patients and similar diagnostic delay and incidence of bronchiectasis (54). For early detection of preventable lung involvement, pulmonary function tests and high-resolution computerized lung tomography are essential to prevent development and/or progression of bronchiectasis (9). Our strong recommendation is to always consult a clinical immunologist for performing immunological evaluation.

Current front-line treatment strategies for CLL including combination chemotherapy and biologic agents, such as fludarabine plus cyclophosphamide plus Rtx (FCR) or bendamustine plus Rtx (BR) or alemtuzumab, have greatly improved overall survival, complete response rate and progression-free survival (61–65). However, it has been suggested that these combination regimens increase synergistically myelosuppression and immunosuppression, which comes to the prize of an increased risk of infections (66, 67). Severe or unusual infections, with higher rates of global infections compared with the historical group of patients treated with FC alone but without significant influence in infection-related mortality have been reported (62). Particularly, the growing use of Rtx in front-line combination chemotherapy with fludarabine (FCR) increases the risk of all types of infections with respect to Rtx in combination with cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP) or with cyclophosphamide, vincristine, and prednisone (R-CVP) chemotherapy (68). FCR regimen results in a significant myelosuppression and high rates of early and late infections, especially in elderly patients (61). With respect to maintenance therapy, Rtx alone for 2 years after induction therapy was associated with higher rate of all types of infections than the observation alone group (66 versus 50%) without influence in infection-related mortality (69). Anti-CD20 monoclonal antibodies have been associated with CMV infections and other virus reactivations (hepatitis B and polyomavirus JC) (63, 70, 71). Reactivation of herpes viruses have also been described related to treatment with purine analogs (as Flu) and alemtuzumab (8). With FR combination but also alemtuzumab, opportunistic infections by Pneumocystis, Listeria, mycobacteria, and Candida have become a real concern and occur frequently during the first 3 months of therapy and in previously treated patients (7, 72–74). Second- and third-generation anti-CD20 monoclonal antibodies are humanized or fully human antibodies used in CLL and indolent NHL with either higher complement-dependent cytotoxicity or antibody-dependent cell-mediated cytotoxicity or direct cytotoxic effects than Rtx (75), but their infections profile in the long-term respect to that of Rtx remain to be explored. Obinutuzumab use for hematological malignancy was associated with severe infection in 7% of patients in a recent clinical trial (76).

The impact of new biologicals for hematological malignancies that inhibit B-cell receptor (BCR) signaling pathway, B cell leukemia/lymphoma-2 (Bcl-2) antagonists, and chimeric antigen receptor (CAR) T cells in terms of long-term infectious complications has not been established. BCR inhibitors’ approved to date include the Bruton’s tyrosine kinase (BTK) inhibitor ibrutinib; the selective inhibitor of the phosphatidylinositol 3-kinase (PI3K) delta isoform idelalisib; as well as Syk inhibitor fostamatinib disodium (FosD) and SRC family kinases inhibitors (dasatynib, bafetinib). A recent study reports ibrutinib to have a better impact on humoral reconstitution and hence lower infection rates at 12 months (77). However, in a large multicenter, open-label, clinical trial evaluating ibrutinib versus anti-CD20 (ofatumumab) therapy in refractory CLL, severe infections (mainly pneumonia, urinary tract infections, and pyrexia) were similar in both groups (24 versus 22% of patients), with higher infections in general in the ibrutinib group (70 versus 54%, respectively), although patients on ibrutinib had longer drug exposure (median, 8.6 versus 5.3 months) and more than 3 months longer adverse event reporting period (78). The rate of severe infections was lower when ibrutinib was given as first-line therapy (4% pneumonia and 4% diarrhea) (79). In the clinical trial with idelalisib for refractory CLL reported infections included pneumonia in 6% and febrile neutropenia in 5% (80); another clinical trial with idelalisib for indolent refractory NHL included severe neutropenia in 27% and pneumonia in 7% (81), while a clinical trial on treatment-naive CLL patients reported 18% incidence of pneumonia (82). Bcl-2 antagonists, such as venetoclax, have shown in a recent phase II clinical trial in refractory CLL an incidence of pneumonia in 6% and febrile neutropenia in 5% of patients (83). In summary, despite promising clinical efficacy and acceptable adverse effects’ profiles, these later agents inhibit key molecules in B cell activation and differentiation and also other relevant pathways independent of BCR and neutrophils. Thus, the real clinical extent of the humoral and cellular immune and innate defect of these new agents at the long term, as shown by the primary defects of these molecules (84, 85) remains to be determined.

On the other hand, CD19 CAR T cells have been used in relapsed and high-risk CLL with overall response rate of 57% of 14 patients and persistent B cell aplasia in 100% of patients with complete remission (CR) and half of the patients with partial remission (86, 87). One out of the 4 patients who achieved CR died of an infectious surgery complication almost 2 years after CAR T therapy. Severe hypogammaglobulinemia in these patients was managed with IVIg (86). Interestingly, long-lived plasma cells may account for the persistence of secondary responses to previous exposures (88).

Similar front-line treatment regimens as for CLL (mostly FluR combination) and novel biologic agents are being used in iNHL, mostly in the management of follicular lymphoma (FL), mantle-cell lymphoma (MCL), and cases of diffuse large cell lymphoma (DLCL), leading to substantial improvement in prognosis and overall survival (89–92).

The issue of immune hyporesponsiveness during specific intensive chemotherapy is still unexplored. Secondary antibody responses do not seem to be invariably impaired after anti-CD20 monoclonal antibodies, but primary responses, as tested in numerous immunization studies, are clearly impaired (93, 94). Rtx has been associated with prolonged hypogammaglobulinemia (95, 96) and with neutropenia (even with delayed-onset) (97), although difficult to quantify due to confounding factors, namely, concomitant use of these biologicals with other immunosuppressive or chemotherapeutic agents and the underlying conditions, as well as under-reporting (98). The experience learned from Rtx for autoimmune disease demonstrates that the lower the baseline IgM and IgG levels, the lower post-treatment IgM and IgG levels, the higher infection risk (96, 99), with an accumulative effect after repeated cycles (100).

Protecting immunocompromised patients against vaccine-preventable infectious diseases is a commonly missed opportunity (101). Although there are no randomized studies showing that vaccination may alter infection rates or outcomes from acquired infections in CLL, routine vaccinations should be maintained in these patients before initiation of treatment if possible (39, 102). All live vaccines are contraindicated, including zoster, since severe and even fatal complications have been reported. Diverse studies have shown reasonable rates of seroprotection and seroconversion in various immunocompromised hosts, including oncology patients, with very minimal downside (101). Conjugate vaccines have proved to be highly immunogenic and are to be recommended in these patients (46, 103).

Therefore, subunit vaccines against seasonal influenza and against H1N1 are broadly recommended across oncologic patient populations, given the severity of the H1N1 pandemic and the highly severe flu impact respect to general population despite poor responses in CLL, and even two doses regimen (48, 101, 104–107).

Streptococcus pneumoniae is the commonest pathogen in hematological malignancy, and invasive disease is significantly higher in MM and more modestly in CLL and NHL. As discussed earlier, although vaccination of seven-valent pneumococcal vaccine show protection to six of the seven antigens in 40% in series of CLL patients when administered early and prior to therapy (108), there is no data on infection prevention by such immunization in CLL. Despite this, it is now recommended to administer both the 13-valent and the 23-valent vaccine (107, 109), since 58% of untreated CLL patients showed a response (103).

A moderate vaccination response rate of 43% against H. influenzae type b (Hib) conjugate vaccine among adult and elderly patients with CLL has been reported (110) but will only protect against type b H. influenzae and not against the much more common infection by the non-encapsulated pathogen. Normal IgG1 and IgA concentrations were associated with protection, while IgM, in turn, was the best predictor of a significant vaccination response.

Patients receiving intensive chemo- or immunotherapy should be screened for hepatitis B and C infection (grade of evidence A1). Hepatitis B vaccine is indicated for patients with CLL who are negative for antibodies against HBsAg; the combined hepatitis A and B vaccine should be considered (111) in endemic regions. The HPV vaccine may be considered for all HPV-seropositive subjects with hematological malignancies, as these malignancies have been associated with high risk of second primary malignancies in general (112, 113), and with HPV in particular in a recent study (not only skin cancer but also prostate and colon cancer cells were HPV DNA by PCR) (114).

Given the suboptimal immune responses to immunization of immunocompromised patients, the optimal timing dosing, use of adjuvants, and delivery method might maximize the immunologic benefit of vaccination in oncology patients (101). A potential beneficial strategy could be to vaccinate patients early at diagnosis, for instance at MBL stage, when better responses are attained than CLL (115). The same might be true for MGUS patients at risk of developing MM.

The changing spectrum of infections in CLL patients on current therapeutic protocols mandates a newer approach to prophylaxis and therapy. Early recognition of patients susceptible to infection and prophylactic administration of appropriate antibiotics remain the first-line management for symptomatic antibody deficiency in CLL (9, 21, 22).

However, clinical trials are lacking (51). In patients with bronchiectasis, nebulized or low dose oral antibiotics, such as azithromycin, can reduce the incidence of recurrent infection. Replacement Ig should be considered if, despite AB prophylaxis, there are significant breakthrough bacterial infections or if bronchiectasis develops or progresses (51).

Anti-infective prophylaxis is recommended for patients receiving purine-analog and/or alemtuzumab during treatment and thereafter, if tolerated herpes virus (acyclovir or equivalent) and anti-Pneumocystis jirovecii prophylaxis (sulfamethoxazole/trimethoprim or equivalent) (Grade of evidence IV, B) (39, 51, 107, 116). Antivirals are given if the patient is on chemotherapy (FC or FCR) and thus at high risk of opportunistic viral infection. Anti HSV and HZV prophylaxis is recommended in patients requiring intensive and/or immunosuppressive treatment who are seropositive, have a low CD4 count or a history of previous herpes infections (51). The duration of anti-Pneumocystis and herpes prophylaxis is controversial.

Patients on alemtuzumab should be monitored for CMV reactivation (117–119). The current management is controversial; some use ganciclovir (oral or IV) prophylactically if viremia is present, others use ganciclovir only if viral load is rising. CMV viremia should be measured by PCR quantitation at least every 2–3 weeks. Pre-emptive ganciclovir/valganciclovir may be used with CMV viremia or increasing viral load, for 14–21 days until symptoms resolve and PCR tests are negative. Patients’ positive for HBSAg or HBCAg may require antiviral treatment and should be managed jointly with a specialist in viral hepatitis (120).

The myeloid growth factor, granulocyte colony-stimulating factor (G-CSF) or granulocyte macrophage colony-stimulating factor (GM-CSF), has radically changed the approach to the prevention of febrile neutropenia (121–123). Their use is recommended for the first cycle of chemotherapy for patients with more than a 20% risk of febrile neutropenia or in case of poor response to antibiotics with persistent neutropenia should lead to the use of these cytokines.

Given that therapy of hematological malignancies has substantially changed in the last decade with the introduction of drugs with different modes of action, including monoclonal antibodies, drugs with immunomodulatory effects such as lenalidomide, and targeted drugs such as tyrosine kinase inhibitors, the previous clinical trials may not accurately represent the drawbacks and benefits of IVIg/SCIg now.

Current guidelines state the use of IVIg in severe hypogammaglobulinemia and recurrent infections (Level of evidence I, A) (107). The question of the correct dose of IVIg to prevent bacterial infections is still unresolved and is probably patient dependent (9, 37, 124). Much has been learnt from the management of patients with PID, such as the use of clinical measures and trough IgG levels to guide dose adjustments and the need to use higher replacement doses in patients with bronchiectasis (124). As a general guideline, maintenance of trough serum IgG in treated patients above 500–700 mg/dL is a reasonable goal (9).

Treatment should be reconsidered if there is no improvement in the frequency or severity of bacterial infections after 1 year (125). The choices of subcutaneous Ig and self-infusion have reduced costs and improved quality of life (9, 126).

The use of IVIg prophylaxis in hematopoietic stem cell transplantation (HSCT) for MM patients without selection by immunological parameters has shown beneficial results in terms of quality of life, but does not seem to affect CMV infection incidence or other infectious complications (127–129).

Expected mechanisms of action of replacement Ig therapy on B cell cancer include effective reduction in infections by pathogen neutralization, toxin inactivation and opsonization, and complement-mediated bactericidal effects (130). The potential immunomodulatory effects of Ig therapy on the setting of cancer are unclear (131). Ig could induce in vivo beneficial immunomodulatory effects on the malignant B cell clone by several potential mechanisms: inducing unresponsive anergy and subsequent apoptosis in B cells, by inducing proapoptotic molecules expression on tumor cells or by Fas-blocking antibodies in IVIg (132) or through ERK phosphorilation in vitro (133); altering activation and proliferation of B cells (132, 134); inhibition of B cell antigen presentation (135, 136); in vitro differentiation of B cells and Ig secretion (137); and other immune effects [revised by Corbi et al. (131)].

Increases in the immunodeficiency status of CLL and other hematological malignancies, and growing antibiotic resistant pathogens in what is being called the post-antibiotic era, have renewed attention in antibody therapy. In carefully selected patients by clinical infectious history and antibody production defect (9), the criteria for initiating IVIg therapy need to be reassessed in the light of data of PID management to prevent lung damage due to ongoing subclinical infection (138, 139). In PID, replacement IVIg therapy is the mainstay of treatment of subjects with antibody failure, and it has revolutionized the lives of these patients (139).

Futures studies should address the question whether in CLL patients requiring long-standing anti-infectious combined protocols with antiviral, antibiotic, and antifungal treatment, IVIg would add a benefit, given the occurrence of pathogens’ antibiotic resistance, deleterious effects on microbiota and side-effects. Also, comparison studies of the impact of IVIg versus antibiotics with current protocols in terms of lung involvement and prognosis are necessary. In the absence of specific studies, recommendations for IVIg replacement in hematological malignancies should be based on clinical experience in PIDs and only in selected individual patients.