Should mild hypogammaglobulinemia be managed ...

EJINME-02791; No of Pages 6 European Journal of Internal Medicine xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

European Journal of Internal Medicine journal homepage: www.elsevier.com/locate/ejim

Original Article

Should mild hypogammaglobulinemia be managed as severe hypogammaglobulinemia? A study of 389 patients with secondary hypogammaglobulinemia Mathieu Blot a, Pierre Boyer b, Maxime Samson a,c,d, Sylvain Audia a,c,d, Hervé Devilliers c,e, Vanessa Leguy a, Sabine Berthier a, Jean-François Besancenot c,e, Bernard Lorcerie a,c, Daniela Lakomy b,c, Bernard Bonnotte a,c,d,⁎ a

Department of Internal Medicine and Clinical Immunology, Dijon University Hospital, 21079 Dijon, France Laboratory of Immunology, Dijon University Hospital, 21079 Dijon, France c Faculty of Medicine, University of Burgundy, Dijon, France d INSERM, UMR1098, University of Franche-Comté, 25030 Besançon, France e Department of Internal Medicine and Systemic Diseases, Dijon University Hospital, 21079 Dijon, France b

a r t i c l e

i n f o

Article history: Received 23 April 2014 Received in revised form 21 July 2014 Accepted 15 September 2014 Available online xxxx Keywords: Hypogammaglobulinemia Etiology Infection Vaccination

a b s t r a c t Background: Although secondary hypogammaglobulinemia is more frequent than primary hypogammaglobulinemia, its etiology and management are poorly described, particularly for mild hypogammaglobulinemia. Methods: This retrospective observational study included all adult patients with a gammaglobulin level b6.4 g/L on serum electrophoresis identified at Dijon teaching hospital between April and September 2012. Clinicobiological features, etiologies and infectious complications were collected at inclusion and compared between group 1 (gammaglobulin b5 g/L, severe hypogammaglobulinemia), and group 2 (gammaglobulin b6.4 and ≥5 g/L, mild hypogammaglobulinemia). Results: Among the 4011 serum electrophoreses, 570 samples from 389 patients had gammaglobulin levels below 6.4 g/L: 156 (40%) in group 1 and 233 (60%) in group 2. Mean age ± SD was 67 (15) years, and sex ratio was 1.04 (M/F) with no difference between the two groups. An etiology was identified in 79% and 58% of patients in groups 1 and 2, respectively (p b 0.0001). The main etiologies were similar in both groups and included malignant hemopathy treated with cytostatic agents (n = 129, 33%), smoldering or newly-diagnosed hemopathy without treatment (n = 49, 13%) and immunosuppressive treatment (n = 91, 23%). The incidence of hypogammaglobulinemia-related infections was 22/100/year, with no significant difference between the two groups (p = 0.17). Vaccination coverage against pneumococcus was 33%, and higher in group 1 (46% vs. 24%; p b 0.0001). When no cause was known at inclusion, an etiology was discovered in 22/130 patients (17%), 11 in each group. Conclusions: Though mild hypogammaglobulinemia does not meet the classical criteria for hypogammaglobulinemia (b5 g/L), the etiology and infectious risk are similar. It therefore requires investigation and vaccination. © 2014 European Federation of Internal Medicine. Published by Elsevier B.V. All rights reserved.

1. Introduction Hypogammaglobulinemia diagnosed on serum protein electrophoresis (SPE) is a diagnostic challenge for clinicians. After confirmation on a second electrophoresis, it is essential to identify a cause and prescribe a treatment when possible [1,2]. Hypogammaglobulinemia can be due to a variety of conditions, which can be divided into immunoglobulin production diseases such as lymphoid hemopathies (most commonly chronic lymphocytic leukemia (CLL) and multiple ⁎ Corresponding author at: Département de Médecine Interne et Immunologie Clinique, Hôpital du Bocage, CHU de Dijon, 2 Boulevard du maréchal de Lattre de Tassigny, 21079 Dijon Cedex, France. Tel.: +33 3 80 29 34 32; fax: +33 3 80 29 38 46. E-mail address: [email protected] (B. Bonnotte).

myeloma (MM)), infections (human immunodeficiency virus (HIV), parvovirus B19, Epstein–Barr virus (EBV), cytomegalovirus (CMV)) [1, 2], medications (mainly antiepileptic and immunosuppressive therapies (IST)) and immunoglobulin loss diseases such as enteropathies or renal amyloidosis. When no cause is identified, a diagnosis of primary immune deficiency, most frequently common variable immunodeficiency (CVID) or Good syndrome in adults, has to be made [1,2]. While the incidence, findings and infectious consequences of CVID are well documented [3,4], little is known about the epidemiology, management and complications of secondary hypogammaglobulinemia, although these conditions are more frequent than primary hypogammaglobulinemia. Secondary hypoglobulinemia is found in 27% to 52% of CLL, depending on the definition of hypogammaglobulinemia and the stage and duration of CLL [5,6]. Under anti-CD20 therapy,

http://dx.doi.org/10.1016/j.ejim.2014.09.009 0953-6205/© 2014 European Federation of Internal Medicine. Published by Elsevier B.V. All rights reserved.

Please cite this article as: Blot M, et al, Should mild hypogammaglobulinemia be managed as severe hypogammaglobulinemia? A study of 389 patients with secondary hypogammaglobulinemia, Eur J Intern Med (2014), http://dx.doi.org/10.1016/j.ejim.2014.09.009

2

M. Blot et al. / European Journal of Internal Medicine xxx (2014) xxx–xxx

38.5% of patients has transient hypogammaglobulinemia, essentially depending on the duration of treatment [7]. However, hypogammaglobulinemia may promote infections. About 38% of patients suffering from invasive encapsulated Streptococcus pneumoniae or Haemophilus influenzae infections presents hypogammaglobulinemia [8]. Therefore patients with hypogammaglobulinemia should be vaccinated against S. pneumoniae and influenza [9]. Intra-venous immunoglobulin (IVIg) therapy may be administered in secondary hypogammaglobulinemia only when severe and recurrent infections occur, especially in CLL and MM, even if this treatment is still debated [10,11]. Furthermore, the definition of hypogammaglobulinemia is still discussed even though a serum level of immunoglobulin (Ig) under 5 g/L has been arbitrarily defined for CVID diagnosis [2,12]. However, the lower limit of normal for gammaglobulin (γGb) on SPE is usually higher, around 7 g/L, depending on the manufacturer of the SPE kit [6, 7]. In addition, little is known about patients who present mild hypogammaglobulinemia, and there is still no consensus on how such patients should be managed. In this study, we compared etiologies, infectious complications and management (diagnostic strategy and vaccination coverage) of hypogammaglobulinemia in group 1 (severe hypoγGb under 5 g/L) and in group 2 (mild hypoγGb from 5 to 6.4 g/L).

2. Methods

2.2. Data collection Common causes of hypogammaglobulinemia such as hematological diseases (MM, CLL, B-cell lymphoma, AL amyloidosis, monoclonal gammopathy of unknown significance (MGUS)), drugs, Ig leakages (nephrotic syndrome, protein-losing enteropathy, capillary leakage), and infections (HIV, EBV, CMV, parvovirus B19) [1,2,13] were identified. Demographic data, biological findings, vaccination against S. pneumoniae and intravenous Ig (IVIg) therapy were recorded. Infections that occurred during the year before the inclusion were analyzed and ear nose throat (ENT), respiratory or meningeal infections were considered related to the hypogammaglobulinemia. For patients for whom no etiology of hypogammaglobulinemia was identified, the initial clinical, biological presentation and complementary investigations that were performed to find a cause were collected. Blood tests (albumin, creatinine, calcemia, hemoglobin count, SPE, immunoglobulin test, serum immunofixation electrophoresis (IFE), kappa/lambda light chain ratio, lymphocyte immunophenotyping), urinary exams (proteinuria (g/day), urinary electrophoresis and IFE), medullogram, osteomedular biopsy and imaging (tomodensitometry and positron emission tomography scan) were collected. Data from the two groups were compared: patients with γGb level b5 g/L (group 1: severe hypogammaglobulinemia) and those with γGb level ≥ 5 and b6.4 g/L (group 2: mild hypogammaglobulinemia). The study was outside the Huriet act, and in accordance with the local ethics committees of Dijon University Hospital.

2.1. Study design 2.3. Statistics A retrospective, single-center, observational study of all adults with hypogammaglobulinemia under 6.4 g/L, which was the main French national threshold, recorded between April 1st and September 31st 2012 in the biochemistry laboratory of a 1772-bed teaching hospital (Dijon CHU—France) was performed. A second electrophoresis was recorded when available. Clinical and biological data as well as the etiologies of the hypogammaglobulinemia were collected at the time of inclusion. Patients diagnosed with primary immunodeficiency or “false” hypogammaglobulinemia (cryoglobulinemia, hemodilution) were excluded. Serum levels of γGb were analyzed by SPE in the immunology laboratory of Dijon CHU on a Sebia Hydrasys system (Sebia, Evry, France) with agarose gel reagents (Hygragel ref 4140, Sebia). Total protein concentrations were assessed using a modified biuret reaction on a Dimension Vista analyzer (Siemens Healthcare Diagnostics, Deerfield USA).

Fig. 1. Flow chart showing the selection of patients with secondary hypogammaglobulinemia.

Continuous variables were expressed with means ± standard deviations (SD), and categorical variables with frequencies and Table 1 Demographic data and spectrum of etiologies of 389 adult patients with hypogammaglobulinemia: comparison of group 1 (gammaglobulin level b5 g/L) and group 2 (gammaglobulin level ≥5 g/L and b6.4 g/L). Total

Group 1

Group 2

p values

n = 389 n = 156 n = 233 Demographic data Age, mean (SD) Male Known etiology Etiology discovered after investigation No etiology Hemopathy without IST Multiple myeloma Lymphoma Chronic lymphocytic leukemia MGUS Others Hemopathy with chemotherapy or IST Treated multiple myeloma Treated lymphoma Treated chronic lymphocytic leukemia Treated AL amyloidosis Others Drug Corticoid only IST in association Cancer chemotherapy Anti-epileptic drug Leakage of immunoglobulin Nephrotic syndrome Protein-losing enteropathy Infection HIV EBV

67 (15) 198 (51) 259 (67) 22 (6) 108 (28) 49 (13) 16 (4) 12 (3) 9 (2) 9 (2) 3 (1) 129 (33) 89 (23) 21 (5) 10 (3)

68 (13) 81 (52) 123 (79) 11 (7) 22 (14) 24 (15) 9 (6) 5 (3) 7 (5) 2 (1) 1 (1) 74 (47) 51 (33) 9 (6) 8 (5)

67 (16) 0.87 117 (50) 0.74 136 (58) b.0001 11 (5) 0.33 86 (37) b.0001 25 (11) 0.18 7 (3) 0.18 7 (3) 1.0 2 (1) 0.03 7 (3) 0.33 2 (1) 1.0 55 (24) b0.0001 38 (16) b0.0001 12 (5) 0.79 2 (1) 0.02

6 (2) 4 (1) 92 (24) 40 (10) 46 (12) 5 (1) 1

4 (3) 2 (1) 28 (18) 14 (9) 11 (7) 3 (2) 0

2 (1) 2 (1) 64 (28) 26 (11) 35 (15) 2 (1) 1

0.22 1.0 0.03 0.49 0.02 0.39

8 (2) 1

6 (4) 1

2 (1) 0

0.06

1 1

1 0

0 1

All data are represented as number (%), except age: mean (SD). Abbreviations: MGUS: monoclonal gammopathy of undetermined significant; IST: immunosuppressive therapies; HIV: human immunodeficiency virus; EBV: Epstein–Barr virus.



percentages. Continuous data were compared using the Mann–Whitney U-test and categorical data were compared using the chi-square test and Fisher's exact test when appropriate. Statistical analyses were computed using SPSS v19 software. For all statistical analyses, a 2tailed p ≤ 0.05 was considered significant. 3. Results 3.1. Epidemiology of secondary hypogammaglobulinemia From the 4011 SPE performed between April and September 2012, a γGb level under 6.4 g/L was found in 570 samples from 389 patients: 156 (40%) in group 1 and 233 (60%) in group 2 (Fig. 1). A second electrophoresis was available in 226 patients (58%), enabling to confirm hypogammaglobulinemia in 189 patients (49%). The main characteristics of these patients are summarized in Table 1. At inclusion, the mean age ± SD was 67 ± 15 years and the sex ratio was 1.04 (M/F) with no difference between the two groups. 3.2. Etiological spectrum of secondary hypogammaglobulinemia The distribution of the 389 patients according to their level of γGb and the identification of an etiology are represented in Fig. 2. The number of patients with no etiology was higher in the group of mild hypogammaglobulinemia compared to the group of severe hypogammaglobulinemia. However, an etiology of the hypogammaglobulinemia has been identified in most of the patients. In addition, an etiology was discovered after investigation in a few cases, not only for patients in group 1, but also for patients in group 2. The etiology of hypogammaglobulinemia was more frequently identified in group 1 than in group 2 (79% versus 58%; p b 0.0001) (Table 1). Among the three main etiologies identified, malignant hemopathies treated with cytostatic agents were more frequent in group 1 than in group 2 (47% versus 24%; p b 0.0001). Conversely, IST, mainly represented by glucocorticoids alone or by a combination of immunosuppressants, were less frequently identified in group 1 than

3

in group 2 (18% versus 28%; p b 0.03) whereas the proportion of newly diagnosed or smoldering hemopathies was similar in the two groups (15% versus 11%; p = 0.18). Two hundred and seven patients of the 389 (53%) had received glucocorticoids. Among the 86 patients treated with immunosuppressants, 28 (33%) were treated with corticosteroids ± immunosuppressants for a kidney or cardiac transplantation and 41 (48%) for an autoimmune or inflammatory disease (Table 2). Notably, 31 out of the 41 patients (76%) suffering from an autoimmune or an inflammatory disease had no hypogammaglobulinemia before the treatment onset arguing in favor of the role of IST and against the role of the autoimmune disorder in the etiology of hypogammaglobulinemia. For the 10 others, data were not available but underlying diseases were not classically known to trigger hypogammaglobulinemia (Rheumatoïd Arthritis (n = 5), Giant Cell Arteritis (n = 2), PolyMyalgia Rheumatica (n = 1) and other inflammatory rheumatisms (n = 2)). 3.3. Incidence of infections and vaccination coverage The incidence of hypogammaglobulinemia-related infections during the year before the inclusion was high (22/100/year) and not significantly different between the two groups (p = 0.17). Pulmonary infections accounted for 39% of all infections and 78% of hypogammaglobulinemia-related infections, without significant difference between the two groups (p = 0.12). Immunization against S. pneumoniae was low (33%), but more frequent in group 1 than in group 2 (46% versus 24%; p b 0.0001) (Table 3). Furthermore, none of the patients with hypogammaglobulinemia for which no etiology had been identified was vaccinated, while 124 of the 178 patients (70%) with hypogammaglobulinemia related to hemopathy were vaccinated against pneumococcus. Immunoglobulin replacement therapy was used in six patients: one because of an infection due to parvovirus B19 and five because of recurrent severe infections (pulmonary (n = 4) and ENT (n = 1)) during the evolution of a hematological disease (2 CLL, 2 B-cell lymphoma, 1 MM). 3.4. Etiological exams carried out and follow-up of the patients A second electrophoresis was not available in 163 patients out of 389 (42%). However, the etiology of hypogammaglobulinemia was already identified for 76 patients out of 389 (20%) (immunosuppressive treatment (n = 42), hematological disease (n = 28) and nephrotic syndrome (n = 5)). For the 87 remaining patients out of 389 (22%), only one electrophoresis had been performed and no cause of hypogammaglobulinemia was found or sought. Of note, sixty seven of these patients had a gammaglobulin level between 5 and 6.4 g/L and no etiological exam was performed for 54 patients (62%). Of the 130 patients with hypogammaglobulinemia (33%) for which no etiology was identified at the time of inclusion, at least one etiological examination was performed in 60/130 patients (46%), with no

Table 2 Underlying diseases.

Fig. 2. Distribution of 389 patients with hypogammaglobulinemia according to the level of gammaglobulin.

Autoimmune or inflammatory diseases Rheumatoïd Arthritis PolyMyalgia Rheumatica/Giant Cell Arteritis Others inflammatory rheumatisms Systemic vasculitis Connective tissue diseases Idiopathic thrombocytopenic purpura Others Kidney or cardiac transplant Miscellaneous causes

Total

Group 1

Group 2

n = 86

n = 25

n = 61

41 (48) 11 9 5 6 4 2 4 28 (33) 17 (20)

10 (40) 3 1 2 2 1 0 1 6 (24) 9 (36)

31 (51) 8 8 3 4 3 2 3 22 (36) 8 (13)

All data are represented as number and main data as number (%).


4


Table 3 Infectious complications, vaccination coverage and intravenous immunoglobulin therapy in 389 patients with secondary hypogammaglobulinemia. Total

Group 1

Group 2

p values

n = 389 n = 156 n = 233 Incidence of infections during the year before the inclusion (n/100/year): All infections Hypogammaglobulinemia-related infections – Pulmonary tract Vaccination against Streptococcus pneumoniae IVIg substitutive therapy

Finally, 108 patients had no cause discovered or sought. Gammaglobulin levels were monitored in only 21/108 of these patients (19%) and remained low (b6.4 g/L) in 5 patients (24%). None of the patients with persistent hypogammaglobulinemia had recurrent infections. 4. Discussion

43 22

49 26

39 18

16 23 128 (33) 72 (46)

13 56 (24)

6 (2)

1

5 (3)

0.10 0.17 0.12 b0.0001 0.04

Incidence of infection is represented as number of infections/100 patients/year and vaccination and IVIg as number (%). Abbreviations: ENT: ear nose throat; IVIg: intra-venous immunoglobulin.

significant difference between the two groups (58% versus 42%; p = 0.13). Proteinuria was checked in only 42 cases (32%), and was more often high (≥ 3 g/day) in group 1 than in group 2 (54% versus 10%; p = 0.005). Other complementary examinations such as serum IFE, free light chain assay, lymphocyte immunophenotyping, medullogram and imaging were rarely performed but equally effective at finding a cause in both groups (p = NS for all) (Table 4). In fact, the examinations were more effective in identifying an etiology for patients in group 1 than those in group 2 (33% versus 11%; p = 0.004). An etiology was identified in 22 patients (17%), 11 patients in group 1: 6 nephrotic syndromes (1 B-cell lymphoma, 1 parvovirus B19 primo-infection, 1 idiopathic membranous glomerulonephritis, 1 metastatic melanoma, 1 decompensated diabetic glomerulosclerosis and 1 of unknown origin), 3 MM and 2 B-cell lymphoma; and 11 patients in group 2: 3 MGUS, 2 MM, 2 nephrotic syndromes (1 Waldenström macroglobulinemia, 1 decompensated diabetic glomerulosclerosis), 1 CLL, 1 B-cell lymphoma, 1 acute leukemia and 1 acute systemic EBV infection.

Herein, we described 389 adult patients affected by secondary hypogammaglobulinemia and recruited in a French teaching hospital. The gammaglobulin level was below the cut-off (b 6.4 g/L) in 14.2% of the 4011 SPE carried out, highlighting the fact that secondary hypogammaglobulinemia is a common condition, especially in the elderly. At the time of inclusion, the cause of the hypogammaglobulinemia was identified in almost three quarters of our patients, and a result of two main etiologies: immunosuppressive drugs and glucocorticoids on one hand and hematologic diseases on the other hand. These accounted for 96% of causes and were sometimes associated. More than half of the patients with hypogammaglobulinemia had been treated with glucocorticoids, which emphasizes the considerable responsibility of corticoid therapy in inducing hypogammaglobulinemia. Even if autoimmune or inflammatory diseases can be the causes of hypogammaglobulinemia, the involvement of the IST is highly probable since in most of the cases, hypogammaglobulinemia does not exist before IST treatment. Indeed, it is already known that glucocorticoid therapy affects plasma level of IgG and IgA but not IgM, and induces hypogammaglobulinemia in about 12% of treated patients, with a daily dose superior to 12.5 mg/day of equivalent prednisone [14–16]. Interestingly, patients with isolated lymphoid hemopathy were equally distributed in both groups. These results should prompt clinicians to investigate patients with mild hypogammaglobulinemia, since it could precede a diagnosis of hemopathy by as much as 9.8 years [17].

Table 4 Characteristics of 130 adult patients with hypogammaglobulinemia without an identified cause at inclusion: comparison of group 1 (gammaglobulin level b5 g/L) and group 2 (gammaglobulin level ≥5 g/L and b6.4 g/L).

Demographic data Age, mean (SD) Male Symptoms at inclusion Fever General physical health deterioration Pain Biological tests Plasma albumin b 30 g/L Plasma creatinine ≥ 150 μmol/L Hemoglobin b 10 g/dL Calcemia N 2.52 mmol/L Etiological examinations performed Serum IFE Monoclonal protein Proteinuria tests Proteinuria ≥ 3 g/day Urinary IFE Monoclonal protein Free light chain assay Abnormal kappa/lambda ratio Lymphocyte immunophenotyping Clonal population Medullogram Hemopathy discovered Imaging Etiology discovered At least one exam performed Etiology found

Total

Group 1

Group 2

n = 130

n = 33

n = 97

p values

70.2 (17.2) 58 (45)

72.0 (15.4) 16 (49)

69.5 (17.9) 42 (43)

0.58 0.61

14 (11) 26 (20) 33 (25)

8 (24) 10 (30) 8 (24)

6 (6) 16 (17) 25 (26)

0.008 0.09 0.86

31 (24) 25 (20) 29 (23) 2 (2)

17 (52) 7 (22) 10 (30) 0

14 (14) 18 (19) 19 (20) 2 (2)

b0.0001 0.70 0.22 1.0

65 (50) 16 (25) 42 (32) 10 (24) 14 (11) 4 (29) 17 (13) 6 (35) 9 (7) 3 (33) 12 (9) 8 (67) 18 (14) 5 (31) 60 (46) 22 (17)

20 (61) 7 (35) 13 (39) 3 (10) 4 (12) 3 (75) 8 (24) 3 (38) 3 (9) 1 (33) 5 (15) 4 (80) 8 (24) 3 (43) 19 (58) 11 (33)

45 (46) 9 (20) 29 (30) 0.005 10 (10) 1 (10) 9 (9) 3 (33) 6 (6) 2 (33) 7 (7) 4 (57) 10 (10) 2 (22) 41 (42) 11 (11)

0.16 0.22 0.31 0.75 0.04 0.04 1.0 0.69 1.0 0.18 0.58 0.08 0.60 0.13 0.004

All data are reported as number (%) except for age: mean (SD). Abbreviations: IFE: immunofixation electrophoresis.



In our study, an etiology was discovered for 22 patients, with equal number in the two groups. This emphasizes the need to investigate patients with mild hypogammaglobulinemia. In our study, however, the proportion of patients in group 1 for whom an etiology was found was three times greater than that in group 2. In around one quarter of cases, no etiology for the hypogammaglobulinemia was found or sought. It is worth noting that in our study, proteinuria was measured in only one third of patients with no known cause of hypogammaglobulinemia, while it was abnormal in about 25% of these. Therefore, we think that this simple biological examination should be more systematically performed in this situation. In addition, these patients were rarely monitored, while it is crucial to confirm the persistence of the hypogammaglobulinemia before investigations [2]. However, it was unlikely that these patients had primary immune deficiency because none of them had recurrent infections. Normalization of the γGb level in 15/21 (71%) patients who were monitored suggests that some patients may have had transient hypogammaglobulinemia, sometimes explained by an acute infection [12]. In this situation, hypogammaglobulinemia has been reported to be associated with a poor outcome, especially in a context of severe sepsis [18,19]. The incidence of all infectious complications at the inclusion was 43/ 100/year thus reflecting a severe level of immunosuppression in this cohort of patients with secondary hypogammaglobulinemia. This consequence was often the result of the combination for a same patient of several factors of immunosuppression: IST, hematological diseases, old age and hypogammaglobulinemia. Whether primary or secondary, hypogammaglobulinemia is associated with an increased risk of infection, in particular when immunoglobulin levels are very low for long periods of time [13]. Common hypogammaglobulinemia-related infections (ENT, respiratory or meningeal infections) were, in our study, equally distributed in the two groups, thus highlighting the fact that clinicians have to consider vaccination coverage in both groups of patients. Only one third of patients, and mostly patients suffering from hemopathy, were immunized against S. pneumoniae. By contrast, none of the patients with isolated hypogammaglobulinemia were vaccinated against S. pneumoniae. This finding suggests that recommendations [9] about vaccination are not being implemented in secondary hypogammaglobulinemia. In addition, and in accordance with the recommendations, only six patients were treated with IVIg, most of them because of recurrent, severe infections while they were affected by hemopathy [11,20]. It is worth remembering the lack of hard evidence about the costeffectiveness of this practice [21]. Our study has several limitations. Firstly, our cohort has an inherent selection bias linked to the recruitment of patients in a teaching hospital. Secondly, hypogammaglobulinemia could not be confirmed for 163 patients. A second control of the immunoglobulin level was not required for 76 of them with an already known cause of persistent hypogammaglobulinemia. Furthermore, it would not have changed the therapeutic strategy, the vaccination being already imposed by the underlying condition. For the 87 remaining patients with only one electrophoresis, no etiological exam was performed for 62% of them, suggesting that this discovery was not considered by the clinician. Then, we cannot exclude the possibility that among patients for whom no secondary cause was found, some may have had primary hypogammaglobulinemia. Finally, according to the retrospective design of our study, the incidence of infections in this cohort was probably underestimated and needs to be explored in prospective study. 5. Conclusion Hypogammaglobulinemia is a frequent condition in adults treated in a teaching hospital. Such patients are frail and often associate several factors of immunosuppression. The results of this study strongly suggest that clinicians who discover an abnormal γGb level (b 6.4 g/L) in a

5

patient should confirm and investigate this deficiency. The incidence of infectious complications and the effectiveness of vaccination in secondary hypogammaglobulinemia need to be greater explored in prospective studies. Learning points • Secondary hypogammaglobulinemia is a frequent condition. • Mild hypogammaglobulinemia (b6.4 g/L and 5 g/L) and severe hypogammaglobulinemia (i.e. b5 g/L) share close similarities in terms of etiology and infections. • Mild hypogammaglobulinemia requires confirmation, then investigation and vaccination. Abbreviations SPE Ig γGb CLL MM IST HIV EBV CMV ENT CVID IFE IVIg MGUS

serum protein electrophoresis immunoglobulin gammaglobulin chronic lymphocytic leukemia multiple myeloma immunosuppressive therapies human immunodeficiency virus Epstein–Barr virus cytomegalovirus ear nose throat common variable immunodeficiency immunofixation electrophoresis intra-venous immunoglobulin monoclonal gammopathy of unknown significance

Conflicts of interest and funding sources The authors have no funding or conflicts of interest to disclose. No financial support was received. Acknowledgments The authors thank Philip Bastable for his help in reviewing the manuscript. References [1] Rose ME, Lang DM. Evaluating and managing hypogammaglobulinemia. Cleve Clin J Med 2006;73:133–7. [2] Samson M, Audia S, Lakomy D, Bonnotte B, Tavernier C, Ornetti P. Diagnostic strategy for patients with hypogammaglobulinemia in rheumatology. Jt Bone Spine Rev Rhum 2011;78:241–5. [3] Chapel H, Cunningham-Rundles C. Update in understanding common variable immunodeficiency disorders (CVIDs) and the management of patients with these conditions. Br J Haematol 2009;145:709–27. [4] Oksenhendler E, Gérard L, Fieschi C, Malphettes M, Mouillot G, Jaussaud R, et al. Infections in 252 patients with common variable immunodeficiency. Clin Infect Dis 2008;46:1547–54. [5] Freeman JA, Crassini KR, Best OG, Forsyth CJ, Mackinlay NJ, Han P, et al. Immunoglobulin G subclass deficiency and infection risk in 150 patients with chronic lymphocytic leukemia. Leuk Lymphoma 2013;54:99–104. [6] Svensson T, Höglund M, Cherif H. Clinical significance of serum immunoglobulin G subclass deficiency in patients with chronic lymphocytic leukemia. Scand J Infect Dis 2013;45:537–42. [7] Casulo C, Maragulia J, Zelenetz AD. Incidence of hypogammaglobulinemia in patients receiving rituximab and the use of intravenous immunoglobulin for recurrent infections. Clin Lymphoma Myeloma Leuk 2013;13:106–11. [8] Martinot M, Oswald L, Parisi E, Etienne E, Argy N, Grawey I, et al. Immunoglobulin deficiency in patients with Streptococcus pneumoniae or Haemophilus influenzae invasive infections. Int J Infect Dis 2013;19:79–84. [9] HCSP. Vaccinations of immunocompromised and asplenic people. Guidelines. [cited 2014 March 09]. Available from: http://www.hcsp.fr/Explore.cgi/Telecharger? NomFichier=hcspr20120712_vaccinationimmunodeprime.pdf; 2012. [10] Intravenous immunoglobulin for the prevention of infection in chronic lymphocytic leukemia. A randomized, controlled clinical trial. Cooperative Group for the Study of Immunoglobulin in Chronic Lymphocytic Leukemia. N Engl J Med 1988; 319:902–7.


6


[11] Raanani P, Gafter-Gvili A, Paul M, Ben-Bassat I, Leibovici L, Shpilberg O. Immunoglobulin prophylaxis in chronic lymphocytic leukemia and multiple myeloma: systematic review and meta-analysis. Leuk Lymphoma 2009;50:764–72. [12] Riches PG, Hobbs JR. Mechanisms in secondary hypogammaglobulinaemia. J Clin Pathol 1979;13:15–22. [13] Furst DE. Serum immunoglobulins and risk of infection: how low can you go? Semin Arthritis Rheum 2009;39:18–29. [14] Posey WC, Nelson HS, Branch B, Pearlman DS. The effects of acute corticosteroid therapy for asthma on serum immunoglobulin levels. J Allergy Clin Immunol 1978;62:340–8. [15] Kawano T, Matsuse H, Obase Y, Kondo Y, Machida I, Tomari S, et al. Hypogammaglobulinemia in steroid-dependent asthmatics correlates with the daily dose of oral prednisolone. Int Arch Allergy Immunol 2002;128:240–3. [16] Hamilos DL, Young RM, Peter JB, Agopian MS, Iklé DN, Barka N. Hypogammaglobulinemia in asthmatic patients. Ann Allergy 1992;68:472–81.

[17] Tsai H-T, Caporaso NE, Kyle RA, Katzmann JA, Dispenzieri A, Hayes RB, et al. Evidence of serum immunoglobulin abnormalities up to 9.8 years before diagnosis of chronic lymphocytic leukemia: a prospective study. Blood 2009;114:4928–32. [18] Prucha M, Zazula R, Herold I, Dostal M, Hyanek T, Bellingan G. Presence of hypogammaglobulinemia in patients with severe sepsis, septic shock, and SIRS is associated with increased mortality. J Infect 2014;68:297–9. [19] Taccone FS, Stordeur P, De Backer D, Creteur J, Vincent J-L. Gamma-globulin levels in patients with community-acquired septic shock. Shock 2009;32:379–85. [20] Mouthon L, Lortholary O. Intravenous immunoglobulins in infectious diseases: where do we stand? Clin Microbiol Infect 2003;9:333–8. [21] Weeks JC, Tierney MR, Weinstein MC. Cost effectiveness of prophylactic intravenous immune globulin in chronic lymphocytic leukemia. N Engl J Med 1991;325:81–6.
