Secondary Logo

Journal Logo

Online Articles: Original Articles

Clinical Features and Treatment Outcomes of Childhood Autoimmune Hemolytic Anemia

A Retrospective Analysis of 68 Cases

Fan, Junjie MD; He, Hailong MD; Zhao, Wenli MD; Wang, Yi MD; Lu, Jun MD; Li, Jie MD; Li, Jianqin MD; Xiao, Peifang MD; Lu, Ye MD; Chai, Yihuan MD; Hu, Shaoyan MD, PhD

Author Information
Journal of Pediatric Hematology/Oncology: March 2016 - Volume 38 - Issue 2 - p e50-e55
doi: 10.1097/MPH.0000000000000476


Autoimmune hemolytic anemia (AIHA) is caused by autoantibodies against patients’ own erythrocytes leading to hemolysis and anemia.1 Generally, diagnosis is made by 2 criteria: symptoms of hemolysis and anemia, and positive direct antiglobulin test (DAT, Coombs test).2 AIHA can be subdivided into primary (or idiopathic) AIHA and secondary AIHA associated with any potentially causative disorders. AIHA can also present with thrombocytopenia, which is defined as Evans syndrome (ES). Usually ES is considered a distinct type of AIHA.2 There are few reports on childhood AIHA.3–8 As showed in Table 1, recent studies (after 2000) reported that the median age of diagnosis varied from 10.5 months to 11 years in different regions, and there was no sex preference in AIHA. At diagnosis, all patients had varied degrees of anemia, and the majority of them were DAT positive. Secondary AIHA accounted for 23.5% to 63.0% of the cases. Usually glucocorticoid was used as first-line therapy, whereas there were refractory and/or relapse cases even with multitherapies including immunosuppressors and splenectomy, and occasional fatal cases. ES tended to have a chronic and relapsing clinical course.9–11 Here, we present a cohort of 68 pediatric patients diagnosed with AIHA, with noticeably 29.4% cases of ES, as well as the associated clinical and hematologic profiles. In addition, the treatment duration and efficacy without the use of rituximab were analyzed.

Results of Studies After 2000



Seventy-eight consecutive patients diagnosed with AIHA between March 2006 and January 2013 were well screened for the retrospective study. Six patients younger than 3 months were excluded because the hemolytic anemia may be due to transplacental antibody, and the clinical course and treatment outcome are very different from AIHA of older children. In addition, 4 patients with positive DAT actually had underlying congenital hemolytic diseases (3 thalassemia and 1 pyruvate kinase deficiency), therefore excluded. A total of 68 patients were finally included in this study after all screening procedures. This study was reviewed and approved by the Ethical Committee of Human Experimentation of the Children’s Hospital of Soochow University.


Besides symptoms, the diagnosis of AIHA diagnosis is mainly based on a positive DAT, and laboratory evidence of hemolysis2: complete blood count (including reticulocyte count), indirect bilirubin level, lactic dehydrogenase (LDH), and free plasma hemoglobin (Hb) assay. Some cases with negative DAT also can be diagnosed with AIHA if diseases such as hereditary spherocytosis, paroxysmal nocturnal hemoglobinuria, glucose-6-phosphate dehydrogenase deficiency, and elliptocytosis can be excluded, and responded to glucocorticoid. None of the cases presented in this study had clinical findings to fulfill the diagnosis of autoimmune lymphoproliferative syndrome. Human immunodeficiency virus and hepatitis B and C screening were routinely performed, and none of the study patient was positive for human immunodeficiency virus or hepatitis B/C. Differentiation of AIHA based on different serological antibodies (IgG, IgM, IgA, C3, and/or combinations) was performed, as different serological profiles might correlate with the severity and treatment outcome of AIHA.1,3,12 AIHA also can be subdivided into primary (or idiopathic) AIHA and secondary AIHA on the basis of associated and potentially causative disorders, including infections, autoimmune disorders, and malignancies.


As local practice standard, first-line treatment is glucocorticoid with or without intravenous immunoglobulin (IVIG); second-line therapies are immunosuppressors including 6-mercaptopurine, cyclosporine A, or cyclophosphamide; splenectomy is considered as the third-line therapy. In some reports, rituximab was used as second-line therapy and showed benefit13,14; however, in our study, no patient received rituximab (further discussed in the Discussion section). Transfusion as supportive therapy was given in severe/symptomatic anemia, and patients received folic acid orally during treatment to support hematopoiesis. Response criteria2 were classified as: (1) complete remission (CR): clinical symptoms disappeared, normalized erythrocyte count, Hb level, reticulocyte count, bilirubin level, and negative DAT; (2) partial response: clinical symptoms controlled, with Hb>80 g/L, reticulocyte count <5%, bilirubin ≤34 umol/L, as well as negative or reducing titer of DAT; (3) no response: anemia or persistence of symptoms of hemolysis, and the laboratory criteria not achieving partial response.


Patients’ follow-up information was obtained from outpatient records and telephone survey. The cutoff point for follow-up analysis was June 2014. Forty-five cases (66.2%) have complete follow-up information at the conclusion of data collection, whereas 23 (33.8%) cases did not have complete follow-up data due to practical limitations such as lack of cooperation of the parents or changing of medical insurance. The median duration of follow-up is 45 months (range, 16 to 98).

Statistical Analysis

Normally distributed continuous data were analyzed using t-test/t’-test, and nonparametric data were analyzed using the Mann-Whitney U test. The χ2 test was used to analyze categorical data such as relapse rates. Kaplan-Meier survival analysis was performed to illustrate the time of glucocorticoid therapy. All analyses were performed using SPSS version 19.0. Differences were considered significant at P<0.05.


Clinical Features at Presentation

The age at initial diagnosis of patients in this cohort ranged from 4 months to 15 years old with a median age of 16 months, and 43 cases (63.2%) 3 years of age or below (Fig. 1). On the basis of the thrombocyte level, patients were stratified into the AIHA-only group (n=48) or the ES group (n=20, with platelet <100×109/L). All patients presented with symptomatic anemia and disease features as illustrated in Table 2.

Onset age distribution.
Clinical Features at Diagnosis

Associated Diseases

AIHA with established associated diseases were considered as secondary AIHA. Forty-one (60.3% of total) cases had associated diseases, mainly infectious diseases (40 cases, 97.6%) such as respiratory infection (25 cases, 61.0%), enteritis (7 cases, 17.1%), Epstein-Barr virus/cytomegalovirus infection (7 cases, 17.1%), and soft tissue infection (1 case, 2.4%). During follow-up, 1 patient was diagnosed with ulcerative colitis, but the anemia recovered. Six patients presented with positive antinuclear antibody (ANA) at diagnosis and had presumed systemic lupus erythematosus (SLE). During follow-up, all had recovered from anemia and had no symptoms of SLE; repeated screening tests for SLE also showed insufficient evidence of SLE.

Laboratory Findings

Of the patients enrolled, 64 cases (94.1%) were DAT positive with a total of 9 types of antibody combinations (Table 3), whereas 4 cases (5.9%) were DAT negative.

Hemolysis Degrees in Different Groups

All patients’ Hb declined at diagnosis. Among all 68 cases, 7 cases (10.3%) had mild anemia (Hb>90 g/L); 24 cases (35.3%) had moderate anemia (Hb 60 to 90); 35 cases (51.5%) had severe anemia (Hb 30 to 60); 2 cases (2.9%) had very severe anemia (Hb<30). The average Hb was 61.0±21.2 g/L. Reticulocytosis as adequate response to AIHA was seen in 63 cases (92.6%), whereas normal reticulocyte level as inadequate response was seen in only 5 cases (7.4%). Leukocytosis was noted in 34 (50.0%) patients, leukopenia in 4 (5.9%) patients, and 30 (44.1%) cases had normal WBC levels. All ES patients had varied degrees of thrombocytopenia with the median platelet count decreased to 38.5×109/L (10 to 88×109/L).

Forty-five patients (66.2%) had an increased total bilirubin level with the median of 34.1 μmol/L. Forty-four (64.7%) cases had elevated indirect bilirubin and the median was 22.5 μmol/L. In AIHA-only patients, the median indirect bilirubin was higher than that in ES patients (P=0.03).

Of the 68 patients, the median LDH level was 596.4 U/L (229 to 5501) and there were 55 (80.9%) cases with LDH level exceeded the upper normal range of 382 U/L. There was no significant difference of average LDH between AIHA-only and ES groups (P=0.21).

Comparison of the Severity of Hemolysis

We stratified patients on the basis of age (1 y or below vs. above 1 y), initial presentations (AIHA-only vs. ES), antibody detected by DAT (single-antibody vs. combined-antibody), and presence of IgM (IgM-contained vs. non–IgM-contained) to compare the severity of hemolysis. The average Hb level was lower in the 1-year or below age group than in the above 1-year age group, combined-antibody group than single-antibody group, and IgM-contained group than non–IgM-contained group, respectively (P<0.05 for all). For free plasma Hb, there is no statistical difference between the comparison groups (P>0.05 for all) (Table 3).

Treatment Response

All patients were treated with glucocorticoid as the first-line therapy with a starting dose of methylprednisolone 2 to 5 mg/kg/d based on the severity of the anemia. Once a response was seen, the methylprednisolone dose was tapered and changed to prednisone and stopped when the Hb and reticulocyte count normalized. Along with glucocorticoid, IVIG was given to 22 cases concurrently, with a dose of 0.4 to 1 g/kg for 1 to 3 times. All IVIGs were given after initial diagnosis, and no IVIG was used for relapsed cases. Of the 45 patients with complete follow-up data, 40 (88.9%) cases received first-line therapy only, 4 cases received second-line therapy such as 6-mercaptopurine due to poor response to glucocorticoid, and 1 patient received splenectomy after failure to respond to pharmacotherapy of glucocorticoid and cyclosporin A (Figure 2).

Outcomes of treatment.

Of all 68 patients, 44 patients (64.7%) received washed red cell transfusion as supportive treatment due to severe anemia.

After treatment, all 45 cases achieved CR with the first-line therapy. At the cutoff point of follow-up, 29 cases remained in continuous CR (64.4%), whereas 16 cases relapsed (35.6%). Relapse occurred 0.5 to 36 months after initial diagnosis with a median time of 4 months. The treatment outcome is summarized in Figure 2.

On the basis of the follow-up data available of 45 cases, the median duration of therapy is 7.5 months (range, 1 to 83). Duration of therapy was longer in the ES group than in the AIHA-only group (P<0.05). However, the duration of therapy was not significantly impacted by age, IVIG, or types of serologic antibody groups. Relapse rates were also compared between all comparison groups, there was no significant difference (P>0.05 for all) (Table 4).

Durations of Therapy and Relapse Rates of Different Groups

Of the 45 cases, 29 cases achieved continuous CR with glucocorticoid alone. The median duration of therapy was 5.5 months (range, 1 to 31) and the median follow-up time of 50 months (range, 17 to 79). Sixteen cases relapsed, and 9 cases achieved second CR with retreatment of glucocorticoid only and the median duration of therapy was 15 months (range, 7 to 80), and median follow-up time of 59 months (range, 16 to 98). There has been longer duration of glucocorticoid therapy in those patients who relapsed than in those who did not relapse (P<0.05).

We performed Kaplan-Meier survival analysis for time to CR for the patient cohort who entered remission (n=38) as well as the patient cohort with time to second remission after relapse (n=9). The median time is 5.8 months versus 10 months, and there is no statistical difference between these 2 groups (P=0.14) (Supplemental Figure 1, Supplemental Digital Content 1,

Side effects of glucocorticoid occurred during therapy including moon face (26 cases) and hypertension (2 cases). No deep vein thrombosis, pulmonary emboli, hyperglycemia, or gastrointestinal hemorrhage were observed during the duration of follow-up.

Among 20 cases of ES, 13 cases had long-term follow-up data (as discussed in the Results section). Only 1 patient was still on low-dose glucocorticoid 83 months into therapy with platelet count at 60×109/L after initial disease relapse. One patient failed frontline glucocorticoid and responded well to splenectomy in combination with cyclosporine without further therapy. The median duration of therapy for the case cohort was 15 (3 to 83) months. The relapse rate is 38.5% (5/13). In one of the relapsed cases, with only thrombocytopenia without the findings of recurrent AIHA, glucocorticoid was given.


AIHA is an uncommon disease that can affect patients at any age. The incidence in adults is around 1 to 3 cases per 100,000 per year,15,16 whereas it is estimated that the annual incidence in children is around 3 per million.3,17 In 4 previous reports of childhood AIHA, the median age at diagnosis was 10.5 months, 3.8 years, 6.5 years, and 11 years, respectively.3,5–7 In our study, the median age of onset was 16 months (Table 1), and 63.2% cases were 3 years of age or younger. The age distribution varies significantly in published literature. We do not have the information for the patient selection in other publications and, noticeably, the median age of onset was 10.5 months in the Brazil study. In our study, between 2006 and 2013, consecutive cases that fulfilled the diagnosis criteria were enrolled. Furthermore, we observed that patients less than 1-year old tend to present with a lower Hb level at diagnosis as shown in Table 3. It might be partly due to physiological anemia, and should be taken into consideration prior making a diagnosis and management of AIHA.

Positive DAT is the diagnostic hallmark of AIHA; however, it may be absent in some cases (Table 1). Some studies showed that usually, DAT positivity needed a certain number of antibody molecules sensitizing all red blood cells,18–21 therefore AIHA with a lower quantity of autoantibodies may present with negative DAT, which can be a diagnostic pitfall. Four cases in our study were diagnosed as AIHA with negative DAT, yet all other diagnostic measures met the criteria, and the efficacy of glucocorticoid therapy validated the accuracy of the diagnosis.

There are a variety of antibody combination types in DAT. Earlier studies indicated that IgG/IgG+C3 was associated with severe hemolysis and relatively poor treatment outcomes, whereas C3 correlated with lower relapse rates3,22; another report showed that compared with IgG alone, IgG combined with IgM or IgA can lead to more severe hemolysis.23 Consistent with published data, our study showed that patients with combined-antibody or IgM-contained serologic profiles tended to have more severe anemia.

In some prior reports, ES accounted for 13.6% to 64.3% of all cases of AIHA and as reported had a chronic, refractory nature, variable treatment response, and high relapse rate.9–11,24,25 In our study, the degree of anemia was similar between ES group and AIHA-only group; however, the duration of therapy in ES patients was longer than that of isolated AIHA. Therefore, clinically the glucocorticoid therapy should be prolonged in ES and tapered off with more consideration.

The mechanism of glucocorticoid to treat AIHA is probably downregulation of Fc receptors on phagocytes and reduced IL-2 production.26 A prior study showed that in adult AIHA, the response rate of glucocorticoid was 68.6% in the acute phase and 35% in the prolonged treatment.27 In our study, all cases responded to glucocorticoid in acute phase. Even after relapse, 56.3% cases achieved CR with glucocorticoid treatment only. These findings confirmed that glucocorticoid is the mainstay in childhood AIHA treatment.

Second-line therapy includes immunosuppressors such as 6-mercaptopurine, cyclosporine A, and cyclophosphamide.28–30 Recently, rituximab has been proven effective in refractory AIHA.13,14 Splenectomy as third-line therapy would be considered when these medicines failed. However, as the glucocorticoid treatment outcome of childhood AIHA is better than that of adults, these treatment approaches are less common in children, especially the use of rituximab in undeveloped regions due to its high price (Table 1).

AIHA may be primary or secondary to other diseases, including infections, autoimmune disorders, and malignancies. Data published before showed that in adults the main cause of AIHA was systemic autoimmune diseases such as SLE.31 The presence of ANA in the normal population has been reported without evidence of hemolytic anemia or autoimmune disease. In the Chinese population there have been 2 publications to our best knowledge that showed that 2.2% to 5.6% of healthy children were tested positive for ANA.32,33 In our study of childhood population, secondary AIHA was mainly associated with infectious disease, whereas systemic autoimmune disease was rare. Perhaps longer follow-up may reveal more underlying systemic autoimmune diseases.


1. Gehrs BC, Friedberg RC. Autoimmune hemolytic anemia. Am J Hematol. 2002;69:258–271.
2. Zhang ZN, Shen T. Standard of Diagnosis and Therapeutic Effect of Hematopathy. Beijing: Science Press; 2008.
3. Aladjidi N, Leverger G, Leblanc T, et al.. New insights into childhood autoimmune hemolytic anemia: a French national observational study of 265 children. Haematologica. 2011;96:655–663.
4. Gurgey A, Yenicesu I, Kanra T, et al.. Autoimmune hemolytic anemia with warm antibodies in children: retrospective analysis of 51 cases. Turk J Pediatr. 1999;41:467–471.
5. Naithani R, Agrawal N, Mahapatra M, et al.. Autoimmune hemolytic anemia in children. Pediatr Hematol Oncol. 2007;24:309–315.
6. Oliveira MC, Oliveira BM, Murao M, et al.. Clinical course of autoimmune hemolytic anemia: an observational study. J Pediatr (Rio J). 2006;82:58–62.
7. Park DC, Yang CH, Kim KY. Autoimmune hemolytic anemia in children. Yonsei Med J. 1987;28:313–321.
8. Sokol RJ, Hewitt S, Stamps BK, et al.. Autoimmune haemolysis in childhood and adolescence. Acta Haematol. 1984;72:245–257.
9. Anderson D, Ali K, Blanchette V, et al.. Guidelines on the use of intravenous immune globulin for hematologic conditions. Transfus Med Rev. 2007;21:S9–56.
10. Norton A, Roberts I. Management of Evans syndrome. Br J Haematol. 2006;132:125–137.
11. Pui CH, Wilimas J, Wang W. Evans syndrome in childhood. J Pediatr. 1980;97:754–758.
12. Engelfriet CP, Overbeeke MA, von dem Borne AE. Autoimmune hemolytic anemia. Semin Hematol. 1992;29:3–12.
13. Barcellini W, Zaja F, Zaninoni A, et al.. Low-dose rituximab in adult patients with idiopathic autoimmune hemolytic anemia: clinical efficacy and biologic studies. Blood. 2012;119:3691–3697.
14. Dierickx D, Verhoef G, Van Hoof A, et al.. Rituximab in auto-immune haemolytic anaemia and immune thrombocytopenic purpura: a Belgian retrospective multicentric study. J Intern Med. 2009;266:484–491.
15. Pirofsky B. Auto Immunization and the Auto Immune Haemolytic Anaemias. Baltimore: Williams and Wilkins; 1969.
16. Bottiger LE, Westerholm B. Acquired haemolytic anaemia. I. Incidence and aetiology. Acta Med Scand. 1973;193:223–226.
17. Sokol RJ, Hewitt S, Stamps BK. Autoimmune haemolysis: an 18-year study of 865 cases referred to a regional transfusion centre. Br Med J. 1981;282:2023–2027.
18. Chaplin H, Avioli LV. Grand rounds: autoimmune hemolytic anemia. Arch Intern Med. 1977;137:346–351.
19. Fabijanska-Mitek J, Lopienska H, Zupanska B. Gel test application for IgG subclass detection in auto-immune haemolytic anaemia. Vox Sang. 1997;72:233–237.
20. Gilliland BC. Coombs—negative immune hemolytic anemia. Semin Hematol. 1976;13:267–275.
21. Merry AH, Thomson EE, Rawlinson VI, et al.. Quantitation of IgG on Erythrocytes: correlation of number of IgG molecules per cell with the strength of the direct and indirect antiglobulin tests. Vox Sang. 1984;47:73–81.
22. Barcellini W, Fattizzo B, Zaninoni A, et al.. Clinical heterogeneity and predictors of outcome in primary autoimmune hemolytic anemia: a GIMEMA study of 308 patients. Blood. 2014;124:2930–2936.
23. Sokol RJ, Hewitt S, Booker DJ, et al.. Red cell autoantibodies, multiple immunoglobulin classes, and autoimmune hemolysis. Transfusion. 1990;30:714–717.
24. Mathew P, Chen G, Wang W. Evans syndrome: results of a national survey. J Pediatr Hematol Oncol. 1997;19:433–437.
25. Wang WC. Evans syndrome in childhood: pathophysiology, clinical course, and treatment. Am J Pediatr Hematol Oncol. 1988;10:330–338.
26. Collins PW, Newland AC. Treatment modalities of autoimmune blood disorders. Semin Hematol. 1992;29:64–74.
27. Zupanska B, Sylwestrowicz T, Pawelski S. The results of prolonged treatment of autoimmune haemolytic anaemia. Haematologia (Budap). 1981;14:425–433.
28. Liu H, Shao Z, Jing L. The effectiveness of cyclosporin A in the treatment of autoimmune hemolytic anemia and Evans syndrome. Zhonghua Xue Ye Xue Za Zhi. 2001;22:581–583.
29. Moyo VM, Smith D, Brodsky I, et al.. High-dose cyclophosphamide for refractory autoimmune hemolytic anemia. Blood. 2002;100:704–706.
30. Sobota A, Neufeld EJ, Lapsia S, et al.. Response to mercaptopurine for refractory autoimmune cytopenias in children. Pediatr Blood Cancer. 2009;52:80–84.
31. Baek SW, Lee MW, Ryu HW, et al.. Clinical features and outcomes of autoimmune hemolytic anemia: a retrospective analysis of 32 cases. Korean J Hematol. 2011;46:111–117.
32. Guo YP, Wang CG, Liu X, et al.. Analysis of distribution of antinuclear antibodies in a healthy population. Chin J Microbiol Immunol. 2014;34:212–215.
33. Qian PX, Wang L. Effect of age and gender on distribution of antinuclear antibodies in healthy population. Hainan Med J. 2015;26:201–203.

autoimmune hemolytic anemia (AIHA); Evans syndrome (ES); children; glucocorticoid

Supplemental Digital Content

Copyright © 2016 Wolters Kluwer Health, Inc. All rights reserved.