Incidence and Outcomes of Neutropenia in Patients with Celiac Disease- a Consecutive Analysis of 1729 Patients

Background: Celiac disease (gluten sensitive enteropathy) is a systemic disease with hematological manifestations including; iron deficiency anemia, vitamin B12 and folate deficiency, leukopenia/neutropenia, hyposplenism, venous thromboembolism, and the enteropathy associated T cell lymphoma (E-TCL).  Leukopenia/neutropenia is rare, and can result from folate or copper deficiency, but in many cases remains idiopathic. The consequences of neutropenia and neutrophil responses to a gluten free diet are not well described. We carried out this study to assess the incidence and outcomes of neutropenia in patients with celiac disease.

Methods: The Mayo clinic celiac disease database (n=1729) was retrospectively analyzed for all patients with a confirmed diagnosis of celiac disease that had leukopenia (<3.5 x10(9)/L) and/or neutropenia (<1.5 x 10(9)/L).  The diagnosis of celiac disease was established by positive tissue transglutaminase antibodies (TTG) and confirmatory small bowel biopsy findings.  Additional causes of neutropenia such as copper and folate deficiency, drugs, benign ethnic and cyclic neutropenia, chemotherapy, hematological malignancies, combined variable immunodeficiency (CVID), and autoimmune diseases were meticulously excluded.  Data abstracted included demographics, temporal association with diagnosis, nadir counts, associated hematological findings, relationship with infections, the use and effect of G-CSF, response to gluten free diet and survival outcomes. 

Results:  21(1.2%) of 1729 cases screened had idiopathic-celiac related neutropenia; 17 (81%) Caucasian, 15 (48%) females.  In 12(57%) patients the neutropenia preceded or occurred near the time of the diagnosis of celiac disease, while in 7 (33.3%) it occurred subsequently. In 2 cases the temporal association could not be established.  The median age at diagnosis of celiac disease was 46 years (10-67) and the median follow up for this group was 41.7 months (0-207.3).  At last follow up 1 (4.8%) death has been documented; 0 from infectious complications, with no cases of E-TCL.   The median laboratory values at diagnosis of celiac disease were available in 11 (52%) patients and included; hemoglobin 12.9 g/dL (10.5-16.8), MCV 91.5 fl (79.6-102.8), WBC 3.2 x10⁹/L (2.3-5.2), ANC 1.29x10⁹/L (0.33-3.24), ALC 1.34x10⁹/L (0.75-2.17), AMC 0.39x10⁹/L (0.15-0.71), and platelet counts 225x10⁹/L (168-711). Anti-granulocyte antibodies were assessed in 7 patients and were negative in all cases. None of these patients had coexistent hyposplenism and three cases of neutropenia that occurred in the setting of celiac disease and CVID were excluded. In 15 (71%) patients the neutropenia was incidentally detected during routine laboratory work and in 4 (19%) it came to light secondary to infections. Ten (48%) patients had recurrent infections (≥1) as documented by their providers, including 7 with sinopulmonary infections, 1 with urinary tract infections, and 1 with skin and soft tissue infections.  Additional immunological assessments were not available in these patients.

Three (14%) patients had an ANC <500 (significant neutropenia); of which 2 had recurrent sinopulmonary infections. The third patient remained largely asymptomatic.  Two of the 3 patients received G-CSF support during infections and responded adequately (ANC improvement > 1 x 10(9)/L). The ANC in patients in this cohort did not correlate with severity of infections.  Five (23%) of 21 patients had improvement in neutrophil counts after adopting a gluten free diet, while there was no response in 10 and data was unavailable in 6.  The median time to ANC response was 14.5 months (10-15.3) and the median increment in ANC was 0.67 x 10(9)/L (0.48-0.98).   

 Conclusions:  Celiac disease associated neutropenia, especially significant neutropenia (ANC <500) is a rare occurrence (~1%). It can often pre-date the diagnosis of celiac disease or occur subsequently.  It is potentially associated with an increased incidence of infections, with-out good correlation with the severity of neutropenia.  Less than half the patients do seem to respond to a gluten free diet.  G-CSF responses seem to be adequate and G-CSF can be used in the setting of severe infections.