Hematologic complications of inflammatory bowel disease (IBD) include anemia secondary to micronutrient deficiencies, thrombocytosis and leukocytosis from active disease, and myelosuppression from medications. (1–3) Autoimmune antibody-mediated hematologic disorders also have been reported in patients with IBD (2,4–6). Altman et al. (4) described three cases of autoimmune hemolytic anemia in patients with ulcerative colitis and reviewed 16 cases previously reported in the literature. In addition, autoimmune neutropenia has been reported in patients with Crohn disease. (5)
Immune thrombocytopenic purpura (ITP), also referred to as idiopathic thrombocytopenic purpura, has been sporadically reported in individuals with ulcerative colitis or Crohn disease (7). ITP is a bleeding disorder defined by a low platelet count in the absence of other clinically apparent causes of thrombocytopenia. (8) The thrombocytopenia is caused by antibody-mediated platelet destruction; bone marrow biopsy demonstrates megakaryocytes that are normal or increased in the bone marrow. (9) There is a scarcity of literature describing the association between ITP and IBD in children. (7,10) We describe a series of eight children with both inflammatory bowel disease and ITP and propose that the association between these two rare illnesses may reflect a defect in immune regulation.
PATIENTS AND METHODS
To identify an association between ITP and IBD, one of the authors (AB) posted a request for similar cases on the pediatric gastroenterology internet bulletin board in June 1999. At the time, approximately 750 pediatric gastroenterologists worldwide and 50% to 75% of the practicing pediatric gastroenterologists in North America had subscribed to the bulletin board (Colletti R. Personal communication, 2000). From the replies, 10 cases of children with both ITP and IBD were identified; investigators submitted chart reviews of eight patients for inclusion in the case series. In each case, to confirm the accuracy of the diagnosis and to determine the natural history and long-term outcome of the illness, the medical records were reviewed by two pediatric gastroenterologists (L. H. and A. B.) and a pediatric hematologist (S. J.).
The case series is summarized in Table 1, and patients are described in detail in the following sections.
A 12-year-old boy was seen in July of 1994 with a 1-year history of abdominal pain and loose stools. Family history was notable for a paternal aunt with IBD. Initial laboratory evaluation showed a hematocrit concentration of 42%, an erythrocyte sedimentation rate (ESR) of 14 mm/h, and an albumin concentration of 4.0 g/dL. Upper gastrointestinal series with small bowel follow-through was unremarkable. Upper endoscopy with biopsies showed histologic duodenitis and gastritis (neutrophils and eosinophils in the crypts with no Helicobacter pylori present). Colonoscopy showed endoscopic and microscopic inflammation in the descending colon and microscopic inflammation in the terminal ileum and rectum. The patient was diagnosed with Crohn disease and treated with prednisone and sulfasalazine. Because of steroid dependency at 6 months, 6-mercaptopurine was started.
Two years after his initial presentation, while receiving 6-mercaptopurine, routine monitoring of his blood count showed a decreasing platelet count. At the time, the patient was asymptomatic, and there was no evidence of active Crohn disease. Despite the discontinuation of the 6-mercaptopurine, the platelet count continued to drop to less than 10,000/mL. A bone marrow biopsy was performed to differentiate between myelosuppression from 6-mercaptopurine and ITP. The biopsy showed marked thrombocytopenia with a normal number of megakaryocytes, consistent with an increased rate of platelet destruction. The patient received two treatments of prednisone (60 mg/day per treatment) over a 2-week period, with a rise in platelet count to >100,000/mL, then the steroids were tapered. Because of the lack of sustained response to the corticosteroids, intravenous immunoglobulin (IV Ig) was administered (50 g/dose), with a rise in the platelet count to 449,000/mL. The IV Ig therapy was continued at 2-month intervals until June of 1998; 6-mercaptopurine was restarted after the IV Ig was initiated. Platelet count 4 years after IV Ig therapy was 150,000/mL.
A 14-year-old girl was seen in May of 1999 with abdominal pain, epistaxis, and rectal bleeding. Initial laboratory evaluation showed a platelet count of less than 10,000/mL, a hemoglobin concentration of 10.2 g/dL, a white blood cell count of 7,600/mL, a reticulocyte count of 2.6%, and a negative Coombs test. A bone marrow biopsy showed marked thrombocytopenia, with increased megakaryocytes, suggestive of peripheral destruction or a sequestration process. The patient was diagnosed with ITP and treated with one dose of IV Ig (800 mg/kg) with resolution of her thrombocytopenia, abdominal pain, and epistaxis. Two months after her initial presentation, she had bloody diarrhea develop. Flexible sigmoidoscopy showed endoscopic inflammation in the descending colon, sigmoid colon, and rectum. Microscopic evaluation of a colon-rectal biopsy showed active colitis with mild architectural distortion. Upper gastrointestinal series with small bowel follow-through was unremarkable. The patient was diagnosed with indeterminate colitis and treated with a rectal preparation of 5-aminosalicylic acid.
A 2-year-old boy was seen in February of 1997 with diarrhea, rectal bleeding, and a perianal abscess. Family history was notable for a maternal cousin with rheumatoid arthritis. Initial laboratory evaluation showed a hematocrit of 34%, ESR of 27 mm/h, and albumin of 4.3 g/dL. Colonoscopy biopsy specimens showed microscopic chronic inflammation of the ileocecal valve, cecum, and right colon. Initial upper gastrointestinal series with small bowel follow-through was unremarkable. He was diagnosed with Crohn disease and treated with sulfasalazine and metronidazole with good initial response. Approximately 1 month after his initial presentation, routine laboratory evaluation revealed a decreasing platelet count of 125,000/mL. The platelet count continued to decrease to 6,000/mL, despite stopping the sulfasalazine therapy. Other studies performed at the time included a negative Coombs test and an antinuclear antibody of 1:320. Bone marrow biopsy showed normal cellularity with mild megakaryocytosis consistent with ITP. Prednisone (2 mg/kg/day) was initiated with a rapid increase in the platelet count, allowing the resumption of sulfasalazine therapy. After approximately 1 year, his prednisone was weaned to 2.5 mg every other day, and his platelet count remained stable without recurrence of thrombocytopenia.
A 5-year-old girl was seen in 1985 with purpura and mucosal petechiae. Family history was notable for ITP and non-Hodgkins lymphoma in the mother. Initial laboratory evaluation showed a platelet count of 8,000/mL, hemoglobin of 10.1 g/dL, white blood cell count of 8,800/mL, and negative Coombs and antinuclear antibody evaluations. A bone marrow biopsy was consistent with ITP. The patient was treated with prednisolone with an increase in platelet count to 40,000/mL. She was weaned off the steroids 4 months later, with her platelet count stable at 40,000/mL. However, over the next 3 years she required two additional 4- to 6-month courses of prednisolone for platelet counts of 13,000/mL and 10,000/mL, respectively, and four courses of IV Ig to maintain stable platelet counts.
Six years after her initial presentation, the patient had abdominal pain, diarrhea, rectal bleeding, growth failure, and anal tags develop. Laboratory evaluation at that time showed a hemoglobin concentration of 12.7 g/dL, an ESR of 35 mm/h, and a white blood cell count of 9,700/mL. Albumin level was not available. Colonoscopy showed endoscopic findings of aphthous lesions in the rectum and left colon and severe transverse ulceration. Stool evaluation was negative for infectious pathogens. She was diagnosed with Crohn disease. She was initially treated with prednisolone, olsalazine, and metronidazole. During the same year as the Crohn disease diagnosis, she again had thrombocytopenia develop. She was treated with further courses of IV Ig without improvement, and, subsequently, a splenectomy was performed. The patient has had no further thrombocytopenia. Her Crohn disease proved difficult to control, despite the use of enteral nutrition, cyclosporine, tacrolimus, and azathioprine, and she subsequently received a colectomy and ileorectal anastomosis.
A 9-year-old boy was seen in December 1995 with a 2-month history of abdominal pain, diarrhea, rectal bleeding, and weight loss. Family history was unremarkable. Initial laboratory data showed a white blood cell count of 7,000/mL, with a differential of 18% bands, 34% neutrophils, 38% lymphocytes, 7% monocytes, and 3% eosinophils. Hematocrit was 38.5%, and platelet count was low at 59,000/mL. Antinuclear antibody was negative. His ESR rate was increased at 36 mm/h. Albumin concentration was 3.4 g/dL. Stool evaluation results were negative for infectious pathogens. Colonoscopy showed gross and microscopic pancolitis with cryptitis, crypt abscesses, and crypt architecture distortion, extending from the hepatic flexure to the rectum. Upper gastrointestinal studies with small bowel follow-through and ileal biopsy results were normal. The patient was diagnosed with ulcerative colitis and prescribed sulfasalazine (50 mg/kg/day).
Three months after the patient was diagnosed with ulcerative colitis, he was noted to have a platelet count of 17,000/mL on routine laboratory screening. Other laboratory values included a hematocrit of 38.6%, white blood cell count of 4,500/mL, negative Coombs antibody, and negative antiplatelet antibody. He was diagnosed with ITP, treated with Rh o (D) immune globulin [Rh o (D) IV Ig], and within 2 days his platelet count increased to 110,000/mL. Since then, the patient has not required any further therapy for the ITP. Subsequently, he experienced three relapses of his ulcerative colitis requiring corticosteroids. He is now maintained on mesalamine and 6-mercaptopurine.
A 16.5-year-old boy was seen with abdominal pain and bloody diarrhea in April of 1998. Family history was unremarkable. Laboratory evaluation showed a white blood cell count of 6,200/mL with a differential of 24% polymorphonuclear cells, 17% band forms, 42% lymphocytes, 9% monocytes, and 5% eosinophils; a hematocrit concentration of 45% with a reticulocyte count of 2.8%; platelet count of 17,000/mL; and ESR of 20. Upper gastrointestinal series with small bowel follow-through was normal. Colonoscopy to the splenic flexure showed endoscopic evidence of diffuse, continuous colitis. On the basis of these findings and his presentation, he was diagnosed with ulcerative colitis. He was treated with prednisone (20 mg/day). Evaluation of the thrombocytopenia revealed the presence of IgG and IgM antiplatelet antibodies and negative antinuclear antibody and human immunodeficiency virus tests. Bone marrow examination showed an increase in megakaryocytes compatible with ITP. In addition to the prednisone given for his ulcerative colitis, he was administered IV Ig, resulting in a gradual increase in the platelet count to 146,000/mL by 9 months later and has remained normal since then. Ten months after he was initially seen, the patient experienced a spontaneous thrombosis of his iliac vein; evaluation for a hypercoagulable state was unrevealing.
A 5-year-old girl was seen in October of 1997 with increased bruising, hematochezia, and cutaneous petechiae 1 week after a crouplike illness. Family history was only notable for a maternal great aunt with systemic lupus erythematosus. Initial laboratory evaluation showed a hematocrit of 33.7%, white blood cell count of 10,860/mL, and platelet count of 4,000/mL. She was diagnosed with ITP and was given Rh o (D) IV Ig, and her platelet count rose to 80,000/mL by the next week. Ten days later, she had a fever and petechiae develop and had a platelet count of 8,000/mL. A bone marrow biopsy was performed and showed a normal number of megakaryocytes, consistent with the diagnosis of ITP. She was started on prednisolone (2 mg/kg/day). The patient was treated with corticosteroids and subsequently both Rh o (D) IV Ig and IV Ig.
By August of 1998, because of persistent thrombocytopenia, the patient was maintained on corticosteroids and Rh o (D) IV Ig. Over the next several months, the patient experienced intermittent abdominal cramping and bloody diarrhea; this was temporally associated with infusions of Rh o (D) IV Ig. Because of her persistent gastrointestinal symptoms, she underwent colonoscopy in May 1999. At this time, diffuse colitis and histologic findings of chronic active colitis extending from the cecum to the rectum were identified. An upper gastrointestinal series with small bowel follow-through was negative. She was diagnosed with ulcerative colitis with backwash ileitis, and sulfasalazine (50 mg/kg/day) was initiated. Since starting the sulfasalazine, her ulcerative colitis and ITP have remained well controlled, despite weaning off steroids. She has not received any further Rh o (D) IV Ig or IV Ig infusions, and her platelet counts have remained approximately 100,000/mL.
A 14-year-old boy was seen in December of 1998 with rectal bleeding. Initial laboratory evaluation showed a hematocrit of 41.5%, white blood cell count of 11,200/mL, and platelet count of 183,000/mL. Stool evaluation was negative for infectious pathogens. Colonoscopy revealed endoscopic and microscopic evidence of proctitis. Antineutrophil cytoplasmic antibody was positive at 96.9 EU/mL. He was diagnosed with ulcerative proctitis and therapy with hydrocortisone enemas and 5-aminosalicylic acid suppositories was initiated; later olsalazine was added. He had a flare of his disease in September 1999, with increasing abdominal pain, diarrhea, and rectal bleeding. Over the next 2 months, other medications were added to his regimen to induce remission, including prednisone, mesalamine replacing olsalazine, metronidazole, and 6-mercaptopurine. In November 1999, the patient's rectal bleeding persisted, and repeat colonoscopy revealed endoscopic and microscopic evidence of universal ulcerative colitis with extension of his disease to the cecum. He was treated with tacrolimus. Blood work monitoring revealed a platelet count of 51,000/mL, with hematocrit of 37.2%, reticulocyte count of 2%, and white blood cell count of 9,700/mL. (The platelet count had decreased to 97,000/mL before therapy with tacrolimus.) He was diagnosed with ITP and treated with continued steroid therapy and IV Ig infusions. Initially, the patient responded, but over time, the IV Ig stopped working, and dexamethasone was started with some response to the high-dose steroids. In February 2000, the patient had fulminant colitis unresponsive to medical therapy develop and underwent colectomy with loop ileostomy with J pouch and closure of the ostomy 6 weeks later. Tacrolimus therapy was stopped after surgery. Since then, he has continued to have episodes of thrombocytopenia treated with IV Ig and steroids.
In this series, we describe eight children diagnosed with IBD and ITP. Of the eight patients, there were four patients with ulcerative colitis, three with Crohn disease, and one patient with indeterminate colitis. All the patients had IBD involving the colon. Three children had IBD preceding ITP, three children had ITP preceding IBD, and in two children, both the IBD and ITP presented simultaneously. Two children had acute ITP, and the other six children had chronic ITP lasting longer than 6 months (see Table 1). We identified only five other reports of children younger than 18 years, who had both IBD and ITP develop. (7,11)
Childhood ITP is usually a self-limited illness that affects previously healthy children, with the peak age range of 4 to 8 years of age. Traditionally, childhood ITP is divided into acute and chronic forms based on the duration of thrombocytopenia. (9) Persistence of ITP for more than 6 months defines the chronic form, which accounts for only 10% to 20% of children with ITP. (12) In the case of acute ITP, there is a history of a preceding viral illness, and the onset of purpuric symptoms is typically abrupt. In contrast, in chronic ITP, children are usually older than 10 years, antecedent intercurrent illnesses are less common, the onset of purpura is more insidious, and relapses are common.
Children with ITP have purpura and mucosal bleeding without evidence of hepatomegaly, splenomegaly, or lymphadenopathy. Gastrointestinal or renal hemorrhage can occur. Central nervous system bleeding occurs in less than 1% of individuals with ITP. Rectal bleeding is an uncommon presentation in ITP. Laboratory evaluation reveals thrombocytopenia with an otherwise normal complete blood count, unless bleeding is sufficient to cause anemia. Bone marrow aspirate demonstrates normal to increased numbers of megakaryocytes. A test for antiplatelet antibodies (PAIgG) may be positive, but a negative test does not exclude the diagnosis.
Other causes for thrombocytopenia need to be considered in the differential diagnosis, including drug-induced thrombocytopenia. Three of our patients were taking medications at the time of their ITP diagnosis: one patient was being treated with 6-mercaptopurine and mesalamine, another was taking sulfasalazine, and the third was taking tacrolimus. In the two former cases, bone marrow evaluation showed findings consistent with ITP, thus excluding a hypocellular marrow suggestive of a drug-induced thrombocytopenia. In the latter case, with the patient taking tacrolimus, no bone marrow evaluation was performed. However, the platelet count had decreased to 97,000/mL before initiating therapy with tacrolimus, and the thrombocytopenia recurred after the tacrolimus was stopped. Therefore, it is unlikely that the ITP in this case was caused by the tacrolimus.
In contrast to the general pediatric population, in our case series, most children had the chronic form of ITP. Interestingly, the younger children in our case series were the individuals who tended to have chronic ITP; in contrast, chronic ITP in the general population is seen in older children (see Table 1). The more than expected percentage of chronic ITP cases in the IBD population compared with the general population may suggest a unique relationship between these two disorders. Chronic ITP is associated with other disorders with underlying immune dysregulation, including systemic lupus erythematosus, IgA deficiency, autoimmune endocrinopathy, common variable immunodeficiency, and Evans syndrome (13).
Investigators have hypothesized that pathogenesis of both IBD and ITP involves immune dysregulation. However, it is unlikely that a single molecular or cellular abnormality will explain the immune dysregulation in both IBD and ITP. Crohn disease and ulcerative colitis are both characterized by the presence of activated mucosal macrophages and T lymphocytes. In Crohn disease, there is increased mucosal production of TH1-type proinflammatory cytokines (e.g., interleukin-2, interferon-gamma, and tumor necrosis factor-α), whereas in ulcerative colitis, there may be preferential release of TH2-type cytokines (e.g., interleukin-4). (14) Studies of serum or cultured lymphocytes from children with acute ITP suggest increased levels of interleukin-2 and interferon-gamma production, suggesting the possibility that a viral illness may induce a TH1-type response directed against platelets. (15–17) Therefore, current data suggest that a TH1 helper immunologic response may mediate both acute ITP and Crohn disease. Most of our patients with acute and chronic ITP had ulcerative colitis, suggesting that current immunologic hypotheses are inadequate to explain the association of these two diseases. It also has been postulated that an enhanced exposure to luminal antigens, secondary to increased mucosal permeability in IBD, may lead to increased opportunities for the production of antibodies to these antigens which may result in cross-reactivity with platelet antigens. (7) One might argue that this may not account for those cases in which the ITP precedes the IBD. However, subclinical injury of the gut may have coincided with the ITP before clinical diagnosis of the IBD.
In summary, we present the largest case series reported to date of children with both IBD and ITP. Most likely this series underestimates the number of actual cases, given that only a subset of pediatric gastroenterologists subscribe to the pediatric gastroenterology internet bulletin board, and not all subscribers will respond to the request for cases. In our series, ITP was associated with both Crohn disease and ulcerative colitis, but colonic involvement was identified in all cases. In most patients, the ITP was chronic but responded to conventional treatment for ITP and colitis. A prior case report suggests that ITP resolves after colectomy in patients with ulcerative colitis. (18) However, in our series, the one patient with ulcerative colitis who underwent colectomy did not show improvement in his ITP course postoperatively. Thus, it is unclear whether colectomy for ulcerative colitis improves ITP symptoms. Larger based studies may be helpful to better delineate the natural history and outcome of patients with both IBD and ITP. Given the potential association between IBD and ITP, we suggest evaluation for IBD in a patient with ITP and prolonged rectal bleeding.
1. Rankin G. Extraintestinal and systemic manifestations of inflammatory bowel disease
. Med Clin North Am 1990; 74: 39–50.
2. Hyams J. Extraintestinal manifestations of inflammatory bowel disease
in children. J Pediatr Gastroenterol Nutr 1994; 19: 7–21.
3. Talstad I, Rootwelt K, Gjone E. Thrombocytosis in ulcerative colitis
and Crohn's disease
. Scand J Gastroenterol 1973; 8: 135–8.
4. Altman A, Maltz C, Janowitz H. Autoimmune hemolytic anemia in ulcerative colitis
: Report of three cases, review of the literature, and evaluation of modes of therapy. Dig Dis Sci 1979; 24: 282–5.
5. Stevens C, Peppercorn M, Grand R. Crohn's disease
associated with autoimmune neutropenia. J Clin Gastroenterol 1991; 13: 328–30.
6. Gupta S, Saverymuttu S, Hodgson H, et al. Immune thrombocytopenic purpura
, neutropenia and sclerosing cholangitis associated with ulcerative colitis
in an adult. Clin Lab Haematol 1986; 8: 67–9.
7. Zlatanic J, Korelitz B, Wisch N, et al. Inflammatory bowel disease
and immune thrombocytopenic purpura
: Is there a correlation? Am J Gastroenterol 1997; 92: 2285–8.
8. George J, Raskob G. Idiopathic thrombocytopenic purpura: A concise summary of the pathophysiology and diagnosis in children and adults. Semin Hematol 1998; 35: 5–8.
9. Kuhne T, Elinder G, Blanchette V, et al. Current management issues of childhood and adult immune thrombocytopenic purpura
(ITP). Acta Paediatr Suppl 1998; 424: 75–81.
10. Kocoshis S, Gartner J, Gaffney P, et al. Thrombocytopenia
in ulcerative colitis
. J Pediatr 1979; 95: 83–4.
11. Boyne M, Dye K. Crohn's colitis and idiopathic thrombocytopenic purpura. Postgrad Med J 2000; 76: 299–300.
12. Kuhne T, Imbach P. Chronic immune thrombocytopenic purpura
in childhood. Semin Thromb Hemost 1998; 24: 549–53.
13. Casella J, Bowers D, Pelidis M. Disorders of platelets. In: DeAngelis C, Feigin R, Warshaw J, eds. Oski's Pediatrics: Principles and Practice. Philadelphia: Lippincott Williams & Wilkins; 1999: 1475–81.
14. Sartor R. Current concepts of the etiology and pathogenesis of ulcerative colitis
and Crohn's disease
. Gastroenterol Clin North Am 1995; 24: 475–507.
15. Imbach P, Kuhne T, Hollander G. Immunologic aspects in the pathogenesis and treatment of immune thrombocytopenic purpura
in children. Curr Opin Pediatr 1997; 9: 35–40.
16. Semple J. Immunobiology of T helper cells and antigen-presenting cells in autoimmune thrombocytopenic purpura (ITP). Acta Paediatr Suppl 1998; 424: 41–5.
17. Andersson J. Cytokines in idiopathic thrombocytopenic purpura (ITP). Acta Paediatr Suppl 1998; 424: 61–4.
18. Bauer W, Litchtin A, Lashner B. Can colectomy cure immune thrombocytopenic purpura
in a patient with ulcerative colitis
? Dig Dis Sci 1999; 44: 2330–3.