Fetal hemolytic anemia associated with maternal sulfasalazine therapy during pregnancy

Abstract

A 29 year-old woman with a diagnosis of ulcerative colitis for 11 years was admitted to the Department of Obstetrics in pregnancy week 26. During the whole course of her pregnancy, she had been treated with sulfasalazine (Salazopyrin EN) in a 1-g twice-daily dose. She had had an intrauterine fetal death at 25 weeks of gestation three years earlier. No fetal malformation or chromosomal aberration was found. Two years later she gave birth to a healthy girl at term. During both of these pregnancies she was treated with sulfasalazine. Additional medications during the two pregnancies were folic acid (Folacin) 0.4 mg daily and ferrous sulfate (Duroferon) 100 mg daily. Her blood group was 0 Rh-positive without signs of immunization. The patient became pregnant for the third time while she was on sulfasalazine 1-g twice-daily treatment in addition to folic acid in a daily dose of 0.4 mg. After 26 weeks of pregnancy, an ultrasound was performed due to high symphysis–fundus level and fetal ascites was discovered. No malformations were found. Amniocentesis with chromosomal analysis showed normal male karyotype. Due to high blood velocity in the fetal middle cerebral artery, there was a suspicion of severe anemia as an explanation of the noted nonimmune hydrops. A chordocentesis was performed which showed a hemoglobin level of 34 g/l and reticulocytes of 22%. Severe anemia was diagnosed and the fetus was treated with intrauterine blood transfusions. During gestational weeks 26–31, the fetus received in total three intrauterine blood transfusions, and in gestational week 31 the post-transfusion hemoglobin level was 117 g/l. Tests of maternal blood for parvo- and cytomegalovirus infections by PCR, and IgG and IgM antibodies were all negative. Kleihauers test of maternal blood for fetomaternal transfusion was negative. No blood group incompatibility was found and DAT test was negative. No signs of hemolytic anemia were found in maternal blood. The limited amount of fetal blood received from chordocentesis showed normal differential white cell count. There was no glucose-6-phosphatedehydrogenase deficiency and no abnormal hemoglobin chains were found on cationic exchange chromatography. The limited amount of fetal blood available did not allow determination of salazopyrine concentration. However, free hemoglobin concentration was 244 mg/l in fetal blood plasma, confirming the diagnosis of fetal hemolytic anemia and an adverse reaction to sulfasalazine was suspected. The sulfasalazine medication was therefore withdrawn and betamethasone therapy 12 + 12 mg intramuscular was initiated in week 30 of pregnancy. Due to premature contractions, a cesarean section was performed at 31 weeks of pregnancy. A baby boy with a birth weight of 2,590 g was delivered with an Apgar score of 3-6-7. During cesarean section, 7 l of bloodstained amniotic fluid was emptied and there were signs of partial placental abruption. The first hemoglobin value after birth was 90 g/l. A partial exchange transfusion followed by blood transfusions was made postpartum which gradually elevated the hemoglobin value to 173 g/l. The intrauterine anemia was judged to be of hemolytic origin with high levels of reticulocytes indicating an active bone marrow. The infant was intubated and mechanical ventilation was initiated shortly after birth. He was tapped for 100 ml of ascites, and could be extubated to continuous positive airway pressure (CPAP) after two days. He was treated with supplemental oxygen for a total of 12 days. Neurological investigation and viral serology showed normal results. Further growth continued uncomplicated and he was discharged from hospital after one month at which time the hemoglobin value was 111 g/l. Psychomotor evaluation was normal at 5 and 10 months of age. Sulfasalazine is a well established therapeutic alternative in the management of ulcerative colitis and Crohn's disease and this drug has also been evaluated in the treatment of rheumatoid arthritis 1, 2. In patients with ulcerative colitis, sulfasalazine is particularly used in maintenance treatment and also in prevention of relapse 3. When used in pregnant patients, it is often desirable to continue treatment during the whole period of pregnancy and breastfeeding 4. Such therapy has generally been considered safe, and sulfasalazine has generally been associated with no restrictions during use in pregnancy 5. However, previous studies 6, 7 show that sulfasalazine and its metabolites reach the fetus and can be found in concentrations corresponding to those found in the maternal serum. It may therefore be reasonable to assume that the fetus is exposed to similar risks for hematological complications as those arising from the treatment of patients in general. Moreover, the newborn infant may also be exposed to sulfasalazine during breastfeeding, but the concentrations of sulfasalazine and its metabolites in breast milk are much lower than those of maternal serum and are therefore unlikely to cause concentration-dependent harmful side effects 6. When sulfasalazine is administrated to pregnant women in therapeutic doses, about 98–99% of sulfasalazine is protein-bound in plasma 8 and approximately 31–40% of the metabolite sulfapyridine (SP) is protein-bound 9. Thus, during the major portion of gestation, the fetus is exposed to sulfasalazine in concentrations largely similar to those found in the pregnant mother, while the concentration of SP is estimated to be lower in the fetus. Interestingly, as SP has a lower degree of protein binding than sulfasalazine, it should have a larger distribution volume in the pregnant woman than sulfasalazine. Moreover, as sulfasalazine is protein-bound to a higher degree than SP, it may also have a larger bilirubin-displacing capacity than SP 7. The present report illustrates a case with fetal hemolytic anemia occurring during pregnancy associated with sulfasalazine therapy for maternal ulcerative colitis. Although the amount of blood retrieved by chordocentesis was not sufficient to determine the sulfasalazine level in fetal blood, the hemolysis was confirmed and some other possible causes of hemolysis could be excluded based on the limited amount of blood available. Treatment by intrauterine blood transfusions improved fetal hemoglobin concentrations. Betamethasone therapy was instituted to prevent further anemia until delivery. Furthermore, later follow-up of the patient demonstrated that the postnatal period was uneventful. Fetal hemolytic anemia in relation to maternal sulfasalazine treatment as illustrated in this case has not to our knowledge previously been reported in the literature. However, a previous report 10 described a case of fetal death due to hydrops and aplastic anemia developing after maternal sulfasalazine and prednisone treatment of a mother with chronic ulcerative colitis. In this case there was no clear etiology to the fetal cause of death, but sulfasalazine was suspected to be a contributory factor. Furthermore, Levi et al. 11 described a case of reversible congenital neutropenia associated with sulfasalazine and prednisone therapy due to maternal Crohn's disease. Also in this case, the maternal sulfasalazine treatment was a suspected cause of the fetal hematological disorder. However, the majority of cases treated with sulfasalazine during pregnancy result in normal deliveries 12 without known teratogenic sequelae 13. In a review of the Hungarian case control surveillance registry of 22,865 index cases of congenital abnormalities, 0.07% of these were treated with sulfasalazine and a similar proportion of sulfasalazine treatment was observed in the 38,151 control infants without known congenital abnormalities. Thus, from this registry there was no evidence of an increased prevalence of congenital abnormalities in children of women treated with sulfasalazine during pregnancy 13. In an investigation of the Swedish Birth Register (MFR) by Professor Bengt Källén from 1995 to 2001 he found 1,165 children born to mothers on aminosalicyl preparations, mostly Salazopyrin. In one of these cases hemolytic anemia with an initial hemoglobin count of 57 g/l was diagnosed immediately after birth at term. High reticulocyte count (15%) immediately after birth gave suspicion of chronic anemia debuting before birth. The mother had been treated with Salazopyrin for Crohn's disease and had a history of intrauterine fetal death in pregnancy week 30. The newborn baby was treated with two blood transfusions and recovered well. Figure 1 Hemoglobin and reticulocyte values of the fetus/infant. ↑ indicate blood transfusions given to the fetus/infant and ⇓ indicates betamethasone given to the mother. Due to the fact that a history of intrauterine fetal death was reported both by our patient and in the above-mentioned case, we performed a separate investigation in the MFR of intrauterine fetal death. In 546 women on aminosalicylic preparations in 1995–2002, four cases was found. This is a suggestive although not significant (p=0.1) increase from the expected 1.3 cases. In adults, a wide range of hematological adverse reactions to sulfasalazine have been reported such as Heinz body anemia, hemolysis, methemoglobulinemia, and thrombocytopenia 11. Hemolytic anemia is a known adverse effect to sulfasalazine 14, but other hematological complications appear to be rare 11. In adults, a number of drugs have been reported to induce hemolytic anemia such as methyldopa for hypertension, levodopa used for Parkinson's disease, mefenamic acid, a nonsteroidal anti-inflammatory drug, interferon-alpha, used in chronic viral hepatitis, cyclosporine used for prevention of graft rejection and the treatment of certain autoimmune diseases, and fludarabin, used in chronic lymphoid leukemia 15. Thus, drug-induced autoimmune hemolytic anemia requires that the drug have an effect on both autoantigens and on the immune system. Drugs may induce immunohemolytic anemia by either inducing a disorder resembling warm-antibody immunohemolytic anemia or by becoming associated as haptenes with the red blood cells (RBC) surface, which induces formation of antibodies directed against the RBC–drug complex. A positive direct Coombs test is sometimes observed. In the first case, the drug alters and makes immunogenic the protein(s) of the Rh complex and the resulting antibodies cross-react with the Rh protein and not the drug itself. The hemolysis present decreases over the course of several weeks after cessation of drug therapy, but the direct Coombs test may remain positive for a longer period. In the second case, which represents most other cases of drug-induced hemolysis, the antibody is directed against the drug and the membrane glycoprotein combination. In those cases the hemolytic reaction itself is dependent on the drug and usually ceases shortly after discontinuation of the drug therapy. In this form of hemolytic anemia the drug is firmly adhered to the protein of the RBC membrane. The hemolysis in vivo is usually not severe. However, in some cases hemolysis may be more severe, resulting in signs of intravascular hemolysis. Also in such cases there is a rapid resolution 16. In the present case, the most likely mechanism would be that the hemolytic reaction was related to antibody formation to a sulfasalazine–membrane glycoprotein complex, i.e. the second option mentioned. This is also the reaction commonly encountered in sulfonamide hemolytic anemia. Further the clinical cause and relatively rapid resolution of the fetal hemolytic anemia in the present case also favors this explanation. In the management of drug-induced fetal hemolytic anemia, maternal steroid treatment is usually started. In the present case such treatment was instituted in combination with intrauterine blood transfusions. This therapeutic strategy resulted in a clinical remission. Thus, the case presented illustrates a new potential risk of sulfasalazine therapy during pregnancy and a therapeutic strategy in cases with suspected drug-induced hemolytic anemia. Other therapeutic options for the treatment of inflammatory bowel disease in pregnant women should be considered if judged safe for the mother. Close cooperation between gastroenterologists, obstetricians, pediatricians, laboratory specialists, and clinical pharmacologists is of utmost importance in the management of pregnant women with inflammatory bowel disease requiring drug treatment. To Professor Bengt Källén for invaluable help searching the Swedish Birth Register (MFR).