American Journal of Hematology 67:51–53 (2001) Reversible Sideroblastic Anemia Associated With the Tetracycline Analogue COL-3 Michelle A. Rudek,1 McDonald Horne,2 William D. Figg,1* William Dahut,1 Valerie Dyer,1 James M. Pluda,3 and Eddie Reed1 1 Medicine Branch, Division of Clinical Sciences, National Cancer Institute, National Institutes of Health, Bethesda, Maryland 2 Clinical Pathology Department, W.G. Magnuson Clinical Center, Bethesda, Maryland 3 Investigational Drug Branch, Cancer Therapy Evaluation Program, Division of Cancer Treatment and Diagnosis, National Cancer Institute, National Institutes of Health, Bethesda, Maryland Eight of 35 patients with cancer receiving COL-3, a tetracycline derivative with antiangiogenic properties, developed anemia while on treatment. All of these patients were enrolled on an approved Phase I clinical trial at the National Cancer Institute. Three of these patients had bone marrow examinations that revealed ringed sideroblasts. This paper describes these cases. Am. J. Hematol. 67:51–53, 2001. Published 2001 Wiley-Liss, Inc.† Key words: anemia; COL-3; tetracycline; sideroblast; cancer; Phase I INTRODUCTION Drugs cause anemia by shortening the red cell life span or by intoxicating erythropoiesis. An unusual example of the latter is sideroblastic anemia, which is characterized by the accumulation of iron in the mitochondria of erythroblasts and has been associated with antituberculin agents, chloramphenicol, and ethanol. Here we report the development of sideroblastic anemia in 3 patients who received a non-antimicrobial tetracycline analogue, COL-3. COL-3, 6-demethyl-6-deoxy-4-dedimethylaminotetracycline, is an orally administered drug currently undergoing Phase I investigation for treatment of refractory tumors at the National Cancer Institute [1]. Of 35 patients enrolled on study at 36, 50, 70, or 98 mg/m2/day, eight patients developed anemia without leukopenia or thrombocytopenia. Two patients were anemic prior to starting treatment but became more anemic with COL-3 treatment. Six additional patients who had normal hemoglobin levels at baseline had drops in their hemoglobin during COL-3 treatment. One of these six patients had occult blood detected in his stool. The hemoglobin of another patient fell 1.7 g/dl during times of major fluid shifts, suggesting that the etiology of his anemia was largely dilutional. The other four patients had unexplained decreases in hemoglobin of 3.4–5.2 g/dl during the initial 2 months of COL-3 treatment. Three of these patients underwent bone marrow examination. Each of these three Published 2001 Wiley-Liss, Inc. † This article is a U.S. Government work and, as such, is in the public domain in the United States of America. had ringed sideroblasts in the bone marrow and clinical evidence to suggest that the anemia was possibly induced by COL-3. CASES DS is an African-American male with metastatic follicular thyroid carcinoma, who was diagnosed at 56 years of age. After a series of treatments over 8 years, he started oral COL-3 therapy at a dose of 200 mg/day. Previous treatments for his malignancy included thyroidectomy, two courses of I-131, and external beam radiation to 40 Gray. Co-morbid illnesses included mild adult-onset diabetes mellitus, hypertension, hypercholesterolemia, benign prostatic hypertrophy, peptic ulcer disease, and erectile dysfunction. Medications included doxazosin, daily enteric coated aspirin, levothyroxine, omeprazole, pravastatin, and depo-testosterone. Overthe-counter medications that were frequently used included acetaminophen and yohimbine. Contract grant sponsor: U.S. Government; Contract grant sponsor: Intramural Program of the National Cancer Institute and the Office of Research on Minority Health, National Institutes of Health. *Correspondence to: William D. Figg, National Cancer Institute, Building 10, Room 5A01, 9000 Rockville Pike, Bethesda, MD 20892. E-mail: wdfigg@helix.nih.gov Received for publication 18 July 2000; Accepted 15 November 2000 52 Brief Report: Rudek et al. TABLE I. Blood Indices at Time of First Bone Marrow Biopsy Hgb (g/dl) Mean cell volume (MCV) WBC (K/ml) Plt DS TY JM 6.7 76 10.7 637,000 7.4 90 5.2 244,000 8.7 83 5.7 277,000 During the first 3 weeks of treatment, DS’s hemoglobin varied between 9.3 and 10.4 g/dl without any trend. Between Day 25 and Day 53, however, his hemoglobin fell from 9.3 to 6.7 g/dl. His white blood count (10.7 K/ml) and platelet count of (637,000/ml) remained normal (see Table I). The patient received 3 units of packed red blood cells. COL-3 was reduced to 140 mg/day for 6 days and then was stopped because of documented progression of the patient’s malignant disease. Evaluation of the anemia revealed microcytosis (a drop in MCV from 81 to 76 fl by Day 53) and a negative stool test for occult blood. A bone marrow biopsy performed 7 days after COL-3 was stopped showed erythroid hyperplasia with 26% ring sideroblasts. COL-3 was never restarted because of the demonstration of progressive disease. TY is a Caucasian male with nodular sclerosing Hodgkin’s lymphoma diagnosed at age 32. He started COL-3 three years after diagnosis at a dose of 130 mg/day. Prior treatments included doxorubicin, bleomycin, vinblastine, and dacarbazine (ABVD × 6 cycles), two courses of radiation (45 Gray), and an allogenic bone marrow transplantation regimen. TY took occasional aspirin for headaches and lansoprazole 30 mg for 4 days, during COL-3 for the treatment of occasional heartburn symptoms. After 57 days of COL-3 therapy, the patient’s hemoglobin had fallen from 12.6 g/dl, before therapy, to 7.4 g/dL. The white blood count (5.2 × 103 K/ml), and platelet count of (244,000/ml) were normal and unchanged from baseline (see Table I). The patient received 3 units of packed red blood cells. Evaluation of the anemia revealed normal red blood cell indices and a negative test for occult blood in the stool. Bone marrow examination 2 days after COL-3 was stopped showed moderate hypocellularity and 61% ring sideroblasts. The decision not to restart COL-3 was made by the patient, who indicated that he felt better off the drug. JM is a Caucasian female with metastatic low-grade spindle cell sarcoma to the lungs diagnosed at age 41. At this time, the mass was removed by surgery. When the fibrosarcoma relapsed approximately 5 years later, she was treated with two cycles of mesna, doxorubicin, ifosfamide, and dacarbazine (MAID). Six years after diagnosis, JM began COL-3 at a dose of 120 mg/day. She took bactrim DS for 5 days during COL-3 therapy for a urinary tract infection. After 57 days of COL-3, the patient’s hemoglobin had fallen from 12.4 to 8.7 g/dL (see Fig. 1). Her white blood count, platelet Fig. 1. Clinical course of patient JM. Periods of COL-3 administration are shown by the shaded horizontal boxes. Bone marrow examinations were performed at the times indicated by the vertical arrows and revealed the percentages of ring sideroblasts indicated. count, and red blood cell indices remained normal. The reticulocyte count was 2.0%. A test for occult blood in the stool was negative. A bone marrow examination performed 7 days after COL-3 was stopped, when the reticulocyte count was 7.7%, showed moderate erythroid hyperplasia, increased storage iron, and 5% ringed sideroblasts. The anemia improved over the next 3 weeks to a hemoglobin of 11.2 g/dl. At this point, COL-3 was restarted at 90 mg/day. After 75 days at this dose, the hemoglobin had fallen to 9.1 g/dl without a change in the white count or the platelet count. There was no evidence of bleeding or hemolysis. A repeat bone marrow examination again showed erythroid hyperplasia but with 47% ringed sideroblasts (Fig. 2A). Myeloid and megakaryocyte maturation was normal. COL-3 was stopped. The anemia resolved by 35 days off treatment without a transfusion (Fig. 1). A third bone marrow examination was performed, 35 days after stopping COL-3. Evaluation revealed total absence of the ringed sideroblasts (Fig. 2B). COL-3 was not restarted. DISCUSSION COL-3 is a derivative of tetracycline with antiangiogenic properties that lacks the antimicrobial properties of the parent compound. The tetracycline family has been implicated as a cause of anemia. There are several reports of tetracycline-induced aplastic anemia [2], megaloblastic anemia [3], and hemolytic anemia [4–8]. Oxytetracycline has been associated with aplastic anemia [8] while another derivative minocycline has been implicated in hemolytic anemia [9]. There are no previous reports of sideroblastic anemia associated with any of the tetracycline derivatives. In our trial, eight patients became anemic or more anemic while taking COL-3. In five of them the anemia could be explained by clinical events (bleeding, excessive IV fluid, disease progression). For the remaining Brief Report: COL-3 Sideroblastic Anemia 53 Fig. 2. A. Prussian blue stain of the second marrow aspirate from patient JM in which 47% of the nucleated red cells were ring sideroblasts. B. Prussian blue stain of the final marrow aspirate from JM in which ring no ring sideroblasts were identified. three, however, only the administration of COL-3 was temporally related to the anemia, and COL-3 appeared to be the causative agent. These three patients denied current ethanol use. One patient was rechallenged with COL-3 and developed anemia again (Fig. 1). Furthermore, bone marrow from each of these patients contained ringed sideroblasts during the anemia. In JM, this unusual morphologic feature was documented to have disappeared within 4 weeks of discontinuing COL-3, coinciding with resolution of the anemia. Ringed sideroblasts (Fig. 2) are nucleated red cells with stainable iron deposits in their mitochondria. Other than in rare hereditary anemias and primary bone marrow neoplasias, ringed sideroblasts are only associated with exposures to toxins, including ethanol, lead, and a small number of medications (chloramphenicol, cycloserine, ethionamide, pyrazinamide, and isoniazid) [10–13]. Therefore, the presence of ringed sideroblasts in the bone marrows of three anemic patients receiving COL-3 strongly suggests a toxic effect of this agent on the erythropoiesis of these individuals. COL-3 is a derivative of tetracycline. Although tetracycline has been associated with a variety of anemias (aplastic, hemolytic, megaloblastic), it has never been associated with ringed sideroblasts. The molecular mechanisms by which COL-3 may cause sideroblastic anemia are quite unclear. Other toxins associated with sideroblastic anemia appear to impair heme synthesis by interfering with pyridoxine metabolism [11,13]. COL-3, however, is an inhibitor of matrix metalloproteinases, which degrade extracellular proteins during tissue remodeling. Although an effect on erythropoiesis might be mediated by alterations in the marrow stroma, such a proposed mechanism is purely speculative. COL-3 may be a potent divalent cation chelator like other tetracycline analogues and may be responsible for chelating the iron. In conclusion, we feel that COL-3 may be implicated in at least 3 of the 8 cases of anemia seen during this Phase I study of this agent (a total of 35 patients were treated). These 3 cases are described above. We suggest that routine laboratory examination for anemia should be performed when this agent is administered. REFERENCES 1. Rudek MA, Figg WD, Dyer V, Dahut W, Turner ML, Steinberg S, Kohler DR, Pluda JM, Reed E. A phase I clinical trial of oral COL-3, matrix metalloproteinase inhibitor, in patients with refractory metastatic cancer. J Clin Oncol 2001;19:584–592. 2. Lehrner LM, Cooke JH, Enck RE. Tetracycline-induced aplastic anemia. Southern Med J 1979;72:358–361. 3. Jones CC. Megaloblastic anemia associated with long-term tetracycline therapy. Report of a case. Ann Intern Med 1973;78:910–912. 4. Wenz B, Klein RL, Lalezari P. Tetracycline-induced immune hemolytic anemia. Transfusion 1974;14:265–269. 5. Schalow DM. Tetracycline—a case of hemolysis. Am J Med Technol 1980;46:115. 6. Mazza JJ, Kryda MD. Tetracycline-induced hemolytic anemia. J Am Acad Dermatol 1980;2:506–508. 7. Simpson MB, Pryzbylik J, Innis B, Denham MA. Hemolytic anemia after tetracycline therapy. New Engl J Med 1985;312:840–842. 8. Kishore B, Bansal OP, Hazra DK, Seth HC. Aplastic anemia associated with oxytetracycline therapy. Indian J Med Sci 1969;23:137–139. 9. Kudoh T, Nagata N, Suzuki N, Nakata S, Chiba S, Takahashi T. Minocycline-induced hemolytic anemia. Acta Paediatr Jpn 1994;36:701–704. 10. Saidi P, Wallerstein RO, Aggeler PM. Effect of chloramphenicol on erythropoiesis. J Lab Clin Med 1961;57:247–256. 11. MacGibbon BH, Mollin DL. Sideroblastic anemia in man: observations on seventy cases. Br J Haematol 1965;11:59–69. 12. Verwilghen R, Reybrouck G, Callens L, Cosemans J. Antituberculous drugs and sideroblastic anemia. Br J Haematol 1965;11:92–95. 13. Bowman WD. Abnormal (“ringed”) sideroblasts in various hematologic and non-hematologic disorders. Blood 1962;18:662–671.