as well as those reported by others 6 show that splenectomy can be safely performed with good results even in patients who do not respond to PE. Other authors have done splenectomy to prevent relapses in TTP, co nfirming a reductio n o f the relapse rate with minimal mo rbidity. 5,7 Overall, these findings suggest that splenectomy reduces the frequency of relapses and can induce prolonged remissions in refractory patients. The reason for clinical improve- ment after splenectomy remains elusive. Recently, the body of evidence supporting an autoimmune mech- anism as the primary cause of endothelial damage in TTP has increased. 9 If the preliminary reports sug- gesting an auto immune basis fo r TTP are co nfirmed, 10 these findings could contribute to a better under- standing o f the ro le that the spleen plays in the patho - genesis of this disease. Javier de la Rubia, Inmaculada García, Isidro Jarque, Francisco Arriaga, Federico Gomis, Miguel A. Sanz Department of Hematology, La Fe University Hospital, Valencia, Spain Keywords Thrombotic thrombocytopenic purpura, refractory, relapse, splenectomy. Correspondence Dr. Javier de la Rubia, M.D., Servicio de Hematología, Hospital Universitario La Fe, Avda. Campanar 21, 46009 Valencia, Spain. Phone: international +34-6-3862721 – Fax: international +34-6-3868757 – E-mail: delarubia- jav@ gva.es References 1. Amorosi EL, Ultmann JE. Thrombotic thrombocy- to penic purpura. Repo rt o f 16 patients and review o f the literature. Medicine (Baltimore) 1996; 139-59. 2. Geo rge JN, Gilcher RO, Smith JW, Chandler L, Duvall D, Ellis C. Thrombotic thrombocytopenic purpura- hemolytic uremic syndrome: diagnosis and manage- ment. J Clin Apheresis 1998; 13:120-5. 3. Ro ck GA, Shumak KH, Buskard NA, et al. Co mpariso n of plasma exchange with plasma infusion in the treat- ment of thrombotic thrombocytopenic purpura. Canadian Apheresis Study Gro up. N Engl JMed 1991; 325:393-7. 4. Shumak KH, Rock GA, Nair RC. Late relapses in patients successfully treated for thrombotic throm- bocytopenic purpura. Canadian Apheresis Group. Ann Intern Med 1995; 122:569-72. 5. Veltman GA, Brand A, Leeksma OC, ten Bo sch GJ, van Krieken JH, Briët E. The role of splenectomy in the treatment o f relapsing thro mbo tic thro mbo cyto penic purpura. Ann Hematol 1995; 70:231-6. 6. Winslow GA, Nelson EW. Thrombotic thrombocy- topenic purpura: indications for and results of splenectomy. Am J Surg 1995; 170:558-63. 7. Crowther MA, Heddle N, Hayward CP, Warkentin T, Kelton JG. Splenectomy done during hematologic remission to prevent relapse in patients with throm- botic thrombocytopenic purpura. Ann Intern Med 1996; 125:294-6. 8. de la Rubia J, López A, Arriaga F, et al. Response to plasma exchange and steroids as combined therapy for patients with thrombotic thrombocytopenic pur- pura. Acta Haematol 1999; 102:12-6. 9. Porta C, Caporali R, Montecucco C. Thrombotic thrombocytopenic purpura and autoimmunity: a tale of shadows and suspects. Haematologica 1999; 84:260-9. 10. Tsai HM, Lian EC. Antibo dies to vo n Willebrand fac- tor-cleaving protease in acute thrombotic thrombo- cytopenic purpura. N Engl JMed 1998; 339:1585-94. Delay of onset of candidemia and emergence of Candida krusei fungemia in hematologic patients receiving prophylactic fluconazole Invasive fungal infections are an important cause of morbidity and mortality in hematologic patients. In spite of newer antifungal approaches candidemia remains a severe condition associated with a high mortality. Since the introduction of fluconazole pro- phylaxis we registered an increasing prevalence of C. krusei fungemia as well as a significant reduction of early-onset candidemia. Sir, Fungal infections have increased substantially over the past two decades in patients with hematologic malignancies. Since the introduction of fluconazole prophylaxis in the early 1990s, changes in the epi- demiology of candidemia have been reported 1-3 with an increase o f infectio ns caused by species o ther than C. albicans, mainly C. krusei . However, a causal rela- tio nship between prio r use o f fluco nazo le and increas- ing number o f infectio ns due to C. krusei remains co n- tro versial. 5-9 We report a retrospective evaluation of 39 can- didemia episodes which occurred in patients with scientific correspondence 441 Haematologica vol. 85(4):April 2000 Table 1. Splenectomy in thrombotic thrombocytopenic purpura. Patients’ main characteristics. UPN Se x/Age Platelet count No. of PE # Before splenectomy Status Response Status at diagnosis (x10 9 /L) Platelet count (x10 9 /L) Follow-up (months) 1 M/ 4 2 3 20 94 Re fra c to ry CR Alive (26) 2 M/ 3 5 25* 5 120 Relapse CR Alive (24) 3 M/ 3 5 10* 28 61 Re fra c to ry Relapse Alive (4) 4 M/ 5 7 5 28 50 Re fra c to ry CR Alive (11) *Platelet count at relapse; # represents the number of days with PE before splenectomy; PE: plasma exchange; CR: complete response. ©Ferrata Storti Foundation