Fixed Low-Dose Hydroxyurea for the Treatment of Adults with Sickle Cell Anemia in Nigeria

To the Editor

Hydroxyurea has not been prospectively evaluated for adults with sickle cell anemia (SCA) in sub-Saharan Africa. A fixed low-dose regimen may have wide applicability where frequent blood count monitoring is not available. We enrolled 53 SCA adults in Ibadan, Nigeria into a study of hydroxyurea 10 mg/kg/day for 24 weeks and a crossover observation period of 24 weeks. There were significant increases in hemoglobin concentration, mean corpuscular volume (MCV), hemoglobin F percent and weight during hydroxyurea therapy (P≤0.007). Only one patient developed the pre-specified primary adverse outcome of platelets <50,000/μL, which resolved with antimalarial treatment and continued hydroxyurea therapy.

Hydroxyurea therapy for SCA increases hemoglobin F, reduces pain crises, acute chest syndrome and blood transfusions, and possibly increases survival[1]. It is typically initiated at 15 mg/kg/day followed by dose escalations up to 35 mg/kg/day if tolerated. Neutropenia and thrombocytopenia are limitations to achieving maximal dose, and frequent blood count monitoring is required. Hydroxyurea is unlicensed for SCA in sub-Saharan Africa where >75% of SCA children are born[2]. Whereas some progress is being made toward wider use[3], a 2014 survey of Nigerian sickle cell clinics found that only 8 of 18 clinics prescribed hydroxyurea[4]. In centers that prescribe hydroxyurea, only 5% to 25% of patients receive the agent[5–7]. Several recent studies have examined standard-dose hydroxyurea in children with SCA in sub-Saharan Africa. Twenty to 25 mg/kg/day decreased secondary stroke in 13 Nigerian children[8] and lowered transcranial Doppler flow velocity in 31 Nigerian children[9]. A feasibility study of 20 mg/kg/day for 24 months for primary stroke prevention found hydroxyurea was well-tolerated in 29 Nigerian children[10]. A 12-month placebo-controlled trial of 20 mg/kg/day in 207 Ugandan children found hydroxyurea was well tolerated, led to a decreased composite outcome of painful crisis, dactylitis, acute chest syndrome, splenic sequestration, or blood transfusion, and did not increase the malaria incidence[11].

A prospective study of hydroxyurea in African adults with SCA has not been reported. Obstacles to greater use include cost, limited access to basic medical care, and concern for adverse effects[12]. We examined fixed low-dose hydroxyurea prescribed during the course of normal clinical care of SCA adults in Nigeria. We wished to find a regimen that could be given safely in areas where frequent monitoring is not possible and yet be clinically beneficial. In the landmark adult US Multicenter Study of Hydroxyurea in Sickle Cell Anemia (MSH), pain crises decreased during the first three months of therapy, before dose escalation and maximal increase in hemoglobin F[1]. Hydroxyurea 10-15.9 mg/kg/day decreased pain episodes in children and adolescents in Oman[13], and 10 mg/kg/day decreased pain episodes in children and adults with SCA in India[14]. From these observations, a fixed dose of hydroxyurea 10 mg/kg/day was reasonable to investigate in a region where the safety profile in relationship to the resources and infectious exposures is not known.

The study was conducted at the University College Hospital of the University of Ibadan College of Medicine, Ibadan, Nigeria. The Joint Ethical Committee of the University of Ibadan/University College Hospital, which was the primary IRB for the clinical trial, reviewed and approved the study. Inclusion criteria were hemoglobin SS or Sβ⁰ thalassemia genotype, age ≥18 years, hemoglobin >45g/L, absolute neutrophils >1,500/μL, platelets >95,000/μL, serum creatinine <106.1μmol/L, and alanine transaminase less than two times upper limit of normal. Exclusion criteria were pregnancy or planning to become pregnant, breastfeeding, and positivity for human immunodeficiency virus, hepatitis B surface antigen or hepatitis C antibody. Written informed consent was obtained from each patient. Hemoglobin SS was determined by gel electrophoresis (pH 8.5), complete blood count by an automated counter, and hemoglobin F by HPLC.

Patients were assigned to receive hydroxyurea 500 mg/day orally for 24 weeks and then to crossover to an observation period of 24 weeks when they were off hydroxyurea. The onset of the observation period was separated from the end of the hydroxyurea period by at least four weeks. During both the hydroxyurea and observation periods, the patients came for a baseline visit and were instructed to return every four weeks for six additional visits for clinical evaluation. During the hydroxyurea period, patients were provided a four-week supply of hydroxyurea at baseline and a resupply every four weeks for five additional visits. In keeping with University College Hospital policy, patients were instructed to continue taking malaria prophylaxis with proguanil 200 mg orally daily throughout the study. Proguanil was supplied on a four-weekly basis.

The primary outcome variable was severe cytopenia, defined as neutrophils <500/μL, platelets <50,000/μL or reticulocyte count <95,000/μL in the presence of hemoglobin <90g/L. The prospective threshold for a clinically meaningful adverse result rate was defined as the occurrence of the primary outcome variable in ≥10% of the patients. This threshold was chosen empirically but we reasoned the study should be interrupted if such a high occurrence of severe cytopenia was observed. The secondary outcome variable was the development of an infection such as clinical malaria or tuberculosis. Additional outcomes included complete blood count and hemoglobin F. Study power was calculated prospectively based on the improvement in the white blood cell count in the MSH: a presumably beneficial decline in mean (SD) white blood cell count from 12,600 (3,400) to 9,900 (3,100) per μL[1]. A sample size of at least 20 patients was projected to yield a similar change at significance P=0.05 and test power 92%. We compared results obtained during hydroxyurea therapy and results obtained during the observation period with baseline results for each period with the Wilcoxon signed rank test, which does not assume normality of the variables. Missing values were imputed by carrying the last measured value forward.

Fifty-three adult SCA patients with a median (interquartile range) age of 24 (22-27) years were enrolled; 20 (37.8%) were females. The median daily dose of hydroxyurea according to body weight was 9.6 (8.8-10.3) mg/kg. Five patients did not return after receiving the first month’s supply of hydroxyurea, and the analysis of the results of administering fixed low-dose hydroxyurea are based on the 48 remaining patients. Fort-five (94%) of these patients continued on the study through 16 weeks and 33 (69%) through 24 weeks. The common reason for non-compliance with the full 24 weeks of therapy was fear of the effects of hydroxyurea on fertility. Thirty-eight of these patients also participated in the observation period.

One patient developed platelets <50,000/μL during hydroxyurea therapy, which returned to >150,000/μL with antimalarial treatment and continued hydroxyurea therapy. Another patient developed platelets <50,000/μL during the observation period. No patient developed neutrophils <500/μL. The lowest neutrophil count was 1,400/μL while receiving hydroxyurea and 2,200/μL during the observation period. We were unable to measure reticulocyte counts reliably in the laboratory that was used for this study. SCA patients in malaria-endemic areas may have splenomegaly rather than auto-splenectomy, and patients with splenomegaly may have low circulating platelets due to pooling in the spleen rather than true thrombocytopenia. Furthermore, thrombocytopenia is a frequent manifestation of malaria infection. Clinical malaria was unusual in both the hydroxyurea and observation periods, possibly related to the prophylaxis with proguanil. One patient developed recrudescent pulmonary tuberculosis while receiving hydroxyurea and one patient died of hyperhemolysis during the observation period. Disseminated nontuberculous mycobacterial infection and cryptosporidium infection have been described in sickle cell disease patients receiving hydroxyurea[15]. The incidence of active tuberculosis is high in Ibadan, Nigeria, but nevertheless patients should be monitored for active tuberculosis in future studies of hydroxyurea therapy in Nigeria.

At the end of 24 weeks of hydroxyurea therapy, there were median increases in the hemoglobin concentration of 5 g/L, hematocrit of 1.1% and MCV of 8.4 fL (P<0.001) and body weight of 1.3 kg (P=0.061) (Table 1). There were median decreases in the white blood cell count of 1,200/μL (P=0.013), in neutrophils of 900/μL (P=0.086) and in platelets of 56,000/μL (P=0.001). The median (IQ range) hemoglobin F was 4.0% (2.6-6.6%) at baseline and increased to 9.8% (6.9-13.0%) during therapy (P<0.001) (Supplemental Figure 1). One patient experienced an increase in hemoglobin F from 8.0% to 26.4%. During the observation period off hydroxyurea, there were trends to decreases rather than increases in hematocrit and MCV and hemoglobin F, and the other parameters did not change significantly (Table 1).

Table 1.

Response variables during 24 weeks of fixed low-dose hydroxyurea and during 24 weeks of observation off hydroxyurea in 48 sickle cell anemia patients in Ibadan, Nigeria. Results are given in median (interquartile range). P-value represents comparison with pre-hydroxyurea or observation baseline by the Wilcoxon signed rank test.

Variable	Pre-Hydroxyurea baseline (n=48)	Fixed low-dose hydroxyurea therapy				Observation baseline (4 weeks off hydroxyurea) (n=38)	Observation period of being off hydroxyurea
		Hydroxyurea 12 weeks (n=48)	P-value	Hydroxyurea 24 weeks (n=44)	P-value		Observation 12 weeks (n=27)	P-value	Observation 24 weeks (n=26)	P-value
Hemoglobin (g/L)	78 (72-87)	81 (75-90)	0.001	83 (76-92)	<0.001	80 (72-86)	82 (75-88)	0.900	76 (72-82)	0.500
Hematocrit (%)	25.1 (22.9-27.5)	25.8 (24.5-29.1)	<0.001	26.2 (24.4-29.4)	<0.001	25.9 (23.2-27.4)	26.0 (23.9-28.0)	0.200	24.4 (22.8-26.5)7	0.040
Mean corpuscular volume (fL)	87.1 (80.7-90.7)	94.5 (89.7-102.8)	<0.001	95.5 (90.5-102.9)	<0.001	94.2 (85.5-100.9)4	91.9 (86.3-97.2)	0.019	91.7 (87.3-96.3)7	0.002
White blood cells (1000/μL)	10.1 (8.7-11.8)	8.5 (7.2-9.8)	0.007	8.9 (7.3-10.4)	0.013	9.2 (7.7-10.8)	9.9 (8.2-11.8)	0.150	8.9 (8.1-11.7)	0.150
Neutrophils (1000/μL)	5.3 (4.1-6.5)	4.1 (3.6-5.2)1	0.025	4.4 (3.4-5.7)	0.086	4.4 (3.8-5.8)4	5.2 (3.8-6.9)	0.200	4.7 (4.1-5.5)	0.200
Platelets (1000/μL)	467 (363-554)	382 (312-492)	<0.001	411 (298-472)	0.001	386 (293-455)	394 (254-477)	0.200	401 (337-511)7	0.400
Body weight (kg)	50.8 (48.1-55.0)1	52.0 (48.5-55.7)2	0.004	52.1 (46.4-56.1)3	0.061	52.3 (47.4-57.0)5	50.2 (46.6-54.0)6	0.800	49.9 (47.0-55.5)	0.073

¹N=47;

²N=38;

³N=43;

⁴N=37;

⁵N=36;

⁶N=26;

⁷N=25

Limitations to our study are that it was observational rather than a double-blind placebo-controlled trial, that we were unable to determine if fixed low-dose hydroxyurea has an effect on vaso-occlusive pain episodes and other vaso-occlusive complications, and that only about three-fifths of the subjects complied with the planned course of hydroxyurea therapy due in part to patients’ fear of the effects of hydroxyurea on fertility. Nevertheless, our results suggest that fixed low-dose hydroxyurea for adults with SCA in Nigeria is feasible, is associated with a low rate of adverse outcomes and leads to improvements in blood counts, hemoglobin F and body weight. We propose that a large prospective multicenter study including urban and rural areas be organized to determine the effects of fixed low-dose hydroxyurea on vaso-occlusive episodes and other complications in adult sickle cell patients in sub-Saharan Africa[12].

Supplementary Material

Figure1

Supplementary Figure 1. Box plot of hemoglobin F (%) at baseline, during 24 weeks of fixed low-dose hydroxyurea therapy (HU), and during 24 weeks of observation (Observe) in Nigerian adult SCA patients for whom the measurement was made. Hemoglobin F significantly greater during hydroxyurea therapy compared to baseline (P=0.008) and to the observation period (P=0.006).