Acute painful crises of sickle cell disease (SCD) in Egyptian children: predictors of severity for preventive strategy.

Acute painful crises of sickle cell disease (SCD) in Egyptian children: predictors of severity for preventive strategy


Mohammad Al-Haggar † MD, Hala Al-Marsafawy, MD, Nabeel Abdel-Razek, MD Rizk Al-Baz 1 , PhD and Abdel-Hamid Mostafa 2 MD


Paediatrics Department, Faculty of Medicine, Mansoura University, 1 Mansoura University Children’s Hospital (Genetics laboratories), 2 Clinical Pathology Department, Faculty of Medicine, Al-Azhar University, Egypt


This work had been conducted in Mansoura University Children’s Hospital, Dakahlia province, Middle delta, Egypt.


† Correspondence: Mohammad Al-Haggar, MD, Professor of Pediatrics and Genetics, Paediatrics Department, Faculty of Medicine, Mansoura University, Mansoura, Egypt, e-mail: mhajjar2000@yahoo.co.uk


ABSTRACT
Objective; To predict sickle cell infants (SCD) who are prone to develop severe painful crises in future.
Subjects and methods; In mixed hospital-community-based population (76 cases), demographic data, diagnostic parameters of SCD and basal blood counts were correlated to two indices of SCD severity; pain rate (average days of painful episodes per year of follow up) and serious life threatening complications like hyper-hemolytic crises. Data were analyzed blind to these indices; t-test, ANOVA and Pearson correlation were used to determine association to pain rate. Discriminant analysis was the test used for prediction of SCD severity.
Results; Pain rate was significantly high in HB SS patients especially in those with early onset of dactylitis. There were statistically significant negative correlations of pain rate to basal HB level, Hct%, HB F% and arterial oxygen saturation (P < 0.01 for all correlations). The top 3 predictors for SCD severity were genotype, basal HB level and early dactylitis in a descending order.
Conclusion; Severe forms of SCD could be predicted in early infancy with 100% accuracy using the basal diagnostic parameters of the disease; those infants should be closely monitored with special attention to their ventilation status even before the development of dactylitis.
(Keywords; sickle disease, painful crises, prediction)


INTRODUCTION

Acute painful crisis is the most common clinical presentation of children with sickle cell disease (SCD); usually secondary to hypoxia, dehydration and infection. Many studies had been published giving the upper hand to some prognostic parameters over others; some considered early-onset of dactylitis, leucocytosis and low haemoglobin to be of utmost importance [1], others considered the disease genotype and HB F to be more significant [2 - 4]. Nocturnal arterial oxygen desaturation was found in a significant proportion of SCD patients and this explains the night exacerbations in these cases [5 - 8]. Up to our knowledge no available references tried to predict the disease severity from these prognostic parameters in a mathematical fashion.

This work aims to predict infants and young children with SCD who are likely to have severe painful crises using a multivariate analysis. Attempting to define such children will permit accurate prognostication and proper therapy plan in order to minimize the morbidity and mortality associated with SCD.


SUBJECTS AND METHODS

Study Design
This study was conducted on hospital and community based subjects after obtaining a formal written consent from all parents. The hospital-based cases were the newly diagnosed infants and young children with SCD at Paediatric Genetics Unit of Mansoura University Children’s Hospital (MUCH), they constituted 42 cases (23 males and 19 females) with age range 8-34 months (25.4 ± 9.2). The community-based subjects were the old cases with SCD followed at Paediatric Haematology Clinic of MUCH; they were 34 cases (20 males and 14 females) with age range 4-13 years (8.8 ± 3.6).

Data collection
The hospital-based subjects were followed for a period of about 2 years starting at June, 2001 to July, 2003 or until terminal event (start of regular transfusions or analgesic medication or even child death by a relevant cause). We stressed on relevant symptoms like painful crises especially those lasting for > 2 days, increasing pallor, deepening jaundice, number and reasons for hospital admissions. Demographic data, sickling test, HB electrophoresis as well as basal blood counts were the main profile for each child. Enzyme assay was checked whenever possible to rule out the possibility of associated G6PD deficiency. Data of community-based SCD children were retrieved from the files.

Laboratory Diagnosis
Sickling test is a qualitative solubility test for screening HB-S; HB released from lysed RBCs by saponin, is reduced by Dithionate buffer. Reduced HB S is characterized by its very low solubility thus forming nematic liquid crystals (tactoids) causing a turbid system, there is some non-S HBs causing positive sickling [9]. HB electrophoresis depends upon the differential migration of the charged HB variants in an electric field. The migration patterns differ according to pH, temperature, voltage, ionic strength of buffers as well as nature of supporting medium; at alkaline pH (8.6 – 8.8) in a cellulose acetate medium HB SS, Sβ and SA can be differentiated, however to differentiate HB SS from SC, we used acidic buffer (pH 6.2) in agar gel [10]. G6PD is assayed using a semi-quantitative method; enzyme released from lysed RBCs catalyzes conversion of NADP to NADPH which in presence of phenazine methosulphate will reduce the dye colour (dichlorophenol indophenol) to colourless form, the reaction mixture is proportional to RBCs content of G6PD [11].

Patient Categorization
Two dependant variables were concluded from patient’s profile; the 1st one was pain rate (average number of days of hospital stay due to painful episodes or days of extreme relevant illness at home from patient’s own calendar per year of follow up) and the 2nd variable if there was any serious life-threatening complication like protracted haemolysis or hyper-haemolytic crisis in a recently transfused child. Severity was deducted from these 2 variables; cases showing frequent painful episodes (pain rate = > 5 days/year) or any serious events were labelled as severe (15 cases) and those with infrequent hospital attendance (pain rate < 2 days/year) were considered mild (12 cases).

Children presented with frequent nocturnal painful episodes (42 out of 76) were admitted for at least 2 nights; O2 saturation (O2 SAT) was estimated 4 times each night with 2 hours apart, then mean O2 SAT was deducted; cases suggestive of obstructive sleep apnoea were transferred to ENT specialists for further evaluation.

Statistical Analysis
Data of all subjects were merged together regardless to the basis of study to minimize the error of retrospective nature of community-based portion. Data were analyzed blind to the deducted SCD severity using SPSS version 10.0 [12]. Quantitative variables showed preserved normality (Kolmogrov Smirnov test). T-test, one way ANOVA, Pearson’s correlation and Chi square test were done to test for associations to pain rate. At two cut off values of O2 SAT (80% and 90%), pain rate was assessed in the different subgroups. Discriminant analysis was the multivariate test using the different prognostic parameters for prediction of severity in a stepwise approach. Accuracy of prediction was assessed by Wilks’ lambda (X2 for test function was 46.14, with P < 0.001) as well as percentage of correctly classified cases.
RESULTS
Patients with HB SS genotypes (71%) suffered from long painful episodes especially if they had a history of early dactylitis before the age of 2 years (15.7%); the mean pain rate in HB SS patients is the highest with very close and overlapping values of the two other genotypes (table: 1 & figure: 1). Twenty SCD children were presented with dactylitis before age of 2 years (8 of them were SS genotype; X2 for association of early dactylitis to SCD genotype was 9.11 with P = 0.01; not presented in tables).
Pain rate was significantly negative correlated to basal HB, Hct value, HB F%. Also there was a strong negative correlation between pain rate and O2 SAT; note the down slope of the best fit line of scatter diagram with severe cases (solid triangles) being clustered opposite 80% and mild cases (asterisks) being clustered opposite 90% O2 SAT (table: 2 - figure: 2). To determine level of O2 desaturation below which painful episode became severely protracted, patients were categorized according to two cut off values of O2 SAT, mean difference and statistical significance were more evident at 80% O2 SAT (table: 3).

Prediction of SCD severity was determined by discriminant analysis that yielded a predictive equation; formulated from the basal parameters in table 4-a:
Predictive equation = 2.75 X genotype (codes; 1 for HB SS, 2 for HB SC and 3 for HB Sβ) – 0.42 X Early onset of dactylitis (codes; 1 if yes and 0 if no) + 0.74 X Basal HB – 10.95 = Score value.
If score value ≥ 2.88, it will be mild SCD variant, however a score value ≤ - 2.30 will be a severe case. Accuracy of Predictive equation could be assessed by its application to the 27 cases of already known clinical severity (table 4-b). Comparing the original grouping to the predicted grouping, all cases were found correctly classified i.e. no false positive or negative rates, so predictive equation could be considered 100% accurate.

DISCUSSION
Sickle cell disease (SCD) is an inherited autosomal recessive disorder of β-globin chain with high incidence in Middle East due to the high rate of consanguineous marriage. Sickle HB is considered to be the second most common mutation of β globin (after β thalassaemia) accounting for about 12.2% of all mutations of β globin clusters [13 & 14]. Acute painful crisis is the most frequent and intractable problem encountered in this disease [15].
Despite the fact that all subjects with SCD have the same single base pair mutation, the severity of clinical and haematological manifestations is extremely variable from race to race and even within the same race. So, prediction of severity is a hot topic in SCD. Early-onset of dactylitis (defined as pain and tenderness in hands and feet before age of 1 year), leucocytosis (in the absence of infection) and low basal haemoglobin (< 7 gm/dl) were suggested to be the most important prognostic parameters in SCD [1]. Disease genotype and HB F level were considered by other authors to be more important as regards disease severity [2 - 4]; on the other hand some authors lessen the role of HB F level as a predictive parameter of disease severity [16].
In our hospital and community-based study done on Egyptian SCD children, we used multivariate analysis to predict those cases who are liable to develop severe disease in future by analyzing the demographic data of SCD children of known clinical severity. No gender difference in pain severity (P = 0.91) was found in our study opposite to the finding reported on Saudi SCD adults being more severe among males [17]. Analysis of difference between the 3 genotypes of SCD revealed a high statistical significance with the longest pain rate among SS patients (P < 0.001) but with no significant difference between the other 2 genotypes (P 0.09).
In our study, high HB F level was associated with low pain rate irrespective to the disease genotype (r = - 0.29 with P 0.01). Some reports attributed the mild nature of SCD to high HB F; however others found no relation between basal HB F and pain rate [3, 4, 8 & 16]. The mild SCD in most Kuwaitis children was attributed to the elevated HB F although authors described a high incidence of avascular necrosis of femoral head among their population (26.7%) [3]. The high prevalence of Arab-Indian genotype in Iran which is associated with high HB F (30%) accounts for the mild clinical presentation of SCD in Iran [4]. On the other hand, in Lebanese high HB F was not associated with decreased severity of SCD symptoms [16].
In SCD patients having nocturnal painful episodes, O2 SAT was negatively correlated to pain rate denoting that hypoxemia which is possibly due to an underlying upper airway obstruction is a stimulus for bone marrow to correct anaemia. However, in adults there was a paradox of high HB & Hct and low O2 SAT due to the compensated chronic hypoxemia by acclimatization [1].
Prediction of severity was done by discriminant analysis using the relevant parameters as independent variables; the best predictors as determined from their coefficients were genotype, basal HB level and early-onset of dactylitis in a descending order (coefficients were 2.75, 0.737 and -0.420 respectively) with no marked impact for the other parameters. Accuracy of prediction was proved by Wilks Lambda as well as percentage of correctly classified cases (100%). The exclusion of HB F from the top 3 predictors of severity could be explained by the fact that our young cases had a very high HB F level which will be rapidly changing so that statistical analyses would be unlikely to find it as a strong predictor. Further prospective study for these cases should be warranted to check for the importance of HB F at older ages to solve the issue of controversies regarding this parameter [3, 8 & 16].
We can conclude that with the use of this predictive equation, SCD infants beyond age of 6 month could be scored as a high or low risk using the top 3 prognostic parameters incriminated in discriminant analysis. High risk SCD infants should be a target for powerful therapeutic plan that include the availability of newly introduced pain controllers for home use [18] and a community-based educational treatment to increase spiritual well-being and self-efficacy to cope with pain [19]. With the help of these measures altogether SCD morbidity and mortality would be mitigated.
Acknowlegement
We acknowledge all staff members of Genetics unit laboratories, MUCH, Faculty of Medicine, Mansoura University, for blood sampling and laboratory analysis. Particular thanks are due to staff of paediatric haematology unit in MUCH for the efforts in retrieving personal and clinical data of old cases, and recruiting them for further evaluation.

REFERENCES
1. Miller ST, Sleeper LA, Pegelow CH, et al: Prediction of adverse outcomes in children with sickle cell disease. N Engl J Med 2000; 342:83-89.
2. Kaul DK, Fabry ME, Nagel RL: The pathophysiology of vascular obstruction in the sickle syndromes. Blood Rev 1996; 10:29-44.
3. Marouf R, Gupta R, Haider MZ, Al-Wazzan H, Adekile AD: Avascular necrosis of the femoral head in adult Kuwaiti sickle cell disease patients. Acta Hematol 2003; 110(1): 11-15.
4. Rahimi Z, Karimi M, Haghshenass M, Merat A: Beta-globin gene cluster haplotypes in sickle cell patients from southwest Iran. Am J Hematol 2003; 74(3):156-160.
5. Homi J, Levee L, Higgs D, Thomas P, Serjeant G: Pulse oximetry in a cohort study of sickle cell disease. Clin Lab Haematol 1997; 19:17-22.
6. Needleman JP, Franco ME, Varlotta L et al: Mechanisms of nocturnal oxyhemoglobin desaturation in children and adolescents with sickle cell disease. Pediatr Pulmonol 1999; 28:418-422.
7. Wilkey O, Evans JPM, Telfer PT, Kirkham FJ: Predictors of nocturnal hypoxaemia in children with sickle cell disease [abstract]. Arch Dis Child 2002; 86 (suppl 1):A12.
8. Haregrave DR, Wade A, Evans JPM, Hewes DKM, Kirkham, FJ: Nocturnal oxygen saturation and painful sickle cell crises in children. Blood 2003; 101(3): 846-848.
9. Fairbanks VK, Klee GG: In textbook of Clinical Chemistry, Teitz NW (ed) WB Saunders co., Philadelphia, 1986; 1540-1541.
10. Kim HC: Separation of Hemoglobins in Hematology Textbook; 4th edition, Williams WJ and Lecihman MA (eds), McChill, NewYork 1990; 1711-1714.
11. Echier G: Determination of G-6-PD. Am J Med Technol 1983; 49:259.
12. SPSS: Statistical Package for Social Science, standard version 10.0.1; 1999; SPSS Inc., Chicago, IL.
13. Al-Arrayed S, Hafadh N, Amin S, Al-Mukhareq H, Sanad H: Student screening for inherited blood disorders in Bahrain. East Mediterr Health J 2003; 9(3):344-352.
14. Keser I, Sanlioglu AD, Manguoglu E, et al : Molecular analysis of beta-thalassemia and sickle cell anemia in Antalya. Acta Haematol 2004; 111(4):205-210.
15. Platt OS, Thorington BD, Brambilla DJ, et al: Pain in sickle cell disease. Rates and risk factors. N Engl J Med 1991; 325:11-16.
16. Inati A, Taher A, Bou Alawi W, et al: Beta-globin gene cluster haplotypes and HB F levels are not the only modulators of sickle cell disease in Lebanon. Eur J Haematol 2003; 70(2):79-83.
17. Udezue E, Girshab AM: Differences between males and females in adults sickle cell pain crises in eastern Saudi Arabia. Ann Saudi Med 2004; 24(3): 179-182.
18. Graves PA, Kedar A, Koshy M, et al: RheothRx (Poloxamer 188) Injection for the Acute Painful Episode of Sickle Cell Disease: A Pilot Study. Blood 1997; 90(5):2041-2046.
19. Cooper-Effa M, Blount W, Kaslow N, Rothenberg R, Eckman J: Role of spirituality in patients with sickle cell disease. J Am Board Fam Pract 2001; 14(2): 116-122.