Impact of sickle cell trait on the thrombotic risk associated with non-O blood groups in northern Nigeria

Abstract

Background

The non-O blood group is an established risk factor for deep vein thrombosis (DVT), while controversy surrounds the role of sickle cell trait (SCT) as a risk factor for DVT. We hypothesised that if SCT is a risk factor for DVT, individuals with non-O blood groups and SCT (Hb AS) would have a higher risk of DVT than their counterparts with non-O blood groups and normal haemoglobin phenotype (Hb AA).

Materials and methods

We retrospectively analysed the prevalence of SCT and non-O blood groups among 148 DVT patients with control subjects in order to determine the role of SCT as a risk factor for DVT and its impact on the risk of DVT among patients with non-O blood groups.

Results

In comparison with control subjects, DVT patients had significantly higher prevalences of SCT (35.1% vs 27.7%, p=0.04) and non-O blood groups (68.9% vs 45.9%, p=0.02). The odds ratios for DVT due to SCT, non-O blood groups with normal Hb phenotype (Hb AA) and non-O blood groups with SCT (Hb AS) were 1.3, 2.4 and 3.5, respectively.

Discussion

These results suggest that SCT by itself is a weak risk factor for DVT but it has the potential of escalating the DVT risk among patients with non-O blood groups. The combined effects of elevated clotting factors (non-O group effect) and increased clotting factor activation (SCT effect) were responsible for the escalated DVT risk among patients with co-inheritance of non-O blood groups and SCT. Co-inheritance of SCT and non-O blood group is, therefore, an important mixed risk factor for DVT. This should be taken into account when assessing DVT risk profiles of patients in Africa and other parts of the world where the SCT is prevalent.

Introduction

The ABO blood groups have a significant effect on the incidence of venous thrombosis with individuals having non-O blood groups being at higher risk of thromboembolism due to higher levels of von Willebrand factor (vWF) and factor VIII (FVIII)^1,2. The rate of proteolytic clearance of vWF by ADAMTS-13 is relatively lower in the plasma from subjects with non-O blood groups, hence the half-life of vWF is longer in non-O plasma than in group O plasma^3,4. Consequently, vWF levels are 25–30% higher in non-O plasma than in group O plasma⁵. The high levels of vWF in non-O plasma are always associated with concomitant elevations of FVIII levels due to the physiological role of vWF as the carrier of FVIII and its protector from the proteolytic effect of ADAMTS-13⁶. Lower baseline levels of vWF and FVIII were thought to be causally related to high risks of idiopathic epistaxis, postoperative haemorrhage and bleeding during oral anticoagulation seen among patients with blood group O^7–9. Conversely, higher levels of vWF and FVIII seen in subjects with non-O blood groups were strongly correlated with increased risks of venous thrombosis, a situation that made the non-O blood groups be rated as the commonest genetic risk factors for venous thromboembolism⁶.

People with sickle cell trait (SCT) are heterozygous for the sickle β-globin gene and their red cells have the Hb AS genotype expressing both Hb S (20–40%) and Hb A (60–80%)^10,11. The abundance of Hb A prevents sickling under physiological conditions, hence the red cell life span is normal in SCT and affected individuals have a normal life expectancy^11,12. The frequency of SCT is up to 25–30% in tropical Africa, including Nigeria, because SCT protects against malaria^13,14. Nonetheless, SCT might confer its carriers the undesirable risk of venous thromboembolism. An earlier study reported a case of SCT with recurrent deep vein thrombosis (DVT) that could not be associated with any other known risk factor for DVT¹⁵. However, a larger study had shown that SCT was associated with an increased risk of venous thromboembolism, the risk being stronger for pulmonary embolism than for DVT¹⁶. Moreover, a study suggested that hormonal contraceptives might pose a greater risk of venous thromboembolism in black women with SCT than in black women without SCT¹⁷. Yet another study revealed that SCT increased the risk of pulmonary embolism but not the risk of DVT¹⁸. The high risk of venous thromboembolism among people with SCT was thought to be related to increased clotting factor activation resulting from subclinical red cell sickling^16,19.

In contradistinction, one study failed to confirm the existence of increased risk of venous thromboembolism in patients with SCT even under challenging prothrombotic conditions such as pregnancy and the puerperium²⁰. These conflicting reports on the risk of venous thromboembolism in SCT create an area of continuing controversy that merits further research. We, therefore, investigated the role of SCT as a risk factor for DVT and evaluated the impact of SCT on the thrombotic risk associated with non-O blood group, which is a well-established risk factor for DVT. We hypothesized that if SCT is a risk factor for DVT, individuals with non-O blood groups and SCT (Hb AS) would have a higher risk of DVT than their counterparts with non-O blood groups and normal haemoglobin phenotype (Hb AA). To the best of our knowledge, the relationship between SCT and non-O blood groups with regards to the risk of DVT has not been previously studied. Hence, we also conducted a retrospective analysis of the prevalence of SCT and non-O blood groups among patients with DVT in order to determine the possible role of SCT as a risk factor for DVT and its impact on the risk of DVT among patients with non-O blood groups in northern Nigeria.

Materials and methods

This is a retrospective case-control study of patients with DVT managed, between 1996 and 2010, in the Haematology Departments of four Nigerian tertiary health institutions: the University of Maiduguri Teaching Hospital, Maiduguri, northeast Nigeria (1996–2007), Federal Medical Centre Birnin Kudu, northwest Nigeria (2004–2005), Murtala Muhammad Specialist Hospital, Kano, northwest Nigeria (2008–2010) and Aminu Kano Teaching Hospital, Kano, northwest Nigeria (2008–2010).

Inclusion criteria

This study included all DVT patients with characteristic clinical features including lower limb oedema, warmth and/or tenderness associated with ultrasonographically demonstrated intravenous thrombi with or without associated pulmonary embolism.

Exclusion criteria

Patients with clinical and/or radiological features of pulmonary embolism without ultrasonographically detectable DVT were excluded from this study. DVT patients with sickle cell disease, Hb C trait, glucose-6-phospahte dehydrogenase deficiency and/or α-thalassaemia were also excluded. Due to limitations of laboratory facilities, DVT patients in Nigeria are not routinely screened for inherited thrombophilia. Hence, patients with recurrent DVT or a family history of DVT were excluded in order to eliminate the likelihood of inherited thrombophilia in the sample studied.

Ethics, procedures, data collection and analysis

All patients were recruited with informed consent and all procedures were conducted with the approval of local institutional ethics committees and in accordance with ethical standards for human experimentation as enshrined in the Helsinki Declaration of 2013 with its amendments. Demographic and clinical laboratory data including age, sex, clinical risk factors and anatomical distribution of DVT, haemoglobin phenotype and ABO blood groups as determined at the time of diagnosis of DVT were retrieved from the clinical records of the patients. An equal number of consenting age- and sex-matched control subjects were recruited from healthy local populations and their haemoglobin phenotypes and ABO blood groups were determined.

Determination of haemoglobin phenotypes

Hb phenotypes of DVT patients and control subjects were determined by haemoglobin electrophoresis at a pH of 8.6 on cellulose acetate paper, sickling test and haemoglobin quantitation²¹. Based on the electrophoretic patterns, the patients’ or control subjects’ phenotypes were categorised as normal (Hb AA) or SCT (Hb AS)²¹.

Determination of ABO blood groups

ABO blood groups of DVT patients and control subjects were determined manually by using monoclonal anti-A and anti-B against the patients’ red cells suspended in saline in tubes at room temperature and read for agglutination after 15 minutes of incubation in accordance with standard procedures and manufacturer’s guidelines²². Based on the pattern of agglutination, patients/subjects were categorised as having blood group O, A, B or AB²².

Statistical analysis

The data accrued from the four centres were collated. Values of evaluated parameters were compared between subject categories using the X² test and p-values of less than 0.05 were considered statistically significant. The risk of DVT was determined by the odds ratio (OR) for three risk categories: SCT (Hb AS), non-O blood groups with normal Hb phenotype (Hb AA) and non-O blood groups with SCT (Hb AS) as follows:

• - the OR of DVT for SCT = (n. of patients of all ABO groups with SCT ÷ n. of patients of all ABO groups with Hb AA)/(n. of control subjects of all ABO groups with SCT ÷ n. of controls of all ABO groups with Hb AA);

• - the OR of DVT for non-O blood groups with normal Hb phenotype (Hb AA) = (n. of patients with non-O blood groups and Hb AA ÷ n. of patients with blood group O and Hb AA)/(n. of controls with non-O blood groups and Hb AA ÷ n. of controls with blood group O and Hb AA);

• - the OR of DVT for non-O blood group with SCT (Hb AS) = (n. of patients with non-O blood groups and SCT ÷ n. of patients with blood group O and Hb AA)/ (n. of controls with non-O blood groups and SCT ÷ n. of controls with blood group O and Hb AA).

The OR with 95% confidence intervals (CI) were determined by logistic regression analysis with adjustments for age and sex. OR values were considered statistically significant if the lower limit of the 95% CI was greater than 1.0 with a p-value of less than 0.05. Statistical analyses were performed using computer software SPSS version 15.0 (SPSSInc., Chicago, IL, USA).

Results

A total of 148 patients with DVT and an equal number of age- and sex-matched control subjects were studied. The prevalence of clinical risk factors and the anatomical distribution of DVT among the group of patients is shown in Table I. Pregnancy, surgical operations, trauma, hormonal therapy, malignancies, diabetes and hypertension were the commonest clinical risk factors for DVT. The left lower limb was the predominantly affected limb with a relative frequency of DVT of 64.2%. Proximal DVT was the predominant form of DVT with a relative frequency of 67.6%. The distribution of age, sex, haemoglobin phenotypes and ABO blood groups in the DVT patients and control subjects is shown in Table II. In comparison with normal control subjects, DVT patients had significantly higher prevalences of SCT (35.1% vs 27.7%, p=0.04) and non-O blood groups (68.9% vs 45.9%, p=0.02). After adjustments for age and sex, the OR for the risk of DVT were 1.3 (95% CI: 1.1–1.9, p=0.03) for SCT, 2.4 (95% CI: 1.8–3.1, p=0.01) for non-O blood groups with normal Hb phenotype (Hb AA), and 3.5 (95% CI: 2.3–4.2, p=0.005) for non-O blood groups with SCT (Hb AS), as shown in Table III.

Table I.

Prevalence of clinical risk factors and anatomical distribution of DVT.

Risk factors*	N. of patients (%)
Pregnancy/puerperium	34 (23)
Surgical operation	31 (20.9)
Accident/trauma	24 (16.2)
Hormonal therapy/contraceptives	18 (12.2)
Malignancy	13 (8.8)
Diabetes/hypertension	12 (8.1)
Systemic lupus erythematosis/rheumatoid arthritis	9 (6.1)
Paraplegia/hemiparesis	9 (6.1)
Obesity (body mass index >30 kg /m2)	9 (6.1)
No apparent risk factor (unprovoked)	3 (2)
Limbs affected
Right lower limb	53 (35.8)
Left lower limb	95 (64.2)
Veins affected
Proximal veins (iliac, femoral, popliteal)	100 (67.6)
Distal veins (calf)	48 (32.4)

^*Many patients were affected by more than one risk factor.

DVT: deep vein thombosis.

Table II.

Distribution of age, sex, haemoglobin phenotypes and ABO blood groups among DVT patients and control subjects.

Parameters	DVT patients (n=148)	Control subjects (n=148)
Age median [range]	54 [33–84]	52 [30–82]
Sex ratio (Male/Female)	0.7	0.8
Normal Hb phenotype (Hb AA)	96 (64.9)*	107 (72.3)*
Sickle cell trait (Hb AS)	52 (35.1)*	41 (27.7)*
Blood group O	46 (31.1)*	80 (54.1)*
Non-O blood groups (A+B+AB)	102 (68.9)*	68 (45.9)*

^*Differences in corresponding values between patients and controls are statistically significant (p<0.05).

DVT: deep vein thombosis; Hb: haemoglobin.

Table III.

Odds ratio distribution among risk categories showing interactions between SCT and non-O blood groups with respect to risk of DVT.

Risk categories	DVT patients (n=148)	Control subjects (n=148)	OR [CI 95%]
All ABO groups with SCT (Hb AS)	52	41	1.3 [1.1–1.9]
Non-O blood groups with normal Hb phenotype (Hb AA)	66	49	2.4 [1.8–3.1]
Non-O blood groups with SCT (Hb AS)	36	19	3.5 [2.3–4.2]

SCT: sickle cell trait; DVT: deep vein thombosis; OR: odds ratio; CI: confidence interval; Hb: haemoglobin.

Discussion

Apart from a few apparently unprovoked cases, the overwhelming majority of cases in this study were associated with several clinical risk factors that were identified as being aetiologically associated with DVT in the study. Nonetheless, all of the identified risk factors ultimately operate within the pathophysiological triad of hypercoagulability, stasis and/or vascular endothelial injury, all of which predispose to thrombosis as originally described by Virchow²³.

The frequency of SCT among the normal control subjects was 27.7%, which is consistent with the sickle cell gene frequency in Nigeria where SCT occurs in 25–30% of the population²⁴. In comparison to the normal control subjects, patients with DVT had a higher frequency of SCT with a modest but significant OR of 1.3, which suggested that SCT was a risk factor for DVT. This finding is consistent with previous studies that associated SCT with an increased risk of DVT^15,16. However, our finding is at variance with other studies in which SCT was not found to be associated with an increased risk of DVT^17,20. The aetiological basis for the elevated risk of DVT in patients with SCT may be multi-factorial with at least three possible contributory factors. Firstly, as mentioned earlier, SCT has been shown to be associated with increased clotting factor activation and hypercoagulability^16,19. Secondly, previous studies showed that patients with SCT have greater red cell rigidity and blood viscosity in comparison to people with a normal Hb phenotype^25,26. Thirdly, SCT is commonly associated with renal papillary necrosis and hyposthenuria, which would raise plasma osmolality and further increase blood viscosity^11,27. The increased risk of DVT in patients with SCT may, therefore, reflect the combined effects of increased clotting factor activation and blood hyperviscosity leading to hypercoagulability and sluggish blood flow, both of which would predispose to DVT as elements of Virchow’s triad²³. Nonetheless, the finding of an OR of only 1.3 suggests that SCT is a weak risk factor for DVT in this study, a finding that was consistent with a previous report¹⁶.

The relative proportions of O and non-O blood groups among normal control subjects mirrored the ABO blood group frequencies in the general population in northern Nigeria²⁸. In contrast, patients with DVT showed a distorted distribution of ABO blood groups with a significant excess of non-O blood groups, reaffirming the well-known role of the non-O blood group as a risk factor for DVT⁶. The OR for the risk of DVT among our patients who had non-O blood groups with a normal Hb phenotype (Hb AA) was 2.4. However, patients who had non-O blood groups with SCT (Hb AS) had significantly higher OR of 3.5, suggesting a higher risk for DVT. These findings imply that co-inheritance of SCT in patients with non-O blood groups aggravated the risk of DVT in such patients. From a pathophysiological perspective, it can be deduced that SCT can interactively amplify the thrombotic risk associated with non-O blood groups. This is because on the one hand, non-O blood group is associated with elevated levels of clotting factors (vWF and FVIII) while on the other hand, SCT is associated with increased activation of clotting factors. We infer that the combined effects of elevated levels of clotting factors (non-O blood group effect) on a background of increased activation of clotting factors (SCT effect) were responsible for the escalated risk of DVT among patients with non-O blood group who co-inherited SCT^6,19. Our findings suggest that while SCT by itself appears to be a relatively weak risk factor for DVT, it nonetheless has the potential to increase the DVT risk associated with non-O blood groups significantly.

Although every effort was made to ensure data correctness, we acknowledge that this study has important limitations, which include small sample size within the context of a retrospective data appraisal over a long period of recruitment. Nonetheless, the results of this study have potential clinical implications as they highlight the possible amplification of the thrombotic risk of non-O blood groups by SCT. Co-inheritance of SCT and non-O blood group should, therefore, be considered as an important mixed risk factor for DVT, which should be taken into account when evaluating DVT risk with a view to offering prophylactic anticoagulation to patients subjected to surgery, trauma or other prothrombotic challenges in Africa and other parts of the world where the SCT is prevalent.

Conclusions

This study suggests that both SCT and non-O blood groups are independent risk factors for DVT. Although SCT alone appears to be a weak risk factor for DVT, co-inheritance of SCT in patients with non-O blood groups appears to aggravate the risk of DVT in such patients. Co-inheritance of SCT and non-O blood groups should, therefore, be considered as an important mixed factor in DVT risk assessment for Africans and other ethnic populations in which the SCT is prevalent. Although the results of this study validated our hypothesis, given the small sample size and preliminary nature of the study, further, larger, prospective studies are needed to determine precisely the impact of SCT on the thrombotic risk associated with non-O blood groups and the implications for clinical practice.