Intravenous immunoglobulin-related thromboembolic events – an accusation that proves the opposite

Since the first reports of intravenous immunoglobulin (IVIg)-associated thromboembolic events (TE) in 1986 1, a further 443 TE that occurred during or post-IVIg treatment have been reported. TE included stroke, myocardial infarction, pulmonary thromboembolism and deep vein thrombosis. The risk of these events per patient is estimated to be between 0·6 and 4·5% 2.

A temporal relationship was observed between IVIg infusion and the occurrence of TE (during infusion to up to 42 days post-infusion). IVIg supposedly increases blood viscosity and reduces blood flow which, in turn, increases the risk of acute coronary events, particularly in patients with underlying risk factors 3. Based on this hypothesis, each IVIg infusion is a potential trigger for a TE and is expected to increase the risk of TE. Therefore, the prevalence of TE per infusion would be a more informative measure to determine whether IVIg infusion increases TE. The estimated global IVIg consumption from 1986 to the present is 1332 metric tonnes. Assuming a body weight of 73 kg per patient and a dose of 400 mg/kg per infusion, approximately 45·6 million single-dose infusions have been administered to date. Based on these estimations, the expected incidence of TE is one in 103 000 infusions.

The previously mentioned temporal relationship is based on the observation of TE up to 42 days post-infusion. However, when the half-lives of coagulation factors are considered, a 24 h post-infusion cut-off point may be more appropriate 4. Of the 443 cases of TE reported following IVIg infusion, 269 patients had TE within 24 h. Using these values, overall prevalence decreases to one event in 302 000 infusions. Of the 269 patients who experienced TE 24 h post-infusion, underlying risk factors for TE were present in 80% of the patients. The observed TE post-infusion may be due to natural progression of the underlying condition. This confounds the reliability of the results, as there is no matched group that can be used for comparison with these studies. Only two case reports describe an increase in blood viscosity post-IVIg infusion, but fail to establish cause-and-effect relationship. Furthermore, no increase in blood viscosity was observed in a case-series of seven patients 5.

These factors suggest that the mechanism leading to TE remains hypothetical, and the consensus that IVIg infusion causes an increased risk of TE may be unwarranted, as supporting evidence does not exist.

We compared the incidence of stroke among 1127 primary immunodeficiency (PID) patients on IVIg from the USIDNET registry with Centers for Disease Control and Prevention (CDC) data on frequency of stroke in the general population (GP) of the United States 6. CDC data were obtained using the Behavioral Risk Factor Surveillance System (BRFSS), a random telephone survey of the GP (aged ≥18 years) who reported stroke occurrence. The same demographic characteristics as in CDC BRFSS were used to stratify PID patients: age groups (18–44, 45–65 and >65 years), sex (male, female) and race/ethnicity (white, black, Hispanic, Asian or Native Hawaiian/other Pacific Islander and American Indian/Alaska Native).

The overall prevalence of stroke in PID patients was approximately four times lower than in the GP (0·62 versus 2·6%). In both the 18–44 and 45–65-year age groups, stroke was approximately two times less frequent in PID patients than in the GP. There were no reports of stroke in PID patients who were >65 years, whereas the incidence of stroke in the corresponding segment of the GP was 8·3% (Table 1).

Table 1.

Stroke in intravenous immunoglobulin (IVIg) recipients versus the general population

Age group (years)	No. of patients	No. of stroke cases	Prevalence of stroke in PID patients (%)	Prevalence of stroke in the general population (%)
18–44	671*	2*	0·29	0·7
45–64	321*	5*	1·55	2·9†
≥65	135*	0*	0·00	8·3†
Overall	1,127*	7*	0·62	2·6†

^*USIDNET registry;

^†CDC data on the frequency of stroke. PID = primary immunodeficiency.

Our data suggest that IVIg may have protective effects against stroke, particularly in elderly patients (>65 years), where stroke prevalence in the GP correlates with increasing age. The lower incidence of stroke observed in the older PID patients compared to the younger groups may be due to the increased duration on IVIg therapy. Clinical trials are needed to confirm the neuroprotective effect of IVIg for stroke in PID patients.

There is also a wide base of evidence that IVIg can be used to prevent TE in specific conditions. Atherosclerosis is a major risk factor associated with major adverse cardiac events (MACE). The effect of IVIg in the prevention and reduction of atherosclerosis has been demonstrated in a low-density lipoprotein (LDL)-receptor-deficient mouse model. Accelerated atherosclerosis was induced in this mouse model, characterized by plaques containing increased T cells and macrophages. IVIg therapy was found to reduce macrophage and CD4⁺ T cell accumulation in the atherosclerotic plaque 7. The results from another study found that, in mice with diet-induced atherosclerosis, arterial fatty lesions were reduced by 50%, decreasing the risk of MACE as a result of atherosclerosis 8.

Anti-phospholipid syndrome (APS) causes thrombosis in arteries and veins as well as pregnancy-related complications. A review of the available literature shows that in six animal studies, IVIg was associated with decreased mortality, inhibition of thrombosis, reduced fetal resorption, improved pregnancy outcomes and amelioration of endothelial cell inflammatory and thrombogenic phenotypes. None of these studies reported drug-related TE. Similarly, studies in humans with APS have shown no association between IVIg and TE. IVIg was used in 20 studies involving 252 patients and more than 3000 individual infusions. Compared to conventional therapy alone, IVIg was effective in preventing recurrent thrombosis, increasing live birth rates and reducing antepartum complications such as pre-eclampsia, fetal growth restriction and fetal distress due to placental insufficiency.

Sickle cell disease is another example of a condition caused by hyperviscosity and hypercoagulation. The hallmark of this disease is thrombotic vaso-occlusion leading to multiple organ infarctions, caused by adherent sickle-shaped erythrocytes. In mice with sickle cell disease, IVIg treatment reduced the number of adherent neutrophils and subsequently improved blood flow in capillaries and survival rates. IVIg also altered adhesion pathways and allowed an increase in leucocyte extravasation which could reduce inflammation and decrease the risk of TE 9. The use of IVIg in human sickle cell disease has been investigated in 11 case reports and a case series. These studies concluded that IVIg resolved haemolysis without the need for additional blood transfusions 10.

No serious side effects were apparent in any of the animal models or clinical studies. An ongoing clinical trial investigating the effect of IVIg treatment in sickle cell disease patients with pain crises is being conducted.

Inflammation triggered by complement activation plays an important role in the occurrence of cardiovascular disease (CVD). The anti-inflammatory effects of IVIg may be elicited by the ability of Ig molecules to scavenge activated complement fragments and prevent their binding to target cells, thus attenuating immune damage. Different regions of the Ig molecule mediate complement scavenging – the constant region of the Fab fragment binds and neutralizes anaphylatoxins (C3a and C5a) 11, whereas the Fc portion binds larger fragments of complement components C3b and C4b. This prevents their incorporation into the membrane attack complex (C5b-9) which causes cell destruction and exaggerated inflammation, leading to an increased risk of CVD. This mechanism can explain the lower rate of stroke observed in the USIDNET study and how IVIg decreases the likelihood of TE.

In our literature review, we found no in vitro or clinical evidence that IVIg infusion can cause TE. In fact, IVIg treatment in animal models that mimic thrombogenic conditions, such as atherosclerosis, sickle cell anaemia and APS, was beneficial. Similarly, the use of IVIg in human counterparts of these disorders demonstrated improved clinical outcomes and no drug-related TE. Complement fragments C3a, C5a and C5b-9 mediate inflammatory and thrombotic pathophysiology in atherosclerosis, sickle cell disease and APS. IVIg scavenging of potentially harmful complement fragments may explain its anti-thrombogenic effect. Carefully designed clinical trials are needed to confirm this possible new mechanism of action of IVIg.

Disclosures

M. B. has received research grants and consulting fees from CSL Behring.