Abstract
Immune thrombocytopenia (ITP) is an autoantibody-mediated disorder characterized by accelerated platelet destruction and impaired platelet production. Antibody-opsonized platelets engage Fcγ receptors (FcγRs) such as FcγRIIIa (CD16a) on monocytes, macrophages, and NK cells, triggering phagocytosis and inflammatory cytokine release. Preclinical studies in chimpanzees blocked FcγRIIIa and slowed IgG-mediated chromium-labelled red cell clearance (Clarkson JEM1986). Therapeutic value in ITP was limited to a case report (Clarkson NEJM 1986). Intravenous immunoglobulin (IVIG), a mainstay of ITP therapy, rapidly raises platelet counts with the mechanism thought to be saturating FcγRs and thus blocking clearance of antibody-coated platelets. Yet the exact FcγR subtype responsible for this effect remains uncertain. FcγRIII is a logical candidate, but evidence also points to blocking FcγRI, FcγRIIa or upregulation of inhibitory FcγRIIb. Understanding these mechanisms is important for designing targeted therapies, eg anti-RhoD (Salama Lancet 1983), that replicate IVIG's benefits without its cost, infusion burden, or reliance on human plasma.
Two investigator-initiated pilot studies explored selective FcγRIII blockade as therapy for chronic, refractory ITP. In the 1st study (1985–1989), 11 adults received 3G8, a murine IgG1 anti-CD16 monoclonal antibody. In the 2nd (2006–2008), 10 adults were treated with GMA161, a humanized, Fc deglycosylated 3G8 variant engineered to reduce infusion reactions and immunogenicity. All patients had longstanding, severe ITP and had failed multiple prior treatments: steroids, splenectomy, vincristine, immunosuppressives, and, for the 2nd study, rituximab. Almost all patients had received IVIG repeatedly with no or short-lived responses. Platelet response was defined as ≥20 × 10⁹/L increase within 7 days. Safety, immunogenicity, pharmacokinetics, and laboratory correlates were evaluated.
Fourteen of 21 patients (67%) achieved transient platelet responses after anti-CD16 therapy. Platelet counts rose rapidly, peaking days 4–5 after infusion, with mean increments in responders of 159 × 10⁹/L for 3G8 and 121 × 10⁹/L for GMA161. There was no correlation of IVIG response and anti-CD16 response. 3G8 infusions caused acute, severe reactions (fever, chills, nausea/vomiting) which required steroids, acetaminophen, anti-histamine AND additional medication to control. High titer human anti-mouse antibodies (HAMA) prevented redosing. In contrast, GMA161 markedly improved tolerability: only two patients developed low titer, transient anti-drug antibodies, and infusion reactions were milder but still significant. Both antibodies produced predictable laboratory effects, including transient leukopenia (WBC nadirs: 36% for 3G8; 51% for GMA161) and reversible suppression of neutrophil and NK cell activity; high levels of human neutrophil elastase were seen immediately post-infusion reflecting interaction of anti-FcγRIII with neutrophil FcγRIIIb. Infusion of 3G8 Fab fragments, at doses that slowed antibody-coated chromium-labelled red cell clearance in chimpanzees, failed to elicit platelet responses in 2 patients who also both failed to respond to subsequent intact antibody infusion on day 7. Soluble(s) FcγRIII levels were high, confounding RBC clearance studies: certain patients actually accelerated their clearance following 3G8, concomitant with dramatic reduction in sFcγRIII levels.CONCLUSIONS: Targeting FcγRIII with monoclonal antibodies produced rapid platelet increases, sometimes substantial, in most heavily-pretreated, refractory ITP patients in the pre-TPO era. The failure of IVIG and anti-FcγRIII responses to correlate both validates FcγRIIIa as a mechanistic driver of platelet clearance and that it is not the primary FcγR targeted by IVIG. The failure of correlation of IVIG and 3G8 response implies that IVIG's platelet raising mechanism cannot be primarily attributed to FcγRIII blockade. The short-lived responses limit the durability of benefit. Another limitation is that the Fc piece of anti-FcγRIII might engage other FcγR after binding and the failure of the Fab fragments to increase the platelet counts support this. However, the two patients not responding to intact 3G8 and the good results seen with the deglycosylated GMA161 make this less likely. These studies highlight FcγRIII blockade as a therapeutic strategy in ITP but demonstrate the difficulties in addressing it.
