DISCUSSION
This is the first prospective trial describing the successful use of rhC1-INH as prophylactic treatment in one out of six patients with InH-AAE (patient 1). Restart resulted in an equally fast and beneficial response. pdC1-INH also appeared to be effective. We hypothesize that the bradykinin route is pathomechanistically involved although we found no formal proof after biomarker investigations. Post-trial, patients 1, 2, 5 and 6 received omalizumab treatment. Patients 2 and 6 experienced a good, and a near complete response respectively. Patients 1 and 5 reported no clinical effect of omalizumab.
We observed an almost complete and fast improvement of AE activity in patient 1, with only 3 mild attacks in the treatment period versus 19 attacks in the observation period. Equal effectiveness was found during post-trial follow-up upon restart of rhC1-INH treatment on two different treatment periods with a similarly fast and near complete response. This was unlikely a placebo effect, since post-trial follow-up data showed a similar response to pdC1-INH, whereas treatment with omalizumab had no effect and resulted in an increased attack rate. The effect of rhC1-INH was in line with that in HAE, in which a sustained effect of 72 hours or longer was observed in 93% of HAE patients despite the short half-life of approximately 3 hours.23 Clinical response to recombinant (r)hC1-INH in patients with InH-AAE was not published before. Three previous case reports describe successful use of plasma derived (p)dC1-INH prophylaxis in four patients with InH-AAE.16-18 However, from these reports, it is not clear to what extent the observed responses were due to a placebo effect, and what proportion of patients may respond to such therapy. Our data show that only a proportion of the patients with InH-AAE may respond to C1-INH treatment. Failure of efficacy of rhC1-INH in the other 5 InH-AAE patients may suggest that bradykinin is not involved in generating attacks in these patients. However, we favour another explanation. A dose of 50 IU rhC1-INH per kg increases circulating C1-INH levels by approximately 2-fold. This increase may have been too low, especially since factor XIIa bound to an activator is less well inhibited by C1-INH.24
None of the biomarkers of the bradykinin route (C1-INH, FXII, PK, HK) differentiated the responding patient from the others. Moreover, none of these biomarkers pointed to an increased activation of the contact system. One may argue that levels of bradykinin could provide a better biomarker for such activation. This is supported by the finding of elevated bradykinin levels in InH-AAE patients at the time of AE attacks.10 Considering the technical issues, such as processing of plasma samples, we did not assess bradykinin levels in the patients described here.
After the study period, treatment of patient 1 with both rhC1-INH and pdC1-INH resulted in an immediate effect, though the effective treatment interval appeared to be 3-4 days for pdC1-INH versus 5-6 days for rhC1-INH. Previous studies show that there is no fundamental difference in efficacy between pdC1-INH and rhC1-INH in patients with HAE, when similar doses, expressed as IU, are given.34 However, doses given to patient 1, who had a body weight of 100 kg, were different, 4200 IU of rhC1-INH versus 1000 IU of pdC1-INH per administration. Thus, the increase of C1-INH activity, which was approximately 1 U per ml in case of rhC1-was considerably lower, about 0.25 IU, when pdC1-INH was administered. Therefore, C1-INH levels may have decreased below a critical level more rapidly in cases of pdC1-INH administration, in spite of its longer half-life.
The restart of C1-INH treatment in patient 1 resulted in an immediate effect with a clear difference of the effective treatment interval of 3-4 days for pdC1-INH versus 5-6 days for rhC1-INH. Previous studies show that there is no fundamental difference in efficacy between pdC1-INH and rhC1-INH in patients with HAE.25 The observed difference in treatment interval is therefore most likely explained by dosage inequivalence since rhC1-INH is dosed at 50IU/kg with a maximum of 4200IU and pdC1-INH at a fixed dose of 1000IU.26, 27 In our patient with body weight 100kg, 1000IU pdC1-INH might have been relatively under dosed compared to 4200IU rhC1-INH resulting in lower effectivity and the need for shorter treatment intervals.28
C1-INH treatment is generally well tolerated, though recent studies suggest a small risk of trombotic events upon treatment with rhC1-INH and pdC1-INH in HAE patients.29, 30 We did not observe thrombotic or any other adverse events in the patients studied, even though we supplemented C1-INH in patients with normal levels.
A review of omalizumab for InH-AAE in six small case series, reports a complete response in all 20 patients, with time to response ranging from one day to 16 weeks.31 Four patients with InH-AAE who did not respond to rhC1-INH treatment, were subsequently treated with omalizumab. Two patients responded to this therapy. Patient 1 and 5 did not show any benefit from three months omalizumab treatment, although the possibility that the treatment period with omalizumab was too short cannot be excluded.32 Prospective evaluation in an unselected population is needed to gain insight about the real percentage of responders. Response to omalizumab may be interpreted as evidence for histamine as the main mediator of AE in these patients.32  One should be careful about making this conclusion, as marked activation of FXII, PK, and kininogen has previously been found during anaphylaxis, indicating bradykinin cannot be ruled out as the main mediator of AE even in conditions commonly associated with mast cell activation.33
In conclusion, rhC1-INH treatment was effective in 1 of 6 InH-AAE patients. Response to omalizumab and tranexamic acid during follow-up in some of the other patients points to a heterogeneous pathogenesis of this disease and, as a consequence, the need for a personalised treatment approach.