Association of anti-IgA antibodies with adverse reactions to gamma globulin infusion

Serum IgG anti-IgA antibody has been associated with the development of adverse reactions, (including non-IgE mediated anaphylaxis) to intravenous immunoglobulin (IVIG) in patients with undetectable IgA (here defined as <7 mg/dL).¹ Class-specific anti-IgA antibodies bind both IgA1 and IgA2; they are found in 25–30% of IgA-deficient individuals, but not in those with IgA> 7mg/dL.² Subclass-specific (or limited specificity) anti-IgA reacts with only one subclass [IgA1, IgA2m(1) or IgA2m(2)].² It is estimated that up to 40% of patients with selective IgA deficiency and 9–25% of patients with common variable immunodeficiency (CVID) may have IgG anti-IgA.³ IgE anti-IgA has been found much less frequently. Three of four patients with IgE anti-IgA had anaphylaxis to gammaglobulin or other blood products.^4,5

We conducted a retrospective and prospective observational study to evaluate the possible association with adverse reactions of IgG and/or IgE anti-IgA in IgA-deficient patients receiving IVIG or subcutaneous immunoglobulin (SCIG) at Children's Hospital Boston, Boston, MA, and Brigham and Women's Hospital, Boston, MA. All investigations were conducted according to the policies and procedures of the Institutional Review Boards of both institutions.

Among 425 immunodeficient patients receiving IgG therapy, we identified 35 (8.2%) with undetectable IgA; 22 were enrolled. Twenty of these were studied retrospectively (medical record review for one year prior to enrollment) and prospectively (6 months) for symptoms related to IgG infusion. Thirteen patients (11 CVID, 2 IGAD) were excluded due to increase in IgA level (3), lack of consent (7), or non-adherence with study procedure (3). Blood was collected at enrollment and after the 6 month observation period. Subjects #22 and #23 were only studied prospectively since they did not receive IgG in the year prior to obtaining the serum specimens. Data were collected with respect to the IgG product, lot number, dose, rate of infusion, use of premedications, and symptoms recorded within 1 hour after the end of the infusion (acute) or up to 72 hours later (delayed). Symptoms were graded as mild (treated by the patients or medical staff without discontinuing the infusion), moderate (requiring telephone contact with, or a visit to an outpatient setting for assistance from a health-care provider, or requiring cessation of the infusion), or severe (requiring an emergency room visit or hospitalization, or resuscitation).

We measured serum levels of IgA1, IgA2 and IgG anti-IgA1 and anti-IgA2 by ELISA using myeloma controls to establish approximate concentration standard curves. Serum specimens were also studied at the Mayo Clinic Laboratory (Rochester MN) using a Luminex®-based assay system and in the Red Cross Laboratory (Philadelphia, PA) by passive hemagglutination. IgE anti-IgA1 and anti-IgA2 were measured by ImmunoCAP assays at ViraCor-IBT Laboratories (Lenexa, KS).

In all patients throughout the study, only mild symptoms were reported, including headache, fatigue and malaise. These were of a nature and frequency commonly seen in many clinical trials of IgG therapy that routinely exclude IgA-deficient patients. Therefore no attempt was made to correlate these symptom data with results of immunoassays.

Results of the immunochemical analyses are shown in Table E1 (available online). Background levels for IgG anti-IgA1 or IgA2 ELISA ranged from 51–240 ng/mL (lower limit of assay detection, sera diluted 1:100). Three subjects had levels well above this range (504–4,528 ng/mL) (Table 1). Two individuals (#23 and #32) had class-specific IgG anti-IgA by ELISA. One subject (#33) had subclass-specific IgG anti-IgA2. Subjects #32 and #33 have only ever received SCIG and have never had adverse reactions. Their ability to tolerate IVIG is unknown. Seven years prior to enrollment, patient #23 had anaphylaxis requiring epinephrine during infusion of an IVIG product containing <10 mcg/mL IgA and ceased IgG therapy. The specimen for measurement of IgG anti-IgA was obtained at the time of enrollment in the prospective study, when she began receiving SCIG, which she has now tolerated for 2 ½ years without any adverse effects. She had a level of IgG anti-IgA (mean of anti-IgA1 and anti-IgA2) of 3,946 ng/mL. The Mayo lab reported a total IgG anti-IgA level >1,000 arbitrary units and in the Red Cross Laboratory IgG anti-IgA was detected. Her IgG anti-IgA level repeated 2 ½ years later at the Mayo Laboratory was still >1000 AU.

Table E1.

Results of immunochemical analysis.

Study #	Sex	Agea (years)	Dx., SC/IVb	Spec. No.	IgA1 (ng/mL)	IgA2 (ng/mL)	CHB IgG anti-IgA1 (ng/mL)	CHB IgG anti-IgA2 (ng/mL)	Mayo IgG anti-IgA (AU)c	RC IgG anti-IgAd
01	F	8	CID, IV	1	983	1261	127	170	91	ND
				2	824	344	98	164	67	ND
02	F	43	CVID, IV	1	1030	632	60	63	24	ND
				2	445	275	63	68	33	ND
03	F	43	CVID, IV	1	436	220	138	126	84	ND
				2	375	297	170	128	63	ND
04	F	31	CVID, SC	1	185	96			53	ND
				2	550	286	103	92	67	ND
06	F	40	IGAD, IV	1	327	305	129	135	56	ND
				2	318	287	126	155	55	ND
07	F	37	CVID, IV	1	263	159	68	93	45	ND
				2	469	330	241	232	114	ND
08	F	17	CVID, IV	1	459	165	67	93	35	ND
				2	380	150	78	96	48	ND
13	F	15	G4/IGAD, IV	1	342	418	97	164	36	ND
14	M	13	Hypogam, IV	1	343	175	124	108	52	ND
15	M	35	CVID, IV	1	661	232	145	174	73	ND
				2	393	109	59	58	22	ND
18	M	17	CVID, IV	1	245	82	71	81		ND
				2	238	75	87	89	45	ND
19	M	6	Hypogam, IV	1	2473	621	91	100	41	ND
				2	1166	764	118	65	25	ND
20	F	23	CVID, IV	1	236	110	82	106
				2	225	114	80	93	68	ND
21	F	55	CVID, IV	1	185	59	60	73	57	ND
22	M	6	Hypogam, IV	1	190	53	49	58	14	ND
				2	222	89	74	75	26	ND
23	F	26	CVID, IV	1	52	20	4528	3363	>1000	Anti-IgA
30	F	53	CVID, SC	1	1017	323	115	109	82	ND
				2	773	337	125	115	75	ND
31	M	55	CVID, IV	1	95	4	97	75	54	ND
				2	58	0	75	70	46	ND
32	F	60	CVID, SC	1	150	32	2821	1859	684	ND
				2	121	20	4398	3563	>1000	ND
33	M	60	CVID, SC	1	204	116	167	755	172	Anti-IgA2
				2	253	71	130	504	133	Anti-IgA2
34	F	80	CVID, IV	1	130	64	51	98	80	ND
				2	127	47	91	94	77	ND
35	F	29	Hypogam, SC	1	124	84	79	81	65	ND
				2	120	92	63	52	55	ND

An empty cell indicates the test was not done.

^aAge when first serum specimen obtained

^bG2=IgG2 deficiency, G4=IgG4 deficiency, CID=combined immunodeficiency (see list for other abbreviations); hypogam=unspecified hypogammaglobulinemia, consisting of low total IgG level with apparent adequate vaccine antibodies; therapy at time when serum specimen obtained, IV = intravenous infusion; SC = subcutaneous infusion;

^cAU = arbitrary units, <100 = negative, 100–200 = equivocal, >200 = positive; maximum result reported >1,000

^dQualitative test, results reported as type of anti-IgA detected, or not detected (ND)

Table 1.

Characteristics of patients with positive anti-IgA antibodies.

Study #	Sex	Agea (years)	Dx., SC/IVb	Spec. No.	IgA1 (ng/mL)	IgA2 (ng/mL)	CHB IgG anti-IgA1 (ng/mL)	CHB IgG anti-IgA2 (ng/mL)	Mayo IgG anti-IgA (AU)c	RC IgG anti-IgAd
23	F	26	CVID, IV	1	52	20	4528	3363	>1000	Anti-IgA
32	F	60	CVID, SC	1	150	32	2821	1859	684	ND
				2	121	20	4398	3563	>1000	ND
33	M	60	CVID, SC	1	204	116	167	755	172	Anti-IgA2
				2	253	71	130	504	133	Anti-IgA2

^aAge when first serum specimen obtained

^bTherapy at time when serum specimen obtained, IV = intravenous infusion; SC = subcutaneous infusion;

^cAU = arbitrary units, <100 = negative, 100–200 = equivocal, >200 = positive; maximum result reported >1,000

^dQualitative test, results reported as type of anti-IgA detected, or not detected (ND)

The two patients that had class-specific IgG anti-IgA in our assay also tested positive in the Mayo laboratory. The one individual with subclass-specific IgG anti-IgA2 did not test positive in the Mayo laboratory (result in the “equivocal” range). Subject #23 also tested positive in the Red Cross laboratory; subject #32 did not. The test was repeated with the same result. The reason for the discrepancy is unknown. The Red Cross laboratory did detect the subclass-specific IgG anti-IgA2 that was also found by our ELISA in pt. #33. Overall, there is good agreement between our results and these two clinical reference laboratories. IgE anti-IgA was not detected in any patient.

Anaphylaxis is very rare among patients receiving IVIG.⁶ Therefore, a possible increased risk associated with IgG anti-IgA remains very difficult to quantify. None of the patients evaluated developed a significant reaction during the study period. We could not address the potential clinical significance of the subclass-specific anti-IgA2 with respect to IVIG, since the only such patient we identified had only ever received SCIG.

Is IgG anti-IgA a “biomarker” of increased risk of non-IgE mediated anaphylaxis to gammaglobulin infusion containing IgA? Several anecdotal reports (including this one) suggest this may be the case. The mechanism whereby anti-IgA antibodies might cause an adverse reaction to IgG that contains some IgA is open to speculation. It has been reported that some individuals react to products with “high” IgA and tolerate products with “low” IgA.^{7, 8} This could be interpreted as an indication that the reaction mechanism involves interaction of IgG anti-IgA with infused IgA. However, patient #23 appears not to tolerate IVIG products containing even trace amounts of IgA and some patients with IgG anti-IgA tolerate IVIG without symptoms.⁹

Given the rarity of anaphylaxis to IVIG, much larger prospective studies are required to establish more clearly any possibility of increased risk associated with IgA deficiency. It is not clear that class or sub-class specific IgG anti-IgA antibodies have clinical relevance. Our data and others' suggest that further study of a possible association is warranted. In an individual patient, the presence of IgG anti-IgA may indicate a need for closer monitoring, or consideration of alternative therapy such as SCIG, which appears to be tolerated in many of these patients.

Capsule summary.

In a cohort of 22 IgA-deficient patients receiving gammaglobulin, we found three with IgG anti-IgA antibodies, but only one had ever received intravenous infusion, and had non-IgE mediated anaphylaxis. All three tolerate subcutaneous IgG.

Abbreviations

CVID

common variable immunodeficiency

IGAD

IgA deficiency

IMIG

intramuscular immunoglobulin

IVIG

intravenous immunoglobulin

SCIG

subcutaneous immunoglobulin