Immune responses to alglucosidase in infantile Pompe disease: recommendations from an Italian pediatric expert panel

Two types of immune reactions against ERT have been reported: 1) infusion associated reactions (IARs) including hypersensitivity responses with or without increase of specific immunoglobulin E (IgE) and 2) development of specific immunoglobulin G (IgG) which reduces treatment efficacy by way of two different mechanisms: a) uptake of the administered enzyme in Fc receptor expressing cells such as monocytes and macrophages (binding, non-neutralizing antibodies) and b) targeting functional or catalytic domains (neutralizing antibodies) of the replaced enzyme [12, 14].

Antibodies and antigens can also form immune complexes and trigger a cascade of adverse events [15]. In addition to an antibody response, ERT has also been shown to induce a T cell response [16].

Based on anti-rhGAA IgG antibody titers, patients are classified into 3 groups [17]:

HSAT is closely associated with clinical decline [18]. Moreover, there is evidence that antibodies levels ≥12,800 (lower range of SIT) over a certain period of time also reduce the enzyme’s efficacy [5]. It appears that the persistence of sustained titers for a long period of time, rather than the absolute values of the titers, impairs the ERT efficacy and thus the clinical outcome [19]. Approximately 90% of IOPD patients treated with alglucosidase alfa develop IgG antibodies [5, 20], with a large majority developing immunological tolerance with continued treatment [18, 21]. Most pediatric patients develop antibodies in the first 3 months of therapy, but this is not always the case [11]. Sometimes, antibodies become positive or peak several years after beginning ERT [22].

Several factors need to be considered in assessing immunogenicity risk. Index signs of the severity of the clinical phenotype such as early onset of symptoms, severe cardiomyopathy, and residual enzyme activity (< 1%), have been considered by some authors to be risk factors for the development of a high antibody titer, while there seems to be no link with ERT dose, infusion rate, and age at the time of therapy inception, contrary to what was assumed in the past [12, 22]. The most important factor indicated appears to be CRIM status [20, 25, 26].

Patients with no detectable GAA protein are classified as CRIM negative (CN), so that their immune system fails to tolerate ERT and recognizes it as foreign (about 30% of patients with c-IOPD). Patients with some residual GAA protein (active or inactive forms) are classified as CRIM positive (CP), and usually do not mount an immune response or else mount a transient low titer response [20, 25]. Previous studies have demonstrated that in CRIM negative patients antibody titers were higher, seroconversion occurred earlier (by 4 weeks), titers were sustained at high levels, and neutralizing antibodies often developed. Conversely, CRIM positive patients showed a variable time to seroconversion (4–64 weeks, median 8 weeks) with either no antibody response or a non-neutralizing antibody response with a low peak titer that diminished with continued therapy [20, 26].

Serotiter levels play a role in the clinical decline of the CRIM negative patients. Kishnani et al. reported that both CRIM positive and CRIM negative groups showed similar cardiological status (left ventricular mass index), gross motor development, and biomarker levels (eg tetrasaccharide) at baseline. After 26 weeks of ERT both groups showed improvement but at 52 weeks, CRIM positive patients demonstrated additional improvement, while the CRIM negative group worsened in correlation with the increased and persistent antibody response. By 27.1 months of age, all CRIM negative patients were deceased or on ventilators compared to 19% of the CRIM positive patients [20]. Similar results were reported by Van Gelder et al. [27].

A previous Italian experience on 28 patients, followed for a median period of 6 years, also demonstrated that, compared to CRIM negative, CRIM positive patients showed a better outcome. In particular, the risk of death in CRIM positive patients was 1/4 and ventilator free survival risk was 1/5 compared to CRIM negative patients. CRIM positive patients also showed improved motor and cardiac outcomes. Of note, data concerning the titers of anti-rhGAA antibodies were available in 13 patients (7 CRIM positive, 5 CRIM negative, 1 unknown) and all CRIM positive patients have no antibodies or were LT, while all CRIM negative patients were SIT or HSAT [28].

Immune tolerance induction (ITI) protocols can be used for prophylaxis, to preempt immune response in ERT naive patients (prophylactic approach), and for therapy, to decrease existing antibodies in ERT treated patients with previously established immune responses. For this purpose, it is important to know CRIM status before starting ERT. The gold standard to determining CRIM status is the western blot analysis of skin fibroblast lysates. However, skin biopsies are invasive and require several weeks. A more rapid method for determining CRIM status using western blot analysis of blood mononuclear cells (PBMCs) was successfully tested on a small number of patients (n = 8) [29] but its sensitivity was assessed at only around 82% in a subsequent study including 33 patients [30]. CRIM status can also be predicted based on GAA gene mutations in more than 90% of patients. The types of mutations in Pompe disease are shown in Table 1 [31, 32]. If, based on the mutation, CRIM status is not already foreseen, a western blot on fibroblasts should be performed [31].

Most CRIM negative patients develop HSAT, but limited information is available to accurately predict those who are least likely to develop it. Accordingly, a risk-benefit ratio supports initiating an immunotolerance protocol at the time of ERT inception. Advantages of preemptive immunotolerance induction are exposure to fewer drugs, shorter duration of therapy and improved clinical outcomes by preventing prolonged exposure to HSAT [33]. Several protocols have been used as described below and summarized in Table 2.

Banugaria et al. first reported this protocol in 7 CRIM negative patients. Four patients never seroconverted, one remained with a LT of antibodies, while other two patients had a peak of 1:6400 and receiving a second course of the same ITI regimen. In fact, their protocol provided that patients with antibody titers of ≥6400 at 2 or more time-points after CD19+ B cell recovery at 5 months were administered another cycle of the ITI regimen with or without a plasma cell targeting agent [23].

Successively, Kazi et al. reported the same protocol in 19 CRIM negative patients. 15 out of 19 patients (79%) either did not develop rhGAA IgG antibodies (n = 8, never in ERT monotherapy group) or maintained low titers (≤6400) even after B lymphocytes reappeared. Four patients required further cycles of ITI (1 HSAT, 3 SIT) [34]. This may be due to a Fcγ receptor polymorphism which reduces the efficacy of rituximab [40]. No significant difference in baseline characteristics was noted between patients who had broken tolerance to the ERT compared with those who had maintained low/no titers. Patients who had been undergoing immunomodulation regimen, when compared to those on ERT monotherapy (n = 10), showed improved overall and ventilator-free survival, improved cardiomyopathy, motor function and feeding (like CRIM positive survivors on ERT), evident after the first 6 months. Conversely, during the same period of time, the monotherapy group worsened because of development of antibodies within the HSAT range. Although serious infections were noted in 4 patients, no immunomodulation regimen was withheld due to safety concerns. Titer against routine vaccines was performed after B cell recovery in 4 patients, which demonstrated the ability to mount a humoral response to other antigens, although not in all patients nor for all antigens. This data has been confirmed by other authors as well [22, 41]. Poelman et al. used the same protocol, but with a different dosage of methotrexate (Table 2) on 1 CRIM negative and 2 CRIM positive patients [36]. However, the immune regimen failed to induce immune tolerance: 2 patients developed an high and sustained titer (1 CRIM negative and 1 CRIM positive) and 1 CRIM positive patient developed a titer of 1:6250. More recently, Li et al. reviewed a cohort of CRIM negative patients, among which 20 were immunomodulated before starting ERT. All remained immunotolerant, 13 no seroconverted, 7 developed LT. Compared to historical cohort treated with ERT monotherapy (n = 10), of whom median peak titer was 204,800 (range 25,600-1,638,400), patients treated with ERT + ITI had significantly lower median peak titer of 0 (n = 20, 0–51,200). Surprisingly, within the cohort treated with ERT + ITI, although early treated patients (≤4 weeks) showed a better outcome than late treated, the number of immunotolerant patients (titre ≤6400) did not differ significantly between these groups. However, early treated patients were significantly more likely to remain seronegative than late treated patients [42].

Elder et al. described treatment of 5 infantile onset patients (4 CRIM negatives, 1 CRIM positive) using this protocol [37]. It prevented antibody formation with improved outcomes and no adverse events except in 1 CRIM negative patient (who had not received maintenance rituximab, but only mycophenolate) who developed HSAT and subsequent clinical decline. One limitation is that this protocol delays initiation of ERT for 3 weeks, which can lead to irreversible muscle damage.

A subset of CRIM positive patients (about 40%) develop HSAT. These patients are defined as high-titer CRIM positive (HTCP patients) and have a similar pattern of antibody response and clinical decline as CRIM negative patients. The most marked difference between CRIM positive HSAT and CRIM negative patients regards the development of neutralizing antibodies which seem to be a common occurrence in CRIM negative groups. The basis for this finding is not clear; however high titers of binding antibodies such as those present in HTCP patients may also abrogate product efficacy by several mechanisms. HCTP patients have similarly poor outcomes compared to CRIM negative patients [18, 43]. The main characteristics of immune responses in different patient groups is shown in Table 3.

Kazi et al. used a transient low dose methotrexate protocol to prevent the developing of antibodies in naive CRIM positive patients (Table 2) [17]. In a study on 14 patients, 2 (14.2%) developed titers in the HSAT or SIT range versus 32.4% in the comparison group (n = 37).

A more recent work of the same group retrospectively studied the effects of a more complex protocol (short-course of rituximab, methotrexate, and/or IVIG, as described for CRIM negative patients, see above) in a cohort of 9 ERT-naive CRIM positive patients. None developed SIT or HSAT, after a median timepoint following ERT initiation of 104 weeks [38].

The elimination of anti-ERT antibodies, particularly those within the range of HSAT, has been challenging as patients require prolonged and higher intensity immune modulation and sometimes immune suppression is unsuccessful [13, 45]. Furthermore, since antibodies take weeks to decline to clinically insignificant levels, the enzyme is less likely to have an effect during this period of HSAT and the disease continues to progress [40]. The protocols used are summarized in Table 4.

This protocol was shown to be effective by Mendelsohn et al. in 1 CRIM negative patient with titer 1:1600 [46], Messinger et al. in 2 CRIM negative patients with titers 1:1600 and 1:12,800 respectively [35], Markic et al. in 1 CRIM positive patient with titer 1:6400 [49], and Blasco-Alonso et al. in 1 CRIM negative patients with titer 1:32000 [50], but it often required prolonged drug administration. Moreover, the success of eliminating anti-drug antibodies is demonstrated only in patients with non-HSAT titers. It is possible that patients with HSAT may not respond to this regimen due to the use of drugs that do not target antibody producing plasma cells [43].

To eliminate long-memory plasma cells, especially in HSAT patients, some authors added bortezomib, a plasma cell-targeting agent [23, 47]. These authors described 1 CRIM negative and 2 CRIM positive HSAT who after treatment showed a rapid sustained decline of the antibody titer associated with clinical improvement and reduction of urinary tetrasaccharide levels without serious adverse events or infections. Similar results were obtained by Stenger et al. in 1 CRIM positive patient with HSAT [41]. Conversely, Owens et al. reported 2 patients (1 CRIM positive and 1 CRIM negative) in whom bortezomib failed to eliminate HSAT despite the reduction in IgG antibody titer. This could be explained by the administration of a limited number of bortezomib cycles and the already advanced disease progression at the time of bortezomib initiation [51].

Poelman et al. described this protocol in two CRIM negative and 1 CRIM positive patients with HSAT. The protocol did not eliminate anti-ERT antibody titers although the titers decreased and the neutralizing effects and infusion-associated reactions were no longer present. None of the patients experienced a serious adverse event [48].

Anecdotal reports have been made on the use of plasma exchange to remove pathogenic antibodies and immune complexes from plasma and high-dose IVIG (1–3 g/kg/week) in patients with adverse reactions to other drugs, but with modest results, and in patients with low antibody titers involving, in the case of plasma exchange, additional risks related to the procedure [52, 53].

Lastly, ERT dosage can have a role in determining the efficacy of the drug in patients with an antibody response: with a higher dosage (40 mg/kg/week), more antibody-free ERT should be available and the neutralizing effects of the same titer are likely to be less severe than in patients receiving 20 mg/kg every other week. For example, van Gelder et al. calculated that an ELISA titer of 1:156,250 may bind as much as 54% of the administered enzyme at a dosage of 20 mg/kg (about 10 mg/kg) [27]. Theoretically, if a similar amount is bound upon administration of 40 mg/kg, about 30 mg/kg would still be available for uptake in target tissues [27, 48]. Recently, an observational study of the European Consortium project group on 124 c-IOPD concluded that ERT dosage of 40 mg/kg/week significantly improves survival compared to standard dosage (20 mg/kg every other week) [54]. Further studies on patients with positive antibodies are needed.

The immunomodulation regimens target immune cells at different levels including elimination of B cells, T cells, plasma cells and induction of GAA-specific B_reg and T_reg. Table 5 summarizes the mechanisms of immune tolerance and potential adverse events of pharmaceutical agents that are commonly used. It should be noted that the duration of rituximab-induced immunosuppression is about 5 months (based on the half-life of the drug) and the increase in CD19 can be associated with increased specific antibody response [23].

Immunomodulation regimens are relatively non-antigen specific and have the potential to cause generalized immune suppression with a risk of infection, also with opportunistic pathogens, and malignancy [23, 37]. Patients with Pompe disease are already susceptible to lung involvement with higher risk of infections. Surprisingly, very few reports of severe life-threatening infections have been reported. This may be due to the use of intravenous immunoglobulins [23] and to the improvement of the underlying disease [12]. There are anecdotal reports of the use of preventive antibiotic therapy (e.g. azithromycin) during immunomodulation regimens [48]. Vaccination response might be diminished during immunotolerance induction and during the following 6 months [23]. There is little documented experience on the long-term effects of immune modulating therapies in children over time and long-term assessment is needed.

About half of patients with Pompe disease experience infusion associated reactions which can involve a wide range of clinical symptoms especially cardiac, respiratory, cutaneous and/or gastrointestinal manifestations. Approximately 1% of patients develop anaphylactic shock and/or cardiac arrest that requires life-support measures, while 5–14% of patients develop significant allergic reactions that involve at least 2 or 3 body systems [5]. These can appear at various times, even years after the start of the ERT [27]. Risk factors include the presence of IgE (which does not seem to correlate with IgG), acute illness at the time of the infusion, or a very high-rate infusion regimen, while there does not seem to be a correlation with dosage or IgG levels [22, 69, 70]. In a previous Italian observational study, among 128 patients eight experienced hypersensitivity reactions (6.25%) [28].

Hypersensitivity reactions can be classified as allergic and non-allergic. Allergic hypersensitivity, either IgE- or non-IgE mediated, is antigen specific. In this case skin tests and intradermal tests are usually positive and tryptase increases within 1 h after clinical reaction (peaks at 60–90 min after the onset of symptoms and remain elevated for up to 5 h), as the reaction is mast cell/basophil mediated [71‐73]. A non-allergic reaction (more frequent) is not antigen specific. An immunologic pathogenic mechanism has not been demonstrated, but may be related to mechanisms of complement activation, cytokine release, and direct mast cell stimulation [5].

As infusion-associated reactions of various mechanisms have similar presentation, clinicians are unable to use clinical history to predict these events and guide management [74]. After a potential hypersensitivity reaction has been identified, three approaches are possible: standard prevention and management, immunomodulation, and desensitization [75].

Standard prevention and management include: slowing down or interrupting infusion and then restored it at the last safe rate (then, gradually increasing dose, rate, and concentration), monitoring of vital signs, symptom-specific medical intervention (e.g. antipyretics, antihistamines, corticosteroids, inhaled short-acting beta2-agonists if bronchoconstriction, epinephrine if anaphylaxis) and drawing blood for laboratory tests [3, 10, 11, 76‐78]. After an adverse reaction, a premedication protocol can be applied. Most of the premedication protocols provide for the use of antihistamines, glucocorticoids, and antipyretics [79]. However, prophylactic antihistamines are not recommended in patients with a history of IgE positive hypersensitivity reactions, as this may mask early symptoms of a hypersensitivity reaction such as a cutaneous reaction [69]. Some authors have suggested the use of tranexamic acid (500 mg/day), especially in patients with angioedema or complement-mediated infusion-associated reactions, since it inhibits plasmin activation of kallikrein and interrupts the kinin generating cycle [79]. A novel approach for IgE mediated anaphylaxis uses an IgE monoclonal antibody, namely omalizumab. One patient had been successfully treated with omalizumab, but he appeared to require its chronic administration to continue ERT. Interestingly, this patient, who was CRIM negative, also maintained a low IgG antibody titer suggesting that omalizumab, perhaps via signaling through the Fcε receptor, had also limited development of the IgG response [80]. Nonetheless, further studies are needed.

Some immunomodulation approaches, as those described above, have also demonstrated to alleviate hypersensitivity reactions, if it is sustained by an antibody response. Thus, they are a limited role in cases with positive anti-rhGAA antibodies. Moreover, these regimes result in significant risks and reappearance of antibodies following successful initial suppression has been reported [13, 78].

If these measures fail or are unapplicable, desensitization is strongly indicated. Desensitization is the induction of a temporary state of unresponsiveness to a compound responsible for a hypersensitivity reaction (IgE and non-IgE mediated) by increasing sub-therapeutic doses over a short period of time until the total cumulative therapeutic dose is delivered. It is a high-risk procedure and should be used only in patients in whom there is no safe, effective alternative drug treatment (e.g. ERT in Pompe disease) and after careful risk-benefit analysis [72].

Infants with Pompe disease subject to a desensitization protocol require attention to special aspects such as the need for administration of a large amount of enzyme (compared to patients on ERT for other disorders), and the restricted amount of fluid intake allowed due to underlying cardiomyopathy [69]. Particular attention should be paid to patients at increased risk because of a comorbidity such as hemodynamically unstable patients, those with uncontrolled cardiac disease, or subjects who have experienced severe anaphylaxis. No specific guidelines yet exist. General rules are suggested by the “EAACI consensus statement on rapid desensitization for drug hypersensitivity” [61]. Any concomitant medication must be continued however drugs such as beta-blockers, which can interfere with the treatment of a severe hypersensitivity reaction, should be discontinued whenever possible. Generally it is preferable to use desensitization protocols after they have been successfully proven on at least 10 patients [61] although only a few Pompe patients having ERT with desensitization have been described.

To our knowledge, 10 cases (8 pediatric, 2 adults) affected by Pompe disease and subjected to desensitization protocol have been reported [14, 56, 74, 75, 79, 81, 82] (Table 6). Of note, recently Emecen Sanli et al. reported the first case of successful concomitant immunotolerance induction and desensitization protocol in a CRIM negative 7-month-old male patient, that have developed IARs and anti-rhGAA antibodies (1:12,800) [83].

Most of these protocols (6/10) administered a half dose (10 mg/kg) once a week. Only one patient was on a very low dose of ERT for a longer period (8 weeks) [75]. Breakthrough reactions are drug hypersensitivity reactions that occur despite the desensitization procedure. These most often occur during the first course of desensitization [72]. Such reactions have been reported in 7 of 10 patients and were mild/moderate in 4, while 3 patients experienced severe infusion-associated reactions including anaphylactic reactions requiring epinephrine injections in 2 cases (Table 6). After a breakthrough reaction, some authors did not modify the protocol, while others decelerated the dose escalation introducing intermediate dosing steps, or going back one or two steps, depending, however, on the severity of the reaction.