Perinatal hypophosphatasia refractory to asfotase alfa with neutralizing antibodies that affected bone mineralization: a case report

Abstract.

Hypophosphatasia (HPP) is a rare osteometabolic disease. Enzyme replacement therapy (ERT) for HPP was approved in 2015 and has significantly improved the survival and quality of life of patients. Poor responses to ERT have been reported; however, detailed information is limited. We encountered a case of severe perinatal HPP refractory to ERT with neutralizing antibody (NAb) expression that possibly affected bone mineralization. Asfotase alfa (AA) (6 mg/kg/wk) was initiated 2 mo after birth and resulted in complete resolution of rickets by age 7 mo. However, rickets recurred at age 18 mo without any other identifiable cause than NAbs detected. The AA dose was increased to 9 mg/kg/wk based on the United States prescribing guidelines when the patient was age 3 yr. By ages 4 yr and 8 yr, rickets in the upper limbs and lower limbs, respectively, had nearly disappeared. NAbs were not detected at age 6 yr. We reduced the AA dose (6 mg/kg/wk) at age 8 yr. Rickets recurrence was not observed. Changes in bone mineralization corresponded to NAb expression, suggesting that NAbs may have influenced the therapeutic effect. The optimal AA dose may vary based on clinical findings and the NAb status.

Highlights

● Detailed reports of HPP refractory to AA are limited.

● The efficacy of AA may be affected by NAb expression.

● The optimal dose of AA for HPP may vary with the presence of NAbs.

Introduction

Hypophosphatasia (HPP) is a rare osteometabolic disease characterized by low serum alkaline phosphatase (ALP) levels caused by mutations in the ALPL gene, which encodes tissue-nonspecific ALP (TNSALP) (1, 2). Because TNSALP controls skeletal and dental mineralization, its deficiency can result in chest wall instability and respiratory complications (2). The severity and age of onset of HPP vary, and it has several clinical forms, including perinatal, prenatal benign, infantile, childhood, adult, and odontohypophosphatasia (1).

Asfotase alfa (AA), which is a genetically engineered human recombinant TNSALP, was approved as a novel enzyme replacement therapy (ERT) in 2015 (3). The 1-yr and 5-yr survival rates of patients with perinatal HPP or infantile HPP improved from 42% and 27% without AA treatment to 95% and 84% with AA treatment, respectively (2). AA administration can cause neutralizing antibody (NAb) expression, which, according to several reports, does not significantly alter its efficacy (4,5,6); however, the proportion of patients with a positive NAb status among those without a response to AA treatment (AA nonresponders) is consistently higher than that among those with a response to AA treatment (AA responders) (6). Additionally, one case report suggested that NAb development is associated with the loss of AA efficacy (7). Nevertheless, detailed reports of the clinical courses of refractory cases, strategies for their management, and the impact of NAbs on such cases are limited.

Here, we report the detailed 11-yr clinical course of a perinatal HPP case treated with AA that involved concurrent worsening of radiographic findings of rickets and NAb development and improved after the disappearance of NAbs.

Patients and Methods

The patient was an 11-yr-old Japanese girl who was born via scheduled cesarean delivery at 38 wk and 4 d of gestation with a birth weight of 2543 g. The patient required intubation and mechanical ventilation because of severe chest retractions. Radiographic imaging at birth revealed a large anterior fontanelle, defective cranial ossification, thin ribs, and cupping of the humeri and femurs (Fig. 1). A blood test revealed an extremely low serum ALP level of 5 U/L and a high urinary phosphoethanolamine (PEA) level of 1,286 µmol/L (Table 1, Fig. 2), resulting in the clinical diagnosis of HPP. Genetic testing revealed homozygous c.1559delT deletions in the ALPL gene, which is the most common variant in Japanese patients with HPP (8). AA (6 mg/kg/wk divided into three doses) was initiated at age 54 d (Fig. 2). Osteogenesis was observed 2 wk later, and radiographic findings indicated complete resolution of rickets at age 222 d; the Rickets Severity Score (RSS; 10-point Thacher score) (9) decreased from 10 to 0 (Fig. 2). Tracheostomy was performed at age 144 d, mechanical ventilation was discontinued at age 164 d, and cannula removal was performed at age 300 d. At age 10 mo, the patient was discharged home with injectable AA.

Fig. 1.

Radiographic images of the patient at birth. (A) Head. (B) Chest and upper arms. (C) Abdomen and lower extremities.

Table 1. Status of the anti-AA antibodies and laboratory data during the clinical course.

Fig. 2.

. Graphs showing time-course changes in the Rickets Severity Score (RSS), the dose of asfotase alfa (AA), the urinary phosphoethanolamine to creatinine ratio (U-PEA/Cre; solid line), the urinary phosphoethanolamine level (U-PEA; dotted line), and the urinary calcium to creatinine ratio (U-Ca/Cre). Radiographic images of the left upper and lower limbs obtained during the clinical course, with the corresponding wrist and knee RSS values shown below each panel, are incorporated into the graphs. The timing of neutralizing antibody (NAb) testing is indicated by arrows, and the corresponding NAb status is shown using the symbols (+) or (–).

Measurements of anti-AA antibodies and NAbs

Anti-AA antibodies and NAbs were measured using an electrochemiluminescence immunoassay and an enzymatic activity inhibition assay, respectively, by PPD Laboratory Services (10). The positive status was defined as ≥ 21.4% and ≥ 4.478% inhibition, respectively. These antibodies were measured in response to clinical deterioration or improvement or at clinician’s discretion. The actual timing and frequency of the measurements are shown in Fig. 2.

Results

At age 1 yr and 6 mo, radiographic imaging indicated rickets recurrence (the RSS increased from 0 to 5.5) without an identifiable cause such as vitamin D deficiency or hypophosphatemia (Table 1). The urinary PEA level initially decreased (70 µmol/L at age 1 yr and 4 mo), but subsequently increased to 312 µmol/L at age 1 yr and 7 mo (Table 1, Fig. 2). The ratios of pyridoxal-5’-phosphate (PLP) to pyridoxal (PL) and to 4-pyridoxic acid (PA) was 0.76 and 4.83, respectively (Table 1). The ratio of urinary calcium (Ca) to creatinine (Cre) increased up to 0.96 (Table 1, Fig. 2) and nephrocalcinosis was detected by renal ultrasonography around the same time. Although nephrocalcinosis persisted throughout the follow-up period, no renal calculi were observed and the renal function remained normal. Anti-AA antibodies were positive at age 8 mo and remained positive since that time (Table 1). NAbs were initially negative; however, they were positive at age 1 yr and 7 mo (Fig. 2). Therapy comprising vitamin D and zinc was ineffective. At age 2 yr and 9 mo, the frequency of AA administration was temporarily increased from three to six times per wk to prevent decreased trough levels, without improvement over 3 mo. From age 3 yr and 2 mo, the dose was gradually increased to 9 mg/kg/wk (Fig. 2) based on the United States prescribing guidelines (11). Rickets slightly improved but did not completely resolve (Fig. 2). No apparent adverse events were observed after dose escalation except for lipodystrophy at the injection sites. Around age 4 yr and 6 mo, the ratio of urinary PEA to Cre slightly decreased (Fig. 2). At age 6 yr and 10 mo, NAbs spontaneously became negative, and by age 8 yr and 8 mo, rickets gradually resolved completely (Fig. 2). The dose was reduced to 6 mg/kg per wk at age 9 yr. At age 11 yr, rickets recurrence was not observed (Fig. 2).

Discussion

We encountered a case of HPP with skeletal symptoms that correlated to the presence of NAbs. Although AA was initially effective, rickets concurrently appeared with the emergence of NAbs. We attempted dose escalation of AA to alleviate worsening findings, which resulted in only partial resolution. However, subsequently, rickets resolution and NAb disappearance simultaneously occurred. One possible explanation is that the presence of NAbs influenced the effectiveness of AA. The optimal AA dose may therefore vary based on NAb expression.

Some studies have documented the emergence of NAbs and their influence on AA treatment for patients with HPP. During a phase 2 study of AA, anti-AA antibodies and NAbs were detected in 88% and 67% of infants with HPP, respectively (6). In another study that included 5-yr follow-up of adolescents and adults with HPP, 23.5% of those with anti-AA antibodies also had NAbs; however, in all patients, the NAb status was negative by the end of the follow-up period (12). According to several studies, the efficacy of AA is not significantly associated with the presence of NAbs (4,5,6, 13); however, the sample sizes of these studies were small. Notably, the proportion of patients with a positive NAb status among AA nonresponders was consistently higher (specifically, 1.5- to 3-times higher) than that among AA responders, although this difference was not always statistically significant (6). An HPP case in which NAb emergence led to a poor response to AA even after dose escalation, and resolved only after initiation of immune tolerance induction (ITI) has been reported (7). In our case, improvement and deterioration of rickets corresponded to the absence and presence of NAbs, respectively. We considered that NAbs may have weakened the efficacy of AA, similar to the previously reported case (7), and that the impact of NAbs may vary depending on their activity and affinity, which may explain the favorable treatment response to AA in some patients with a positive NAb status.

PLP, PL, and PA are key forms of vitamin B6 (14). PLP, a TNSALP substrate, is dephosphorylated to PL (15), but this process is impaired in HPP, resulting in a higher ratio of PLP to PL (15). Akiyama et al. reported, in a study involving more than 100 HPP subjects, that the ratios of serum PLP to PL and to PA clearly separated patients on ERT from those not on ERT (16). A ratio of PLP to PL of less than 4.0 is known as a good indicator of the effect of ERT (15). In our case, upon recurrence of rickets, the ratios of PLP to PL and to PA were 0.76 and 4.83, respectively (reference ranges: 1.5–3.4 and 1.1-6.5, respectively) (16), indicating improved vitamin B6 metabolism; however, the bone findings worsened. Possible explanations for this discrepancy include that serum PLP levels are not always elevated in untreated patients with HPP, even in the perinatal form, and that the ratio of PLP to PL is also within the normal range in some patients with other forms of HPP (16). In our case, PLP, PL, and PA were measured only twice, and no pre-treatment data were available. Although the serum PLP level after the recurrence of rickets was 51.2 nmol/L (reference range: 12.5–132 nmol/L) (16), comparison with pre-treatment levels could have provided further insight. However, we acknowledge that these explanations for the discordance between the biomarkers and radiographic findings remain speculative.

Urinary PEA levels have been notably elevated in patients with perinatal severe HPP (16). In our case, during the NAb-positive period, the ratio of urinary PEA to Cre was slightly elevated compared with the subsequent NAb-negative period; this aligns with findings suggesting that urinary PEA levels can be valuable for monitoring treatment response in patients with HPP (17). The ratio of urinary PEA to Cre may therefore be useful for evaluating clinical efficacy, even in treatment-resistant HPP cases; however, the difference observed in our case was modest.

Although the ratio of urinary Ca to Cre fluctuated, an overall trend was observed over the long term: it was relatively high when the NAb status was positive and rickets worsened. Hypercalciuria was reported to be not useful as a diagnostic biomarker because its extent varies widely (18). Fortunately, renal function was preserved in our case. However, given that nephrocalcinosis and renal calculi can affect renal function, monitoring parameters related to calcium metabolism is important in the management of HPP cases, especially during periods of exacerbation. It remains unclear whether the reduced treatment effect observed in our case was associated with the development of nephrocalcinosis.

Although few cases have been reported, management options for refractory HPP may include dose escalation of AA and ITI. Although Japanese prescribing guidelines do not mention dose escalation options (19), the United States prescribing guidelines permit dose increases up to 9 mg/kg/wk for refractory perinatal and infantile HPP cases (11). One case report suggested that temporary dose escalation to more than 9 mg/kg/wk may be effective when clinical deterioration occurs (20). In our case, dose escalation successfully prevented further deterioration of rickets. Subsequently, NAbs became negative and rickets gradually resolved; therefore, the higher dose of AA was no longer necessary. The finding that NAb disappearance was observed after long-term administration of AA is consistent with that of the previous report (12). These observations suggest that prolonged exposure to AA may have induced immune tolerance. Another potential strategy may be ITI, as described in the previous case report (7). Immunomodulation including ITI has been used for other diseases that are treated with ERT (21, 22). The efficacy of ERT for Pompe disease (23) and Fabry disease (24) has been reduced by the presence of anti-drug antibodies or NAbs, but it can be improved by immunomodulation. Information on management strategies for refractory HPP is limited. In particular, evidence for ITI is currently limited to a single case report (7) and no established therapeutic approach exists.

The optimal dose of AA for HPP is case-dependent and difficult to predict based on the genotype. One case report described patients with the same ALPL genotype who presented with markedly different phenotypes (25). Another report described an HPP case, in which a low dose of 2 to 2.4 mg/kg/wk successfully maintained improvement of rickets, whereas the standard dose of 6 mg/kg/wk caused hyperphosphatemia, which was considered a side effect of AA (26). In contrast, dose escalation was implemented for our case to address deterioration of rickets and resulted in improvement, although the response was not complete. Because NAbs tend to emerge in AA nonresponders (6), and because the optimal AA dose for HPP may vary according to the presence of NAbs, dose escalation of AA may be considered for patients with a poor treatment response and positive NAb status.

Because this is a single-case report, the observed temporal association between the positive NAb status and the reduced treatment efficacy does not establish a causal relationship. We acknowledge, as another limitation, that detailed functional characterization of NAbs is not available. However, considering that few detailed reports of the long-term clinical course of HPP refractory to AA are available, our report provides useful information that may enhance the management of such cases. Further studies are warranted to elucidate the association between the presence of NAbs and therapeutic response to AA as well as inform optimal treatment strategies.

Conclusion

We encountered a case of perinatal HPP that was treated with AA; however, the therapeutic response to AA appeared to be influenced by the presence of NAbs. Further deterioration was successfully prevented by a temporary increase in the dose of AA. Appropriate strategies for other refractory cases may differ and are difficult to predict. Additional reports of refractory HPP cases are essential to guiding clinical decision-making.

Conflict of interests

All authors have no conflicts of interest to disclose.