Monoclonal insulin autoimmune syndrome in pregnancy and postpartum

Abstract

Summary

In insulin autoimmune syndrome (IAS), high levels of anti-insulin antibodies cause hypoglycemic episodes. Pregnancy in patients with IAS raises several concerns. The mother may experience glucose imbalances affecting the mother and fetus, transfer of antibodies may induce glucose disturbances in the fetus, and the antibodies may cause IAS in the newborn child. We present the first case of pregnancy and the postpartum period in monoclonal IAS, discuss pathogenetic mechanisms, and recommend how to handle future cases. A patient with monoclonal IAS was treated successfully with daratumumab to decrease anti-insulin antibodies from 1,840 U/mL (<0.4) to 180 U/mL. Treatment was paused to facilitate pregnancy, and after four months she conceived. Her antibody levels remained stable, and she was euglycemic during her pregnancy. The fetus grew at a normal rate. At delivery, the anti-insulin antibody level in the mother was 74 U/mL. The newborn had a level of 38 U/mL, which disappeared within 21 weeks. The child experienced no abnormal glucose fluctuations. The mother’s IAS relapsed a few days postpartum, with hypoglycemia and anti-insulin antibody levels rising to 946 U/mL. Treatment with daratumumab was resumed efficaciously. The pregnancy went well for the mother, fetus, and child, probably mainly because of the relatively low levels of anti-insulin antibodies. We discuss possible explanations for the recurrence of IAS in the mother postpartum.

Learning points

• Future pregnancies in patients with IAS should be planned.

• Treatment to reduce levels of anti-insulin antibodies before pregnancy is advised.

• We recommend close monitoring of the mother and fetus during pregnancy and of the mother and child postpartum.

Background

Insulin autoimmune syndrome (IAS), also known as Hirata’s disease, is a rare condition in which high levels of anti-insulin antibodies cause spontaneous episodes of hypoglycemia (1, 2). There are two subtypes of the disease, monoclonal IAS and polyclonal IAS (3, 4). We present the first case of monoclonal IAS in pregnancy and the consequences for the mother, fetus, and child.

In the literature, we identified two cases of pregnancy in patients with IAS triggered by carbimazole. In one, the patient had an early miscarriage, which could be attributed to underlying hyperthyroidism (5). In the other case, the patient had IAS six months before pregnancy and recovered from IAS once carbimazole was discontinued and her hyperthyroidism was treated with radioiodine. Her hyperthyroidism relapsed at eight weeks of pregnancy, and, due to the impossibility of reinstating carbimazole, she underwent thyroidectomy. The anti-insulin antibodies were reported as negative at 24 weeks of gestation (6). There are no data on the fetus or child in either of these cases. Both cases share the typical characteristics of polyclonal IAS.

Pregnancy in patients with IAS raises several concerns: the mother may suffer from hyperglycemic and hypoglycemic episodes, which may affect both the mother and fetus; transfer of anti-insulin antibodies from mother to fetus may induce glucose disturbances in the fetus; and the newborn child may suffer from IAS during the first months of life because of the transferred antibodies. We present data related to these concerns in our patient and her fetus and child, and we discuss implications for future similar pregnancies.

Case presentation

The diagnosis and treatment of a young female patient with monoclonal IAS have been described previously (3, 4). She presented with low glucose values of 1.5 mmol/L and was diagnosed with IAS after finding anti-insulin antibodies of 5,305 U/mL (<0.4), S-insulin > 6,945 pmol/L (fasting < 160), and C-peptide 4,042 pmol/L (fasting 300–1,480), and ruling out other causes of hypoglycemia. Treatments reported in earlier cases were unsuccessful, and she eventually went into remission when treated with daratumumab, a monoclonal IgG₁ kappa antibody targeting CD38 on plasma cells.

Treatment with daratumumab is not recommended during pregnancy. IgG antibodies pass the blood–placental barrier, and daratumumab was therefore paused preconception. Her antibody levels were low but still present at titers well above normal. At discontinuation, she was symptom-free and remained so when she tested positive for pregnancy four months later. Her anti-insulin antibody levels before pregnancy, during pregnancy, and after delivery are illustrated in Fig. 1 and Table 1. The pregnancy was unremarkable, with no hypoglycemia (lowest measured value of 4.2 mmol/L and highest 6.4 mmol/L). She wore a continuous glucose monitor (CGM) for parts of the pregnancy. Due to many false low alarms, for which glucometer measurements were within the normal range, she opted to assess her blood glucose manually only if she felt unwell.

Figure 1.

Patient’s anti-insulin antibodies during and after pregnancy. The point marked as week 0 was taken an estimated three days before the point of conception.

Table 1.

Mother’s anti-insulin antibody development. The table shows a marked reduction in anti-insulin antibody titers with daratumumab treatment, low levels during pregnancy, relapse postpartum, and a subsequent decline following reinstatement of daratumumab treatment.

Time (description)	Antibody titer (<0.4 U/mL)
Pre-daratumumab	1,221
Preconception	160
Gestational week 4	200
Gestational week 12	161
Gestational week 18	157
Gestational week 24	97
Gestational week 29	86
Postpartum day 3	74
Postpartum day 14 (time of relapse)	946
Daratumumab 4 weeks	118
Daratumumab 20 weeks	15

The fetus grew at a normal rate on the 50th percentile and had no abnormalities on ultrasounds throughout gestation. We had to prepare for the possibility of the child having significant glucose fluctuations after birth with fulminant IAS. The birth was therefore scheduled at a tertiary hospital, with preparations being made for possible neonatal intensive care needs.

The baby (female) was born at 38 + 1 weeks by caesarean section due to breach presentation, with a birthweight of 3,664 g (90th percentile), and a head circumference of 36 cm (96th percentile). The length at birth was not measured due to the breech position at birth. Apgar scores were 9 at 1 min and 10 at 5 min. Umbilical cord blood was taken at birth to screen for anti-insulin antibodies. These were present at delivery at 38 U/mL. The initial blood glucose level (taken at approximately 1 h after birth) was 2.0 mmol/L. The blood glucose was measured, and the baby was breast-fed every 3 h during the first couple of days, in concordance with Norwegian guidelines on neonatal hypoglycemia. The subsequent blood glucose values were all above the threshold level (and the majority were above 3 mmol/L) during the three-day observation period after delivery.

A few days after delivery, the patient started to experience symptoms of hypoglycemia, with glucose values below 3 mmol/L 3–4 times daily. The insulin-antibody titers increased from 74 U/mL at delivery to 946 U/mL 14 days postpartum. Daratumumab 1,800 mg subcutaneously weekly was reinstated shortly after the recurrence of symptomatic hypoglycemia, and was thereafter administered every 2 weeks. She ceased breast-feeding the baby with the initiation of daratumumab because transfer of daratumumab to the mother’s milk is unknown. The patient was in remission again after two months of treatment. After seven cycles, the interval was reduced to every 4 weeks, and she remains in remission 24 months after delivery, with the goal of having another pregnancy in the near future.

The baby remained unaffected. The baby was given a CGM (a Freestyle Libre 3), but due to issues in such small babies with adhesion and inappropriate readings and alarms, this was discontinued. The patient monitored the baby’s glucose by blood glucose measurements from the heel if the child seemed discontent or less agile. There was no measured hypoglycemia. The development of the child’s anti-insulin antibodies is listed in Table 2. The child has followed its own growth curve at the 90th percentile for height and above the 97th percentile for weight during the first 7 months of life.

Table 2.

The child’s anti-insulin antibody development. The table shows complete clearance of the child’s anti-insulin antibodies within 21 weeks.

Time (days from delivery)	Antibody titer (<0.4 U/mL)
0	38
9	41
14	>50
44	16
77	6.7
147	0

Discussion

There are no previous reports on monoclonal IAS in pregnancy. The novelty of this situation warranted many theoretical considerations both preconception and during pregnancy and postpartum for the mother, fetus, and child. The patient’s wish to become pregnant was major. Daratumumab, being an IgG kappa antibody, is not recommended during pregnancy because it will pass the placental barrier and likely reduce CD38+ plasma cells in the fetus and possibly other fetal cells expressing CD38. It was therefore discontinued. We aimed for close follow-up after discontinuation and throughout the pregnancy and postpartum period.

Our main concern for the mother was relapse of IAS before getting pregnant, which would have made pregnancy difficult to recommend, and relapse during pregnancy, mainly because of the effects this might have on the fetus. Her anti-insulin antibody titers were lower than they had been during the full course of her illness, but they were still present. They were not high enough to give the patient symptoms but possibly high enough to give the fetus and child symptoms. With a potential relapse with major blood glucose fluctuations, we planned to use conservative measures (close monitoring by CGM and small, frequent meals) primarily and, if needed, prednisolone, as the effects on the fetus are known and would be manageable.

Pregnancy tilts the maternal metabolism in the direction of insulin resistance. It is possible that the fetus can, to some extent, buffer the maternal glucose fluctuations by receiving glucose from the mother when she is hyperglycemic and giving glucose to the mother when she is hypoglycemic (7). Our patient was in remission with stable blood glucose values before pregnancy, and she remained so during pregnancy, which led us to assume that the fetus would not be exposed to major fluctuations from her.

With regard to the fetus, we had many further concerns. A fetus has endogenous insulin production, which starts around the end of the first trimester. Insulin does not pass the blood-placental barrier, but glucose passes easily from mother to fetus. Therefore, the fetus will have the same blood glucose as the mother at any given time in a normal pregnancy.

The fetus was monitored with ultrasounds during pregnancy because we were concerned about its growth and whether it could possibly be large for gestational age due to exposure to transient hyperglycemia in the womb. Transient or mild hypoglycemia has not been shown in human studies to be associated with fetal embryopathy, although severe- or prolonged hypoglycemia in the mother can affect fetal heart rate and fetal movements (8).

We had to assume that the fetus would have the same anti-insulin antibody values as the mother in late pregnancy and that it could, in theory, have IAS. If the mother had very stable glucose values (as she did), the fetus would not be receiving the stimuli required for a major release of insulin. One could assume that the antibodies present would also have roughly the same amount of insulin bound to them at any given time and that there would not be a significant release of insulin with resulting hypoglycemia. On the other hand, if the mother had large glucose fluctuations, the fetus would have roughly the same fluctuations as the mother and, with its anti-insulin antibody positivity, could also develop its own IAS fluctuations.

We believe the presence of pathogenic maternal IgG antibodies to be the most important factor determining whether the fetus or child will have clinical symptoms. Immunoglobulin G (IgG) antibodies constitute the only antibody class to pass over the placental barrier in any significant way (9), and transport starts as early as 13 weeks of gestation and peaks in the third trimester. At delivery, the concentration of IgG is similar in the fetus and the mother. Cord blood IgG has a good correlation with maternal levels, as demonstrated in our case. The maternal levels in our case were too low to give the mother hypoglycemic fits, and similarly low levels in the child could be the main reason that glucose imbalance did not occur. This is why we recommend reducing maternal anti-insulin antibodies preconception.

There are other factors at play than antibody levels. In mouse models of autoimmune skin blistering diseases, administering large amounts of exogenous IgG (IVIG) to saturate the FcRn receptor (the neonatal receptor responsible for the specific transfer of IgG from the maternal circulation into the fetal blood supply) and to dilute the pathogenic antibodies has had a positive effect (10). Our patient had some cycles of IVIG as part of her treatment regimen before pregnancy, but the last dose was given a year before conception and is therefore likely not relevant in this case but could be a potential treatment option in other cases.

The gender of the baby, how the placental absorption of maternal antibodies functions, the presence of blocking maternal antibodies, antigen density and distribution in the fetal cells and tissues (in children born to mothers with SLE, the density of binding sites in the fetal heart determines if the baby will have congenital heart block or not), the specificity of the maternal antibody, and factors relating to the infant’s immune system and its synthesis of proteins also contribute (11). Some of these factors may explain why the baby had no hypoglycemia at birth, despite anti-insulin antibodies being present. Most of these factors will not be modifiable and therefore not of practical importance in the management of these patients.

The definition of hypoglycemia requiring treatment in newborns, according to Norwegian national guidelines, is as follows: blood glucose <1.5 mmol/L in babies 0–4 h of age; <2.0 mmol/L in babies 4–24 h of age; and <2.6 mmol/L in babies >24 h of age. If the blood glucose level is >2 mmol/L (2.0–2.6 mmol/L) and the baby shows no symptoms of hypoglycemia, it is recommended to provide breast milk or infant formula every 2–3 h and monitor blood glucose levels until stabilized at >2.6 mmol/L. With the initial value of 2 mmol/L at birth, the glucose values were measured every 3 h, and after 24 h, every second meal. The baby’s anti-insulin antibodies cleared at a rate that could be comparable to the clearance of other known fetal IgG antibody syndromes (12). Usually, the majority of IgG antibodies clear within 3 months and totally by 6 months. After 21 weeks in this case, the values were normal, which is in line with the experience from other antibody syndromes.

As transfer of daratumumab to the mother’s milk is unknown, the patient ceased breast-feeding. There are, however, data on rituximab (another IgG1 monoclonal antibody) showing low concentrations in mother’s milk (<0.4%) (13), and antibodies in mother’s milk will be degraded in the gastrointestinal tract of the baby (as with any other protein), which indicates that breast-feeding could have been continued.

Postpartum, the mother had a rapid rise in anti-insulin antibodies and recurrence of symptoms. This puts into question her monoclonality, which has been championed in previous articles (3, 4). On the one hand, one could argue that the recurrence of her antibodies and symptoms postpartum is more compatible with polyclonal autoimmunity. In such diseases, for example, autoimmune thyroiditis, Graves’ disease, or multiple sclerosis, termination of the immunosuppressive effect of pregnancy probably triggers recurrence (14). On the other hand, the clinical and laboratory findings indicating that the patient has monoclonal IAS are also strong: a long disease duration without spontaneous remission; no use of drugs or presence of disease often associated with polyclonal IAS; a HLA type different from those typical for polyclonal IAS; no effect of anti-lymphocyte treatment, while plasma cell-directed treatment proved effective; and anti-insulin antibodies restricted to IgG kappa, indicating monoclonality. A few patients with plasma cell neoplasia have been reported to show no significant progression during pregnancy but to progress shortly after delivery (15).

We cannot exclude small heterogeneities of the antibodies, which could lead to the term oligoclonal IAS. However, small heterogeneities of the antibodies also occur during the course of the disease in patients with a typical monoclonal start due to mutations over time in immunoglobulin genes. Still, the disease should be considered monoclonal.

The bulk of evidence (the IgG kappa restriction of the antibodies, and the lack of clinical and laboratory characteristics of patients with polyclonal IAS described above) indicates that our patient has a syndrome quite different from polyclonal IAS (3, 4, 16). We therefore maintain the term monoclonal IAS, even if we cannot exclude small antibody heterogeneities.

In our case, the pregnancy went well for the mother, fetus, and child. We recommend that patients with IAS plan their pregnancies in collaboration with their health care professionals and that both the mother and fetus are monitored closely during pregnancy. We believe it is important to treat IAS to reduce the level of anti-insulin antibodies as much as possible before the start of pregnancy. We also recommend close monitoring of the child at birth and measurement of anti-insulin antibodies at the time of birth to assess the child’s risk of hypoglycemia. One should also be aware of the risk of IAS relapse after delivery. It is reasonable to assume that these recommendations are also of value for pregnancies in patients with polyclonal IAS.

Declaration of interest

The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.

Funding

This research did not receive any specific grant from any funding agency in the public, commercial, or not-for-profit sector.

Patient consent

Written informed consent for the publication of the clinical details was obtained from the patient on behalf of herself and her daughter.

Author contribution statement

ACPW developed the concept, drafted the original manuscript, and treated the patient. HDM treated the child postpartum, drafted a section of the discussion in the original manuscript, and reviewed the manuscript. GET was involved in the treatment of the patient and reviewed the manuscript. IF developed the concept, drafted sections of the discussion in the original manuscript, and reviewed the original manuscript.