How 100 years of insulin have changed pregnancy for women with type 1 diabetes

Mark B. Landon, MD: The discovery of insulin in 1921 by Dr. Frederick Banting and Dr. Charles Best and its introduction into clinical practice may well be the most significant achievement in the care of pregnant women with diabetes mellitus in the last century. Why was this advance so monumental?

Steven G. Gabbe, MD: Insulin is the single most important drug we use in taking care of diabetes in pregnancy. It is required not only by all patients with type 1 diabetes, but also by the majority of patients with type 2 diabetes. Moreover, at least a third of our patients with gestational diabetes require more than lifestyle change. The American College of Obstetricians and Gynecologists and the American Diabetes Association recommend that insulin be considered as the first-line pharmacologic therapy.

Courtesy Dr. Patricia Gabbe

Before insulin, the most prudent option for women who had glucose in their urine early in pregnancy, which was called “true diabetes,” was deemed to be termination of the pregnancy. The chances of surviving a pregnancy, and of having a surviving infant, were low.

Pregnancies were a rarity to begin with because most women of reproductive age died within a year or two of the onset of their illness. Moreover, most women with what we now know as type 1 diabetes were amenorrheic and infertile. In fact, before insulin, there were few cases of pregnancy complicated by diabetes reported in the literature. A summary of the world literature published in 1909 in the American Journal of the Medical Sciences reported: 66 pregnancies in 43 women; 50% maternal mortality (27% immediate; 23% in next 2 years); and a 41% pregnancy loss (Obstet Gynecol. 1992;79:295-9, Cited Am J Med Sci. 1909;137:1).

The first injection of insulin was administered in 1922 to a 13-year-old Canadian boy, and for several years the focus was on children. (Some of them had been kept alive with 450 calories/day long enough to benefit from the new treatment.)

For women with what we now know as type 1 diabetes, insulin kept them alive, restored their fertility, and enabled them to survive a pregnancy. Maternal mortality dropped dramatically, down to a few percent, once pregnant women became beneficiaries of insulin therapy.

Perinatal outcomes remained poor, however. In the early years of insulin therapy, more than half of the babies died. Some were stillbirths, which had been the primary cause of perinatal deaths in the pre-insulin era. Others were spontaneous preterm births, and still others were delivered prematurely in order to avert a stillbirth, and subsequently died.

Dr. Landon: A significant improvement in perinatal outcomes was eventually realized about two decades after insulin was introduced. By then Dr. Priscilla White of the Joslin Clinic had recorded that women who had so-called ‘normal hormonal balance’ – basically good glucose control – had very low rates of fetal demise and fetal loss compared with those who did not have good control. You had the opportunity to work alongside Dr. White. How did she achieve these results without all the tools we have today?

Dr. Gabbe: In 1925, the perinatal mortality in pregnancies complicated by type 1 diabetes was about 40%. By 1965 it was 10%, and when I began my residency at the Joslin Clinic and Boston Hospital for Women in 1972 it was closer to 5%

In those days we didn’t have accurate methods for dating pregnancies or assessing fetal size or well-being. We didn’t have tools to monitor blood glucose levels, and our insulins were limited to regular insulins and NPH (neutral protamine Hagedorn) as a basal insulin.

Dr. White had concluded early on, and wrote in a 1928 paper, that controlling diabetes was essential to fetal welfare and that the “high glucose content of placental blood” was probably linked to excessive fetal growth. She also wrote about the importance of “close and persistent supervision” of the patient by both an internist and obstetrician.

When I began working with her in the 1970s, her program involved antepartum visits every week or two and a team approach. Patients would be seen by Dr. White and other diabetologists, by head obstetrician Dr. Luke Gillespie, and by nurses and nutritionists. At the end of each day, after all the patients had been seen, we’d gather in Dr. White’s office and look at each patient’s single morning blood glucose measurement and the histories we’d obtained, and we’d make adjustments to their insulin regimens.

Dr. White’s solution to the problem of monitoring blood glucose was a program of hospitalization throughout pregnancy. Patients were hospitalized for a week initially to achieve blood glucose control, and then again around 20 weeks of gestation for monitoring and improvement. Hospitalizations later in the pregnancy were timed according to her classification of obstetric diabetes, which had been published in a landmark paper in 1949. In that paper Dr. Priscilla White wrote: “It is evident that age at onset of diabetes, duration, severity, and degree of maternal vascular disease all influence the fetal survival unfavorably”(Obstet Gynecol. 1992;79:295-9 / Am J Med. 1949;7:609-16).

The classification system considered age of onset, duration of diabetes, need for insulin, and presence of vascular disease. Women in higher classes and at greater risk for intrauterine death were admitted at 32 weeks, while those at less risk could wait until about 34 weeks. The timing of delivery was somewhat arbitrary, but the goal was to choose a time at which the fetus could survive in the nursery and, as Dr. White had written, “before the dreaded late intrauterine accident could occur.” (In the early ’70s, approximately half of newborns admitted to [newborn intensive care unites] at 32 weeks would survive.)

We did measure estriol levels through 24-hour urine collections as a marker for fetal and placental well-being, but as we subsequently learned, a sharp drop was often too late an indicator that something was wrong.

Dr. Landon: Dr. White and others trying to manage diabetes in pregnancy during the initial decades after insulin’s discovery were indeed significantly handicapped by a lack of tools for assessing glucose control. However, the 1970s then ushered in a “Golden Era” of fetal testing. How did advances in antepartum fetal monitoring complement the use of insulin?

Dr. Gabbe: By the mid-1970s, researchers had recognized that fetal heart rate decelerations in labor signaled fetal hypoxemia, and Dr. Roger Freeman had applied these findings to the antepartum setting, pioneering development of the contraction stress test, or oxytocin stress test. The absence of late decelerations during 10 minutes of contractions meant that the fetus was unlikely to be compromised.

 

When the test was administered to high-risk patients at Los Angeles County Women’s Hospital, including women with diabetes, a negative result predicted that a baby would not die within the next week. The contraction stress test was a major breakthrough. It was the first biophysical test for fetal compromise and was important for pregnancies complicated by diabetes. However, it had to be done on the labor and delivery floor, it could take hours, and it might not be definitive if one couldn’t produce enough contractions.

In the mid-1970s, the nonstress test, which relied on the presence of fetal heart rate accelerations in response to fetal movement, was found to be as reliable as the contraction stress test. It became another important tool for prolonging gestation in women with type 1 diabetes.

Even more predictive and reliable was the biophysical profile described several years later. It combined the nonstress test with an assessment using real-time fetal ultrasound of fetal movements, fetal tone and breathing movements, and amniotic fluid.

So, in a relatively short period of time, antepartum surveillance progressed from the contraction stress test to the nonstress test to the biophysical profile. These advances, along with advances in neonatal intensive care, all contributed to the continued decline in perinatal mortality.

Dr. Landon: You have taught for many years that the principal benefit of these tests of fetal surveillance is not necessarily the results identifying a fetus at risk, but the reassuring normal results that allow further maturation of the fetus that is not at risk in the pregnancy complicated by type 1 diabetes.

You also taught – as I experienced some 40 years ago when training with you at the University of Pennsylvania – that hospitalization later in pregnancy allowed for valuable optimization of our patients’ insulin regimens prior to their scheduled deliveries. This optimization helped to reduce complications such as neonatal hypoglycemia.

The introduction of the first reflectance meters to the antepartum unit eliminated the need for so many blood draws. Subsequently, came portable self-monitoring blood glucose units, which I’d argue were the second greatest achievement after the introduction of insulin because they eliminated the need for routine antepartum admissions. What are your thoughts?

Dr. Gabbe: The reflectance meters as first developed were in-hospital devices. They needed frequent calibration, and readings took several minutes. Once introduced, however, there was rapid advancement in their accuracy, size, and speed of providing results.

Other important advances were the development of rapid-acting insulins and new basal insulins and, in the late 1980s and early 1990s, the development of insulin pumps. At Penn, we studied an early pump that we called the “blue brick” because of its size. Today, of course, smaller and safer pumps paired with continuous glucose monitors are making an enormous difference for our patients with type 1 diabetes, providing them with much better outcomes.

Dr. Landon: A century after the discovery of insulin, congenital malformations remain a problem. We have seen a reduction overall, but recent data here and in Sweden show that the rate of malformations in pregnancy complicated by diabetes still is several-fold greater than in the general population.

The data also support what we’ve known for decades – that the level of glucose control during the periconceptual period is directly correlated with the risk of malformations. Can you speak to our efforts, which have been somewhat, but not completely, successful?

Dr. Gabbe: This is one of our remaining challenges. Malformations are now the leading cause of perinatal mortality in pregnancies involving type 1 and type 2 diabetes. We’ve seen these tragic outcomes over the years. While there were always questions about what caused malformations, our concerns focused on hyperglycemia early in pregnancy as a risk factor.

Knowing now that it is an abnormal intrauterine milieu during the period of organogenesis that leads to the malformations, we have improved by having patients come to us before pregnancy. Studies have shown that we can reduce malformations to a level comparable to the general population, or perhaps a bit higher, through intensive control as a result of prepregnancy care.

The challenge is that many obstetric patients don’t have a planned pregnancy. Our efforts to improve glucose control don’t always go the way we’d like them to. Still, considering where we’ve come from since the introduction of insulin to the modern management of diabetes in pregnancy, our progress has been truly remarkable.

Mark B. Landon, MD: The discovery of insulin in 1921 by Dr. Frederick Banting and Dr. Charles Best and its introduction into clinical practice may well be the most significant achievement in the care of pregnant women with diabetes mellitus in the last century. Why was this advance so monumental?

Steven G. Gabbe, MD: Insulin is the single most important drug we use in taking care of diabetes in pregnancy. It is required not only by all patients with type 1 diabetes, but also by the majority of patients with type 2 diabetes. Moreover, at least a third of our patients with gestational diabetes require more than lifestyle change. The American College of Obstetricians and Gynecologists and the American Diabetes Association recommend that insulin be considered as the first-line pharmacologic therapy.

Courtesy Dr. Patricia Gabbe

Before insulin, the most prudent option for women who had glucose in their urine early in pregnancy, which was called “true diabetes,” was deemed to be termination of the pregnancy. The chances of surviving a pregnancy, and of having a surviving infant, were low.

Pregnancies were a rarity to begin with because most women of reproductive age died within a year or two of the onset of their illness. Moreover, most women with what we now know as type 1 diabetes were amenorrheic and infertile. In fact, before insulin, there were few cases of pregnancy complicated by diabetes reported in the literature. A summary of the world literature published in 1909 in the American Journal of the Medical Sciences reported: 66 pregnancies in 43 women; 50% maternal mortality (27% immediate; 23% in next 2 years); and a 41% pregnancy loss (Obstet Gynecol. 1992;79:295-9, Cited Am J Med Sci. 1909;137:1).

The first injection of insulin was administered in 1922 to a 13-year-old Canadian boy, and for several years the focus was on children. (Some of them had been kept alive with 450 calories/day long enough to benefit from the new treatment.)

For women with what we now know as type 1 diabetes, insulin kept them alive, restored their fertility, and enabled them to survive a pregnancy. Maternal mortality dropped dramatically, down to a few percent, once pregnant women became beneficiaries of insulin therapy.

Perinatal outcomes remained poor, however. In the early years of insulin therapy, more than half of the babies died. Some were stillbirths, which had been the primary cause of perinatal deaths in the pre-insulin era. Others were spontaneous preterm births, and still others were delivered prematurely in order to avert a stillbirth, and subsequently died.

Dr. Landon: A significant improvement in perinatal outcomes was eventually realized about two decades after insulin was introduced. By then Dr. Priscilla White of the Joslin Clinic had recorded that women who had so-called ‘normal hormonal balance’ – basically good glucose control – had very low rates of fetal demise and fetal loss compared with those who did not have good control. You had the opportunity to work alongside Dr. White. How did she achieve these results without all the tools we have today?

Dr. Gabbe: In 1925, the perinatal mortality in pregnancies complicated by type 1 diabetes was about 40%. By 1965 it was 10%, and when I began my residency at the Joslin Clinic and Boston Hospital for Women in 1972 it was closer to 5%

In those days we didn’t have accurate methods for dating pregnancies or assessing fetal size or well-being. We didn’t have tools to monitor blood glucose levels, and our insulins were limited to regular insulins and NPH (neutral protamine Hagedorn) as a basal insulin.

Dr. White had concluded early on, and wrote in a 1928 paper, that controlling diabetes was essential to fetal welfare and that the “high glucose content of placental blood” was probably linked to excessive fetal growth. She also wrote about the importance of “close and persistent supervision” of the patient by both an internist and obstetrician.

When I began working with her in the 1970s, her program involved antepartum visits every week or two and a team approach. Patients would be seen by Dr. White and other diabetologists, by head obstetrician Dr. Luke Gillespie, and by nurses and nutritionists. At the end of each day, after all the patients had been seen, we’d gather in Dr. White’s office and look at each patient’s single morning blood glucose measurement and the histories we’d obtained, and we’d make adjustments to their insulin regimens.

Dr. White’s solution to the problem of monitoring blood glucose was a program of hospitalization throughout pregnancy. Patients were hospitalized for a week initially to achieve blood glucose control, and then again around 20 weeks of gestation for monitoring and improvement. Hospitalizations later in the pregnancy were timed according to her classification of obstetric diabetes, which had been published in a landmark paper in 1949. In that paper Dr. Priscilla White wrote: “It is evident that age at onset of diabetes, duration, severity, and degree of maternal vascular disease all influence the fetal survival unfavorably”(Obstet Gynecol. 1992;79:295-9 / Am J Med. 1949;7:609-16).

The classification system considered age of onset, duration of diabetes, need for insulin, and presence of vascular disease. Women in higher classes and at greater risk for intrauterine death were admitted at 32 weeks, while those at less risk could wait until about 34 weeks. The timing of delivery was somewhat arbitrary, but the goal was to choose a time at which the fetus could survive in the nursery and, as Dr. White had written, “before the dreaded late intrauterine accident could occur.” (In the early ’70s, approximately half of newborns admitted to [newborn intensive care unites] at 32 weeks would survive.)

We did measure estriol levels through 24-hour urine collections as a marker for fetal and placental well-being, but as we subsequently learned, a sharp drop was often too late an indicator that something was wrong.

Dr. Landon: Dr. White and others trying to manage diabetes in pregnancy during the initial decades after insulin’s discovery were indeed significantly handicapped by a lack of tools for assessing glucose control. However, the 1970s then ushered in a “Golden Era” of fetal testing. How did advances in antepartum fetal monitoring complement the use of insulin?

Dr. Gabbe: By the mid-1970s, researchers had recognized that fetal heart rate decelerations in labor signaled fetal hypoxemia, and Dr. Roger Freeman had applied these findings to the antepartum setting, pioneering development of the contraction stress test, or oxytocin stress test. The absence of late decelerations during 10 minutes of contractions meant that the fetus was unlikely to be compromised.

 

When the test was administered to high-risk patients at Los Angeles County Women’s Hospital, including women with diabetes, a negative result predicted that a baby would not die within the next week. The contraction stress test was a major breakthrough. It was the first biophysical test for fetal compromise and was important for pregnancies complicated by diabetes. However, it had to be done on the labor and delivery floor, it could take hours, and it might not be definitive if one couldn’t produce enough contractions.

In the mid-1970s, the nonstress test, which relied on the presence of fetal heart rate accelerations in response to fetal movement, was found to be as reliable as the contraction stress test. It became another important tool for prolonging gestation in women with type 1 diabetes.

Even more predictive and reliable was the biophysical profile described several years later. It combined the nonstress test with an assessment using real-time fetal ultrasound of fetal movements, fetal tone and breathing movements, and amniotic fluid.

So, in a relatively short period of time, antepartum surveillance progressed from the contraction stress test to the nonstress test to the biophysical profile. These advances, along with advances in neonatal intensive care, all contributed to the continued decline in perinatal mortality.

Dr. Landon: You have taught for many years that the principal benefit of these tests of fetal surveillance is not necessarily the results identifying a fetus at risk, but the reassuring normal results that allow further maturation of the fetus that is not at risk in the pregnancy complicated by type 1 diabetes.

You also taught – as I experienced some 40 years ago when training with you at the University of Pennsylvania – that hospitalization later in pregnancy allowed for valuable optimization of our patients’ insulin regimens prior to their scheduled deliveries. This optimization helped to reduce complications such as neonatal hypoglycemia.

The introduction of the first reflectance meters to the antepartum unit eliminated the need for so many blood draws. Subsequently, came portable self-monitoring blood glucose units, which I’d argue were the second greatest achievement after the introduction of insulin because they eliminated the need for routine antepartum admissions. What are your thoughts?

Dr. Gabbe: The reflectance meters as first developed were in-hospital devices. They needed frequent calibration, and readings took several minutes. Once introduced, however, there was rapid advancement in their accuracy, size, and speed of providing results.

Other important advances were the development of rapid-acting insulins and new basal insulins and, in the late 1980s and early 1990s, the development of insulin pumps. At Penn, we studied an early pump that we called the “blue brick” because of its size. Today, of course, smaller and safer pumps paired with continuous glucose monitors are making an enormous difference for our patients with type 1 diabetes, providing them with much better outcomes.

Dr. Landon: A century after the discovery of insulin, congenital malformations remain a problem. We have seen a reduction overall, but recent data here and in Sweden show that the rate of malformations in pregnancy complicated by diabetes still is several-fold greater than in the general population.

The data also support what we’ve known for decades – that the level of glucose control during the periconceptual period is directly correlated with the risk of malformations. Can you speak to our efforts, which have been somewhat, but not completely, successful?

Dr. Gabbe: This is one of our remaining challenges. Malformations are now the leading cause of perinatal mortality in pregnancies involving type 1 and type 2 diabetes. We’ve seen these tragic outcomes over the years. While there were always questions about what caused malformations, our concerns focused on hyperglycemia early in pregnancy as a risk factor.

Knowing now that it is an abnormal intrauterine milieu during the period of organogenesis that leads to the malformations, we have improved by having patients come to us before pregnancy. Studies have shown that we can reduce malformations to a level comparable to the general population, or perhaps a bit higher, through intensive control as a result of prepregnancy care.

The challenge is that many obstetric patients don’t have a planned pregnancy. Our efforts to improve glucose control don’t always go the way we’d like them to. Still, considering where we’ve come from since the introduction of insulin to the modern management of diabetes in pregnancy, our progress has been truly remarkable.

Mark B. Landon, MD: The discovery of insulin in 1921 by Dr. Frederick Banting and Dr. Charles Best and its introduction into clinical practice may well be the most significant achievement in the care of pregnant women with diabetes mellitus in the last century. Why was this advance so monumental?

Steven G. Gabbe, MD: Insulin is the single most important drug we use in taking care of diabetes in pregnancy. It is required not only by all patients with type 1 diabetes, but also by the majority of patients with type 2 diabetes. Moreover, at least a third of our patients with gestational diabetes require more than lifestyle change. The American College of Obstetricians and Gynecologists and the American Diabetes Association recommend that insulin be considered as the first-line pharmacologic therapy.

Courtesy Dr. Patricia Gabbe

Before insulin, the most prudent option for women who had glucose in their urine early in pregnancy, which was called “true diabetes,” was deemed to be termination of the pregnancy. The chances of surviving a pregnancy, and of having a surviving infant, were low.

Pregnancies were a rarity to begin with because most women of reproductive age died within a year or two of the onset of their illness. Moreover, most women with what we now know as type 1 diabetes were amenorrheic and infertile. In fact, before insulin, there were few cases of pregnancy complicated by diabetes reported in the literature. A summary of the world literature published in 1909 in the American Journal of the Medical Sciences reported: 66 pregnancies in 43 women; 50% maternal mortality (27% immediate; 23% in next 2 years); and a 41% pregnancy loss (Obstet Gynecol. 1992;79:295-9, Cited Am J Med Sci. 1909;137:1).

The first injection of insulin was administered in 1922 to a 13-year-old Canadian boy, and for several years the focus was on children. (Some of them had been kept alive with 450 calories/day long enough to benefit from the new treatment.)

For women with what we now know as type 1 diabetes, insulin kept them alive, restored their fertility, and enabled them to survive a pregnancy. Maternal mortality dropped dramatically, down to a few percent, once pregnant women became beneficiaries of insulin therapy.

Perinatal outcomes remained poor, however. In the early years of insulin therapy, more than half of the babies died. Some were stillbirths, which had been the primary cause of perinatal deaths in the pre-insulin era. Others were spontaneous preterm births, and still others were delivered prematurely in order to avert a stillbirth, and subsequently died.

Dr. Landon: A significant improvement in perinatal outcomes was eventually realized about two decades after insulin was introduced. By then Dr. Priscilla White of the Joslin Clinic had recorded that women who had so-called ‘normal hormonal balance’ – basically good glucose control – had very low rates of fetal demise and fetal loss compared with those who did not have good control. You had the opportunity to work alongside Dr. White. How did she achieve these results without all the tools we have today?

Dr. Gabbe: In 1925, the perinatal mortality in pregnancies complicated by type 1 diabetes was about 40%. By 1965 it was 10%, and when I began my residency at the Joslin Clinic and Boston Hospital for Women in 1972 it was closer to 5%

In those days we didn’t have accurate methods for dating pregnancies or assessing fetal size or well-being. We didn’t have tools to monitor blood glucose levels, and our insulins were limited to regular insulins and NPH (neutral protamine Hagedorn) as a basal insulin.

Dr. White had concluded early on, and wrote in a 1928 paper, that controlling diabetes was essential to fetal welfare and that the “high glucose content of placental blood” was probably linked to excessive fetal growth. She also wrote about the importance of “close and persistent supervision” of the patient by both an internist and obstetrician.

When I began working with her in the 1970s, her program involved antepartum visits every week or two and a team approach. Patients would be seen by Dr. White and other diabetologists, by head obstetrician Dr. Luke Gillespie, and by nurses and nutritionists. At the end of each day, after all the patients had been seen, we’d gather in Dr. White’s office and look at each patient’s single morning blood glucose measurement and the histories we’d obtained, and we’d make adjustments to their insulin regimens.

Dr. White’s solution to the problem of monitoring blood glucose was a program of hospitalization throughout pregnancy. Patients were hospitalized for a week initially to achieve blood glucose control, and then again around 20 weeks of gestation for monitoring and improvement. Hospitalizations later in the pregnancy were timed according to her classification of obstetric diabetes, which had been published in a landmark paper in 1949. In that paper Dr. Priscilla White wrote: “It is evident that age at onset of diabetes, duration, severity, and degree of maternal vascular disease all influence the fetal survival unfavorably”(Obstet Gynecol. 1992;79:295-9 / Am J Med. 1949;7:609-16).

The classification system considered age of onset, duration of diabetes, need for insulin, and presence of vascular disease. Women in higher classes and at greater risk for intrauterine death were admitted at 32 weeks, while those at less risk could wait until about 34 weeks. The timing of delivery was somewhat arbitrary, but the goal was to choose a time at which the fetus could survive in the nursery and, as Dr. White had written, “before the dreaded late intrauterine accident could occur.” (In the early ’70s, approximately half of newborns admitted to [newborn intensive care unites] at 32 weeks would survive.)

We did measure estriol levels through 24-hour urine collections as a marker for fetal and placental well-being, but as we subsequently learned, a sharp drop was often too late an indicator that something was wrong.

Dr. Landon: Dr. White and others trying to manage diabetes in pregnancy during the initial decades after insulin’s discovery were indeed significantly handicapped by a lack of tools for assessing glucose control. However, the 1970s then ushered in a “Golden Era” of fetal testing. How did advances in antepartum fetal monitoring complement the use of insulin?

Dr. Gabbe: By the mid-1970s, researchers had recognized that fetal heart rate decelerations in labor signaled fetal hypoxemia, and Dr. Roger Freeman had applied these findings to the antepartum setting, pioneering development of the contraction stress test, or oxytocin stress test. The absence of late decelerations during 10 minutes of contractions meant that the fetus was unlikely to be compromised.

 

When the test was administered to high-risk patients at Los Angeles County Women’s Hospital, including women with diabetes, a negative result predicted that a baby would not die within the next week. The contraction stress test was a major breakthrough. It was the first biophysical test for fetal compromise and was important for pregnancies complicated by diabetes. However, it had to be done on the labor and delivery floor, it could take hours, and it might not be definitive if one couldn’t produce enough contractions.

In the mid-1970s, the nonstress test, which relied on the presence of fetal heart rate accelerations in response to fetal movement, was found to be as reliable as the contraction stress test. It became another important tool for prolonging gestation in women with type 1 diabetes.

Even more predictive and reliable was the biophysical profile described several years later. It combined the nonstress test with an assessment using real-time fetal ultrasound of fetal movements, fetal tone and breathing movements, and amniotic fluid.

So, in a relatively short period of time, antepartum surveillance progressed from the contraction stress test to the nonstress test to the biophysical profile. These advances, along with advances in neonatal intensive care, all contributed to the continued decline in perinatal mortality.

Dr. Landon: You have taught for many years that the principal benefit of these tests of fetal surveillance is not necessarily the results identifying a fetus at risk, but the reassuring normal results that allow further maturation of the fetus that is not at risk in the pregnancy complicated by type 1 diabetes.

You also taught – as I experienced some 40 years ago when training with you at the University of Pennsylvania – that hospitalization later in pregnancy allowed for valuable optimization of our patients’ insulin regimens prior to their scheduled deliveries. This optimization helped to reduce complications such as neonatal hypoglycemia.

The introduction of the first reflectance meters to the antepartum unit eliminated the need for so many blood draws. Subsequently, came portable self-monitoring blood glucose units, which I’d argue were the second greatest achievement after the introduction of insulin because they eliminated the need for routine antepartum admissions. What are your thoughts?

Dr. Gabbe: The reflectance meters as first developed were in-hospital devices. They needed frequent calibration, and readings took several minutes. Once introduced, however, there was rapid advancement in their accuracy, size, and speed of providing results.

Other important advances were the development of rapid-acting insulins and new basal insulins and, in the late 1980s and early 1990s, the development of insulin pumps. At Penn, we studied an early pump that we called the “blue brick” because of its size. Today, of course, smaller and safer pumps paired with continuous glucose monitors are making an enormous difference for our patients with type 1 diabetes, providing them with much better outcomes.

Dr. Landon: A century after the discovery of insulin, congenital malformations remain a problem. We have seen a reduction overall, but recent data here and in Sweden show that the rate of malformations in pregnancy complicated by diabetes still is several-fold greater than in the general population.

The data also support what we’ve known for decades – that the level of glucose control during the periconceptual period is directly correlated with the risk of malformations. Can you speak to our efforts, which have been somewhat, but not completely, successful?

Dr. Gabbe: This is one of our remaining challenges. Malformations are now the leading cause of perinatal mortality in pregnancies involving type 1 and type 2 diabetes. We’ve seen these tragic outcomes over the years. While there were always questions about what caused malformations, our concerns focused on hyperglycemia early in pregnancy as a risk factor.

Knowing now that it is an abnormal intrauterine milieu during the period of organogenesis that leads to the malformations, we have improved by having patients come to us before pregnancy. Studies have shown that we can reduce malformations to a level comparable to the general population, or perhaps a bit higher, through intensive control as a result of prepregnancy care.

The challenge is that many obstetric patients don’t have a planned pregnancy. Our efforts to improve glucose control don’t always go the way we’d like them to. Still, considering where we’ve come from since the introduction of insulin to the modern management of diabetes in pregnancy, our progress has been truly remarkable.