Insulin lispro and the development of proliferative diabetic retinopathy during pregnancy

Abstract

To the Editors:In 1999 we observed a possible association between change to insulin lispro treatment in early pregnancy and the rarely described development of proliferative diabetic retinopathy (PDR) by the third trimester in 3 women with type 1 or 2 diabetes who had normal retinal examinations at the beginning of pregnancy.1Kitzmiller J Main E Ward B Theiss T Peterson D. Insulin lispro and the development of proliferative diabetic retinopathy during pregnancy.Diabetes Care. 1999; 22 ([letter]): 874-876Crossref PubMed Scopus (73) Google Scholar In your November 2000 issue, Buchbinder et al2Buchbinder A Miodovnik M McElvy S Rosenn B Kranias G Khoury J et al.Is insulin lispro associated with the development or progression of diabetic retinopathy during pregnancy?.Am J Obstet Gynecol. 2000; 183: 1162-1165Abstract Full Text Full Text PDF PubMed Scopus (42) Google Scholar report the results of retinal examinations after pregnancy in 8 women with type 1 diabetes treated with insulin lispro before and during gestation—3 changed to insulin lispro in the first trimester, and 1 changed in the second trimester. Of their total group of 12, six had a normal result at the initial retinal examination before 24 weeks’ gestation, 4 had background diabetic retinopathy, 2 had PDR, and none showed progression at the 6- through 16-week postpartum examination. These results were compared with historical control subjects who used regular insulins. No individual data were reported for the insulin lispro group, but they had elevated glycohemoglobin A1 in the first trimester at 3.1 SDs above the normal mean, and 2 had rapid normalization of glycemic control during pregnancy defined as a decrease from enrollment to second trimester in glycohemoglobin A1 concentrations of > 4 SD. To compare their data with our original observation, it would be helpful if Buchbinder et al would report the details of retinal status, glycemic control, and hypertension in the 3 patients who changed to insulin lispro in the first trimester.Buchbinder et al also reported the second case known to me of a pregnant woman with diabetes who was treated with regular insulins who had a negative initial retinal examination and who progressed to PDR at the postpartum examinations. Their patient had many risk factors for progression, including diabetic nephropathy, chronic hypertension, superimposed preeclampsia, elevated glycohemoglobin in the first trimester, and rapid normalization of glycemic control during pregnancy.In their paper, Buchbinder et al cite the letter of Bhattacharyya and Vice,3Bhattacharyya A Vice PA. Insulin lispro, pregnancy, and retinopathy.Diabetes Care. 1999; 22 ([letter]): 2101-2102Crossref PubMed Scopus (29) Google Scholar which reported no worsening of retinopathy in 16 well-controlled women with type 1 or 2 diabetes mellitus who were treated with insulin lispro in the prepregnancy clinic and who continued the treatment during pregnancy. Buchbinder et al ignored my published response to Bhattacharyya and Vice in which I agreed that the subgroup with poor diabetic control at the onset of pregnancy is at the highest risk of developing PDR during pregnancy, perhaps because they may already have undetected peripheral retinal ischemia.4Kitzmiller J. Response to Bhattacharyya and Vice.Diabetes Care. 1999; 22: 2102-2104Crossref Scopus (86) Google ScholarI also noted that to avoid the trap of anecdotal cases and uncontrolled observations seeming to infer a cause-effect relationship, properly controlled studies should be designed. I proposed (on the basis of the data available in 1999) that 110 women with diabetes with initial negative retinal examinations would be required in each group (insulins with no growth factor homogeneity versus insulin lispro) to have an 80% power of finding a significant difference in development of PDR on the basis of insulin therapy at the.05 level, 1-tailed.4Kitzmiller J. Response to Bhattacharyya and Vice.Diabetes Care. 1999; 22: 2102-2104Crossref Scopus (86) Google ScholarBuchbinder et al calculated that 319 diabetic pregnant subjects would be required in each group to demonstrate a 100% increase in rates of progression of retinopathy from 6% in women treated with regular insulins to 12% in a group treated with insulin lispro (β =.2; α =.05; 1-tailed). A properly controlled trial could be enriched by focusing on pregnant women with diabetes with initial poor glycemic control, but it would have to exclude women with the additional risk factors of nephropathy and hypertension. On the basis of their report of 12 pregnant women treated with insulin lispro, it is unwarranted for Buchbinder et al to suggest that insulin lispro is not associated with the development of PDR during pregnancy. We can only answer the question with a prospective controlled trial. To the Editors:In 1999 we observed a possible association between change to insulin lispro treatment in early pregnancy and the rarely described development of proliferative diabetic retinopathy (PDR) by the third trimester in 3 women with type 1 or 2 diabetes who had normal retinal examinations at the beginning of pregnancy.1Kitzmiller J Main E Ward B Theiss T Peterson D. Insulin lispro and the development of proliferative diabetic retinopathy during pregnancy.Diabetes Care. 1999; 22 ([letter]): 874-876Crossref PubMed Scopus (73) Google Scholar In your November 2000 issue, Buchbinder et al2Buchbinder A Miodovnik M McElvy S Rosenn B Kranias G Khoury J et al.Is insulin lispro associated with the development or progression of diabetic retinopathy during pregnancy?.Am J Obstet Gynecol. 2000; 183: 1162-1165Abstract Full Text Full Text PDF PubMed Scopus (42) Google Scholar report the results of retinal examinations after pregnancy in 8 women with type 1 diabetes treated with insulin lispro before and during gestation—3 changed to insulin lispro in the first trimester, and 1 changed in the second trimester. Of their total group of 12, six had a normal result at the initial retinal examination before 24 weeks’ gestation, 4 had background diabetic retinopathy, 2 had PDR, and none showed progression at the 6- through 16-week postpartum examination. These results were compared with historical control subjects who used regular insulins. No individual data were reported for the insulin lispro group, but they had elevated glycohemoglobin A1 in the first trimester at 3.1 SDs above the normal mean, and 2 had rapid normalization of glycemic control during pregnancy defined as a decrease from enrollment to second trimester in glycohemoglobin A1 concentrations of > 4 SD. To compare their data with our original observation, it would be helpful if Buchbinder et al would report the details of retinal status, glycemic control, and hypertension in the 3 patients who changed to insulin lispro in the first trimester.Buchbinder et al also reported the second case known to me of a pregnant woman with diabetes who was treated with regular insulins who had a negative initial retinal examination and who progressed to PDR at the postpartum examinations. Their patient had many risk factors for progression, including diabetic nephropathy, chronic hypertension, superimposed preeclampsia, elevated glycohemoglobin in the first trimester, and rapid normalization of glycemic control during pregnancy.In their paper, Buchbinder et al cite the letter of Bhattacharyya and Vice,3Bhattacharyya A Vice PA. Insulin lispro, pregnancy, and retinopathy.Diabetes Care. 1999; 22 ([letter]): 2101-2102Crossref PubMed Scopus (29) Google Scholar which reported no worsening of retinopathy in 16 well-controlled women with type 1 or 2 diabetes mellitus who were treated with insulin lispro in the prepregnancy clinic and who continued the treatment during pregnancy. Buchbinder et al ignored my published response to Bhattacharyya and Vice in which I agreed that the subgroup with poor diabetic control at the onset of pregnancy is at the highest risk of developing PDR during pregnancy, perhaps because they may already have undetected peripheral retinal ischemia.4Kitzmiller J. Response to Bhattacharyya and Vice.Diabetes Care. 1999; 22: 2102-2104Crossref Scopus (86) Google ScholarI also noted that to avoid the trap of anecdotal cases and uncontrolled observations seeming to infer a cause-effect relationship, properly controlled studies should be designed. I proposed (on the basis of the data available in 1999) that 110 women with diabetes with initial negative retinal examinations would be required in each group (insulins with no growth factor homogeneity versus insulin lispro) to have an 80% power of finding a significant difference in development of PDR on the basis of insulin therapy at the.05 level, 1-tailed.4Kitzmiller J. Response to Bhattacharyya and Vice.Diabetes Care. 1999; 22: 2102-2104Crossref Scopus (86) Google ScholarBuchbinder et al calculated that 319 diabetic pregnant subjects would be required in each group to demonstrate a 100% increase in rates of progression of retinopathy from 6% in women treated with regular insulins to 12% in a group treated with insulin lispro (β =.2; α =.05; 1-tailed). A properly controlled trial could be enriched by focusing on pregnant women with diabetes with initial poor glycemic control, but it would have to exclude women with the additional risk factors of nephropathy and hypertension. On the basis of their report of 12 pregnant women treated with insulin lispro, it is unwarranted for Buchbinder et al to suggest that insulin lispro is not associated with the development of PDR during pregnancy. We can only answer the question with a prospective controlled trial. In 1999 we observed a possible association between change to insulin lispro treatment in early pregnancy and the rarely described development of proliferative diabetic retinopathy (PDR) by the third trimester in 3 women with type 1 or 2 diabetes who had normal retinal examinations at the beginning of pregnancy.1Kitzmiller J Main E Ward B Theiss T Peterson D. Insulin lispro and the development of proliferative diabetic retinopathy during pregnancy.Diabetes Care. 1999; 22 ([letter]): 874-876Crossref PubMed Scopus (73) Google Scholar In your November 2000 issue, Buchbinder et al2Buchbinder A Miodovnik M McElvy S Rosenn B Kranias G Khoury J et al.Is insulin lispro associated with the development or progression of diabetic retinopathy during pregnancy?.Am J Obstet Gynecol. 2000; 183: 1162-1165Abstract Full Text Full Text PDF PubMed Scopus (42) Google Scholar report the results of retinal examinations after pregnancy in 8 women with type 1 diabetes treated with insulin lispro before and during gestation—3 changed to insulin lispro in the first trimester, and 1 changed in the second trimester. Of their total group of 12, six had a normal result at the initial retinal examination before 24 weeks’ gestation, 4 had background diabetic retinopathy, 2 had PDR, and none showed progression at the 6- through 16-week postpartum examination. These results were compared with historical control subjects who used regular insulins. No individual data were reported for the insulin lispro group, but they had elevated glycohemoglobin A1 in the first trimester at 3.1 SDs above the normal mean, and 2 had rapid normalization of glycemic control during pregnancy defined as a decrease from enrollment to second trimester in glycohemoglobin A1 concentrations of > 4 SD. To compare their data with our original observation, it would be helpful if Buchbinder et al would report the details of retinal status, glycemic control, and hypertension in the 3 patients who changed to insulin lispro in the first trimester. Buchbinder et al also reported the second case known to me of a pregnant woman with diabetes who was treated with regular insulins who had a negative initial retinal examination and who progressed to PDR at the postpartum examinations. Their patient had many risk factors for progression, including diabetic nephropathy, chronic hypertension, superimposed preeclampsia, elevated glycohemoglobin in the first trimester, and rapid normalization of glycemic control during pregnancy. In their paper, Buchbinder et al cite the letter of Bhattacharyya and Vice,3Bhattacharyya A Vice PA. Insulin lispro, pregnancy, and retinopathy.Diabetes Care. 1999; 22 ([letter]): 2101-2102Crossref PubMed Scopus (29) Google Scholar which reported no worsening of retinopathy in 16 well-controlled women with type 1 or 2 diabetes mellitus who were treated with insulin lispro in the prepregnancy clinic and who continued the treatment during pregnancy. Buchbinder et al ignored my published response to Bhattacharyya and Vice in which I agreed that the subgroup with poor diabetic control at the onset of pregnancy is at the highest risk of developing PDR during pregnancy, perhaps because they may already have undetected peripheral retinal ischemia.4Kitzmiller J. Response to Bhattacharyya and Vice.Diabetes Care. 1999; 22: 2102-2104Crossref Scopus (86) Google Scholar I also noted that to avoid the trap of anecdotal cases and uncontrolled observations seeming to infer a cause-effect relationship, properly controlled studies should be designed. I proposed (on the basis of the data available in 1999) that 110 women with diabetes with initial negative retinal examinations would be required in each group (insulins with no growth factor homogeneity versus insulin lispro) to have an 80% power of finding a significant difference in development of PDR on the basis of insulin therapy at the.05 level, 1-tailed.4Kitzmiller J. Response to Bhattacharyya and Vice.Diabetes Care. 1999; 22: 2102-2104Crossref Scopus (86) Google Scholar Buchbinder et al calculated that 319 diabetic pregnant subjects would be required in each group to demonstrate a 100% increase in rates of progression of retinopathy from 6% in women treated with regular insulins to 12% in a group treated with insulin lispro (β =.2; α =.05; 1-tailed). A properly controlled trial could be enriched by focusing on pregnant women with diabetes with initial poor glycemic control, but it would have to exclude women with the additional risk factors of nephropathy and hypertension. On the basis of their report of 12 pregnant women treated with insulin lispro, it is unwarranted for Buchbinder et al to suggest that insulin lispro is not associated with the development of PDR during pregnancy. We can only answer the question with a prospective controlled trial.