Beta-cell Reserve and Pancreatic Volume in Survivors of ALL
This study provides evidence for the first time of reduced β-cell reserve and smaller pancreatic volume in young adult survivors of childhood ALL treated with BMT/TBI. Previous studies investigating β-cell function in childhood BMT/TBI survivors were nonconclusive and limited by the heterogeneity of study cohort and methodology. D'Annunzio et al. concluded there was no evidence of β-cell dysfunction using HOMA-β and autoimmune antibodies in a cohort of 21 children treated with allogeneic or autologous BMT/TBI for leukaemia. Although transmission of autoimmune type 1 diabetes has been reported in patients post-BMT/TBI, β-cell dysfunction from treatment toxicity is unlikely to be an autoimmune process and HOMA-β has been shown to markedly underestimate the magnitude of β-cell defect in individuals with impaired glucose tolerance and diabetes. Immunosuppressants are known to cause hyperglycaemia by reduction in insulin secretion, but literature on the hyperglycaemic effect of immunosuppressants in patients postrenal transplantation shows that this is dose-dependent, nonprogressive and reversible. Therefore, abnormal glucose tolerance in the BMT/TBI survivors in this study is unlikely to be associated with previous immunosuppressants as all patients had discontinued any immunosuppressant treatment several years before this study.
In this study, the role of β-cell dysfunction in the development of impaired glucose tolerance and diabetes was explored, and this was assessed allowing for the degree of insulin resistance. In healthy subjects, β-cell insulin secretion occurs in a biphasic pattern. During a meal, insulin is secreted from a readily releasable pool of granules in the β-cell within 2 min of nutrient ingestion. This phase is completed by 10–15 min and is followed by a second phase of insulin secretion that is sustained until normoglycaemia is restored. When insulin sensitivity is reduced, there is initially an increase in endogenous insulin production to maintain normal glycaemia. However, this compensatory mechanism becomes blunted as disease progresses resulting in hyperglycaemia. Therefore, assessments of β-cell function must recognize the prevailing insulin sensitivity. The auxological measurements of our subjects showed that despite a lower BMI, BMT/TBI survivors have greater central-to-total fat distribution as indicated by higher waist-to-hip ratio. Associations between abnormal body composition and reduced insulin sensitivity in BMT survivors have been previously reported. Therefore, the interpretation of AIR (β-cell function) was performed allowing for insulin sensitivity.
Previous data on insulin secretion in BMT/TBI survivors have been limited. Cortona et al. reported normal AIR from IVGTT in 13 post-BMT paediatric patients treated for a mixture of malignant and nonmalignant conditions compared with population reference ranges. Lorini et al. demonstrated raised AIR from IVGTT in a heterogeneous group of paediatric patients who were 9 months to 10·2 years post-BMT with and without TBI. Neither of these studies evaluated insulin secretion with adjustment for insulin sensitivity. Frisk et al. reported a reduction in insulin sensitivity and raised AIR, but no difference in the glucose disposition index between young adult survivors of childhood ALL/lymphoma treated with BMT compared with chemotherapy alone. Although the glucose disposition index, a surrogate marker of β-cell function adjusted for insulin sensitivity, has been shown to predict progression to diabetes in population studies, this has not been validated in individuals with chronic health conditions associated with an increased risk of diabetes. In addition, whereas allogeneic BMT has been reported to be associate with a higher risk of diabetes than autologous BMT, a high proportion (83%) of the cohort by Frisk et al. received autologous transplantation. In our study, β-cell reserve was adjusted for the degree of insulin sensitivity by ancova, and results showed that abnormalities in AIR were present in BMT/TBI subjects without existing biochemical manifestation of impaired glucose tolerance. This implies evolving pathophysiology and is similar to at risk individuals with early β-cell decompensation in the general population who develop type 2 diabetes mellitus at a relatively early age and low BMI. Reductions in both phases of insulin response, to a greater extent in the first phase, have been demonstrated in individuals with a positive family history of diabetes. Our study did not demonstrate a significant reduction in the second phase insulin response in the BMT/TBI survivors, but this may reflect insufficient numbers as the power calculation was only based on previously reported outcomes of the first phase insulin response.
There are no previous published data on pancreatic size after BMT. MRI has been validated as a reliable and reproducible method in the assessment of pancreatic volume both in normal patients and in patients with type 1 diabetes mellitus. Pancreatic volume was reduced by a mean of 28·5% in BMT/TBI survivors compared with the controls. This is similar to a 26% reduction described in patients newly diagnosed with type 1 diabetes. The results of the pancreatic volume measured in Group 2 were highly comparable to normative population data. Reduction in pancreatic volume has been demonstrated in patients with different forms of diabetes mellitus including type 1, type 2, monogenic, cystic fibrosis related diabetes and diabetes in patients with thalassaemia major, but not previously in BMT/TBI survivors with diabetes. It has been suggested that the development of pancreatic atrophy may occur many years before the onset of clinical disease due to reduction in insulinotropic effect of the acinar cells and atrophy from chronic inflammation with β-cell destruction. Pancreatic atrophy has been less consistently reported in type 2 diabetes, with studies reporting from no change to 39% reduction in size.
Although an association has been shown between reduction in pancreatic size and time from diagnosis in patients with type 1 diabetes mellitus, the number of patients with diabetes in our study was too small to assess this. Consistent with previous findings, pancreatic volume in this study correlated with mean height, weight, BMI and BSA. The association between body size and pancreatic volume reflects the increase in pancreatic parenchyma associated with somatic growth during childhood.
The effect of radiation therapy on pancreatic volume is reported to be dose-dependant in animal studies. However, an association between TBI dose and pancreatic volume could not be assessed in our study in view of the homogeneity of dose delivered (19/21 received 14·4 Gy). Pancreatic volume differences cannot be explained by the use of l-asparaginase, a drug with a known pancreatic toxicity profile, as this was included in the primary treatment of all participants in both groups.
A number of previous studies have reported the lack of association between pancreatic volume and function. Sequeiros et al. described a group of nondiabetic cystic fibrosis patients with very small pancreatic size and suggested that sufficient insulin production only requires very small amounts of β-cell mass. Williams et al. found no associations between pancreatic volume and fasting C-peptide, glucose, glycated haemoglobin and islet cell antibodies levels in patients with type 1 diabetes mellitus. However, the relationship between pancreatic volume and β-cell function represented by the AIR has not been investigated previously. In the present study, AIR postarginine stimulation (AIRarg) adjusted for insulin sensitivity (ISIcomp) correlated positively with pancreatic volume.
Discussion
This study provides evidence for the first time of reduced β-cell reserve and smaller pancreatic volume in young adult survivors of childhood ALL treated with BMT/TBI. Previous studies investigating β-cell function in childhood BMT/TBI survivors were nonconclusive and limited by the heterogeneity of study cohort and methodology. D'Annunzio et al. concluded there was no evidence of β-cell dysfunction using HOMA-β and autoimmune antibodies in a cohort of 21 children treated with allogeneic or autologous BMT/TBI for leukaemia. Although transmission of autoimmune type 1 diabetes has been reported in patients post-BMT/TBI, β-cell dysfunction from treatment toxicity is unlikely to be an autoimmune process and HOMA-β has been shown to markedly underestimate the magnitude of β-cell defect in individuals with impaired glucose tolerance and diabetes. Immunosuppressants are known to cause hyperglycaemia by reduction in insulin secretion, but literature on the hyperglycaemic effect of immunosuppressants in patients postrenal transplantation shows that this is dose-dependent, nonprogressive and reversible. Therefore, abnormal glucose tolerance in the BMT/TBI survivors in this study is unlikely to be associated with previous immunosuppressants as all patients had discontinued any immunosuppressant treatment several years before this study.
In this study, the role of β-cell dysfunction in the development of impaired glucose tolerance and diabetes was explored, and this was assessed allowing for the degree of insulin resistance. In healthy subjects, β-cell insulin secretion occurs in a biphasic pattern. During a meal, insulin is secreted from a readily releasable pool of granules in the β-cell within 2 min of nutrient ingestion. This phase is completed by 10–15 min and is followed by a second phase of insulin secretion that is sustained until normoglycaemia is restored. When insulin sensitivity is reduced, there is initially an increase in endogenous insulin production to maintain normal glycaemia. However, this compensatory mechanism becomes blunted as disease progresses resulting in hyperglycaemia. Therefore, assessments of β-cell function must recognize the prevailing insulin sensitivity. The auxological measurements of our subjects showed that despite a lower BMI, BMT/TBI survivors have greater central-to-total fat distribution as indicated by higher waist-to-hip ratio. Associations between abnormal body composition and reduced insulin sensitivity in BMT survivors have been previously reported. Therefore, the interpretation of AIR (β-cell function) was performed allowing for insulin sensitivity.
Previous data on insulin secretion in BMT/TBI survivors have been limited. Cortona et al. reported normal AIR from IVGTT in 13 post-BMT paediatric patients treated for a mixture of malignant and nonmalignant conditions compared with population reference ranges. Lorini et al. demonstrated raised AIR from IVGTT in a heterogeneous group of paediatric patients who were 9 months to 10·2 years post-BMT with and without TBI. Neither of these studies evaluated insulin secretion with adjustment for insulin sensitivity. Frisk et al. reported a reduction in insulin sensitivity and raised AIR, but no difference in the glucose disposition index between young adult survivors of childhood ALL/lymphoma treated with BMT compared with chemotherapy alone. Although the glucose disposition index, a surrogate marker of β-cell function adjusted for insulin sensitivity, has been shown to predict progression to diabetes in population studies, this has not been validated in individuals with chronic health conditions associated with an increased risk of diabetes. In addition, whereas allogeneic BMT has been reported to be associate with a higher risk of diabetes than autologous BMT, a high proportion (83%) of the cohort by Frisk et al. received autologous transplantation. In our study, β-cell reserve was adjusted for the degree of insulin sensitivity by ancova, and results showed that abnormalities in AIR were present in BMT/TBI subjects without existing biochemical manifestation of impaired glucose tolerance. This implies evolving pathophysiology and is similar to at risk individuals with early β-cell decompensation in the general population who develop type 2 diabetes mellitus at a relatively early age and low BMI. Reductions in both phases of insulin response, to a greater extent in the first phase, have been demonstrated in individuals with a positive family history of diabetes. Our study did not demonstrate a significant reduction in the second phase insulin response in the BMT/TBI survivors, but this may reflect insufficient numbers as the power calculation was only based on previously reported outcomes of the first phase insulin response.
There are no previous published data on pancreatic size after BMT. MRI has been validated as a reliable and reproducible method in the assessment of pancreatic volume both in normal patients and in patients with type 1 diabetes mellitus. Pancreatic volume was reduced by a mean of 28·5% in BMT/TBI survivors compared with the controls. This is similar to a 26% reduction described in patients newly diagnosed with type 1 diabetes. The results of the pancreatic volume measured in Group 2 were highly comparable to normative population data. Reduction in pancreatic volume has been demonstrated in patients with different forms of diabetes mellitus including type 1, type 2, monogenic, cystic fibrosis related diabetes and diabetes in patients with thalassaemia major, but not previously in BMT/TBI survivors with diabetes. It has been suggested that the development of pancreatic atrophy may occur many years before the onset of clinical disease due to reduction in insulinotropic effect of the acinar cells and atrophy from chronic inflammation with β-cell destruction. Pancreatic atrophy has been less consistently reported in type 2 diabetes, with studies reporting from no change to 39% reduction in size.
Although an association has been shown between reduction in pancreatic size and time from diagnosis in patients with type 1 diabetes mellitus, the number of patients with diabetes in our study was too small to assess this. Consistent with previous findings, pancreatic volume in this study correlated with mean height, weight, BMI and BSA. The association between body size and pancreatic volume reflects the increase in pancreatic parenchyma associated with somatic growth during childhood.
The effect of radiation therapy on pancreatic volume is reported to be dose-dependant in animal studies. However, an association between TBI dose and pancreatic volume could not be assessed in our study in view of the homogeneity of dose delivered (19/21 received 14·4 Gy). Pancreatic volume differences cannot be explained by the use of l-asparaginase, a drug with a known pancreatic toxicity profile, as this was included in the primary treatment of all participants in both groups.
A number of previous studies have reported the lack of association between pancreatic volume and function. Sequeiros et al. described a group of nondiabetic cystic fibrosis patients with very small pancreatic size and suggested that sufficient insulin production only requires very small amounts of β-cell mass. Williams et al. found no associations between pancreatic volume and fasting C-peptide, glucose, glycated haemoglobin and islet cell antibodies levels in patients with type 1 diabetes mellitus. However, the relationship between pancreatic volume and β-cell function represented by the AIR has not been investigated previously. In the present study, AIR postarginine stimulation (AIRarg) adjusted for insulin sensitivity (ISIcomp) correlated positively with pancreatic volume.
Source...