Abstract
Individuals with type 1 diabetes (T1D) or permanent neonatal diabetes (PND) due to an INS gene mutation (INS-PND) have a marked reduction in pancreas volume by MRI compared with control individuals with no diabetes (ND). One possible explanation for this is loss of islet-acinar insulin signaling in these forms of severe insulin deficiency. To test the hypothesis that insulin deficiency drives the loss of pancreas volume in diabetes, we used a standardized and validated MRI protocol to measure pancreas volumes in individuals with various forms of monogenic diabetes, including maturity-onset diabetes of the young (MODY) and PND (HNF4A-MODY, GCK-MODY, HNF1A-MODY, HNF1B-MODY, INS-MODY, or INS-PND; n = 37), and compared their pancreas volumes with those of previously reported individuals with T1D (n = 93) or healthy control participants with ND (n = 90). Across all monogenic diabetes groups, individuals receiving insulin therapy had significantly smaller pancreas volume compared with those not requiring insulin. These results support the hypothesis that insulin signaling to the exocrine pancreas determines pancreas volume in multiple types of diabetes.
Article Highlights
Individuals with type 1 diabetes (T1D) have a markedly smaller pancreas, but the mechanism responsible for the reduction in size is unknown.
How pancreas volume differs in individuals with specific forms of monogenic diabetes and how pancreas volume relates to the severity of insulin deficiency are unknown.
Measured by MRI, individuals with permanent neonatal diabetes due to an INS gene mutation (INS-PND) or the HNF1B gene associated with maturity-onset diabetes of the young had smaller pancreas than individuals without diabetes. Across all types of monogenic diabetes, individuals receiving insulin replacement therapy had smaller pancreas than individuals not using insulin.
These results support the conclusion that insulin deficiency is a major factor contributing to changes in pancreas volume in T1D, INS-PND, and other forms of monogenic diabetes.
Introduction
Pancreas volume, as measured by MRI, is markedly reduced in individuals with autoimmune type 1 diabetes (T1D) starting before diagnosis and progressing after diagnosis and through long-standing disease (1–5). Our recent finding that pancreas volume is also reduced in people with permanent neonatal diabetes (PND) due to an INS gene mutation (INS-PND) (6) implicates islet-exocrine insulin signaling as a regulator of pancreas size. Mutations in certain genes known to be expressed in the pancreas and/or pancreatic islets cause varying degrees of nonautoimmune insulin deficiency that result in monogenic diabetes, including maturity-onset diabetes of the young (MODY) and PND. Studying individuals with these rare forms of diabetes provides a unique opportunity to investigate if insulin deficiency from forms of diabetes other than PND or T1D affects pancreas volume.
The most common forms of monogenic diabetes are caused by autosomal dominant missense or nonsense variations in the HNF4A (formerly MODY-1), GCK (MODY-2), HNF1A (MODY-3), HNF1B (MODY-5), or INS genes (MODY-10 and INS-PND) (7). GCK-MODY usually results in mild, nonprogressive, elevated fasting glucose levels that do not require treatment. HNF4A-MODY and HNF1A-MODY both have a progressive insulin secretory defect that may be responsive to sulfonylurea therapy. Of note, HNF4A regulates HNF1A. HNF1B-MODY has varying degrees of insulin deficiency and can be associated with genitourinary abnormalities, kidney dysfunction, and pancreatic atrophy. INS gene variants are associated with INS-PND or adult-onset INS-MODY, depending on the nature of the specific mutation. INS-PND requires lifelong insulin therapy, whereas INS-MODY may or may not require insulin treatment. To better understand the effects of islet-exocrine signaling, we examined pancreas volume in individuals with monogenic diabetes who do or do not require insulin therapy, compared with individuals with T1D or no diabetes (ND).
Research Design and Methods
Study Participants
Informed consent was obtained and documented in accordance with the institutional review boards at Vanderbilt University Medical Center and The University of Chicago. Study participants were enrolled for MRI imaging (clinicaltrials.gov identifier NCT03585153) as an extension of the Multicenter Assessment of the Pancreas in Type 1 Diabetes (https://medsites.vumc.org/map-t1d). All individuals in this study were recruited at Vanderbilt University Medical Center or through the University of Chicago Kovler Diabetes Center’s monogenic diabetes registry. Diagnoses were confirmed according to previously reported criteria (8). Gene sequencing was performed at the University of Chicago, and gene variant information was interpreted based on classifications of the Clingen-Monogenic Diabetes Expert Panel and the American College of Medical Genetics and Genomics and Association for Molecular Pathology (9). The ND (n = 90) and T1D (n = 93) cohorts and four individuals in the INS-PND cohort have been reported on in prior publications (1,3,4,6).
MRI
MRI was performed according to our standardized and validated protocol (10) on a Philips 3T Achieva or Ingenia scanner using a 16-channel receive torso coil. The image acquisition consisted of a T2-weighted fast-spin echo sequence with 1.5 × 1.5 × 5 mm spatial resolution and repetition time, echo time, and flip angle of 840 ms, 70 ms, and 90°, respectively. Imaging was performed in the axial plane over two breath holds with a total imaging time of 25 s. Spectral attenuated inversion recovery (SPAIR) fat suppression was performed to enhance contrast between pancreas and visceral fat. T1-weighted images were also used to guide pancreas segmentation on the T2-weighted image. To minimize risk to the participants, intravenously-administered contrast was not used. Pancreas volume (in mL), delineated by a board-certified radiologist (M.A.H.) who was blinded to the individuals’ diagnoses, was divided by individual body weight (in kg) to yield the pancreas volume index (PVI). PVI accounts for changes in pancreas size with body habitus and was used for consistency between our published studies and those of others (5). Some groups prefer to normalize pancreas volume to BMI (2) or body surface area, but both yielded results consistent with PVI in the present study (data not shown).
Statistical Analysis
Statistical analysis was performed using GraphPad Prism statistical software. P values were generated using ordinary, unpaired one-way ANOVA, corrected for multiple comparisons using the Šidák method, with P < 0.05 accepted as statistically significant.
Data and Resource Availability
The data sets generated during and/or analyzed during this study are available from the corresponding author on reasonable request. The resources used to generate the data are available from the corresponding author on reasonable request.
Results
Participant Characteristics
Participants with monogenic diabetes were 9–71 years of age, with age, sex, and BMI distribution comparable to that of the ND and T1D groups (Table 1). Available HbA1c data showed a wide range of glycemic control, particularly among those with HNF1A-MODY and HNF1B-MODY. Nineteen of 37 individuals with monogenic diabetes were treated with insulin therapy. As expected, no patients with GCK-MODY were using insulin, whereas all participants with INS-PND were receiving insulin therapy. HbA1c levels prior to starting treatment and insulin dosages were not available. Specific genetic mutation data are provided in Supplementary Table 1.
Table 1.
Participant characteristics by diabetes type and treatment regimen
| Group | Age, years | Sex, F/M | BMI, kg/m2 | HbA1c, % [mmol/mol] | Receiving insulin therapy |
|---|---|---|---|---|---|
| ND (n = 90) | 19 (8–71) | 43/47 | 22.8 (13.7–38.3) | NA | 0 |
| T1D (n = 93) | 17 (8–65) | 43/50 | 23.0 (14.9–40.1) | 10.1 (5.5–15.8) [87, 37–149] | 93 |
| HNF4A-MODY (n = 5) | 37 (25–71) | 5/0 | 23.2 (20.8–25.4) | 5.9 (5.5–7.0) [41, 37–53] | 2 |
| GCK-MODY (n = 5) | 16 (10–55) | 2/3 | 21.4 (18.6–35.8) | 7.0 (6.6–7.0) [53, 49–53] | 0 |
| HNF1A-MODY (n = 10) | 25 (18–60) | 6/4 | 27.9 (21.6–31.4) | 7.2 (5.6–11.9) [55, 38–107] | 4 |
| HNF1B-MODY (n = 6) | 31 (27–67) | 5/1 | 21.3 (19.1–25.6) | 6.8 (5.0–10.5) [51, 31–91] | 3 |
| INS-MODY (n = 3) | 39 (37–63) | 3/0 | 19.2 (18.5–22.1) | NA | 2 |
| INS-PND (n = 8) | 32 (9–62) | 4/4 | 26.3 (17.0–31.2) | 6.2 (5.8–6.9) [44, 40–52] | 8 |
| MODY, no insulin (n = 18) | 33 (10–55) | 11/7 | 22.4 (18.5–35.8) | 6.6 (5.0–11.9) [49, 31–107] | 0 |
| MODY, using insulin (n = 19) | 33 (9–71) | 14/5 | 25.0 (17.0–31.4) | 7.8 (5.5–10.5) [62, 37–91] | 19 |
All values shown as number or as median and range. F, female; M, male; NA, not available.
Pancreas Volume in Monogenic Diabetes
To determine if specific forms of monogenic diabetes have altered pancreas volume, we compared PVI in each group with those previously reported for ND and T1D participants. PVI scores in individuals with HNF4A-MODY, GCK-MODY, HNF1A-MODY, and INS-MODY were not significantly different from the ND group (mean score ± SD: 0.95 ± 0.35, 0.97 ± 0.22, 0.90 ± 0.19, and 1.07 ± 0.37 mL/kg respectively, compared with ND, 0.96 ± 0.25); those with HNF1B-MODY and INS-PND had PVI comparable to T1D (0.63 ± 0.53 and 0.45 ± 0.14 mL/kg, compared with T1D, 0.59 ± 0.24) and significantly smaller than that of the ND group (P < 0.05 for HNF1B-MODY and P < 0.001 for INS-PND) (Fig. 1).
Figure 1.
Pancreas volume is variable in monogenic diabetes. Pancreas volume (A) and PVI (B) for individuals with HNF4A-MODY (n = 5), GCK-MODY (n = 5), HNF1A-MODY (n = 10), HNF1B-MODY (n = 6), INS-MODY (n = 3), or INS-PND (n = 8), compared with individuals with ND (n = 90) or with T1D (n = 93). *P < 0.05, **P < 0.01, ***P < 0.001 by ANOVA, corrected for multiple comparisons by Sidak method.
Because insulin deficiency likely contributes to loss of pancreas volume in T1D and PND, we postulated that severity of insulin deficiency, rather than the specific mutations, would correlate with pancreas volume in monogenic diabetes. To test this hypothesis, we grouped all participants according to whether they were treated with insulin, as a surrogate for more severe insulin deficiency. Individuals with monogenic diabetes who were receiving insulin therapy had significantly lower PVI compared with those with monogenic diabetes not requiring insulin (mean score ± SD: 0.61 ± 0.33 vs 0.97 ± 0.29 mL/kg, P < 0.001) or those with ND (0.96 ± 0.25 mL/kg, P < 0.001), and this was similar in size to the PVI in the T1D group (0.59 ± 0.24 mL/kg) (Fig. 2). These findings suggest smaller pancreas volume in monogenic diabetes correlates with, and is possibly caused by, impaired endogenous insulin production.
Figure 2.
Pancreas volume is reduced in individuals who require insulin therapy. Pancreas volume (A) and PVI (B) for individuals with monogenic diabetes treated with insulin (n = 19) or not (n = 18), compared with individuals with ND (n = 90) or T1D (n = 93). *P < 0.05, **P < 0.01, ***P < 0.001 by ANOVA, corrected for multiple comparisons by Sidak method.
Discussion
Insulin deficiency is a hallmark of autoimmune T1D as well as PND, and both forms of diabetes are accompanied by markedly reduced pancreas volume. In the case of T1D, loss of pancreas volume is observed in the progression from stage 1 to stage 2 to overt stage 3 T1D (3), suggesting that the timing of initial insulin deficiency is very closely aligned with reduction in pancreas size. This work further emphasizes the critical link between these two processes through the discovery that individuals with insulin-dependent monogenic diabetes also have smaller pancreas volume compared with people with noninsulin-dependent monogenic diabetes or ND.
Our study cohort included individuals with the most common forms of monogenic diabetes (HNF4A, HNF1A, HNF1B, and GCK) in addition to multiple INS variants, representing a wide range of diabetes severity. As expected, individuals with GCK-MODY were not using insulin therapy, because they are the most clinically unaffected. However, HNF4A, HNF1A, HNF1B, and even INS-MODY cohorts were all heterogeneous in terms of the need for insulin therapy and common noninsulin therapies. All individuals with INS-PND had been insulin-dependent since infancy, except for one; this individual, who was diagnosed at age 22 years and is the father of three of the other participants, carries an INS variant strongly associated with INS-PND, as previously reported (6). We suspect that his later onset of hyperglycemia is due to variable expression or mosaicism of the de novo INS mutation he carries. Interestingly, his pancreas volume was similarly reduced to those of his children diagnosed in infancy, suggesting that the effects of islet-exocrine signaling persist in adulthood. Unanswered questions are whether individuals who develop autoimmune diabetes as adults, in whom insulin develops more slowly and thus may not be completely insulin deficient, or individuals with type 2 diabetes, some of whom take insulin, also have a smaller PVI.
When we did not consider treatment regimen, there was no significant difference in pancreas volume between individuals with monogenic diabetes and ND, with the exceptions of INS-PND, all of whom were receiving insulin therapy, and HNF1B-MODY, which is associated with pancreas atrophy and exocrine insufficiency (11). To our knowledge, pancreas volumes in HNF4A-MODY, GCK-MODY, and INS-MODY have not been reported (12), but our results differ from those of Vesterhus et al. (13), who reported reduced pancreas volume in HNF1A-MODY using computed tomography in the Norwegian MODY Registry. We speculate that the difference between our studies may be related to differences in patient population, stage of clinical disease, or diabetes treatment regimen, which was not reported in the Vesterhus et al. study.
Loss of a trophic effect of insulin on acinar cells has been proposed as a cause of reduced pancreas size in diabetes (14,15), but direct evidence has been lacking in humans. Combined with our prior findings (6) and those of others (16), the findings in the present work support the idea that insulin deficiency leads to a reduction in pancreas volume whether that insulin deficiency is caused by genetic variation or by autoimmune destruction of β-cells. It is also possible that insulin therapy itself suppresses remaining endogenous insulin secretion (17), resulting in lower local trophic effects of insulin in the pancreas, and contributes to the smaller PVI observed in insulin-treated participants. Although we were not able to distinguish how exogenous insulin may contribute to the observed changes in this study, we recently reported that smaller pancreas volume predicts rapid progression to clinical disease in individuals with presymptomatic stages of T1D (3). Importantly, all individuals at these presymptomatic stages of T1D do not require insulin, suggesting that pancreas atrophy precedes the need for exogenous insulin, though further suppression of endogenous insulin after initiation of insulin therapy remains plausible.
It is possible that hyperglycemia directly affects exocrine tissue, though two factors in our data argue against a primary role for hyperglycemia in this process. First, the GCK-MODY group, which all have mild hyperglycemia that generally does not require therapy, had no difference in pancreas volume compared with ND individuals. Second, we observed no correlation between pancreas volume and HbA1c, a measure of severity of hyperglycemia.
Whether insulin influences exocrine tissues via paracrine effects or through intercompartmental vascular communication (18,19) remains unknown. Other islet-derived factors also likely affect pancreas volume. For example, mouse studies have demonstrated an islet-hepatic axis, in which glucagon signaling from islets regulates serum amino acid levels via hepatocytes, which influences pancreatic acinar cell proliferation (20). Whether altered glucagon signaling or amino acid metabolism contributes to changes seen in states of insulin deficiency like T1D or monogenic diabetes requires further study.
Our study, though large in total size, is limited in number of individual MODY types, so correlation between pancreas volume and other clinical features, including specific gene variants, was not possible. Other limitations of our study include lack of details about the participants’ age at diagnosis and treatment history, including the factors that led their clinical providers to initiate insulin therapy. We also do not have a direct evaluation of islet or exocrine function, which may vary widely even in people receiving insulin therapy, or longitudinal data to evaluate pancreas volume before and after starting insulin. Pairing pancreatic imaging with direct measures of residual β-cell function, such as by oral glucose tolerance testing, in individuals with nonautoimmune insulin deficiency would strengthen the connection between insulin signaling and pancreas size.
Overall, these data support the conclusion that loss of pancreas volume is driven by insulin deficiency, predicts those likely to progress to overt diabetes, and eventually coincides with the need for insulin therapy. More studies are needed to clearly define the nature of communication between pancreatic islets and surrounding exocrine tissue, how endocrine-exocrine intercommunication affects disease progression, and how these changes can help guide intervention in T1D prevention and assessment of atypical forms of diabetes.
This article contains supplementary material online at https://doi.org/10.2337/figshare.28945121.
Article Information
Acknowledgments. The authors thank the Kovler Diabetes Center and University of Chicago for their help in identifying study participants; the Vanderbilt University Institute of Imaging Science for their role in performing the MRI scans; and the study participants and their families for their dedication to diabetes research.
Duality of Interest. No potential conflicts of interest relevant to this article were reported.
Author Contributions. J.J.W., J.V., D.J.M., and A.C.P. designed the experiments. J.M.W. and J.J.W. performed the research, recruited participants, analyzed the data, and wrote the manuscript. L.R.L.-F. and D.B. helped recruit participants. M.A.H. read and outlined the MRI images. J.J.W. performed statistical analysis and data interpretation. D.J.M., J.V., A.C.P., L.P., S.A.W.G., L.R.L., and B.K. critically reviewed the manuscript. All authors approved of the final version. J.J.W. is the guarantor of this work and, as such, has full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Prior Presentation. Parts of this study were presented virtually in abstract form at the 80th Scientific Sessions of the American Diabetes Association, 12–16 June 2020.
Funding Statement
Research support was received from The Leona M. and Harry B. Helmsley Charitable Trust (grant 2207-05374 to D.J.M.), Breakthrough T1D (formerly JDRF) (grants 3-SRA-2015-102-M-B and 3-SRA-2019-759-M-B to A.C.P.); the U.S. Department of Veterans Affairs (grant BX005910 to J.J.W.); the National Institutes of Health National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) (grant DK129979 to J.V. and grant K08DK133691 to J.J.W.); and the Vanderbilt Diabetes Research & Training Center (grant DK20593). REDCap was used for this work, which is supported by the NIH National Center for Advancing Translational Sciences (grant UL1TR000445). The University of Chicago Kovler Diabetes Center monogenic diabetes registry is supported by the NIDDK (grants R01DK104942 and P30DK020595).
Footnotes
Clinical trial reg. no. NCT03585153, clinicaltrials.gov
Supporting information
References
- 1. Virostko J, Williams J, Hilmes M, et al. Pancreas volume declines during the first year after diagnosis of type 1 diabetes and exhibits altered diffusion at disease onset. Diabetes Care 2019;42:248–257 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Campbell-Thompson ML, Filipp SL, Grajo JR, et al. Relative pancreas volume is reduced in first-degree relatives of patients with type 1 diabetes. Diabetes Care 2019;42:281–287 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Virostko J, Wright JJ, Williams JM, et al. Longitudinal assessment of pancreas volume by MRI predicts progression to stage 3 type 1 diabetes. Diabetes Care 2024;47:393–400 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Wright JJ, Dulaney A, Williams JM, et al. Longitudinal MRI shows progressive decline in pancreas size and altered pancreas shape in type 1 diabetes. J Clin Endocrinol Metab 2023;108:2699–2707 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Williams AJK, Thrower SL, Sequeiros IM, et al. Pancreatic volume is reduced in adult patients with recently diagnosed type 1 diabetes. J Clin Endocrinol Metab 2012;97:E2109–E2113 [DOI] [PubMed] [Google Scholar]
- 6. Wright JJ, Williams JM, Letourneau-Freiberg LR, et al. Insulin deficiency from insulin gene mutation leads to smaller pancreas. Diabetes Care 2023;46:773–776 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. ElSayed NA, Aleppo G, Bannuru RR, et al.; American Diabetes Association Professional Practice Committee . 2. Diagnosis and Classification of Diabetes: Standards of Care in Diabetes—2024. Diabetes Care 2024;47:S20–S42 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Greeley SAW, Naylor RN, Cook LS, Tucker SE, Lipton RB, Philipson LH. Creation of the web-based University of Chicago Monogenic Diabetes Registry: using technology to facilitate longitudinal study of rare subtypes of diabetes. J Diabetes Sci Technol 2011;5:879–886 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Richards S, Aziz N, Bale S, et al.; ACMG Laboratory Quality Assurance Committee . Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. Genet Med 2015;17:405–424 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Virostko J, Craddock RC, Williams JM, et al. Development of a standardized MRI protocol for pancreas assessment in humans. PLoS One 2021;16:e0256029. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Edghill EL, Bingham C, Slingerland AS, et al. Hepatocyte nuclear factor-1 beta mutations cause neonatal diabetes and intrauterine growth retardation: support for a critical role of HNF-1beta in human pancreatic development. Diabet Med 2006;23:1301–1306 [DOI] [PubMed] [Google Scholar]
- 12. Haldorsen IS, Ræder H, Vesterhus M, Molven A, Njølstad PR. The role of pancreatic imaging in monogenic diabetes mellitus. Nat Rev Endocrinol 2011;8:148–159 [DOI] [PubMed] [Google Scholar]
- 13. Vesterhus M, Haldorsen IS, Raeder H, Molven A, Njølstad PR. Reduced pancreatic volume in hepatocyte nuclear factor 1A-maturity-onset diabetes of the young. J Clin Endocrinol Metab 2008;93:3505–3509 [DOI] [PubMed] [Google Scholar]
- 14. Chiarelli F, Verrotti A, Altobelli E, Blasetti A, Morgese G. Size of the pancreas in type I diabetic children and adolescents. Diabetes Care 1995;18:1505–1506 [PubMed] [Google Scholar]
- 15. Gilbeau JP, Poncelet V, Libon E, Derue G, Heller FR. The density, contour, and thickness of the pancreas in diabetics: CT findings in 57 patients. AJR Am J Roentgenol 1992;159:527–531 [DOI] [PubMed] [Google Scholar]
- 16. Lu J, Guo M, Wang H, et al. Association between pancreatic atrophy and loss of insulin secretory capacity in patients with type 2 diabetes mellitus. J Diabetes Res 2019;2019:6371231. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Elahi D, Nagulesparan M, Hershcopf RJ, et al. Feedback inhibition of insulin secretion by insulin: relation to the hyperinsulinemia of obesity. N Engl J Med 1982;306:1196–1202 [DOI] [PubMed] [Google Scholar]
- 18. Dybala MP, Kuznetsov A, Motobu M, et al. Integrated pancreatic blood flow: bidirectional microcirculation between endocrine and exocrine pancreas. Diabetes 2020;69:1439–1450 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Rizk AA, Dybala MP, Rodriguez KC, Slak Rupnik M, Hara M. Pancreatic regional blood flow links the endocrine and exocrine diseases. J Clin Invest 2023;133:e166185. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Dai C, Zhang Y, Gong Y, et al. Hyperaminoacidemia from interrupted glucagon signaling increases pancreatic acinar cell proliferation and size via mTORC1 and YAP pathways. iScience 2024;27:111447. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.