The case of the patient presented herein highlights the importance of considering ICI-related DM as a potential adverse event of ICI treatment as well as the utility of DWI for assessment of pancreatic involvement in such cases. Individuals diagnosed with ICI-related DM often manifest abnormal results in blood and urine tests that can lead to a definitive diagnosis. However, pancreatic MRI findings in these individuals are rarely reported. Our report therefore provides a highly valuable description of pancreatic MRI results relevant to the treatment of individuals diagnosed with ICI-related DM.

DWI is a diagnostic technique that measures the movement of water molecules within organs, with ADC values calculated based on DWI providing a quantitative assessment of such water movement. Regions of high signal intensity on DWI images correspond to a decrease in ADC values and restricted diffusion, and such decreased ADC values have been found to result from the presence of fibrosis or the infiltration of inflammatory cells [16]. We therefore hypothesize that the regions of high signal intensity revealed by DWI in the pancreas of the present patient are indicative of inflammation and fibrosis.

Recent findings suggest that quantification of the ADC may prove informative for assessment of disease activity in individuals with other autoimmune diseases, such as autoimmune pancreatitis and Crohn’s disease [17, 18]. Pancreatic DWI has also yielded important results in individuals with type 1 DM, which is caused by lymphocytic infiltration of the pancreas and the rapid destruction of pancreatic β cells [11, 19, 20]. In addition, ADC values for the pancreas were found to be substantially lower in individuals with type 1 DM than in nondiabetic control individuals [11]. In the present case, the ADC values for all parts of the pancreas were lower than the previously reported optimal cutoff values for diagnosis of fulminant type 1 DM [11], suggesting that rapid destruction of pancreatic β cells may also occur in ICI-related DM as it does in individuals with fulminant type 1 diabetes.

In the case of pancreatitis associated with ICI treatment, the pancreas typically presents with diffuse or focal enlargement [21]. Pancreatic injury induced by ICIs can lead to reduced endocrine and exocrine pancreatic function and consequent metabolic and nutritional disturbances, and it is often accompanied by abdominal pain, pancreatic enlargement, and elevated circulating levels of pancreatic enzymes [22]. Whereas pancreatic enzyme levels have also been found to be elevated in certain individuals with ICI-related DM [6], atrophy of the pancreas has been detected either at the onset or later in the progression of this disease [23]. ICI-induced acute pancreatitis may also result in pancreatic atrophy [24], however, which can present a challenge to differentiation between ICI-related DM and pancreatitis on the basis of pancreatic morphology.

In the current case, no morphological irregularities of the pancreas were apparent on MRI at the early stage of disease onset, although that does not preclude the eventual development of pancreatic atrophy. The absence of obvious morphological changes in the pancreas together with the fact that the patient did not experience abdominal pain suggests that endocrine function was primarily affected. High signal intensity in DWI of the pancreas may therefore serve as an imaging biomarker reflecting islet damage and inflammatory infiltration at an early stage of ICI-related DM, even in the absence of morphological abnormalities of the pancreas.

Recent pathological findings have revealed increased infiltration of T lymphocytes and macrophages in pancreatic islets, including regions containing α cells and β cells, together with a downregulation of islet PD-L1 expression in individuals with ICI-related DM [25, 26]. The observed destruction of pancreatic β cells in such individuals may be linked to the decrease in PD-L1 expression in these cells resulting from ICI treatment. Further studies are required to investigate the relation between the pancreatic regions of high signal intensity identified by DWI and corresponding pathological features.

A prompt and accurate diagnosis of ICI-related DM is important in individuals treated with ICIs who develop hyperglycemia. However, individuals with type 2 diabetes treated with ICIs or those receiving high-dose glucocorticoid treatment may also experience hyperglycemia [27]. The correct diagnosis of ICI-related DM can therefore be challenging, especially during its early stages or in the absence of symptoms. In such instances, MRI of the pancreas can provide valuable information by detecting the rapid destruction of pancreatic β cells and thereby aid in the early diagnosis of ICI-related DM. Given that the clinical manifestations of ICI-related DM vary widely in terms of rate of onset and timing, additional studies are warranted to evaluate the utility of pancreatic MRI for diagnosis of this condition. Furthermore, it is of importance to investigate the relationship between the regions of high signal intensity revealed by pancreatic DWI and islet damage, including measurement of insulin secretion and its fluctuations over time.

In this study, MRI was not performed prior to the initiation of nivolumab therapy. Moreover, long-term alterations in MRI findings remain uncertain, with the exception of those observed 3 months after the onset of DKA. Hence, further studies are needed to elucidate the association between the initiation of immune processes and specific manifestations observed in MRI and to gain deeper insight into the distinctions between ICI-related DM and conventional fulminant type 1 DM.

Moreover, pancreatic islet autoantibodies were negative and susceptible HLA haplotypes associated with conventional type 1 DM or ICI-related DM were absent in this case. Previous research showed that islet autoantibodies and genotypes susceptible to conventional type 1 DM were negative in 47% and 39% of individuals with ICI-related DM, respectively [28]. In the absence of islet autoantibodies or susceptible HLA, as observed in this case, MRI findings may contribute to the diagnosis of ICI-DM; however, further research is required to elucidate the relationship between autoantibodies or HLA haplotypes and imaging findings.

In conclusion, we here report an individual with ICI-related DM characterized by hyperglycemia and DKA who showed elevated signal intensity with low ADC values in the pancreas on DWI performed at the early stage of disease onset. Quantification of ADC values may therefore prove beneficial for assessment of disease activity in individuals with ICI-related DM. The application of pancreatic MRI may also prove valuable for elucidation of the pathophysiology of hyperglycemia and islet damage in individuals treated with ICIs.