4.9 million people in the United Kingdom (UK) have diabetes characterised by progressive loss of beta-cell mass and/or function. There are broadly two main classifications of diabetes mellitus; Type 1 (T1DM) and Type 2 (T2DM) but sometimes it is difficult to precisely distinguish between the two. The first simultaneous pancreas and kidney transplant was performed in 1966 and initially reserved for patients with T1DM [1]. As the techniques and indications have evolved it was soon realised that some patients with T2DM would also benefit [2–5].

Approximately 90% of the diabetic population have been classified as T2DM compared to only 8% with T1DM [6]. Previously it was thought that T1DM was a disease with onset always in the young, whereas T2DM affected only older adults who were overweight. However, with increasing understanding about diabetes, binary classification of T1DM and T2DM has become increasingly difficult [7]. Studies using historic data comparing various cohorts of diabetic patients is therefore subject to different interpretations when considering the complexities of categorisation. The complex aetiology also makes planning the best management of these patients challenging when they are referred for beta-cell replacement therapy. T2DM is an extremely heterogenous disease. For example, life threatening severe hypoglycaemic unawareness is rare in T2DM but more common in patients with T1DM. Consequently, both Pancreas transplant alone (PTA) and Islet transplant alone (ITA), indicated in the UK solely for recurrent life-threatening hypoglycaemia has never been undertaken for T2DM patients. In the current study outcomes after SPK, as opposed to solitary pancreas transplantation, were investigated in patients with T2DM.

The current UK listing criteria for SPK in a potential T2DM recipient includes; 1) the need for insulin treatment and dependence 2) a BMI of ≤30 kg/m² and 3) patients must be receiving dialysis or have a GFR ≤20 mLs/min [8]. The presence of C-peptide is not an absolute contraindication because of inaccuracies in evaluation in patients with renal failure [9]. In essence potential T2DM recipients need to be fit for surgery, not overtly obese, on insulin treatment with end stage renal disease. Numerous previous studies have shown that patient survival after SPK transplant is superior to those patients on dialysis or those having deceased donor kidney transplant alone (KTA) [10–13].

The aim of this study was to compare outcomes in the NHSBT database between patients with either T2DM or T1DM after SPK transplantation.

NHS Blood and Transplant UK registry data was obtained for all simultaneous pancreas and kidney (SPK) transplants that took place between 2003–2019, n = 2,236. Cases where the aetiology of diabetes was missing or had been classified as “other” rather than specifically Type 1 or Type 2 diabetes were excluded, as were recipients who had received a re-transplant, resulting in a final cohort of n = 2,154. The type of diabetes was predefined by the centre listing the patient for transplant.

Recipient characteristics; age, sex, body mass index (BMI—categorised by the WHO classification) [14], ethnic group (categorised as white or BAME—black, Asian and minority ethnic), waiting time for transplant, pre-transplant insulin requirements and dialysis status were analysed for variations between our two cohorts. Donor characteristics; age, sex, ethnic group, donor type (DBD/DCD), warm ischaemic time (WIT) and cold ischaemic time (CIT) were also analysed for variation.

Recipient survival and death-censored pancreas and kidney graft survival were analysed at 1, 3, 5 and 10 years. Pancreas graft failure was defined by the recipient follow-up centre based on the resumption of insulin treatment. Kidney graft survival is defined as resumption of dialysis.

We further delineated our groups by BMI into; T1DM < 30 kg/m², T1DM > 30 kg/m², T2DM < 30 kg/m² and T2DM > 30 kg/m² and performed recipient survival and death-censored pancreas and kidney graft and patient survival at 10 years. We also further delineated our groups by ethnic group into; T1DM-White, T1DM-BAME, T2DM-White, T2DM-BAME.

Any patient outside standard listing criteria is discussed through an exemptions panel. Expert opinion within this group guided potential listing.

Common complications after pancreas transplant were analysed between our two cohorts including; incidence of post-operative myocardial infarction (MI), cerebrovascular accident (CVA), anastomotic leak, urinary tract infection (UTI), systemic infection (further delineated into bacterial, viral or fungal), pancreatitis, rejection at 3 months and resumed insulin use at 1 year.

This study aims and methodology were submitted to the NHS Blood and Transplant Research Advisory Group (RAG) and approved prior to gaining access to the registry data.

The majority, (95.6%) of simultaneous pancreas and kidney (SPK) transplants were performed in recipients with Type 1 diabetes mellitus (T1DM) (n = 2,060). Only 3.4% (n = 94) of SPK transplants have been performed between 2003 and 2019 in recipients with type 2 diabetes mellitus (T2DM). Over the past 15 years we have seen an increasing trend in the percentage of SPK transplants being performed in T2DM recipients (1.6% in 2004 to 5.8% in 2018), Table 1. However, numbers remain comparatively small when compared to T1DM recipients. The median follow-up of all patients in this study was 1900 days, which was until death in 193 patients (8.96%).

In the UK categorisation of diabetes is primarily a clinical diagnosis. For a diagnosis of T1DM, the current criteria includes; hyperglycaemia (random plasma glucose >11 mmol) with one or more of the following features; ketosis, rapid weight loss, age <50, BMI < 25 kg/m² and/or a personal/family history of autoimmune disease [15]. For a diagnosis of T2DM, the patient should have persistent hyperglycaemia (inferred using a HbA1c > 48 mmol/mol as a surrogate marker) [16], symptoms of; polyuria, polydipsia, unexplained weight loss, recurrent infections or tiredness in the context of known risk factors (i.e., obesity, family history, ethnicity, metabolic syndrome) and the absence of T1DM features (i.e., rapid onset, young age, insulin dependence, ketoacidosis). Unlike other countries, biomarkers such as auto-antibodies or c-peptide are not routinely used for diagnosis or classification in the United Kingdom. The National Institute of Health and Care Excellence (NICE) guidance on diagnosis of diabetes (updated in 2022) recommends using clinical features to make the diagnosis of diabetes, to not routinely use C-peptide, and if using diabetes-specific autoantibodies to take into account the false negative rate of this test [15].

Irrespective of how a patient’s diabetes is classified the unifying result is hyperglycaemia which leads to down-stream micro and macrovascular complications. Early detection and changing medical management of diabetes mellitus undoubtedly helps delay the onset of complications associated with hyperglycaemia however, retinopathy, vasculopathy and nephropathy still remain serious and common afflictions in these patients [6]. Pancreas transplant remains the only realistic, long-term insulin-independent treatment for diabetes [10]. During 2020 there were 198 patients on the UK SPK transplant waiting list compared to 16 who need simultaneous islet and kidney (SIK) transplant [17]. Although the indications are the same the patient groups are likely to be different in terms of associated co-morbidities. SPK transplant is the favoured treatment in those that are fit when considering long-term insulin independence, this also applies to those patients with T2DM.

This study has shown comparable death censored pancreas graft survival and kidney graft survival after simultaneous pancreas and kidney transplants regardless of type of diabetes with the caveat that the diagnosis and type of diabetes was pre-defined by the listing centre using the UK criteria highlighted above. This study has also shown comparable patient survival at 1 and 3 years regardless of diabetes type. Interestingly at 5 years we see a statistically significant decrease in T2DM patient survival when compared to their T1DM counterparts. This trend is not borne out at 10 years which again shows no statistically significant difference in patient survival. We believe this may be partly explained by the older, heavier T2DM having a poorer initial patient survival and the younger, lighter T2DM recipients surviving out to 10 years.

In 2020, an American single-centre study (n = 323) demonstrated comparable outcomes in terms of pancreas graft survival (death censored) and incidence of post transplantation diabetes mellitus (PTDM) between recipients with T2DM (n = 39) compared with T1DM (n = 284). Patients in this study were defined as T1DM and T2DM using clinical assessment at the time of initial evaluation as well as utilising a novel scoring system assessing; pre-transplant insulin requirement, pre-transplant fasting c-peptide levels (assigning a score of +2 if C-peptide <0.5 ng/L, −1 if 0.5–2 ng/L and −2 if >2 ng/L), family history and the presence of diabetes-associated antibodies. A score from −9 to +9 was created, and a negative score correlated with T2DM and a positive score with T1DM [18]. This scoring system was not used in our study.

The largest reported study to date (n = 6,756), utilised the United Network for Organ Sharing (UNOS) database. Again, the majority of patients (90.8%, n = 6,141) had T1DM and only 8.2% of SPK transplant’s were performed in T2DM (n = 582). This study also showed no statistically significant difference in graft and patient survival in patients with T2DM [19]. Type of diabetes was predefined by the listing centre, and no further information regarding this was offered in this publication which makes it harder to draw any more useful detail between the two cohorts.

A smaller single centre study was reported in 2013 by an Austrian group (n = 248) comparing T1DM undergoing SPK transplant (n = 195) with T2DM SPK transplant (n = 21) and T2DM kidney transplant alone (KTA) (n = 32) [20]. They defined T2DM using detectable C-peptide levels. This study demonstrated comparable rates of graft survival between T1DM and T2DM recipients undergoing SPK. A statistically significant difference in patient survival was seen when comparing T2DM recipients (both SPK and KTA) with T1DM who underwent SPK (p < 0.001). This finding contrasts with the other literature discussed. It is also important to note this paper does not differentiate KTA by donor brain death (DBD), donor circulatory death (DCD) or living related donor (LRD) making it difficult to interpret. However, there is a large American study utilising a National Registry for T2DM patients (n = 37,117) where T2DM recipients were shown to have better statistically significant patient survival and kidney allograft survival after SPK when compared to those receiving a KTA alone, irrespective of whether the kidney was from a deceased donor or living donor [21].

A further single centre study in Argentina (n = 45), showed no statistically significant difference in patient survival when comparing T1DM (n = 35) to T2DM (n = 11) after SPK. They classified patients type of diabetes clinically; those who were diagnosed in childhood, with a lower BMI and requiring immediate insulin treatment were classified as T1DM whereas patients who were diagnosed with diabetes aged >30 years/old and with metabolic features were classified as T2DM.

A final study from Washington classifying diabetes by C-peptide >/<0.8 ng/mL (n = 136) showed comparable outcomes between their Type and Type 2 recipients. They state that C-peptide status does not influence outcomes after SPK transplant and this treatment option should be offered regardless of their C-peptide level [22].

Whilst the majority (90%) of the UK diabetic population have T2DM, only 3.4% of this population had an SPK. Other countries have comparable proportions of T2DM; in the US 91% of the diabetic population have T2DM [23], in Germany 90%–95% [24] and 90% in the Netherlands [25]. In 2010 the International Pancreas Transplant Registry, IPTR (which receives data from both UNOS and Eurotransplant) showed 8% of SPK’s were performed in patients presumed to have T2DM [26, 27]. Despite comparable proportions of T2DM within these national populations one can extrapolate that the percentage of SPKs performed in recipients with T2DM in the UK is well below that of our American and other European counterparts [3, 28]. However we accept there is no uniform consensus on the criteria used for a diagnosis of T2DM which could explain this observation.

From 2019–2021 a consensus group was formed to deliberate on current pancreas transplant outcomes in an effort to provide evidence to support current practice (28), After removal of duplicate papers and by applying exclusion criteria, 31 studies regarding SPK in T2DM patients were reviewed. The consensus concluded that SPK transplant improved both quality of life and long-term survival in suitable T1DM and T2DM recipients. For T2DM, the authors state that evidence is insufficient to suggest SPK transplant provides greater survival when compared with living donor kidney transplant alone. We did not analyse solitary kidney transplants in this study and to our knowledge this analysis has never been done. It would be interesting to see if SPK transplant is better than PAK transplant with a living donor kidney for patients with T2DM. Numbers in our national dataset would be too small for a useful comparison.

From our cohort we can see that those patients who received a pancreas transplant with T2DM were more likely to be older, male and with a higher BMI. Factors associated with the development of T2DM are obesity and smoking [29–31] which have been typically associated with a male cohort [32, 33]. Currently the UK SPK transplant patient selection policy contains a selection criteria of a BMI <30 kg/m² for T2DM recipients and does not define a BMI restriction for those with T1DM [8]. Although there is some selectively amongst UK centres where >30 kg/m² may be considered as a relative contra-indication by some.

In total 8.2% (n = 176) of our cohort had a BMI > 30 kg/m². A previous study from our group showed that whilst BMI does affect outcomes, those who received a pancreas transplant (SPK, PTA and PAK) with a BMI > 30 kg/m² had comparable outcomes with recipients with a BMI < 30 kg/m² and concluded that assigning a cut off of <30 kg/m² as a gatekeeper to pancreas transplantation had the potential to prevent good candidates accessing this treatment option [18,34]. For our study we delineated BMI by type of diabetes to better analyse the data in the context of this selection policy. 8.5% (n = 8) of the SPK transplants performed in T2DM had a BMI > 30 kg/m². As this goes against the standard selection policy their case went to an exemptions panel prior to being placed on the waiting list and so were excellent candidates in terms of other parameters. We saw no statistically significant difference in overall graft survival, however recipients with T2DM and a BMI > 30 kg/m² had poorer patient survival than the other categories. Whilst numbers are small in our T2DM > 30 kg/m² group, this would suggest that the combination of T2DM and obesity is of concern. Many T2DM patients are not obese and, in this study, have been shown to have comparable outcomes. Those with T1DM showed comparable graft and patient survival outcomes independent of BMI.

Our study also found that those patients with T2DM receiving a transplant were much more likely to be from a BAME (Black, Asian and minority ethnic) community rather than their T1DM counterparts. This is not unsurprising because from epidemiological studies we know that UK BAME communities have a 3–5× higher prevalence of T2DM with an earlier age of onset [28, 35, 36]. We further analysed our dataset to better understand the role ethnicity played in recipient outcomes. We found no statistically significant difference in either graft or patient survival regardless of ethnicity.

An analysis of common complications after pancreas transplantation was performed. This included incidence of myocardial infarction, cerebrovascular accident, anastomotic leaks, urinary tract infections, systemic infections, pancreatitis, graft rejection at 3 months and resumed insulin use (at 1 year and also 5 years). Both T1DM and T2DM have been associated with an increased incidence of infection [37], poorer wound healing [38] and thrombotic events [39, 40]. When compared against each other we found comparable rates of all complications regardless of type of diabetes.