Simultaneous pancreas-kidney transplantation (SPKT) in Type I diabetes (T1D)with end-stage renal disease (ESRD) has produced significant improvement in prolongation and quality of life. Patient survival approaches 97% and 92% at 1 year and 3 years, respectively [1]. The half-life of pancreas allografts has increased to 15.5 years [2] secondary to advances in immunosuppressive therapy, surgical techniques, and immune monitoring [1, 3–5]. SPKT is also associated with improved kidney graft survival [6, 7] and improved preservation of kidney graft ultrastructure and function [8] compared to deceased donor kidney transplant alone.

Concerning SPKT in Type II diabetes mellitus (T2D) patients with ESRD, many studies have addressed the outcomes of pancreas transplantation for such patients [9]. Such studies have found comparable results between the two types of recipients regarding various endpoints including insulin resistance and β-cell function [3], kidney and pancreas graft survival [9–16], post-transplant glycemic control, BMI control [9, 17], and patient survival [6, 11, 18].

However, the effect of diabetes type on graft rejection after pancreas transplantation is less well understood. Differing rates of allograft rejection are observed in other abdominal solid organ transplants based on the primary etiology of the organ failure, especially with autoimmune components [19–30]. Several studies have evaluated the effects of donor-specific anti-HLA antibodies (DSA) on graft outcomes [31–33] and noted significantly decreased kidney and pancreas allograft survival [33–36]. None of these studies, however, account for the type of diabetes as a distinguishing factor.

In addition, T2D patients may be obese and consequently may have an increased risk of surgical site infections [37, 38] and worse graft outcomes [39]. The inflammatory milieu of T2D may impact the risk of surgical infections, thrombosis, etc., [40–42]. While these theoretical risks may exist, the outcomes of T1D and T2D SPKT recipients with respect to important specific surgical and infection-related outcomes have not been thoroughly evaluated.

Thus, in this study, we sought to comprehensively examine whether the type of diabetes impacts the rates of acute biopsy-proven rejection and DSA development as well as other key surgical and infectious complications. Additionally, we globally analyze factors contributing to these outcomes in the T1D and T2D SPKT populations.

A single center retrospective review of prospectively collected data from a comprehensive in-house Transplant Database, electronic medical records, and the UNOS/OPTN STAR file was approved by the local Institutional Review Board. Analysis included primary SPKT recipients from 2006–2017 with 1-year minimum post-transplant follow-up. Diabetes mellitus types were determined by a holistic assessment with a grading system that included factors of patients’ age at diabetes onset, need for immediate use of insulin, pre-transplant fasting C-peptide, family history of diabetes, and the presence of autoantibodies (GAD65, Insulin- and Islet-antibodies) [43]. Primary outcomes included patient and graft survival, incidence of biopsy proven pancreas and kidney rejection and dnDSA, readmissions, infections, and surgical complications, including bleeding, pancreatic graft thromboses and other surgical site complications (Figure 1).

Whereas the majority of studies focus on patient and graft survival outcomes between T1D and T2D recipients, few address key infectious, surgical, and immunological outcomes. The current study addresses this gap and demonstrates that similar post-transplant outcomes, such as the incidence of acute BPR, readmissions, infections, UTIs, thrombosis, and other major surgical complications can be achieved between T1D and T2D SPKT recipients. Also, consistent with findings from previous studies demonstrating improvement in patient survival with advancing eras [9, 10, 43], the present study demonstrates acceptable and comparable patient-, pancreas allograft- and kidney allograft-survival in T2D versus T1D SPKT recipients.

Organ transplant recipients whose primary etiology of organ failure is autoimmune in nature may have higher rates of rejection and recurrence, especially in kidney transplantation [19–23] and liver transplantation [24–30]. However, a UNOS registry review did not find a significant association of rejection between T2D and T1D when combining kidney and pancreas rejection outcomes [10]. This study has the caveat however that kidney and pancreas rejection were not analyzed separately, and the majority of centers did not perform routine pancreas allograft biopsies in SPKT recipients, thereby potentially leading to underreporting of pancreas rejection. Thus, we posited that T1D SPKT recipients may experience higher rates of pancreas rejection than T2D SPKT recipients given the autoimmune nature of diabetes in the former. However, we did not observe a significantly different 1-year pancreas, or kidney, BPR rate in T2D vs. T1D patients. The overall rates of rejection in our population are consistent with those previously reported in the literature (4%–38%) [48–53]. The current study provides greater granularity particular to rejection type and severity compared to prior studies [6, 9, 10, 14, 15, 52]. These prior studies, additionally, did not meticulously categorize T1D and T2D recipients [14, 15, 54], assess pancreas BPR separately from kidney BPR [10, 14, 15, 55, 56] or specifically look at pancreas BPR [6, 9, 52, 57]. Thus, the current study adds a more comprehensive assessment of the rejection risk confronting T2D SPKT recipients. It also suggests a very low overall incidence of pancreas antibody-mediated rejection (ABMR) based on the ∼6% overall incidence of C4d>5% staining on biopsies in both T1D and T2D recipients, which is consistent with previously reported data [50].

While not definitive, our data does suggest a possible signal with regard to more rejection in T1D recipients. For example, we observed a greater number of episodes of Grade 2-3 ACR, indeterminate ACR, and C4d+ rejection, and numerically more patients with these rejection diagnoses within the first year in T1D patients. Moreover, we observed a higher incidence of dnDSA in T1D patients compared to T2D patients. Thus, this type of diabetes may be associated with an increased risk of pancreas rejection endpoints. Though we observed a higher rate of pancreas rejection signals by univariate analysis, we failed to detect a significant difference in multivariable analyses. Given the lack of major differences between T1D and T2D SPKT recipients, the current study suggests that a primary autoimmune pathology does not pose a substantially increased risk of BPR, nor does it suggest T2D confers higher rates of pancreas or kidney BPR. Thus, the type of diabetes thus should not affect candidacy for SPKT from the rejection perspective.

Similar patient and graft survival outcomes have been described with both T-cell depleting and non-depleting agents for SPKT [53, 58–60]. Overall, lower rates of early acute rejection have been described in SPKT with T-cell-depleting agents versus non-depleting agents [53]. Comparing types of T-cell-depleting therapies, ALEM versus ATG has been associated with comparable surgical complications, readmissions, thromboses, and bleeding [61]. These studies however involve very few T2D recipients. The results of our study are congruent with these findings and indicate that, compared to BAS, ALEM induction might be beneficial for pancreas graft rejection and was associated with lower risk of dnDSA development, while ATG induction was associated with reduced kidney graft rejection. We also did not find an association between ALEM and kidney graft rejection, consistent with Sampaio et al [10]. Induction trends in our cohort are also consistent with those reported in T2D recipients represented in registry data, with an increasing trend toward use of T-cell-depleting antibodies in more recent eras [9]. Larger cohorts of T2D SPKT recipients are needed to make definitive conclusions regarding any differences in the rejection rate between induction regimens based on diabetes type.

The development of dnDSA after pancreas and SPK transplantation has been identified as a significant risk factor for pancreas and kidney rejection, and for graft failure [33–35]. We demonstrated a significantly lower incidence of dnDSA within the first year in T2D versus T1D SPKT recipients. This result may be explained by differences in induction immunosuppression mentioned earlier (i.e., more BAS induction in T1D vs. T2D recipients), and therefore should not necessarily be construed as definitively indicating T2D SPKT recipients would require less intensive immunosuppression or less vigorous postoperative-immune monitoring, though these benefits remain a possibility.

Previous analysis of SPKT registry data from over a decade ago [10] and more recent UK registry data [55] has suggested that the type of diabetes did not significantly impact the rate of surgical complications including abscess formation, anastomotic leak, pancreatitis, and primary non-function. Obesity, frequently associated with T2D, on the other hand, has been associated with increased risk of postoperative infections, a need for postoperative invasive procedures [62, 63], increased risk of patient death, pancreas graft loss, and kidney graft loss [39]. Our findings demonstrate no difference in risks of major surgical complications (bleeding and non-bleeding), surgical site infections, incidental image-identified pancreatic graft thrombotic lesions, and pancreatic graft losses secondary to thrombosis in T2D vs. T1D recipients. In the absence of significantly worse infectious and surgical complications and similar rejection rates between T2D and T1D SPKT recipients, it seems very reasonable to continue to offer selected IDDM/CKD patients an SPKT regardless of their diabetes labels. Prospective trials would also be valuable to definitively compare efficacy and safety outcome endpoints, but await a significant multi-center effort to accrue a sufficient number of patients. In the meantime, we recommend a careful and systematic center-specific approach to offering SPKT to T2D/CKD patients.

Given the rising rates of T2D-associated CKD and obesity, safe criteria for SPKT in the T2D/CKD population should be established [64, 65]. Though we found some marginal protective effect associated with older age with regard to pancreas and kidney rejection, elderly patients tend to preform poorly due to having more comorbidities. UNOS/OPTN policy still requires patients to be insulin-dependent, though weight or BMI restrictions were recently eliminated [44]. Consequently, the indication for SPKT for T2D and CKD at most centers in the US is quite narrow and the majority of T2D/CKD patients presenting to centers are not considered candidates for SPKT but are generally offered a kidney transplant alone. Morbidly obese patients with CKD who do not require insulin most likely have residual beta cell mass, and their diabetes could be reversed by bariatric surgery [66–70]. However, if they have undetectable or minimal C-peptide, their diabetes is unlikely reversed by bariatric surgery alone. CKD patients whose diabetes is controlled by non-insulin oral or injectable agents, diet or exercise are not eligible for pancreas transplantation currently in the US based on allocation policy. However, it is well understood by the transplant community that once they receive a kidney transplant and the requisite immunosuppression, the patient’s diabetes will worsen and ultimately require long-term insulin for control. In this situation, they may benefit from a pancreas-after-kidney transplant, but would it be reasonable to offer a “preemptive” SPKT to this population, preempting their requirement for insulin, just as we offer kidneys preemptively in patients with CKD prior to dialysis? Understanding the relative mortality risk of T2D/CKD waiting list patients who are controlled without insulin to those who are on insulin may support future policy decisions.

We recognize potential limitations to the broader applicability of the results presented here given the non-randomized, single-center, and retrospective nature of our study. Despite using an objective multiparametric approach to classify diabetes type, mis-categorization is possible as not all patients fit neatly into the classically defined T1D and T2D categories, though we believe that this approach is more holistic and objective. We also acknowledge that dnDSA and rejection may still develop after our 1 year minimum follow-up period. Therefore, to ensure valid conclusions can be made, we limited our incidence analysis of immunological, surgical, and infectious complications to the first year or less so that every patient had an equal chance to realize these complications. Consequently, we cannot describe medium or longer-term outcomes relative to these complications. Lastly, our T2D population is relatively small compared to registry data, albeit one of the larger single-center experiences presented to date. However, we feel that the granularity of our data, the recent cohort, and the greater homogeneity of candidate selection, surgical technique, immunosuppression, and post-operative practices at a single center than exists in registry data are benefits to teasing out differences between these populations and to provide updated information. Nonetheless, we believe these data provide useful guidance by comprehensively examining immunological, infectious, and surgical complications after SPKT in T2D recipients.

In conclusion, with the increasing prevalence of T2D related ESRD and an increasing trend of SPKT performed in T2D(9) this study found similar outcomes regarding rejection, major surgical complications, infections, and readmissions between SPKT T1D and T2D recipients. It further demonstrates the success that SPKT can achieve in carefully selected T2D recipients, and provides valuable reassurance to the transplant community for continued careful protocolized application of SPKT to low cardiovascular risk T2D/CKD patients.