Islet dimension and its impact on transplant outcome: A systematic review

Abstract

BACKGROUND

Not all islet transplants desirably achieve insulin independence. This can be attributed to the microarchitecture and function of the islets influenced by their dimensions. Large islets enhance insulin secretion through paracrine effects but are more susceptible to hypoxic injury post-transplant, while small islets offer better viability and insulin independence. In vivo studies suggest large islets are essential for maintaining euglycemia, though smaller islets are typically preferred in transplantation for better outcomes.

AIM

To document the impact of islet dimension on clinical and preclinical transplant outcomes to optimize procedures.

METHODS

PubMed, Scopus and EMBASE platforms were searched for relevant literature up to 9 April 2024. Articles reported on either glucose-stimulated insulin-secreting (GSIS) capacity, islet viability and engraftment, or insulin independence based on the islet dimension were included. The risk of bias was measured using the Appraisal Tool for Cross-Sectional Studies. Extracted data was analyzed via a narrative synthesis.

RESULTS

Nineteen studies were included in the review. A total of sixteen studies reported the GSIS, of which nine documented the increased insulin secretion in the small islet, where the majority reported insulin secretion per islet equivalent (IEQ). Seven studies documented increased GSIS in large-sized islets that measure insulin secretion per cell or islet. All the articles that compared small and large islets reported poor viability and engraftment of large islets.

CONCLUSION

Small islets with a diameter < 125 µm have desired transplantation outcomes due to their better survival following isolation. Large-sized islets receive blood supply directly from arterioles in vivo to meet their higher metabolic demands. The large islet undergoes central necrosis soon after the isolation (devascularization); failing to maintain the viability and glucose stimuli leads to a decline in GSIS and the overall function of the islet. Improved preservation of large islets after islet isolation, enhances the islet yield (IEQ), thereby reducing the likelihood of failed islet isolation and potentially improves transplant outcome.

Key Words: Islet diameter; Transplantation; Islet size; Insulin-secretion; Viability; Engraftment; Insulin independence; Islet transplantation

Core Tip: This systematic review examines the impact of islet size on transplantation outcomes in clinical and preclinical studies. Small islets (< 125 µm) demonstrate superior viability, glucose-stimulated insulin secretion, and engraftment post-transplantation compared to large islets, which suffer from hypoxic injury and poor viability. However, large islets, essential for maintaining euglycemia in vivo, require improved preservation techniques to enhance their post-isolation survival and function. Optimizing islet size and preservation could significantly improve the success of islet transplantation and insulin independence.

INTRODUCTION

Islet transplantation is a promising treatment option for type 1 diabetes mellitus (DM), yet its success rate has not met expectations. In the late 1960s, Lacy et al[1] first isolated pancreatic islets and successfully transplanted them into chemically induced diabetic rodent models[2]. Subsequent experiments focused on refining the islet isolation process, transplantation and administration of immunosuppressive drugs regimen to prevent graft rejection. Human islet transplantation was initiated in the 1980s, with the preferred site being the portal vein[3]. However, in the initial period, due to the unpurified nature of isolated islets and prolonged use of immunosuppressive agents like corticosteroids, adverse reactions, including disseminated intravascular coagulation, portal vein thrombosis and islet cell toxicity, limiting the success rate of islet transplantation[4,5]. By the late 1990s, approximately 450 islet transplantations had been performed worldwide, with fewer than 8% resulting in insulin independence[6]. In 1999, Shapiro et al[7] achieved a significant breakthrough by introducing the Edmonton protocol for islet transplantation, which employed a corticosteroid-free immunosuppressive regimen comprising sirolimus, tacrolimus, and daclizumab. This protocol markedly improved outcomes, with 80% of patients maintaining insulin independence during the one-year follow-up[7].

Despite significant advancements in the last few decades in islet isolation techniques, immunosuppressive regimens, cryopreservation methods, and transplantation sites, challenges in islet engraftment persist[8]. The Edmonton trial identified islet volume as crucial for achieving insulin independence. Islet equivalent (IEQ) is utilized to standardize the calculation of islet volume, representing the mathematical conversion of variable islet sizes and volumes to the standard IEQ of 150 µm diameter[7,9]. The Edmonton protocol recommends a transplantation of 12000 IEQ for optimal outcomes[10,11]. However, it is believed that only 20% to 40% of this IEQ is enough to achieve insulin independence. A significant portion of transplanted islets is lost in the immediate postoperative period due to immune-mediated and hypoxia-induced injuries in the venous compartment[12,13]. Islets, being endocrine tissues with a rich vascular supply, experience a sudden loss of vascular support following isolation and rely on diffusion for oxygen and nutrients. This diffusion may be sufficient for small-sized islets, but those with a diameter exceeding 150 µm (large islets) undergo hypoxic-mediated injury soon after isolation[14]. Thus, the size of the islet, a previously neglected factor, plays a significant role in transplant outcomes, gaining attention in the past two decades[9,15-17].

In vivo, studies on both animal and human pancreas have suggested that large islets are necessary to maintain euglycemia[18-20]. The intermingled architecture of alpha and beta cells within large islets enhances insulin-secreting capacity through a paracrine effect[18]. Notably, Kilimnik et al[19] observed a preferential loss of larger islets in individuals with type 2 DM (T2DM) compared to non-diabetic counterparts. However, in the realm of islet transplantation, conventional wisdom has leaned towards the belief that small islets offer optimal outcomes regarding insulin secretion, viability, engraftment[9,15,17], and achieving insulin independence[21-25].

Thus, our systematic review aims to comprehensively understand the impact of islet size on these critical clinical transplant parameters in animal and human subjects, providing valuable insights into optimizing islet transplantation procedures.

MATERIALS AND METHODS

Literature search

This review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement. The literature search was performed independently by two authors. The search plan was fixed before the beginning. Electronic databases such as PubMed, Scopus, and EMBASE were searched for relevant literature published up to 9^th April 2024 in English. Keywords included in the search were "islets of langerhans transplantation", "islet size", "large islet", "small islet", "isolation", "graft survival", "insulin/biosynthesis", "insulin-secreting capacity", "insulin secretion", "treatment outcome" and "insulin independence" without language restriction. We included multiple truncation terms to widen the literature search. Filters were used to remove articles other than English, and review articles were removed. The search strategy followed in the various search engine were outlined in the supplementary document (Supplementary Table 1).

Assessment of the study eligibility

All the relevant literature identified from the various databases was imported into the Rayyan Software. After removing the duplicates, two independent reviewers screened all the articles with titles and abstracts. Discussed the conflicts with the senior reviewer and sorted the eligibility.

Eligibility criteria

Preclinical studies: (1) Original research articles where human or animal islets are isolated and segregated depending on the size of the islets; and (2) Should have one of the desired outcomes, such as graft survival or insulin-secreting capacity (static or dynamic).

Clinical studies: (1) Original research articles where human or animal islets are isolated and segregated depending on the size of the islets; and (2) Should have one of the desired outcomes, such as graft survival, insulin-secreting capacity (static or dynamic), insulin independence, reversal of diabetes or engraftment.

Exclusion criteria: (1) Studies where the segregation of islets by size was not done; (2) Review article; and (3) Studies without parameters such as graft survival, insulin-secreting capacity and insulin independence.

Data collection process

Data were extracted from the preselected articles that met the selection criteria. Two independent reviewers have extracted the data using the prefixed spreadsheet template containing data items. Inconsistencies in data were subjected to the judgment of a third independent reviewer, and final data were ultimately chosen based on consensus.

Data from the graphs of the included study were extracted using web-based plot digitization software, specifically WebPlotDigitizer Version 4.8, available at https://apps.automeris.io/wpd4/. Graphs were included if they presented relevant data necessary for this systematic review's objectives. The software was calibrated using the scale bar provided within each graph. A known reference point on the graph was used to verify the calibration accuracy. This step ensured that the extracted data were aligned correctly with the original data points presented in the graphs.

Risk of bias assessment

The risk of bias in included studies was assessed using the Appraisal Tool for Cross-Sectional Studies (AXIS)[26]. The AXIS consists of eight questions designed to evaluate the methodological quality of cross-sectional studies. Each question was rated as "yes", "no", "not sure", or "not applicable", and the overall risk of bias for each study was determined based on these ratings. Yes: Indicates a low risk of bias. No: Indicates a high risk of bias. Not sure: Indicates uncertainty about the risk of bias. Not applicable: Indicates that the criterion does not apply to the study design.

Statistical analysis

The outcomes of the study were reported in a narrative synthesis. This included islet dimension-based glucose-stimulated insulin-secreting (GSIS), viability and graft engraftment or insulin independence. In exploring this data, the hypothesis or justification for the better functioning of islets with a particular dimension over others was discussed, along with the measures to utilize the islet of various dimensions effectively for desired outcomes of the clinical islet transplantation. The tables, generated in Microsoft Office-Word, are designed to outline the included studies' characteristics and highlight reported GSIS, viability and graft engraftment or insulin independence of various islet dimensions from the included studies. Figures and graphs are generated using Microsoft Office to summarise the key findings and results of the risk of bias assessments.

RESULTS

Study selection

The comprehensive procedure undertaken for literature searches and study selection is illustrated in Figure 1, following the PRISMA flow chart guidelines. A total of 251 records were identified from databases including PubMed (n = 125), Scopus (n = 32), EMBASE (n = 93) and handpicked (n = 1). After using an automation tool (Rayyan software) to remove 109 duplicate records, 142 studies were screened for title and abstract. One hundred and one (n = 101) studies were excluded based on irrelevant outcomes, inappropriate publication types (such as letters, editorials, reviews, and conference papers), or language other than English. Approximately 41 articles were retrieved for full-text review, of which 40 were reassessed for eligibility based on the inclusion criteria. Twenty-one (n = 21) articles were excluded as they did not discuss islet sizes (small, large, medium) or referenced pseudoislets, IEQs, or pathological studies involving islets. After these exclusions, nineteen (n = 19) studies were included in the current systematic literature analysis.

Figure 1  Preferred Reporting Items for Systematic Reviews and Meta-Analyses (2020) flow diagram of studies identified, screened, excluded and included in the review.

Study characteristics

Nineteen studies were included in this systematic literature review and were summarized in Table 1[9,15-17,21-25,27-36]. Six studies involved human pancreatic tissue from brain-dead donors or integrated islet distribution programs, while thirteen involved tissues from rats (n = 9), goats (n = 1), mice (n = 2), and hamsters (n = 1). Islet isolation was primarily performed using standard collagenase digestion of the pancreas, followed by mechanical or enzymatic separation. Islets were sorted by size using an automated Complex Object Parametric Analyzer Sorter, image analysis (Cell Sans, Saisam; Microvision Instruments) or manual sorting under microscopes. Thirteen studies categorized islets into small and large groups, while six studies used small, medium, and large groups. Small islets ranged from 50 µm to 250 µm, large islets from 125 µm to 400 µm, and medium islets from 150 µm to 300 µm. GSIS was measured under static and dynamic conditions for different islet sizes. In various studies, islet viability has been observed to range from 24 hours to approximately 14 days for both small and large islets. Six of the 19 studies have transplanted isolated islets and assessed insulin independence based on islet size.

Table 1 Study characteristics.

No	Ref.	Tissue	Source of pancreatic tissue sample	Procedure for isolation of islets	Method for islet dimension sorting
1	Farhat et al[22], 2013	Human	IIDP	Islets were maintained in CMRL 1066 medium with 2 mmol/L glutamine, 10% FBS and 1% antibiotic/anti-mycotic at 37 °C in a culture chamber with 5% CO2	Automated separation: COPAS (Union Biometrica). Manual sorting: COPAS not applicable
2	Lehmann et al[9], 2007	Human	Brain dead donors	Mechanical and enzymatic dissociation	Islets were counted, and the diameter of every islet was measured using the image analysis system. (Saisam; Microvision Instruments, Evry, France)
3	Huang and Stehno-Bittel[21], 2015	Rats	Animal	Collagenase digestion method	Manually separated by size based on the criteria
4	Vakhshiteh et al[24], 2013	Caprine (goat)	Animal	Collagenase digestion	Manually-handpicked
5	Su et al[15], 2010	Mice	Animal	Collagenase digestion	Manually sorted under the microscope
6	Kitahara and Adelman[34], 1979	Rats	Animal	Collagenase digestion	Manually under the dissection microscope
7	Sakonju et al[35], 1995	Wistar rats	Animal	Not applicable	Manually under an inverted microscope
8	Williams et al[27], 2012	Human (n = 20)	IIDP and BetaPro (Gordonsville, VA, United States)	Isolated islets were maintained in CMRL 1066 medium with 2 mmol/L glutamine, 10% FBS and 1% antibiotic/anti-mycotic at 37 °C in a culture chamber containing 5% CO2	Automated separation: COPAS (Union Biometrica). Manual sorting: COPAS not applicable
9	Steffen et al[33], 2011	Human	Donors	Collagenase digestion	Manually: Islets handpicked based on size using an eyepiece grid and sorted. Automated sorting: COPAS
10	Komatsu et al[16], 2020	Human	Human	Collagenase digestion	All islets are captured in the micrographs. Islet area measurements were converted into a 2-dimensional circular model using Cell Sens imaging software. Islet diameter calculated assuming it is spherical
11	Zorzi et al[25], 2015	C57/BL10 male mice	Animal	Collagenase digestion	Separated by size into three different groups using stainless steel mesh filtrations by gravity confirmed by light microscopy
12	Huang et al[28], 2011	Rats	Animal	Collagenase digestion	Automated separation: COPAS (Union Biometrica). Manual sorting: COPAS not applicable
13	Cui et al[36], 2005	Syrian golden hamster	Animal	Collagenase digestion	Manual sorted
14	Williams et al[29], 2010	Rats	Animal	Collagenase digestion	Manually sorted. Automated: COPAS (Union Biometrica, Holliston, MA)
15	Li et al[17], 2014	Rats	Animal	Collagenase digestion	Manually sorted
16	MacGregor et al[23], 2006	Rats	Animal	Collagenase digestion	Sedimentation method
17	Hopcroft et al[32], 1985	Wistar rats	Animal	Collagenase digestion	Manually under an inverted microscope
18	Warnock et al[31], 1988	Human	Human	Collagenase digestion	Manually under stereomicroscope
19	Aizawa et al[30], 2001	Wistar rats	Animal	Collagenase digestion	Manually under microscope

COPAS: Complex Object Parametric Analyzer Sorter; IIDP: Integrated islet distribution program.

GSIS

GSIS was reported by 16 (84.21%) out of 19, including studies were summarized in Table 2[9,15-17,21-25,27-36]. Static insulin secretion was reported in eight (50%) studies, perfusion (dynamic) insulin secretion was reported in five (31.25%) studies and both methods were performed in three (18.75%) studies. Nine (56.25%) studies documented the increased insulin secretion in the small islets when compared with medium or large islets out of 88.89% studies measured insulin secretion per IEQ[9,15,17,22-24,27-29]. Seven (43.75%) studies documented increased insulin secretion in large-sized islets[21,30], which measures insulin secretion per cell, islet, or DNA; none of the studies reported or measured IEQ[31-35]. Vakhshiteh et al[24] reported a significant decrease in insulin secretion in large islet after 48 hours of isolation compared to fresh insulin secretion; however, no significant change was observed in small-sized islet following 48 hours of culture.

Table 2 Outcome of the included studies.

Ref.	Population	Dimension (μm)			Insulin secretion (GSIS)-static		Insulin secretion (GSIS)–dynamic		Viability of islets		Transplantation	Insulin independence
		Small	Medium	Large	Small	Large	Small	Large	Small	Large
Farhat et al[22], 2013	Human	< 125	N/A	> 125	Low glucose conc: 0.0035 mIU/cell. High glucose conc: 0.014 mIU/cell	Low glucose conc: 0.0030 mIU/cell. High glucose conc: 0.006 mIU/cell	At 30 minutes (high glucose conc): 0.094 µIU × 10-5/minute/cell. At 60 minutes (high glucose conc.): 0.762 µIU × 10-5/minute/cell. At 120 minutes (low glucose conc.): 0.135 µIU × 10-5/minute/cell	At 30 minutes (high glucose conc): 0.012 µIU × 10-5/minute/cell. At 60 minutes (high glucose conc.): 0.069 µIU × 10-5/minute/cell. At 120 minutes (low glucose conc.): 0.012 µIU × 10-5/minute/cell	N/A		N/A	N/A
Lehmann et al[9], 2007	Human	50-150	N/A	150 and 300	N/A		Peak: 3.098 (fmol/minute)/IE. At 120 minutes: 1.924 (fmol/minute)/IE	Peak: 1.263 (fmol/minute)/IE. At 120 minutes: 1.0353 (fmol/minute)/IE	At 24 hours < 50 μ: 96.8%. 50-100 μ: 94.2%. > 100 μ: 91.2%. At 48 hours < 50 μ: 95%. 50-100 μ: 93.2%. > 100 μ: 93.4%	N/A	N/A	N/A
Huang and Stehno-Bittel[21], 2015	Rats	< 100	N/A	> 200	Low glucose: 0.101 ng/cell. High glucose: 0.099 ng/cell	Low glucose: 0.111 ng/cell. High glucose: 0.138 ng/cell	N/A		N/A		N/A	N/A
Vakhshiteh et al[24], 2013	Caprine (goat)	≤ 150	N/A	> 150	(48 hours) Low glucose: 1.39 ng/IE ± 0.20 ng/IE. High glucose: 2.95 g/IE ± 0.33 ng/IE. Shortly after isolation: Low glucose: 1.84 ng/IE ± 0.22 ng/IE. High glucose: 3.97 ng/IE ± 0.38 ng/IE	(48 hours) Low glucose: 0.48 ng/IE ± 0.20 ng/IE. High glucose: 1.01 ng/IE ± 0.26 ng/IE. Shortly after isolation: Low glucose: 1.69 ng/IE ± 0.20 ng/IE. High glucose: 3.64 ng/IE ± 0.33 ng/IE	N/A		92.46% ± 0.92% (48 hours)	63.16% ± 2.19% (48 hours)	N/A	N/A
Su et al[15], 2010	Mice	< 250	N/A	> 250	30 minutes: 16.90 ng/mL ± 1.15 ng/mL. 60 minutes: 24.5 ng/mL ± 1.50 ng/mL. 120 minutes: 27.66 ng/mL ± 2.52 ng/mL	30 minutes: 11.00 ng/mL ± 1.11 ng/mL. 60 minutes: 13.40 ng/mL ± 0.53 ng/mL. 120 minutes: 16.33 ng/mL ± 1.52 ng/mL	N/A		94.15% ± 1.05% (24 hours )	77.17% ± 1.95% (24 hours )	Yes	The 100% in smaller islets transplanted group. The 62.50% in large islet transplanted group
Kitahara and Adelman[34], 1979	Rats	50-80	N/A	350-400	N/A		2 months of age: 88.5 μunits/islet/2 hours ± 3.7 μunits/islet/2 hours. 24 months of age: 46.1 μunits/islet/2 hours ± 6.7 μunits/islet/2 hours	2 months of age: 464 μunits/islet/2 hours ± 30 μunits/islet/2 hours. 24 months of age: 361 μunits/islet/2 hours ± 29 μunits/islet/2 hours	N/A		N/A	N/A
Sakonju et al[35], 1995	Wistar rats	100-200	201-300	> 300	N/A		At 16.7 mmol/L: 1.05 ng/islet/hour ± 0.26 ng/islet/hour. At 2.8 mmol/L: 0.38 ng/islet/hour ± 0.13 ng/islet/hour	At 16.7 mmol/L: 2.43 ng/islet/hour ± 0.37 ng/islet/hour. At 2.8 mmol/L: 1.13 ng/islet/hour ± 0.29 ng/islet/hour	N/A		N/A	N/A
							At 16.7 mmol/L: 1.73 ng/islet/hour ± 0.36 ng/islet/hour. At 2.8 mmol/L: 0.75 ng/islet/hour ± 0.14 ng/islet/hour
Williams et al[27], 2012	Human	< 125	N/A	> 125	At low glucose: 0.0037 μIU/mL/cell. At high glucose: 0.0153 μIU/mL/cell	At low glucose: 0.0033 μIU/mL/cell. At high glucose: 0.0062 μIU/mL/cell	Peak at high glucose conc: 3.92 μIU/1000 cells	Peak at high glucose conc: 0.43 μIU/1000 cells	92.11% (24 hours), 90.76% (4 days), 87.88% (6 day), 82.5% (9 day)	84.81 (24 hours), 71.92% (4 days), 74.81% (6 days), 43.85% (9 days)	N/A	N/A
Steffen et al[33], 2011	Human	< 150	N/A	> 150	3.3 mmol/L glucose for 1 hour secreted 3.31 g insulin/µg ± 0.82 ng insulin/µg DNA. Upon stimulation with 16.7 mmol/L glucose for 1 hour insulin secretion increased to 5.52 ng insulin/µg ± 1.26 ng insulin/µg DNA for handpicked	large human islets secreted at rest 5.41 ng insulin/µg ± 0.55 ng insulin/µg DNA. After stimulation with 16.7 mmol/L glucose their insulin secretion increased to 10.52 ng insulin/µg ± 2 ng insulin/µg DNA	N/A		Immediately after isolation: 92% ± 1.5%	Immediately after isolation: 91% ± 1.3%	N/A	N/A
Komatsu et al[16], 2020	Human	50-150	150-250	> 250	N/A		N/A		N/A		Yes	Increasing the proportion of large sized islet in transplantation decreases the outcome (poor insulin independence)
Zorzi et al[25], 2015	C57/BL10 male mice	< 150	150–300	> 300	N/A		N/A		N/A		Yes	The 86% of the cases obtained insulin independence when small islets were transplanted. Whereas only 56% and 33% of cases obtained insulin independence when medium and large islets were transplanted
Huang et al[28], 2011	Rats	50	N/A	200	N/A		At 10 minutes: 0.00370 ng/minute/IE. 20 minutes: 0.00327 ng/minute/IE. 30 minutes: 0.00460 ng/minute/IE. Peak: 0.00613 ng/minute/IE	At 10 minutes: 0.00086 ng/minute/IE. 20 minutes: 0.00148 ng/minute/IE. 30 minutes: 0.00204 ng/minute/IE. Peak: 0.00222 ng/minute/IE	N/A		N/A	N/A
Cui et al[36], 2005	Syrian golden hamster	< 100	100-200, 200-300	> 300	N/A		N/A		After 14 days of culture (37 °C). Core cell death: 0%. Core cell death: 2.8%. Core cell death: 14.78%	After 14 days of culture at 37 °C. Core cell death: 33.74%	N/A	N/A
Williams et al[29], 2010	Rats	< 125	N/A	> 150	At 3 mmol/L glucose: 0.729 ng/IE. At 20 nM glucose: 2.017 ng/IE	At 3 mmol/L glucose: 0.156 ng/IE. At 20 nM glucose: 0.342 ng/IE	N/A		After 7 days of culture: 90.39%	After 7 days of culture: 59.85%	Yes	The 80% of the cases obtained independence when small islets were transplanted. None of the cases obtained independence when large islets were transplanted
Li et al[17], 2014	Rats	< 125	N/A	> 150	At 2.8 mmol/L glucose: 21.18 μIU/IE. At 16.7 nM glucose: 69.52 μIU/IE	At 2.8 mmol/L glucose: 16.042 μIU/IE. At 16.7 nM glucose: 47.70 μIU/IE	N/A		120 hours: Cell death is less apparent. Approximately 95% viable	120 hours: Cell death is 50%.	Yes	Better glycaemic control in the group which received small islets when compared to large islets
MacGregor et al[23], 2006	Rats	< 125	N/A	> 150	At basal: 0.262 ng/100 IE/minute. At glucose: 1.288 ng/100 IE/minute. At glucose + high Potassium: 1.976 ng/100 IE/minute	At basal: 0.082 ng/100 IE/minute. At glucose: 0.558 ng/100 IE/minute. At glucose + high Potassium: 0.695 ng/100 IE/minute	N/A		After 24 hours: 95.66%	After 24 hours: 76.59%	Yes	The 80% of the cases obtained independence when small islets were transplanted. None of the cases obtained independence when large islets were transplanted
Hopcroft et al[32], 1985	Rats	80-100	100–200, 200-300	300-400	Islet DNA content: 4.269 ng DNA/islet; 10.676 ng DNA/islet; 20.774 ng DNA/islet; 44.595 ng DNA/islet		< 100-300 μ: 4.272 pmoles DNA/5 minutes ± 0.38 pmoles DNA/5 minutes. > 300 μ: 2.659 pmoles DNA/5 minutes ± 0.34 pmoles DNA/5 minutes. 200-300 μ: 0.107 pmoles/islet/5 minutes ± 0.01 pmoles/islet/5 minutes. > 300 μ: 0.117 pmoles/islet/5 minutes ± 0.02 pmoles/islet/5 minutes		N/A		N/A	N/A
Warnock et al[31], 1988	Human	< 100	N/A	> 100	N/A		Cultured Small at Peak: 0.175 μU/islet/minute. Fresh Small at Peak: 0.078 μU/islet/minute. Cultured Small at 120 minutes: 0.0.081 μU/islet/minute. Fresh Small at 120 minutes: 0.121 μU/islet/minute	Cultured Large at Peak: 0.448 μU/islet/minute. Fresh Large at Peak: 0.405 μU/islet/minute. Cultured Large at 120 minutes: 0.244 μU/islet/minute. Fresh Large at Peak: 0.413 μU/islet/minute	N/A		N/A	N/A
Aizawa et al[30], 2001	Rats	> 200	≥ 200-300	≥ 300	At lower glucose concentrations. Small islet: 0.058 ng/islet/30minute. Medium islet: 0.140 ng/islet/30minute. Large islet: 0.315 ng/islet/30minute. At higher glucose concentrations. Small islet: 0.968 ng/islet/30minute. Medium islet: 1.983 ng/islet/30minute. Large islet: 2.823 ng/islet/30minute		At the end of 10 minutes. Small islet: 29.565 pg/islet/minute. Medium islet: 66.667 pg/islet/minute. Large islet: 110.145 pg/islet/minute. At the end of 30 minutes. Small islet: 64.928 pg/islet/minute; Medium islet: 113.043 pg/islet/minute; Large islet: 160.580 pg/islet/minute		N/A		N/A	N/A

GSIS: glucose-stimulated insulin-secreting; N/A: Not applicable.

Viability of the islet

Viability is crucial for assessing islet functionality and health, ensuring suitability for culture with maximum yield post-transplantation. The timeline in which these articles reported the viability is highly variable, ranging from immediately after isolation to 14 days of culture. Nine (56.25%) articles reported islet viability or cell necrosis. Lehmann et al[9], 2007 reported the viability of small islets without comparing them with the large islet. All eight studies that compared small and large islets reported a reduction in the viability of large islets either in culture or cryopreservation. Viability assays were measured using Sytox and calcein fluorophores, fluorometric assays, or staining with fluorescein diacetate and propidium iodide to distinguish between live and dead cells within islet preparations. Williams et al[29] found no significant improvement in insulin secretion or content with the viability assay but noted that removing the diffusion barrier enhanced cellular viability without increasing insulin secretion. Hopcroft et al[32] utilized a modified fluorometric assay for islet DNA. William et al[29] reported fewer dead cells in porous islets with a similar percentage of live cells as in intact small islets. Li et al[17] quantified viable islet cells using the colorimetric-based cell counting kit 8. Cui et al[36] employed the terminal deoxynucleotidyl transferase dUTP nick end labeling assay to identify apoptosis in isolated islets and found that culturing under a lower temperature prevented core damage. Vakhshiteh et al[24] used the Annexin assay to report 5.21% apoptotic death in small islets compared to 7.34% in large islets.

The eight studies' mean viability of small and large islets is 93.62% ± 2.26% and 71.11% ± 13.68%, respectively. The large islets are reported to have been less viable when compared to the small islets (P = 0.003). The features of central necrosis in large islets were evident soon after the isolation.

Insulin independence (engraftment)

Insulin independence or engraftment is usually defined as the state in which the transplanted human or animal will remain insulin-independent within the specific time frame of the islet transplantation. It is the most essential desired clinical outcome that depends on the GSIS and viability of the small and large islets. Six (37.5%) articles reported insulin independence post-transplantation. All the (100%) articles reported the superiority of small islets in terms of better engraftment. Out of which, five articles (83.33%) were performed with small and large in different groups. Whereas one (16.67%) study was conducted with varying proportions of small, medium or large islets, which shows a large proportion of small-sized islets in the transplanted islet volume leading to better clinical outcomes[16].

Risk of bias assessment

The risk of bias was assessed for various components related to islet research based on predefined criteria. The assessment outcomes across the included studies were summarised in the Figure 2. The majority of studies demonstrated clear definitions for islet size (16 studies), adherence to standard isolation procedures (17 studies), clearly outlined islet size sorting procedures (16 studies) and defined protocols for GSIS (15 studies) indicating robust methodological practices.

Figure 2 Appraisal Tool for Cross-Sectional Studies. A: Risk of bias assessment of the included studies; B: Summary of risk of bias assessments. GSIS: Glucose-stimulated insulin-secreting.

However, challenges were noted in several areas. GSIS measurement and expression in IEQ showed variability, with only eight studies achieving clarity, while eight studies exhibited unclear or inconsistent reporting. Similarly, viability measurement of islets under optimal conditions lacked clarity in a significant proportion of studies (4 studies unclear). Only three studies clearly defined criteria for engraftment or insulin independence, with three presenting unclear criteria. The absence of a high bias rating across all assessed components indicates a generally acceptable level of methodological quality in the included studies.

DISCUSSION

Islet size remains constant across rodents and humans with unique cellular arrangements

Islets are small micro-organs ranging from a few small clusters to hundreds of cells (800 cells to 3000 cells per islet)[37], roughly contributing to 1% to 2% of total pancreatic cells. Islets are composed of three major cell types: (1) Alpha (30% to 50%); (2) Beta (50% to 60%); and (3) Delta (approximately 5%) cells[38,39]. Other minor endocrine cells include pancreatic polypeptide and Epsilon cells, which contribute less than 1% of the islet volume[38]. The size of the islet doesn't increase with the organism's size; it's almost constant and maintained within the particular limit from the rodent to human[18,40]. The arrangement of cells within the islet varies from animal to human or monkey. In the lower animals, the beta cells are confined to the core, and the alpha and delta are arranged in the peripheral mantel zone[18,41]. However, in humans the size matters. In the small islet, the cellular arrangement resembles that of the rodent[41]. However, the larger islets show an intermingled arrangement of alpha and beta cells with a marginal increase in alpha-beta ratio[19,40,42]. Several authors hypothesize that the intermingled arrangement within larger islets enhances alpha-to-beta cell contact, thereby boosting the insulin-secreting capacity of beta cells through the paracrine effect (Figure 3)[18,43]. This intermingled cytoarchitecture is also reported in rodents with higher insulin demand, such as obesity and pregnancy[18,41]. Subsequently, a cascade of changes in cellular arrangement ensues, serving as a compensatory mechanism to further enhance insulin secretion, indicative of a finely regulated process in response to metabolic needs. Studies documented that the reduction in the number of larger islets in the T2DM compared to healthy controls supports the above hypothesis[19].

Figure 3 Characteristic differences in small and large islets play a significant role in their survival, function and transplant outcome. GSIS: Glucose-stimulated insulin-secreting; IEQ: Islet equivalent.

Contrary to the above finding[9,15-17], studies on islet transplantation have shown that a higher proportion of small islets in the transplanted tissue yields more promising results in achieving insulin independence than a higher proportion of large islets[23,25,29]. In vitro islet studies reported the superiority of small islets compared to large islets in insulin-secreting capacity, viability, and post-transplant insulin independence (Table 2)[9,15-17,21-25,27-36].

A smaller proportion of large islets contributes to a higher volume of the islet

Large islets have been reported to be in lesser proportion than small islets. Despite that, the larger islets contribute to most of the total islet volume. Sakonju et al[35] reported that 43% of the total islet volume is contributed by 16.7% of the islet with a diameter of > 300 µm. Similarly, Bonner-Weir[44] reported that small islets (< 160 µm) constitute nearly 75% of the total islet count but account for only approximately 15% of the total islet volume. Conversely, large islets (> 250 µm) make up approximately 15% of the total islet count but represent approximately 60% of the total islet volume[35,44]. Therefore, effectively utilizing the large islet by successful cryopreservation or culturing methods under optimal viable conditions would enable us to overcome the islet shortage with increased yield[25,35]. Further standard islet volume parameter measured during transplantation (IEQ) is calculated using islet diameter based on the assumption that the islets are perfectly spherical[45]. However, several studies have shown that islets are ellipsoidal using the circularity index[46]. Especially, large islets are predominantly ellipsoidal, with the circularity index ranging from 0.6 to 0.74[47]. In addition, the subjective nature of islet binning with the 50 µm increment leads to an overestimation of IEQ up to 50%[48,49]. Our unpublished data also indicates that the overestimation of IEQ is significantly higher when the islet size is larger. Transplanted islet volume in terms of IEQ is the crucial parameter determining the transplant outcome. An increased proportion of large islets may result in an overestimation of IEQ, leading to unfavourable transplant outcomes due to the actual lower volume of islets[16,46].

Large islets meet their metabolic demand via arterioles rather than capillaries

Islets are highly vascularised structures that receive 20% of the total pancreatic blood supply even though their volume corresponds to 1% to 2% of total pancreatic volume (Figure 4E)[39,50]. Studies documented that islets with direct contact with the arterioles tend to be large-sized and fewer in number than smaller ones[51,52]. Therefore, a large-sized islet with an intermingled arrangement essentially requires an increased blood supply to function. This observation offers valuable insight into potential factors contributing to the suboptimal performance of large islets in transplantation studies[16]. Since the islets were isolated from the pancreas, they abruptly cut off from the vascular supply. Limited O₂ diffusion in the isolated islet, especially in the large islet, induces central necrosis, leading to decreased functionality[53,54]. Studies show that a transplanted human islet requires three days to regain vascularisation and takes 10 days to 14 days to complete[39,55,56]. The large islets exhibit features of cell death soon after isolation, more so than small islets[14]. This might be one of the reasons that more than 50% of the transplanted islets are lost immediately due to immune-mediated injuries and hypoxia[12].

Figure 4 Isolated human islets at various stages of assessment before transplantation. A: Unstained isolated human islets of different sizes (S: Small-sized islet; L: Large-sized islet); B: Dithizone (Diphenylthiocarbazone)-stained human islets, freshly isolated, showing the purity of the preparation; C: Unstained normal human islet under higher magnification; D: Unstained human islet with central necrosis after culture for 5 days, viewed under higher magnification; E: Haematoxylin and Eosin-stained image of large-sized islet in native pancreas with multiple intra-islet capillaries (black arrowhead) (higher magnification); F: Fluorescein diacetate/propidium iodide staining to assess the viability of the isolated islets (Bright green fluorescence indicates live cells, while bright red or orange fluorescence indicates dead or necrotic cells).

Cultured islets exhibit enhanced transplantation success compared to fresh islets

In the early period of Edmonton protocol, the islets were transplanted soon after isolation, i.e., two hours to four hours (fresh islets) (Figure 4A and C)[7,57]. Later studies documented the beneficial effect of the cultured islet over a fresh islet in terms of reduction in the pro-inflammatory markers when the islets were cultured under specific conditions[58,59]. The culturing conditions allow the islets to sensitize to the hypoxic environment, leading to a reduction in the hypoxic injury post-transplantation[59,60].

Cultured islets provide better insulin independence when cultured for 12 hours to 24 hours and currently, islets are cultured for 36 hours to 72 hours post-isolation to assess the viability and its functionality[59,61-63]. In general, it is recommended that, the highly purified islet can be cultured at 37 °C with 5% CO₂ for the first 12–24 hours and at 22 °C with 5% CO₂ for the remaining time. The lower temperature is advised when the purity level of the islet is low[64]. Culture at a lower temperature (26 °C) was advocated to minimize the central necrosis in larger islets compared to culture at 37 °C in animal islets[36]. Even culturing the islets with certain islet basement membrane proteins like Collagen-IV, Laminin-521 and Nidogen-1 have reduced the pro-inflammatory mediators with overall improvement in the islets' viability, structural and functional integrity[65]. Brandhorst et al[65] reported that using islet basement membrane protein in cultures almost always rescued 90% of suboptimal pancreatic islet isolation to proceed with the clinical islet transplantation.

The diffusion barrier is the most crucial determinant of viability in larger islets post-transplant

Recent studies reported the size-dependent viability of islets. Small-sized islets were reported to have better survivability than large islets (Table 2, Figure 4B)[9,15-17,21-25,27-36]. The diffusion barrier is the most crucial factor that decreases the viability of larger islets since post-isolation islets solely depend on diffusion[24,27,29]. The Diffusion barrier of both small and large human islets are two to three times stronger than the rodent islets, allowing only substances up to 10 kDa[27,29]. The diffusion barrier limits two critical substances, oxygen and glucose, essential for cell viability and GSIS, respectively. The oxygen consumption of a large islet was limited because of the diffusion barrier and documented as less than 45% to 50% of that of a small islet per IEQ[23]. The glucose diffusion rate of the islet was measured with fluorescent-labelled glucose [2-(N-(7-nitrobenz-2-oxa-13-diazol-4-yl) amino)-2-deoxyglucose], showing a 33% reduction in the diffusion rate of the large islet when compared to the small islet. Williams et al[29] reported, glucose never reaches the core of the large islet when the diameter is more than 150 µm. Barriers also limit the excretion of metabolic waste from the islet. The above phenomenon explains the lower viability and GSIS of large islets post-isolation. Damaging the islet barrier by over-digestion leads to increased viability of islet cells[29]. However, it doesn't increase the GSIS, probably due to the inherent ability of large islets or disruption of the islet micro-environment[14,25,27,29].

Small islets secrete more insulin per IEQ, while in large islets, diffusion barriers and cell death reduce insulin secretion, even affecting the paracrine effect of glucagon-like peptide 1 in GSIS

Glucose acts as a stimulus for insulin release and activates the cascade for insulin secretion by the transcription of new insulin mRNA and the translation of existing insulin mRNA into proinsulin[66]. Increased intracellular glucose levels activate the insulin gene transcription factors via peroxiredoxin 1, NeuroD and musculoaponeurotic fibrosarcoma oncogene homolog A, which transcribe pre-pro-insulin mRNA[21,66]. The immediate effect of GSIS starts with insulin release and translation of existing pre-pro-insulin to proinsulin synthesis. Finally, the cleavage of c-peptide from the proinsulin leads to insulin synthesis[21]. The long-term effect of GSIS starts with the formation of new pre-pro-insulin mRNA[67]. Higher glucose concentration increases the large islet proinsulin level per cell more than the small islet, with no significant changes in the final insulin secretion[21]. Thus, Huang and Stehno-Bittel[21], hypothesized it as either large islet has better potentiality to convert the proinsulin mRNA into proinsulin or poor potentiality to convert further it into insulin[21].

Studies exploring GSIS on islets of varying sizes are relatively few in the literature. In addition, the heterogeneous methodologies and reporting units make comparison and analysis difficult. The common finding[9,15,17,22,23] in the available literature is that the small islets secrete more insulin when an equal volume of islets is taken in both groups, i.e., reported as insulin secretion per IEQ[24,27-29]. However, when the studies where insulin secretion were reported in the unit per cell or islet[21], large islets seemed to secrete more insulin[31-34,68]. When the insulin secretion was documented per IEQ, equal volumes of small and large islets were included, i.e., more number of small islets compared with equal volume of larger islets (less in number). This nullifies the volume difference, but in large islets, there is the central core of cell death, which decreases the number of viable cells in the large islets group (Figure 4D and F). This discrepancy in cell viability might contribute to the variability observed in the results.

Glucose is the essential stimulus for the insulin secretion. As described earlier, the diffusion barrier in the large islet slows down the glucose perfusion[27,29]. It never allows glucose to reach the core, leading to decreased stimulation of cells of large islets[29]. In addition, central cell death further causes a decline in insulin secretion. Even alpha cells enhance the GSIS of beta cells via the paracrine effect by releasing glucagon-like peptide 1[38]. Maintaining the microarchitecture of the islet is necessary to preserve the paracrine effect. Thus, the porous small or large islet failed to improve the insulin-secreting capacity[27]. Over-digested islets might lose the cells in the periphery, disrupting the microarchitecture of islets and affecting the paracrine effect. Thus, the viability of the over-digested islets increases, but not the GSIS[27,29].

As the age advances, the GSIS of the islet decreases, and reduction is more predominant in small islets than in large islets[34]. Kitahara and Adelman[34] reported that an increase in large islets during ageing is the compensatory mechanism to tackle the decreased insulin secretion of the small islets.

Small islets excel in viability and GSIS, leading to better engraftment and insulin independence; prompt circulation restoration post-isolation may boost larger islet success

Achieving insulin independence post-transplantation is the most desirable outcome, which is better when an increased proportion of small islets are transplanted[9,15-17], which is directly linked with the better viability and GSIS of smaller islets[23,25,29]. In the above two scenarios, a small islet functions better, leading to better engraftment of the islet and achieving insulin independence.

The isolated small islets reported to have better insulin secreting capacity since it can survive better with the diffusion. For better comparison of insulin secretion between the small and large islets, we need to develop the protocol where the blood circulation of the large islets should be restored soon after the isolation which might provide the higher success rate post-transplantation.

In vivo studies documented (1) Large islets have better insulin secreting capacity even as the lower glucose concentration via paracrine effect[18,43]; (2) In T2DM there is loss of large islets[19]; and (3) Intermingled architecture of large islets were observed in animal to overcome insulin demand in the conditions such as obesity and pregnancy[18,41]. The above findings justify the importance and functionality of large islets. The majority of the islet volume (75%) is contributed by the smaller proportion of the large islets (25%); in other words, 10% of large islets houses half of the total beta cell volume[41,50]. Thus, the optimal way of utilization of large islets solves the shortage of islet for transplantation and improves the chance of engraftment.

Small islets contain more cells than large islets per IEQ

Apart from the calculation error in IEQ, it is necessary to know the number of cells contributing to the volume. We consider one IEQ to contain a fixed number of cells irrespective of the islet dimension. However, studies show that small islets (50 µm) (83 cells per islet) contain 2235 cells per IEQ [approximately requires 27 small islets (50 µm) to make one IEQ], whereas large islets (2071 cells per islet) (200 µm) contain 874 cells per IEQ (because one 200 µm large islet is equal to 2.37 IEQ)[28]. Similarly, Huang and Stehno-Bittel[21] reported that the traditional IEQ overestimates the volume of the large islet by 100% compared to their actual islet cell counts[47]. This is possible since the large islet contains a rich intra-islet capillary network, which is comparatively less in the small islet. Thus, in terms of cells, small islets possess higher cells per IEQ.

Cell counts are considered a more reliable marker for predicting the success of the transplant outcome than the IEQ. An animal study was conducted with three dosages of IEQ (2500 IEQ, 3500 IEQ, and 4000 IEQ), and it shows no correlation between IEQ and favourable transplant outcomes. However, when the volume is calculated as the number of cells per mouse, irrespective of the IEQ, the mice that received a cell count of more than 5.13 M have a favourable outcome[46]. In larger islets, both IEQ and cell counts tend to be overestimated. This combined overestimation leads to a reduction in the volume of islet cells actually transplanted when the islet size index is ≥ 1. To address this issue and minimize the risk of underestimation, it is advisable to adjust the IEQ to the upper range when the islet size index is ≥ 1, thereby compensating for the potential overestimation of islet volume. The lower islet size index is reported to have better transplantation outcome in terms of the insulin independence[69].

Navigating large islet challenges with innovative solutions

Various strategies have been hypothesized to improve the vascularity of the islets in the literature including (1) Co-culturing or co-transplantation with other cell types such as mesenchymal stem cells, fibroblast, Endothelial cells and endothelial progenitors[56,70,71]; (2) Release of pro-angiogenesis factors[72]; (3) Pre-vascularisation of transplant site[56,70]; (4) Use of scaffolding materials promoting vascularisation[56,73]; (5) Co-encapsulation of islets with others cellular types or with metalloproteins[70]; (6) Culturing with mesentery[74]; (7) Three-dimensional bioprinting of islets with capillaries (organoids)[75]; and (8) Use of pharmacological agents[70]. Experimentally, most of these methods show promising results; however, the long-term effects of these drugs or organoids or scaffoldings were uncertain.

Even after isolation, islets have a capillary channel in them[76]. The immediate action to restore the blood flow will help us keep the large islets viable. We hypothesize that creating a pressure difference across the islet soon after isolation could establish the blood flow within the islets, improving the islet's viability, preserving the microarchitecture of islets and helping in faster vascularisation (Figure 5). Several other researchers have used a similar principle to maintain perfusion by regulating pressure differences in tissues like the heart, kidney, and limbs[77-81]. Also, islets can be transplanted into a similar environment with an existing pressure difference in the blood flow, like an Arterio-venous fistula, to ensure maximum blood flow. However, a unique encapsulated device is required to hold it in position. The islet isolation technique needs to be optimized to be more friendly to the intra-islet capillary endothelium, which helps establish the blood flow sooner.

Figure 5 Strategies to maintain the viability of large islet. A: Creating the pressure difference across the islet might increase the blood flow in the capillaries present in the isolated islets, keeping it viable and accelerating the vascularisation; B: Magnified view showing single islet and direction of the flow. Arrow shows the direction of the blood flow. H: High pressure; L: Low pressure.

Effective utilization of large islets enables us to (1) Maximize the islet yield since only a small number of large islets constitute a larger volume, providing higher IEQ for transplantation. Thus, it will reduce the shortage of islets; (2) Reduce the immediate postoperative loss of the islet due to hypoxia-mediated injury; and (3) Increase the period of post-transplant insulin independence because of its paracrine effect.

This systematic review has certain limitations that are important to consider. First, the included studies exhibit considerable heterogeneity in islet size classification, measurement techniques for GSIS and viability, and outcome reporting, which limits direct comparability and synthesis of findings. Due to this heterogeneity, a meta-analysis was not feasible, and results were instead synthesized narratively. Second, despite including all the articles from various databases, the majority of the included studies originate from specific geographical regions, which may limit the generalizability of the findings to broader populations. Few studies that represented data in graphs were extracted using WebPlotDigitizer, which may have introduced minor variations in the data. However, these data were utilized solely for narrative synthesis. We incorporated the known point on the scale bar of each graph after calibration to confirm the accuracy of the data extracted. Furthermore, the exclusion of non-English studies introduces a potential language bias, as relevant research published in other languages may have been overlooked. We recommend conducting further research on islet dimension-related studies by incorporating enhanced islet preservation techniques with standard islet dimensions of < 125 µm (small), 125 µm to 300 µm (medium), and > 300 µm (large) islets to elucidate the impact of islet dimension on transplantation outcomes. Additionally, the GSIS capacity reported per IEQ will be recommended for improved comparison with subgroups and other studies. Future research should prioritize validating these findings by analysing clinical data from transplant centres. This includes assessing the proportion of transplanted islets across various dimensions, the total IEQ transplanted, and transplantation outcomes, particularly in terms of insulin independence.

CONCLUSION

Post isolation, small islet (less than 125 µm) has shown better survivability, insulin-secreting capacity, engraftment and insulin independence on transplantation. They survive better on diffusion with a comparatively lesser cell volume per islet. Thus, smaller islets provided better transplant outcomes in in vitro studies and clinically. Nonetheless, in vivo studies demonstrate that large islets (greater than 150 µm) are manufactured to handle increased insulin demand, which consequently necessitates increased blood flow to match their metabolic requirements. Yet the diffusion barrier restricts the glucose to reach the core of the larger islet resulting in poor performance due to acute disruption in their blood supply. Therefore, restoring blood supply shortly after isolation would enhance the utilization of large islets, increasing the islet yield. This would also reduce the need for a higher IEQ required for transplantation. Moreover, the error in IEQ calculation which depends on cell count per IEQ would be lower for larger islets. Additionally, large islets, designed to handle increased insulin demand, may also have greater longevity in terms of achieving insulin independence. To sum up, while smaller islets generally performed better, the prospective of large islets cannot be dismissed if utilized effectively.

ACKNOWLEDGEMENTS

We extend our deepest gratitude to the courageous families who generously donated their loved one’s organs and tissues for biomedical research. Such important research like this would not be possible without this selfless gift of hope. Thanks to Network for Hope (Louisville and Cincinnati) and Lifeline of Ohio, Columbus for supporting these special families and providing human research pancreases to Dr. Balamurugan's islet lab. We also thank members of Dr. Balamurugan's islet lab for their contributions to islet research and transplant and thanking funding agencies.