Intraportal Islet Oxygenation

Abstract

Islet transplantation (IT) is a promising therapy for the treatment of diabetes. The large number of islets required to achieve insulin independence limit its cost-effectiveness and the number of patients who can be treated. It is believed that >50% of islets are lost in the immediate post-IT period. Poor oxygenation in the early post-IT period is recognized as a possible reason for islet loss and dysfunction but has not been extensively studied. Several key variables affect oxygenation in this setting, including (1) local oxygen partial pressure (pO₂), (2) islet oxygen consumption, (3) islet size (diameter, D), and (4) presence or absence of thrombosis on the islet surface. We discuss implications of oxygen-limiting conditions on intraportal islet viability and function. Of the 4 key variables, the islet size appears to be the most important determinant of the anoxic and nonfunctional islet volume fractions. Similarly, the effect of thrombus formation on the islet surface may be substantial. At the University of Minnesota, average size distribution data from clinical alloislet preparations (n = 10) indicate that >150-µm D islets account for only ~30% of the total islet number, but >85% of the total islet volume. This suggests that improved oxygen supply to the islets may have a profound impact on islet survivability and function since most of the β-cell volume is within large islets which are most susceptible to oxygen-limiting conditions. The assumption that the liver is a suitable islet transplant site from the standpoint of oxygenation should be reconsidered.

Islet transplantation (IT) remains a promising therapeutic option for the treatment of diabetes. This is primarily because it has the potential to permanently reverse the hyperglycemic state without involving major surgery. The most experienced centers currently report insulin independence rates of approximately 50% at 5 years post-IT, which are comparable to that of pancreas transplant.¹ However, to achieve insulin independence following IT, islets from 2-3 pancreata are usually required. This is problematic because donor pancreas supply is already limited and multiple islet transplants increase cost and risk to the patient. Consequently, IT is currently only offered to a small subset of patients with hypoglycemic unawareness. Based on our understanding that the human pancreas has a significant insulin-producing reserve,² it should not require multiple donors to obtain adequate amounts of islet tissue to achieve diabetes reversal.

It is believed that >50% of intraportally transplanted islets are lost within the early post-IT time frame.³ However, it is unclear which mechanisms are involved and whether they are primarily immunologic or nonimmunologic. Many investigators agree that the liver may not be the optimal site for islet survival or function and that inadequate oxygen supply may be an issue.^4-7

Reasons for why the liver may not represent the optimal site include the following:

1. Instant blood-mediated inflammatory reaction (IBMIR),which consists of thrombus formation on the islet surface, complement-mediated islet cell lysis, and a local inflammatory response⁸

2. Immunosuppressant toxicity,⁹ which may be exacerbated by higher drug levels in the portal venous blood¹⁰

3. Poor reestablishment of extracellular matrix connections, which may exacerbate islet cell death by anoikis¹¹

4. Difficulty with tracking or monitoring the islet graft,¹² which complicates early detection of rejection¹³

5. Poor islet oxygenation¹⁴

Intraportal islet oxygenation in the very early post-IT time period is generally underappreciated by the field, possibly because of the assumption that direct contact with the blood stream should provide an adequate oxygen supply. Data exist indicating that islets in the liver experience poor oxygen supply and may not completely vascularize.^15-17 Olsson et al used pimonidazole, an oxygen-sensitive intracellular dye that accumulates under conditions with pO₂ < 10 mmHg, to illustrate that ~70%, ~60%, and ~30% of intraportally transplanted syngeneic murine islets stained positive for reduced oxygenation at 1 day, 1 month, and 3 months post-IT, respectively.¹⁶ Carlsson et al used modified Clark microelectrodes to measure the pO₂ in an islet transplanted under the liver capsule to be ~5 mmHg at 9-12 weeks.¹⁵ Of note, they did not take any early post-IT measurements. Furthermore, microelectrode measurements only provide pO₂ values for a single point in space (and no information regarding oxygen gradients, which can be 15-25 mmHg within a fully viable and oxygenated 150-µm diameter [D] non-revascularized transplanted islet) and these measurements may be inaccurate due to their susceptibility to stray current artifact. Nevertheless, in the same study, the authors measured the pO₂ of a native islet to be ~40 mmHg, which suggests that there may be at least a large discrepancy in the oxygenation of the native pancreatic islet versus the transplanted islet in the liver. Even considering the results of Olsson et al¹⁶ or Carlsson et al,¹⁵ it is unknown what is the pO₂ at the site of intraportal IT, particularly during the early post-IT time period.

There are major differences in the route by which oxygen is delivered when comparing the native and intraportally transplanted islet prior to it becoming revascularized. The native islet is perfused directly with oxygenated arterial blood, whereas the intraportal islet is not perfused at all and relies on oxygen diffusion from its surface (Figure 1). Furthermore, most native islet cells are located within 1 cell D (~10-15 µm) away from the nearest capillary,¹⁸ but many cells located within the core of a larger islet may be >200 microns away from the nearest oxygen source located at the surface of an islet (or even further away from the islet surface in some cases), which itself is believed to be characterized by low pO₂.

Figure 1.

Sketch illustrating an islet in the native pancreas (left) and an islet that has been intraportally transplanted (right). The native islet is well perfused with arterial blood, whereas the transplanted islet obtains oxygen from the surrounding intraportal blood and relies solely on oxygen diffusion from its surface before it becomes revascularized and possibly indefinitely if not revascularized well.

Effect of Poor Oxygenation on Islet Survival and Function

Adequate oxygenation is important for the survival and healthy functioning of pancreatic islets. It is well-known that β-cells are poorly equipped to handle hypoxic conditions.¹⁹ Under conditions of low pO₂, insulin secretion is reduced significantly. Dionne et al presented data on perifused rat islets showing that the insulin secretory rate decreased to 50% and 2% of normoxic values at 27 and 5 mmHg bulk perfusate pO₂, respectively.²⁰ Even if conditions do not result in the formation of an anoxic core, there may be a large region of the islet that remains viable but is poorly functioning or nonfunctioning. There are recent data indicating that early hypoxic exposure may result in a persistent change in the human islet cell gene signature and longer-lasting loss of glucose-stimulated insulin secretion.^21,22 This becomes even more important if the islet does not completely revascularize, since this would mean that the islet would rely on passive oxygen transport for several days or weeks and possibly indefinitely. It is generally believed that revascularization takes 7-10 days.²³ However, there is evidence indicating that intraportal islets revascularize poorly. Following work previously done by Carlsson et al on human islets transplanted under the kidney capsule,²⁴ Lau et al showed that the human islets transplanted via the portal vein exhibited poor vascular density at 1-month post-IT as compared to the native pancreatic islet.¹⁴ The authors found that revascularization in human islets was similar when comparing transplant sites (intraportal vs renal subcapsule),¹⁴ which has implications for the design of a strategy to provide long-term adequate oxygenation to an islet graft at an extrahepatic site.

Oxygenation of the Intraportally Transplanted Islet

From the standpoint of oxygenation, islets transplanted into the liver can find themselves in very different circumstances. Some islets may experience adequate oxygenation, whereas some may not. As discussed, oxygenation has profound implications on islet survival and function.

Several key variables affect the early oxygenation of the intraportal islet and they include (but are not limited to) the following:

1. Local pO₂ at the site of engraftment

2. Islet tissue oxygen consumption rate (OCR)

3. Islet size (D)

4. Presence or absence of thrombus on the islet surface

The importance of local pO₂ is intuitive but the pO₂ required to provide adequate oxygenation is dependent on the other variables. Islet OCR represents the oxygen sink and the size of this sink is defined by the size (or D) of the viable islet cell cluster. Thrombus formation on the surface of the islet represents an additional oxygen transfer resistance. The best-case scenario would be an islet of small D with low OCR, under conditions of high pO₂ and no thrombosis. Figure 2 depicts this ideal case and cases in which these key variables are unfavorably modified within a realistic range of values to show the formation of an anoxic core. The relative sizes of these anoxic cores can be estimated using previously derived mathematical theory.²⁵ The worst-case scenario could be characterized by the combination of large islet D, high OCR, low pO₂, and thrombosis to result in very poor oxygenation and a large anoxic core that involves most of the islet volume. It must be noted that the OCR would be lower in the case there is more nonviable tissue. Although low OCR is advantageous in terms of oxygen availability to transplanted islets, it is not desirable to have low OCR since it is indicative of poor fractional viability. The lower the OCR, the higher is the fraction of nonviable islet tissue. The nonviable tissue represents an oxygen diffusion barrier and a substrate that triggers or enhances an immune reaction against the graft (in the case of allo- or xeno-IT). Each individual islet will experience unique circumstances following intraportal transplant, and thus may experience very different oxygenation when compared to another islet. It is impossible (and unnecessary) to predict the nature of the local microenvironment surrounding each individual islet but we can begin to understand the factors most likely to influence islet graft survival and function. Apart from the local oxygen supply, which is clearly important, islet size (or D) may be the most important variable affecting oxygenation since the ΔpO₂ from the surface to the center of a fully oxygenated spherical islet exhibits D² dependence. Most of these key variables affecting oxygenation are nonmodifiable, making critical the thorough evaluation of islet oxygenation in the liver. Other variables that would negatively affect islet oxygenation but whose impact is difficult to quantify include islet clumping, leukocyte infiltration (inflammation), and the presence of cotransplanted acinar tissue. Islet clumping, or clustering of multiple islets to form a larger islet tissue aggregate, is particularly problematic from the standpoint of oxygenation and it can be thought of as an increase to the effective islet size. For example, most of a large islet tissue aggregate with a D of 1000-µm is likely to be anoxic. Inflammation and acinar impurity represent additional oxygen sinks which make less oxygen available to islets, but may also contribute to islet cell loss and dysfunction via other mechanisms (eg, release of proteolytic enzymes, immune rejection). It may be that IT into an extrahepatic site (intraperitoneal, subcutaneous, intramuscular) could provide the best opportunity to adequately oxygenate an entire islet graft. However, it is unclear which extrahepatic site could inherently provide the best native oxygen supply, and even these sites would likely require enhanced early oxygenation via engineered prevascularized chambers,^26-28 oxygen-generating biomaterials,²⁹ a tethered oxygen reservoir or supply,^30-32 or an in situ oxygen-generating device.³³ A key advantage of certain extrahepatic sites is that the entire graft could be localized in a single location (rather than dispersed throughout a large organ like a liver), enabling directed oxygen delivery. Some of these approaches could provide the additional benefit of immunoisolation which would eliminate or reduce the need for immunosuppression, and may even enable islet xenotransplantation. If fully oxygenating the islet graft is not possible, perhaps the islets can be protected from the effects of hypoxia³⁴ as part of a strategy to preserve graft survival.

Figure 2.

Schematic depicting the effect of different variables on intraportal islet oxygenation and anoxic core formation. The ideal case in terms of oxygenation involves a small diameter (D) islet with minimal metabolic demand (or low oxygen consumption rate [OCR]), high local oxygen partial pressure (pO₂) (>40 mmHg) and no thrombosis. Unfortunately, most cases are likely to be worse in terms of oxygenation and involve unfavorable changes to some of these key variables. Cases 1 to 4 depict approximate and relative sizes of the anoxic core that would form if 1 of the key variables were unfavorably adjusted within a realistic range of values. Case 1 involves a small D islet under conditions of low pO₂ (<10 mmHg). Case 2 involves a small D islet with a high OCR. Case 3 involves a small D islet with formation of a thrombus on the surface of the islet. Case 4 involves a large D islet. Finally, if all variables are affected unfavorably, then most of the islet would be anoxic; the worst case would involve a large D islet with high OCR, low local pO₂, and thrombus formation on its surface.

Importance of Islet Size Distribution in Clinical Preparations

As indicated, the islet size (or D) may be a particularly important determinant of adequate oxygenation. In fact, there are data already published indicating that small islets perform better than large islets following IT.^35,36 Fujita et al also showed that the amount of insulin secreted per islet equivalent in response to a glucose stimulus was significantly higher in smaller versus larger islets,³⁷ which may reflect differences in oxygenation. All of these data support the hypothesis that the insulin-secreting β-cells located within the core of larger islets may be nonviable or dysfunctional.

These concepts can be appreciated on the scale of an entire clinical islet preparation. When reviewing actual islet size distribution data from high-purity, postculture clinical alloislet preparations produced at the University of Minnesota from February 2011 to January 2012 (n = 10), it was found that the majority of islets exhibit small D (Figure 3). However, if the islet numbers are converted to islet volumes, based on their approximate D and assuming a spherical islet shape, the vast majority of the transplantable islet tissue volume is attributable to larger-sized islets. Based on these data, islets of >150 µm in D account for only 32% of the total islet number but 86% of the total islet volume. This means that most of the β-cell volume is also located within larger islets, which are most susceptible to the negative effects of poor oxygenation. The impact of average islet product size on clinical IT outcomes should be studied further, particularly in recipients of marginal islet doses. Outcomes in these cases would be most likely influenced by factors such as islet size distribution.

Figure 3.

Islet size distribution as stratified by ranges of islet diameter, D (A). Mean (± standard error, SE) number fractions are representative data averaged from 10 human alloislet preparations (high-purity, cultured fractions) produced at the University of Minnesota from February 2011 to January 2012. Mean (± SE) volume fractions are estimated from number fraction data by calculating the mean islet volumes under the assumption that the islets are spherical with a median D for that size range. Of the total number of islets, 32% were >150-µm in D, but these islets account for 86% of the total transplantable volume of islet tissue.

Conclusions

Oxygenation of the intraportally transplanted islet particularly in the early post-IT time period has not been studied extensively and may be an important contributor to early islet loss and dysfunction. Future study of islet oxygenation should involve the development of new methods for the accurate measurement of the local pO₂ near the intraportal islet. Other related studies should assess for a correlation between the average islet size in a preparation and clinical IT outcomes, particularly in recipients of marginal islet doses. It may be that extrahepatic transplant sites could provide the opportunity to engineer approaches for enhanced oxygen delivery and thus improve islet survival and function.