| Site | Local environment challenges | Glyco-insular response | Portal versus systemic insulin delivery | Immune exposure | Surgical accessibility | Preclinical evidence | Clinical evidence |
| Intraportal | ++ | +++ Mimics physiological insulin release | ? | 0 IBMIR | +++ Minimally invasive | [58, 74, 76, 81, 91, 105, 115–117] | [118–120] | Renal subcapsule | + Hypoxia | 0 Need many islets | Systemic | ++ | 0 invasive | [74, 76, 91, 115, 116, 121, 122] Good rodent evidence | [75] | Omental pouch | + Hypoxia | + | Portal | ++ | ++ | [85, 123, 124] | | Gastrointestinal wall | ++ | +++ Physiological glucose entry site | Portal | ++ | +++ Endoscopic access | [92–94] | | Subcutaneous | + Prevascularization required | + | Systemic | + | +++ | [60, 125–127] | | Muscle | + Prevascularization required | + | Systemic | + | +++ | [125, 128–130] | [110, 131] | Bone marrow | +++ Highly vascular | + | Systemic | + | 0 Invasive | [108, 109, 132] | | Adipose | ++ Vascular | + | Systemic | + | +++ | [111, 112] | | Pancreas | +++ Native site | +++ Native site | Portal | + | 0 Invasive | [91, 133] | | Spleen | +++ Highly vascular | +++ | Portal | 0 IBMIR | 0 Hemorrhagic risk | [76, 134–137] Good canine evidence | [124] | Lung (intravenous) | + Venous supply is hypoxic | ++ | Systemic | 0 IBMIR | 0 Widely dispersed | [138, 139] | | Brain | +++ Highly vascular | ++ | Systemic | +++ Immune-privileged | 0 Cerebral ischemia risk | [101, 102] | | Testis | + | + | Systemic | +++ Immune privileged | ++ | [97, 98, 140] | | Thymus | + | + | Systemic | +++ Immune-privileged | ++ | [104–107] | [141] | Celiac artery | + Infarct in terminal end-arteries | ++ | Portal | + IBMIR | + | [117] | |
|
|