A first-in-human phase 0 clinical study of RNA interference-based spherical nucleic acids in patients with recurrent glioblastoma

Glioblastoma (GBM) is one of the most difficult cancers to effectively treat, in part because of the lack of precision therapies and limited therapeutic access to intracranial tumor sites due to the presence of the blood-brain and blood-tumor barriers. We have developed a precision medicine approach for GBM treatment that involves the use of brain-penetrant RNA interference-based spherical nucleic acids (SNAs), which consist of gold nanoparticle cores covalently conjugated with radially oriented and densely packed small interfering RNA (siRNA) oligonucleotides. On the basis of previous preclinical evaluation, we conducted toxicology and toxicokinetic studies in nonhuman primates and a single-arm, open-label phase 0 first-in-human trial (NCT03020017) to determine safety, pharmacokinetics, intratumoral accumulation and gene-suppressive activity of systemically administered SNAs carrying siRNA specific for the GBM oncogene Bcl2Like12 (Bcl2L12). Patients with recurrent GBM were treated with intravenous administration of siBcl2L12-SNAs (drug moniker: NU-0129), at a dose corresponding to 1/50th of the no-observed-adverse-event level, followed by tumor resection. Safety assessment revealed no grade 4 or 5 treatment-related toxicities. Inductively coupled plasma mass spectrometry, x-ray fluorescence microscopy, and silver staining of resected GBM tissue demonstrated that intravenously administered SNAs reached patient tumors, with gold enrichment observed in the tumor-associated endothelium, macrophages, and tumor cells. NU-0129 uptake into glioma cells correlated with a reduction in tumor-associated Bcl2L12 protein expression, as indicated by comparison of matched primary tumor and NU-0129-treated recurrent tumor. Our results establish SNA nanoconjugates as a potential brain-penetrant precision medicine approach for the systemic treatment of GBM.

The Effects of Prostacyclin on the Coagulation of Whole Blood

The effects of prostacyclin (PGI2) on mechanical properties of forming clots were investigated by testing human blood samples on a Thrombelastograph. Concentrations greater than 50 ng/ml (blood) caused a biphasic development of clot stiffness. During the first phase, PGI2 partially inhibited the platelet involvement in coagulation causing initial clot formation at a normal time but with reduced clot stiffness. The second phase occurred after neutralization of PGI2 activity and was characterized by recovery of platelet activity to produce a final clot with normal shear modulus. The duration of the inhibitory effects depended on PGI2 concentration and hematocrit. With a normal hematocrit, a PGI2 concentration of 60 ng/ml caused an inhibition for about 40 min whereas a concentration of 100 ng/ml caused inhibition for about 75 min.

Comparison of Thrombelastography with Common Coagulation Tests

Thrombelastography, although proven as a useful research tool has not been evaluated for its clinical utility against common coagulation laboratory tests. In this study we compare the thrombelastographic measurements with six common tests (the hematocrit, platelet count, fibrinogen, prothrombin time, activated thromboplastin time and fibrin split products). For such comparisons, two samples of subjects were selected, 141 normal volunteers and 121 patients with cancer. The data was subjected to various statistical techniques such as correlation, ANOVA, canonical and discriminant analysis to measure the extent of the correlations between the two sets of variables and their relative strength to detect blood clotting abnormalities. The results indicate that, although there is a strong relationship between the thrombelastographic variables and these common laboratory tests, the thrombelastographic variables contain additional information on the hemostatic process.

Latent inhibition in 35-day-old rats is not an "adult" latent inhibition: implications for neurodevelopmental models of schizophrenia

RATIONALE

Latent inhibition (LI) refers to retarded conditioning to a stimulus as a consequence of its inconsequential preexposure. Amphetamine-induced disruption of LI and its potentiation by antipsychotic drugs (APDs) in the adult rat are well-established models of schizophrenia and antipsychotic drug action, respectively. It is not clear whether LI can be similarly modulated at prepubertal age.

OBJECTIVES

In view of the notion that schizophrenia is a neurodevelopmental disorder whose overt expression depends on postpubertal brain maturational processes, we investigated whether several manipulations known to modulate LI in adult rats, including systemic administration of amphetamine and the atypical APD clozapine, are capable of producing the same effects in prepubertal (35-day-old) rats.

METHODS

LI was measured in a thirst motivated conditioned emotional response (CER) procedure in which rats received 10 or 40 tone preexposures followed by 2 or 5 tone-footshock pairings.

RESULTS

Like in adults, LI was present with 40 preexposures and 2 conditioning trials. In contrast to findings in adults, LI was resistant to disruption by amphetamine at a dose (1 mg/kg) that significantly increased locomotor activity, as well as by reducing the number of preexposures to ten, increasing the number of conditioning trials to five, or changing the context between preexposure and conditioning. Clozapine (5 mg/kg) and the selective 5HT2A antagonist M100907 (0.3 mg/kg) administered in conditioning were without an effect on "persistent" LI with extended conditioning, but were capable of disrupting LI when administered in the preexposure stage, as found in adults.

CONCLUSION

The results point to functionality within brain systems regulating LI acquisition but not those regulating LI expression in periadolescent rats, further suggesting that postpubertal maturation of the latter systems may underlie schizophrenia-mimicking LI disruption reported in adult rats following perinatal manipulations and possibly disrupted LI observed in schizophrenia.

Antiphospholipid antibodies, proteins C and S, and coagulation changes in sickle cell disease

The significance, interactions, and sources of coagulation abnormalities and their relationship to clinical severity and painful episodes in sickle cell disease are not clear. To evaluate this, we have examined various measures of coagulation in 37 patients with sickle cell disease (20 patients with HbSS disease and 17 patients with HbSC disease). Measurements have included isotypes of antiphospholipid antibodies (IgG, IgM, IgA) to specific phospholipids; proteins C (activity, total antigen) and S (activity, total and free antigen); measures of coagulation activation (prothrombin fragment 1.2, thrombin-antithrombin, fibrinopeptide A, d-dimers); indicators of clinical severity; and studies obtained during steady states and painful episodes. Results in HbSS disease showed that antiphospholipid antibodies were increased, with IgG phosphatidylserine showing the highest and most frequently increased levels (37% of patients). Protein C (activity) and protein S (activity, total, free antigen) were decreased (P<.01), and all measures of coagulation activation were increased (P<.001). In HbSC disease, antiphospholipid antibodies were normal, protein C (activity) and protein S (free antigen) were decreased (P<.001), and all measures of coagulation activation were increased (P<.02). A strong correlation was observed in HbSS disease between IgG-PS and d-dimers. Moderate correlations occurred between protein C activity and thrombin-antithrombin and fibrinopeptide A, between protein S activity and prothrombin fragment 1.2 and d-dimers, and between protein C and protein S activity. In HbSC disease, moderate and fewer correlations occurred. Significant differences between HbSS disease and HbSC disease were observed in aPLs, proteins C and S, and measures of coagulation activation. Measurements during steady states and during painful episodes were not significantly different. We conclude that the antiphospholipid antibody IgG-PS may contribute to coagulation activation in HbSS disease and that IgG-PS, protein C, and protein S relate to each other and jointly to measures of coagulation activation. The increased level of IgG-PS in HbSS disease most likely reflects exposure of the procoagulant phosphatidylserine on the surfaces of red cell-shed vesicles and sickle red cells, which would further affect coagulation activation. The significant differences in coagulation measures between HbSS disease and HbSC disease are consistent with differences in clinical severity between the diseases. The development of painful episodes does not appear to be related to the coagulation changes.

The CD28 ligand, B7, enhances IL-2 production by providing a costimulatory signal to T cells

Previous studies demonstrated that a human pre-B acute lymphoblastic leukemia cell line, NALM-6, failed to stimulate a primary MLR, despite expression of class II MHC and adhesion molecules. Here we demonstrate that this is the result of the fact that NALM-6 cells do not express the ligand for CD28, namely B7. NALM-6 transfectants that expressed high levels of B7 gained the capacity to stimulate IL-2 production by class II MHC molecule-specific alloreactive T cells and to costimulate a polyclonal population of purified T cells cultured with immobilized anti-CD3 mAb. In the presence of PMA, NALM-6 cells transfected with B7 polyclonally stimulated T cells in a cyclosporine A-resistant fashion, a property previously attributed only to agonistic anti-CD28 mAb. The gain of these functions could not be explained solely by an increased capacity of the transfectants to form conjugates with T cells, suggesting that the CD28/B7 interaction transduces a costimulatory signal in T cells.

The CD28 ligand, B7, enhances IL-2 production by providing a costimulatory signal to T cells

Previous studies demonstrated that a human pre-B acute lymphoblastic leukemia cell line, NALM-6, failed to stimulate a primary MLR, despite expression of class II MHC and adhesion molecules. Here we demonstrate that this is the result of the fact that NALM-6 cells do not express the ligand for CD28, namely B7. NALM-6 transfectants that expressed high levels of B7 gained the capacity to stimulate IL-2 production by class II MHC molecule-specific alloreactive T cells and to costimulate a polyclonal population of purified T cells cultured with immobilized anti-CD3 mAb. In the presence of PMA, NALM-6 cells transfected with B7 polyclonally stimulated T cells in a cyclosporine A-resistant fashion, a property previously attributed only to agonistic anti-CD28 mAb. The gain of these functions could not be explained solely by an increased capacity of the transfectants to form conjugates with T cells, suggesting that the CD28/B7 interaction transduces a costimulatory signal in T cells.

The effects of prostacyclin on the coagulation of whole blood

The effects of prostacyclin (PGI2) on mechanical properties of forming clots were investigated by testing human blood samples on a Thrombelastograph. Concentrations greater than 50 ng/ml (blood) caused a biphasic development of clot stiffness. During the first phase, PGI2 partially inhibited the platelet involvement in coagulation causing initial clot formation at a normal time but with reduced clot stiffness. The second phase occurred after neutralization of PGI2 activity and was characterized by recovery of platelet activity to produce a final clot with normal shear modulus. The duration of the inhibitory effects depended on PGI2 concentration and hematocrit. With a normal hematocrit, a PGI2 concentration of 60 ng/ml caused an inhibition for about 40 min whereas a concentration of 100 ng/ml caused inhibition for about 75 min.

Breast carcinoma: expected results with prevailing therapy

Computer-stored tumor registry data were scanned to produce the set of patients whose breast carcinoma was either diagnosed or first treated at Evanston Hospital between the years 1973 and 1977, inclusive. Of 560 evaluable patients, 184 were less than or equal to age 50 (LE 50), and 376 were greater than age 50 (GT 50). Comparisons of projected survival show that the survival of all patients is 8% at 5 years and 72% at 7 years. The survival of patients LE 50 is nearly identical to that of the GT 50 age group. Survival of Stage III GT 50 patients, however, is better than Stage III LE 50 (P less than 0.001). Comparison with earlier studies shows an historical trend toward better 5-year survival. Data were generated in this study which allow prediction of prognosis based upon age and stage at diagnosis.

Comparison of thrombelastography with common coagulation tests

Thrombelastography, although proven as a useful research tool has not been evaluated for its clinical utility against common coagulation laboratory tests. In this study we compare the thrombelastographic measurements with six common tests (the hematocrit, platelet count, fibrinogen, prothrombin time, activated thromboplastin time and fibrin split products). For such comparisons, two samples of subjects were selected, 141 normal volunteers and 121 patients with cancer. The data was subjected to various statistical techniques such as correlation, ANOVA, canonical and discriminant analysis to measure the extent of the correlations between the two sets of variables and their relative strength to detect blood clotting abnormalities. The results indicate that, although there is a strong relationship between the thrombelastographic variables and these common laboratory tests, the thrombelastographic variables contain additional information on the hemostatic process.

Plasma cholinesterase activity in bank blood

Plasma cholinesterase activity was periodically measured in bank blood. Of the initial plasma cholinesterase activity 87% was retained after 21 days of storage at 4 C in citrate-phosphate-dextrose solution. In addition, plasma frozen for 7 weeks at -70 C showed no decrease in enzyme activity. No difference could be demonstrated in the stability of the enzyme related to whether the blood was stored in citrate-phosphate-dextrose or in acid-citrate-dextrose solution. Bank blood transfusion prior to or during succinylcholine administration should not significantly alter the duration of action of succinylcholine, and would probably be effective in the management of prolonged apnea resulting from a deficiency of this enzyme.

The colon influences ileal resection diarrhea

Fecal mass and electrolyte concentrations from 25 ileectomy and/or colectomy patients on known diets were used to assess those factors most responsible for their diarrhea. In 18 ileectomy patients the severity of diarrhea, expressed as a fecal weight, was a function of both percent of colon and centimeters of ileum removed. Linear regression analysis, however, showed that the extent of missing colon had three times the effect of missing ileum on fecal weight. Patients who lost the ileocecal valve and part of the right colon had more diarrhea than those who lost comparable lengths of ileum but had this area preserved. Fecal ion concentrations seemed independent of diet but were related to fecal weight and the amount of colon and ileum removed. Potassium concentration was strongly dependent on the amount of colon lost, while sodium concentration was more influenced by the length of resected ileum. Choloride was most dependent on fecal weight. As expected, fecal fat correlated strongly with the extent of ileum removed. Regresison equations were constructed from the electrolyte data which described and predicted the extent of lost ileum or colon. Our data were also used to separate patients with less than 100 cm of ileum removed from those with more extensive resections. The severity of diarrhea following ileal resection depends primarily on the amount of contiguous colon removed. Varying loss of ileum and colon produced predictable effects on fecal weight and electrolyte composition. Surgeons should preserve the maximum amount of colon possible to reduce the severity of diarrhea in these patients.

Hematologic Analysis Following Intravenous Calcium and Sodium Heparin Administration

Twenty normals received in randomized order a dose of 1000 units each of calcium and sodium heparin intravenously seven days apart. The heparin used for preparation of the calcium salt was from the same batch as the sodium salt. Blood tests were obtained pre, 15 and 60 minutes after injection. Values of hematocrit, hemoglobin, RBC, WBC, prothrombin time and platelet adhesive index showed no significant response to the quantity of heparin. Lee White, activated partial thromboplastin time, activated clot time, thrombelastograph and thrombin calcium clot time doubled at 15 minutes. Heparin levels reached 0.4 U/ml (Anti X assay) at 15 minutes while factor X was decreased by 15% for both heparin salts. A differential response to the heparin salts was seen at 15 minutes where the FSP was increased (average difference ± standard error of the difference) by 1.6±. 6 ug/ml, fibrinogen decreased 49 ± 20 mg/ml and plasminogen dropped 2. 3 ± .6 mg/ml for the sodium salt compared to the calcium preparation. With the exception of fibrinogen, these statistically significant differences between the salts disappeared by 60 minutes. At 60 mins ATIII levels were slightly lower for the sodium heparin. Significant changes in C′3 , C′4 and serum calcium were noted with both heparins. These results suggest similar hematologic effects following intravenous sodium or calcium heparin except for an apparent increased fibrinolytic response at 15 minutes when the sodium salt was administered. Due to the small magnitudes of these fibrinolytic changes, these findings may be of questionable clinical significance, but merit further investigation at higher dose levels.

Hypercoagulability Associated with Breast Cancer

Previous thrombelastographic (TEG) studies in 262 individuals (not on medications) has resulted in the derivation and verification of a discriminant function equation which can classify individuals into a normal or accelerated coagulability range based on the TEC resultsThe significance of the separation between the groups was related to the combination individual’s TEG parameters for whole blood (WB) and their celite activated WB. The centite activated rate of clot formation appeared to be the most discriminating of the TEG variables. The equation separated the individuals with known malignancies (121) from the benign or healthy volunteers (141) with an overall accuracy of 99%. This breaks down into a 97-98% sensitivity and a 100% specificity (false positives). These findings of hypercoagulability are not specific for cancer end can be seen in other disease states e.g. sepsis or even the result of some medications. Therefore the present study was done to apply this analysis to a consecutive series of consenting patients (119) just prior to open breast biopsy, without selecting out patients on medications. 50/68 patients with ( 1 benign lesions had normal TEG results (false positive rate 26%), while 48/51 patients with pathologic evidence of malignancy had positive results. In two of the three tumor patients with a negative analysis, no invasion of the tumor was seen. The reduced specificity tourul with this group may be attributable to the presence of patients on medications or a higher than normal incidence of other disease states in this population. The results demonstrate one clinical application of TEG analysis to the identification of breast malignancies.

Thrombelastographic Effects of Dextran 70 on Human Blood

The effects of dextran was studied by thrombelastography (TEG) because it can graphically depict fibrin, formation and dissolution and plate 1st-fibrin interaction during clothing. Dextran 70 in concentrations of 1-8% ware added to samples of, native whole blood (WB), recalcified whole blood (RWB), platelet-rich plasma (PRP) and platelet-poor plasma (PPP) using buffered saline as control. Dextran produced an 13% reduction in clot Stiffness (MA) with native and celite activated WB. The same native WB shewed a 23? prolongation of clotting (R) and a 24% decrease in the rate of clot formation (α), while the celite activated WB R and α were reduced by 17% and 19% respectively. However, using RWB we found only a 10% reduction in MA as a result of dextran addition, suggesting that citrating and recalcification diminished the dextran effects. With FRF (platelets >560°000 dextran did not show any significant reduction in the measured parameters, uhiU FFP gavs the greatest response with a 36% decrease in MA. Results vere similar if FRF was obtained from subjects who had taken aspirin. This indicates the main action of dextran but to be on the fibrin network rather than on platelets or the plate let-fibrin interaction. though this effect is best seen when platelet concent: rations are low. Clot lysis by urokinase or streptokinase was accelerated in the presence of dextran but was reduced by increased platelet concentrations. These results suggest a direct effect of dextran on fibrin formation and may explain the antithrombotic effect of dextran in thit defective fibrin formation occurs.

Effect of platelet count and hematocrit on the coagulability of heparinized blood

The coagulability of heparinized blood is shown to be a function of platelet and red cell concentrations. Thrombelastrographic tests were conducted on celite-activated fabricated whole blood samples with heparin concentrations ranging from zero to 1.46 units per ml of plasma, hematocrits from zero to 50% and platelet counts from zero to 300,000/mm3. The thrombelastograph provides three useful parameters of coagulability in native and heparinized blood: clot time, rate of clot stiffening and final clot stiffness. Of these, only the clot time is determined with the common clotting tests, while the rate and final clot stiffness are measures of clot quality that are found to be particularly sensitive to thrombocytopenia. The test results indicate (i) increases of platelet concentration produce the expected decrease in clot time, increase in rate of stiffening, and increase in final stiffness; (ii) increases of the red cell concentration from zero to 20% hematocrit, produce nonlinear increases in clotability, while further increases beyond 20% hematocrit produce no additional effects if the heparin concentration is less than 1.2 units/ml. Quantative relations are derived between the thrombelastographic variables, heparin concentration, and platelet concentration for those samples with hematocrits greater than 20%. These data indicate further experiments for the study of possible interaction between cellular blood elements and heparin.

Disseminated intravascular multiple systems activation (DIMSA) following thermal injury

Seventy-seven major thermal injury victims were studied with a number of hematologic and immunologic tests initially and sequentially during the first postburn month. The patients were grouped by initial prognostic index as well as by ultimate survival. Pairs were grouped by initial prognostic index as well as by ultimate survival. Pairs of test data from subjects studied at successive time intervals were compared with prognostic index and ultimate survival. Statistically significant changes in coagulation, fibrinolytic, complement, and kinin tests all occurred within these groups. These findings strongly suggest that intravascular contamination occurs following thermal injury in proportion to the extent of the burn, because of the occurrence and persistence of statistically significant multiple system changes. Regrouping all of the blood data according to ultimate death or survival reinforced the concept of intravascular contamination and provided the basis by which one can compute laboratory prognostic indices. The combination of plasminogen, C'3 complement, C'H50, one-minute kinin, and TEG index discriminated between death and survival with 91% accuracy by the end of the first postburn week. These data offer the potential for select blood measurements in refining current prognostic indicators. This may provide an objective data base for the analysis of new treatment programs in thermal injury victims.

Variation of cold agglutinin levels

Cold agglutinin levels were determined in 120 individuals ranging in age from 1 to 64 years using untreated and ficin-treated group O erythrocytes. The geometric mean titer for untreated cells was 5.7 +/- 1.1 SEM and for ficin-treated cells 18.7 +/- 1.1 SEM. The mean titers and scores for various blood groups were not significantly different. Titers and scores correlated significantly to total serum IgM concentration (p less than 0.001). Females had significantly higher cold agglutinin levels than males (p less than 0.001). A significant decrease of cold agglutinins occurred in the 26- to 64-year age group as compared to the 11- to 25-year age group (p less than 0.01). The data should contribute to a clearer understanding of the biological impact of cold agglutinins.

In vitro testing of surface thrombogenicity using the thrombelastograph

Synthetic surfaces can be assessed for their ability to promote or inhibit whole blood coagulation with improved accuracy using a new protocol developed for the Thrombelastograph instrument. Results obtained from the analysis of the hydrodynamic state of blood clotting in Thrombelastograph chambers have enabled a quantitative interpretation of the output of this instrument to be made. The thrombogenic properties of a segmented polyurethane and a polydimethyl siloxane are compared with a preparation of reconstituted elastin. It is seen that a four-fold delay in clotting time is exhibited by reconstituted elastin when it is compared with other biomedical synthetic materials tested.

Diffusion of oxygen in plasma and blood

D in fresh human plasma has been found to decrease almost linearly with total protein content over a wide range of concentration and to vary only +/-4% in normals and +/-13% in abnormals. The average values for D in normal human plasma, at 25 and 37 degrees C, are 1.62 and 2.18 X 10(-5) cm2/sec respectively. For normal human blood at 42% hematocrit, the values of D, at 25 and 37 degrees C, are 1.20 and 1.62 X 10(-5) cm2/sec respectively.

Gastric operations and glucose homeostasis. II. glucagon and secretin

Alimentary hyperglycemia in patients who have undergone gastric operations may be due, in part, to altered intestinal signals for glucose disposition. We measured glucose, immunoreactive insulin (IRI) pancreatic glucagon (IRG), and glucagon-like immunoreactivity (GLI) after oral glucose in patients with prior antrectomy or vagotomy and pyloroplasty and in normal individuals. All subjects had normal assimilation coefficients for intravenous glucose, which suggests that the responsiveness of the pancreatic beta-cells had not been altered by the surgical procedures. The early hyperglycemic response to oral glucose and the associated elevation of plasma GLI were much greater and the IRI levels slightly higher in both experimental groups in comparison to normal subjects. A decrease in the level of IRG, albeit not statistically significant, was noted in all groups after the ingestion of glucose. In gastrectomy patients, secretin infusion during repeated oral glucose tolerance tests partially corrected the hyperglycemia and lowered plasma GLI and IRI levels. The responses of the vagotomy and pyloroplasty patients and of the normal subjects were not altered by secretin infusion. We conclude that the intolerance or oral glucose after gastric surgery may be related to elevated GLI levels, and that the beneficial effect of secretin may be due to its ability to decrease these levels.

Alcohol inhibition of cyclic AMP-induced insulin release

In order to define the role of alcohol in cAMP-induced insulin release, the two agents were infused into the pancreas or liver of normal dogs separately or in combination. Blood levels of insulin and glucose were measured in the portal vein. CAMP infusion in the pancreatic artery provoked a greater early insulin response than did a similar infusion into the portal vein when glycemic levels were comparable. The addition of alcohol suppressed insulin response to intrapancreatic cAMP. Alcohol in the absence of the stimulus failed to suppress insulin below basal levels. It is concluded that pancreatic alcohol directly inhibits cAMP-related insulin release and that this is the mechanism by which glucagon-induced insulin release is blunted by alcohol. Glucose-induced release, which is potentiated by intravenous alcohol, apparently involves another mechanism.

Profile of insulin release due to intrapancreatic glyburide infusion

Intrapancreatic glyburide infusion in dogs (0.02 mg. per kg. of body weight) for five minutes provoked immediate, marked insulin release into the portal venous blood. Compared to control animals, the normal intrinsic insulin levels were exaggerated at intervals of approximately eighty minutes. Lack of immediate hypoglycemia suggests that the insulin release is merely secondary in contributing to the anti-diabetic property of this agent. An amount of tolbutamide fifty times as great caused a weak, delayed insulin response. The delayed blood glucose response is probably caused by the combined action of glyburide and circulating insulin.

Pancreatic insulin secretion following intrapancreatic infusion of amino acid

Both protein ingestion and amino acid infusion are known to stimulate insulin secretion. The mechanism of this response has been attributed to increased plasma amino acid concentration. The present studies were undertaken to determine whether this-secretory response is due to a direct effect of amino acids on the islet cells or whether it is mediated through some other indirect mechanism. Accordingly, pancreatic infusion studies were conducted to help resolve this question.

Casein hydrolysate (1 gm.) infused into the pancreatico-duodenal artery of dogs for one-half hour provoked an immediate secretion of insulin into the portal vein. Femoral arterial insulin concentration also rose within five minutes, although hyperglycemia was not observed until the end of the infusion. An insulin response of similar magnitude was achieved by pancreatic glucose infusion (1 gm.). Simultaneous administration of epinephrine (30 μg.) inhibited the insulin responses to both glucose and amino acid. Combined infusion of casein hydrolysate and glucose (0.5 gm. each) produced a greater insulin response than the sum of the individual responses, suggesting glucose-amino acid synergism. Intraportal infusion of casein hydrolysate produced no significant insulin response in either portal venous or femoral arterial plasma.

These studies suggest that the insulin response associated with ingested or parenteral protein represents a direct amino acid effect on the islets of Langerhans. They also suggest that the amino acid concentration in the pancreas may be as important as glucose in control of insulin secretion.