Secondary Logo

Journal Logo

Successful Islet Transplantation from Nonheartbeating Donor Pancreata Using Modified Ricordi Islet Isolation Method

Matsumoto, Shinichi1,7; Okitsu, Teru2; Iwanaga, Yasuhiro2; Noguchi, Hirofumi2; Nagata, Hideo2; Yonekawa, Yukihide2; Yamada, Yuichiro3; Fukuda, Kazuhito3; Shibata, Toshiya4; Kasai, Yasunari5; Maekawa, Taira5; Wada, Hiromi6; Nakamura, Takayuki6; Tanaka, Koichi2

doi: 10.1097/
Original Articles: Cell Therapy and Islet Transplantation

Background. Current success of islet transplantation has led to donor shortage and the need for marginal donor utilization to alleviate this shortage. The goal of this study was to improve the efficacy of islet transplantation using nonheartbeating donors (NHBDs).

Methods. First, we used porcine pancreata for the implementation of several strategies and applied to human pancreata. These strategies included ductal injection with trypsin inhibitor for protection of pancreatic ducts, ET-Kyoto solution for pancreas preservation, and Iodixanol for islet purification.

Results. These strategies significantly improved both porcine and human islet isolation efficacy. Average 399,469±36,411 IE human islets were obtained from NHBDs (n=13). All islet preparations met transplantation criteria and 11 out of 13 cases (85%) were transplanted into six type 1 diabetic patients for the first time in Japan. All islets started to secrete insulin and all patients showed better blood glucose control without hypoglycemic loss of consciousness. The average HbA1c levels of the six recipients significantly improved from 7.5±0.4% at transplant to 5.1±0.2% currently (P<0.0003). The average insulin amounts of the six recipients significantly reduced from 49.2±3.3 units at transplant to 11±4.4 units (P<0.0005) and five out of six patients reduced to less than half dose. The first patient is now insulin free, the first such case in Japan.

Conclusion. This demonstrates that our current protocol makes it feasible to use NHBDs for islet transplant into type 1 diabetic patients efficiently.

1 Kyoto University Hospital Transplantation Unit, Kyoto, Japan.

2 Department of Transplantation and Immunology, Kyoto University Graduate School of Medicine, Kyoto, Japan.

3 Department of Diabetes and Clinical Nutrition, Kyoto University Graduate School of Medicine, Kyoto, Japan.

4 Kyoto University Hospital Radiology and Nuclear Medicine Service, Kyoto, Japan.

5 Kyoto University Hospital Department of Transfusion Medicine and Cell Therapy, Kyoto, Japan.

6 Kyoto University Hospital Department of Thoracic Surgery, Kyoto, Japan.

This work was supported in part by the Ministry of Education, Science, and Culture, the Ministry of Health, Labour and Welfare and the 21st Century Center of Excellence Program, Japan.

7 Address correspondence to: Shinichi Matsumoto, M.D., Ph.D., Kyoto University Hospital Transplantation Unit, 54 Kawara-cho Shogoin, Sakyo-ku, Kyoto 606-8507, Japan.


Received 2 April 2005. Revision requested 2 May 2005.

Accepted 30 March 2006.

Islet transplantation has proven to be highly effective for reversing the diabetic state of patients with type 1 diabetes (1, 2). Because the number of cadaveric donors is inadequate to treat the number of qualifying type 1 diabetic patients, the utilization of marginal donor pancreata for islet isolation is needed to alleviate donor shortage. A University of Pennsylvania group recently demonstrated that type 1 diabetes could be cured with islet transplantation from nonheartbeating donors (NHBDs) (3). Their study is particularly important for countries like Japan where isolating islets from heart-beating brain-death donor pancreata is not allowed. Their study also indicated that islet isolation was more difficult from NHBDs than from heart-beating brain death donors. Indeed, after their first case, there had been no clinical islet transplantation using NHBDs until we initiated series of islet transplantations with NHBDs reported in this paper.

For this study, we first used a porcine model to evaluate several strategies to improve the efficacy of islet isolation from NHBDs and then applied these strategies to clinical islet isolations. They included the two-layer (modified ET-Kyoto solution/perfluorocarbon [PFC]) method of pancreas preservation, pancreatic ductal protection and new islet purification solution (Iodixanol-based solution).

It was found that the two-layer method (TLM) of pancreas preservation improved the efficacy of human islet isolation from marginal donors (4–6), whereas the ET-Kyoto solution proved to be superior to the University of Wisconsin solution (UW solution) for pancreas preservation as part of the TLM before islet isolation (7).

Recently, it was shown that ductal injection of UW solution at the time of pancreas procurement protected pancreatic ducts and improved the efficacy of islet isolation in a rodent model (8). In addition, trypsin inhibition during pancreas digestion significantly improved islet isolation efficacy with marginal donors (4). On the basis of these findings, we used the ET-Kyoto solution and trypsin inhibitor to enhance its effect.

The Iodixanol-based purification solution has been proven superior to Ficoll for porcine islet purification (9, 10). We combined Iodixanol with the ET-Kyoto solution to generate a new purification solution.

With these strategies, we successfully isolated human islets from NHBDs and islets from 11 out of 13 donors were transplanted into six type 1 diabetic patients with promising results for the first time in Japan.

Back to Top | Article Outline


This study was approved by the Ethics Committee of the Kyoto University Graduate School and Faculty of Medicine.

Back to Top | Article Outline

Porcine Islet Isolation

Porcine pancreata were obtained at a local slaughterhouse. Warm ischemic time (WIT) was defined as the time elapsed between cessation of heart beat and placement of the pancreas into the preservation solution. Cold ischemic time (CIT) was defined as the time elapsed between placement of the pancreas into the preservation solution and the start of islet isolation. After obtaining a pancreas, we put it immediately into a two-layer (UW/PFC) preservation container or immediately inserted cannula into the main pancreatic duct for infusion with the modified ET-Kyoto solution for ductal protection (8), and put the pancreas into a two-layer (modified ET-Kyoto solution/PFC) preservation container. The modified ET-Kyoto solution consisted of 50,000U of Ulinastatin (Mochida Pharmaceutical Co Ltd, Tokyo, Japan) added to 1 L of the solution. On arrival at the islet isolation laboratory at Kyoto University, the pancreata were processed according to Edmonton protocol with some modifications (1, 11). Briefly, after the pancreas had been decontaminated, the ducts were perfused in a controlled fashion with a cold enzyme of Liberase HI (Roche Molecular Biochemicals, Indianapolis, IN). The distended pancreas was then cut into nine pieces and transferred to a Ricordi chamber. The pancreas was digested by repeatedly circulating the enzyme solution through the Ricordi chamber at 37°C. The Phase I period was defined as the time between placement of the pancreas in the Ricordi chamber and the start of collection of the digested pancreas. The Phase II period was defined as the time between the start and the end of the collection.

Islets were purified with a continuous density gradient of high-osmolality Ficoll based on the Edmonton islet isolation protocol (1) or with the Iodixanol-ET-Kyoto solution in an apheresis system (COBE 2991 cell processor, Gambro Laboratories, Denver, CO). For solution, low-density and high-density Iodixanol-ET-Kyoto solutions were produced by changing the volumetric ratio of Iodixanol and ET-Kyoto solution.

Back to Top | Article Outline

Islet Isolation Protocols

We used three islet isolation protocols for porcine islet isolations. For the first group, we used the two-layer (UW/PFC) method of pancreas preservation without ductal protection and islet purification with Ficoll (UW-Ficoll). For the second group, we used the two-layer (modified ET-Kyoto solution/PFC) method of pancreas preservation with ductal protection and islet purification with Ficoll (Kyoto-Ficoll). For the third group, we used the two-layer (modified ET-Kyoto solution/PFC) method of pancreas preservation with ductal protection and islet purification with Iodixanol-ET-Kyoto solution (Kyoto-Iodixanol).

Back to Top | Article Outline

Procurement of Human Pancreata

Thirteen human pancreata were obtained with the informed consent from relatives of NHBDs and procured through the Central Japan Region (n=12, Aichi Prefecture, Japan) and the West Japan Region (n=1, Osaka Prefecture, Japan) of the Japan Organ Transplantation Network between January 17, 2004 and January 8, 2005. After brain death had been confirmed, we inserted cannula into the iliac vessels for rapid cooling of the pancreas (Nagata et al., manuscript in preparation). Cold lactate ringer solution was perfused after cessation of heart beating until removal of pancreata via the cannula. Modified ET-Kyoto solution was infused into the main pancreatic duct for ductal protection at the time of pancreas preservation (8). The pancreata were preserved with the two-layer (modified ET-Kyoto solution/oxygenated perfluorocarbon) method and transported to our human islet isolation GMP facility (Center for Cell and Molecular Therapy).

Back to Top | Article Outline

Human Islet Isolation

Islet isolation was performed according to the Edmonton protocol (1, 2, 11) and modified on the basis of the results of our porcine experiments. The second islet isolation protocol (Kyoto-Ficoll) was used for the first two cases and the third islet isolation protocol (Kyoto-Iodixanol) for the other 11 cases.

Back to Top | Article Outline

Human and Porcine Islet Evaluation

Islet preparations were evaluated for yield and purity by means of dithizone staining (12). Gross morphology was qualitatively assessed by two independent investigators scoring the islets for shape (flat vs. spherical), border (irregular vs. well-rounded), integrity (fragmented vs. solid/compact), uniformity of stain (not uniform vs. perfectly uniform) and diameter (all <100 μl vs. >10% >200 μl) (11, 12). Each parameter was graded from zero to two with zero equaling the worst and two the best score, so that the worst islet preparations were given a score of zero and the best a score of 10. Spherical, well-rounded, solid/compact, uniformly stained, and large islets were characterized as the best islets.

Islet viability after purification was assessed by using acridine orange (10 μmol/L) and propidium iodide (15 μmol/L) (AO/PI) staining to visualize living and dead islet cells simultaneously (11–13). Fifty islets were inspected and their individual viability was determined visually, followed by calculation of their average viability (11).

Islet function was assessed by monitoring the insulin secretory response of the purified islets during glucose stimulation according to a procedure described by Shapiro and colleagues (1). Briefly, after overnight culture with CMRL at 37°C with 5% CO2, 100 Islet Equivalents (IE) were incubated with either 2.8 mM or 20 mM glucose in RPMI 1640 (GIBCO BRL, Tokyo, Japan) for 2 hr at 37°C and 5% CO2. The supernatant was then collected and insulin levels were determined with an insulin ELISA kit (Morinaga Seikagaku Co Ltd., Tokyo, Japan). The stimulation index was calculated by determining the ratio of insulin released from the islets at a high glucose concentration to that released at a low concentration.

Back to Top | Article Outline

Islet Transplantations into Diabetic Patients

We used 11 out of the 13 islet preparations to perform islet transplantations into six type 1 diabetic patients between April 7, 2004 and January 7, 2005. The two remaining islet preparations were cryopreserved for future transplantation. Four patients received multiple islet preparations each and two patients received only one.

Patients were sedated and a percutaneous transhepatic approach was used to gain access to the portal vein for all six patients. Once access was confirmed, the Seldinger technique was used to place the Kumpe catheter within the main portal vein. Islets were infused by gravity and using the bag technique (14).

Back to Top | Article Outline

Assessment of Transplanted Islet Function

Islet functioning was assessed in terms of daily serum glucose levels, amount of insulin requirement, and HbA1c before and after islet transplantation. We performed the glucagon stimulation test before and after transplantation. For the glucagon stimulation test, we drew blood for C-peptide measurement before and after six min of 1 mg glucagon infusion.

Back to Top | Article Outline


Values for the data collected represent means±SE. Three groups were compared by means of analysis of variance (ANOVA) followed by Fisher’s PLSD posthoc test. P values less than 0.05 were considered significant.

Back to Top | Article Outline


Porcine Islet Isolation Characteristics

Porcine islet isolation characteristics in terms of isolation protocols are shown in Table 1. There were no significant differences in pancreas size or operation time among the three groups. WIT was significantly shorter for the UW-Ficoll group since the ductal injection process was eliminated for this group. However, CIT was significantly longer for the UW-Ficoll group since this group needed additional cannulation at the laboratory. Phase I was significantly longer for the UW-Ficoll group but Phase II was similar for all three groups.



Islet yield before purification was significantly higher in the Kyoto-Iodixanol and Kyoto-Ficoll groups compared with the UW-Ficoll group (Kyoto-Iodixanol: 10,247±637 IE/g, Kyoto-Ficoll: 8,846±904 IE/g, UW-Ficoll; 4,809±454 IE/g) (Fig. 1A). Islet yield after purification was significantly higher in the Kyoto-Iodixanol group than the UW-Ficoll and Kyoto-Ficoll groups (Kyoto-Iodixanol: 7,253±915 IE/g, Kyoto-Ficoll: 3,527±795 IE/g, UW-Ficoll: 2,486±394 IE/g) (Fig. 1B). Other porcine islet characteristics are shown in Table 2. The Kyoto-Iodixanol group had a significantly high morphological score, postpurification recovery rate, and islet size (Table 2), whereas the stimulation index of the glucose challenge test was similar for the three groups (Table 2).





Back to Top | Article Outline

Human Donor and Pancreas Characteristics

The average age was 44±4 years, period of intensive care unit (ICU) stay was 11±3 days, body mass index (BMI) was 21±1 kg/m2, pancreas size was 87±6 g, WIT was 7±3 min, and CIT was 256±18 min.

The WIT was minimal for all pancreata due to immediate perfusion with cold lactate Ringer solutions and total CIT was less than six hr in all cases. All average blood chemistry values were abnormal.

Back to Top | Article Outline

Human Islet Characteristics

Islet characteristics in terms of islet isolation protocol are shown in Table 3. There were no apparent differences in islet yield before and after purification between the two groups. The values for purity and viability and the scores were similar for the two groups. The postpurification recovery rate was higher for the Kyoto-Iodixanol group than for the Kyoto-Ficoll group as porcine experiments.



Our current criteria for the approval of clinical transplantation are that islets yield more than 5000 IE/kg body weight, purity more than 30%, viability more than 70%, tissue volume less than 10 mL, endotoxin level less than 5 EU/kg body weight, and they show a negative gram stain based on the Edmonton protocol (1). According to these criteria, all cases met transplant criteria except for islet yields (Table 3). We transplanted the islets, one set from the Kyoto-Ficoll group and 10 sets from the Kyoto-Iodixanol group to type 1 diabetic patients.

Back to Top | Article Outline

Postoperative Course of All Cases

All transplanted islets started to secrete insulin based on C-peptide measurements and all patients improved blood glucose control without experiencing any hypoglycemic loss of consciousness. Average insulin amount at the transplant was 39.2±3.2 unit which was reduced to 11.0±4.4 unit (P<0.0005). Two patients became insulin free, two other patients reduced insulin doses less than 10 units, and the other two patients also reduced insulin amounts. HbA1c levels of all six patients gradually decreased and reach less than 6% within three months irrespective of single or multiple islet transplantation (Fig. 2A). The average HbA1c levels of the six recipients significantly improved from 7.5±0.4% at transplant to 5.1±0.2% currently (P<0.0003). All patients had undetectable C-peptide levels (<0.1 ng/ml) before transplantation and all patients had positive C-peptide after transplantation. Basal and stimulated C-peptide levels after the first transplantation was 0.29±0.06 ng/ml and 0.52±0.11 ng/ml respectively (n=6) (Fig. 2B). Both basal and stimulated C-peptide levels after the second transplantation increased compared with after the first transplantation and actual average was 0.75±0.12 ng/ml (P<0.01) and 1.45±0.26 ng/ml (P<0.005), respectively (n=3) (Fig. 2B).



Back to Top | Article Outline

Islet Transplantation and Postclinical Course of the Japanese First Case

We performed the first clinical allogenic islet transplantation in Japan on April 7, 2004 and this patient received a second islet preparation on June 2, 2004. Islet yield for the first transplant was 354,384 IE and 474,234 IE for the second transplant. The patient was a 36-year-old female who had had type 1 diabetes for 22 years with frequent severe hypoglycemic episodes and diabetic retinopathy. Renal function was normal with a creatinine level of 0.7 mg/dl. Immunosuppressants were administered based on the Edmonton protocol (1) but we used 20 mg of basiliximab instead of daclizumab on postoperative day (POD) 0 and 4.

Blood glucose levels before breakfast and dinner were measured from April 4, 2002 to May 28, 2002 for the control of pretransplant blood glucose values. Pretransplant blood glucose levels were highly unstable and ranged were from 20 mg/dl to over 400 mg/dl. After the first transplantation, blood glucose levels were kept within a narrow range (Fig. 3A). The amount of insulin required before transplantation was between 30 and 40 unit/day, but the insulin dosage could gradually be reduced to 10 unit/day at 1 month after the first transplantation and was completely stopped 20 days after the second transplantation (Fig. 3B).



Transaminase blood levels increased temporarily after the first transplantation up to but no more than 100 mg/dl, but returned to normal levels within 3 weeks. Creatinine levels stayed at less than 1.0 mg/dl and blood urea nitrogen (BUN) maintained normal levels throughout the study period.

Back to Top | Article Outline


In this study, we modified the Ricordi islet isolation protocol for pancreata from NHBDs and improved the efficacy of islet transplantation. With this modified islet isolation protocol, we were able to perform the first allogenic islet transplantation in Japan for the treatment of type 1 diabetes. The first recipient became insulin independent after the second transplantation.

It is known that results of human islet isolation depend on the quality of donor pancreata (15, 16), so that it is not easy to isolate suitable human islets successfully from NHBDs.

We first examined several protocols using a porcine model in order to improve the efficacy of islet isolation from marginal donors. Our first protocol followed a current standard procedure for clinical islet isolation (4–6). Pancreata were preserved with the TLM, islets were isolated with the Ricordi method and purified with a Ficoll-based solution. With this protocol, average postpurification islet yield was 2,486 IE/g, and, assuming that the average human pancreas weighs about 80 g, total islet yield would be about 200,000 IE. This yield is not enough for transplantation for patients of average body weight. We therefore used ductal injection with the ET-Kyoto solution supplemented with a trypsin inhibitor and a modified TLM using the ET-Kyoto solution. The ET-Kyoto solution was proven to be superior to the UW solution as a component of the TLM before islet isolation (7). Trypsin inhibition was also shown to improve islet isolation efficacy when marginal donors are used (4). Ductal injection was demonstrated to protect the pancreatic duct and improve islet isolation efficacy in a rat model (8). We used these three protocols to improve the efficacy of islet isolation from NHBDs. Because ductal injection needs additional time, WIT was significantly longer for the Kyoto-Ficoll and Kyoto-Iodixanol groups. However, the phase I period was significantly shorter for the Kyoto-Ficoll and Kyoto-Iodixanol groups, suggesting that use of these strategies enhanced pancreatic digestion. We therefore speculated that modified TLM and ductal injection improved efficacy of collagenase delivery.

The Kyoto-Ficoll protocol significantly increased islet yield before purification; but the average postpurification recovery rate was only 38%. Because it was previously shown that an Iodixanol-based solution significantly improved porcine islet purification (9,10), we also used it. Iodixanol has a lower viscosity than Ficoll then the islets should suffer less force. The fact that the average islet size was significantly larger and the morphological score better for the Kyoto-Iodixanol group seems to support this concept. An important finding was that the postpurification recovery rate was significantly improved with the Iodixanol-based solution. For these reasons, our current protocol uses Iodixanol for islet purification.

We isolated human islets with the Kyoto-Ficoll protocol in two cases and with the Kyoto-Iodixanol protocol in 11 cases. The postpurification recovery rate was improved with Iodixanol like porcine experiments. Total islet yield was 339,480 IE with Ficoll and this increased to 410,376 IE with Iodixanol. The important outcome is that with the Kyoto-Iodixanol protocol we were able to transplant 10 cases out of 11 cases (91%) using NHBDs. Because even when heart-beating brain-dead donors and TLM pancreas preservation were used, the transplant rates were about 50% (6); therefore, we believe that our Kyoto-Iodixanol protocol substantially improved islet isolation using marginal donors. The fact that all transplanted islets started to secrete insulin and blood glucose control of all patients improved underscores the suitability of our current islet isolation method.

All six transplanted patients improved glycemic control and reduced insulin dosage with positive C-peptide. Two patients became insulin free and other two patients reduced insulin dosage less than 10 units. Because we focused on better glycemic control even using insulin, the latter two patients might be insulin free if we might accept HbA1c more than 6.0%. These findings suggest that it should be possible to cure type 1 diabetes by means of islet transplantation using our current islet isolation method with NHBDs.

This islet isolation protocol could be applied for islet isolation with brain death donor and living donor pancreata. Indeed, we used this protocol for the first successful living donor islet transplantation (17).

Further study might be necessary to evaluate our islet isolation and transplantation protocol. Nevertheless, the results reported here encourage us to make the pass for the treatment of type 1 diabetes by islet transplantation with NHBDs.

Back to Top | Article Outline


The authors wish to thank Chikao Yamazaki, Osamu Kato, Akika Uwano, Tomoko Asai, Masaki Otsuka, Katsunori Miyake, and Miyuki Hara for their assistance and arrangement of organ procurement, Michiko Ueda, Akemi Ishii, Yusuke Nakai, Emi Yabunaka, Yurika Uchida, Yoko Nakagawa, Hiroko Muramatsu, Hiroko Matsumura, and Hiroaki Tsuji for their technical assistance, Tomosaki Masuda for sirolimus concentration measurements and all members of the Japan Society for Pancreas and Islet Transplantation, as well as Mike D. Strong and Laura Schreiber for their careful review of this manuscript.

Back to Top | Article Outline


1. Shapiro AMJ, Lakey JRT, Ryan EA, et al. Islet transplantation in seven patients with type1 diabetes mellitus using a glucocorticoid-free immunosuppressive regimen. NEJM 2000; 343: 230.
2. Ricordi C. Islet transplantation: a brave new world. Diabetes 2003; 52: 1595.
3. Markmann JF, Deng S, Desai NM, et al. The use of nonheartbeating donors for isolated pancreatic islet transplantation. Transplantation 2003; 75: 1423.
4. Matsumoto S, Rigley HT, Reems JA, et al. Improved islet yields from macaca nemestrina and marginal human pancreata after two-layer method preservation and endogenous trypsin inhibition. Am J Transplant 2003; 3: 53.
5. Matsumoto S, Rigley T, Qualley S, et al. Efficacy of the oxygen-charged static two-layer method for short-term pancreas preservation and islet isolation from nonhuman primate and human pancreata. Cell Transplant 2002; 11: 769.
6. Ricordi C, Fraker C, Szust J, et al. Improved human islet isolation outcome from marginal donors following addition of oxygenated perfluorocarbon to the cold-storage solution. Transplantation 2003; 75: 1524.
7. Ridgway DM, Swift SM, White SA. Improvement of islet yield by pancreas preservation using trehalose containing solution. Acta Diabetol. 2003; 40: 37.
8. Sawada T, Matsumoto I, Nakano M, Kirchhof N, Sutherland DER, Hering BJ. Improved islet yield and function with ductal injection of University of Wisconsin solution before pancreas preservation. Transplantation 2003; 75: 1965.
9. van der Burg MP, Basir I, Bouwan E. No porcine islet loss during density gradient purification in a novel iodixanol in University of Wisconsin solution. Transplant Proc 1998; 30: 362.
10. Matsumoto S, Zhang HJ, Gilmore T, et al. Large-scale isopycnic islet purification utilizing non-toxic, endotoixn-free media facilitates immediate single-donor pig islet allograft function. Transplantation 1998; 66: S30.
11. Matsumoto S, Qualley S, Goel S, et al. Effect of the two-layer (University of Wisconsin solution-perfluorochemical plus O2) method of pancreas preservation on human islet isolation, as assessed by the Edmonton Isolation Protocol. Transplantation 2002; 74: 1414.
12. Ricordi C, Gray DW, Hering BJ, et al. Islet isolation assessment in man and large animals. Acta Diabetol Lat 1990; 27: 185.
13. Bank. Rapid assessment of islet viability with acridine orange and propidium iodide. In Vitro Cell Dev Biol 1988; 24: 266–273.
14. Baidal DA, Froud T, Ferreira JV, et al. The bag method for islet cell infusion Cell Transplant 2003; 12: 809.
15. Lakey JRT, Warnock GL, Rajotte RV, et al. Variables in organ donors that affect the recovery of human islets of Langerhans. Transplantation 1996; 61: 1047.
16. Matsumoto S, Zhang G, Qualley S, et al. Analysis of donor factors affecting human islet isolation with current isolation protocol. Transplant Proc 2004; 36: 1034.
17. Matsumoto S, Okitsu T, Iwanaga Y, et al. Insulin independence after living-donor distal pancreatectomy and islet allotransplantation. Lancet 2005; 365: 1642.

Islet transplantation; Islet isolation; Nonheartbeating donor; ET-Kyoto solution; Iodixanol

© 2006 Lippincott Williams & Wilkins, Inc.