A Comparison of Islet Cell Transplantation With Medical Therapy for the Treatment of Diabetic Eye Disease

Study Purpose

The purpose of the study is to determine if islet cell transplantation (ICT) is more effective than intensive medical therapy at reducing the progression of diabetic retinopathy.

Hypothesis

ICT will result in less progression of diabetic retinopathy than intensive medical therapy. The null hypothesis states that there is no difference between ICT and intensive medical therapy in the chance of progression of diabetic retinopathy.

Study design: prospective, single-centre, randomized clinical trial

Outcome Assessment

Main Efficacy Outcome The primary outcome measure is a composite consisting of progression of DR, development of significant ME, or moderate visual loss in patients who have significant ME at randomization. The definition of each term is described below. The unit of analysis is an individual eye and an eye can only be counted as progressing once.

Rationale for Primary outcome Progression of either DR or DME is clinically meaningful since each can threaten vision. Since the chance of progression in one of these manifestations of diabetic eye disease is independent of the chance of progression in the other, the combined endpoint of progression in either is designated as the primary outcome. Progression of each condition is assessed differently and will be done using generally accepted criteria. The primary eligibility criteria is the level of DR since this is where the benefit found in our initial study was identified. We are attempting to determine the benefit of our intervention (ICT) at levels of DR before PRP is indicated and therefore before vision is threatened. For this reason, two of the three definitions of progression use a surrogate outcome, and one uses the clinical outcome of visual acuity. Surrogate outcomes can be used if they predict the clinical outcome of interest and this has been established for the levels of DR. The appropriate outcome measure for ME depends on its severity. For patients without CSME, the surrogate outcome is the proportion progressing to CSME. For those with CSME, the appropriate outcome is the clinical measure of the incidence of moderate visual loss. (PKC-DRS2 2006, Hauser 2008)

Progression of Diabetic Retinopathy is defined as either:

1. An increase of 2 grades in an individual eye
2. Development of PDR (grade ≥ 61)

The progression must be present on 2 consecutive visits separated by at least 4 months.

Progression of Macular Edema is defined as either:

1. In eyes with central subfield mean thickness (CSMT) measured by OCT < 300 μm at baseline, the development of CSMT > 300 μm AND and increase of > 50 μm.
2. In eyes with CSMT measured by OCT > 300 μm at baseline, a vision loss of ≥ 15 letters.

VA is measured using the ETDRS protocol. (ETDRS 1991c). The outcome is the development of moderate visual loss, a decrease in the ability to read 15 letters on this 90 letter scale. This corresponds to a loss of 3 lines of visual acuity or the doubling of the visual angle (eg from 20/20 to 20/40).

The progression must be present on 2 consecutive visits separated by at least 4 months.

Sample Size Calculation Metabolic control, baseline retinopathy and diabetes duration are the only independent predictors of progression to PDR (Porta 2001, Skrivarhaug 2006). A recent study examining the effect of protein kinase C inhibition on the progression of DR enrolled subjects a similar spectrum of DR severity as is planned for our study (PKC-DRS 2005). The progression rate over 3 years in the study was 50%, lower than their estimated 64% (ETDRS 1991c). We will use the more conservative estimate of 50% for our sample size calculation.

We anticipate that 30% of our subjects will have significant ME at the time of randomization and that 25% of this group will experience moderate visual loss over 3 years. (Klein 1998, PKC-DRS2 2006, PKC-DRS 2005) Of the anticipated 70% without significant ME at randomization, 25% can be expected to develop ME over 3 years (ETDRS 1985). Since the risk of progression is 25% whether or not ME is present at entry, our sample size calculation will not be influenced if the proportions of the recruited population is different than the projected 30:70

This literature analysis suggests that the composite endpoint can be expected to occur in 75% of the control group over 3 years. For conservative purposes, a rate of 65% will be used for the calculation. We wish to be able to detect a 50% reduction to a rate of 32.5%. For α = .05 and a power of 80%, 36 eyes (18 subjects) are needed in each group. We plan to recruit 20 subjects for each group (total 40) to allow a dropout of 10%.

Ability to perform study Our rate of donor pancreases suitable for ICT has averaged 1/month since we began in March 2003 (Thompson 2008). Although 2 pancreases are usually required to allow a subject to become insulin independent, we have found that the benefits of ICT on microvascular complications begin after a single islet infusion, whether or not the subject is able to stop insulin. Therefore, we should be able to have sufficient donors to allow 20 subjects to receive an islet infusion in 2 years. For conservative purposes, to allow for unexpected events such as a decrease in the availability of donor pancreases, we will allow 3 years for all subjects to be randomized.

Step 1 - Recruitment

It is anticipated that most subjects will be identified during routine-care examinations with their endocrinologist or ophthalmologist. The Islet Transplant Program will make ophthalmologists and endocrinologists in British Columbia aware of this study and we believe that all recruitments will come from these physicians. They will be asked to provide any subjects who are potentially eligible and express interest with contact information. Subjects will be able to contact the program by phone (604-875-5997), fax (604-875-5925) or email (isletlab@interchange.ubc.ca).

Step 2 - Initial Meeting

When a subject has identified themself as being interested in the study, they will be contacted by phone by the study coordinator and asked to schedule a preliminary meeting with the PI and study coordinator who will provide further information about the study. At the end of the meeting they will be given the informed consent document to read and consider. There is no time limit given for a subject to determine if they wish to enter the study. Subjects who remain interested will proceed to the next step.

Step 3 - Further Assessment: the components of step 3 occur simultaneously

3a Suitability for ICT Meeting with transplant team to ensure eligibility and that the subject understands the risks of the transplant procedure and immunosuppression.

3b Intensive medical therapy Subjects to begin intensive insulin management, as with current program, to ensure that they will be able to comply with this aspect of the study prior to randomization and after if they are randomized to the medical arm. This is necessary so that ICT is compared with the best available current treatment. Failure to maintain regular contact with the nurse-clinicians, perform regular glucose monitoring and to achieve an A1c < 8.0 by the end of this assessment will be considered a reason to exclude the subject from the study.

3c Ocular assessment Complete eye examination including 7-field fundus photos, visual acuity and OCT to confirm eligibility. The photography will be performed by certified photographers at the Eye Care Centre. These photos will be read by one of the investigators (ISB) locally.

The time to complete step 3 is estimated to be approximately 2 months.

Step 4 Team meeting of the investigators to make a decision about each subject regarding enrollment in study. Eligible subjects will be asked to sign an informed consent at this time.

Step 5 Subjects are now considered to be enrolled in the study and will continue intensive medical management based on the 2008 CDA guidelines. Failure to comply with the requirements of intensive therapy will disqualify a patient from the study.

Randomization will be in blocks: 8 subjects per block for blood groups A and O and 4 subjects per block for groups B and AB. Eight subjects are chosen for the common blood groups to make it highly likely that at least 4 would be cross-match negative with a particular donor and thus eligible to receive the islet infusion from that donor. The smaller number is used for the less common groups so that they can participate in the study. All subjects remain on intensive medical therapy until a block is full.

Steps 6 and 7 A completed block is activated. Prior to activation, the subjects understanding of the trial, willingness to accept the assigned treatment group and commitment to the follow-up schedule will be confirmed.

Half of these subjects will randomly be assigned to receive and ICT and the other half will continue with intensive medical therapy for the duration of the study. If the study demonstrates that ICT is more effective than medical therapy in preventing progression of diabetic retinopathy, subjects who have not received an ICT as part of this study will be offered the chance to have an ICT as part of their ongoing care.

The random number sequence will be generated by computer and the numbers placed in numbered, opaque, sealed envelopes which will be stored in the locked islet transplantation laboratory. These will be prepared before any subjects have been enrolled. Each numbered, sealed envelope will contain two additional sealed but unmarked envelopes - one containing a randomization number for medical treatment and the other for ICT.

Randomization will occur at the time a donor becomes available. The two best matched subjects for the donor will be identified. This identification requires the agreement of at least two of the investigators. In the presence of at least 2 investigators, a numbered envelope will be obtained from the islet lab and opened. The name of one subject will be written on one of the sealed, unmarked envelopes and the name of the other subject on the other envelope. Only then will the envelopes be opened to reveal the code. One of these subjects will then be randomized to receive the ICT and the other will continue medical therapy.

No new subjects of this blood group will be activated until half have received an ICT and the other half have been randomized to medical therapy. At this time, another group of subjects of this blood group will be activated.

Assessment of study outcomes will begin from the time of randomization of each 2 subjects (one ICT, one medical). A complete ocular assessment will be performed within 1 week of randomization.

Follow-up Phase

Blinding The subjects and investigators administering the interventions by necessity will be aware of the treatment allocation. The outcome assessment will be blinded in the following manner. The eye photographers and technicians performing the photographs and VA testing will not be aware if the subject has received a transplant. Once collected this information will be marked with a unique identifier and shipped to the University of Wisconsin Fundus Photograph Reading Center for analysis. The assessors at the center will have only the unique identifier number and therefore will not know which treatment the subject is receiving.

Outcome Assessment Ocular assessment (for complete eye exam, ETDRS best corrected VA, Fundus photos, OCT) will be performed every 4 months ± 1 month from the time of randomization.

Islet transplantation Islet isolation and transplantation are performed as described (Warnock 2005). Subjects will receive ≥ 5,000 IE/kg for their initial infusion and are considered to have received a completed islet transplant when they have received ≥ 12,000 IE/kg in 1 - 3 infusions.

Post-transplantation Care Subjects are followed in the Solid Organ Transplant Clinic at Vancouver General Hospital as described (Warnock 2005)

Diabetes Management Diabetes care is provided according to the standards recommended in the Canadian Diabetes Association 2008 Clinical Practice Guidelines. Subjects are followed closely as described (Warnock 2005)

Duration of Follow-up: 3 years from time the last patient is randomized

Statistical Analysis The unit of analysis is an individual eye. Analysis will be by intention-to-treat. At the end of the study, the proportion of eyes in each treatment group that have reached the primary outcome will be compared by Chi-square analysis.

There will be no interim data analysis. The DCCT found that transient early worsening of DR can occur in the first two years of intensive therapy and that analysis during this time would provide inaccurate information as to the long-term benefit of treatment that begins to appear after three years of follow-up (DCCT 1995b).