A Trial of Autologous Bone Marrow Derived Stem Cells in Paediatric Heart Failure

This pilot study is of a randomized, crossover, placebo controlled design. The Consolidated Standards of Reporting Trials (CONSORT) statement was used to guide the development of the protocol. Ethical Approval was obtained for this project in April 2008.

A CONSORT flow diagram was produced and a project plan for a crossover study designed. Although the primary end point of the study was safety and feasibility the investigator aimed to look at functional outcome as a secondary endpoint. Statistical review of the available literature indicated a 3% change in ejection fraction (EF) as measured by magnetic resonance imaging (MRI) would represent a meaningful change and calculated 10 patients would be sufficient to detect this mean change in EF, assuming a SD of difference of 3% with 80% power and 5% significance. For all calculations it was assumed that the distribution of differences was Normal. Calculation of a sample size for the crossover study required a definition of the minimum difference to be detected.

Children aged 1 year to 16 years at review, either attending Heart Failure Clinic at Great Ormond Street Hospital (GOSH) or referred for admission and management of their heart failure at GOSH will be invited to participate in the study. The lower age limit was selected to permit coronary artery catheterisation to be feasibly performed taking into account the size of the patient.

On recruitment all patients have impaired systolic function as determined by reduced shortening fraction on echocardiogram (more than 2 SD below normal). The functional status of the recruited patients was class 2-4 as defined by the NYHA or Ross classification.

Exclusion criteria were the need for high dependency or intensive care and congenital heart disease, viral infection that would preclude the use of hospital Cell Therapy Laboratory facilities, active malignancy and unstable cardiac drug therapy.

Potential patients will be identified from case notes, and they and their families approached either by telephone or following clinic review to evaluate their interest in participation. Any patients who appeared interested in participating in the study will be given age and language appropriate written information; and also given a contact telephone numberso that additional information could be supplied if required. Each family will be given the opportunity to discuss the study with the research team before consent for inclusion was obtained.

Protocol:

Ten patients who met the eligibility criteria will be recruited for this study. Each patient is randomized at entry to the study to determine whether they would receive stem cells at Period One or Period Two of the protocol. Randomisation was performed using a computer-generated random number table, Graph pad.

Following recruitment and randomisation, patients have their individual admission time-line planned. Those patients who are to receive stem cells at their first study admission are reviewed within the 28 days preceding that date. At this review, screening for infection, as per our hospital Cell Therapy Laboratory protocol (including HIV and hepatitis B and C) will be performed, following full written consent.

On the day of the procedure the patient will be admitted to the cardiac day-care facility for baseline observations and clinical assessment. The study procedure will be performed in the Cardiac MRI theatre under general anaesthetic. Blood samples will be taken at the time of anaesthesia for routine investigation (full blood count, urea, electrolytes, liver function and NTproBNP).

Those patients receiving intracoronary stem cells at this admission have 20mL of bone marrow aspirated from the posterior iliac crest under aseptic conditions immediately after induction of anaesthesia.

Bone marrow aspiration technique:

Those patients on anticoagulation therapy are highlighted in advance of all procedures and admissions, and optimal coagulation management implemented prior to, during and following the bone marrow aspiration. All bone marrow processing is undertaken at the Great Ormond Street Hospital cell therapy processing facility.

Following anaesthetic induction and the securing of a stable airway the patient is positioned to permit bone marrow aspiration. Full aseptic conditions are observed throughout the procedure. A volume of 20 mL of bone marrow is aspirated from a single needle puncture site using heparinized 10mL syringes attached to the bone marrow needle following removal of the central trocar. The aspirated bone marrow is transferred to sterile heparinized universal specimen bottles. Immediately after aspiration the bone marrow sample is taken to the cell therapy laboratory for mononuclear cell separation using standard techniques of gradient centrifugation.

The mononuclear cell separation protocol takes approximately 90 minutes to complete. The bone marrow sample is processed to provide 1mL for intra-coronary injection and 1mL to be cryogenically frozen for further phenotypic analysis at a later date. A small aliquot of stem cell suspension is analysed to obtain total mononuclear cell counts per millilitres. This cell count represents the total number of mononuclear cells injected via intra-coronary catheterisation. In addition cell viability was also assessed.

Whilst waiting for the prepared stem cells, the cardiac MRI scan is performed using a breath hold protocol for image acquisition. Following the MRI scan the patient is transferred by means of the mechanized sliding table back into the angiography suite.

On return to the angiography suite the patient is maintained under general anaesthesia until notice is given by the stem cell laboratory that the stem cell processing is nearing completion. Following contact with the stem cell laboratory an arterial sheath is placed in the femoral artery of the study patient and advanced until its tip is sited in the left main coronary artery. Partial occlusion of the coronary artery was observed as demonstrated by ST segment changes and damping of the pressure trace is routinely monitored during such procedures. The 1mL volume is slowly injected via the coronary artery catheter into the left main coronary artery followed by a 1mL 0.9% saline flush over a period of 2 minutes.

Following the intracoronary injection of stem cell suspension the cardiac catheter is removed and haemostasis at the skin entry site achieved. The patient is then woken and recovered in a standard manner.

Patients are admitted to the cardiac ward overnight with routine monitoring of the arterial catheterisation access site and bone marrow aspiration site; cardiac care observations are continued throughout. Patients are discharged the following day and electively reviewed in the outpatient's clinic at three months post-procedure or advised to contact the Research Fellow sooner if any concerns.

Cross-over of interventions occurs at six months with the stem cell group receiving placebo via the cardiac catheterisation (no bone marrow harvest is required for the placebo arm of the study) on the second occasion. The same admission, monitoring and follow up will be arranged as for the stem cell stage of the study. The placebo used for the intracoronary injection used is 1mL 0.9% saline, consistent with the placebo solution used in adult studies. The 1mL of placebo is followed with a 1mL 0.9% saline flush for correlation with the stem cell arm of the study. An interim clinic review is again arranged at three months post procedure and a follow up cardiac MRI scan at six months. The time scale was chosen after review of the available adult literature which showed that effects on the left ventricle were usually seen within 6 months.

Those patients randomly allocated placebo at stage one of the study cross over to receive stem cells at stage two. Interim and final follow up arrangements are the same for the stem cell and placebo groups.

Anaesthesia was standardized for all patients and the same anaesthetist used for all procedures. Bone marrow preparation was also standardized and performed by the same scientist in each case. Computer data entry will be double-checked. Patient identifiers are removed from the anonymous images to allow blind analysis later on; in this way unbiased image analysis is performed. A single researcher using Osirix software performs off-line analysis of the data. As a second check of these data the researcher's results are evaluated, again blind to patient identifiers, by a second analyser. The key to the blinding of the MRI data is devised, applied and stored securely by a clinician not otherwise involved in the study. Indexed values correct the data for age, sex, height and weight and permit true comparisons to be made between the series of studies of the same individual and between different individuals.

Primary outcome measures:

The primary outcome measures were freedom from death and transplantation or any complication which could be considered related to bone marrow injection or anaesthesia (infection/malignancy/anaphylaxis/renal deterioration). A safety monitoring committee, including a cardiologist from Great Ormond Street Hospital not involved in the trial, an external cardiologist and a lay representative, will be established. Any adverse events will be reported to the committee which was empowered to halt the trial at any time if concerns arose.

Secondary outcome measures:

Ejection fraction, left ventricular volumes and mass are derived from cardiac MRI. Blood is taken for NT proBNP. In addition functional status is assessed by the New York Heart Association classification or the Ross classification in younger children.

Each patient has a cardiac MRI under general anaesthetic. The anaesthetist and anaesthesia protocol are the same for all patients.

All images are obtained with a 1.5-T MR scanner (Avanto, Siemens, Erlangen, Germany) using a 12-element phased-array coil for signal reception and the body coil for signal transmission. A vector ECG system is used for cardiac gating.

Left ventricular (LV) volumes are measured from contiguous short-axis steady state free precession cine covering both ventricles (7-13 slices, depending on the size of the child). Each slice is acquired in a single 10- to 15-second breath hold as previously described. All image processing is performed using in-house plug-ins for the open-source Osirix DICOM (digital imaging and communications in medicine) software. The LV end-diastolic volume (EDV) and end-systolic volume (ESV) are measured by manual segmentation of the endocardial borders in the short-axis data. Careful segmentation of the basal slices in conjunction with 4-chamber and LV long-axis views is performed to overcome problems with delineating the mitral valve. Ventricular stroke volume (SV) was the difference between the EDV and ESV, and ventricular ejection fraction (%) was (SV/EDV)×100.

For each of the secondary end points, EDV, ESV and EF, comparisons between treatments are made on an intention to treat basis. A crossover analysis on post treatment measures is conducted using an analysis of variance model, considering sequence, period and treatment effects. Carry-over effect is tested first and where there is no evidence of a significant effect, the treatment effect is tested. A subsequent generalized estimating equation (GEE) model was fitted which allowed for an adjustment for baseline measures.

Where the distribution of the data was clearly non-Normal the investigators use a logarithmic transformation. Stata is used for all statistical analyses and all tests were two-sided. A P value < 0.05 was considered significant.