|
|
| |||||||||||||||||||||||||||||||||
Background Myoblast transfer has been proposed as a technique to replace dystrophin, the skeletal-muscle protein that is deficient in Duchenne's muscular dystrophy. Donor myoblasts injected into muscles of affected patients can fuse with host muscle fibers, thus contributing their nuclei, which are potentially capable of replacing deficient gene products. Previous controlled trials involving a single transfer of myoblasts have been unsuccessful.
Methods We injected donor muscle cells once a month for six months to the biceps brachii muscles of one arm of each of 12 boys with Duchenne's muscular dystrophy. The opposite arms served as sham-injected controls. In each procedure 110 million cells donated by fathers or brothers were transferred. The patients were randomly assigned to receive either cyclosporine or placebo. Strength was measured by quantitative isometric muscle testing. Six months after the final myoblast transfer, the presence of dystrophin was assessed with the use of peptide antibodies specific to the deleted exons of the dystrophin gene.
Results There was no significant difference in muscle strength between arms injected with myoblasts and sham-injected arms. In one patient, 10.3 percent of muscle fibers expressed donor-derived dystrophin after myoblast transfer. Three other patients also had a low level of donor dystrophin (<1 percent); eight had none.
Conclusions Myoblasts transferred once a month for six months failed to improve strength in patients with Duchenne's muscular dystrophy. The value of exon-specific peptide antibodies in the interpretation of myoblast-transfer results was demonstrated in a patient with Duchenne's muscular dystrophy who had a high percentage of donor-derived dystrophin. Specific variables affecting the efficiency of myoblast transfer need to be identified in order to improve upon this technique.
This study differs from previous controlled investigations in several ways. First, instead of a one-time transfer, myoblasts were transferred from donor to host muscles once a month for six months. One study had suggested that myoblasts could assume a satellite-cell position and be available for fusion at a subsequent time.13 Second, dystrophin expression was measured by the use of peptide antibodies specific to the deleted exons of the dystrophin gene. This method allows one to distinguish between dystrophin-positive fibers derived from donor DNA and host fibers that had reverted to normal ("revertant" fibers).14,15 Third, patients were randomly assigned to receive cyclosporine or placebo, permitting a direct assessment of the need for antirejection drugs and of the effect of cyclosporine on muscle strength.
Methods
Patient Population
Twelve boys with Duchenne's muscular dystrophy participated in this one-year randomized, double-blind trial of myoblast transfer (Table 1). Informed consent was obtained from the parents of all subjects; clinical outliers were excluded.16 To be eligible for the study the boys had to be ambulatory; be 5 to 10 years of age; have an identifiable deletion in the dystrophin gene in peripheral-blood leukocytes17,18; have no dystrophin on Western blot analysis of a muscle-biopsy specimen2; be able to cooperate for efficacy testing; have an available donor free of serologic evidence of the human immunodeficiency virus, human T-cell lymphotrophic virus type I, hepatitis B virus, cytomegalovirus, Epstein–Barr virus, or syphilis; and have a haplotype compatible with that of the donor (Table 1).19,20
|
Selected heterogeneous populations of normal human myoblasts were isolated without the aid of mechanical sorting, according to established culture techniques.21,22 The method yielded approximately 1x108 cells by the second passage. Muscle samples of 1 g were removed from the biceps brachii of all donors. Half the sample was minced and dissociated immediately; the other half was maintained at 4°C for seven days before dissociation. The dissociation cocktail consisted of 0.01 percent trypsin (Worthington Biochemical, Freehold, N.J.), collagenases I and II (Sigma, St. Louis), and 0.02 percent EDTA. The delay before isolation provided a greater yield of myoblasts in some but not all cases. After the dissociation step, the muscle cultures were treated identically at every stage. The growth medium for all cultures was Ham's MCDB 121 (Clonetics, San Diego, Calif.) supplemented with 15 percent fetal-calf serum. For each culture we verified that more than 95 percent of cells expressed the myoblast-specific neural-cell adhesion molecule using antibody 5H.11 (courtesy of Frank Walsh, Ph.D.)23 with a Coulter Elite fluorescence-activated cell sorter.
Before myoblast transfer all donor cultures were tested for fusion competency. Each culture was plated in growth medium at 1x104 per square centimeter, allowed to reach 80 percent confluency, and then incubated for 14 days in a fusion medium (Ham's F10) of which 5 percent consisted of the patient's own serum (previously harvested), with 10 µg of insulin per milliliter.22 After trypsinization, the cells were counted with a phase microscope with a standard hemocytometer. For the first transfer the percentage of cells capable of fusing (the number of multinucleated cells/the total number of cells) was as follows: Donor 1, 58 percent; Donor 2, 48 percent; Donor 3, 53 percent; Donor 4, 67 percent; Donor 5, 70 percent; Donor 6, 37 percent; Donor 7, 50 percent; Donor 8, 50 percent; Donor 9, 62 percent; Donor 10, 74 percent; Donor 11, 74 percent; and Donor 12, 67 percent. We demonstrated the ability of the multinucleated cells (myotubes) to express dystrophin using standard immunostaining methods.14,15
Twenty-four hours before each myoblast transfer, bovine proteins in the growth medium were replaced with the patient's own serum (15 percent). Immediately before injection the myoblasts were harvested, washed in phosphate-buffered saline, and resuspended in phosphate-buffered saline with 1 percent human serum albumin (Baxter Laboratories, Deerfield, Ill.), at a final concentration of 1x106 cells per 100 µl. Two hundred microliters, representing 2 million cells, or an equivalent volume of diluent without cells was loaded into individual 1-ml precoded (myoblast injection or sham injection) syringes for injection. An aliquot of each prepared suspension of myoblasts demonstrated more than 98 percent viability on the basis of the trypan-blue exclusion method.
Before the myoblasts were transferred, additional tests for adventitious agents (bacteria, fungi, mycoplasma, and virus) and pyrogenicity were performed by the Ohio State University Hospitals Department of Microbiology, in accordance with the Public Health Service Diagnostic Laboratory practices and the guidelines of the American Type Culture Collection for quality control of cell lines.24,25,26
Myoblast-Transfer Procedure
Only the biceps brachii muscles were injected. The patients' arms were randomly assigned to receive either myoblast injections or sham injections (Table 1). Before each transfer procedure, the patients received intramuscular meperidine hydrochloride (2.0 mg per kilogram of body weight) and intravenous lorazepam (0.05 mg per kilogram). Fifty-five sites, each 5 mm apart, distributed in 11 rows and 5 columns, were injected throughout the depth of the muscle with a 1.5-inch, 27-gauge needle. The patients and investigators were unaware of which arm received the myoblasts and which the sham injections.
Immunosuppression
The patients were randomly assigned to receive placebo or oral cyclosporine (5 mg per kilogram per day in divided doses) beginning the day before the initial transplantation and continuing throughout the entire 12 months of the study. On the day before each transplantation and for three days thereafter, all patients (irrespective of whether they had been assigned to placebo or cyclosporine) received prednisone at a dose of 2 mg per kilogram per day to diminish inflammatory infiltration at the injection sites in the peritransplantation period.
Clinical Evaluation
The primary efficacy variable for myoblast transfer was the maximal voluntary isometric strength of elbow flexion (Biomedical Designs, Edmonton, Alberta, Canada). The system uses a force transducer with a load of 0.5 to 1000 N (approximately 0.05 to 100.0 kg). For each muscle group the final value recorded represented the average of three consecutive trials.
The effects of cyclosporine on muscle strength at sites distant from the site of the myoblast transfers were determined by recording the maximal voluntary isometric strength of elbow extension and by manual muscle testing of 34 muscle groups (excluding the injected muscles) to obtain the average muscle score.11,16
Muscle Biopsies after Myoblast Transfer
Biopsies of bilateral biceps brachii muscle were performed at one year (six months after the final myoblast transfer). Superficial and deep pieces of vertically aligned muscle, each measuring 0.5 by 1 cm, were removed, and the samples were processed for histochemical analysis and immunostaining for dystrophin as previously described.14,15,27,28 Dystrophin antibodies included those directed to the carboxy terminal (NCL-DYS2, Novacastra Laboratories, Newcastle upon Tyne, United Kingdom) and polyclonal and monoclonal antibodies (Table 2) raised against exon-specific peptides of the deleted exons of the patients. The polyclonal antibodies included Ab2 and Ab4 for exons 8 and 51, respectively,28 and 9218 for exons 48 through 52.15 The monoclonal antibodies were as follows: MANDYS1-15 for exons 31 and 32,27 MANDYS17 for exons 26 and 27,27 MANDYS106 for exon 43,27 MANEX45A for exon 45,29 and MANEX50 for exon 50.29
|
Statistical Analysis
Data were collected with MS Excel and transferred to SAS JMP for analysis. Before conducting the statistical analysis, we divided the study into five periods and calculated aggregate data for each response variable as follows: period 0, mean of two base-line measurements; period 1, mean of months 1 through 3; period 2, mean of months 4 through 6; period 3, mean of months 7 through 9; and period 4, mean of months 10 through 12. After the aggregate data were calculated, a second manipulation was performed. The difference from base line was calculated for each period, and the differences from base line for periods 2 through 4 were used in the statistical analysis. All analyses used a repeated-measures analysis of variance model in which the response variable was the calculated difference from base line for each quantitative measure (i.e., quantitative isometric muscle strength of elbow flexor, average muscle score, and so on). In the analysis-of-variance model, the treatment period was included as a nominal factor, immunosuppression (cyclosporine vs. no cyclosporine) was included as a treatment factor, and subjects were included as random factors within each treatment group. An interaction term involving the treatment factor and time factor was also included. A P value of 0.05 or less was considered to indicate statistical significance.
Results
Maximal Voluntary Isometric Strength
Myoblast transfer had no effect on muscle strength (Figure 1, Figure 2A), and Figure 2B). Patients underwent myoblast transfer at base line (month 0) and months 1, 2, 3, 4, and 5 for a total of six transfers. In each case, the maximal voluntary isometric strength was measured the day before myoblast transfer. The final test of efficacy was performed at 12 months. Figure 1 shows data from all patients irrespective of whether they received immunosuppressive therapy. At no point was there a statistically significant difference in the degree of change from base line between arms injected with myoblasts and sham-injected arms. Figure 2A and Figure 2B shows the results in the two groups according to whether cyclosporine was given. Cyclosporine therapy had no beneficial effect on myoblast transfer.
|
|
The expression of dystrophin was initially examined after myoblast transfer with a monoclonal antibody to the carboxy terminal. There were no significant differences in the mean (±SD) number of dystrophin-positive fibers between the myoblast-injected arms (superficial region, 2.57±3.18 fibers per square millimeter; deep region, 2.68±2.7 fibers per square millimeter) and sham-injected arms (superficial region, 2.69±3.87 fibers per square millimeter; deep region, 3.96±5.23 fibers per square millimeter). Without the use of exon-specific peptide antibodies for deleted exons it was not possible to determine whether any dystrophin-positive fibers in the arms receiving myoblasts expressed donor-derived dystrophin or represented revertant fibers.14,15 A panel of peptide-specific antibodies raised against deleted exons of the patients unequivocally identified muscle fibers expressing donor-derived dystrophin. As shown in Table 2, in 4 of the 12 patients muscle fibers expressing donor-derived dystrophin were identified. Only a single positive fiber was found in the entire cross section of muscle from Patients 3 and 6, and two positive fibers were identified in muscle from Patient 12. This small number is less than the 1 percent of revertant fibers that express dystrophin that are usually observed in patients with Duchenne's muscular dystrophy.14,15 In contrast, Patient 5 had both small and large fibers expressing donor-derived dystrophin (Figure 3A, Figure 3B, and Figure 3C): 10.3 percent of the muscle fibers exhibited dystrophin contributed by the patient's brother (420 dystrophin-positive fibers of a total of 4080 fibers). The number of dystrophin-positive fibers was nearly equally distributed in the superficial and deep regions of the transplanted muscle. Patient 5 and his mother both had the same deletion of exon 45 (data not shown), excluding somatic mosaicism as an explanation for the patient's dystrophin-positive fibers after myoblast transfer.
|
Six boys were treated with 5 mg of cyclosporine per kilogram per day, and six boys received placebo for one year. As shown in Figure 4, there was no significant difference in the average muscle scores between the cyclosporine-treated boys and those given placebo. The loss of muscle strength in both groups was similar to the previously established natural rate of decline.16 In addition to manual muscle testing, we studied the right and left elbow extensors using maximal voluntary isometric strength as a measure of efficacy in the cyclosporine-treated and placebo groups. This sensitive measure also failed to identify significant differences in strength between the two groups (Figure 5).
|
|
Blood levels of cyclosporine were monitored every month and adjusted to maintain therapeutic levels. Most patients had transient soreness in the injected muscles that typically resolved within three to four days of the procedure. No local or systemic evidence of immune-mediated rejection was seen.
There was no significant difference in the number of side effects between the patients treated with cyclosporine and those given placebo, although excessive hair growth was more common in the cyclosporine-treated group (reported at eight follow-up visits, as compared with one visit in the placebo group) (Table 3). Mild upper respiratory tract infections or otitis media that resolved with antibiotic treatment was reported at six follow-up visits in the cyclosporine group and seven follow-up visits in the placebo group. One patient in each group had transient urticaria lasting two to three weeks that resolved without treatment. In a single patient (treated with cyclosporine) a focal cellulitis developed over the buttocks, requiring drainage and antibiotic treatment. None of the monitored laboratory values differed significantly between the two groups. No patient had to discontinue cyclosporine therapy prematurely, and none had dose-limiting changes in blood pressure or renal function.
|
The present study evaluated myoblast transfer as a treatment for Duchenne's muscular dystrophy.31 We used a multiple-injection protocol in which myoblasts were transferred once a month for six months. In a previous report, two transfers of myoblasts to the same muscle in one patient seemed to increase the number of dystrophin-positive fibers.32 Our patients received twice the number of myoblasts per injection site as were given in previous controlled studies.6,7,9 A total of 660 million myoblasts were delivered to the biceps brachii muscle of each patient. Law et al. delivered a total of 5 billion cells through 48 injections into the major muscle groups in both legs of patients in an uncontrolled trial.10 We observed no improvement in muscle strength in any patient, including the boy with 10 percent dystrophin-positive fibers.
An important step in assessing myoblast therapy is determining the extent of expression of donor-derived dystrophin. Gussoni et al. used a method involving the reverse-transcriptase polymerase chain reaction and concluded that 1 percent of host muscle fibers in the transplanted region expressed donor-derived dystrophin.6 Although messenger RNA was detected, the levels were low and could have arisen from transcription in myoblasts that had not fused to host muscle fibers. Indeed, these investigators found no difference in the number of dystrophin-positive fibers between muscle injected with myoblasts and sham-injected muscle. The dystrophin-positive fibers on the sham-injected side presumably represented revertant fibers, believed to arise from a second-site mutation in the dystrophin gene that corrects the reading frame.14,15 The problems in distinguishing the source of dystrophin in the transplanted region are further illustrated by a patient described by Karpati et al.; in this patient 5 percent of fibers were positive for dystrophin, yet no donor-derived DNA or RNA was found.7 In the present study, antibodies were raised to peptides from the deleted DNA exons of each patient. Four patients had donor-derived dystrophin. In three, the proportion of dystrophin-positive fibers was meager, but in a single patient the proportion was considerable (10.3 percent of all fibers). The dystrophin-positive fibers were distributed equally throughout the superficial and deep regions of the biceps muscle, probably reflecting the method of myoblast injection throughout the belly of the muscle. Our observation is a convincing demonstration of the transcription and translation of dystrophin by donor myoblast cells. The 10 percent level of dystrophin expression observed in one patient also has particular importance, since Cox et al. estimated that a 20 percent level of dystrophin expression was potentially protective in experiments with transgenic mice.33 Unfortunately, we found no accompanying increase in muscle strength.
It would be of interest to know what factors governed the survival, fusion, and expression of donor myoblasts in our patients.7 Connective-tissue barriers alone cannot account for the lack of success, since the patient with the highest level of dystrophin expression was one of the two oldest in the study (age, 9 years) (Table 1). The older age of this patient also tends to negate the possibility that favorable results of myoblast transfer depend on the presence of a greater number of random clusters of regenerating fibers, a finding more prevalent in young patients. The likelihood that the presence of basal lamina hinders myoblast fusion is also not supported by our successful results in the older patient. Immune-mediated rejection probably does have a role, as illustrated in a series of studies by Huard et al.8,13 and Tremblay et al.9 In our study the only patient with positive results was taking cyclosporine. Whether this is the predominant factor underlying the relative success of the procedure in a single patient cannot be determined. We did not measure serum antibodies to dystrophin.13 Possibly other immunosuppressive regimens could increase the success of myoblast transfer, as indicated by improved results in mdx mice given tacrolimus (FK 506).34
The use of prednisone at the time of myoblast transfer could have potential inhibitory effects on myoblast fusion, but in vivo results might be difficult to predict. In cultures of muscle from patients with Duchenne's or Becker's muscular dystrophy, methylprednisolone inhibits myoblast fusion.35 In contrast, glucocorticoids have no effect on normal cultures35 but enhance the expression of dystrophin in cultures of normal muscle and muscle from patients with Becker's muscular dystrophy in which fusion is arrested.36
A secondary outcome variable of our study concerned the effects of cyclosporine in patients with Duchenne's muscular dystrophy. Sharma et al. reported that treatment with cyclosporine (5 mg per kilogram per day) increased the strength of tibialis anterior muscles, as measured by tetanic force and maximal voluntary isometric contraction.12 The improvement occurred as early as two weeks after treatment began and persisted for eight weeks. In that open-label study, the response in cyclosporine-treated patients was compared with the natural rate of decline of muscle strength in Duchenne's muscular dystrophy. We were unable to confirm a cyclosporine-induced improvement in strength. We found no difference in strength between cyclosporine-treated patients and patients given placebo by using manual muscle testing11 or maximal voluntary isometric strength of elbow extensors. Although because of our small sample (12 patients) the results of this negative study should be interpreted with caution, the accuracy of the findings is supported by the fact that neither group showed significant deviations from our previously established natural rate of decline in muscle strength in Duchenne's muscular dystrophy.11,17,37
In conclusion, our study indicates that myoblast transfer in its current state of development does not improve strength in patients with Duchenne's muscular dystrophy despite the transfer of a large number of cells. However, we also demonstrated that myoblast transfer can be used to deliver cells capable of fusing with host muscle fibers and expressing donor-derived dystrophin. Only through further work will it be possible to identify the important variables affecting the efficiency of transfer and thus making the use of myoblasts a potential form of cell therapy for muscle diseases.
Supported by the Muscular Dystrophy Association, the Muscular Dystrophy Group of Great Britain and Northern Ireland, and a grant from the National Institutes of Health (Mo1-RR-00034).
We are indebted to the patients and their families; tremendous demands were placed on them in terms of time, effort, and expense in complying with all the complexities of the study.
Source Information
From the Departments of Neurology (J.R.M., J.T.K., A.A.A., W.K., L.S., T.W.P., Z.S., S.B., P.E.M., R.R., A.H.M.B.), Pathology (J.R.M., T.W.P., R.S., L.L.), Medical Biochemistry (A.H.M.B.), Statistics (H.N.), and Molecular Genetics (P.E.M., A.H.M.B.), Ohio State University, Columbus; and the North East Wales Institute, Research Division, Deeside, Clwyd, United Kingdom (G.E.M.).
Address reprint requests to Dr. Mendell at the Department of Neurology, Rm. 463, Means Hall, Ohio State University, 1654 Upham Dr., Columbus, OH 43210.
References
| |||||||||||||||||||||||||||||||||
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2008 Massachusetts Medical Society. All rights reserved. |
Previous
