The NEW ENGLAN D
`JOURNAL of M ED I CIN E
`
`VOL. 357 NO. 26
`
`(
`
`ESTABLISHED IN 1812
`
`DECEMBER 27, 2007
`
`WWW.NEJM.ORG
`
`2656 THIS WEEK IN THE JOURNAL
`
`2652
`
`PERSPECTIVE
`2649 Comparing Physicians on Efficiency A. Milstein
`and T.H. Lee
`Is Quality Improvement Improving Quality?
`A View from the Doctor's Office M. Vonnegut
`2653 One Step Forward, Two Steps Back - Will There
`Ever Be an AIDS Vaccine? R. Steinbrook
`
`ORIGINAL ARTICLES
`2657 Prophylactic Catheter Ablation for the Prevention
`of Defibrillator Therapy
`V.Y. Reddy and Others
`
`2666 Paclitaxel plus Bevacizumab versus Paclitaxel Alone
`for Metastatic Breast Cancer
`I<. Miller and Others
`
`2677 Local Dystrophin Restoration with Antisense
`Oligonucleotide PRO0Sl
`J.C. van Deutekom and Others
`
`2687 COL4Al Mutations and Hereditary Angiopathy,
`Nephropathy, Aneurysms, and Muscle Cramps
`E. Plaisier and Others
`
`CLINICAL PRACTICE
`2696 Localized Prostate Cancer
`P.C. Walsh, T.L. Deweese, and M.A. Eisenberger
`
`Owned & published by the MASSACHUSETTS MEDICAL SOCIETY© 2007.
`All ri9hts reserved. ISSN 0028-4793.
`
`000300833
`VOUR CUSTOMER HUMBER:
`1.11I111 II. I. 11 Ill., ;.1. I ••• I. I II I. I •• I I I 11, I I •• II .1
`HBXHBVJG *****************CAR- RT LOT**C- 017
`tt 000300833/5 # 20071227 20071227 0016
`28
`EBLIHG LIB IHFO RES UHIT
`1823
`UHIY WISCOHSIH-MADISOH
`9
`HLTH SCI LEARHIHG CTR
`750 HIGHLAND AVE
`MADISON WI
`53705-2221
`
`IMAGES IN CLINICAL MEDICINE
`2706 Mapping the Atrioventricular Node
`A.E. Epstein andj.l<. Kirklin
`
`e30 Small-Bowel lntussusception
`C.H . Wilson and S.A. White
`
`CASE RECORDS OF THE MASSACHUSETTS
`GENERAL HOSPITAL
`2707 A Man with Weakness in the Hands
`W.J. Triggs and D. Cros
`
`EDITORIALS
`2717 Ablation after ICD Implantation - Bridging
`the Gap between Promise and Practice
`N.A.M. Estes Ill
`
`2719 Skipping toward Personalized Molecular Medicine
`E.P. Hoffman
`
`SPECIAL REPORT
`2723 Military-Civilian Collaboration in Trauma Care
`and the Senior Visiting Surgeon Program
`E.E. Moore and Others
`
`2728 CORRESPONDENCE
`Effectiveness of Influenza Vaccination
`Sexuality and Health among Older Adults
`Ventricular Pacing in Sinus-Node Disease
`Autoimmune Diseases after Stem-Cell Transplantation
`Aspirin and Hormone Therapy for Prostate Cancer
`
`2739 BOOK REVIEWS
`
`2743 NOTICES
`2745 CONTINUING MEDICAL EDUCATION
`
`EDW·IG Ll89ARY
`UNIVERSITY OF WISCONSIN
`
`JAN
`
`2 2008
`
`750 Highland Avenue
`Madison, WI 53705
`
`

`

`The NEW ENGLAND JOURNAL of MEDICINE
`
`This material may be protected by Copyright law (Title 17 U.S. Code)
`
`1 ... 1 ______ __ _ o_ R_r_G_r_N_A_L_ A_R_T_r_c_L_E _________ ....... I I
`
`Local Dystrophin Restoration
`with Antisense Oligonucleotide PRO0Sl
`
`Judith C. van Deutekom, Ph.D., AnnekeA.Janson , B.S., leke B. Ginjaar, Ph.D.,
`Wendy S. Frankhuizen, B.S., Annemieke Aartsma-Rus, Ph.D.,
`Mattie Bremmer-Bout, B.S.,Johan T. den Dunnen, Ph.D., Klaas Koop, M.D.,
`AnnekeJ. va~ der Kooi, M.D., Ph.D., Nathalie M. Goemans, M.D., Ph.D.,
`SjefJ. de Kimpe, Ph.D., Peter F. Ekhart, M.Sc., Edna H. Venneker, M.D.,
`GerardJ. Platenburg, M.Sc.,JanJ. Verschuuren, M.D., Ph.D.,
`and Gert-Jan B. van Ommen, Ph.D.
`
`ABSTRACT
`
`BACKGROUND
`Duchenne's muscular dystrophy is associated with severe, progressive muscle weak(cid:173)
`ness and typically leads to death between the ages of 20 and 35 years. By inducing
`specific exon skipping during messenger RNA (mRNA) splicing, antisense com(cid:173)
`pounds were recently shown to correct the open reading frame of the DMD gene and
`thus to restore dystrophin expression in vitro and in animal models in vivo. We
`explored the safety, adverse-event profile, and local dystrophin-restoring effect of a
`single, intramuscular dose of an antisense oligonucleotide, PRO051, in patients with
`this disease.
`
`METHODS
`Four patients, who were selected on the basis of their mutational status, muscle con(cid:173)
`dition, and positive exon-skipping response to PRO051 in vitro, received a dose of
`0.8 mg ofPRO051 injected into the tibialis anterior muscle. A biopsy was performed
`28 days later. Safety measures, composition of mRNA, and dystrophin expression
`were assessed.
`
`From the Departments of Human and
`Clinical Genetics (J.C.D., A.A.j ., 1.B.G.,
`W.S.F., A.A.·R., M.B.-B.,J.T.D., G.-J .B.O.),
`Pathology (K.K.), and Neurology (J .J,V.),
`Leiden University Medical Center; and
`Prosensa B.V. (J .C.D., S.J.K., P.F.E., G.J.P.)
`both in Leiden, the Netherlands; the
`-
`Department of Neurology, Academic Med(cid:173)
`ical Center, University of Amsterdam, Am(cid:173)
`sterdam (A.J.K.) ; the Department of Pe(cid:173)
`diatric Neurology, University of Leuven,
`Leuven , Belgium (N .M.G.); and Afforce
`Healthcare, The Hague, the Netherlands
`(E.H.V.). Address reprint requests to Dr.
`van Deutekom at Prosensa B.V,, Was(cid:173)
`senaarseweg 72, 2333 AL Leiden, the Neth(cid:173)
`erlands, or at j.vandeutekom@prosensa.nl.
`
`N Englj Med 2007;357:2677-86.
`Copyright © 2007 Massachusetts Medical Soci,ty.
`
`RESULTS
`PRO051 injection was not associated with clinically apparent adverse events. Each
`patient showed specific skipping of exon 51 and sarcolemmal dystrophin in 64 to
`97% of myofibers. The amount of dystrophin in total protein extracts ranged from
`3 to 12% of that found in the control specimen and fro in 17 to 35% of that of the
`control specimen in the quantitative ratio of dystrophin to laminin al.
`
`CONCLUSIONS
`Intramuscular injection of antisense oligonucleotide PRO051 induced dystrophin
`synthesis in four patients with Duchenne's muscular dystrophy who had suitable
`mutations, suggesting that further studies might be feasible.
`
`N ENGLJ MEO 357;26 WWW. NEJM .ORG DECEMBER 27, 2007
`
`2677
`
`

`

`T h, N EW ENGLAND JOURNAL of MEDICINE
`
`UCHENNE'S MUSCULAR DYSTROPHY IS
`a severely debilitating childhood neuro(cid:173)
`muscular disease that affects 1 in 3500
`newborn boys.1 Progressive weakness of the skel(cid:173)
`etal muscles, cardiomyopathy, and respiratory
`failure are the most prominent features, but the
`brain can also be affected. 2,3 Virtually all patients
`are wheelchair-dependent by the age of 12 years,
`and most die in early adulthood. Improved venti(cid:173)
`lation techniques and glucocorticoid treatment
`have substantially improved fitness and muscle
`strength, prolonged mobility, and extended the
`expected lifespan from less than 20 years to 25 to
`35 years.4·6 However, there has been no treatment
`to prevent the eventual fatal outcome.
`Duchenne's muscular dystrophy is caused by
`deletions (approximately 72%) and duplications
`(approximately 7%) of one or more exons or
`point mutations (20%) in the 2.4-Mb DMD gene,7
`which encodes a protein, dystrophin, that is cru(cid:173)
`cial for sarcolemmal integrity. s-12 In patients with
`the disease, such mutations disrupt the open read(cid:173)
`ing frame and abrogate dystrophin synthesis. In
`contrast, mutations in the same gene that con(cid:173)
`serve the reading frame but lead to internally
`•10 often
`truncated or slightly altered dystrophins8
`cause the milder Becker's muscular dystrophy,
`which is characterized by a life expectancy that
`is longer than that of patients with Duchenne's
`muscular dystrophy.9,11
`Up to 50% of patients with Duchenne's mus(cid:173)
`cular dystrophy show evidence of rare, dystrophin(cid:173)
`positive fibers (revertant fibers) caused by spon(cid:173)
`taneous, clonal, frame-restoring skipping of
`stretches of exons.13-16 This finding has prompted
`the investigation of the potential for therapeutic
`conversion of Duchenne's muscular dystrophy
`into its nearest in-frame counterpart (i.e., Becker's
`muscular dystrophy) with the use of antisense
`techniques. Because of their capacity to skip an
`exon specifically by blocking its inclusion during
`splicing, 12 antisense oligonucleotides can correct
`the reading frame of DMD transcripts, yielding
`internally truncated dystrophins such as those
`associated with Becker's muscular dystrophy
`(Fig. lA and lB). Although in principle such a
`process is mutation-specific, the skipping of par(cid:173)
`ticular exons is theoretically therapeutic in a se(cid:173)
`ries of different mutations. Thus, a judicious choice
`of 10 exons may eventually correct more than
`85% of mutations in patients with Duchenne's
`muscular dystrophy.
`
`Two main types of compounds are being in(cid:173)
`vestigated for antisense-induced exon skipping:
`2'-0-methyl-modified ribose molecules with a
`fulHength phosphorothioate backbone (2OMePS)
`and phosphorodiamidate morpholino oligomers.
`Preclinical proof..of..concept has been obtained
`with both types of molecules in cultured muscle
`cells from a series of patients with Duchenne's
`muscular dystrophy with various mutations, 13,14,17-19
`as well as in the mdx mouse model and the gold(cid:173)
`en retriever muscular dystrophy (GRMD) dog mod(cid:173)
`el.15•16,20 In the mdx mouse, which carries a non(cid:173)
`sense mutation in exon 23, systemic delivery of
`exon-23-skipping antisense compounds restored
`up to 50% of dystrophin expression in various
`muscle groups.15•16 This treatment led to improved
`muscle force and reduced creatine kinase levels
`without tissue toxicity.15,16
`This recent preclinical progress has led to ini(cid:173)
`tiatives toward preliminary clinical studies in pa(cid:173)
`tients. Our previous studies in cultured cells from
`patients showed that an intraexonic 2OMePS anti(cid:173)
`sense oligonucleotide, PRO0Sl, efficiently induced
`specific exon-51 skipping.19 On the basis of the
`frequency of mutations in patients with Du(cid:173)
`chenne's muscular dystrophy in the Leiden data(cid:173)
`base,7 we concluded that PRO0Sl might correct
`the reading frame in 16% of all patients with the
`in other words, 25% of deletions, in(cid:173)
`disease -
`cluding exon 50 (Fig. 1), exon 52, exons 45 to 50,
`exons 48 to 50, and exons 49 to 50. Obviously,
`the actual therapeutic benefit would depend on
`the functionality of the resulting modified dys(cid:173)
`trophin.
`Our exploratory, open-label, single-center study
`involved four patients with Duchenne's muscular
`dystrophy, each of whom received a single injec(cid:173)
`tion of PRO0Sl into the tibialis anterior muscle.
`The primary outcome of this trial was adverse
`events in the four subjects; secondary outcomes
`were specific exon-51 skipping and dystrophin
`expression.
`
`METHODS
`
`PATIENTS AND STUDY DESIGN
`Patients with Duchenne's muscular dystrophy who
`were between the ages of 8 and 16 years were
`eligible to participate in the study. All patients
`had deletions that were correctable by exon-51
`skipping and had no evidence of dystrophin on
`previous diagnostic muscle biopsy. Concurrent
`
`2678
`
`N EN G L J MED ~7;26 WWW. NEJM . 0RG DECEMBER 2 7, 2007
`
`

`

`LOCAL DYSTROPHIN RESTORATION IN PATIENTS WITH DMD
`
`glucocorticoid treatment was allowed. Written in(cid:173)
`formed consent was obtained from the patients
`or their parents, as appropriate. During the pre(cid:173)
`screening period (up to 60 days), each patient's
`mutational status and positive exon-skipping re(cid:173)
`sponse to PRO0Sl in vitro were confirmed, and
`the condition of the tibialis anterior muscle was
`determined by T1-weighted magnetic resonance
`imaging (MRI).21 For patients to be included in
`the study, fibrotic and adipose tissue could make
`up no more than 50% of their target muscle.
`During the baseline visit, safety measures were
`assessed. In eacli patient, the leg that was to be
`injected was fixed with a tailor-made plastic mold
`and its position was recorded. A topical eutectic
`mixture of local anesthetics (EMLA) was used to
`numb the skin. Four injections of PRO0Sl were
`given along a line measuring 1.5 cm running
`between two small skin tattoos with the use of a
`2.5-cm electromyographic needle (MyoJect Dis(cid:173)
`posable Hypodermic Needle Electrode, TECA Ac(cid:173)
`cessories) to ensure intramuscular delivery. The
`volume of each injection was 200 µl containing
`200 µg of PRO0Sl, which was dispersed in equal
`portions at angles of approximately 30 degrees.
`At day 28, safety measures were assessed again.
`The leg that had been injected was positioned
`with the use of the patient's own mold, and a
`semiopen muscle biopsy was performed between
`the tattoos under local anes.thesia with a forceps
`with two sharp-edged jaws (Blakesley Concho(cid:173)
`toma, DK Instruments).22 The biopsy specimen
`was snap-frozen in 2-methylbutane cooled in
`liquid nitrogen.
`Patients were treated sequentially from May
`2006 through March 2007 and in compliance
`with Good Clinical Practice guidelines and the
`provisions of the Declaration of Helsinki. The
`study was approved by the Dutch Central Com(cid:173)
`mittee on Research Involving Human Subjects
`and by the local institutional review board at
`Leiden University Medical Center. All authors
`contributed to the study design, participated in
`the collection and analysis of the data, had com(cid:173)
`plete and free access to the data, jointly wrote
`the manuscript, and vouch for the completeness
`and accuracy of the data and analyses presented.
`
`A
`
`B
`
`DMD deletion at exon 50
`
`•
`. . . •
`
`No dystrophin
`
`Pre-mRNA
`
`Out-of-frame mRNA
`
`PRO051
`
`I iron 49 5.0
`
`Pre-mRNA
`
`Splicing
`
`In-frame mRNA
`
`BMD-like dystrophin
`
`Figure 1. Schematic Representation of Exon Skipping.
`In a patient with Duchenne's muscular dystrophy who has a deletion of
`exon SO, an out-of-frame transcript is generated in which exon 49 is spliced
`to exon 51 (Panel A). As a result, a stop codon is generated in exon 51, which
`prematurely aborts dystrophin synthesis. The sequence-specific binding of
`the exon-internal antisense oligonucleotide PRO0Sl interferes with the cor(cid:173)
`rect inclusion of exon 51 during splicing so that the exon is actually skipped
`(Panel B). This restores the open reading frame of the transcript and allows
`the synthesis of a dystrophin similar to that in patients with Becker's muscu(cid:173)
`lar dystrophy (BMD).
`
`ribose molecules and phosphorothioate internu(cid:173)
`cleotide linkages. The drug was provided by Pro(cid:173)
`sensa B.V. in vials of 1 mg of freeze-dried mate(cid:173)
`rial with no excipient. It was dissolved and
`administered in sterile, unpreserved saline (0.9%
`sodium chloride). PRO0Sl was not found to be
`mutagenic by bacterial Ames testing. In regula(cid:173)
`tory Good Laboratory Practice safety studies, rats
`that received a single administration ofup to 8 mg
`per kilogram of body weight intramuscularly and
`SO mg per kilogram intravenously showed no ad(cid:173)
`verse effects; monkeys receiving PRO0Sl for
`1 month appeared to tolerate doses up to 16 mg
`per kilogram per week when the drug was admin(cid:173)
`istered by intravenous 1-hour infusion or by sub(cid:173)
`cutaneous injection, without clinically relevant ad(cid:173)
`verse effects.
`
`DESCRIPTION OF PRO0Sl
`PRO0Sl is a synthetic, modified RNA molecule
`with sequence 5'-UCAAGGAAGAUGGCAUUUCU-
`3'.23 It carries full-length 2'-0-methyl-substituted
`
`IN VITRO PRESCREENING
`A preexisting primary myoblast culture19 was used
`for the prescreening of Patient 4. For the other
`three patients, fibroblasts were converted into
`
`N ENGLJ MED 357;26 WWW.NEJM.ORG DECEMBER 27, 2007
`
`2679
`
`

`

`The NEW ENGLAND JOURNAL of MEDICINE
`
`myogenic cells after infection with an adenoviral
`vector containing the gene for the myogenic tran(cid:173)
`scription factor (MyoD) as desci;ibed previous(cid:173)
`ly.'9•24•25 Myotube cultures were transfected with
`PRO051 (100 nM) and polyethylenimine (2 µ.1 per
`microgram of PRO0Sl), according to the manu(cid:173)
`facturer's instructions for ExGenS00 (MBI Fermen(cid:173)
`tas). RNA was isolated after 48 hours. Reverse
`transcriptase-polyrnerase chain reaction (R:f-PCR),
`immunofluorescence, and Western blot analyses
`were performed as reported previously.19•23 PCR
`fragments were analyzed with the use of the 2100
`Bioanalyzer (Agilent) and isolated for sequencing
`by the Leiden Genome Technology Center.
`
`SAFETY ASSESSMENT
`At baseline and at 2 hours, 1 day, and 28 days
`after injection, all patients received a full physical
`examination (including the measurement of vital
`signs) and underwent electrocardiography. In ad(cid:173)
`dition, plasma and urine were obtained to deter(cid:173)
`mine renal and liver function, electrolyte levels,
`complete cell counts, the activated partial-throm(cid:173)
`boplastin time, and complement activity values in
`the classical (CHS0) and alternative (APS0) routes.
`The use of concomitant medications was record(cid:173)
`ed. At baseline and on day 28, the strength of the
`tibialis anterior muscle was assessed with the use
`of the Medical Research Council scale26 to evalu(cid:173)
`ate whether the procedures had affected muscle
`performance. (On this scale, a score of O indicates
`no movement and a score of S indicates normal
`muscle strength.) Since only a small area of the
`muscle was treated, clinical benefit in terms of
`increased muscle strength was not expected. At
`each visit, adverse events were recorded.
`
`RNA ASSESSMENT
`Serial sections (SO µ.m) of the frozen muscle(cid:173)
`biopsy specimen were homogenized in RNA-Bee
`solution (Campro Scientific) and MagNA Lyser
`Green Beads (Roche Diagnostics). Total RNA was
`isolated and purified according to the manufac(cid:173)
`turer's instructions. For complementary DNA,
`synthesis was accomplished with Transcriptor re(cid:173)
`verse transcriptase (Roche Diagnostics) with the
`use of 500 ng of RNA in a 20-µ.I reaction at 55°C
`for 30 minutes with human exon 51 or 54 specific
`reverse primers. PCR analyses were performed as
`described previously.19•23 Products were analyzed
`on 2% agarose gels and sequenced. In addition,
`RT-PCR with the use of a primer set for the pro-
`
`Figure 2 (facing page). Prescreening Studies of the Four
`Patients.
`Magnetic resonance images of the lower legs of the
`/ four patients (tlie left leg-of Patlent3 and right legs of
`the other three patients) show the adequate condit ion
`of the tibialis anterior muscle (less than 50% fat in·
`filtration and fibrosis) (Panel A). The diagnosis of
`Duchenne's muscular dystrophy° in these patients was
`confirmed by diaminobenzidl ne tetrahydrochloride
`staining of cross sections of biopsy spedmens obtained
`previously from the quadriceps muscle (Panel 8). No
`dystrophin expression was observed, with the excep(cid:173)
`tion of one dystrophln-pos1tive, or revertant, fiber in
`Patient 2 (arrow). Reverse-transcriptase-polymerase(cid:173)
`chain-reaction (RT-PCR) analysis of the transcript re(cid:173)
`gion flanking the patients' mutations and exon 51 con•
`firmed both the individual mutations in nontreated
`myotubes (NT) and the positive response-to PRO051
`(i.e., exon 51 skipping) in treated myotubes (T) on the
`RNA level (Panel'C). The efficiencies of exon skipping
`were 49%for Patient 1, 84% fo r Patient 2, 58% for Pa(cid:173)
`tient 3, and 90% for Patient 4. A cryptic splice site
`within exon 51 is sometimes activated by PRO051 in
`cell culture, resulting in an extra aberrant splicing
`product, as seen in the treated sample from Patient 4.
`Lane M shows a 100-bp size marker, and lane C RNA
`from healthy control muscle. Sequence analysis of the
`RT-PCR fragi:nents from treated and untreated myo(cid:173)
`tubes identified the precise skipping of exon 51 for
`each patient (Panel D). The new in-frame transcripts
`led to substantial dystrophin synthesis, as detected by
`immunofluorescence analysis of treated myotubes with
`the use of monoclonal antibody NCL-DYS2 (Panel E).
`No dystrophin was detected before treatment.
`
`tein-truncation test27 was used to rapidly screen
`for aspecific aberrant splicing events throughout
`the DMD gene.
`
`ASSESSMENT OF PROTEIN LEVEL
`For immunofluorescence analysis, acetone-fixed
`sections were incubated for 90 minutes with
`monoclonal antibodies against the central rod
`domain (MANDYS106, Dr. G. Morris, United King(cid:173)
`dom) at a dilution ofl:60, the C-terminal domain
`(NCL-DYS2, Novocastra Laboratories) at a dilution
`ofl:30, or (as a reference) laminin a2 (Chemicon
`International), a basal lamina protein that is un(cid:173)
`affected by dystrophin deficiency, at a dilution of
`1:150, followed by Alexa Fluor 488 goat anti(cid:173)
`mouse IgG (H+L) antibody (Molecular Probes) at
`a dilution of 1:250 for 1 hour. Sections were
`mounted with Vectashield Mounting Medium (Vec(cid:173)
`tor Laboratories). ImageJ software (W. Rasband,
`National Institutes of Health, http://rsb.info.nih.
`gov/ij) was used for quantitative image analysis
`
`2680
`
`N ENGLJ MED 357;26 WWW.NEJM.ORG DECEMBER 27, 2OD]
`
`

`

`A
`
`LOCAL DYSTROPHIN RESTORATION IN PATIENTS WITH DMD
`
`B
`
`Patient 1
`
`Patient 2
`
`~ I]
`ra,~
`~ ~
`
`Patient 3
`
`Patient 4
`
`Patient 1
`
`Patient 2
`\ "-
`' -.;
`~ ~~
`
`~ )
`
`j•
`
`Patient 3
`
`Patient 4
`
`C
`Patient 1, '150
`M C
`
`600 bp-
`
`NT T
`
`D
`Patient 1, '150
`Exon 49 Exon 51
`a CC'AO -ro ... .--op toe T,'IC' TCA
`
`E
`Patient 1, '150
`NT
`
`-f49 1s11.s2153 1
`
`- !49152)53!
`
`Exon 49 Exon 52
`
`T
`
`T
`
`T
`
`T
`
`Patient 2, '148-50
`M C NT T
`
`600bp _ -~
`-
`----
`
`!:: - -
`-·
`
`- 147151 152)53!
`
`-!47 152 1531
`
`84%
`
`Patient 3, '149-50
`
`M C
`
`600 bp _
`
`NT T :(cid:173)- ---58%
`
`- !48l511s21531
`
`- 148152 1s3 1
`
`Patient 4, '152
`
`M
`
`NT
`
`600 bp_
`
`-
`
`~- - -
`~ ... =
`
`T
`
`~--90%
`
`- I SOI 51 153 I 54 I
`- 1501s3 154 I
`
`Patient 2, '148-50
`Exon 47 Exon 51
`
`Patient 2, '148-50
`NT
`
`Exon 47 Exon 52
`
`OT OO At.u..AO@ e MOM ;i"O
`
`~
`
`Patient 3, '149-50
`Exon 48 Exon 51
`
`Patient 3, '149-50
`NT
`
`Exon 48 Exon 52
`
`Patient 4, '152
`Exon 51 Exon 53
`
`Patient 4, '152
`
`NT
`
`Exon SO Exon 53
`
`N ENG LJ M ED 357;26 WWW.N EJ M .0RG D ECE M BER 2 7 , 2 007
`
`2681
`
`

`

`The NEW ENGLAND JOURNAL of MEDICINE
`
`29 Entire cross sections
`as described previously.28
`•
`were subdivided into series of 6 to 10 adjacent
`images, depending on the size of the section. To
`ensure reliable measurements, staining of the
`sections and recording of all images were per(cid:173)
`formed during one session with the use of fixed
`exposure settings and the avoidance of pixel sat(cid:173)
`uration. The lower-intensity threshold was set at
`background for Duchenne's muscular dystrophy,
`and positive fluorescence was quantified for each
`section (area percentage), both for dystrophin and
`laminin a2.
`Western blot analysis was performed as de(cid:173)
`scribed previously19 with the use of pooled homo(cid:173)
`genates from sets of four serial 50-µ,m sections
`throughout the biopsy specimen. For each pa(cid:173)
`tient, two amounts of total protein - 30 µ,g and
`60 µ,g - were applied, and for the control sam(cid:173)
`ple, 3 µ,g. The Western blot was incubated over(cid:173)
`night with dystrophin monoclonal antibody
`NCL-DYSl (Novocastra Laboratories) at a dilution
`of 1:125, followed by horseradish-peroxidase(cid:173)
`labeled goat antimouse IgG (Santa Cruz Biotech(cid:173)
`nology) at a dilution of 1:10,000 for 1 hour. Im(cid:173)
`munoreactive bands were visualized with the use
`of the ECL Plus Western blotting detection sys(cid:173)
`tem (GE Healthcare) and Hyperfilm ECL (Amer(cid:173)
`sham Biosciences). Signal intensities were mea(cid:173)
`sured with the use of ImageJ software.
`
`RESULTS
`
`PRESCREENING OF PATIENTS
`The study was planned to include four to six pa(cid:173)
`tients. Six patients were invited to participate, and
`one declined. The remaining five patients were
`
`prescreened. First, the condition of the tibialis
`anterior muscle was evaluated on MRI. The mus(cid:173)
`cle condition of four patients was deemed to be
`adequate for the study (Fig. 2A), an_d the absence
`of dystrophin was confirmed in the patients'
`original biopsy specimens (Fig. 2B). Second, the
`mutational status and positive exon-skipping re(cid:173)
`sponse to PRO051 of these four patients were
`confirmed in fibroblast cultures. PRO051 treat(cid:173)
`ment generated a novel, shorter fragment of mes(cid:173)
`senger RNA for each patient, representing 46%
`(in Patient 4) to 90% (in Patient 1) of the total
`RT-PCR product (Fig. 2C). Precise exon-51 skip(cid:173)
`ping was confirmed by sequencing (Fig. 2D). No
`other transcript regions were found to be ~ltered.
`Immunofluorescence analyses showed a prepon(cid:173)
`derance of dystrophin-positive myotubes (Fig. 2E),
`a finding that was confirmed by Western blot
`analysis (not shown). Thus, the four patients were
`judged to be eligible for PRO051 treatment. Their
`baseline characteristics are shown in Table 1.
`
`SAFETY AND ADVERSE EVENTS
`All patients had one or more adverse events. How(cid:173)
`ever, only one patient reported mild local pain at
`the injection site, which was considered to be an
`adverse event related to the study drug. Other
`events included mild-to-moderate pain after the
`muscle biopsy. Two patients had blistering under
`the bandages used for wound closure. In the pe(cid:173)
`riod between injection and biopsy, two patients
`reported a few days of flulike symptoms, and one
`patient had mild diarrhea for 1 day. At baseline,
`the muscle-strength scores of the treated tibialis
`anterior muscle in Patients 1, 2, 3, and 4 were 4,
`2, 3, and 4, respectively, on the Medical Research
`
`Table 1. Baseline Characteristics of the Patients.
`
`Variable
`
`Age (yr)
`
`Deletion
`
`Age at loss of ambulation (yr)
`
`Previous glucocorticoid treatment
`
`Previous diagnostic biopsy
`
`Age at ti me of biopsy (yr)
`
`Patient 1
`
`10
`
`Exon 50
`
`9
`
`Yes
`
`4
`
`Patient 2
`
`13
`
`Patient 3
`
`13
`
`Exons 48-50
`
`Exo ns 49-50
`
`11
`
`Yes
`
`8
`
`7
`
`No
`
`4
`
`Patient 4
`
`11
`
`Exon 52
`
`10
`
`Stopped 8 months
`before study
`
`4
`
`Type of muscle specimen
`
`Quadriceps
`
`Dystrophin-positive fibers in
`biopsy specimen (%)
`
`0
`
`Quadriceps
`<l
`
`Quadriceps
`
`Q uadriceps
`
`0
`
`0
`
`2682
`
`N EN GL J MEO 357 ;26 W W W. N EJM .O RG DECEMB ER 2 7 , 2007
`
`

`

`LOCAL DYSTRO P HIN RESTORATION IN PATIENT S W I TH D M D
`
`Council scale. None of the patients showed chang(cid:173)
`es in the strength of this muscle during the study
`or significant alterations in standard laboratory
`measures or increased measures of complement
`split products or activated partial-thromboplas(cid:173)
`tin time. No local inflammatory or toxic response
`was detected in the muscle sections of the pa(cid:173)
`tients (data not shown). Patient 3 successfully
`underwent preplanned surgery for scoliosis in the
`month after the study was completed.
`
`RNA AND PROTEIN LEVEL
`At day 28, a biopsy of the treated area was per(cid:173)
`formed in each patient. Total muscle RNA was
`isolated from serial sections throughout the bi(cid:173)
`opsy specimen. In all patients, RT-PCR identified
`a novel, shorter fragment caused by exon-51 skip(cid:173)
`ping, as confirmed by sequencing (Fig. 3). Fur(cid:173)
`ther transcript analysis showed no other alter(cid:173)
`ations (data not shown). Immunofluorescence
`analyses of sections throughout the biopsy speci-
`
`men of each patient showed clear sarcolemmal
`dystrophin signals in the majority of muscle fibers
`(Fig. 4A and 4B). Dystrophin antibodies proximal
`and distal to the deletions that were used includ(cid:173)
`ed MANDYS106 (Fig. 4A and 4B) and NCL-DYS2
`(similar to MANDYS106, not shown). The fibers
`in each section were manually counted after stain(cid:173)
`ing for laminin a2.30 The individual numbers var(cid:173)
`ied, consistent with the size of the biopsy speci(cid:173)
`men and the quality of the muscle. In the largest
`sections, Patient 2 had 726 fibers, of which 620
`were dystrophin-positive, whereas Patient 3 had
`120 fibers, of which 117 were dystrophin-positive
`(Fig. 4A and 4C). The dystrophin intensities were
`typically lower than those in a healthy muscle
`biopsy specimen (Fig. 4B). The single fibers with
`a more intense dystrophin signal in Patients 2
`and 3 could well be revertant fibers (Fig. 4B).
`Western blot analysis confirmed the presence
`of dystrophin in varying amounts (Fig. 4E). The
`dystrophin signals were scanned and correlated
`
`M
`
`C
`
`Patient 1
`
`M
`
`C
`
`Patient 2
`
`600 bp
`
`600 bp_
`
`-l411s1 Js21
`
`!49!52!
`
`M
`
`C
`
`Patient 3
`
`M
`
`C Patient 4
`
`600 bp
`
`600 bp_
`
`- l4sl s1 I s21
`
`!SO!S1)53!
`
`Figure 3. RT-PCR Analysis of RNA Isolated from Serial Sections of Biopsy Specimens from the Patients.
`After treatment with PRO051, reverse-transcriptase- polymerase-chain-reaction (RT-PCR) analysis shows novel,
`shorter transcript fragments for each patient. Both the size and sequence of these fragments confirm the precise
`skipping of exon 51. No additional splice va riants were observed. At 28 days, still significant in-frame RNA transcripts
`were detected, suggesting prolonged persistence of PRO051 in muscle. Owing to the small amount of section mate(cid:173)
`rial, high-sensitivity PCR conditions were used; this process precluded the accurate qua ntificatio n of skipping effi(cid:173)
`ciencies and the meaningful correlation between levels of RNA and protein. M denotes size marker, and C control.
`
`N ENGL) M ED 357 ;26 WWW. N EJM.O RG DECEMBER 2], 2007
`
`268 3
`
`

`

`The NEW ENGLAND JOURNAL of MEDICINE
`
`to the control (per microgram of total protein).
`The amounts varied from 3% in Patient 3, who
`had the most-dystrophic muscle, to 12% in Pa(cid:173)
`tient 2, who had the best-preserved muscle. Since
`such comparison on the basis of total protein
`does not correct for the varying amounts of fi(cid:173)
`brotic and adipose tissue in patients with Duch(cid:173)
`enne's muscular dystrophy, we also quantified
`the dystrophin fluorescence signal (Fig. 4A and
`4B) relative to that of the similarly located lam(cid:173)
`inin a2 in each section by ImageJ analysis. When
`the ratio of dystrophin to laminin a2 was set at
`100 for the control section, Patients 1, 2, 3, and
`4 had ratios of 33, 35, 17, and 25, respectively
`(Fig. 4D).
`
`DISCUSSION
`
`Our study showed that local intramuscular injec(cid:173)
`tion of PRO0Sl, a 2OMePS antisense oligoribo(cid:173)
`nucleotide complementary to a 20-nucleotide se(cid:173)
`quence within exon 51, induced exon-51 skipping,
`corrected the reading frame, and thus introduced
`dystrophin in the muscle in all four patients with
`Duchenne's muscular dystrophy who received
`therapy. Dystrophin-positive fibers were found
`throughout the patients' biopsy specimens, indi(cid:173)
`cating dispersion of the compound in the inject(cid:173)
`ed area. Since no delivery-enhancing excipient
`was used, PRO0Sl uptake did not seem to be a
`major potentially limiting factor. We cannot rule
`out that increased permeability of the dystrophic
`fiber membrane had a favorable effect. The pa(cid:173)
`tients produced levels of dystrophin that were 3 to
`12% of the level in healthy control muscle, as
`shown on Western blot analysis of total protein.
`Since the presence of fibrosis and fat may lead to
`some underestimation of dystrophin in total pro(cid:173)
`tein extracts, we determined the ratio of dystro(cid:173)
`phin to laminin al in the cross sections, which
`ranged from 17 to 35, as compared with 100 in
`control muscle. The dystrophin-restoring effect of
`PRO051 was limited to the treated area, and no
`strength improvement of the entire muscle was
`observed. Future systemic treatment will require
`repeated administration to increase and maintain
`dystrophin expression at a higher level and to ob(cid:173)
`tain clinical efficacy.
`Because of medical-ethics regulations regard(cid:173)
`ing interventions in minors, we could not obtain
`a biopsy specimen from the patients' contralat-
`
`Figure 4 (facing page). Dystrophin•Restoring Effect
`ofa Single Intramuscular Dose of PRO0Sl.
`lmmunofluorescence analysis with the use of the dys·
`trophin antibody MANDYS106 clearly shows dystrophin
`expression at the membranes of the majority of fibers
`throughout the biopsy specimen obtained from each
`patient (Panel A). The areas indicated by the squares
`are shown in higher magnification in Panel B. For
`comparison, a sample from an untreated patient with
`Duchenne's muscular dystrophy (DMD) and a healthy
`control sample from gastrocnemius muscle (HC) are
`included with the samples from the patients. Putative
`revertant fibers are indicated by arrows. The total num(cid:173)
`ber of muscle fibers that contained dystrophin and
`laminin a2 were counted manually and the ratios of
`dystrophin to laminin a2 were plotted (Panel C). West(cid:173)
`ern blot analysis of total protein extracts isolated from
`the patients' biopsy specimens with the use ofNCL-DYSl
`antibody show restored dystrophin expression in all
`patients (Panel E). For each patient, 30 µg (right lane)
`and 60 µg (left lane) were loaded; for comparison, 3 µg
`of total protein from a healthy gastrocnemius muscle
`sample was also loaded (to avoid overexposure). Be(cid:173)
`cause of the relatively smal l deletions in the DM