Impact of Transforming Growth Factor-β1 on Atrioventricular Node Conduction Modification by Injected Autologous Fibroblasts in the Canine Heart
Background— Atrioventricular (AV) nodal ablation for management of atrial fibrillation (AF) is irreversible and requires permanent pacemaker implantation. We hypothesized that as an alternative, implantation of autologous fibroblasts in the perinodal region would focally modify AV nodal conduction and that this modulation would be enhanced by pretreatment with transforming growth factor-β1 (TGF-β1), a stimulant of fibroblasts.
Methods and Results— Skin biopsies were taken from 12 mongrel dogs, and derived fibroblasts were dissociated and grown in culture for 2 weeks. Multiple injections (0.25 mL) were made through an 8F NOGA catheter along the fast/slow AV nodal pathways as guided by an electroanatomic mapping system. Seven dogs received fibroblasts alone (1×106 cells/mL), 7 dogs received TGF-β1 (5 μg), 4 dogs received fibroblasts and TGF-β1 (1×106 cells/mL+5 μg), and 4 dogs received saline only. AV node function was assessed at baseline and after 4 weeks. Saline (80 mL) with assigned therapy (0.25 mL per injection) was injected into the peri-AV nodal region in each dog. At baseline, the AH interval (66±3 ms) and the average RR interval (331±17 ms) in pacing-induced AF were similar in each cohort. The increase in AH interval in normal sinus rhythm was longer after fibroblast (23±4 versus 5±5 ms; P=0.05) and fibroblast plus TGF-β1 (50±5 versus 5±5 ms; P<0.001) injections than with saline alone, with similar findings during high right atrium and distal coronary sinus pacing. The AH interval was not significantly increased after TGF-β1 injections. The AH interval was significantly longer after fibroblast plus TGF-β1 injections than with either therapy (TGF-β1 or fibroblasts) alone. The RR interval during AF was increased in dogs that received fibroblasts alone (110±36 versus −41±34 ms) and to a greater extent with the addition of TGF-β1 (294±108 versus −41±34 ms). No AV block was seen in any cohort at 4 weeks. Labeled fibroblasts that expressed vimentin were identified in all dogs that received cell injections at 4 weeks.
Conclusions— AV nodal modification can be achieved with injected fibroblasts without the creation of AV block. The effect on AV node conduction is substantially enhanced by pretreatment of fibroblasts with TGF-β1. These data have therapeutic potential for the management of rapid ventricular rate during AF without pacemaker implantation.
Received June 22, 2005; revision received February 17, 2006; accepted March 9, 2006.
Recently, much effort has been directed toward the development of nonpharmacological treatments for atrial fibrillation (AF). Many different strategies have been used either to restore sinus rhythm or to optimize rate control to minimize the morbidity of the disease.1 In many patients, rate control with anticoagulation is an acceptable therapy for AF, with long-term outcomes similar to antiarrhythmic therapy.2–5 Nevertheless, pharmacological therapies often are poorly tolerated. Although the side-effect profiles of rate control medications are less than with rhythm control agents, efficacy with these medications over time may be inadequate.5,6
Editorial p 2474
Clinical Perspective p 2494
In patients who are intolerant of medical management for AF, radiofrequency ablation of the atrioventricular (AV) node is a validated long-term rate control option.7 This approach electrically disconnects the atrium and ventricle but results in long-term pacemaker dependence. Although the outcomes after pacemaker implantation are largely favorable, the results are permanent.7,8 Furthermore, right ventricular pacing may have a detrimental impact on left ventricular function.9,10 Partial alteration of AV nodal conduction with radiofrequency ablation to avoid pacemaker implantation has resulted in improvement in ventricular rate control, although high-grade AV block is a frequent complication.11–17
These limitations provide the incentive for the development of new treatment strategies. One such approach would be to use cell therapy to interrupt or to modify conduction into the AV node. The deposition of electrically inactive collagen and resulting fibrosis might serve as a barrier to conduction as mediated in part by fibroblasts. We therefore sought to determine whether cultured autologous fibroblast cells embedded in the peri-AV nodal area would focally modify AV node conduction. Second, we hypothesized that pretreatment of these fibroblasts with transforming growth factor-β1 (TGF-β1), a known stimulant of fibroblasts, would enhance the physiological effect of the fibroblasts on AV nodal conduction in both sinus rhythm and AF.18,19
The experimental protocol was approved by the Mayo Foundation Institutional Animal Care and Use Committee. The authors had full access to the data and take full responsibility for their integrity. All authors have read and agree to the manuscript as written.
General Study Design
Twenty-two male mongrel dogs (weight, 30 to 40 kg; age, 1 to 3 years) were used for this study. These dogs received 1 of 4 therapies in an identical volume of saline: (1) autologous fibroblasts alone, (2) autologous fibroblasts with TGF-β1, (3) TGF-β1 alone, or (4) normal saline. The therapy assignments were not random because of staff and equipment availability to perform cell culture. AV nodal function in sinus rhythm and pacing-induced AF were measured in each dog before the therapy was given, immediately afterward, and 4 weeks after recovery.
Animal Monitoring/Surgical Care
The dogs were anesthetized with intravenous ketamine (10 mg/kg) and diazepam (0.5 mg/kg), intubated, and ventilated with a pressure-cycled ventilator. Each animal was monitored continuously with 8 surface ECG leads, along with blood pressure monitoring through a Teflon sheath in a femoral artery. Vascular access was obtained as previously described.20
Each dog assigned to receive a cell-based therapy underwent skin biopsy. A skin specimen ≈2×2 cm was removed from the right shoulder and immediately placed in “culture media” containing Dulbecco’s low-glucose modified Eagle’s medium (GIBCO BRL) with 20% FBS (GIBCO BRL), 1% amphotericin (GIBCO BRL), and 1% penicillin, streptomycin, and glutamine (GIBCO BRL). The epidermis and underlying connective and fatty tissue were dissected from the specimen, and the tissue was sliced with a sterile scalpel to the consistency of a paste and mixed in 10 mL of 0.25% trypsin at 37°C for 15 minutes. The fluid was then withdrawn with a sterile pipette and placed in culture media containing bovine serum to stop digestion. This process was repeated after adding 10 mL of 0.25% trypsin to the tissue specimen until the solution no longer became cloudy after 15 minutes of exposure. The dissociated cells were placed in culture media and plated for culture. The culture media was replaced after 2 hours and every other day thereafter.
Cell Identification Labeling
The plated cells were assessed daily for growth. Fibroblasts reaching a population of 200×106 were resuspended in culture media and labeled with the fluorescent marker CM-DiI (Molecular Probes, Eugene, Ore) for identification 4 weeks after implantation. A concentration of 1 mg/mL CM-DiI was used because it was the minimum dose that resulted in label uptake of each cell as seen after 1 week of growth in culture. After exposure to CM-DiI, the cells were washed in serum-free Dulbecco’s low-glucose modified Eagle’s medium 3 times and resuspended in normal saline at a concentration of 1×106 cells/mL. This concentration was chosen on the basis of prior growth rate studies showing that 8 mL of 1×106 cells/mL (cell culture media) injected though the delivery catheter directly into a cell culture dish produced cell growth confluence at 1 week.
Clinical Animal Studies
Intracardiac Ultrasound Imaging
A 10F, 5.5- to 10.0-MHz ultrasound catheter with multidirection tip deflectability was positioned in the right atrium (RA) and interfaced to the input stage of an Acuson Sequoia imaging platform for complete intracardiac imaging. Intracardiac ultrasound imaging (intracardiac echocardiography [ICE]) was used to guide the injectable catheter placement. Color flow and spectral Doppler imaging was used to assess cardiac function and valve integrity before and after injections.
AV Node Assessment
AV nodal function, as reflected by the AH interval, was assessed in sinus rhythm and during pacing from the high RA and distal coronary sinus (CS) at a 400-ms cycle length. Six consecutive AH intervals were recorded and averaged. The HV interval also was measured in both sinus rhythm and during pacing. Anterograde decremental pacing was performed from the high RA, beginning at a cycle length 20 ms shorter than the sinus rate down to 200 ms, and the cycle lengths at which Wenckebach and 2:1 AV node block occurred were established. Finally, AF was induced by burst pacing in the high RA, and 15 consecutive RR intervals were measured and averaged.
The catheter/guidance mapping system consisted of an external magnetic field, a deflectable mapping catheter, and a computer processing unit (Carto, Biosense Webster Inc, Diamond Bar, Calif). Three-dimensional electroanatomic maps of the RA, superior vena cava, and inferior vena cava were created during sinus rhythm. The positions of the CS ostium and His bundle were annotated on the surface geometries. The position of the tricuspid annulus was established with 3 anatomic points obtained with electroanatomic mapping. Injection points were labeled with 3-mm lesion markers in a color distinct from the anatomic labels (Figure 1).
Therapy Assignment and Preparation
The animals were injected with 1 of 4 therapies: fibroblasts alone (n=7), TGF-β1 alone (n=7), fibroblasts with TGF-β1 (n=4), or normal saline (n=4). Initially, the fibroblasts were suspended in 80 mL normal saline to a concentration of 1.0×106/mL. Then, 5 μg TGF-β1 (R&D Systems, Minneapolis, Minn) was mixed in 80 mL normal saline. For combination therapy, fibroblasts were suspended at a concentration of 1.0×106/mL of normal saline and 5 μg of TGF-β1 added to the 80-mL solution. Finally, control animals were injected with 80 mL normal saline alone.
All therapies were injected in the peri-AV nodal region, specifically along the expected sites of the slow and fast pathways, using a specially designed injection catheter developed by Biosense-Webster (NOGA catheter, a mapping catheter that contains a 27-gauge retractable needle) advanced from the right femoral vein. The needle depth was set at 2 to 3 mm. The catheter was primed with the assigned therapy solution before it was advanced into the atrium under ICE and biplane fluoroscopic guidance. The catheter was withdrawn periodically during the procedure to reassess the needle depth. The catheter was initially placed above the tendon of Todaro in the region of the expected fast pathway to begin injections; then, it was drawn back slowly to a position superior and ultimately posterior to the His lesion marker. Injections were made along the tricuspid annulus in the region of the expected slow pathway, back to the anterior rim of the CS orifice (Figure 1). At each point, the endocardial electrogram recorded by the NOGA catheter was examined, and injections were withheld if a His electrogram was present. Then, 0.25 mL solution was delivered at each injection site with an indeflator. The procedure was repeated until 80 mL solution (≈320 injections) was injected into the peri-AV nodal area and along the expected slow and fast pathways. Further studies are required to determine the minimal number of cells required to affect AV nodal function. In this study a large number of cells was used to determine the feasibility of the approach.
Follow-Up Assessment of the AV Node
The AV node was reassessed immediately after the injections were complete. With ICE, the region of the AV node, tricuspid valve, and interatrial septum were examined for acute injury. The ICE catheter was advanced just below the tricuspid annulus to assess ventricular function and to screen for an effusion.
All animals were restudied 4 weeks after the peri-AV nodal injections. Vascular access was obtained through the left-sided femoral vessels. AV nodal function was then reassessed using pacing techniques similar to those used at baseline. Through the use of ICE, the low medial RA and regions of the AV node, tricuspid valve, and intra-atrial septum were examined for indication of more chronic injury.
Histological Characterization of Lesions
Thereafter, the animals were placed under deep anesthesia and exsanguinated. The epicardial and endocardial surfaces along the sites of injection in the atrium were examined. The treated atrium was removed and immediately frozen with liquid nitrogen to facilitate identification of implanted fibroblasts. The tissue was sectioned finely in 5-μm slices to minimize potential CM-DiI marker bleed-through. A coverslip was mounted over the tissue with 1 drop of Vectorshield containing DAPI (Vector Labs, Burlingame, Calif). CM-DiI signals were enumerated with a Zeiss Axioplan microscope equipped with a triple-pass filter (Vysis). Additional sections were prepared in a similar manner; then, immunofluorescence staining was performed with antibodies against α-smooth muscle actin (Sigma, St. Louis, Mo; 1:800) and vimentin (Chemicon International, Temecula, Calif; 1:200). Additional sections of the perinodal atrium injection sites also were fixed in formalin and embedded in paraffin, and 4-μm sections were stained with hematoxylin and eosin and Masson’s trichrome.
The averages of continuous variables are expressed as mean±SE. A linear mixed model was used to compare the effect of study therapy across the 3 study time points on each AV nodal electrical function parameter. Subject characteristics were defined by animal and treatment time point, with repeated measures defined for the multiple readings at each time point under a heterogeneous first-order autoregressive covariance design. A no-intercept full factorial model was evaluated for the fixed effects of therapy and time point. Post hoc Bonferroni comparisons were then performed to identify which therapy groups accounted for any significances. Further modeling used the linear mixed model constrained to a specific time point to evaluate differences between specific treatments at the individual study times. A value of P≤0.05 was considered significant.
Twenty-two dogs were studied. At baseline, the average AH interval was 66±3 ms in sinus rhythm, 72±3 ms during high RA pacing, and 67±3 ms during distal CS pacing in all dogs. The average RR interval in pacing-induced AF was 331±17 ms. There were no significant differences between the baseline AV node conduction characteristics across any of the 4 therapy cohorts. A total of 80 mL of solution with the assigned therapy (≈320 injections) was injected into the peri-AV nodal region in the RA (Figure 1) in each animal. There were no signs of perforation during therapy delivery. There were no immediate observable adverse reactions to the fibroblasts or combination therapy.
In the cohort that received TGF-β1, 2 dogs developed transient complete heart block without a ventricular escape and required temporary pacing. At the time of study completion, both animals were in sinus rhythm without AV block. Both animals were in sinus rhythm after 4 weeks. No atrial or AV nodal arrhythmias were seen at the time of injections or restudy.
AV Nodal Modification
Fibroblast Versus Normal Saline Injections
The baseline AH interval was similar in fibroblast-injected and control dogs. After 4 weeks, fibroblast-injected dogs had a greater increase in AH interval (23±4 versus 5±5 ms; P=0.050) during sinus rhythm (the Table) than in controls. The change in the AH interval also was statistically greater in fibroblast animals than in controls for high RA pacing (45±5 versus −5±6 ms; P=0.020) and tended to be greater but was not significant during distal CS pacing (27±4 versus 7±5 ms; P=0.11). The ventricular response to AF was slower in the fibroblast-injected group as noted by a prolongation of the RR interval compared with the control animals (110±36 versus −41±34 ms), but this did not reach statistical significance (P=1.0 after Bonferroni adjustment). The HV interval was similar in both groups in sinus rhythm and with distal CS and high RA pacing. The heart rate at which Wenckebach and 2:1 heart block occurred was increased for fibroblast dogs compared with the control dogs, but these changes were not significantly different, as seen in the Table.
TGF-β1 Versus Normal Saline Injections
The baseline AH interval was similar in the TGF-β1–injected and control dogs. At 4 weeks, TGF-β1–injected dogs had an increase in their AH (18±4 versus 5±5 ms) interval during sinus rhythm (the Table) compared with control dogs, but this was not significant (P=0.13). The change in the AH interval was not statistically different (P=1.0 after Bonferroni correction) in the TGF-β1 group during either high RA (1±4 versus −5±6 ms) or distal CS (14±3 versus 7±55 ms) pacing. The RR intervals in AF decreased in both the TGF-β1 group and the control cohort (−49±39 versus −41±34 ms; P=1.0) compared with the baseline measurements. The HV interval was similar between groups in sinus rhythm and with pacing. The heart rate at which Wenckebach and 2:1 heart block occurred was not significantly different in the TGF-β1 dogs compared with controls.
Fibroblast/TGF-β1 Versus Normal Saline Injections
In contrast, at 4 weeks, dogs injected with fibro-blasts+TGF-β1 had a greater increase in their AH (50±5 versus 5±5 ms; P<0.001) interval in sinus rhythm compared with control dogs and nearly double that seen with fibroblast injections alone. An example of AH interval prolongation with combination therapy is shown in Figure 2. The changes in these intervals also were statistically increased in dogs injected with fibroblasts+TGF-β1 during both high RA (50±6 versus −5±6 ms; P=0.040) and distal CS (51±5 versus 7±5 ms; P=0.001) pacing compared with controls (Figure 3). The increase in the AF RR interval appeared to be greater in the fibroblasts+TGF-β1 group compared with controls (294±108 versus −41±34 ms; P=0.023), but this was not significant after adjustment for multiple comparisons (P=0.068; Figure 4). An example of the RR interval prolongation during pacing-induced AF in a dog that received combination therapy is shown in Figure 4. The heart rate at which Wenckebach (60±31 versus −5±31 ms) and 2:1 heart block (63±32 versus −3±32 ms) occurred was increased nonsignificantly (P=1.0) for fibroblasts+TGF-β1 dogs compared with the control dogs, respectively.
In all animals, endocardial scar was present in the expected atrial regions targeted for injection (Figure 5). In dogs that received fibroblasts or fibroblasts+TGF-β1, discrete subendocardial areas of fibroblastic proliferation were seen at the injection sites (Figure 6A, 6B, 6D, and 6E). This type of response was seen in 1 control dog (25%), 1 of the dogs that received TGF-β1 only (14%), and all 11 dogs that received fibroblasts with and without TGF-β1 (100%). The other control animals had a more typical form of injury response similar to that seen at biopsy sites (Figure 6C). In dogs that received a combination of fibroblasts and TGF-β1, there also was a mononuclear infiltrate (Figure 6F).
Injected fibroblasts were identified by CM-DiI fluorescence in all animals receiving fibroblasts or fibro-blasts+TGF-β within the atrium and peri-AV nodal region (Figure 7A). No CM-DiI–labeled cells were found in control animals that received saline. The labeled cells were localized to the injection lines as shown by CM-DiI labeling around DAPI-labeled nuclei. Sections of the tricuspid valve, aortic root, and ventricular septum below the injection sites, examined grossly and histologically, revealed no injection injuries at 4 weeks. CM-DiI cells did not consistently express α-smooth muscle actin (Figure 7B) but did express vimentin (Figure 7C).
In the present study, fibroblasts could be strategically injected into the atrium through the use of a percutaneous approach. These cells, identified 4 weeks later, were localized within injection lines in the expected anatomic targets. Fibroblast injections, with or without TGF-β1, modified cardiac electrophysiological properties of the AV node without creation of high-grade block. Although TGF-β1 alone did not significantly affect AV node conduction, pretreatment of fibroblasts with TGF-β1 significantly decrease AV nodal conduction, suggesting that this growth factor works through the fibroblast cell line.
The use of fibroblasts as a cell-based therapy in the treatment of burns, skin diseases, and head and neck tumors is well established.21–23 In these cases, the cells are used to replace damaged or injured tissue or to occupy space after tumor resection. In general, fibroblasts engraft well after implantation in noncardiac tissue, with minimal toxic degradation or inflammatory reactions. Fibroblasts as a cell source are advantageous in that they are easily isolated and cultured even from small biopsy specimens and can be multiplied to large numbers in cell culture.24
Our results expand prior studies and show that these cells not only engraft in atrial tissue but also can be strategically implanted from a percutaneous approach. The injected fibroblasts remain within the injection lines, resulting in very focal effects without broad extension outside the injection site or migration to remote tissues. The chondroid matrix response seen in cell-injected dogs is not fully understood and has previously been reported using other therapeutic modalities in canine models.25,26 We observed this in 1 control dog, which most likely reflects the use of mongrel dogs with uncontrolled host variances in immune system composition and cytokine milieu. Nonetheless, the confined injury pattern is encouraging compared with ablative techniques that can result in broad injury, excessive scarring, and damage to remote tissues.20,27,28
These data demonstrate reproducible conduction slowing through the AV node after injections with fibroblasts and fibroblasts+TGF-β1 as manifested by AH prolongation and changes in the RR interval in AF. In this study, we extensively targeted both the anterior and posterior AV nodal inputs, creating focal scar within the conduction tissue. The mechanism behind RR interval prolongation is mostly likely complex, involving an alteration of the summation of multiple atrial signals through AV nodal inputs. However, the pathophysiology underlying the conduction slowing remains to be determined. Collagen and fibrosis within the myocardium from disease and ablation interrupt electrical conduction.29,30 Fibroblasts may also mechanically and electrically couple to myocytes, which can result in long conduction delays in vitro.31,32
We did not observe a significant effect on AV nodal conduction after TGF-β1 injections alone. We postulated that because TGF-β1 increases the percentage of secreted collagen and matrix metalloproteinase expression and stimulates extracellular matrix deposition and protein synthesis in vivo while inhibiting matrix metalloproteinase inhibitor expression, it should enhance the injury response of native fibroblasts.19,33,34 Nevertheless, this absence of AV nodal conduction modification was not seen after TGF-β1 injections alone and may stem from many potential possibilities, including limited exposure of the atrial tissue to TGF-β1 after injections, inefficient interaction of the human-derived growth factor and canine tissue, or an insufficient quantity delivered with each injection.
The impact of fibroblasts pretreated with TGF-β1 before injection on AV nodal conduction was greater than with fibroblasts alone. This observation suggests that the effect of TGF-β1 in this model is mediated through the fibroblast itself. The mechanism of this interaction in an in vivo setting is unclear. Fibroblasts that are exposed to TGF-β1 differentiate into myofibroblasts.35,36 Myofibroblasts exposed to TGF-β1 induce several proteins responsible for the development of stress fibers and connective tissue contraction that are part of scar formation.37 Our immunofluorescent staining with antibodies against vimentin suggests these cells are of a stromal or mesenchymal phenotype; however, the lack of α-smooth muscle actin does not refute or support differentiation into a myofibroblast. Fibroblasts exposed to TGF-β1 also stimulate the production of collagen type 1.18 Hence, TGF-β1 controls 2 fundamental processes in the development of granulation or scar tissue: the capacity of modifying tissue shape through stress fiber deposition and contraction and the formation of extracellular matrix.38 It is possible that by slightly altering the scar produced by fibroblasts through pretreatment with TGF-β1, along with targeted interruption of AV nodal inputs critical for conduction, these small changes result in much greater physiological effects, as we observed.
The number of autologous cells identified 4 weeks after injection was relatively low compared with the overall physiological effect on AV nodal conduction. This finding has been seen with other cell-based therapies and seems to be independent of the cell type used.39,40 Nonetheless, conduction through the AV node, a function of the summation of multiple right- and left-sided inputs, can be altered significantly by affecting these nodal inputs specifically by these few cells, suggesting that a little goes a long way in affecting the underlying physiology. In addition, noncellular processes may contribute significantly to the overall effect such as the release of TGF-β1 or other growth factors/cytokines as part of the injury response.41,42
The data and techniques provided here are in contrast to other methods attempted to focally modify the AV node that have resulted in unacceptably high rates of high-grade block.11–17 These studies used radiofrequency ablation techniques. Despite focal delivery of energy, untoward injury to distant and collateral tissues and organs remains a common problem.20,27,28 Fibroblasts are an attractive alternative in that they are autologous and thus should not create extensive local inflammation, they can be delivered with minimal injury, and they do not appear to result in progressive or distant injury after implantation. These characteristics in a therapy are favorable when approaching structures such as the AV node where uncontrolled injury may result in unpredictable outcomes.
The present study should be considered in the context of several limitations. First, these data derived from an animal model may not be directly applicable in humans. However, the dog is a well-established model in electrophysiology. In addition, these animals were large, allowing the use of techniques that could be directly used in humans. Second, animals were euthanized at 4 weeks, and the injury response may continue and produce further fibrosis and injury. The animals were euthanized at 4 weeks to increase the likelihood of injected cell identification with the CM-DiI label. With the proof of concept that these cells survive, can be identified at 4 weeks, and remain within injection lines, longer-duration studies can be performed. Next, the injections themselves could have altered AV nodal conduction. Despite the large number of injections and volume of solution delivered in the myocardium, there was minimal effect on AV nodal conduction with saline alone. This observation suggests that the mechanical injury of the procedure had little long-term influence. Fourth, the dogs were studied under general anesthesia, which may directly affect AV nodal conduction. However, all dogs, including those in the control cohort, were managed with a similar anesthesia approach throughout the study. Finally, the study contained a relatively small number of animals in each cohort. This limitation stems from the cost of this large animal model, which employed minimally invasive techniques that can be easily translated into a human laboratory.
Sources of Funding
Dr Bunch received a postdoctoral fellowship grant from the American Heart Association and a National Institutes of Health (NIH) Clinical Research Scholarship. Dr Packer received an unrestricted research grant from Symphony Medical, Inc, and receives research grants from Biosense-Webster.
Dr Bunch received honoraria for travel expenses to present these data from Symphony Medical, Inc, and the American Heart Association, Greater Midwest Affiliate, Postdoctoral Fellowship. Dr Packer has served on the advisory board of Symphony Medical, Inc, and has served on the advisory board of Biosense-Webster. The other authors report no conflicts.
Wyse DG, Waldo AL, DiMarco JP, Domanski MJ, Rosenberg Y, Schron EB, Kellen JC, Greene HL, Mickel MC, Dalquist JE, Corley SD. Atrial fibrillation follow-up investigation of rhythm management, I: a comparison of rate control and rhythm control in patients with atrial fibrillation. N Engl J Med. 2002; 347: 1825–1833.
Ozcan C, Jahangir A, Friedman PA, Patel PJ, Munger TM, Rea RF, Lloyd MA, Packer DL, Hodge DO, Gersh BJ, Hammill SC, Shen WK. Long-term survival after ablation of the atrioventricular node and implantation of a permanent pacemaker in patients with atrial fibrillation. N Engl J Med. 2001; 344: 1043–1051.
Wood MA, Brown-Mahoney C, Kay GN, Ellenbogen KA. Clinical outcomes after ablation and pacing therapy for atrial fibrillation: a meta-analysis. Circulation. 2000; 101: 1138–1144.
Huang SK, Bharati S, Graham AR, Gorman G, Lev M. Chronic incomplete atrioventricular block induced by radiofrequency catheter ablation. Circulation. 1989; 80: 951–961.
Chen SA, Lee SH, Chiang CE, Tai CT, Wu TJ, Cheng CC, Wen ZC, Chiou CW, Ueng KC, Chang MS. Electrophysiological mechanisms in successful radiofrequency catheter modification of atrioventricular junction for patients with medically refractory paroxysmal atrial fibrillation. Circulation. 1996; 93: 1690–1701.
Feld GK, Fleck RP, Fujimura O, Prothro DL, Bahnson TD, Ibarra M. Control of rapid ventricular response by radiofrequency catheter modification of the atrioventricular node in patients with medically refractory atrial fibrillation. Circulation. 1994; 90: 2299–2307.
Bunch TJ, Bruce GK, Johnson SB, Sarabanda A, Milton MA, Packer DL. Analysis of catheter-tip (8-mm) and actual tissue temperatures achieved during radiofrequency ablation at the orifice of the pulmonary vein. Circulation. 2004; 110: 2988–2995.
Lamme EN, van Leeuwen RT, Mekkes JR, Middelkoop E. Allogeneic fibroblasts in dermal substitutes induce inflammation and scar formation. Wound Rep Regen. 2002; 10: 152–160.
Yang W, Arii S, Mori A, Furumoto K, Nakao T, Isobe N, Murata T, Onodera H, Imamura M. sFlt-1 gene-transfected fibroblasts: a wound-specific gene therapy inhibits local cancer recurrence. Cancer Res. 2001; 61: 7840–7845.
Rodriguez LM, Leunissen J, Hoekstra A, Korteling BJ, Smeets JL, Timmermans C, Vos M, Daemen M, Wellens HJ. Transvenous cold mapping and cryoablation of the AV node in dogs: observations of chronic lesions and comparison to those obtained using radiofrequency ablation. J Cardiovasc Electrophysiol. 1998; 9: 1055–1061.
Taylor GW, Kay GN, Zheng X, Bishop S, Ideker RE. Pathological effects of extensive radiofrequency energy applications in the pulmonary veins in dogs. Circulation. 2000; 101: 1736–1742.
Pappone C, Oral H, Santinelli V, Vicedomini G, Lang CC, Manguso F, Torracca L, Benussi S, Alfieri O, Hong R, Lau W, Hirata K, Shikuma N, Hall B, Morady F. Atrio-esophageal fistula as a complication of percutaneous transcatheter ablation of atrial fibrillation. Circulation. 2004; 109: 2724–2726.
Okuyama Y, Miyauchi Y, Park AM, Hamabe A, Zhou S, Hayashi H, Miyauchi M, Omichi C, Pak HN, Brodsky LA, Mandel WJ, Fishbein MC, Karagueuzian HS, Chen PS. High resolution mapping of the pulmonary vein and the vein of Marshall during induced atrial fibrillation and atrial tachycardia in a canine model of pacing-induced congestive heart failure. J Am Coll Cardiol. 2003; 42: 348–360.
Spach MS, Dolber PC. Relating extracellular potentials and their derivatives to anisotropic propagation at a microscopic level in human cardiac muscle: evidence for electrical uncoupling of side-to-side fiber connections with increasing age. Circ Res. 1986; 58: 356–371.
Feld Y, Melamed-Frank M, Kehat I, Tal D, Marom S, Gepstein L. Electrophysiological modulation of cardiomyocytic tissue by transfected fibroblasts expressing potassium channels: a novel strategy to manipulate excitability. Circulation. 2002; 105: 522–529.
Gaudesius G, Miragoli M, Thomas SP, Rohr S. Coupling of cardiac electrical activity over extended distances by fibroblasts of cardiac origin. Circ Res. 2003; 93: 421–428.
Ignotz RA, Massague J. Transforming growth factor-beta stimulates the expression of fibronectin and collagen and their incorporation into the extracellular matrix. J Biol Chem. 1986; 261: 4337–4345.
Varga J, Rosenbloom J, Jimenez SA. Transforming growth factor beta causes a persistent increase in steady-state amounts of type I and type III collagen and fibronectin mRNAs in normal human dermal fibroblasts. Biochem J. 1987; 247: 597–604.
Shi Y, O’Brien JE, Fard A, Zalewski A. Transforming growth factor-beta 1 expression and myofibroblast formation during arterial repair. Arterioscler Thromb Vasc Biol. 1996; 16: 1298–2305.
Desmouliere A, Geinoz A, Gabbiani F, Gabbiani G. Transforming growth factor-beta 1 induces alpha-smooth muscle actin expression in granulation tissue myofibroblasts and in quiescent and growing cultured fibroblasts. J Cell Biol. 1993; 122: 103–111.
Malmstrom J, Lindberg H, Lindberg C, Bratt C, Wieslander E, Delander EL, Sarnstrand B, Burns JS, Mose-Larsen P, Fey S, Marko-Varga G. Transforming growth factor-beta 1 specifically induce proteins involved in the myofibroblast contractile apparatus. Mol Cell Proteomics. 2004; 3: 466–477.
Roberts AB, Sporn MB, Assoian RK, Smith JM, Roche NS, Wakefield LM, Heine UI, Liotta LA, Falanga V, Kehrl JH. Transforming growth factor type beta: rapid induction of fibrosis and angiogenesis in vivo and stimulation of collagen formation in vitro. Proc Nat Acad Sci U S A. 1986; 83: 4167–4171.
Silva GV, Litovsky S, Assad JA, Sousa AL, Martin BJ, Vela D, Coulter SC, Lin J, Ober J, Vaughn WK, Branco RV, Oliveira EM, He R, Geng YJ, Willerson JT, Perin EC. Mesenchymal stem cells differentiate into an endothelial phenotype, enhance vascular density, and improve heart function in a canine chronic ischemia model. Circulation. 2005; 111: 150–156.
Yoshioka T, Ageyama N, Shibata H, Yasu T, Misawa Y, Takeuchi K, Matsui K, Yamamoto K, Terao K, Shimada K, Ikeda U, Ozawa K, Hanazono Y. Repair of infarcted myocardium mediated by transplanted bone marrow–derived CD34+ stem cells in a nonhuman primate model. Stem Cells. 2005; 23: 355–364.
Atrioventricular (AV) nodal ablation for the management of AF is irreversible and requires permanent pacemaker implantation. We hypothesized that as an alternative, implantation of autologous fibroblasts in the perinodal region would focally modify AV nodal conduction and that this modulation would be enhanced by pretreatment with transforming growth factor-β1, a stimulant of fibroblasts. These data have several implications that are of potential clinical relevance. First, the observed injury was well localized to targeted injection areas. Second, we were able to focally modify AV nodal conduction without long-term high-grade AV block. Third, the AV nodal modification favorably altered the ventricular rate during pacing-induced AF. In the combination therapy cohort, the effect translated to a heart rate reduction from an average of 207 bpm in AF to 103 bpm after treatment. These findings suggest that this therapy has potential in the management of AF with rapid ventricular rates seen in clinical practice. These findings are encouraging because other approaches to AV node modification have been complicated by high rates of AV block and the requirement for pacemaker implantation.
Dr Bunch was the recipient of the Samuel A. Levine Young Investigator Award from the American Heart Association. These data were presented in part in abstract form for this competition at the 2005 AHA Scientific Sessions, Dallas, Tex.