| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2005;80:1362-1369
© 2005 The Society of Thoracic Surgeons
a Department of Cardiovascular Surgery, Okayama, Japan
b First Department of Pathology, Okayama University Graduate School of Medicine and Dentistry, Okayama, Japan
Accepted for publication April 6, 2005.
* Address reprint requests to Dr Ishino, Department of Cardiovascular Surgery, Okayama University Graduate School of Medicine and Dentistry, 2-5-1 Shikata-cho, Okayama-City, 00-8558 Japan (Email: ishino{at}tb3.so-net.ne.jp).
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsCommentReferences |
|---|
METHODS: Dogs with normal myocardium underwent either holmium:yttrium-aluminum-garnet laser TMR, TMR using an ultrasonic activated surgical blade, or a thoracotomy as sham operation. The responses of mean arterial pressure to topical application of bradykinin were examined at 3 time points: before, 1 hour after, and 2 weeks after surgery. The hearts were excised for Western blot and immunohistochemical analysis.
RESULTS: The response of mean arterial pressure to bradykinin was similarly attenuated in both TMR groups 1 hour after treatment and decreased to almost none after 2 weeks compared with pretreatment values. By comparison, the sham group showed persistent responses at both time points. Tissue tyrosine hydroxylase content of the treated area decreased significantly compared with the non-treated area in both TMR groups. Immunohistochemistry using anti-Protein Gene Product 9.5 and anti-synaptophysin antibodies showed a significant decrease in the number of positive nerve fibers in both TMR treatment groups compared with the sham group.
CONCLUSIONS: Transmyocardial revascularization caused partial alteration in myocardial innervation immediately after TMR. Tissue responses may continue to occur for the first 2 weeks after treatment. Tissue responses may also contribute to the development of denervation regardless of the energy source in non-ischemic canine myocardium.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
Myocardial denervation is believed to be a possible mechanism that might improve angina in the early postoperative period when there is still no influence of increased regional myocardial perfusion [4]. Some clinical studies support this mechanism by showing the correlation between clinical symptoms and decreased uptake of carbon-11 hydroxyephedrine positron emission tomography [5] and iodine 123-labeled meta-iodobenzylguanide scintigraphy [6, 7], in addition to experimental evidence from animal studies [4]. On the other hand, there are some experimental studies showing no functional alteration of cardiac innervation [8, 9]. The precise consequences to the myocardial innervation within the first few weeks remain unclear as the time points and the methodology used were different in the various experimental studies. Moreover, regional myocardial denervation is known to occur when channels are created with radiofrequency energy [10], which suggests that regional myocardial denervation is potentially nonspecific to laser TMR as is neoangiogenesis [11, 12].
The purpose of this study was to investigate the time at which regional myocardial denervation occurs in the early postoperative period and to evaluate whether or not it is a specific effect of selective energy sources. We compared holmium:yttrium-aluminum-garnet (Ho:YAG) laser TMR system with the ultrasonically activated scalpel, which we previously reported to produce a similar TMR channel morphologically and a similar degree of neoangiogenesis to the Ho:YAG laser TMR [13] immediately after the procedure and 2 weeks after the treatment.
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
Surgical Instruments
Ultrasound generator: Harmonic scalpel (Ethicon Endo-Surgery, Inc, Cincinnati, OH) with a 2-mm diameter rod-shaped blade was used to create ultrasound TMR channels. As reported in our previous study [13], the generator was set to deliver 100 µm amplitude of vibration to achieve similar sizes of acute thermal damage zone (4.63 ± 0.44 vs 4.43 ± 0.28 mm2; p = not significant) and similar vascular density to laser TMR, as determined by factor VIII and proliferating cell nuclear antigen immunostaining in the channel neighboring area (30.0 ± 3.6 vs 41.9 ± 5.6 and 11.4 ± 1.4 vs 15.0 ± 2.8, respectively; p = not significant) 2 weeks after treatment.
Laser generator: Ho:YAG laser (TMR 2000 Laser System; CardioGenesis Corporation, Inc, Foothill Ranch, CA) with a 1-mm diameter optical fiber tip of the type that is currently used for clinical laser TMR was used to create laser TMR channels. The output was set to 7 watts/5 pulses.
Surgical Procedures
Eighteen mongrel dogs, each weighing 18 to 23 kg, with normal myocardium were used for the experiments. Animals were anesthetized with intramuscular injection of ketamine (10 mg/kg) and intravenous injection of pentobarbiturate (3 to 4 mg/kg), and they were intubated and maintained with inhaled isoflurane (1.0% to 2.0%). Surface electrocardiogram and right femoral arterial blood pressure were monitored. A left thoracotomy was performed, the pericardium was opened, and epicardial application of bradykinin was performed. Bradykinin was kept in a 36°C water bath and delivered onto a 1 x 1 cm gauze (150 µg/gauze). Bradykinin-soaked gauzes were randomly placed onto the surfaces of the anterior basal (base), mid-anterior (mid), and apical (apex) regions of the left ventricle. Gauze was kept on the epicardial surface for 2 minutes and then removed. After removal, the area was washed with normal saline. At least 10 minutes was allowed between different applications in order to avoid the effect of tachyphylaxis. Gauze was placed onto each site once. Changes in mean arterial blood pressure (MAP) and surface electrocardiogram were recorded during the stimulation. After initial observation of the responses to the application of bradykinin, animals were randomly assigned to one of the following three groups: (1) laser TMR group (n = 6) in which channels were created by using Ho:YAG laser system; (2) ultrasound TMR group (n = 6) in which channels were created by using the Harmonic Scalpel; and the sham group (n = 6) in which thoracotomy and pericardiotomy were performed but no TMR channels were created. Channels for TMR groups were created from the epicardial surface as previously described [13]. Channels were made with a density of approximately 1 channel/cm2 in the same area of the base, mid, and apex regions where the bradykinin soaked gauzes were applied. One hour after creating the channels, another set of bradykinin soaked gauze pieces were applied onto each region, and the MAP response was recorded again in the same fashion. After the second MAP measurement, the pericardium was reapproximated, the chest was closed, and the animals recovered from the anesthesia. After 2 weeks from the initial surgery, the animals were anesthetized, and the surface of the heart was re-exposed in the same fashion. The MAP response to the topical application of bradykinin was recorded. The animals were then sacrificed by bolus injections of potassium chloride and the heart was excised. Separate samples were obtained for immunoblotting and immunohistochemistry.
Immunoblotting
Approximately 500 mg of full-thickness tissue samples, including the channel remnant of TMR-treated and untreated regions (the lateral wall of the left ventricle of treated animals and corresponding regions of sham-operated animals), were excised and frozen at –80°C before analysis. Tissue content of tyrosine hydroxylase (TH), which is known to be an enzyme contained in sympathetic efferent nerve fibers [14], was measured. Immunoblotting was performed according to the method described by Kwong and colleagues [4]. Densitometry of the bands was performed on a Macintosh computer using the public domain NIH Image program developed at the U.S. National Institutes of Health and available on the Internet at http://rsb.info.nih.gov/nih-image/.
Tissue Fixation
Myocardial samples of the base, mid, and apex were obtained from each animal. All samples were fixed overnight in 10% neutral buffered formalin, dehydrated, and embedded in paraffin. Serial 4 µm sections were cut and stained with Masson's trichrome procedure to evaluate the general morphology.
Immunohistochemical Analysis
A 4 µm section of each sample was stained with mouse monoclonal anti-human Protein Gene Product 9.5 (PGP 9.5) IgG (1:200 [Novocastra, Newcastle, UK]), and mouse monoclonal anti-human Synaptophysin IgG (1:200 [Novocastra]). The PGP 9.5 is a soluble protein that is a major component of the neuronal cytoplasm. It is known to be a marker for assessing the general pattern of functional innervation of the mammalian cardiovascular system [15, 16]. Synaptophysin is a component of presynaptic vesicles, which is also a useful marker for estimating the general distribution of nerve terminals within the mammalian heart [17, 18]. Sections were fixed, hydrated, and blocked for 15 minutes with hydrogen peroxide (3%) in phosphate buffered saline. Primary antibodies were added and incubated for 2 hours at room temperature. Biotinylated secondary antibodies were added for 30 minutes at room temperature. Sections were reacted with horse radish peroxidase-conjugated streptavidin for 30 minutes at room temperature and were developed with 3, 3'-diaminobenzidine. In order to quantify the distribution pattern of the innervation, three random fields of each section were chosen in the myocardium as great 5 mm from the center of each channel remnant. The numbers of immunoreactive neurons were counted manually by an investigator who was blinded to the treatments, and these were confirmed by a second investigator. The results are shown as the number of immunoreactive neurons/mm2.
Statistical Analysis
All data are given as mean ± standard deviation. Serial measurements were compared using the one-way analysis of variance test. When overall significance was detected, Fisher's post hoc test was used to delineate which comparisons were significantly different. A p value of < 0.05 was considered to be significant.
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
Morphology of the TMR Channel
Examples of TMR channels created by laser and ultrasound scalpel are shown in Figure 1. Both channels showed similar size of fibrosis (1.3 ± 0.2 mm2 vs 1.4 ± 0.2 mm2, respectively; p = not significant).
|
|
|
Two weeks after treatment, the percent in decrease of MAP to bradykinin application was further decreased in both TMR-treated groups when compared with baseline conditions of the corresponding group and also at 1 hour after treatment, whereas untreated regions in sham-operated animals showed persistent percent in decrease of MAP. There were no statistically significant differences between the two treated groups when similar regions were compared. Also, no statistically significant differences were noted among treated regions.
Immunoblotting
There was a significant decrease of tissue TH content in all TMR-treated regions. Untreated regions (the lateral wall of the left ventricle of TMR treated animals) showed similar TH content with those of the sham group (see Fig 3A for example).Densitometry showed almost 80% decrease in all treated regions. The TH content was preserved in all untreated regions (Fig 3B). There were no statistically significant differences among treated regions of each group.
|
|
|
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
Several studies have shown alteration of the cardiac afferent nerves after laser TMR in animal experiments. Different methodologies were performed at various time points after treatment [4, 8, 9, 19]. However, in all previous studies the measurements were performed only at a single time point in the early postoperative period after laser TMR. In this study, we made measurements at two different time points postoperatively (ie, 1 hour after treatment and 2 weeks after treatment in the same animal in order to determine the changes occurring in the afferent nerves over this time period). There was a trend that the magnitude of the percent in decrease of MAP was maximal at the basal and minimal at the apical region of the baseline measurement. This finding correlates with the anatomical distribution of nerves in a normal canine heart showing a decrease in the neural density from the base toward the apex of the heart. This trend, although there was no statistical difference, was preserved at 1 hour post-TMR treatment in both groups. This result suggests that the attenuation of the percent in decrease of MAP 1 hour post-treatment is mainly due to the direct damage of the nerve terminals exposed to thermal or acoustic energy. At 2 weeks post-treatment, this trend was not observed and the percent in decrease of MAP was similarly absent in all TMR-treated regions. This finding suggests that cardiac denervation was caused not only by acute damage of nerve fibers that were directly exposed to energy during channel creation, but also by some type of reactive response after treatment that occurred within the first 2 weeks. Time-dependent tissue injury after Ho:YAG laser TMR has been noted by several investigators. Hughes and colleagues [20] showed variations in tissue-water content over time after tissue injury. Tissue injury after Ho:YAG laser TMR is also known to induce neoangiogenesis in the surrounding tissue of the channels as a result of a general healing process [11, 12]. Thus, some types of regional tissue responses, which possibly include a general healing process, may have disrupted the function of the regional afferent nerves. However, this hypothesis was not tested by our current study.
Because we had planned to measure MAP responses to topical bradykinin application at two different time points (ie, 1 hour and 2 weeks after treatment) in the same animals, we followed the methodology of Kwong and colleagues [4] . Arora and colleagues [9] and Hirsch and colleagues [19] suggested that the amount of bradykinin used in Kwong and colleagues' experiment [4] was excessive and some of the chemical might have entered the systemic circulation to cause vasodilation, resulting in decrease in MAP. Minisi and colleagues [8] suggested that in Kwong and colleagues' model [4] the animal hearts did not undergo sinoaortic denervation or vagotomy that might have otherwise caused sinoaortic baroreflex and the vagal cardiopulmonary reflex. We used 150 µg bradykinin per 1 x 1 cm2 gauze for this study. Minisi and colleagues [8] had used a similar amount of bradykinin as in this study (500 µg/2 x 2 cm2 gauze; 125 µg/cm2) for topical application in a canine model with sinoaortic denervation and vagotomy and found a significant increase in MAP. This finding suggests that even the amount of bradykinin that Minisi and colleagues [8] had used could be considered a "high dose" as Arora and colleagues [9] and Hirsch and colleagues [19] had suggested; it did not cause direct systemic vasodilation to show MAP decrease. Thus we consider that the amount of bradykinin we used could not have had an effect on the systemic vasculature. Because we planned to measure two different time points after TMR treatments, we did not perform sinoaortic denervation and vagotomy because it is not theoretically possible to maintain the autoregulation of the animals postoperatively if either of them is performed. Minisi and colleagues [8] found that animals with sinoaortic denervation and vagotomy showed an increase in MAP in the baseline state application of bradykinin, whereas we observed a decrease in MAP at the same time point. This conflicting finding may be related to whether sinoaortic denervation and vagotomy were performed or not. In addition, the anesthetics used in our current study were different from those in Minisi and colleagues' [8] study. The anesthesia induced in our current study might have been relatively deep, because we also did not observe changes in heart rate during bradykinin application. Our findings at the baseline measurement might have been a complex combination of such reflexes. However, because the regions treated and tested were predominantly innervated by the cardiac afferent nerve terminals, our findings still support that the regional sympathetic afferent nerves were altered by both treatments.
In this study, TH content showed significant reduction in both laser and ultrasound TMR-treated regions. Even though we followed the methodology described by Kwong and colleagues [4], and measured the TH content at the same time point, we could not reproduce their observation of the "absence" of TH content, even in laser TMR-treated samples. Although transmural TMR channel creation by either energy source causes a significant degree of anatomic destruction within the first few weeks, our results suggest that some nerve fibers may still remain viable. These observations may support the findings of Arora and colleagues [9], which showed no functional alteration of both sympathetic and parasympathetic efferent nerve function after laser TMR.
The interregional distribution pattern of both PGP 9.5 and synaptophysin positive neurons in the sham-operated group showed that there were more positive neurons present in the basal region than in both the mid and apical regions. This finding also correlates with the anatomical distribution of the nerve terminal of a normal canine heart as well as the findings in the MAP response. Of note, the numbers of immunoreactive nerve fibers were similarly reduced in each treated region with both energy sources. This finding suggests that the cardiac distribution of innervation had been altered similarly regardless of the energy source or the treated location in the heart. These results are in agreement with the tissue TH content data in which complete absence of immunoreactive nerve fibers was not observed.
We found that a comparable degree of denervation was achieved with both methods of TMR (laser and ultrasound) when a similar degree of angiogenesis [13] was induced irrespective of the differences in the energy source used to create the TMR channels. Yamamoto and colleagues [10] have shown a similar reduction in percent in decrease of MAP and reduced TH content in treated regions by using a radiofrequency ablation system to create TMR channels. Taken together with the findings of current study, regional denervation could be a nonspecific response to channel creation by any kind of energy source. The important point, regardless of the device, is that all channels created were transmural. Experimental findings suggest that non-transmural TMR channels produced by percutaneous myocardial revascularization provide less denervation effects [21, 22] because these channels have less influence upon the sympathetic afferent nerves by which the pain signal is considered to be mainly carried [23, 24].
In summary, time-dependent regional denervation of the afferent nerves could be responsible for angina relief in the early postoperative period after TMR, and this effect is nonspecific to laser TMR. There was significant destruction of cardiac efferent nerves; however, there were still viable nerves present and their function could be preserved.
There were some limitations to this study. First, although laser TMR is clinically utilized in the treatment of the ischemic, hibernating myocardium, we treated normal, non-ischemic myocardium in this study. This was specifically chosen to avoid the effect of ischemia to the myocardial nervous system. However, because the histologic findings are similar in normal and chronically ischemic myocardium [25], the results of the current study may be expected to apply to chronically ischemic hearts. Second, the diameter of the tip used for the laser TMR and ultrasound TMR was different (ie, 1 and 2 mm, respectively). This difference was primarily due to the availability of the ultrasound probe. However, in our previous study we reported the exact parameters to create ultrasound channels, which are histologically similar to the Ho:YAG laser channels.
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  P. Atluri, C. M. Panlilio, G. P. Liao, E. E. Suarez, R. C. McCormick, W. Hiesinger, J. E. Cohen, M. J. Smith, A. B. Patel, W. Feng, et al. Transmyocardial revascularization to enhance myocardial vasculogenesis and hemodynamic function. J. Thorac. Cardiovasc. Surg., February 1, 2008; 135(2): 283 - 291. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. A. Horvath Light and Ice Cream Ann. Thorac. Surg., August 1, 2006; 82(2): 771 - 771. [Full Text] [PDF]
|
|
|
|
|
|
J. F. Beek and J. A.P. van der Sloot Invited commentary Ann. Thorac. Surg., October 1, 2005; 80(4): 1369 - 1370. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
]) 2 weeks after treatment. The PGP 9.5 and synaptophysin immunoreactive fibers are present in the interstitial spaces in the sham-operated myocardium (stained in brown and indicated by arrowheads [
]), whereas there are almost no immunoreactive fibers seen in either TMR groups (original magnification, x100, scale bar: 200 µm).