Annals of Vascular Surgery
Volume 23, Issue 6 , Pages 758-763, November 2009

Extra-Anatomical Revascularization of the Adamkiewicz Artery Using the Internal Mammary Artery: Preliminary Anatomical Study

  • Serguei Malikov

      Affiliations

    • Department of Vascular Surgery, Marseille Medical School, Mediterranean University, Marseille APHM-Timone University Hospital Center, 13385 Marseille cedex 05, France
    • Corresponding Author InformationCorrespondence to: Dr. S. Malikov, Service de Chirurgie Vasculaire, Hôpital de la Timone 264 rue Saint-Pierre, 13385 Marseille Cedex 05, France
    • ,
  • Pierre-Edouard Magnan

      Affiliations

    • Department of Vascular Surgery, Marseille Medical School, Mediterranean University, Marseille APHM-Timone University Hospital Center, 13385 Marseille cedex 05, France
    • ,
  • Alain Branchereau

      Affiliations

    • Department of Vascular Surgery, Marseille Medical School, Mediterranean University, Marseille APHM-Timone University Hospital Center, 13385 Marseille cedex 05, France
    • ,
  • Jean-Michel Bartoli

      Affiliations

    • Department of Radiology, Marseille Medical School, Mediterranean University, Marseille, France, APHM-Timone University Hospital Center, 13385 Marseille cedex 05, France
    • ,
  • Pierre Champsaur

      Affiliations

    • Laboratory of Anatomy, Marseille Medical School, Mediterranean University, 13385 Marseille cedex 05, France

Article Outline

Ischemic spinal cord injury remains a major complication of both open and endovascular repairs of extensive lesions of the thoracic or thoracoabdominial aorta. Patients undergoing endovascular treatment cannot benefit from direct revascularization of the Adamkiewicz artery (AA). Primary revascularization of the intercostal artery (ICA) giving rise to the AA using the internal mammary artery (IMA) could ensure uninterrupted flow in the AA even if the origin of the feeding ICA was obstructed. The purpose of this study was to assess the anatomical feasibility of revascularization of the ICA giving rise to the AA using the IMA. Twenty-four dissections were carried out on 12 cadavers (eight men, four women) with a mean age of 76 at the time of death. Preparation consisted of intra-arterial injection of polymethylsiloxane (Rhodorsil®, Rhodia, France). For each IMA, the following parameters were determined: diameter in relation to the ICA in the paravertebral region before division, length, and level of the intercostal space in which direct anastomosis was possible. Dissection showed that the mean diameter at the end of the IMA was 1.8mm (range 1.2-2.4). The mean diameter of the ICA in the paravertebral region was compatible with that of the IMA, i.e., 1.6mm (range 0.9-2.5). The mean length of the IMA was 185mm (range 165-230). The lowest intercostal space available in the paravertebral region for direct anastomosis between the IMA and ICA was the seventh space in one case, the eighth in 12, the ninth in eight, and the tenth in three. The findings of this preliminary study document the feasibility of using the IMA to revascularize the ICA in the paravertebral region. This technique could provide a means of preserving spinal cord vascularization during endovascular treatment of thoracic or thoracoabdominal aortic lesions.

 

Back to Article Outline

Introduction 

The reported risk of neurological defects due to spinal cord ischemia during open repair ranges 5-16% for thoracoabdominal aortic aneurysm1, 2 and 2.7-6.5% for isolated thoracic aortic aneurysm.3, 4 The importance of intercostal artery (ICA) reattachment in preserving vascularization of the Adamkiewicz artery (AA) has been documented.5 Although the superior dorsal artery, middle dorsal artery, and terminal cone artery contribute, the AA (arteriaradicularis magna) is the main source of spinal cord blood flow at the thoracolumbar level. Although more difficult to evaluate, the risk of spinal cord injury is also present during endovascular repair of thoracic aortic aneurysm using the techniques currently under investigation. Reported risk rates during endovascular repair have ranged 3.7-12%.6, 7, 8 By definition, use of endovascular techniques rules out direct revascularization of the AA. We speculate that primary revascularization of the ICA giving rise to the AA using the internal mammary artery (IMA) would ensure uninterrupted blood flow to the spinal cord even if the ostium of that ICA was obstructed (Fig. 1). The purpose of this study was to evaluate the anatomical feasibility of revascularization of the ICA giving rise to the AA using the IMA, to determine the length and diameter of the IMA, and to compare the diameters of the IMA and ICA in the paravertebral region.

Back to Article Outline

Materials and Methods 

Twenty-four anatomical dissections were carried on 12 cadavers (eight men, four women) with a mean age of 76 years at the time of death. The cadavers were preserved by the Winckler method. Preparation consisted of intra-arterial injection of colored latex. Dissections were carried out using microsurgical instrumentation and a binocular magnifying loupe. The IMA was dissected from the origin up to the superior epigastric artery. The following parameters were noted for each IMA: origin, branching, length, distal diameter, and relationship with the phrenic nerve.

For the ICA, the posterior segment was exposed between the projections of the paravertebral and middle axillary lines (Fig. 2). The posterior segment was chosen because of its accessibility since the ICA is located in the middle of the intercostal space at that level and its diameter is regular and fairly large. The parietal and intercostal perforating branches arise distally. All ICAs studied included between the fourth intercostal space and the lower edge of the twelfth rib (twelfth ICA or subcostal artery, level D12). All IMAs were mobilized and transposed into the paravertebral region to determine the feasibility of anastomosis with any of the ICAs that might give rise to the AA. The following parameters were noted for each IMA: length, level in intercostal spaces where direct anastomosis between the IMA and ICA was possible, and diameter in comparison with that of the ICA.

  • View full-size image.
  • Fig. 2 

    ICAs were studied in the area between the paravertebral region (A) and middle axillary line (B). 1, dorsospinal branch; 2, lateral and inferior perforating branch.

Back to Article Outline

Results 

The origin of the IMA was located on the proximal part of the subscapular artery in all 24 dissections. In 22 dissections, the origin was a single trunk. In the remaining two dissections, there was a common trunk for the IMA and suprascapular artery. The second variant was observed on both sides. The mean distance between the origin of the internal mammary vein and IMA was 25mm (range 10-30), and this arrangement was more pronounced on the right.

The phrenic nerve crossed the IMA in most cases (15 of 24 observations), and its direction was most often oblique from the outside to inside. The phrenic nerve crossed the IMA in the front in 12 cases and in the back in three cases (Fig. 3). The mean distance between the origin of the IMA and the phrenic nerve crossing point nerve was 22mm (range 10-35).

A lateral costal branch was observed in six cases including one in which its diameter was equal to that of the IMA. The pericardiophrenic artery was present in most cases (n=21). The mean distance between the origin of the IMA and the beginning of the pericardiophrenic artery was 52mm (range 25-65). In four cases the diameter of the IMA decreased significantly after the beginning of the pericardiophrenic artery.

The mean diameter of the distal part of the IMA was 1.8mm (range 1.2-2.4). The mean length of the IMA was 185mm (range 165-230). In 17 cases the IMA was longer on the left than on the right. In most cases the end of the IMA reached the rib cage at the level of the sixth intercostal space (Table I).

Table I. Level of termination of the IMA
LevelLeftRight
Fifth rib01
Fifth space12
Sixth rib67
Sixth space1312
Seventh rib32
Seventh space10
Total2424

In 22 cases, the IMA bifurcated into the superior epigastric artery and musculophrenic artery. In the remaining three cases, the IMA trifurcated with an additional musculophrenic artery. Dissection of the epigastric artery down to the intersection with diaphragm (Fig. 4) resulted in a mean increase of 40mm (range 25-60) in the length of the IMA.

  • View full-size image.
  • Fig. 4 

    Dissection of the distal segment of the IMA. Dissection was extended in the direction of the epigastric artery (arrow) to the point of intersection with the diaphragm.

Exposure of the intercostal arteries between the paravertebral and middle axillary lines was easy. The artery ran in the middle part of the intercostal space and was not covered by intercostal muscle. The mean diameter of the ICA in the paravertebral region from D5 to D12 was 1.6mm (range 0.9-2.5). The mean diameters measured at each level are listed in Figure 5. Only three ICAs had mean diameters <1mm at the level of the paravertebral line. This finding, i.e., diameter <1mm, involved the fourth ICA in two cases and the fifth ICA in one case. In all dissections, the diameter of the ICA was observed to increase gradually from D5 to D12.

  • View full-size image.
  • Fig. 5 

    Mean diameter of the ICAs facing the paravertebral line (PVL) and middle axillary line (MAL) at the D5 and D12 levels. A progressive increase in the diameter of the ICAs was observed in all dissections.

Comparison of the anatomical features of the IMA and ICA showed that the lowest intercostal space in which direct anastomosis could be performed between the IMA after mobilization and transposition in the paravertebral region and the ICA in this territory was the seventh intercostal space in one case, the eighth in 12, the ninth in eight cases, and the tenth in three cases. In five cases, anastomosis in the eleventh intercostal space could be achieved by extending dissection of the IMA in the direction of the epigastric artery (Fig. 6).

  • View full-size image.
  • Fig. 6 

    Anatomical preparation showing transposition of the mammary artery into the paravertebral region. The ICAs were also dissected. The lowest level at which direct anastomosis between the IMA and ICA was feasible was the eleventh intercostal space.

Back to Article Outline

Discussion 

This study demonstrates the anatomical feasibility of revascularization of the ICA giving rise to the AA using the IMA. Our findings showed that the available length of the IMA after transposition into the paravertebral region was adequate and that the diameter of the IMA was comparable to that of the ICA for microsurgical anastomosis. Recent minimally invasive techniques either by thoracoscopy9 or by video-assisted minithoracotomy10 could be used for harvesting the IMA for extra-anatomical revascularization of the AA.

This technique raises several unresolved hemodynamic issues. The main issue involves creation of competitive flow in the ICA supplying the AA. Indeed, this technique consists of making an anastomosis between the IMA and ICA downstream from the origin of the spinal artery. An animal model study will be needed to answer this question.

The first extra-anatomical revascularization of visceral arteries was proposed by Macierewicz et al.11 to enable endovascular treatment of some types of aortic aneurysm. This principle led us to evaluate the possible advantages of primary revascularization of the AA. Data from recent anatomical12, 13, 14 and angiographic15, 16, 17 studies determined that the origin of the AA was usually located between D5 and L3, with the greatest frequency being between D8 and D12. The surgical anatomy of the ICA in the posterior region of the thorax was specifically studied by Larichev in 1990.18 Findings demonstrated that the anatomical relations of the ICA at the paravertebral level were highly stable and consistent. Anatomoclinical studies designed to evaluate the feasibility of transferring a costal graft also demonstrated the feasibility of performing microsurgical anastomosis on the posterior segment of the ICA.19, 20 Our study confirms data from the literature and shows that the diameter of the posterior segments of the ICA is greater than 1mm and that these vessels can be mobilized. These features are important for performing microsurgical anastomosis under favorable conditions.

Although surgical anatomy of the IMA has been described in several recent studies focusing on myocardial revascularization,21, 22, 23 there has been no information on transposition of the IMA into the paravertebral region or on its relationship with the ICA. Transposition requires mobilization of the IMA all the way to its origin in order to rotate the vessel backward as far as possible. Mobilization is difficult due to the presence of the proximal branches of the IMA (the lateral costal branch and pericardiophrenic branch), to the anatomy of the origin of the mammary vein, and to the intersection with the phrenic nerve. In some cases, transposition may be facilitated by the presence of a common trunk between the IMA and the suprascapular artery since this arrangement results in a posterior course of the IMA. This feature was noted in two of our dissections (8%). In a series of 100 dissections, Henriquez-Pino et al.21 noted the presence of a common trunk in 30% of cases.

The lateral costal branch noted in six of our dissections (25%) was studied by Sutherland and Desai,24 who noted the presence of this “redundant” artery in one-third of their dissections. In our dissections, the presence of the pericardiophrenic artery was observed much more consistently. Henriquez-Pino et al.21 reported the same anatomical arrangement, with the distance between the IMA and beginning of the pericardiophrenic artery being 43±9mm.

Numerous anatomical studies have been carried out to evaluate the risk of paralysis of the diaphragm after mobilization of the IMA.21, 25, 26 In agreement with previously published data, our findings show that the phrenic nerve crosses the IMA in 50% of cases. The distance between the origin of the IMA and the phrenic nerve crossing point is around 20mm (range 5-45).

To improve mobilization of the IMA, some authors27 have advised section of the internal mammary vein at the origin. In agreement, our dissection showed that the origin of the internal mammary vein is located at a mean distance of 25mm in front of the origin of the IMA.

Measurements made in the larger series published by Henriquez-Pino et al.21 and Lachman and Satyapal22 showed that the mean length of the IMA was 180±6mm and 240±21mm, respectively. According to an angiographic study carried out on 80 patients,28 the distal diameter of the IMA was 2.19±0.24mm. Measurement of the same diameter in a morphometric study on 62 cadavers22 gave 1.98±0.2mm. Despite the differences in techniques, these measurements are comparable with those in our study. A bifurcation IMA ending was observed in 93% of cases in the series of Henriquez-Pino et al.21 and 97.5% in that of Peric et al.28

Based on comparison of our anatomical data with those in the literature12, 13, 14 and on those from our previous study,29 extra-anatomical revascularization of the AA using the proposed technique should be possible in about 60% of cases. Preoperative location of the AA is an indispensable requirement for this technique. Improvements in spinal cord arteriographic techniques17 and especially the use of recent noninvasive methods such as three-dimensional magnetic resonance angiography30 and multi-detector row computed tomography31 have opened new perspectives in this domain.

This report presents the results of an exclusively anatomical feasibility study. Further anatomosurgical research and animal studies will be necessary to develop the surgical technique for clinical application.

Back to Article Outline

Conclusion 

This anatomical study demonstrates the feasibility of revascularization of the ICA in the paravertebral region using the IMA. This technique could be useful for preservation of spinal cord blood supply in patients undergoing endovascular repair of thoracic or thoracoabdominal aortic lesions.

Back to Article Outline

References 

  1. Svensson LG, Crawford ES, Hess KR, et al. Experience with 1509 patients undergoing thoracoabdominal aortic operations. J Vasc Surg. 1993;17:357–370
  2. Safi HJ, Miller CC, Carr C, et al. Importance of intercostal artery reattachment during thoracoabdominal aortic aneurysm repair. J Vasc Surg. 1998;27:58–66
  3. Estrera AL, Miller CC, Huynh TT, Porat E, Safi HJ. Neurologic outcome after thoracic and thoracoabdominal aortic aneurysm repair. Ann Thorac Surg. 2001;72:1225–1230
  4. Livesay JJ, Cooley DA, Ventemiglia RA, et al. Surgical experience in descending thoracic aneurysmectomy with and without adjuncts to avoid ischemia. Ann Thorac Surg. 1985;39:37–46
  5. Estrera AL, Rubenstein FS, Miller CC, et al. Descending thoracic aortic aneurysm: surgical approach and treatment using the adjuncts cerebrospinal fluid drainage and distal aortic perfusion. Ann Thorac Surg. 2001;72:481–486
  6. Mitchell RS, Miller DC, Dake MD. Stent-graft repair of thoracic aortic aneurysms. Semin Vasc Surg. 1997;10:257–271
  7. Gravereaux EC, Faries PL, Burks JA, et al. Risk of spinal cord ischemia after endograft repair of thoracic aortic. J Vasc Surg. 2001;34:997–1003
  8. Greenberg R, Resch T, Nyman U, et al. Endovascular repair of descending thoracic aortic aneurysms: an early experience with intermediate-term follow-up. J Vasc Surg. 2000;31:147–156
  9. Watanabe G, Misaki T, Kotoh K, et al. Bilateral thoracoscopic minimally invasive direct coronary artery bypass grafting using internal thoracic arteries. Ann Thorac Surg. 1998;65:1673–1675
  10. Mack MJ, Acuff T, Osborne J. Minimally invasive direct coronary artery bypass: technical considerations and instrumentation. J Card Surg. 1998;13:290–296
  11. Macierewicz JA, Jameel MM, Whitaker SC, et al. Endovascular repair of perisplanchnic abdominal aortic aneurysm with visceral vessel transposition. J Endovasc Ther. 2000;7:410–414
  12. Rodriguez-Baeza A, Muset-Lara A, Rodriguez-Pazos M, et al. The arterial supply of the human spinal cord: a new approach to the arteriaradicularis magna of Adamkiewicz. Acta Neurochir (Wien). 1991;109:57–62
  13. Alleyne CH, Cawley CM, Shengelaia GG, et al. Microsurgical anatomy of the artery of Adamkiewicz and its segmental artery. J Neurosurg. 1998;89:791–795
  14. Koshino T, Murakami G, Morishita K, et al. Does the Adamkiewicz artery originate from the larger segmental arteries? J. Thorac. Cardiovasc. Surg. 1999;117:898–905
  15. Williams GM, Perler BA, Burdick JF, et al. Angiographic localization of spinal cord blood supply and its relationship to postoperative paraplegia. J Vasc Surg. 1991;13:23–33
  16. Bert S, Iyriboz AT, Barret F, et al. An angiographic study of spinal vascularization at the thoracic-lumbar level. J Neuroradiol. 1995;22:12–19
  17. Kieffer E, Fukui S, Chiras J, et al. Spinal cord arteriography: a safe adjunct before descending thoracic or thoracoabdominal aortic aneurysmectomy. J Vasc Surg. 2002;35:262–268
  18. Larichev VB. A technique of retrograde catheterization of the posterior intercostal arteries. Klin Khir. 1990;12:58–59
  19. Kerrigan CL, Daniel RK. The intercostal flap: an anatomical and hemodynamic approach. Ann Plast Surg. 1979;2:411–421
  20. Donald PJ. Costal blood supply in free flap grafting. Arch Otolaryngol. 1984;110:99–102
  21. Henriquez-Pino JA, Gomes WJ, Prates JC, et al. Surgical anatomy of the internal thoracic artery. Ann Thorac Surg. 1997;64:1041–1045
  22. Lachman N, Satyapal KS. Morphometry of the internal thoracic arteries. Surg Radiol Anat. 1998;20:243–247
  23. Pietrasik K, Bakon L, Zdunek P, et al. Clinical anatomy of internal thoracic artery branches. Clin Anat. 1999;12:307–314
  24. Sutherland FW, Desai JB. Incidence and size of lateral costal artery in 103 patients. Ann Thorac Surg. 2000;69:1865–1866
  25. Owens WA, Gladstone DJ, Heylings DJ. Surgical anatomy of the phrenic nerve and internal mammary artery. Ann Thorac Surg. 1994;58:843–844
  26. Henriquez-Pino J, Mandiola-Lagunas E, Prates JC. Origin of the internal thoracic artery and its relationship to the phrenic nerves. Surg Radiol Anat. 1993;15:31–34
  27. Green GE. Use of internal thoracic artery for coronary artery grafting. Circulation. 1989;79:I30–I33
  28. Peric M, Sagic D, Miric M, et al. Angiographic characteristics of the internal thoracic artery—anatomic variations and their surgical importance. Med Pregl. 2000;53:245–249
  29. Malikor S, Rosset E, Paraskevas N, et al. Extraanatomical revascularization of the artery of Adamkiewicz: anatomical study. Ann Vasc Surg. 2002;16:723–729
  30. Hachiro Y, Kawaharada N, Morishita K, et al. Thoracoabdominal aortic aneurysm repair after detection of the Adamkiewicz artery by magnetic resonance angiography; a way to shorten operating time and improve outcome. Kyobu Geka. 2004;57:280–283
  31. Yoshioka K, Niinuma H, Ohira A, et al. MR angiography and CT angiography of the artery of Adamkiewicz: non-invasive preoperative assessment of thoracoabdominal aortic aneurysm. Radiographics. 2003;23:1215–1225

 Presented at the Sixteenth Annual Meeting of the French Language Vascular Surgery Society, Marseille, France, June 28, 2004.

PII: S0890-5096(09)00246-5

doi:10.1016/j.avsg.2009.09.001

Annals of Vascular Surgery
Volume 23, Issue 6 , Pages 758-763, November 2009

Article Tools

* Image Usage & Resolution