The branching pattern of pulmonary vessels varies

Transcription

1 GENERAL THORACIC Right Upper Lobe Venous Drainage Posterior to the Bronchus Intermedius: Preoperative Identification by Computed Tomography Katsuyuki Asai, MD, Norikazu Urabe, MD, PhD, Kiyoshige Yajima, MD, Kazuya Suzuki, MD, PhD, and Teruhisa Kazui, MD, PhD Department of Thoracic Surgery, Numazu City Hospital, Numazu, Shizuoka, First Department of Surgery, Hamamatsu University School of Medicine, Hamamatsu, Shizuoka, Japan Background. Anatomical variations of the pulmonary vessels relevant to pulmonary surgery are of concern to thoracic surgeons. Among such variations, the right upper lobe vein posterior to the bronchus intermedius (UVPBI) has received little attention thus far. Methods. Chest computed tomographic images and medical records of 725 patients were retrospectively reviewed. The frequency, drainage pattern, diameter, and associated anatomical characteristics of the UVPBI were assessed, and our right thoracotomy cases with the UVPBI were examined. Results. The UVPBI was found in 41 (5.7%) of 725 computed tomography cases, and in 9 (3.9%) of 230 right thoracotomy cases. Three UVPBI drainage sites were observed: (1) the superior pulmonary vein group, 55%; (2) the inferior pulmonary vein group, 41%; and (3) the superior segmental vein group, 4%. The diameter of the UVPBI at the level of the bronchus intermedius ranged from 1 to 7 mm ( mm). The diameter of the UVPBI in the superior pulmonary vein group was significantly greater than that in the inferior pulmonary vein group (p < 0.01). The prevalence of a central vein was 43.9% for all UVPBI cases and 15.8% for large UVPBI cases (> 5 mm in diameter). Of the 9 right thoracotomy patients, 1 suffered UVPBI injury; this patient s UVPBI was not identified either preoperatively or intraoperatively. Conclusions. The UVPBI is not as rare as was previously believed. It can be a main drainage route of the right upper lobe. Preoperative identification of this venous variation by computed tomography is useful for safe and accurate surgical procedures. (Ann Thorac Surg 2005;79: ) 2005 by The Society of Thoracic Surgeons The branching pattern of pulmonary vessels varies widely, and some patterns associated with a potential risk of intraoperative bleeding can only be identified at surgery. Preoperative foreknowledge of individual vessel configurations is thus useful for ensuring a safe and accurate procedure during pulmonary surgery. Few reports have discussed anatomical variations of the right upper lobe vein posterior to the bronchus intermedius (UVPBI) relevant to pulmonary resection [1, 2]. Recent advances in noninvasive imaging methods such as multi-detector computed tomography and magnetic resonance angiography have facilitated the detection of pulmonary vessels [3, 4]. Because of its unique anatomical location, UVPBI can be identified easily by chest computed tomography (CT). The purpose of this study was to determine the frequency of UVPBI and to evaluate its associated anatomical characteristics on CT, the most familiar imaging method in our daily clinical practice. Patients and Methods Chest computed tomographic images and medical records of 780 consecutive patients who had undergone a chest CT Accepted for publication Dec 20, Address reprint requests to Dr Asai, Department of Thoracic Surgery, Numazu City Hospital, 550 Harunoki Higashi-shiiji, Numazu, Shizuoka, Japan; asaik@msc.biglobe.ne.jp. between January 1997 and March 2004 at the Department of Thoracic Surgery at Numazu City Hospital were retrospectively reviewed. We used our institution s routine protocol to analyze these computed tomographic images in a lung window setting. Computed tomographic scans were performed with either a helical scanner (ProSeed [GE Yokogawa Medical Systems, Tokyo, Japan]) or a conventional scanner (Quantex [GE Yokogawa Medical Systems]). Computed tomographic images were obtained with either 5-mm or 10-mm collimation and 10-mm intervals. Additional contrast-enhanced scanning was performed as needed. Computed tomographic images were analyzed by two thoracic surgeons (KA, NU), and a final interpretation of the findings was reached by consensus. The CT analysis for UVPBI identification consisted of the following three processes: (1) evaluation of the presence of nodular opacities attached posterior to the bronchus intermedius; (2) confirmation that the nodular opacities were veins by tracing the continuity of the nodule both cranially and caudally on contiguous CT slices; and (3) determination as to which lobe from which the vein drained (ie, the right upper or lower lobe). Depending on the drainage lobe, the vein was classified as either a UVPBI or as a right lower lobe vein posterior to the bronchus intermedius (LVPBI). The LVPBI was regarded as a branch of the right superior segmental vein (V 6 ). The UVPBI was rigorously limited to the pulmonary vein, 2005 by The Society of Thoracic Surgeons /05/$30.00 Published by Elsevier Inc doi: /j.athoracsur

2 Ann Thorac Surg ASAI ET AL 2005;79: CT IDENTIFICATION OF RIGHT PULMONARY VEIN VARIATION which could apparently be followed peripherally into the lung parenchyma of the right upper lobe. Furthermore, after identifying the UVPBI, we grouped them according to their courses towards the left atrium (LA) beyond the posterior wall of the bronchus intermedius. We evaluated the frequency of the UVPBI both in the whole population and in patients who had undergone a right thoracotomy, and we also measured the diameter of the UVPBI in each course group. In UVPBI cases, paying special attention to the fact that the central vein was prevalent in the right upper lobe, which was easily identified by CT, we checked for the presence of a central vein to determine the influence of the UVPBI on the whole drainage pattern in the right upper lobe, to the extent that we could determine this from CT study. The diameter of the UVPBI was measured at the level posterior to the bronchus intermedius. Whenever the nodular opacity was not round, the short-axis diameter was used for measurements. When the bronchus intermedius and the attached nodular opacities were observed in two or more contiguous CT slices, the maximum diameter was used. From a total of 780 computed tomographic images, 55 were excluded because of the absence of normal lung parenchyma necessary to identify anatomical structures in the relevant area. Thus, the study consisted of 725 eligible computed tomographic examinations (457 men, 268 women; age range, 6 to 89 years; mean age, 60 years). The reasons for exclusions were as follows: right pneumothorax (19), pleural effusion (3), tumor (11), inflammation (7), bulla (4), contusion (5), previous right-side pulmonary resection (4), right-sided thoracic aorta (1), and dextrocardia (1). Of the total 725 patients, 422 (58.2%) underwent surgery, 230 (31.7%) had a right thoracotomy (including surgical procedures by thoracoscopy and median sternotomy), and 137 (18.9%) had a right-sided major pulmonary resection (segmentectomy, lobectomy, and pneumonectomy). All vessel diameter measurements were recorded as mean standard deviation. Student s t test and Fisher s exact probability test were used for comparisons between groups. A p value of less than 0.05 was considered statistically significant. Results Abbreviations and Acronyms CT computed tomography IPV inferior pulmonary vein LA left atrium LVPBI right lower lobe vein posterior to the bronchus intermedius SPV superior pulmonary vein UVPBI right upper lobe vein posterior to the bronchus intermedius VATS video-assisted thoracic surgery V 6 superior segmental vein In all 725 eligible CT subjects, the bronchus intermedius and its posterior wall could be easily identified. Nodular opacities attached to the posterior wall of the bronchus intermedius were observed in 63 patients (8.7%). All the nodular opacities were identified as branches of pulmonary veins. Of the 725 computed tomographic images, the UVPBI was found in 41 (5.7%) and the LVPBI in 22 (3.0%). With respect to surgical findings, the UVPBI was found in 9 (3.9%) of 230 right thoracotomy cases, and in 6 (4.4%) of 137 cases of right major pulmonary resections. UVPBI was broadly classified according to its course towards the LA beyond the posterior wall of the bronchus intermedius. The superior pulmonary vein (SPV) group included cases in which the UVPBI ran horizontally into the mediastinum. In the remaining UVPBI cases, the UVPBI ran vertically into the inferior pulmonary vein (IPV) or V 6 while in contact with the posterior wall of the bronchus intermedius and the lower lobe bronchus. These cases were classified into two groups: (1) the IPV group (when the UVPBI joined the inferior pulmonary vein), and (2) the V 6 group (when a common trunk of UVPBI and V 6 joined the LA independently). Thus, the UVPBI cases comprised three drainage variation groups: (1) SPV group, (2) IPV group, and (3) V 6 group (Fig 1). Concerning the SPV group, since the exact course of the UVPBI within the mediastinum could not be detected by the lung window setting CT study, that portion of its course had to be deduced and supported by our limited contrast-enhanced, thin-section CT study, previous studies, and case reports. Anatomical schemata of the SPV, IPV, and V 6 groups and their diameters at the level posterior to the bronchus intermedius are shown in Fig 2. Among the UVPBI, the vessel diameter was significantly larger in the SPV group than that in the IPV group (p 0.01). The prevalence of the central vein in each of the UVPBI groups is shown in Table 1. In UVPBI cases, the prevalence of the central vein was not generally high. The prevalence of a central vein was significantly lower for cases with a large UVPBI ( 5 mm in diameter) than for cases with a small UVPBI ( 5 mm in diameter) (p 0.01). In particular, in large UVPBI cases within the SPV group, the prevalence of a central vein was only 8.3%. All UVPBI cases were drainage veins from the posterior segment. Nine patients with the UVPBI underwent a right thoracotomy (Fig 3). Clinical details of these patients are shown in Table 2. Intraoperative injury of the UVPBI occurred in one case from an electrocautery while a fissure between the right upper and lower lobe was being dissected during a lower lobectomy. In this case, the UVPBI was not identified preoperatively or intraoperatively. Comment 1867 Numerous anatomical variations of pulmonary arteries, veins and bronchi and their interrelations have been observed. In particular, pulmonary veins have great variability in their branching patterns [5 7]. Some vessel variations affect surgical procedures; the most familiar variations involve the middle lobe or lingular artery branching from the basal artery, the mediastinal branch of the lingular artery, the middle lobe vein into the IPV or GENERAL THORACIC

3 GENERAL THORACIC 1868 ASAI ET AL Ann Thorac Surg CT IDENTIFICATION OF RIGHT PULMONARY VEIN VARIATION 2005;79: Fig 1. Computed tomographic findings of each drainage course of the UVPBI (arrowheads): (A) superior pulmonary vein group, (B) inferior pulmonary vein (IPV) group, and (C) superior segmental vein (V 6 ) group. (BSV basal segmental vein; UVPBI right upper lobe vein posterior to the bronchus intermedius.) LA, and the common pulmonary vein [8]. The UVPBI is a less familiar variation, and its frequency is not well known. Thus far only a handful of studies and discussions have focused on the UVPBI [9 12]. Identifying pulmonary vessels by CT, especially at the hila, can sometimes be very difficult because of their complex anatomy. While three-dimensional images can more clearly depict pulmonary vessels [13], CT can also identify pulmonary vessels in certain areas. For instance, CT can clearly depict the posterior wall of the bronchus intermedius, sharply outlined by air in the bronchial lumen and lungs, and usually no other structure is behind it, except lung parenchyma. When CT visualizes a nodular opacity behind the tangentially depicted posterior wall of the bronchus intermedius, the UVPBI or LVPBI can be identified without much difficulty by careful tracing of the continuity of the nodule cranially and caudally. Before the advent of CT, Boyden [14] in 1955 and Yamashita [15] in 1978 described the details of the segmental anatomy of the lung in their textbooks. In his description of 50 cadaver specimens, Boyden [14] reported one case (2%) of segmental vein 2 (V 2 ) draining into the IPV, which he named the posterior segment of the right upper lung (S 3 ); therefore he actually called the vein V 3, not V 2, in his textbook) [14]. Similarly, Yamashita [15] stated that the frequency of V 2 draining into the IPV was 2.4% (4 cases out of 170 cadaver specimens). The frequency of UVPBI in the IPV group reported in these studies agrees with our data. However, neither author mentioned the drainage pattern into the SPV and the LA, posterior to the bronchus intermedius. Computed tomography studies of the UVPBI consist of only a few series [9 11]. In the largest one, Kim and colleagues [9] analyzed nodularity in the posterior wall of the bronchus intermedius on 280 normal CT scans. According to their report, the frequency of UVPBI was 3.6%, whereas that of LVPBI was 1.7%. Thus, the overall prev-

4 Ann Thorac Surg ASAI ET AL 2005;79: CT IDENTIFICATION OF RIGHT PULMONARY VEIN VARIATION 1869 GENERAL THORACIC Fig 2. Anatomical schema of each group: superior pulmonary vein (SPV) group, inferior pulmonary vein (IPV) group, and superior segmental vein (V 6 ) group. (AV azygos vein; BI bronchus intermedius; BSV basal segmental vein; E esophagus; LA left atrium; PA pulmonary artery; SVC superior vena cava; ULB upper lobe bronchus; UVPBI right upper lobe vein posterior to the bronchus intermedius.) alence of pulmonary veins posterior to the bronchus intermedius was 5.3%. In addition, with respect to the drainage pattern of the UVPBI, they reported that 30% (7 of 10) drained into the SPV and 70% (3 of 10) drained into the IPV. Jardin and Remy [11] reported that the frequency of the UVPBI in the IPV group was 9.3% (10 cases from 107 CT images of the right lower lobe). In all the reported cases in which the UVPBI was surgically confirmed, the preoperative CT also clearly detected the UVPBI [1, 2, 16]. Venous drainage patterns of the right upper lobe are usually classified as follows according to the degree of presence of a central vein: central vein type (V 1 and V 2 3 ), semi-central vein type (V 1 2 and V 3 ), or noncentral vein type. Of these, the central vein type is most common, with a reported frequency of 78.3% to 88% [15, 17]. The central vein is easily identified by routine CT. If the central vein is present, it is usually located tangentially within the angle formed by the bifurcation of the right upper lobe bronchus into B 2 and B 3. The CT level of this bifurcation is the key slice for identifying the central vein. In our study, the presence of UVPBI, especially the Table 1. The Prevalence of the Central Vein in Cases of UVPBI UVPBI (Diameter) n Central Vein (%) Total UVPBI Total (43.9) ( 5 mm) 19 3 (15.8) ( 5 mm) (68.2) SPV group Total 22 7 (31.8) ( 5 mm) 12 1 (8.3) ( 5 mm) 10 6 (60.0) IPV group Total (58.8) ( 5 mm) 5 1 (20.0) ( 5 mm) 12 9 (75.0) V 6 group Total 2 1 (50.0) ( 5 mm) 2 1 (50.0) IPV inferior pulmonary vein; SPV superior pulmonary vein; UVPBI right upper lobe vein posterior to the bronchus intermedius; V 6 superior segmental vein. Fig 3. Operative view after a right lower lobectomy in a patient with a right upper lobe vein posterior to the bronchus intermedius (UVPBI) (case 8 [see Table 2], superior pulmonary vein group). (BI bronchus intermedius; Bs bronchial stump; PAs pulmonary arterial stumps; UL upper lobe.)

5 GENERAL THORACIC 1870 ASAI ET AL Ann Thorac Surg CT IDENTIFICATION OF RIGHT PULMONARY VEIN VARIATION 2005;79: Table 2. Right Thoracotomy Cases of UVPBI Case Age/Gender Diagnosis Operation Approach UVPBI Type 1 71/F IN PR (S 4 ) VATS IPV 2 70/F LC ML PL SPV 3 44/M PTx PR (S 1 ) VATS IPV 4 52/M LC ML PL SPV 5 60/M LC LL PL SPV 7 61/M LC UL PL IPV 6 82/M LC UL PL SPV 8 72/F LC LL PL SPV 9 78/M LC PR (S 3 ) Ax IPV Diameter (mm) Preoperative Identification Intraoperative Identification UVPBI Injury UVPBI Treatment 5 no no no - 4 no no no - 1 no no no - 5 no no no - 3 no no yes divided 2 no no no - 6 yes yes no divided 5 yes yes no preserved 6 yes no no - Data are presented in chronological order. Ax axillary thoracotomy; IN inflammatory nodule; LC lung cancer; LL lower lobectomy; ML middle lobectomy; PL posterolateral thoracotomy; PR partial resection; PTx pneumothorax; UL upper lobectomy; UVPBI right upper lobe vein posterior to the bronchus intermedius; VATS video-assisted thoracic surgery. presence of a large UVPBI with drainage to the SPV, was significantly associated with a reduced frequency of a central vein. These data suggest that some of the UVPBI are the dominant drainage route of the right upper lobe, especially of the posterior segment, and that a venous drainage pattern containing a large UVPBI represents another independent type. The most probable operative procedure causing UVPBI injury is the dissection of the posterior aspect of the major fissure. Dissection of the subcarinal, right hilar, or interlobar lymph nodes can also cause UVPBI injury. These situations primarily involve right upper or lower lobectomies. Furthermore, because the UVPBI runs just inside the mediastinal visceral pleura, if the pleural space cannot be identified due to adhesions between the visceral and parietal pleurae, special care must be taken when dissecting the adhesion. Recently, video-assisted thoracic surgery lobectomy has been gaining popularity as a potential alternative to conventional thoracotomy for early stage lung cancer [18, 19]. Vascular injury within such minimally invasive approaches would make hemostasis difficult and force the surgeon to convert to a larger thoracotomy. In particular, massive bleeding from an unknown origin could rapidly lead to a critical situation. The UVPBI and some other vessel variations may thus pose a great danger during video-assisted thoracic surgery. In addition, during median sternotomy or an anterior thoracotomy, because the visibility around the UVPBI is poor, careful manipulation is required. In most cases, dividing the UVPBI first may make subsequent surgical maneuvers safer and easier. However, in cases of a large UVPBI in the SPV group, and especially with a limited pulmonary reserve, preserving the UVPBI may be preferable during a right lower lobectomy. We found two of the UVPBI in the V 6 group on CT. To our knowledge, this drainage variation has only been reported by Maciejewski [20]. For this group, he emphasized that V 6 and V 2 must be ligated separately during lobectomy. The importance of the UVPBI during esophageal resection as well as pulmonary resection, has been recognized [16]. Our study has several limitations. First, unlike thinsection CT, it is not always possible to completely discriminate between the upper lobe and the lower lobe on routine computed tomographic images, even if the lung has a complete interlobar fissure [21]. In the present study, we limited the UVPBI strictly to those veins that were apparently derived from the upper lobe. Thus, the true frequency of the UVPBI might actually be higher than our estimation because some of the veins we categorized as LVPBI might actually have been UVPBI. Second, two scan protocols were utilized in our CT examination: 5-mm collimation at 10-mm intervals, or 10-mm collimation at 10-mm intervals. The first of these two protocols might have made it difficult to detect the horizontal component of the UVPBI, especially in the SPV group, because of the 5-mm gap between two contiguous sections. Finally, because we analyzed computed tomographic images only at lung window settings,

6 Ann Thorac Surg ASAI ET AL 2005;79: CT IDENTIFICATION OF RIGHT PULMONARY VEIN VARIATION we could not visualize the exact courses of the UVPBI within the mediastinum. Thus, our SPV group might also include another group in which the UVPBI drains into the LA independently, a drainage course which Spaggiari and colleagues [1] have previously reported. Determination of a more detailed course of the UVPBI within the mediastinum, especially in the SPV group, would require additional contrast-enhanced CT studies. In conclusion, the UVPBI is not as infrequent as was previously believed. It can be the main drainage route of the right upper lobe. Preoperative identification of this variation is useful for decreasing the incidence of unexpected intraoperative bleeding. Computed tomography is an effective means of identifying the UVPBI and anticipating its course to the LA. A somewhat greater awareness of anatomical variations while interpreting CT studies may make pulmonary resection safer and more anatomically accurate. We thank Dr Hajime Fujimoto from the Department of Radiology, Numazu City Hospital, for his valuable suggestions; and Dr Koichi Mori from the Department of Radiology, Tsuchiura Kyodo General Hospital, for the motivation for this study. References 1. Spaggiari L, Solli P, Leo F, Veronesi G, Pastonno U. Anomalous segmental vein for right upper lobe: an unusual anatomical variation. Ann Thorac Surg 2002;74: Tsuboi M, Asamura H, Naruke T, Nakayama H, Kondo H, Tsuchiya R. A VATS lobectomy for lung cancer in a patient with an anomalous pulmonary vein: report of a case. Surg Today 1997;27: Schoepf UJ, Costello P. CT angiography for diagnosis of pulmonary embolism: state of the art. Radiology 2004;230: Vrachliotis TG, Bis KG, Shetty AN, Ravikrishan KP. Contrastenhanced three-dimensional MR angiography of the pulmonary vascular tree. Int J Cardiovasc Imaging 2002;18: Yazar F, Ozdogmus O, Tuccar E, Bayramoglu A, Ozan H. Drainage pattern of middle lobe vein of right lung: an anatomical study. Eur J Cardiothorac Surg 2002;22: Kato R, Lickfett L, Meininger G, et al. Pulmonary vein anatomy in patients undergoing catheter ablation of atrial fibrillation: lessons learned by use of magnetic resonance imaging. Circulation 2003;107: Marom EM, Herndon JE, Kim YH, et al. Variations in pulmonary venous drainage to the left atrium: implications for radiofrequency ablation. Radiology 2004;230: Sugimoto S, Izumiyama O, Yamashita A, Baba M, Hasegawa T. Anatomy of inferior pulmonary vein should be clarified in lower lobectomy. Ann Thorac Surg 1998;66: Kim JS, Choi D, Lee KS. CT of the bronchus intermedius: frequency and cause of a nodule in the posterior wall on normal scans. AJR 1995;165: Webb WR, Hirji M, Gamsu G. Posterior wall of the bronchus intermedius: radiographic CT correlation. AJR 1984;141: Jardin M, Remy J. Segmental bronchovascular anatomy of the lower lobes: CT analysis. AJR 1986;147: Austin JH. Right upper lobe venous drainage posterior to the bronchus intermedius. Ann Thorac Surg 2003;75: Watanabe S, Arai K, Watanabe T, Koda W, Urayama H. Use of three-dimensional computed tomographic angiography of pulmonary vessels for lung resections. Ann Thorac Surg 2003;75: Boyden EA. In: Boyden, ed. Segmental anatomy of the lungs. New York: McGraw-Hill, 1955: Yamashita H. Variations in the pulmonary segments and the bronchovascular trees. In: Yamashita H, ed. Roentgenologic anatomy of the lung. Tokyo: Igaku-shoin, 1978: Matsubara T. Rare but dangerous anomaly of the right pulmonary vein in subcarinal dissection. Ann Thorac Surg 2003;75: Lee KS, Bae WK, Lee BH, Kim IY, Choi EW, Lee BH. Bronchovascular anatomy of the upper lobes: evaluation with thin section CT. Radiology 1991;181: Gharagozloo F, Tempesta B, Margolis M, Alexander EP. Video-assisted thoracic surgery lobectomy for stage I lung cancer. Ann Thorac Surg 2003;76: Walker WS, Codispoti M, Soon SY, Stamenkovic S, Carnochan F, Pugh G. Long-term outcomes following VATS lobectomy for non-small cell bronchogenic carcinoma. Eur J Cardiothorac Surg 2003;23: Maciejewski R. The venous drainage of the apical segment of the right lower pulmonary lobe. Acta Anat 1994;150: Glazer HS, Anderson DJ, DiCroce JJ, et al. Anatomy of the major fissure: evaluation with standard and thin-section CT. Radiology 1991;180: GENERAL THORACIC