- Email: [email protected]
Observation of thoracic duct morphology in portal hypertension by endoscopic ultrasound
Observation of thoracic duct morphology in portal hypertension by endoscopic ultrasound Vinod K. Parasher, MD, Emanuele Meroni, MD, Alberto Malesci, MD, Pasquale Spinelli, MD, Maurizio A. Tommasini, MD, Ronald Markert, PhD, Manoop S. Bhutani, MD Lewes, Delaware, Milan, Italy, Dayton, Ohio
Background: Thoracic duct dilation has been demonstrated in portal hypertension and hepatic cirrhosis by lymphangiography and laparotomy and at autopsy. It is thought to be secondary to increased hepatic lymph flow and has been described in the absence of ascites or esophageal varices. The aim of the present study was to observe thoracic duct morphology by endoscopic ultrasound in various subsets of patients with portal hypertension and hepatic cirrhosis and also to validate existing radiologic/surgical data. Methods: The thoracic duct of 33 patients with cirrhosis and portal hypertension was studied by endoscopic ultrasound. Patients were divided into four groups: 1, patients with ascites and esophageal varices; 2, esophageal varices without ascites; 3, without esophageal varices or ascites; 4, extrahepatic portal hypertension due to pancreatic malignancy. The thoracic duct diameter was also measured in 14 control subjects (group 5). Results: When the thoracic duct diameter for the five groups was compared with analysis of variance, significance was p < 0.0001; by pairwise comparison, group 1 differed from the other four groups (p < 0.05). Thoracic duct dilation (5.61 mm) was seen in group 1 patients, whereas no dilation was present in groups 2 through 4. Additionally, thoracic duct diameter in 33 portal hypertensive and/or cirrhotic patients was significantly different from that in the 14 control subjects (p = 0.003). Conclusion: The thoracic duct can be reliably identified by EUS in patients with hepatic cirrhosis and portal hypertension. Dilation of the duct is seen only in patients with hepatic cirrhosis, ascites, and esophageal varices. No thoracic duct dilation is present in extrahepatic portal hypertension. Contrary to existing radiologic/surgical data, thoracic duct dilation is not seen in all patients with hepatic cirrhosis and portal hypertension signifying advanced disease. A dilated thoracic duct by endoscopic ultrasound should be considered yet another sign of portal hypertension. (Gastrointest Endosc 1998;48:588-92.)
Increased formation of lymph which contains an abundance of formed elements is characteristic of hepatic cirrhosis with portal hypertension. These formed elements consist of intact red blood cells and protein. The increase in hepatic lymph flow is due to Received June 25, 1996. For revision August 30, 1996. Accepted June 4, 1998. From the Beebe Medical Center, Lewes, Delaware; Istituto Clinico Humanitas and Nazionale Tumori, Milan, Italy; Wright State University, Dayton, Ohio. Reprint requests: Vinod K. Parasher, MD, Director of Endoscopy, Beebe Medical Center, 424 Savannah Rd., Lewes, DE 19958. Copyright © 1998 by the American Society for Gastrointestinal Endoscopy 0016-5107/98/$5.00 + 0 37/1/92175 588
GASTROINTESTINAL ENDOSCOPY
the obstruction of the hepatic venules as a result of cirrhosis.1-5 The normal flow of hepatic lymph is toward the hilum, and there the lymphatics traverse the hepatoduodenal and hepatogastric ligaments to join the cisterna chyli and the thoracic duct. In portal hypertension, the increased lymph flow causes dilation of the hilar lymphatics and the thoracic duct1,3,6 as shown by lymphangiography2 and postmortem studies and at laparotomy.6 It is noteworthy that the dilation of the thoracic duct has been noted even in the absence of varices and ascites,2,6 suggesting that thoracic duct dilation may be an early sign of portal hypertension. Unfortunately, there are as yet no noninvasive methods to study the thoracic duct. Endoscopic VOLUME 48, NO. 6, 1998
Thoracic duct morphology in portal hypertension by EUS
V Parasher, E Meroni, A Malesci, et al.
Figure 1. EUS demonstrates nondilated thoracic duct in patient with hepatic cirrhosis without ascites or esophageal varices (frequency 7.5 MHz, range 6 cm). Arrowhead, Thoracic duct; large arrow, azygos vein; small arrow, aorta.
Figure 2. EUS demonstrates dilated thoracic duct lying between the azygous vein and aorta in a patient with ascites and esophageal varices (frequency 7.5 MHz, range 4 cm).
ultrasound (EUS) is a sensitive imaging technique that is increasingly being used to study intraluminal and extraluminal disorders of the gastrointestinal tract. More recently, we have described the normal anatomy of the thoracic duct by EUS, thereby establishing a semi-invasive method to observe this structure. The thoracic duct was identified as an anechoic 3 to 4 mm structure that is best seen near the mid-esophagus.7 The aim of our study was to answer the following questions: (1) Can the thoracic duct be reliably identified in patients with hepatic cirrhosis? (2) Can EUS validate existing radiologic/surgical data that indicate that the thoracic duct is dilated in intrahepatic portal hypertension? (3) Is the thoracic duct dilated in all cases of hepatic cirrhosis, irrespective of the presence of ascites or varices?
Ascites was diagnosed by clinical examination, transcutaneous ultrasound, EUS, or CT. Esophageal varices were diagnosed by endoscopy and EUS. Patients with active variceal bleeding or a recent history of bleeding (less than 1 month), hepatic encephalopathy, or severe ascites (causing respiratory distress) were excluded from the study. After examining the stomach, the echoendoscope was withdrawn gradually into the esophagus. The balloon was inflated with deaerated water and the thoracic duct was examined in a systematic fashion. Thoracic duct diameter was recorded at its widest part. The diameter of the thoracic duct was also measured in 14 normal control subjects (group 5) undergoing EUS for other indications. There were 10 men and 4 women ranging in age from 42 to 81 years (mean 53.6 years). The differentiation between the thoracic duct and varices was made by following the thoracic duct inferiorly and superiorly over the length of the esophagus. The thoracic duct generally extended longitudinally as a continuous anechoic structure, whereas periesophageal varices were often multiple, noncontinuous, and serpiginous.
PATIENTS AND METHODS Thirty-three patients with biopsy-proven cirrhosis (etiology included alcohol, hepatitis B virus, and hepatitis C virus) were studied prospectively. The study was approved by our institutional review board. EUS was performed with the GF UM20 (Olympus America, Inc., Melville, N.Y.) echogastroscope. There were 28 men and 5 women. The ages ranged from 45 to 73 years (mean 58.5 years). Ten patients had ascites and esophageal varices (group 1), 8 had esophageal varices without ascites (group 2), and 10 had neither ascites nor esophageal varices (group 3). Five patients who had pancreatic malignancy with varices due to invasion of the portal system were also studied as a group with extrahepatic portal hypertension (group 4). Ascites was not present in these patients, but all had gastric varices and portal vein obstruction. VOLUME 48, NO. 6, 1998
Statistical analysis Mean thoracic duct diameters among the 5 groups were compared by analysis of variance (ANOVA) and Tukey’s multiple range test for pairwise differences. The t-test was used to compare patients with portal hypertension and/or cirrhosis with control subjects. Inferences were made at the 0.05 level of significance.
RESULTS The thoracic duct was seen in all patients as an anechoic thin-walled round structure. The duct was more clearly visualized in the mid-esophagus where the esophagus was anterior to it, the spine posteriGASTROINTESTINAL ENDOSCOPY
589
V Parasher, E Meroni, A Malesci, et al.
Thoracic duct morphology in portal hypertension by EUS
Table 1. Thoracic duct diameter in patient groups and control subjects Group 1
Group 2
Group 3
Group 4
Patients
TDD (mm)
Patients
TDD (mm)
Patients
TDD (mm)
Patients
01. 02. 03. 04. 05. 06. 07. 08. 09. 10.
5.0 5.1 6.0 5.4 7.0 5.3 5.0 6.2 5.1 6.0
01. 02. 03. 04. 05. 06. 07. 08.
3.0 4.0 2.3 3.0 3.0 2.1 2.4 3.0
01. 02. 03. 04. 05. 06. 07. 08. 09. 10.
1.0 2.0 2.3 3.1 1.4 2.4 2.5 3.2 3.0 3.5
01. 02. 03. 04. 05.
Mean 5.61 mm 95% CI [5.13, 6.09] Range 5.0-7.0 mm
Mean 2.85 mm 95% CI [2.35, 3.35] Range 2.1-4.0 mm
Mean 2.44 mm 95% CI [1.86, 3.02] Range 1-3.5 mm
Group 5
TDD (mm)
Patients
2.0 2.0 3.3 3.1 2.5
TDD (mm)
01. 1.5 02. 2.0 03. 3.0 04. 4.0 05. 2.1 06. 1.5 07. 3.0 08. 2.4 09. 2.0 10. 2.0 11. 2.1 12. 2.4 13. 4.0 14. 2.0 Mean 2.43 mm 95% CI [1.97, 2.89] Range 1.5-4.0 mm
Mean 2.58 mm 95% CI [1.83, 3.33] Range 2.0-3.3 mm
Groups are as follows: 1, cirrhotic patients with ascites and varices; 2, cirrhotic patients with varices only; 3, cirrhotic patients without varices or ascites; 4, patients with extrahepatic portal hypertension; 5, represents normal control subjects. TDD, Thoracic duct diameter.
or, the azygos vein to its right, and the aorta to the left side of the duct. A single duct was visualized in all cases. The diameter of the thoracic duct in 14 control subjects (group 5) ranged from 1.5 to 4.0 mm (mean 2.43 mm, Fig.1). Thoracic duct dilation was noted only in patients with ascites and esophageal varices (group 1). Diameter of the thoracic duct ranged from 5 to 7 mm (mean 5.61 mm) in this group (Fig. 2). The thoracic duct was not dilated in cirrhotic patients with esophageal varices who did not have ascites (group 2). Here the diameter ranged from 2.1 to 4.0 mm (mean 2.85 mm). Furthermore, the thoracic duct was also not dilated in cirrhotic patients with neither esophageal varices nor ascites (group 3); the diameter ranged from 1 to 3.5 mm (mean 2.44 mm). Also, the duct was not dilated in patients with extrahepatic portal hypertension (group 4); the diameter ranged from 2.0 to 3.3 mm (mean 2.58 mm). No complications occurred during EUS. When the thoracic duct diameter for the five groups (Table 1) was compared with analysis of variance (ANOVA), significance was p < 0.0001. Furthermore, pairwise comparisons with Tukey’s multiple range test found that group 1 was significantly different (p < 0.05) from all other groups. No other pairwise comparisons were significant. When all 33 patients were compared with control subjects, the mean thoracic duct diameter (3.52 mm) was significantly different from that for control subjects (2.43 mm, p = 0.003). 590
GASTROINTESTINAL ENDOSCOPY
DISCUSSION The main function of the lymphatic system is the collection and transport of large molecules consisting of plasma proteins, particulate matter, tissue debris, bacteria, and foreign substances.8 The liver lymph originates from the interstitial fluid in the intercellular spaces of Disse. From here it extends to the limiting plates and then to the portal spaces. Eighty percent of the hepatic lymph passes through the hilar lymphatics to the cisterna chyli and into the thoracic duct. Under normal circumstances, half of the lymph flowing through the thoracic duct originates in the liver and half in the gastrointestinal tract.9 Hepatic hilar lymphatics are distended in patients with portal hypertension and discharge large volumes of lymph when transected. The thoracic duct is greatly enlarged, and lymph flow through the thoracic duct is often 10 to 15 times the normal flow rate.1,3,6 The distended hilar encapsular lymphatics reflect the hepatic venous outflow obstruction and increased hepatic lymph formation. Decompression of the thoracic duct either externally or by a shunt may ameliorate the portal pressure, ascites, and esophageal varices.3 Indeed, a peritoneal venous shunt may be considered as an artificial thoracic duct. Thoracic duct dilation has been shown in hepatic cirrhosis at autopsy and laparotomy and by lymphangiography.2-6 Additionally, thoracic duct dilation has also been demonstrated in the setting of hepatic cirrhosis without ascites and also in cases of VOLUME 48, NO. 6, 1998
Thoracic duct morphology in portal hypertension by EUS
hepatic cirrhosis without ascites or esophageal varices. Schieber2 demonstrated a dilated thoracic duct (diameter 6 to 9 mm, mean 7.5 mm) in 5 patients who had no ascites or major variceal bleeding. Normal duct diameter by lymphangiography is reported to be 1 to 4 mm (mean 2.5 mm). The thoracic duct was not only dilated but was also increased in length, probably because of the increased tortuosity. Dumont and Mulholland6 showed similar findings in one patient who had neither ascites nor varices. We have demonstrated that the thoracic duct can be identified by EUS. It appears as an anechoic structure, 3 to 4 mm in diameter, best seen close to mid-esophagus.7 EUS has been used to evaluate patients with portal hypertension; findings include esophageal and periesophageal varices, gastric and perigastric varices, dilated azygos veins, and small dilated vessels in the gastric wall in patients with portal hypertensive gastropathy and ascites.10,11 This is the first study of thoracic duct by EUS in patients with portal hypertension. The current study demonstrates that the thoracic duct can be reliably identified in patients with cirrhosis; furthermore, thoracic duct dilation is limited to patients who appear to have advanced disease with coexistent ascites and varices. Thoracic duct dilation was not identified in patients who had no ascites or varices. A dilated thoracic duct in the setting of hepatic cirrhosis and ascites is yet another sign of portal hypertension identifiable by EUS. Because of the relatively small sample size in this study, there is a possibility of beta error. The study may lack the statistical power to detect differences in patients with less severe disease. Because of multicenter nature of the study, interobserver variability may exist. However, we found group 1 to be different from each of the other four groups. Larger sample sizes might eventually result in statistically significant differences among the other four groups. If the mean value for groups 1 through 4 remained about the same, the small differences would not be clinically significant. Our study confirms the previous radiologic/surgical data indicating that the thoracic duct is dilated in hepatic cirrhosis; however, it does not confirm that dilation is seen in all cases of hepatic cirrhosis and portal hypertension. The reason for this discrepancy is not known. It can be hypothesized that lymphangiography may alter lymphatic dynamics by the injection of contrast under pressure. This may cause spurious duct dilation.12 It is also known that cannulation of the thoracic duct so as to block the outflow may cause higher pressure.8,13 The length of the thoracic duct was not determined in VOLUME 48, NO. 6, 1998
V Parasher, E Meroni, A Malesci, et al.
our study because the anatomic length cannot be accurately measured by EUS. Periesophageal varices may cause confusion; however, we found that by following the thoracic duct longitudinally in a dynamic fashion, this problem could be avoided. There has been no study of thoracic duct diameter in extrahepatic portal hypertension. We did not find a dilated thoracic duct in a selected group of patients with extrahepatic portal hypertension; however, none of these patients had ascites. More patients with other causes of extrahepatic portal hypertension need to be studied before conclusions can be drawn concerning thoracic duct diameter in this group of patients. However, studies done in patients with presinusoidal obstruction,5 particularly in portal vein constriction, have shown that there is no additional increase in thoracic duct lymph flow and pressure. The increased production of lymph in the small intestine is probably balanced by a decrease in production of lymph in the liver caused by a reduction in hepatic sinusoidol pressure and decreased sinusoidol perfusion by portal venous blood. In conclusion, EUS will reveal a dilated thoracic duct in cirrhotic patients with ascites and esophageal varices. In our study, the endosonographic thoracic duct diameter in cirrhotic patients without ascites or varices and patients with extrahepatic portal hypertension was not significantly different from that of normal control subjects. Endosonography affords a unique method to study the thoracic duct (without manipulating it) and a dilated duct found at EUS may be a sign of advanced cirrhosis with ascites and varices. Further, studies of the thoracic duct by EUS could result in a better understanding of pathophysiologic and hemodynamic alterations in portal hypertension. REFERENCES 1. Dumont AE, Mulholland JH. Alteration in thoracic duct lymph flow in hepatic cirrhosis; significance in portal hypertension. Ann Surg 1962;156:668-77. 2. Schieber W. Lymphangiographic demonstration of thoracic duct dilation in portal cirrhosis. Surgery 1965;57:522-4. 3. Dumont AE, Witte MH. Contrasting patterns of thoracic duct lymph formation in hepatic cirrhosis. Surg Gynecol Obstet 1966;122:524-8. 4. Witte CL, Witte MH, Cole WR, Chung YC, Bliesch VR, Dumant AE. Dual origins of ascites in hepatic cirrhosis. Surg Gynecol Obstet 1969;129:1027-33. 5. Witte MH, Dumont AE, Cole WR, Witte CL, Kinter K. Lymph circulation in hepatic cirrhosis: effect of portacaval shunt. Ann Intern Med 1969;70:303-10. 6. Dumont AE, Mulholland JH. Flow rate and composition of thoracic duct lymph in patients with cirrhosis. N Engl J Med 1960;263:471-4. GASTROINTESTINAL ENDOSCOPY
591
V Parasher, E Meroni, A Malesci, et al.
7. Parasher VK, Meroni E, Spinelli P. Anatomy of thoracic duct: an endosonographic study. Gastrointest Endosc 1995;42: 188-9. 8. Hyman C. Physiologic functions of the lymphatic system. Cancer Chemoth Reports 1968;52:25-30. 9. Cain JC, Grindlay JH, Bollman JL, Flock FU, Mann FC. Lymph flow from the liver and thoracic duct. Surg Gynecol Obstet 1947;85:559-62. 10. Caletti GC, Bolondi L, Zani L, Brocchi E, Guizzardi G, Labo G. Detection of portal hypertension and esophageal varices by
Thoracic duct morphology in portal hypertension by EUS
means of endoscopic ultrasonography. Scand J Gastroenterol 1986;21:74-7. 11. Caletti GC, Brocchi E. Baraldini M, Ferrari A, Gibilaro M, Barbara L. Assessment of portal hypertension by endoscopic ultrasonography. Gastrointest Endosc 1990;36:S21-7. 12. Wallace S. Dynamics of normal and abnormal lymphatic systems as studied with contrast media. Cancer Chemoth Reports 1968;52:31-58. 13. Lee FC. Some observations on lymph pressures. Am J Physiol 1923;67:498-513.
Moving? To ensure continued service please notify us of a change of address at least 6 weeks before your move. Phone Subscription Services at 800-453-4351 (outside the U.S. call 314-453-4351) or fax your information to 314-432-1158.
592
GASTROINTESTINAL ENDOSCOPY
VOLUME 48, NO. 6, 1998
Background: Thoracic duct dilation has been demonstrated in portal hypertension and hepatic cirrhosis by lymphangiography and laparotomy and at autopsy. It is thought to be secondary to increased hepatic lymph flow and has been described in the absence of ascites or esophageal varices. The aim of the present study was to observe thoracic duct morphology by endoscopic ultrasound in various subsets of patients with portal hypertension and hepatic cirrhosis and also to validate existing radiologic/surgical data. Methods: The thoracic duct of 33 patients with cirrhosis and portal hypertension was studied by endoscopic ultrasound. Patients were divided into four groups: 1, patients with ascites and esophageal varices; 2, esophageal varices without ascites; 3, without esophageal varices or ascites; 4, extrahepatic portal hypertension due to pancreatic malignancy. The thoracic duct diameter was also measured in 14 control subjects (group 5). Results: When the thoracic duct diameter for the five groups was compared with analysis of variance, significance was p < 0.0001; by pairwise comparison, group 1 differed from the other four groups (p < 0.05). Thoracic duct dilation (5.61 mm) was seen in group 1 patients, whereas no dilation was present in groups 2 through 4. Additionally, thoracic duct diameter in 33 portal hypertensive and/or cirrhotic patients was significantly different from that in the 14 control subjects (p = 0.003). Conclusion: The thoracic duct can be reliably identified by EUS in patients with hepatic cirrhosis and portal hypertension. Dilation of the duct is seen only in patients with hepatic cirrhosis, ascites, and esophageal varices. No thoracic duct dilation is present in extrahepatic portal hypertension. Contrary to existing radiologic/surgical data, thoracic duct dilation is not seen in all patients with hepatic cirrhosis and portal hypertension signifying advanced disease. A dilated thoracic duct by endoscopic ultrasound should be considered yet another sign of portal hypertension. (Gastrointest Endosc 1998;48:588-92.)
Increased formation of lymph which contains an abundance of formed elements is characteristic of hepatic cirrhosis with portal hypertension. These formed elements consist of intact red blood cells and protein. The increase in hepatic lymph flow is due to Received June 25, 1996. For revision August 30, 1996. Accepted June 4, 1998. From the Beebe Medical Center, Lewes, Delaware; Istituto Clinico Humanitas and Nazionale Tumori, Milan, Italy; Wright State University, Dayton, Ohio. Reprint requests: Vinod K. Parasher, MD, Director of Endoscopy, Beebe Medical Center, 424 Savannah Rd., Lewes, DE 19958. Copyright © 1998 by the American Society for Gastrointestinal Endoscopy 0016-5107/98/$5.00 + 0 37/1/92175 588
GASTROINTESTINAL ENDOSCOPY
the obstruction of the hepatic venules as a result of cirrhosis.1-5 The normal flow of hepatic lymph is toward the hilum, and there the lymphatics traverse the hepatoduodenal and hepatogastric ligaments to join the cisterna chyli and the thoracic duct. In portal hypertension, the increased lymph flow causes dilation of the hilar lymphatics and the thoracic duct1,3,6 as shown by lymphangiography2 and postmortem studies and at laparotomy.6 It is noteworthy that the dilation of the thoracic duct has been noted even in the absence of varices and ascites,2,6 suggesting that thoracic duct dilation may be an early sign of portal hypertension. Unfortunately, there are as yet no noninvasive methods to study the thoracic duct. Endoscopic VOLUME 48, NO. 6, 1998
Thoracic duct morphology in portal hypertension by EUS
V Parasher, E Meroni, A Malesci, et al.
Figure 1. EUS demonstrates nondilated thoracic duct in patient with hepatic cirrhosis without ascites or esophageal varices (frequency 7.5 MHz, range 6 cm). Arrowhead, Thoracic duct; large arrow, azygos vein; small arrow, aorta.
Figure 2. EUS demonstrates dilated thoracic duct lying between the azygous vein and aorta in a patient with ascites and esophageal varices (frequency 7.5 MHz, range 4 cm).
ultrasound (EUS) is a sensitive imaging technique that is increasingly being used to study intraluminal and extraluminal disorders of the gastrointestinal tract. More recently, we have described the normal anatomy of the thoracic duct by EUS, thereby establishing a semi-invasive method to observe this structure. The thoracic duct was identified as an anechoic 3 to 4 mm structure that is best seen near the mid-esophagus.7 The aim of our study was to answer the following questions: (1) Can the thoracic duct be reliably identified in patients with hepatic cirrhosis? (2) Can EUS validate existing radiologic/surgical data that indicate that the thoracic duct is dilated in intrahepatic portal hypertension? (3) Is the thoracic duct dilated in all cases of hepatic cirrhosis, irrespective of the presence of ascites or varices?
Ascites was diagnosed by clinical examination, transcutaneous ultrasound, EUS, or CT. Esophageal varices were diagnosed by endoscopy and EUS. Patients with active variceal bleeding or a recent history of bleeding (less than 1 month), hepatic encephalopathy, or severe ascites (causing respiratory distress) were excluded from the study. After examining the stomach, the echoendoscope was withdrawn gradually into the esophagus. The balloon was inflated with deaerated water and the thoracic duct was examined in a systematic fashion. Thoracic duct diameter was recorded at its widest part. The diameter of the thoracic duct was also measured in 14 normal control subjects (group 5) undergoing EUS for other indications. There were 10 men and 4 women ranging in age from 42 to 81 years (mean 53.6 years). The differentiation between the thoracic duct and varices was made by following the thoracic duct inferiorly and superiorly over the length of the esophagus. The thoracic duct generally extended longitudinally as a continuous anechoic structure, whereas periesophageal varices were often multiple, noncontinuous, and serpiginous.
PATIENTS AND METHODS Thirty-three patients with biopsy-proven cirrhosis (etiology included alcohol, hepatitis B virus, and hepatitis C virus) were studied prospectively. The study was approved by our institutional review board. EUS was performed with the GF UM20 (Olympus America, Inc., Melville, N.Y.) echogastroscope. There were 28 men and 5 women. The ages ranged from 45 to 73 years (mean 58.5 years). Ten patients had ascites and esophageal varices (group 1), 8 had esophageal varices without ascites (group 2), and 10 had neither ascites nor esophageal varices (group 3). Five patients who had pancreatic malignancy with varices due to invasion of the portal system were also studied as a group with extrahepatic portal hypertension (group 4). Ascites was not present in these patients, but all had gastric varices and portal vein obstruction. VOLUME 48, NO. 6, 1998
Statistical analysis Mean thoracic duct diameters among the 5 groups were compared by analysis of variance (ANOVA) and Tukey’s multiple range test for pairwise differences. The t-test was used to compare patients with portal hypertension and/or cirrhosis with control subjects. Inferences were made at the 0.05 level of significance.
RESULTS The thoracic duct was seen in all patients as an anechoic thin-walled round structure. The duct was more clearly visualized in the mid-esophagus where the esophagus was anterior to it, the spine posteriGASTROINTESTINAL ENDOSCOPY
589
V Parasher, E Meroni, A Malesci, et al.
Thoracic duct morphology in portal hypertension by EUS
Table 1. Thoracic duct diameter in patient groups and control subjects Group 1
Group 2
Group 3
Group 4
Patients
TDD (mm)
Patients
TDD (mm)
Patients
TDD (mm)
Patients
01. 02. 03. 04. 05. 06. 07. 08. 09. 10.
5.0 5.1 6.0 5.4 7.0 5.3 5.0 6.2 5.1 6.0
01. 02. 03. 04. 05. 06. 07. 08.
3.0 4.0 2.3 3.0 3.0 2.1 2.4 3.0
01. 02. 03. 04. 05. 06. 07. 08. 09. 10.
1.0 2.0 2.3 3.1 1.4 2.4 2.5 3.2 3.0 3.5
01. 02. 03. 04. 05.
Mean 5.61 mm 95% CI [5.13, 6.09] Range 5.0-7.0 mm
Mean 2.85 mm 95% CI [2.35, 3.35] Range 2.1-4.0 mm
Mean 2.44 mm 95% CI [1.86, 3.02] Range 1-3.5 mm
Group 5
TDD (mm)
Patients
2.0 2.0 3.3 3.1 2.5
TDD (mm)
01. 1.5 02. 2.0 03. 3.0 04. 4.0 05. 2.1 06. 1.5 07. 3.0 08. 2.4 09. 2.0 10. 2.0 11. 2.1 12. 2.4 13. 4.0 14. 2.0 Mean 2.43 mm 95% CI [1.97, 2.89] Range 1.5-4.0 mm
Mean 2.58 mm 95% CI [1.83, 3.33] Range 2.0-3.3 mm
Groups are as follows: 1, cirrhotic patients with ascites and varices; 2, cirrhotic patients with varices only; 3, cirrhotic patients without varices or ascites; 4, patients with extrahepatic portal hypertension; 5, represents normal control subjects. TDD, Thoracic duct diameter.
or, the azygos vein to its right, and the aorta to the left side of the duct. A single duct was visualized in all cases. The diameter of the thoracic duct in 14 control subjects (group 5) ranged from 1.5 to 4.0 mm (mean 2.43 mm, Fig.1). Thoracic duct dilation was noted only in patients with ascites and esophageal varices (group 1). Diameter of the thoracic duct ranged from 5 to 7 mm (mean 5.61 mm) in this group (Fig. 2). The thoracic duct was not dilated in cirrhotic patients with esophageal varices who did not have ascites (group 2). Here the diameter ranged from 2.1 to 4.0 mm (mean 2.85 mm). Furthermore, the thoracic duct was also not dilated in cirrhotic patients with neither esophageal varices nor ascites (group 3); the diameter ranged from 1 to 3.5 mm (mean 2.44 mm). Also, the duct was not dilated in patients with extrahepatic portal hypertension (group 4); the diameter ranged from 2.0 to 3.3 mm (mean 2.58 mm). No complications occurred during EUS. When the thoracic duct diameter for the five groups (Table 1) was compared with analysis of variance (ANOVA), significance was p < 0.0001. Furthermore, pairwise comparisons with Tukey’s multiple range test found that group 1 was significantly different (p < 0.05) from all other groups. No other pairwise comparisons were significant. When all 33 patients were compared with control subjects, the mean thoracic duct diameter (3.52 mm) was significantly different from that for control subjects (2.43 mm, p = 0.003). 590
GASTROINTESTINAL ENDOSCOPY
DISCUSSION The main function of the lymphatic system is the collection and transport of large molecules consisting of plasma proteins, particulate matter, tissue debris, bacteria, and foreign substances.8 The liver lymph originates from the interstitial fluid in the intercellular spaces of Disse. From here it extends to the limiting plates and then to the portal spaces. Eighty percent of the hepatic lymph passes through the hilar lymphatics to the cisterna chyli and into the thoracic duct. Under normal circumstances, half of the lymph flowing through the thoracic duct originates in the liver and half in the gastrointestinal tract.9 Hepatic hilar lymphatics are distended in patients with portal hypertension and discharge large volumes of lymph when transected. The thoracic duct is greatly enlarged, and lymph flow through the thoracic duct is often 10 to 15 times the normal flow rate.1,3,6 The distended hilar encapsular lymphatics reflect the hepatic venous outflow obstruction and increased hepatic lymph formation. Decompression of the thoracic duct either externally or by a shunt may ameliorate the portal pressure, ascites, and esophageal varices.3 Indeed, a peritoneal venous shunt may be considered as an artificial thoracic duct. Thoracic duct dilation has been shown in hepatic cirrhosis at autopsy and laparotomy and by lymphangiography.2-6 Additionally, thoracic duct dilation has also been demonstrated in the setting of hepatic cirrhosis without ascites and also in cases of VOLUME 48, NO. 6, 1998
Thoracic duct morphology in portal hypertension by EUS
hepatic cirrhosis without ascites or esophageal varices. Schieber2 demonstrated a dilated thoracic duct (diameter 6 to 9 mm, mean 7.5 mm) in 5 patients who had no ascites or major variceal bleeding. Normal duct diameter by lymphangiography is reported to be 1 to 4 mm (mean 2.5 mm). The thoracic duct was not only dilated but was also increased in length, probably because of the increased tortuosity. Dumont and Mulholland6 showed similar findings in one patient who had neither ascites nor varices. We have demonstrated that the thoracic duct can be identified by EUS. It appears as an anechoic structure, 3 to 4 mm in diameter, best seen close to mid-esophagus.7 EUS has been used to evaluate patients with portal hypertension; findings include esophageal and periesophageal varices, gastric and perigastric varices, dilated azygos veins, and small dilated vessels in the gastric wall in patients with portal hypertensive gastropathy and ascites.10,11 This is the first study of thoracic duct by EUS in patients with portal hypertension. The current study demonstrates that the thoracic duct can be reliably identified in patients with cirrhosis; furthermore, thoracic duct dilation is limited to patients who appear to have advanced disease with coexistent ascites and varices. Thoracic duct dilation was not identified in patients who had no ascites or varices. A dilated thoracic duct in the setting of hepatic cirrhosis and ascites is yet another sign of portal hypertension identifiable by EUS. Because of the relatively small sample size in this study, there is a possibility of beta error. The study may lack the statistical power to detect differences in patients with less severe disease. Because of multicenter nature of the study, interobserver variability may exist. However, we found group 1 to be different from each of the other four groups. Larger sample sizes might eventually result in statistically significant differences among the other four groups. If the mean value for groups 1 through 4 remained about the same, the small differences would not be clinically significant. Our study confirms the previous radiologic/surgical data indicating that the thoracic duct is dilated in hepatic cirrhosis; however, it does not confirm that dilation is seen in all cases of hepatic cirrhosis and portal hypertension. The reason for this discrepancy is not known. It can be hypothesized that lymphangiography may alter lymphatic dynamics by the injection of contrast under pressure. This may cause spurious duct dilation.12 It is also known that cannulation of the thoracic duct so as to block the outflow may cause higher pressure.8,13 The length of the thoracic duct was not determined in VOLUME 48, NO. 6, 1998
V Parasher, E Meroni, A Malesci, et al.
our study because the anatomic length cannot be accurately measured by EUS. Periesophageal varices may cause confusion; however, we found that by following the thoracic duct longitudinally in a dynamic fashion, this problem could be avoided. There has been no study of thoracic duct diameter in extrahepatic portal hypertension. We did not find a dilated thoracic duct in a selected group of patients with extrahepatic portal hypertension; however, none of these patients had ascites. More patients with other causes of extrahepatic portal hypertension need to be studied before conclusions can be drawn concerning thoracic duct diameter in this group of patients. However, studies done in patients with presinusoidal obstruction,5 particularly in portal vein constriction, have shown that there is no additional increase in thoracic duct lymph flow and pressure. The increased production of lymph in the small intestine is probably balanced by a decrease in production of lymph in the liver caused by a reduction in hepatic sinusoidol pressure and decreased sinusoidol perfusion by portal venous blood. In conclusion, EUS will reveal a dilated thoracic duct in cirrhotic patients with ascites and esophageal varices. In our study, the endosonographic thoracic duct diameter in cirrhotic patients without ascites or varices and patients with extrahepatic portal hypertension was not significantly different from that of normal control subjects. Endosonography affords a unique method to study the thoracic duct (without manipulating it) and a dilated duct found at EUS may be a sign of advanced cirrhosis with ascites and varices. Further, studies of the thoracic duct by EUS could result in a better understanding of pathophysiologic and hemodynamic alterations in portal hypertension. REFERENCES 1. Dumont AE, Mulholland JH. Alteration in thoracic duct lymph flow in hepatic cirrhosis; significance in portal hypertension. Ann Surg 1962;156:668-77. 2. Schieber W. Lymphangiographic demonstration of thoracic duct dilation in portal cirrhosis. Surgery 1965;57:522-4. 3. Dumont AE, Witte MH. Contrasting patterns of thoracic duct lymph formation in hepatic cirrhosis. Surg Gynecol Obstet 1966;122:524-8. 4. Witte CL, Witte MH, Cole WR, Chung YC, Bliesch VR, Dumant AE. Dual origins of ascites in hepatic cirrhosis. Surg Gynecol Obstet 1969;129:1027-33. 5. Witte MH, Dumont AE, Cole WR, Witte CL, Kinter K. Lymph circulation in hepatic cirrhosis: effect of portacaval shunt. Ann Intern Med 1969;70:303-10. 6. Dumont AE, Mulholland JH. Flow rate and composition of thoracic duct lymph in patients with cirrhosis. N Engl J Med 1960;263:471-4. GASTROINTESTINAL ENDOSCOPY
591
V Parasher, E Meroni, A Malesci, et al.
7. Parasher VK, Meroni E, Spinelli P. Anatomy of thoracic duct: an endosonographic study. Gastrointest Endosc 1995;42: 188-9. 8. Hyman C. Physiologic functions of the lymphatic system. Cancer Chemoth Reports 1968;52:25-30. 9. Cain JC, Grindlay JH, Bollman JL, Flock FU, Mann FC. Lymph flow from the liver and thoracic duct. Surg Gynecol Obstet 1947;85:559-62. 10. Caletti GC, Bolondi L, Zani L, Brocchi E, Guizzardi G, Labo G. Detection of portal hypertension and esophageal varices by
Thoracic duct morphology in portal hypertension by EUS
means of endoscopic ultrasonography. Scand J Gastroenterol 1986;21:74-7. 11. Caletti GC, Brocchi E. Baraldini M, Ferrari A, Gibilaro M, Barbara L. Assessment of portal hypertension by endoscopic ultrasonography. Gastrointest Endosc 1990;36:S21-7. 12. Wallace S. Dynamics of normal and abnormal lymphatic systems as studied with contrast media. Cancer Chemoth Reports 1968;52:31-58. 13. Lee FC. Some observations on lymph pressures. Am J Physiol 1923;67:498-513.
Moving? To ensure continued service please notify us of a change of address at least 6 weeks before your move. Phone Subscription Services at 800-453-4351 (outside the U.S. call 314-453-4351) or fax your information to 314-432-1158.
592
GASTROINTESTINAL ENDOSCOPY
VOLUME 48, NO. 6, 1998