- Email: [email protected]
Detection of embryologic ventral pancreatic parenchyma with endoscopic ultrasound
0016-5107/96/4301-001455.00 + 0 GASTROINTESTINAL ENDOSCOPY Copyright © 1996 by the American Society for Gastrointestinal Endoscopy
Detection of embryologic ventral pancreatic parenchyma with endoscopic ultrasound Thomas J. Savides, MD, Frank G. Gress, MD, Syed A. Zaidi, MD Steven O. Ikenberry, MD, Robert H. Hawes, MD Indianapolis, Indiana
Background: The prevalence of detecting the embryologic ventral pancreas (ventral anlage) with endoscopic ultrasound (EUS) is unknown. Purpose: To determine the frequency of, and factors associated with, EUS findings consistent with the ventral anlage. Methods: One hundred patients undergoing upper gastrointestinal EUS for any indication were prospectively evaluated for the presence of a focal, hypoechoic area in the pancreatic head using a radial scanning echoendoscope. Multiple clinical and EUS variables were tested against the ability to detect the ventral anlage. Results: The overall detection rate of the ventral anlage was 59%. The ventral anlage was detected in 75% of patients undergoing EUS for nonpancreatic indications, compared to 40% of patients undergoing EUS to evaluate suspected pancreatic disease (p < 0.001). EUS detected the ventral anlage in 72% of patients with a normal EUS-appearing pancreatic head, compared to 29% of patients who had abnormal pancreatic head parenchyma (mass or chronic pancreatitis) on EUS (p < 0.001). Multivariate analysis revealed the only variable associated with detecting the ventral anlage was abnormal pancreatic head parenchyma on EUS. Conclusion: The ventral anlage is frequently detected during pancreatic EUS, with a significantly lower rate of detection in patients with EUS findings of a pancreatic head mass or diffuse chronic pancreatitis. (Gastrointest Endosc 1996 ;43:14-9.)
During embryogenesis, the ventral bud of the pancreas rotates to form the posterior and inferior portions of the pancreatic head I (Fig. 1). Pancreatic imaging with transabdominal ultrasound (US), CT, and endoscopic ultrasound (EUS) can detect differences in echogenicity or density t h a t are suspected to represent the embryologic-derived ventral pancreatic parenchyma (ventral anlage). 2-7 Correlative histoReceived November 17, 1994. For revision January 3, 1995. Accepted April 27, 1995. From the Divisions of Gastroenterology and Departments of Medicine, Indiana University Medical Center, Indianapolis, Indiana, and University of California--San Diego (UCSD) Medical Center, San Diego, California. Reprint requests: Thomas J. Savides, MD, Gastroenterology (8413), UCSD Medical Center, 200 West Arbor Drive, San Diego, CA 92103-8413. 37/1/66867 14
GASTROINTESTINAL ENDOSCOPY
pathologic studies have demonstrated t h a t the ventral anlage seen with transabdominal US in the pancreatic head is hypoechoic because of diminished fat content relative to the embryologic dorsal pancreas. 2, 4 Prospective studies using abdominal CT or transabdominal US suggest t h a t distinct echopatterns of the ventral and dorsal pancreas can be seen in approximately 25% of patients without pancreatic disease, 3, 4 compared to reports using EUS in which the hypoechoic embryologic ventral pancreas was detected in 75% of patients. 5 The recognition of the normal embryologic ventral pancreas area is important to endosonographers because the findings of a focal hypoechoic area in the pancreatic head on EUS can be mistaken for focal pancreatitis or malignancy. The purpose of this study was to evaluate the frequency of detection of the ventral anlage in patients undergoing EUS and to deterVOLUME 43, NO. 1, 1996
m i n e t h e influence of v a r i o u s clinical c h a r a c t e r i s t i c s in t h e detection of this finding.
D
PATIENTS AND METHODS
From J a n u a r y 1, 1994, to June 30, 1994, 100 nonconsecutive patients at Indiana University Medical Center who underwent upper gastrointestinal EUS for any indication also had EUS imaging of their pancreatic heads. Patients who underwent upper EUS during the study period but were not included in the study consisted of those enrolled in research studies using a linear array echoendoscope alone, patients in whom the transducer could not be passed into the duodenum because of esophageal or pyloric obstruction, patients who had EUS for pancreatic indications prior to the decision to include suspected pancreatic disease in the study, patients who had prior resection of the pancreatic head, and patients with complex nonpancreatic findings in whom time constraints prevented imaging of the pancreas. EUS was performed with an Olympus GF-UM20 radial scanning echoendoscope (Olympus America Inc., Melville, N.Y.) at 7.5 and 12 MHz. EUS imaging was performed from a transduodenal position beginning distal to the papilla and then imaging during scope withdrawal until the transducer was pulled through the pylorus. Repeated imaging of the pancreatic head was performed to confirm the presence and correctly identify the pancreatic duct, common bile duct, superior mesenteric artery, superior mesenteric vein, splenic artery, splenic vein, and any hypoechoic area in the pancreatic head. In cases where a suspected pancreatic abnormality was not the indication for EUS, no more than 10 minutes were spent trying to determine the presence or absence of a hypoechoic area in the pancreatic head. The ultimate decision regarding the presence or absence of a hypoechoic area in the pancreatic head consistent with the embryologic ventral pancreas was made by a single experienced pancreatic endosonographer (RH) during the EUS examination. Data were prospectively recorded on standard data sheets and consisted of age, sex, self-reported height and weight, indication for EUS, and pancreatic EUS findings. To estimate body fat accumulation, the body mass index (BMI) was calculated using the formula of weight in kilograms divided by height in meters squared (kg/m2). s Overweight was defined as a BMI greater than or equal to 27.3 for women and greater than or equal to 27.8 for men. These values represent the sex-specific eighty-fifth percentile of BMI for American adults, ages 20 to 29 years, as estimated from the second National Health and Nutrition Examination Survey, 1976-1980 (NHANES II). 9 The presence of distinct embryologic pancreas parenchyma was based on EUS detection of a round or triangular, well-demarcated, hypoechoic region in the pancreatic head, as previous]y described. 2-~ This area was assessed in terms of shape (round or triangular), distinction between hypoechoic and normal pancreatic parenchyma (very, intermediate, slight), contour of border between hypoechoic and normal pancreatic tissue (sharp, intermediate, jagged), the presence of the pancreatic duct coursing through the hypoechoic region of the pancreas, and the presence of the superior mesenteric artery, superior mesenteric vein, or common bile duct in the same view. EUS findings of an abnormality in the pancreatic head, VOLUME 43, NO. 1, 1996
I
V
2"
ventral bud
Figure 1. Diagram of the rotation of the embryologic dorsal and ventral pancreas. Note the ventral bud is posterior and inferior to the dorsal bud.
either chronic pancreatitis or a hypoechoic mass, were recorded. Chronic pancreatitis was diagnosed on the basis of at least three of the following: a hyperechoic main pancreatic duct, a thickened main pancreatic duct, an irregular main pancreatic duct, a dilated main pancreatic duct, visible pancreatic duct side branches, pancreatic duct stones, heterogeneous echopattern of the pancreatic parenchyma, echogenic bands, echogenic foci, cystic cavities, or lobular architecture.I° The presence or absence of pancreas divisum was determined by reviewing the Indiana University Medical Center ERCP computer database or from reports of ERCPs performed at outside institutions. Data are expressed in terms of mean _+ standard error of measure (SEM). Statistical analysis was performed as appropriate using chi-squared analysis, univariate and multivariate logistic regression, kappa statistics, and the Wilcoxon rank-sum test. RESULTS
B e t w e e n J a n u a r y 1, 1994, a n d J u n e 30, 1994, 220 p a t i e n t s u n d e r w e n t u p p e r g a s t r o i n t e s t i n a l E U S studies. One h u n d r e d p a t i e n t s w e r e prospectively e n t e r e d into this study. T h e indication for E U S in t h e s e 100 p a t i e n t s w a s to e v a l u a t e a s u s p e c t e d a b n o r m a l i t y in t h e p a n c r e a s in 47 (47%), e s o p h a g u s in 21 (21%), s t o m a c h in 15 (15%), m e d i a s t i n u m in 15 (15%), duoden u m in 1 (1%), a n d p e r i t o n e a l cavity in 1 (1%). T h e indication for E U S in t h e 120 p a t i e n t s who w e r e not included in t h e s t u d y w a s to e v a l u a t e a s u s p e c t e d a b n o r m a l i t y in t h e p a n c r e a s in 79 (66%), e s o p h a g u s in 27 (22%), s t o m a c h in 7 (6%), m e d i a s t i n u m in 6 (5%), a n d d u o d e n u m in 1 (1%). Fifty n i n e of 100 p a t i e n t s (59%) w e r e found to h a v e a hypoechoic a s p e c t of the h e a d of the p a n c r e a s t h a t GASTROINTESTINAL ENDOSCOPY
15
Figures 2 to 5. Endoscopic ultrasonogram examples showing the hypoechoic embryologic ventral pancreas, and the more hyperechoic embryologic dorsal pancreas. V, Embryologic ventral pancreas; D, embryologic dorsal pancreas; PD, pancreatic duct; SMA, superior mesenteric artery; SMV, superior mesenteric vein; AO, aorta; IVC, inferior vena cava. Arrows show the border between the embryologic ventral and dorsal pancreas in Fig. 2.
was believed consistent with the embryologic ventral pancreas. Examples of the presumed hypoechoic ventral anlage are shown in Figs. 2 to 5. Table 1 shows the EUS characteristics associated with detecting the ventral anlage. In general, the hypoechoic embryologic ventral pancreatic parenchyma was triangular and was imaged in the same plane as the pancreatic duct, the SMA, and the SMV. Table 2 shows the ability to detect distinct embryologic ventral and dorsal pancreatic parenchymabased on different clinical and EUS variables. Chi-squared analysis was conducted to test for the association of detecting the ventral anlage with the variables of 16
GASTROINTESTINAL ENDOSCOPY
gender, overweight, pancreas divisum, suspected pancreatic disease as the indication for EUS, and EUS findings of a pancreatic head abnormality (either mass or diffuse chronic pancreatitis). The only two variables associated with the ability to detect the embryologic ventral pancreas on EUS were suspected pancreatic disease prior to EUS (p < 0.001), and pancreatic head disease found on EUS (p < 0.001). Also, there was 80% agreement (kappa = 0.5907,p < 0.0001) between suspected pancreatic disease prior to EUS and pancreatic head abnormality found on EUS. A logistic regression using the variables "suspected pancreatic disease prior to EUS" and "pancreatic head VOLUME 43, NO. 1, 1996
abnormality found on EUS" to predict the likelihood of detecting the ventral anlage resulted in the variable "pancreatic head abnormality found on EUS" as the only significant covariate (p < 0.0001). Among patients who underwent EUS for nonpancreatic indications, 75% (40 of 53) had distinct embryologic ventral and dorsal pancreatic parenchyma detected, as compared with 40% (19 of 47) for patients whose indication for EUS was presumed pancreatic disease (p < 0.001). Of those patients who had EUS for suspected pancreatic disease, 62% (29 of 47) had either diffuse chronic pancreatitis or a pancreatic head mass, in contrast to 4% (2 of 53) of patients with no suspected pancreatic disease prior to EUS. Among patients who had no EUS findings of a pancreatic head mass or diffuse chronic pancreatitis in the head, 72% (50 of 69) had a distinct embryologic ventral pancreas parenchyma detected, compared to 29% (9 of 31) of patients who had EUS findings of a pancreatic head mass or diffuse chronic pancreatitis in the head (p < 0.001).
Table 1. EUS characteristics of the focal hypoechoic area in the pancreatic head believed to represent the embryologic ventral pancreas in 59 patients Characteristic Number (%) Shape Triangular 57 (97) Round 2 (3) Distinction Very 19 (32) Intermediate 19 (32) Slightly 21 (36) Contour Sharp 15 (25) Intermediate 37 (63) Jagged 7 (12) Anatomic landmarks Pancreatic duct 52 (88) SMA 38 (64) SMV 40 (68) CBD 10 (17) None 2 (3) SMA, Superiormesentericartery;SMV,superiormesentericvein; CBD, commonbile duct.
DISCUSSION The EUS detection of the hypoechoic area in the pancreatic head corresponding with the embryologic ventral pancreas has been reported anecdotally, but this is the first study to prospectively evaluate the prevalence of this finding and clinical or ultrasonographic features associated with its detection. Our results confirm previous observations 5 t h a t 75% of"normal patients" (i.e., those undergoing EUS for nonpancreatic indications) have a detectable hypoechoic ventral anlage. In contrast, the ventral anlage was detected in only 40% of our patients undergoing EUS to evaluate suspected pancreatic disease. The reason for the decreased ability to detect the ventral anlage in patients undergoing EUS for suspected pancreatic disease was most likely because 62% of these patients had EUS abnormalities of the pancreatic head (either chronic pancreatitis or a hypoechoic mass), compared to only 4% of the patients with no suspected pancreatic disease. When there was either chronic pancreatitis or a hypoechoic mass seen on EUS in the pancreatic head, the ventral anlage could only be detected 29% of the time, compared to 72% of the time when there was a normal=appearing pancreatic head on EUS. When there was diffuse chronic pancreatitis or a hypoechoic mass in the pancreatic head, it was much more difficult to detect a distinct characteristic triangular hypoechoic area because of an inability to tell if a focal hypoechoic area was part of the chronic pancreatitis or the hypoechoic mass. It has been suggested t h a t increased fat deposition occurs in the embryologic dorsal pancreas with advancing age, resulting in a statistically higher detecVOLUME 43, NO. 1, 1996
tion rate of the ventral anlage by CT scan. 4 In our study, however, univariate and multivariate analysis showed that neither age, gender, BMI, overweight (based on sex-specific BMI threshold), nor pancreas divisum was associated with detecting the ventral anlage. The only variables on univariate analysis t h a t affected detection of ventral anlage were performing the EUS to evaluate suspected pancreatic disease or finding pancreatic disease at the time of EUS. As would be expected, these two variables were highly correlated with each other. When multivariate analysis was performed, only the finding of abnormality in the pancreatic head (hypoechoic mass or chronic pancreatitis) was associated with decreased ability to detect the ventral anlage. A limitation of this study is that not all patients were consecutively enrolled because at the start of the study only patients without indications of suspected pancreatic disease were entered. As the study progressed, it was decided to include all patients. The frequencies of suspected anatomic areas of abnormality were similar between the patients included and not included in the study, as might occur with random sampling. However, given t h a t patients were not consecutively enrolled, it is possible that a selection bias occurred t h a t could influence the frequency of detection of the ventral anlage. The ability to detect the hypoechoic embryologic ventral pancreas depends on extensive experience with pancreatic EUS. We generally find the embryologic ventral pancreas when we are withdrawing the transducer from the second portion of the duodenum. When the tubular inferior vena cava or aorta is seen, GASTROINTESTINAL ENDOSCOPY 17
Figure 6. Left panel, Schematic drawing of the positioning of the EUS transducer to obtain the imaging plane to detect the distinct embryologic ventral and dorsal pancreas parenchyma. Middle panel, Schematic drawing of EUS image in the same plane as the transducer in the left panel. This shows the embryologic ventral pancreas located posterior to the embryologic dorsal pancreas. Right panel, 360-degree EUS image showing the same anatomic landmarks and positions of embryologic ventral and dorsal pancreatic parenchyma as in the center panel drawing. V, Embryologic ventral pancreas; D, embryologic dorsal pancreas; PD, pancreatic duct; CBD, common bile duct; SMA, superior mesenteric artery; SMV, superior mesenteric vein; AO, aorta; IVC, inferior vena cava; B, EUS transducer balloon; Ant., anterior; Post., posterior; Rt., right; Lt., left.
Table 2. Characteristics of patients with or without EUS detection of the ventral anlage with EUS Characteristic
Ventral anlage detected (n = 59)
No ventral anlage detected (n = 41)
p value
Age (y) 56.8 _+1.9 59.7 + 2.5 0.45 Male (%) 41 56 0.24 BMI (kg/m2) 25.4 __0.8 25.9 -+ 0.9 0.51 Overweight (%) 32.2 31.7 0.96 Suspected pancreatic abnormality as indication 32.2 68.3 <0.001 for performing EUS (%) Pancreatic head mass or chronic pancreatitis 15.3 53.7 <0.001 found with EUS (%) Pancreas divisum on ERCP (%)* 14.2 4.3 0.28 Results expressed in terms of mean + SEM. BM1, Body mass index. *Based on ERCP results in 14 patients with ventral anlage detected and 23 patients with no ventral anlage detected.
the p a n c r e a t i c p a r e n c h y m a is found b e t w e e n the t r a n s d u c e r a n d the v a s c u l a r structures. As the t r a n s ducer is pulled back further, the superior m e s e n t e r i c a r t e r y a n d vein come into view. This inferior-posterior portion of the p a n c r e a t i c h e a d is generally w h e r e the t r i a n g u l a r , hypoechoic v e n t r a l anlage will be seen 18 G A S T R O I N T E S T I N A L ENDOSCOPY
(Fig. 6). With subtle m o v e m e n t s , the p a n c r e a t i c duct can u s u a l l y be seen crossing from the v e n t r a l anlage into the more hyperechoic dorsal anlage. The interface b e t w e e n the v e n t r a l p a n c r e a s a n d dorsal p a n c r e a s is u s u a l l y distinct a n d fairly linear. This hypoechoic area m a y not be appreciated as easily u s i n g a linear a r r a y VOLUME 43, NO. 1, 1996
endoscope because of the ultrasound imaging oriented parallel to the endoscope and a narrower imaging field, or a catheter-based ultrasound probe with limited resolution and depth of penetration. The EUS appearance of the ventral anlage can vary a great deal, and detection may be subtle. In this study, the appearance of the hypoechoic ventral anlage was equally noted as %ery distinct," "intermediate distinct," or "slightly distinct." These are subjective terms meant to express its spectrum of appearances. Usually the ventral anlage was not immediately apparent unless a slow, careful inspection was performed on pull-back and one was specifically looking for a hypoechoic area. Several sweeps of the echoendoscope around the duodenal curve were usually needed for us to be confident in the presence of the ventral anlage. In summary, a distinct hypoechoic triangular area in the posterior aspect of the pancreatic head is commonly seen with EUS and presumably represents the ventral anlage. The ability to detect the ventral anlage decreases in the presence of pancreatic head disease on EUS. This finding should not be confused with a pancreatic mass or focal chronic pancreatitis. ACKNOWLEDGMENTS The authors would like to thank Raji Sridhara, PhD, and Mark Hanna, MS, for statistical analysis and Lydia V. Gerbig for medical illustration.
VOLUME 43, NO. 1, 1996
REFERENCES 1. Gray SW, Skandalakis JE, Skandalakis LJ. Embryology and congenital anomalies of the pancreas. In: Howard JM, Jordan GL, Reber HA, eds. Surgical diseases of the pancreas. Philadelphia: Lea & Febiger, 1987:37-45. 2. Marchal G, Verbeken E, Van Steenbergen W, et al. Uneven lipomatosis: a pitfall in pancreatic sonography. Gastrointestinal Radiology 1989;14:223-7. 3. Donald JJ, Shorvon PJ, Lees WR. A hypoechoic area within the head of the pancreas--a normal variant. Clin Radiol 1990;41: 337-8. 4. Atri M, Nazarnia S, Mehio A, et al, Hypoechogenic embryologic ventral aspect of the head and uncinate process of the pancreas: in vitro correlation of US with histopathologic findings. Radiology 1994;190:441-4. 5. Lees WR. Endoscopic ultrasonography of the pancreas. In: Sivak MV, ed. Endoscopic ultrasonography symposium. Cleveland: Cleveland Chnic Syllabus, 1989:197-208. 6. Rosch T, Classen M. Gastroenterologic endosonography. New York: Thieme Medical Publishers, 1992:39-42. 7. Caletti G, Odegaard S, Rosch T, et al. Endoscopic ultrasonography (EUS): a summary of the conclusions of the working party for the tenth World Congress of Gastroenterology, Los Angeles, California, October 1994. Am J Gastroenterol 1994; 89:$138-43. 8. Van Itallie TB. Health implications of overweight and obesity in the United States. Ann Intern Med 1985;103:983-8. 9. National Center for Health Statistics. Plan and operation ofthe national health and nutrition examination survey, 1976-80. Washington, D.C.: U.S. Public Health Service, 1981. DHHS publication No. (PHS) 81-1317. (Vital and Health Statistics, series 1, no. 15). 10. Lees WR. Endoscopic ultrasonography of chronic pancreatitis and pancreatic pseudocysts. Scand J Gastroenterol 1986;21 (suppl 123):123-9.
GASTROINTESTINAL ENDOSCOPY
19
Detection of embryologic ventral pancreatic parenchyma with endoscopic ultrasound Thomas J. Savides, MD, Frank G. Gress, MD, Syed A. Zaidi, MD Steven O. Ikenberry, MD, Robert H. Hawes, MD Indianapolis, Indiana
Background: The prevalence of detecting the embryologic ventral pancreas (ventral anlage) with endoscopic ultrasound (EUS) is unknown. Purpose: To determine the frequency of, and factors associated with, EUS findings consistent with the ventral anlage. Methods: One hundred patients undergoing upper gastrointestinal EUS for any indication were prospectively evaluated for the presence of a focal, hypoechoic area in the pancreatic head using a radial scanning echoendoscope. Multiple clinical and EUS variables were tested against the ability to detect the ventral anlage. Results: The overall detection rate of the ventral anlage was 59%. The ventral anlage was detected in 75% of patients undergoing EUS for nonpancreatic indications, compared to 40% of patients undergoing EUS to evaluate suspected pancreatic disease (p < 0.001). EUS detected the ventral anlage in 72% of patients with a normal EUS-appearing pancreatic head, compared to 29% of patients who had abnormal pancreatic head parenchyma (mass or chronic pancreatitis) on EUS (p < 0.001). Multivariate analysis revealed the only variable associated with detecting the ventral anlage was abnormal pancreatic head parenchyma on EUS. Conclusion: The ventral anlage is frequently detected during pancreatic EUS, with a significantly lower rate of detection in patients with EUS findings of a pancreatic head mass or diffuse chronic pancreatitis. (Gastrointest Endosc 1996 ;43:14-9.)
During embryogenesis, the ventral bud of the pancreas rotates to form the posterior and inferior portions of the pancreatic head I (Fig. 1). Pancreatic imaging with transabdominal ultrasound (US), CT, and endoscopic ultrasound (EUS) can detect differences in echogenicity or density t h a t are suspected to represent the embryologic-derived ventral pancreatic parenchyma (ventral anlage). 2-7 Correlative histoReceived November 17, 1994. For revision January 3, 1995. Accepted April 27, 1995. From the Divisions of Gastroenterology and Departments of Medicine, Indiana University Medical Center, Indianapolis, Indiana, and University of California--San Diego (UCSD) Medical Center, San Diego, California. Reprint requests: Thomas J. Savides, MD, Gastroenterology (8413), UCSD Medical Center, 200 West Arbor Drive, San Diego, CA 92103-8413. 37/1/66867 14
GASTROINTESTINAL ENDOSCOPY
pathologic studies have demonstrated t h a t the ventral anlage seen with transabdominal US in the pancreatic head is hypoechoic because of diminished fat content relative to the embryologic dorsal pancreas. 2, 4 Prospective studies using abdominal CT or transabdominal US suggest t h a t distinct echopatterns of the ventral and dorsal pancreas can be seen in approximately 25% of patients without pancreatic disease, 3, 4 compared to reports using EUS in which the hypoechoic embryologic ventral pancreas was detected in 75% of patients. 5 The recognition of the normal embryologic ventral pancreas area is important to endosonographers because the findings of a focal hypoechoic area in the pancreatic head on EUS can be mistaken for focal pancreatitis or malignancy. The purpose of this study was to evaluate the frequency of detection of the ventral anlage in patients undergoing EUS and to deterVOLUME 43, NO. 1, 1996
m i n e t h e influence of v a r i o u s clinical c h a r a c t e r i s t i c s in t h e detection of this finding.
D
PATIENTS AND METHODS
From J a n u a r y 1, 1994, to June 30, 1994, 100 nonconsecutive patients at Indiana University Medical Center who underwent upper gastrointestinal EUS for any indication also had EUS imaging of their pancreatic heads. Patients who underwent upper EUS during the study period but were not included in the study consisted of those enrolled in research studies using a linear array echoendoscope alone, patients in whom the transducer could not be passed into the duodenum because of esophageal or pyloric obstruction, patients who had EUS for pancreatic indications prior to the decision to include suspected pancreatic disease in the study, patients who had prior resection of the pancreatic head, and patients with complex nonpancreatic findings in whom time constraints prevented imaging of the pancreas. EUS was performed with an Olympus GF-UM20 radial scanning echoendoscope (Olympus America Inc., Melville, N.Y.) at 7.5 and 12 MHz. EUS imaging was performed from a transduodenal position beginning distal to the papilla and then imaging during scope withdrawal until the transducer was pulled through the pylorus. Repeated imaging of the pancreatic head was performed to confirm the presence and correctly identify the pancreatic duct, common bile duct, superior mesenteric artery, superior mesenteric vein, splenic artery, splenic vein, and any hypoechoic area in the pancreatic head. In cases where a suspected pancreatic abnormality was not the indication for EUS, no more than 10 minutes were spent trying to determine the presence or absence of a hypoechoic area in the pancreatic head. The ultimate decision regarding the presence or absence of a hypoechoic area in the pancreatic head consistent with the embryologic ventral pancreas was made by a single experienced pancreatic endosonographer (RH) during the EUS examination. Data were prospectively recorded on standard data sheets and consisted of age, sex, self-reported height and weight, indication for EUS, and pancreatic EUS findings. To estimate body fat accumulation, the body mass index (BMI) was calculated using the formula of weight in kilograms divided by height in meters squared (kg/m2). s Overweight was defined as a BMI greater than or equal to 27.3 for women and greater than or equal to 27.8 for men. These values represent the sex-specific eighty-fifth percentile of BMI for American adults, ages 20 to 29 years, as estimated from the second National Health and Nutrition Examination Survey, 1976-1980 (NHANES II). 9 The presence of distinct embryologic pancreas parenchyma was based on EUS detection of a round or triangular, well-demarcated, hypoechoic region in the pancreatic head, as previous]y described. 2-~ This area was assessed in terms of shape (round or triangular), distinction between hypoechoic and normal pancreatic parenchyma (very, intermediate, slight), contour of border between hypoechoic and normal pancreatic tissue (sharp, intermediate, jagged), the presence of the pancreatic duct coursing through the hypoechoic region of the pancreas, and the presence of the superior mesenteric artery, superior mesenteric vein, or common bile duct in the same view. EUS findings of an abnormality in the pancreatic head, VOLUME 43, NO. 1, 1996
I
V
2"
ventral bud
Figure 1. Diagram of the rotation of the embryologic dorsal and ventral pancreas. Note the ventral bud is posterior and inferior to the dorsal bud.
either chronic pancreatitis or a hypoechoic mass, were recorded. Chronic pancreatitis was diagnosed on the basis of at least three of the following: a hyperechoic main pancreatic duct, a thickened main pancreatic duct, an irregular main pancreatic duct, a dilated main pancreatic duct, visible pancreatic duct side branches, pancreatic duct stones, heterogeneous echopattern of the pancreatic parenchyma, echogenic bands, echogenic foci, cystic cavities, or lobular architecture.I° The presence or absence of pancreas divisum was determined by reviewing the Indiana University Medical Center ERCP computer database or from reports of ERCPs performed at outside institutions. Data are expressed in terms of mean _+ standard error of measure (SEM). Statistical analysis was performed as appropriate using chi-squared analysis, univariate and multivariate logistic regression, kappa statistics, and the Wilcoxon rank-sum test. RESULTS
B e t w e e n J a n u a r y 1, 1994, a n d J u n e 30, 1994, 220 p a t i e n t s u n d e r w e n t u p p e r g a s t r o i n t e s t i n a l E U S studies. One h u n d r e d p a t i e n t s w e r e prospectively e n t e r e d into this study. T h e indication for E U S in t h e s e 100 p a t i e n t s w a s to e v a l u a t e a s u s p e c t e d a b n o r m a l i t y in t h e p a n c r e a s in 47 (47%), e s o p h a g u s in 21 (21%), s t o m a c h in 15 (15%), m e d i a s t i n u m in 15 (15%), duoden u m in 1 (1%), a n d p e r i t o n e a l cavity in 1 (1%). T h e indication for E U S in t h e 120 p a t i e n t s who w e r e not included in t h e s t u d y w a s to e v a l u a t e a s u s p e c t e d a b n o r m a l i t y in t h e p a n c r e a s in 79 (66%), e s o p h a g u s in 27 (22%), s t o m a c h in 7 (6%), m e d i a s t i n u m in 6 (5%), a n d d u o d e n u m in 1 (1%). Fifty n i n e of 100 p a t i e n t s (59%) w e r e found to h a v e a hypoechoic a s p e c t of the h e a d of the p a n c r e a s t h a t GASTROINTESTINAL ENDOSCOPY
15
Figures 2 to 5. Endoscopic ultrasonogram examples showing the hypoechoic embryologic ventral pancreas, and the more hyperechoic embryologic dorsal pancreas. V, Embryologic ventral pancreas; D, embryologic dorsal pancreas; PD, pancreatic duct; SMA, superior mesenteric artery; SMV, superior mesenteric vein; AO, aorta; IVC, inferior vena cava. Arrows show the border between the embryologic ventral and dorsal pancreas in Fig. 2.
was believed consistent with the embryologic ventral pancreas. Examples of the presumed hypoechoic ventral anlage are shown in Figs. 2 to 5. Table 1 shows the EUS characteristics associated with detecting the ventral anlage. In general, the hypoechoic embryologic ventral pancreatic parenchyma was triangular and was imaged in the same plane as the pancreatic duct, the SMA, and the SMV. Table 2 shows the ability to detect distinct embryologic ventral and dorsal pancreatic parenchymabased on different clinical and EUS variables. Chi-squared analysis was conducted to test for the association of detecting the ventral anlage with the variables of 16
GASTROINTESTINAL ENDOSCOPY
gender, overweight, pancreas divisum, suspected pancreatic disease as the indication for EUS, and EUS findings of a pancreatic head abnormality (either mass or diffuse chronic pancreatitis). The only two variables associated with the ability to detect the embryologic ventral pancreas on EUS were suspected pancreatic disease prior to EUS (p < 0.001), and pancreatic head disease found on EUS (p < 0.001). Also, there was 80% agreement (kappa = 0.5907,p < 0.0001) between suspected pancreatic disease prior to EUS and pancreatic head abnormality found on EUS. A logistic regression using the variables "suspected pancreatic disease prior to EUS" and "pancreatic head VOLUME 43, NO. 1, 1996
abnormality found on EUS" to predict the likelihood of detecting the ventral anlage resulted in the variable "pancreatic head abnormality found on EUS" as the only significant covariate (p < 0.0001). Among patients who underwent EUS for nonpancreatic indications, 75% (40 of 53) had distinct embryologic ventral and dorsal pancreatic parenchyma detected, as compared with 40% (19 of 47) for patients whose indication for EUS was presumed pancreatic disease (p < 0.001). Of those patients who had EUS for suspected pancreatic disease, 62% (29 of 47) had either diffuse chronic pancreatitis or a pancreatic head mass, in contrast to 4% (2 of 53) of patients with no suspected pancreatic disease prior to EUS. Among patients who had no EUS findings of a pancreatic head mass or diffuse chronic pancreatitis in the head, 72% (50 of 69) had a distinct embryologic ventral pancreas parenchyma detected, compared to 29% (9 of 31) of patients who had EUS findings of a pancreatic head mass or diffuse chronic pancreatitis in the head (p < 0.001).
Table 1. EUS characteristics of the focal hypoechoic area in the pancreatic head believed to represent the embryologic ventral pancreas in 59 patients Characteristic Number (%) Shape Triangular 57 (97) Round 2 (3) Distinction Very 19 (32) Intermediate 19 (32) Slightly 21 (36) Contour Sharp 15 (25) Intermediate 37 (63) Jagged 7 (12) Anatomic landmarks Pancreatic duct 52 (88) SMA 38 (64) SMV 40 (68) CBD 10 (17) None 2 (3) SMA, Superiormesentericartery;SMV,superiormesentericvein; CBD, commonbile duct.
DISCUSSION The EUS detection of the hypoechoic area in the pancreatic head corresponding with the embryologic ventral pancreas has been reported anecdotally, but this is the first study to prospectively evaluate the prevalence of this finding and clinical or ultrasonographic features associated with its detection. Our results confirm previous observations 5 t h a t 75% of"normal patients" (i.e., those undergoing EUS for nonpancreatic indications) have a detectable hypoechoic ventral anlage. In contrast, the ventral anlage was detected in only 40% of our patients undergoing EUS to evaluate suspected pancreatic disease. The reason for the decreased ability to detect the ventral anlage in patients undergoing EUS for suspected pancreatic disease was most likely because 62% of these patients had EUS abnormalities of the pancreatic head (either chronic pancreatitis or a hypoechoic mass), compared to only 4% of the patients with no suspected pancreatic disease. When there was either chronic pancreatitis or a hypoechoic mass seen on EUS in the pancreatic head, the ventral anlage could only be detected 29% of the time, compared to 72% of the time when there was a normal=appearing pancreatic head on EUS. When there was diffuse chronic pancreatitis or a hypoechoic mass in the pancreatic head, it was much more difficult to detect a distinct characteristic triangular hypoechoic area because of an inability to tell if a focal hypoechoic area was part of the chronic pancreatitis or the hypoechoic mass. It has been suggested t h a t increased fat deposition occurs in the embryologic dorsal pancreas with advancing age, resulting in a statistically higher detecVOLUME 43, NO. 1, 1996
tion rate of the ventral anlage by CT scan. 4 In our study, however, univariate and multivariate analysis showed that neither age, gender, BMI, overweight (based on sex-specific BMI threshold), nor pancreas divisum was associated with detecting the ventral anlage. The only variables on univariate analysis t h a t affected detection of ventral anlage were performing the EUS to evaluate suspected pancreatic disease or finding pancreatic disease at the time of EUS. As would be expected, these two variables were highly correlated with each other. When multivariate analysis was performed, only the finding of abnormality in the pancreatic head (hypoechoic mass or chronic pancreatitis) was associated with decreased ability to detect the ventral anlage. A limitation of this study is that not all patients were consecutively enrolled because at the start of the study only patients without indications of suspected pancreatic disease were entered. As the study progressed, it was decided to include all patients. The frequencies of suspected anatomic areas of abnormality were similar between the patients included and not included in the study, as might occur with random sampling. However, given t h a t patients were not consecutively enrolled, it is possible that a selection bias occurred t h a t could influence the frequency of detection of the ventral anlage. The ability to detect the hypoechoic embryologic ventral pancreas depends on extensive experience with pancreatic EUS. We generally find the embryologic ventral pancreas when we are withdrawing the transducer from the second portion of the duodenum. When the tubular inferior vena cava or aorta is seen, GASTROINTESTINAL ENDOSCOPY 17
Figure 6. Left panel, Schematic drawing of the positioning of the EUS transducer to obtain the imaging plane to detect the distinct embryologic ventral and dorsal pancreas parenchyma. Middle panel, Schematic drawing of EUS image in the same plane as the transducer in the left panel. This shows the embryologic ventral pancreas located posterior to the embryologic dorsal pancreas. Right panel, 360-degree EUS image showing the same anatomic landmarks and positions of embryologic ventral and dorsal pancreatic parenchyma as in the center panel drawing. V, Embryologic ventral pancreas; D, embryologic dorsal pancreas; PD, pancreatic duct; CBD, common bile duct; SMA, superior mesenteric artery; SMV, superior mesenteric vein; AO, aorta; IVC, inferior vena cava; B, EUS transducer balloon; Ant., anterior; Post., posterior; Rt., right; Lt., left.
Table 2. Characteristics of patients with or without EUS detection of the ventral anlage with EUS Characteristic
Ventral anlage detected (n = 59)
No ventral anlage detected (n = 41)
p value
Age (y) 56.8 _+1.9 59.7 + 2.5 0.45 Male (%) 41 56 0.24 BMI (kg/m2) 25.4 __0.8 25.9 -+ 0.9 0.51 Overweight (%) 32.2 31.7 0.96 Suspected pancreatic abnormality as indication 32.2 68.3 <0.001 for performing EUS (%) Pancreatic head mass or chronic pancreatitis 15.3 53.7 <0.001 found with EUS (%) Pancreas divisum on ERCP (%)* 14.2 4.3 0.28 Results expressed in terms of mean + SEM. BM1, Body mass index. *Based on ERCP results in 14 patients with ventral anlage detected and 23 patients with no ventral anlage detected.
the p a n c r e a t i c p a r e n c h y m a is found b e t w e e n the t r a n s d u c e r a n d the v a s c u l a r structures. As the t r a n s ducer is pulled back further, the superior m e s e n t e r i c a r t e r y a n d vein come into view. This inferior-posterior portion of the p a n c r e a t i c h e a d is generally w h e r e the t r i a n g u l a r , hypoechoic v e n t r a l anlage will be seen 18 G A S T R O I N T E S T I N A L ENDOSCOPY
(Fig. 6). With subtle m o v e m e n t s , the p a n c r e a t i c duct can u s u a l l y be seen crossing from the v e n t r a l anlage into the more hyperechoic dorsal anlage. The interface b e t w e e n the v e n t r a l p a n c r e a s a n d dorsal p a n c r e a s is u s u a l l y distinct a n d fairly linear. This hypoechoic area m a y not be appreciated as easily u s i n g a linear a r r a y VOLUME 43, NO. 1, 1996
endoscope because of the ultrasound imaging oriented parallel to the endoscope and a narrower imaging field, or a catheter-based ultrasound probe with limited resolution and depth of penetration. The EUS appearance of the ventral anlage can vary a great deal, and detection may be subtle. In this study, the appearance of the hypoechoic ventral anlage was equally noted as %ery distinct," "intermediate distinct," or "slightly distinct." These are subjective terms meant to express its spectrum of appearances. Usually the ventral anlage was not immediately apparent unless a slow, careful inspection was performed on pull-back and one was specifically looking for a hypoechoic area. Several sweeps of the echoendoscope around the duodenal curve were usually needed for us to be confident in the presence of the ventral anlage. In summary, a distinct hypoechoic triangular area in the posterior aspect of the pancreatic head is commonly seen with EUS and presumably represents the ventral anlage. The ability to detect the ventral anlage decreases in the presence of pancreatic head disease on EUS. This finding should not be confused with a pancreatic mass or focal chronic pancreatitis. ACKNOWLEDGMENTS The authors would like to thank Raji Sridhara, PhD, and Mark Hanna, MS, for statistical analysis and Lydia V. Gerbig for medical illustration.
VOLUME 43, NO. 1, 1996
REFERENCES 1. Gray SW, Skandalakis JE, Skandalakis LJ. Embryology and congenital anomalies of the pancreas. In: Howard JM, Jordan GL, Reber HA, eds. Surgical diseases of the pancreas. Philadelphia: Lea & Febiger, 1987:37-45. 2. Marchal G, Verbeken E, Van Steenbergen W, et al. Uneven lipomatosis: a pitfall in pancreatic sonography. Gastrointestinal Radiology 1989;14:223-7. 3. Donald JJ, Shorvon PJ, Lees WR. A hypoechoic area within the head of the pancreas--a normal variant. Clin Radiol 1990;41: 337-8. 4. Atri M, Nazarnia S, Mehio A, et al, Hypoechogenic embryologic ventral aspect of the head and uncinate process of the pancreas: in vitro correlation of US with histopathologic findings. Radiology 1994;190:441-4. 5. Lees WR. Endoscopic ultrasonography of the pancreas. In: Sivak MV, ed. Endoscopic ultrasonography symposium. Cleveland: Cleveland Chnic Syllabus, 1989:197-208. 6. Rosch T, Classen M. Gastroenterologic endosonography. New York: Thieme Medical Publishers, 1992:39-42. 7. Caletti G, Odegaard S, Rosch T, et al. Endoscopic ultrasonography (EUS): a summary of the conclusions of the working party for the tenth World Congress of Gastroenterology, Los Angeles, California, October 1994. Am J Gastroenterol 1994; 89:$138-43. 8. Van Itallie TB. Health implications of overweight and obesity in the United States. Ann Intern Med 1985;103:983-8. 9. National Center for Health Statistics. Plan and operation ofthe national health and nutrition examination survey, 1976-80. Washington, D.C.: U.S. Public Health Service, 1981. DHHS publication No. (PHS) 81-1317. (Vital and Health Statistics, series 1, no. 15). 10. Lees WR. Endoscopic ultrasonography of chronic pancreatitis and pancreatic pseudocysts. Scand J Gastroenterol 1986;21 (suppl 123):123-9.
GASTROINTESTINAL ENDOSCOPY
19