- Email: [email protected]
Sonographically guided coeliac plexus block
Clinical Radiology (1992) 45, 401-403
Sonographically Guided Coeliac Plexus Block K. M. DAS and A. H. C H A P M A N
Department of Radiology, St James's University Hospital, Leeds' A sonographically guided anterior approach to block the coefiac plexus with local anaesthetic was attempted on supine patients before hepatobifiary interventioual procedures. A satisfactory block was achieved in 8/9 patients and the minor compfication of transient hypotension was encountered in two which were managed conservatively. Comparison with controls showed the coeliac block patients to require significantly less intravenous sedation and analgesia to control pain during these procedures. Das, K.M. & Chapman, A.H. (1992). Clinical Radiology 45, 401-403. Sonographically Guided Coeliac Plexus Block
Computed tomography (CT) and fluoroscopically guided anterior coeliac axis blocks have been successfully used for the alleviation of deep visceral pain related to various interventional hepatobiliary procedures (Whiteman et al., 1986). Unfortunately, every procedure cannot be performed in the CT room (Haaga et al., 1984) and with fluoroscopic guidance (Moore et al., 1981) the needle tip cannot always be precisely placed using bone landmarks due to normal variation in soft tissue anatomy. We describe the simple technique of US-guided coeliac axis block, which utilizes soft tissue landmarks and can be performed on the fluoroscopic table immediately before the interventional procedure.
both sides of the SMA root half the dose was injected on each side (Fig. 2). The time taken from inserting to removing the needle was recorded. An intercostal block was also performed to anaesthetize the skin and the parietal peritoneum. The intercostal block was made 5 cm posterior to the biliary puncture site. The needle was
MATERIALS AND M E T H O D S Over the last 6 months a total of nine US-guided coeliac axis blocks were performed on nine different patients who were candidates for various interventional procedures. The procedures for which the blocks were used included single-step internal-external percutaneous biliary drainage (3), percutaneous and endoscopic combined stent placement (2), external biliary drainage (1) and rotary lithotripsy (3). The block was performed with the patient supine on the fluoroscopy table. After pre-medication with 5 mg of IV diazepam, a skin puncture site in the epigastrium was selected with sonography. After sterile preparation and local skin anaesthesia, a 20 G 15 cm long spinal needle was advanced under US guidance (Fig. 1). In the first two cases the injection was made on both sides of the root of the superior mesenteric artery (SMA). In the remaining cases a single injection was made immediately anterior to the aorta between the roots of the coeliac axis and the superior mesenteric artery. The needle was passed through the left lobe of the liver, the stomach was avoided and in most patients it was possible to pass the needle above the pancreas, but in two the needle was passed through the pancreas and splenic vein. The position of the needle tip was checked by fluoroscopy and considered satisfactory if it was overlying the body of LI. The needle was aspirated to ensure that the tip was not in a blood vessel before injecting 40 ml of 1% lignocaine containing adrenaline at a concentration of 1:200 000 slowly over 1-2 min. In the two cases when the injection was placed on Correspondence to: A. H. Chapman, Consultant Radiologist, Department of Radiology, St James's University Hospital, Leeds LS9 7TF.
(a)
(b) Fig. 1-(a) Transverse sonogram at the level of the pancreas. Needle tip seen as a hyperechoic dot (arrow). (b) An ill-defined echogenic spot (arrow) to the right of the superior mesenteric artery (SMA) root represents an air bubble injected with the local anaesthetic. A, Aorta; S, SMA root; P, pancreas. Arrow heads indicate the needle and curved arrow, artefact.
402
CLINICAL RADIOLOGY
patient had an IV line and the blood pressure was monitored. After the procedure a Tepeat US examination was performed to check for any complications that might have developed as a result of the block, such as haemorrhage. To substantiate the usefulness of the coeliac block we also carried out control studies on nine randomly selected patients undergoing similar procedures. These procedures were started with an initial IV dose of 5 mg of diazepam, an intercostal block and an injection of local anaesthetic to the puncture site. Further sedation and analgesia were given when the patient complained of pain during the procedure and vital signs were monitored.
RESULTS
(a)
The results are shown in Table 1. Coeliac block provided excellent regional anaesthesia in eight patients and partial anaesthesia in one. In this patient rotary lithotripsy was being performed and as the procedure lasted 2 h supplemental IV pethidine was required. In one of the eight patients additional intercostal infiltration of local anaesthetic was required during an internal-external bitiary drainage procedure but the visceral pain relief from the coeliac block was satisfactory. Two patients experienced hypotension within 5 min of the coeliac block which was managed by increasing the rate of infusion of IV fluid and raising the lower end of the X-ray table. Both patients responded well to this conservative management. The additional time taken to perform the coeliac block in all cases was less than 5 min. In the control group all nine patients required systemic analgesia for comfortable completion of the procedure. Details are given in Table 2.
DISCUSSION
(b) Fig. 2 Coeliac block with bilateral injection of local anaesthetic mixed with 5 ml of Omnipaque at sides of the superior mesenteric artery (SMA) root. (a). Posterior-anterior view showing spread of contrast medium from the injection sites at the level of the L2 vertebra on each side of the aorta. (b) In the lateral view a curvilinear collection of contrast medium is seen anterior to the aorta and vertebral body (arrow). Subsequent injections were made between the coeliac axis and the S M A root at the level of L1.
advanced to touch the rib, withdrawn a little and advanced to just under the rib when 2 ml of the 1% lignocaine and adrenaline mixture was injected. The puncture site was also infiltrated with lignocaine and adrenaline, taking care that the total dose of lignocaine used did not exceed 50 mg. Following the procedure each patient was asked to grade the pain that was experienced as follows: grade 0, no pain; grade I, mild; grade II, moderate; and grade III, severe pain. Sedatives and narcotics were used as required and a record kept of the drugs and the amounts used. Each
Abdominal viscera, including the biliary tract, are innervated by splachnic nerves, which are derived from sympathetic ganglia (Moore, 1965). The coeliac plexus consists of a right and left coeliac ganglion lying anterior to the aorta between the coeliac axis and the root of the SMA. Kappis first described the coeliac axis block in 1919 and since then it has been used for treating intractable pain caused by pancreatitis and pancreatic carcinoma (Hegedus, 1979; Moore et al., 1981). In recent years this procedure has been infrequently performed using CT for guidance with a view to providing regional anaesthesia and has been performed with fluoroscopic guidance to provide anaesthesia for interventional procedures. As an alternative to CT and fluoroscopy, US can be used to guide the needle allowing soft tissue structures to be clearly identified so needle tip placement can be exact. We found that the time taken for the coeliac block, from needle insertion to withdrawal, was always less than 5 min. Although the anterior approach in the supine position is advantageous to both the interventionalist and to the patient, Lieberman et al. (l 988) reported a low success rate using this approach. In their series only 55% of the cases had complete analgesia compared with 75-95% with the posterior approach used by others (de Takats, 1927; Moore et al., 1981). Lieberman's group believe that the reasons for the poor results were anatomical variation in the position of the coeliac plexus and inadequate spread of local anaesthetic. They therefore advocate the use of a
403
SONOGRAPHICALLY GUIDED COELIAC PLEXUS BLOCK Table 1 - Comparative evaluation of pain in US-guided eoeliac block and control group
Procedure
No.
Pain level None (0)
Mild (I)
Moderate (II)
Severe (III)
Rotary lithotripsy Combined stent External biliary drain Int/Ext biliary drain Biliary stent
3 2 (5) 1 (I) 3 (2) 0 (1)
2 2 (3) 1 (1) 2 (1) 0 (I)
0 0 (2) 0 0 0
Total
9 (9)
7 (6)
0 (2)
2(1)
0
1
0
0 0
0 0
1 (1)
0
0
0
Figures for controls in brackets.
Table 2 - Comparative evaluation of intravenous drug required in the control group and coeliac block group
Procedure
No.
Drugs additional to 5 mg of diazepam Pethidine (mg)
Diazepam (mg)
50
100
5
10
Rotary lithotripsy Combined stent External biliary drain lnt/Ext biliary drain Biliary stent
0 (3) 5 (2) 1 (1) 2 (3) 1
0 (1) 3 1 2 1
0 2 0 0 0
0 4 0 2 0
0 1 0 0 0
Total
9 (9)
7 (1)
2
6
1
Figures for coeliac block patients in brackets.
larger volume (60 ml) of local anaesthetic with the addition of hyaluronidase or injecting at two vertebral levels to widen the distribution of the anaesthetic agent. We feel that with the use of ultrasound, the needle tip can be placed close to the anatomical location of the nerve plexus. As the coeliac plexus lies on both sides of the aorta, bilateral injection has been recommended but most authors now agree that a single accurately placed injection will usually produce an adequate block (Moore, 1965; Bonica, 1981; Carron et al., 1984). This has also been our experience as although in two of our patients we used bilateral injections (with success in one), in the remaining seven a successful result was achieved with a single injection. With wider use of fine-needle aspiration cytology it has become apparent that when structures such as the pancreas and major arteries and veins are transfixed, complications are rare (Lieberman et al., 1980). Nevertheless, sonography allows blood vessels to be identified and avoided. In two of our cases where no other safe route was available, the needle was passed through the retropancreatic splenic vein to reach the root of the SMA and both patients had a successful block without complication. Transient hypotension is a frequently noted complication seen in up to 80% of cases because of splanchnic vasodilatation (Moore, 1965) and such hypotension usually indicates a successful block. An IV line and constant blood pressure monitoring are essential but these patients usually respond quickly to IV fluid and leg raising. The m a x i m u m recommended dose of lignocaine is 200 nag or 500 mg when adrenaline is added to slow absorption (Deacon and Simpson, 1964). The total dose used in the coeliac group of patients approached 500 mg, although a small reduction in this dose could be achieved by omitting the intercostal block or the injection at the skin puncture site. It remains to be seen whether a satisfactory block can
be achieved using a smaller dose oflignocaine. Lignocaine overdosage causes excitement, disorientation and convulsions and it is believed that epileptic patients have a lower threshold for convulsions and so a lignocaine coeliac block is best avoided in this group (Deacon and Simpson, 1964). Inadvertent colonic puncture, pancreatic puncture and abscess formation are potential complications that were considered but the colon usually lies caudal to the site of coeliac block and puncture of the colon by a fine needle is thought to be harmless. Complications from fine-needle pancreatic puncture are rare (Lieberman et al., 1988). Abscess formation from contaminated local anaesthetic is another potential complication but can be avoided by providing antibiotic cover. Coeliac block is easily and quickly performed using US guidance; compared with the control group the coeliac block group required significantly less systemic analgesia, enabling the procedure to be performed on an awake and cooperative patient.
REFERENCES
Bonica, JJ (1981). Sympathetic Nerve Blocks j o t Pain Diagnosis" and Therapy, pp. 78 90. Breon Laboratories, New York. Carton, H, Korbon, GA & Rowlingson, JC (1984). Regional Anesthesia, pp. 138 141. Grune & Stratton, New York. Deacon, ARD & Simpson, WT (1964). Fatal reactions to lignocaine. Anaesthesia, 19, 217-221. de Takats, G (1927). Splanchnic anaesthesia: a critical review of the theory and practice of this method. Surgery Gynecology and Obstetrics, 44, 501 519. Haaga, JR, Kori, SH, Eatwood, DW & Borokowski, GP (1984). Improved technique for CT guided celiac ganglia block. American JournalofRoentgenology, 142, 1201 1204. Hegedus, V (1979). Relief of pancreatic pain by radiography-guided block. American Journal of Roentgenology, 133, 1101-1103. Kappis, M (1919). Sensibilit/it and Lokale Anaeesthesie im Chirurgischen Gebeit der Bauchkule mit Besonderer Ber/icksichtigung der Splanchicus Anasthesie. Beitr Klin Chir, 115, 151 175. Lieberman, RP, Crummy, AB & Matallana, RH (1980). Invasive procedures in pancreatic disease. Seminars in Ultrasound, Computed Tomography and Magnetic Resonance, 1, 192 208. Lieberman, RP, Nance, PN & Cuka, DJ (1988). Anterior approach to celiac plexus block during interventional biliary procedures. Radiology, 167, 562-564. Moore, DC (1965). Regional Block." A Handbook for Use in Clinical Practice of Medicine and Surgery, pp. 145 162. Thomas, Springfield, IL. Moore, DC, Bush, LC & Burnett, LC (1981). Celiac plexus block: a roentgenographic, anatomic study of technique, and spread of solution in patients and corpses. Anesthesia and Analgesia, 60, 369--379. Whiteman, MS, Rosenberg, H, Haskin, PH & Teplick, SK (1986). Celiac plexus block for interventional radiology. Radiology, 161,836 838.
Sonographically Guided Coeliac Plexus Block K. M. DAS and A. H. C H A P M A N
Department of Radiology, St James's University Hospital, Leeds' A sonographically guided anterior approach to block the coefiac plexus with local anaesthetic was attempted on supine patients before hepatobifiary interventioual procedures. A satisfactory block was achieved in 8/9 patients and the minor compfication of transient hypotension was encountered in two which were managed conservatively. Comparison with controls showed the coeliac block patients to require significantly less intravenous sedation and analgesia to control pain during these procedures. Das, K.M. & Chapman, A.H. (1992). Clinical Radiology 45, 401-403. Sonographically Guided Coeliac Plexus Block
Computed tomography (CT) and fluoroscopically guided anterior coeliac axis blocks have been successfully used for the alleviation of deep visceral pain related to various interventional hepatobiliary procedures (Whiteman et al., 1986). Unfortunately, every procedure cannot be performed in the CT room (Haaga et al., 1984) and with fluoroscopic guidance (Moore et al., 1981) the needle tip cannot always be precisely placed using bone landmarks due to normal variation in soft tissue anatomy. We describe the simple technique of US-guided coeliac axis block, which utilizes soft tissue landmarks and can be performed on the fluoroscopic table immediately before the interventional procedure.
both sides of the SMA root half the dose was injected on each side (Fig. 2). The time taken from inserting to removing the needle was recorded. An intercostal block was also performed to anaesthetize the skin and the parietal peritoneum. The intercostal block was made 5 cm posterior to the biliary puncture site. The needle was
MATERIALS AND M E T H O D S Over the last 6 months a total of nine US-guided coeliac axis blocks were performed on nine different patients who were candidates for various interventional procedures. The procedures for which the blocks were used included single-step internal-external percutaneous biliary drainage (3), percutaneous and endoscopic combined stent placement (2), external biliary drainage (1) and rotary lithotripsy (3). The block was performed with the patient supine on the fluoroscopy table. After pre-medication with 5 mg of IV diazepam, a skin puncture site in the epigastrium was selected with sonography. After sterile preparation and local skin anaesthesia, a 20 G 15 cm long spinal needle was advanced under US guidance (Fig. 1). In the first two cases the injection was made on both sides of the root of the superior mesenteric artery (SMA). In the remaining cases a single injection was made immediately anterior to the aorta between the roots of the coeliac axis and the superior mesenteric artery. The needle was passed through the left lobe of the liver, the stomach was avoided and in most patients it was possible to pass the needle above the pancreas, but in two the needle was passed through the pancreas and splenic vein. The position of the needle tip was checked by fluoroscopy and considered satisfactory if it was overlying the body of LI. The needle was aspirated to ensure that the tip was not in a blood vessel before injecting 40 ml of 1% lignocaine containing adrenaline at a concentration of 1:200 000 slowly over 1-2 min. In the two cases when the injection was placed on Correspondence to: A. H. Chapman, Consultant Radiologist, Department of Radiology, St James's University Hospital, Leeds LS9 7TF.
(a)
(b) Fig. 1-(a) Transverse sonogram at the level of the pancreas. Needle tip seen as a hyperechoic dot (arrow). (b) An ill-defined echogenic spot (arrow) to the right of the superior mesenteric artery (SMA) root represents an air bubble injected with the local anaesthetic. A, Aorta; S, SMA root; P, pancreas. Arrow heads indicate the needle and curved arrow, artefact.
402
CLINICAL RADIOLOGY
patient had an IV line and the blood pressure was monitored. After the procedure a Tepeat US examination was performed to check for any complications that might have developed as a result of the block, such as haemorrhage. To substantiate the usefulness of the coeliac block we also carried out control studies on nine randomly selected patients undergoing similar procedures. These procedures were started with an initial IV dose of 5 mg of diazepam, an intercostal block and an injection of local anaesthetic to the puncture site. Further sedation and analgesia were given when the patient complained of pain during the procedure and vital signs were monitored.
RESULTS
(a)
The results are shown in Table 1. Coeliac block provided excellent regional anaesthesia in eight patients and partial anaesthesia in one. In this patient rotary lithotripsy was being performed and as the procedure lasted 2 h supplemental IV pethidine was required. In one of the eight patients additional intercostal infiltration of local anaesthetic was required during an internal-external bitiary drainage procedure but the visceral pain relief from the coeliac block was satisfactory. Two patients experienced hypotension within 5 min of the coeliac block which was managed by increasing the rate of infusion of IV fluid and raising the lower end of the X-ray table. Both patients responded well to this conservative management. The additional time taken to perform the coeliac block in all cases was less than 5 min. In the control group all nine patients required systemic analgesia for comfortable completion of the procedure. Details are given in Table 2.
DISCUSSION
(b) Fig. 2 Coeliac block with bilateral injection of local anaesthetic mixed with 5 ml of Omnipaque at sides of the superior mesenteric artery (SMA) root. (a). Posterior-anterior view showing spread of contrast medium from the injection sites at the level of the L2 vertebra on each side of the aorta. (b) In the lateral view a curvilinear collection of contrast medium is seen anterior to the aorta and vertebral body (arrow). Subsequent injections were made between the coeliac axis and the S M A root at the level of L1.
advanced to touch the rib, withdrawn a little and advanced to just under the rib when 2 ml of the 1% lignocaine and adrenaline mixture was injected. The puncture site was also infiltrated with lignocaine and adrenaline, taking care that the total dose of lignocaine used did not exceed 50 mg. Following the procedure each patient was asked to grade the pain that was experienced as follows: grade 0, no pain; grade I, mild; grade II, moderate; and grade III, severe pain. Sedatives and narcotics were used as required and a record kept of the drugs and the amounts used. Each
Abdominal viscera, including the biliary tract, are innervated by splachnic nerves, which are derived from sympathetic ganglia (Moore, 1965). The coeliac plexus consists of a right and left coeliac ganglion lying anterior to the aorta between the coeliac axis and the root of the SMA. Kappis first described the coeliac axis block in 1919 and since then it has been used for treating intractable pain caused by pancreatitis and pancreatic carcinoma (Hegedus, 1979; Moore et al., 1981). In recent years this procedure has been infrequently performed using CT for guidance with a view to providing regional anaesthesia and has been performed with fluoroscopic guidance to provide anaesthesia for interventional procedures. As an alternative to CT and fluoroscopy, US can be used to guide the needle allowing soft tissue structures to be clearly identified so needle tip placement can be exact. We found that the time taken for the coeliac block, from needle insertion to withdrawal, was always less than 5 min. Although the anterior approach in the supine position is advantageous to both the interventionalist and to the patient, Lieberman et al. (l 988) reported a low success rate using this approach. In their series only 55% of the cases had complete analgesia compared with 75-95% with the posterior approach used by others (de Takats, 1927; Moore et al., 1981). Lieberman's group believe that the reasons for the poor results were anatomical variation in the position of the coeliac plexus and inadequate spread of local anaesthetic. They therefore advocate the use of a
403
SONOGRAPHICALLY GUIDED COELIAC PLEXUS BLOCK Table 1 - Comparative evaluation of pain in US-guided eoeliac block and control group
Procedure
No.
Pain level None (0)
Mild (I)
Moderate (II)
Severe (III)
Rotary lithotripsy Combined stent External biliary drain Int/Ext biliary drain Biliary stent
3 2 (5) 1 (I) 3 (2) 0 (1)
2 2 (3) 1 (1) 2 (1) 0 (I)
0 0 (2) 0 0 0
Total
9 (9)
7 (6)
0 (2)
2(1)
0
1
0
0 0
0 0
1 (1)
0
0
0
Figures for controls in brackets.
Table 2 - Comparative evaluation of intravenous drug required in the control group and coeliac block group
Procedure
No.
Drugs additional to 5 mg of diazepam Pethidine (mg)
Diazepam (mg)
50
100
5
10
Rotary lithotripsy Combined stent External biliary drain lnt/Ext biliary drain Biliary stent
0 (3) 5 (2) 1 (1) 2 (3) 1
0 (1) 3 1 2 1
0 2 0 0 0
0 4 0 2 0
0 1 0 0 0
Total
9 (9)
7 (1)
2
6
1
Figures for coeliac block patients in brackets.
larger volume (60 ml) of local anaesthetic with the addition of hyaluronidase or injecting at two vertebral levels to widen the distribution of the anaesthetic agent. We feel that with the use of ultrasound, the needle tip can be placed close to the anatomical location of the nerve plexus. As the coeliac plexus lies on both sides of the aorta, bilateral injection has been recommended but most authors now agree that a single accurately placed injection will usually produce an adequate block (Moore, 1965; Bonica, 1981; Carron et al., 1984). This has also been our experience as although in two of our patients we used bilateral injections (with success in one), in the remaining seven a successful result was achieved with a single injection. With wider use of fine-needle aspiration cytology it has become apparent that when structures such as the pancreas and major arteries and veins are transfixed, complications are rare (Lieberman et al., 1980). Nevertheless, sonography allows blood vessels to be identified and avoided. In two of our cases where no other safe route was available, the needle was passed through the retropancreatic splenic vein to reach the root of the SMA and both patients had a successful block without complication. Transient hypotension is a frequently noted complication seen in up to 80% of cases because of splanchnic vasodilatation (Moore, 1965) and such hypotension usually indicates a successful block. An IV line and constant blood pressure monitoring are essential but these patients usually respond quickly to IV fluid and leg raising. The m a x i m u m recommended dose of lignocaine is 200 nag or 500 mg when adrenaline is added to slow absorption (Deacon and Simpson, 1964). The total dose used in the coeliac group of patients approached 500 mg, although a small reduction in this dose could be achieved by omitting the intercostal block or the injection at the skin puncture site. It remains to be seen whether a satisfactory block can
be achieved using a smaller dose oflignocaine. Lignocaine overdosage causes excitement, disorientation and convulsions and it is believed that epileptic patients have a lower threshold for convulsions and so a lignocaine coeliac block is best avoided in this group (Deacon and Simpson, 1964). Inadvertent colonic puncture, pancreatic puncture and abscess formation are potential complications that were considered but the colon usually lies caudal to the site of coeliac block and puncture of the colon by a fine needle is thought to be harmless. Complications from fine-needle pancreatic puncture are rare (Lieberman et al., 1988). Abscess formation from contaminated local anaesthetic is another potential complication but can be avoided by providing antibiotic cover. Coeliac block is easily and quickly performed using US guidance; compared with the control group the coeliac block group required significantly less systemic analgesia, enabling the procedure to be performed on an awake and cooperative patient.
REFERENCES
Bonica, JJ (1981). Sympathetic Nerve Blocks j o t Pain Diagnosis" and Therapy, pp. 78 90. Breon Laboratories, New York. Carton, H, Korbon, GA & Rowlingson, JC (1984). Regional Anesthesia, pp. 138 141. Grune & Stratton, New York. Deacon, ARD & Simpson, WT (1964). Fatal reactions to lignocaine. Anaesthesia, 19, 217-221. de Takats, G (1927). Splanchnic anaesthesia: a critical review of the theory and practice of this method. Surgery Gynecology and Obstetrics, 44, 501 519. Haaga, JR, Kori, SH, Eatwood, DW & Borokowski, GP (1984). Improved technique for CT guided celiac ganglia block. American JournalofRoentgenology, 142, 1201 1204. Hegedus, V (1979). Relief of pancreatic pain by radiography-guided block. American Journal of Roentgenology, 133, 1101-1103. Kappis, M (1919). Sensibilit/it and Lokale Anaeesthesie im Chirurgischen Gebeit der Bauchkule mit Besonderer Ber/icksichtigung der Splanchicus Anasthesie. Beitr Klin Chir, 115, 151 175. Lieberman, RP, Crummy, AB & Matallana, RH (1980). Invasive procedures in pancreatic disease. Seminars in Ultrasound, Computed Tomography and Magnetic Resonance, 1, 192 208. Lieberman, RP, Nance, PN & Cuka, DJ (1988). Anterior approach to celiac plexus block during interventional biliary procedures. Radiology, 167, 562-564. Moore, DC (1965). Regional Block." A Handbook for Use in Clinical Practice of Medicine and Surgery, pp. 145 162. Thomas, Springfield, IL. Moore, DC, Bush, LC & Burnett, LC (1981). Celiac plexus block: a roentgenographic, anatomic study of technique, and spread of solution in patients and corpses. Anesthesia and Analgesia, 60, 369--379. Whiteman, MS, Rosenberg, H, Haskin, PH & Teplick, SK (1986). Celiac plexus block for interventional radiology. Radiology, 161,836 838.