Annals of Surgical Oncology Cite Track
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Allen, P. J.
Right arrow Articles by Fong, Y.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Allen, P. J.
Right arrow Articles by Fong, Y.
Related Collections
Right arrow Surgery
Annals of Surgical Oncology 9:875-880 (2002)
© 2002 Society of Surgical Oncology

ORIGINAL ARTICLES

The Management of Variant Arterial Anatomy During Hepatic Arterial Infusion Pump Placement

Peter J. Allen, MD, Alexander Stojadinovic, MD, Leah Ben-Porat, MS, Mithat Gonen, PhD, David Kooby, MD, Leslie Blumgart, MD, Philip Paty, MD and Yuman Fong, MD

From the Departments of Surgery (PJA, AS, DK, LB, PP, YF) and Biostatistics (LBP, MG), Memorial Sloan-Kettering Cancer Center, New York, New York.

Correspondence: Address correspondence and reprint requests to: Yuman Fong, MD, Department of Surgery, Memorial Sloan-Kettering Cancer Center, 1275 York Ave., New York, NY 10021; Fax: 212-639-4031; E-mail: fongy{at}mskcc.org


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Background: The success of hepatic arterial infusion pump (HAIP) placement in patients with variant arterial anatomy has not been well described.

Methods: Patients who underwent HAIP placement over a 5-year time period were evaluated. Arterial- and catheter-related pump complication rates and pump survival were compared between patients with normal and variant arterial anatomy.

Results: Pumps were placed in 265 patients. Variant anatomy was present in 98 (37%) patients. The presence of variant versus normal anatomy did not increase pump complication rates (8% vs. 4%; P = .18) or decrease pump survival (P = .12). In all patients with an isolated variant right or left hepatic artery (n = 56), ligation of the variant vessel and cannulation of the gastroduodenal artery (GDA) resulted in complete hepatic perfusion and no pump complications. Cannulation of vessels other than the GDA (n = 22) was associated with increased pump complication rates (27% vs. 4%; P = .0001) and decreased pump survival (P = .002).

Conclusions: In this study, HAIP placement in patients with variant anatomy was not associated with increased pump complication rates or decreased pump survival. An optimal strategy for managing variant anatomy is to ligate isolated variant vessels and cannulate the GDA.

Key Words: Liver • Perfusion • Chemotherapy • Anatomy


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
The success of hepatic arterial infusion pump chemotherapy is dependent on careful patient selection and technical expertise in pump placement. The technical goals to be achieved during pump placement are 3-fold.1 First, the catheter should be inserted so that infusion with the chemotherapeutic agent results in complete hepatic perfusion. Second, arterial branches distal to the catheter and proximal to the liver should be ligated to prevent extrahepatic perfusion. Third, the catheter should be placed to minimize turbulent flow and thus allow long-term patency of both the catheter and the artery through which the infusion will be performed.

The technique for pump placement in the setting of normal arterial anatomy has been well described.1–3 After a cholecystectomy is performed, the common hepatic artery is mobilized 1 cm proximal and distal to the origin of the gastroduodenal artery (GDA). The right gastric artery is then ligated, and an arteriotomy is made in the GDA. The catheter is then carefully inserted up to, but not beyond, the junction of the GDA and common hepatic artery. Finally, after securing the catheter, fluoroscein or methylene blue is injected into the port, and perfusion of the liver is then confirmed.

A variety of techniques for pump placement have been described in the setting of variant arterial anatomy.2,3 Early studies that evaluated the dual-lumen Infusaid pump (Infusaid Corporation, Sharon, MA) described cannulation of both the GDA and variant vessel when either an accessory or replaced hepatic artery was encountered. This technique is no longer feasible because the dual-lumen pump is no longer produced. Others have advocated ligation of the variant vessel and cannulation of the GDA. Although no data have been published regarding the efficacy of this approach, studies have shown significant cross-perfusion of the liver after ligation of the blood supply to one hepatic lobe or segment.4 Additional techniques have also been described for pump placement in patients with variation in the origin of the GDA. When this anatomical variant is encountered, techniques described for catheter insertion include cannulation of the splenic artery, cannulation of the GDA, and cannulation of the hepatic artery.3

The goal of this study was to evaluate whether or not patients with variant arterial anatomy experience increased rates of catheter-related complications after pump placement. In addition, we wanted to identify specific techniques used at our institution for pump placement in patients with variant anatomy and to document the success or failure of these specific techniques.


   METHODS
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
All patients who underwent hepatic arterial infusion pump placement between December 1996 and May 2001 were identified. Patient-, tumor-, and treatment-related variables were then reviewed and entered into a computer database.

Arterial anatomy was determined in each patient through review of preoperative imaging studies and operative reports. Preoperative imaging studies to define anatomy included both conventional arteriography and computed tomographic arteriography. Presently, the most common study used at our institution is the computed tomography arteriogram, which provides excellent definition of arterial anatomy. Variant hepatic arterial anatomy was recorded as being from the left or right hepatic arterial system, and variant hepatic vessels were further categorized as being accessory or replaced. The origin of the GDA was also recorded. This was categorized as normal when the GDA arose from the common hepatic artery proximal to the takeoff of the left and right hepatic arteries. In this study, variant sites of origin of the GDA included the left hepatic artery, the right hepatic artery, and a trifurcated origin with the left and right hepatic artery.

Patients underwent pump placement either as an isolated operative procedure or in conjunction with a colon resection or liver resection. When a liver resection was performed, the extent of resection was recorded. After insertion of the pump, perfusion of the liver was checked in most cases with a bolus injection through the pump by using either methylene blue or fluoroscein. Perfusion of the liver was also checked after surgery in all patients with a macroaggregated albumin scan. The results of these scans were recorded as normal (complete hepatic perfusion with no extrahepatic perfusion) or abnormal (incomplete hepatic perfusion or extrahepatic perfusion).

All complications related to both the operation and the pump were recorded. Pump-related complications were categorized as pocket, pump, catheter, or arterial. Because the primary goal of this study was to evaluate the influence of anatomical variants on the success of pump placement, only catheter- and arterial-related pump complications were included in the analysis of complications. Pocket complications (pocket seroma, pocket infection, and so on) and pump complications (pump malfunction) were not considered to be influenced by arterial anatomy and were not included in the analysis.

Similarly, in the analysis of pump survival, only catheter- and arterial-related pump complications that resulted in a nonfunctional pump were considered as events. When a pump became nonfunctional secondary to a catheter- or arterial-related pump complication and could not be rendered functional with either an interventional or operative procedure, then an event was recorded. Patients were censored when the pump was functional at last follow-up or was nonfunctional secondary to a pocket or pump complication.

The associations between the catheter-specific pump complications and patient-, tumor-, and treatment-related variables were assessed by using Fisher’s exact test. Exact logistic regression was used to assess the influence of arterial anatomy and GDA cannulation on catheter complications in a multivariate setting. Catheter-specific pump survival probabilities were estimated with the Kaplan-Meier method and were compared by using the log-rank test.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Over the 5-year time period, 265 patients underwent hepatic arterial pump placement. The median age of this group of patients was 58 years, and 167 (63%) of the 265 patients were men. Variant arterial anatomy was present in 98 patients (37%), the most common site of variance was within the left hepatic arterial system (n = 47), and the majority of patients with variant anatomy (87%) had only a single variant vessel (Table 1). During the procedure when the pump was placed, half of all patients underwent a liver resection, and 15% of patients underwent colon resection. Once the catheter had been inserted, an intraoperative perfusion scan with either fluoroscein or methylene blue was performed in 214 patients (81%), with 210 of these patients having complete hepatic distribution of tracer (normal scan).


View this table:
[in this window]
[in a new window]
  		TABLE 1. Patient, anatomical, and treatment-related data
 
Pump-related complications occurred in 12% of patients, with half of these complications (6%) being catheter or arterial related. The only patient-related, anatomical, or treatment-related factor found to be associated with a significantly higher rate of catheter- or arterial-specific pump complications was cannulation of a vessel other than the GDA (Table 2). When a vessel other than the GDA was cannulated (n = 22), a complication rate of 27% was observed. The vessels cannulated within this group of 22 patients included the right hepatic artery (n = 7), common hepatic artery (n = 5), cystic artery (n = 3), left hepatic artery (n = 3), splenic artery (n = 3), and right gastric artery (n = 1). GDA cannulation was significant in the multivariate exact logistic regression model controlling for arterial anatomy (P = .0027). GDA cannulation reduced the odds of catheter or arterial pump complications by 9-fold (relative risk, 8.9; 95% confidence interval, 2.1–37.4).


View this table:
[in this window]
[in a new window]
  		TABLE 2. Patient, anatomical, and treatment-related factors and their association with catheter-specific complications (n = 265)
 
Patients with variant arterial anatomy, as a group, did not experience a significantly higher rate of catheter- or arterial-specific pump complications, although most patients (17 of 22; 77%) whose catheter was placed in a vessel other than the GDA had variant anatomy. A vessel other than the GDA was cannulated in a much smaller proportion of patients (3.0%) with normal arterial anatomy than with variant anatomy (17.3%). This difference was statistically significant (P = .0001). In four of the five patients with normal anatomy who had non-GDA cannulation, a liver resection was performed, and the catheter was inserted into the stump of the ligated hepatic artery. The remaining patient with normal anatomy who underwent non-GDA cannulation had retrograde flow within the GDA, and the common hepatic artery was cannulated. Patients with multiple variant vessels experienced a greater rate of catheter complications as compared with patients with a single variant vessel (23% vs. 6%; P = .07).

The median pump follow-up on all patients was 14 months, and the overall 2-year catheter-specific pump survival was 97%. Catheter- or arterial-related complications that rendered the pump initially nonfunctional occurred in 15 cases; however, in 7 of these cases, radiological or surgical intervention was able to render them functional. Therefore, at the time of last follow-up, only eight pumps were nonfunctional secondary to catheter- or arterial-related complications. The one factor associated with decreased pump survival was cannulation of a vessel other than the GDA (log-rank P = .002). Patients with variant anatomy did not experience significantly different pump survival compared with patients with normal anatomy (log-rank P = .12).

The technical approach to patients with variant anatomy was very similar. Patients with isolated variant hepatic arterial vessels, whether accessory or replaced, underwent ligation of the variant vessel and cannulation of the GDA in 52 of the 56 cases, and no catheter- or arterial-related complications were observed (Table 3). A similar approach was also used in patients with isolated variance in the origin of the GDA. Cannulation of the GDA was performed in 23 of the 28 patients, with only 2 patients (9%) experiencing catheter- or arterial-related pump complications. When the GDA arose from one of the hepatic vessels and was used as the vessel for cannulation, ligation was performed of the hepatic vessel distal to the GDA takeoff. In the setting of trifurcated anatomy, hepatic vessels were ligated in only two cases and were not associated with perfusion abnormalities. All three patients with a variant origin of the GDA and who had the splenic artery cannulated experienced catheter-related complications. Extrahepatic perfusion was detected in all three patients on postoperative perfusion scanning.


View this table:
[in this window]
[in a new window]
  		TABLE 3. Specific treatment strategies and complications for patients with isolated variant vessels
 
As noted previously, patients with multiple variant vessels were at an increased risk of developing catheter-related complications. As similar approach was used for these 13 patients, with the GDA being cannulated in 7 cases. The specific variants encountered, operative approach, and complications are listed in Table 4.


View this table:
[in this window]
[in a new window]
  		TABLE 4. Specific treatment strategies and complications for patients with multiple variant vessels
 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
Hepatic arterial chemotherapy has been shown to be a feasible treatment modality for patients with colorectal liver metastases. Directed intra-arterial chemotherapy maximizes the antitumor therapeutic effect while minimizing adverse systemic side effects. Data from seven prospective randomized trials comparing intra-arterial therapy with systemic therapy for the treatment of unresectable metastatic disease have demonstrated significantly improved treatment response with intra-arterial therapy.5–11 In addition, a recent prospective, randomized trial from our own institution which evaluated intra-arterial chemotherapy after complete resection of hepatic metastases reported significantly increased 2-year survival in patients who received combined hepatic arterial and systemic treatment compared with systemic treatment alone.12

The success of hepatic arterial chemotherapy is dependent on careful patient selection and surgical expertise in pump placement. The catheter must be inserted so there is adequate distribution of chemotherapy to the entire liver without perfusion of extrahepatic tissues. In addition, the catheter tip must not create significant turbulence in the hepatic artery, or the long-term patency of the catheter and the cannulated artery will be diminished.

Hepatic arterial anatomy is considered normal when the common hepatic artery originates from the celiac axis and the GDA arises from the common hepatic artery before the bifurcation of the right and left hepatic arteries. Anatomical variations of the hepatic arterial supply occur in 30% to 50% of patients in the form of accessory or replaced hepatic arteries (Table 5). These hepatic arterial anomalies pose technical challenges at the time of pump placement, and multiple techniques have been described for placement of hepatic arterial pumps in patients with variant anatomy. These techniques have included side arterial cannulation, cannulation of both the GDA and the accessory/replaced vessel with the use of two pumps or with the use of dual-catheter implantable pumps, and ligation of the variant vessel and cannulation of the GDA.1,7,13,14


View this table:
[in this window]
[in a new window]
  		TABLE 5. Studies reporting the prevalence of variant hepatic arterial anatomy and rates of specific anatomical variants
 
Rayner et al., in 1986, reported on 15 patients with variant arterial anatomy who had undergone ligation of the variant lobar artery and single-catheter cannulation of the remaining lobar artery.15 In this study, all 15 patients were found to have complete hepatic perfusion on their postoperative perfusion scans. Cohen et al.4 also evaluated the effect of variant hepatic artery ligation on the completeness of regional infusion in a series of 13 patients with variant arterial anatomy. In this study, ligation of a right variant vessel and cannulation of the remaining lobar vessel resulted in complete hepatic perfusion in 90% of patients. However, incomplete crossover perfusion was demonstrated in 5 (48%) of 13 of those undergoing ligation of a replaced left hepatic artery. This incomplete perfusion after ligation of a replaced left hepatic artery has been thought to be secondary to difficulty in blood flow passing beyond the umbilical fissure to the left lateral segment. In total, these results support previous findings that suggest that hepatic lobar arteries are not end arteries and that prompt reconstitution of flow to the contralateral hepatic lobe through collateral vessels develops after occlusion of a variant vessel.16,17 In the setting of variant anatomy, ligation of the variant vessel and perfusion through the remaining hepatic arteries eliminates the need for the implantation of two separate infusion pumps or for the use of dual-lumen pumps (which are no longer in production).

In cases of incomplete perfusion of the liver after ligation of a replaced or accessory left hepatic artery, collateral blood flow will usually develop within 2 to 4 weeks to the hypoperfused lobe of the liver in nearly all patients.18,19 Curley et al.19 identified variant arterial anatomy (replaced and accessory right and left hepatic arteries) in 66 of 180 patients with hepatic arterial infusion devices that was successfully managed with ligation of the variant vessel and single-catheter cannulation of the GDA in all but 1 patient. Incomplete crossover perfusion was identified in seven patients after surgery; repeated perfusion scans performed 2 to 4 weeks after surgery confirmed complete bilobar hepatic perfusion in all cases. In our study, only 1 of 10 patients who underwent ligation of a replaced left hepatic artery was found to have incomplete hepatic perfusion. These results may be secondary to the fact that we do not perform the perfusion scan until just before discharge, which may allow time for crossover perfusion to develop. The median postoperative day on which the scan was performed was day 5.

Overall operative morbidity and pump-related complications were 20% and 12%, respectively, in this study, and patients with variant hepatic arteries were no more likely to experience these complications than those with standard hepatic arterial anatomy. The only patient, anatomical, or technical factor found to significantly correlate with pump-related morbidity was cannulation of an artery other than the GDA. Patients with variant arterial anatomy were more likely to experience pump complications if a vessel other than the GDA was cannulated (28% vs. 4%; P = .001). In addition, patients with variant arterial anatomy who had multiple variant vessels experienced significantly increased pump-related complications as compared with patients with variant anatomy and only a single variant vessel (23% vs. 6%; P = .04).

Our study lends further support to the hypothesis that ligation of the variant hepatic artery and cannulation of the GDA does not adversely affect pump survival. In our analysis, the only factor that correlated significantly with reduced pump survival was cannulation of an artery other than the GDA. Multiple vessel cannulation is rarely indicated in the patient with variant arterial anatomy. Ligation of the variant lobar artery and cannulation of the GDA will result in complete hepatic perfusion via translobar collateral arteries. In this study, this approach resulted in few pump-related complications and excellent pump survival.

In conclusion, the results from this study suggest that patients with variant arterial anatomy do not experience increased rates of pump complications or decreased pump survival. In all patients, cannulation of the GDA will result in excellent pump survival and minimal pump complications. In patients with isolated variant vessels, ligation of the variant vessel and cannulation of the GDA will result in excellent pump survival and minimal pump complications.


   Footnotes
 
In this study of 265 patients undergoing hepatic arterial infusion pump placement, the presence of variant arterial anatomy did not increase pump complication rates. In patients with variant anatomy, cannulation of a vessel other than the gastroduodenal artery was associated with increased pump complications and decreased pump survival. An optimal strategy for managing variant anatomy is to ligate isolated variant vessels and cannulate the gastroduodenal artery.

Received for publication March 15, 2002. Accepted for publication July 5, 2002.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Daly JM, Kemeny N, Oderman P, Botet J. Long-term hepatic arterial infusion chemotherapy. Anatomic considerations, operative technique, and treatment morbidity. Arch Surg 1984; 119: 936–41.[Abstract]
  2. Kemeny MM. The surgical aspects of the totally implantable hepatic artery infusion pump. Arch Surg 2001; 136: 348–52.[Abstract/Free Full Text]
  3. Kemeny NE, Sigurdson ER. Intra-arterial chemotherapy for liver tumors. In: Blumgart LH, ed. Surgery of the Liver and Biliary Tract. Edinburgh: Churchill Livingstone, 1994: 1617–36.
  4. Cohen AM, Higgins J, Waltman AC, Athanasoulis C, McKusick K. Effect of ligation of variant hepatic arterial structures on the completeness of regional chemotherapy infusion. Am J Surg 1987; 153: 378–80.[CrossRef][Medline]
  5. Kemeny N, Daly J, Reichman B, Geller N, Botet J, Oderman P. Intrahepatic or systemic infusion of fluorodeoxyuridine in patients with liver metastases from colorectal carcinoma. A randomized trial. Ann Intern Med 1987; 107: 459–65.
  6. Chang AE, Schneider PD, Sugarbaker PH, Simpson C, Culnane M, Steinberg SM. A prospective randomized trial of regional versus systemic continuous 5-fluorodeoxyuridine chemotherapy in the treatment of colorectal liver metastases. Ann Surg 1987; 206: 685–93.[Medline]
  7. Hohn DC, Stagg RJ, Friedman MA, et al. A randomized trial of continuous intravenous versus hepatic intraarterial floxuridine in patients with colorectal cancer metastatic to the liver: the Northern California Oncology Group trial. J Clin Oncol 1989; 7: 1646–54.[Abstract]
  8. Wagman LD, Kemeny MM, Leong L, et al. A prospective, randomized evaluation of the treatment of colorectal cancer metastatic to the liver. J Clin Oncol 1990; 8: 1885–93.[Abstract]
  9. Martin JK Jr, O’Connell MJ, Wieand HS, et al. Intra-arterial floxuridine vs systemic fluorouracil for hepatic metastases from colorectal cancer. A randomized trial. Arch Surg 1990; 125: 1022–7.[Abstract]
  10. Rougier P, Laplanche A, Huguier M, et al. Hepatic arterial infusion of floxuridine in patients with liver metastases from colorectal carcinoma: long-term results of a prospective randomized trial. J Clin Oncol 1992; 10: 1112–8.[Abstract]
  11. Allen-Mersh TG, Glover C, Fordy C, Mathur P, Quinn H. Randomized trial of regional plus systemic fluorinated pyrimidine compared with systemic fluorinated pyrimidine in treatment of colorectal liver metastases. Eur J Surg Oncol 2000; 26: 468–73.[CrossRef][Medline]
  12. Kemeny N, Huang Y, Cohen AM, et al. Hepatic arterial infusion of chemotherapy after resection of hepatic metastases from colorectal cancer. N Engl J Med 1999; 341: 2039–48.[Abstract/Free Full Text]
  13. Kemeny MM, Hogan JM, Goldberg DA, et al. Continuous hepatic artery infusion with an implantable pump: problems with hepatic artery anomalies. Surgery 1986; 99: 501–4.[Medline]
  14. Hohn DC, Rayner AA, Economou JS, Ignoffo RJ, Lewis BJ, Stagg RJ. Toxicities and complications of implanted pump hepatic arterial and intravenous floxuridine infusion. Cancer 1986; 57: 465–70.[CrossRef][Medline]
  15. Rayner AA, Kerlan RK, Stagg RJ, Price DC, Hohn DC. Total hepatic arterial perfusion after occlusion of variant lobar vessels: implications for hepatic arterial chemotherapy. Surgery 1986; 99: 708–15.[Medline]
  16. Redman HC, Reuter SR. Arterial collaterals in the liver hilus. Radiology 1970; 94: 575–9.[Medline]
  17. Koehler RE, Korobkin M, Lewis F. Arteriographic demonstration of collateral arterial supply to the liver after hepatic artery ligation. Radiology 1975; 117: 49–54.[Abstract]
  18. Chuang VP, Wallace S. Hepatic arterial redistribution for intraarterial infusion of hepatic neoplasms. Radiology 1980; 135: 295–9.[Abstract/Free Full Text]
  19. Curley SA, Chase JL, Roh MS, Hohn DC. Technical considerations and complications associated with the placement of 180 implantable hepatic arterial infusion devices. Surgery 1993; 114: 928–35.[Medline]
  20. Michels NA. Newer anatomy of the liver and its variant blood supply and collateral circulation. Am J Surg 1966; 112: 337–47.[CrossRef][Medline]



This article has been cited by other articles:


Home page
 RadioGraphicsHome page
O. A. Catalano, A. H. Singh, R. N. Uppot, P. F. Hahn, C. R. Ferrone, and D. V. Sahani
Vascular and Biliary Variants in the Liver: Implications for Liver Surgery
RadioGraphics, March 1, 2008; 28(2): 359 - 378.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Roentgenol.Home page
A. M. Covey, L. A. Brody, G. I. Getrajdman, C. T. Sofocleous, and K. T. Brown
Incidence, Patterns, and Clinical Relevance of Variant Portal Vein Anatomy
Am. J. Roentgenol., October 1, 2004; 183(4): 1055 - 1064.
[Abstract] [Full Text] [PDF]

Home page
 Ann. Surg. Oncol.Home page
J. Cheng, D. Hong, G. Zhu, L. L. Swanstrom, and P. D. Hansen
Laparoscopic Placement of Hepatic Artery Infusion Pumps: Technical Considerations and Early Results
Ann. Surg. Oncol., June 1, 2004; 11(6): 589 - 597.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Allen, P. J.
Right arrow Articles by Fong, Y.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Allen, P. J.
Right arrow Articles by Fong, Y.
Related Collections
Right arrow Surgery

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS