- Email: [email protected]
Direct laparoscopic cannula insertion at the left upper quadrant
November 1997, Vol. 4, No. 5
TheJournal of the American Association of Gynecologic Laparoscopists
Direct Laparoscopic Cannula Insertion at the Left Upper Quadrant Fred M. Howard, M.D., Ahmed M. EI-Minawi, M.D., and Victor E. DeLoach, M.D.
Abstract We evaluated the efficacy and safety of direct left upper quadrant (L UQ) cannula insertion for laparoscopic surgery in 23 women with prior pelvic surgery, compared with direct umbilical cannula insertion in a control group of 81 patients. Generally, the laparoscope was retained at the LUQ site throughout the operative procedure. Cannula insertions at the LUQ were successful in the first attempt in 22 patients, compared with a single successful attempt in 78 of 81 umbilical insertions. Nine women had anterior abdominal wall adhesions that extended to the umbilical area. Seven had either a prior midline (1) or Pfannenstiel (6) incision; all seven had direct LUQ cannula insertions. Two patients with umbilical adhesions had no prior surgery. Of the three complications, two were related to cannula insertions and both were in the control group. There were no bowel injuries. More experience is required to prove that L UQ cannula insertion accomplishes its intended aim of avoiding bowel or omental injuries due to adhesions in women with prior abdominopelvic surgery, but initial results were favorable. (l Am Assoc Gynecol Laparosc 4(5):595-600, 1997)
Introducing either a Veress needle or both Veress needle and laparoscopic cannula through the left upper quadrant (LUQ) is not new.1 It was described during laparoscopy in patients with prior surgery and anticipated intraabdominal adhesive disease. 2,3 It was our practice to insert the cannula directly at the umbilicus, but in 1993 we began to perform the insertion at the LUQ as well. To our knowledge, this approach has not been described in published reports.
assistant, or primary surgeon at laparoscopic surgery in which direct cannula insertion was done from December 1993 to June 1995. Previous surgery, type of incision or scar, preoperative and intraoperative diagnoses, presence of adhesions, intraoperative or postoperative complications, method of cannula placement, operative procedures, number of insertion attempts, number of cannulas, postoperative days, and operating times were recorded. Data were entered and analyzed using Epi Info 6 (CDC, Atlanta, GA). The Z 2 or Fischer's exact test was used for comparisons of frequencies of dichotomous variables. Analysis of variance and KruskalWallis tests were used for comparisons of means.
Materials and Methods We reviewed the hospital records of all 104 patients for whom the senior author (FMH) was the preceptor,
From Rochester General Hospital, University of Rochester School of Medicine and Dentistry, Rochester, New York (all authors). Address reprint requests to Fred M. Howard, M.D., Department of Obstetrics and Gynecology, Rochester General Hospital, 1425 Portland Avenue, Rochester, NY 14621; fax 716 338 3798.
595
Laparoscopic CannuJa Insertion at the Left Upper Quadrant Howard et al
Probability of 0.05 or less was interpreted as statistically significant.
LEFT HAND ELEVATING UPPER ABDOMEN ~ k
PROPER ANGL E OF
Operative Procedure Patients selected for this approach could not have hepatomegaly or splenomegaly. The stomach was emptied with a nasogastric or orogastric tube before surgery. The patient was kept flat (i.e., not in Trendelenburg position) for the insertion. The site of cannula insertion was 2 to 3 cm below the subcostal arch at the left midclavicular line, lateral to the rectus muscle (Figure 1). This site is lateral to the superior epigastric vessels of the abdominal wall and to the aorta and vena cava. The cannula was inserted at about a 45degree angle. The abdominal wall was elevated by the surgeon's left hand, or both the surgeon's and assistant's hands, about 5 to 10 cm inferior to the insertion site (Figure 2). The cannula was directed in a plane parallel to the patient's longitudinal anatomic axis (Figure 3). It was important not to aim toward the midline, as the aorta and vena cava are medial to this site. A double "pop" was felt as the cannula passed through the fascial layers and peritoneum, but it was crucial to stop entry as soon as the peritoneum was thought to be traversed. The transverse colon and omentum are located directly below this site, and excessive depth of entry may injure them. Introduction of the laparoscope allowed confirmation of entry into the abdominal cavity with visu-
FIGURE 2. The proper method of elevating the upper abdomen and angulating the cannula during direct LUQ insertion.
alization of bowel, omentum, and anterior abdominal wall. Tipping the intraabdominal end of the laparoscope toward the anterior abdominal wall helped to confirm entry by allowing easy visualization of the anterior wall in the upper visual field and bowel and omentum in the lower visual field. If entry did not occur, the cannula and cannula sleeve were advanced easily by repeating the confirmatory process. As the omentum and descending colon are directly below this site, it is critical that the cannula not be advanced farther than absolutely necessary during the blind part of the procedure. Generally the laparoscope was retained at the LUQ site throughout the operative procedure, even after umbilical adhesions were lysed. Exposure was quite adequate from this perspective (Figure 4), although it was somewhat limited if extensive surgery was
g
CORRECT INSERTION AXIS 9
.'2--P .*
SITE OF INSERTION
"kx~ ~;' Area of abdomen lifted during direct trocar
UMBILICUS
I
insertion
.
FIGURE 1. (A) left upper quadrant site of direct cannula insertion compared with (B) traditional umbilical site. (C) Area of the abdomen elevated during cannula insertion at the LUQ.
FIGURE 3. Correct entry direction parallel to the anatomic axis of the patient during direct LUQ cannula insertion.
596
November 1997, Vol. 4, No. 5
The Journal of the American Association of Gynecologic Laparoscopisls
TABLE 1. Characteristics of 104 Patients Undergoing Operative Laparoscopy with LUQ versus Umbilical Cannula Insertion
Variable Age (yrs) Body mass index No. (%)of patients with previous surgery Number of prior surgical procedures Number (%) of patients with prior laparotomies No. (%) midline scars No. (%) Pfannenstiel scars No. (%) RLQ scars No. (%) laparoscopy
FIGURE 4. The view obtained with LUQ insertion of the laparoscope, with omental adhesions extending from the uterine fundus to the umbilical area.
necessary along the left pelvic sidewall or deep in the cul-de-sac. In such cases the laparoscope was moved to the umbilicus after umbilical insertion of the cannula under direct visualization. The L U Q site was then used as an additional instrument insertion site for the surgeon on the patient's left side.
LUQ (n = 23)
Umbilicus (n = 81)
p Value
38.3 _+ 11.4 26.4 _+6.5 23 (100)
33.4 _+ 10.4 26.1 _+6.3 53 (65.4)
0.05 0.84 0.02
3.1 _+ 1.5
1.3 _+ 1.3
<0.0001
22 (95.6)
17 (20.9)
<0.0001
5 (21.7)
1 (1)
0.002
16 (69.5)
11 (13.5)
<0.0001
3 (13)
3 (3.7)
0.03
12 (52.2)
45 (55.5)
0.51
2 (8.7)
2 (2.5)
0.21
scars
Results
No. (%) RUQ scars
Twenty-three women had direct LUQ and 81 had direct umbilical insertions of the initial laparoscopic cannulas (Table 1). Disposable 12-mm cannulas were used for direct insertion in all cases. Patients with LUQ entry were older but were not of different body mass index. One hundred percent of women in whom the LUQ technique was used had histories of pelvic surgery, compared with 65% of those with umbilical entry. Patients with LUQ insertions also had undergone more procedures (3.1 vs 1.3), including laparotomy in 96% and 21%, respectively. Twelve (14.8%) patients with an umbilical entry had either a midline or Pfannenstiel scar compared with 21 (91.3%) of the LUQ series. Those with prior surgery but not laparotomy had only laparoscopic procedures (Table 2). Those with LUQ direct insertions had a significantly higher proportion of laparotomies. With the LUQ approach, direct cannula insertion was successful with one attempt in 22 cases, compared with a single successful attempt in 78 umbilical insertions. In only one case was three attempts necessary, this with umbilical insertion. In all but one case the
need for more than one insertion attempt was due to preperitoneal placement with full-depth insertion of the cannula. In the remaining case it was due to umbilical insertion of the cannula into omental adhesions that obscured vision. Operating times and postoperative hospital stays were not different in the two groups (Table 3). Periumbilical adhesions were more frequent in the LUQ insertion group. Nine of the 107 women had anterior abdominal wall adhesions that extended to the subumbilical area. Adhesions were dense in all women. Given the retrospective nature of this study, it was not possible to discern whether the adhesions involved omentum or intestine, as this was not consistently recorded in the operative notes. Seven patients with subumbilical adhesions had either a prior midline (1) or Pfannenstiel (6) incision; all had LUQ cannula insertions. Two women with subumbilical adhesions had no prior surgery (Table 4). One had bilateral hydrosalpinges and a history of pelvic inflammatory
597
Laparoscopic Cannula Insertion at the Left Upper Quadrant Howard et al
TABLE 2. Prior Surgical Procedures in Women with LUQ and Umbilical Cannula Insertions Procedures Operative laparoscopies Tubal surgery Bilateral tubal ligation Ovarian surgery Adhesiolysis Endometriosis destruction Diagnostic laparoscopy Other Laparotomies Cesarean section Abdominal hysterectomy + salpingo-oophorectomy Ovarian surgery Tubal surgery Other Vaginal procedures Total vaginal hysterectomy Total number of procedures a % of laparotomies b
LUQ
Umbilical
20 0 1 5 3 4 7 0 53 27 6
82 8 14 10 3 14 27 6 21 5 6
3 1 16 0 0 73 72.6
1 1 8 1 1 104 21.1
TABLE 4. Correlation of Periumbilical Adhesions with Type of Surgical Incision
Type of Incision Pfannenstiel a Midline a Either Pfannenstiel or midline b Neither Pfannenstiel nor midline b
No. (%) of Periumbilical Adhesions
No. of Periumbilical Adhesions
6 (22.2) 1 (16.6) 7 (21.2)
21 5 26
27 6 33
2 (2.8)
69
71
Totals
ap = 1.0. bp = 0.004.
placement of the laparoscope. The rationale for this site is as follows: it is usually an avascular location; it is a relatively thin tissue layer; it allows direction of the Veress needle and cannula into the hollow of the pelvis; it has a low risk of herniation; it affords a midline perspective of the pelvis; and it is cosmetically appealing. Placement of the cannula is often preceded by insertion of a Veress needle through which pneumoperitoneum is established. Although the Veress needle is intended to minimize the risk of major vascular injury, most reported cases of vascular injury appear to have been secondary to the instrument. 44 Direct insertion of the cannula was proposed as a method of avoiding two blind entries into the abdominal cavity, avoiding preperitoneal insuffiation with carbon dioxide, and shortening the time for peritoneal
aTotals add up to more than 104 as some patients had history of more than one procedure. bp = <0.0001.
disease and the other had Crohn disease; both had umbilical cannula insertions. Three complications occurred (Table 5). Two were related to cannula insertions, both in the umbilical insertion group. There were no bowel injuries.
entry. 7
Discussion
Two large case series of direct cannula insertion have been published. In 2000 cases of direct cannula
Gynecologists traditionally have used the umbilicus as the preferred site for introducing a cannula for
TABLE 5. Complications with 104 Laparoscopies TABLE 3. Operating Times, Postoperative Stays, and Periumbilical Adhesions
Operating time (min) Postoperative stay (days) No. (%) periumbilical adhesions
LUQ (n = 23)
Umbilicus (n = 81 )
p Value
156 _+ 64
150 + 61
0.66
1.0 + 1.6
1.3 _+ 1.6
0.53
7 (30.4)
2 (2.5)
0.0003
Laparotomy More than one insertion attempts Bleeding from cannula sitea Pelvic hematoma Trocar injury to mesenteric vessel a Total complications
LUQ (n = 23)
Umbilicus (n = 81)
p Value
1 1
1 3
0.39 0.78
0
1
1 0
0 1
3
6
0.53
aComplications secondary to initial cannula insertions.
598
November 1997, Vol. 4, No. 5
The Journal of the American Association of Gynecologic Laparoscopists
insertion at the umbilicus, no vascular injuries and 3 (0.15%) bowel injuries were reported. 8 Thirty-six percent of patients had prior abdominal surgery. In a series of 937 cases of direct cannula insertion at the umbilicus, no bowel or vascular injuries occurred. 9 Twenty-nine percent of these women had prior abdominal surgery. Survey data from the American Association of Gynecologic Laparoscopists (AAGL) also suggest about a 0.15% rate of bowel injury with closed techniques at the umbilical site.l~ Although open laparoscopy is suggested as a way of avoiding bowel injury, the same survey of AAGL members actually suggested a higher frequency of bowel injury with open laparoscopy, 1.2%, versus 0.15% with closed laparoscopy. This is believed most likely to be due to preference for open laparoscopy in patients with prior surgery and a greater chance of adhesions. However, these data clearly show that open laparoscopy does not completely avoid bowel injury in women with prior pelvic surgery, as injury may still occur at the time of entry into the peritoneal cavity. Adhesions at the umbilicus after pelvic surgery are not uncommon. Bowel or omental adhesions were found posterior to the umbilicus in 27% of 258 patients with a prior Pfannenstiel incision, 55% of 87 with a infraumbilical midline incision, and 67% of 15 with a midline incision extending above the umbilicus, n Similarly in our much smaller series, 22% of 27 patients with a Pfannenstiel scar had umbilical adhesions, as did 17% of 6 with a midline incision. In the larger series, using the umbilical entry site, there were 6 bowel injuries (2 at open laparoscopy) and 15 omental injuries, u Although the L U Q site for laparoscopy was described in 1974, the first detailed description of its use to avoid injury due to umbilical adhesions in patients with prior surgery appears to have been published in 1991.12 The authors suggested using a Veress needle to establish pneumoperitoneum, followed by insertion of a 5-mm cannula and laparoscope to evaluate the subumbilical area for possible adhesions and allow insertion of a 10- or 11-mm cannula and laparoscope at the umbilicus under direct visualization This approach was used in 10 of 41 patients with cancer. 2 In 31 other patients the Veress needle and cannula were inserted at the ninth or tenth intercostal space. All of the patients had prior midline incisions or umbilical hernias. Twenty-eight (68%) had either bowel or omental adhesions at the umbilicus. One patient experienced a small bowel injury during cannula insertion.
599
In that case an 11-mm cannula was used and the injury was ascribed to increased resistance when inserting the larger cannula. A Veress needle was used in 12 patients with left upper quadrant cannula insertion. 3 In three cases the needle was inserted at the umbilicus and in nine at the LUQ. All of the patients had periumbilical adhesions, four involving intestine and eight involving only omentum. No complications were reported. Adhesions at the LUQ are extremely unlikely in women with prior abdominopelvic surgery through a Pfannenstiel or midline incision, so this site theoretically allows avoidance of bowel injuries due to adhesions. We observed no adhesions in the LUQ in any of our 23 patients, similar to earlier reports in 412 and 12 patients. 3 No injuries occurred in 23 cases of direct LUQ cannula insertion, and considering that these women had a mean of 3.1 + 1.5 previous surgeries, this suggests the potential safety of this method. We have found this an easy entry site; however, some precautions and restraint must be stated. The locations of the descending colon and omentum directly below the site suggest the potential for serious injury if care is not taken to avoid excessive force and depth of entry. The enterotomy that occurred in one report 2 may have been the result of excessive force and depth of insertion. Injury is clearly possible to the stomach if it is not empty; thus emptying with an orogastric or nasogastric tube seems mandatory. Also the distance to the aorta is less than in the pelvis, so it seems critical that the cannula not be directed toward the midline. We have continued to have good success with this technique since collecting these data, including in laparoscopic surgery during pregnancy, but this still represents a small series of cases performed by one experienced laparoscopic surgeon. Certainly much more experience is necessary before we can be assured that the method accomplishes its intended aim of totally avoiding bowel or omental injuries due to adhesions in women with prior abdominopelvic surgery.
References
1. Palmer R: Safety in laparoscopy. J Reprod Med 13:1-5, 1974 2. Childers JM, Brzechffa PR, Surwit EA: Laparoscopy using the left upper quadrant as the primary trocar site. Gynecol Oncol 50:221-225, 1993
Laparoscopic Cannula Insertion at the Left Upper Quadrant Howard et al
3. Chang FH, Chou HH, Lee CL, et al: Extraumbilicalinsertion of the operative laparoscope in patients with extensive intraabdominal adhesions. J Am Assoc Gynecol Laparosc 2:335-337, 1995
8. Copeland C, Wing R, Hulka JF: Direct trocar insertion at laparoscopy: An evaluation. Obstet Gynecol 62:655-659, 1983 9. Byron JW, Fujiyoshi CA, Miyazawa K: Evaluation of the direct trocar insertion technique at laparoscopy. Obstet Gynecol 74:423--425, 1989
4. Peterson HB, Greenspan JR, Ory HW: Death following puncture of the aorta during laparoscopic sterilization. Obstet Gynecol 59:133-134, 1982 5. Kurzel RB, Edinger DD Jr: Injury to the great vessels: A hazard of transabdominal endoscopy. South Med J 76:656-657, 1983
10. Levy BS, Hulka JF, Peterson HB, et al: Operative laparoscopy: American Association of Gynecologic Laparoscopists' I993 membership survey. J Am Assoc Gynecol Laparosc 1:301-305, 1994
6. Baadsgaard SE, Bille S, Egebald K: Major vascular injury during gynecologic laparoscopy. Acta Obstet Gynecol Scand 68:283-285, 1989
11. Brill AI, Nezhat FR, Nezhat CH, et al: The incidence of adhesions after prior laparotomy: A laparoscopic appraisal. Obstet Gynecol 85:269-272, 1995
7. Borgatta L, Gruss L, Barad D, et al: Direct trocar insertion vs Veress needle for laparoscopic sterilization. J Reprod Med 35:891-894, 1990
12. Cohen MR, Scoccia B: Double laparoscopy: An alternative two-stage procedure to minimize bowel and blood vessel injury. J Gynecol Surg 7:203-206, 1991
600
TheJournal of the American Association of Gynecologic Laparoscopists
Direct Laparoscopic Cannula Insertion at the Left Upper Quadrant Fred M. Howard, M.D., Ahmed M. EI-Minawi, M.D., and Victor E. DeLoach, M.D.
Abstract We evaluated the efficacy and safety of direct left upper quadrant (L UQ) cannula insertion for laparoscopic surgery in 23 women with prior pelvic surgery, compared with direct umbilical cannula insertion in a control group of 81 patients. Generally, the laparoscope was retained at the LUQ site throughout the operative procedure. Cannula insertions at the LUQ were successful in the first attempt in 22 patients, compared with a single successful attempt in 78 of 81 umbilical insertions. Nine women had anterior abdominal wall adhesions that extended to the umbilical area. Seven had either a prior midline (1) or Pfannenstiel (6) incision; all seven had direct LUQ cannula insertions. Two patients with umbilical adhesions had no prior surgery. Of the three complications, two were related to cannula insertions and both were in the control group. There were no bowel injuries. More experience is required to prove that L UQ cannula insertion accomplishes its intended aim of avoiding bowel or omental injuries due to adhesions in women with prior abdominopelvic surgery, but initial results were favorable. (l Am Assoc Gynecol Laparosc 4(5):595-600, 1997)
Introducing either a Veress needle or both Veress needle and laparoscopic cannula through the left upper quadrant (LUQ) is not new.1 It was described during laparoscopy in patients with prior surgery and anticipated intraabdominal adhesive disease. 2,3 It was our practice to insert the cannula directly at the umbilicus, but in 1993 we began to perform the insertion at the LUQ as well. To our knowledge, this approach has not been described in published reports.
assistant, or primary surgeon at laparoscopic surgery in which direct cannula insertion was done from December 1993 to June 1995. Previous surgery, type of incision or scar, preoperative and intraoperative diagnoses, presence of adhesions, intraoperative or postoperative complications, method of cannula placement, operative procedures, number of insertion attempts, number of cannulas, postoperative days, and operating times were recorded. Data were entered and analyzed using Epi Info 6 (CDC, Atlanta, GA). The Z 2 or Fischer's exact test was used for comparisons of frequencies of dichotomous variables. Analysis of variance and KruskalWallis tests were used for comparisons of means.
Materials and Methods We reviewed the hospital records of all 104 patients for whom the senior author (FMH) was the preceptor,
From Rochester General Hospital, University of Rochester School of Medicine and Dentistry, Rochester, New York (all authors). Address reprint requests to Fred M. Howard, M.D., Department of Obstetrics and Gynecology, Rochester General Hospital, 1425 Portland Avenue, Rochester, NY 14621; fax 716 338 3798.
595
Laparoscopic CannuJa Insertion at the Left Upper Quadrant Howard et al
Probability of 0.05 or less was interpreted as statistically significant.
LEFT HAND ELEVATING UPPER ABDOMEN ~ k
PROPER ANGL E OF
Operative Procedure Patients selected for this approach could not have hepatomegaly or splenomegaly. The stomach was emptied with a nasogastric or orogastric tube before surgery. The patient was kept flat (i.e., not in Trendelenburg position) for the insertion. The site of cannula insertion was 2 to 3 cm below the subcostal arch at the left midclavicular line, lateral to the rectus muscle (Figure 1). This site is lateral to the superior epigastric vessels of the abdominal wall and to the aorta and vena cava. The cannula was inserted at about a 45degree angle. The abdominal wall was elevated by the surgeon's left hand, or both the surgeon's and assistant's hands, about 5 to 10 cm inferior to the insertion site (Figure 2). The cannula was directed in a plane parallel to the patient's longitudinal anatomic axis (Figure 3). It was important not to aim toward the midline, as the aorta and vena cava are medial to this site. A double "pop" was felt as the cannula passed through the fascial layers and peritoneum, but it was crucial to stop entry as soon as the peritoneum was thought to be traversed. The transverse colon and omentum are located directly below this site, and excessive depth of entry may injure them. Introduction of the laparoscope allowed confirmation of entry into the abdominal cavity with visu-
FIGURE 2. The proper method of elevating the upper abdomen and angulating the cannula during direct LUQ insertion.
alization of bowel, omentum, and anterior abdominal wall. Tipping the intraabdominal end of the laparoscope toward the anterior abdominal wall helped to confirm entry by allowing easy visualization of the anterior wall in the upper visual field and bowel and omentum in the lower visual field. If entry did not occur, the cannula and cannula sleeve were advanced easily by repeating the confirmatory process. As the omentum and descending colon are directly below this site, it is critical that the cannula not be advanced farther than absolutely necessary during the blind part of the procedure. Generally the laparoscope was retained at the LUQ site throughout the operative procedure, even after umbilical adhesions were lysed. Exposure was quite adequate from this perspective (Figure 4), although it was somewhat limited if extensive surgery was
g
CORRECT INSERTION AXIS 9
.'2--P .*
SITE OF INSERTION
"kx~ ~;' Area of abdomen lifted during direct trocar
UMBILICUS
I
insertion
.
FIGURE 1. (A) left upper quadrant site of direct cannula insertion compared with (B) traditional umbilical site. (C) Area of the abdomen elevated during cannula insertion at the LUQ.
FIGURE 3. Correct entry direction parallel to the anatomic axis of the patient during direct LUQ cannula insertion.
596
November 1997, Vol. 4, No. 5
The Journal of the American Association of Gynecologic Laparoscopisls
TABLE 1. Characteristics of 104 Patients Undergoing Operative Laparoscopy with LUQ versus Umbilical Cannula Insertion
Variable Age (yrs) Body mass index No. (%)of patients with previous surgery Number of prior surgical procedures Number (%) of patients with prior laparotomies No. (%) midline scars No. (%) Pfannenstiel scars No. (%) RLQ scars No. (%) laparoscopy
FIGURE 4. The view obtained with LUQ insertion of the laparoscope, with omental adhesions extending from the uterine fundus to the umbilical area.
necessary along the left pelvic sidewall or deep in the cul-de-sac. In such cases the laparoscope was moved to the umbilicus after umbilical insertion of the cannula under direct visualization. The L U Q site was then used as an additional instrument insertion site for the surgeon on the patient's left side.
LUQ (n = 23)
Umbilicus (n = 81)
p Value
38.3 _+ 11.4 26.4 _+6.5 23 (100)
33.4 _+ 10.4 26.1 _+6.3 53 (65.4)
0.05 0.84 0.02
3.1 _+ 1.5
1.3 _+ 1.3
<0.0001
22 (95.6)
17 (20.9)
<0.0001
5 (21.7)
1 (1)
0.002
16 (69.5)
11 (13.5)
<0.0001
3 (13)
3 (3.7)
0.03
12 (52.2)
45 (55.5)
0.51
2 (8.7)
2 (2.5)
0.21
scars
Results
No. (%) RUQ scars
Twenty-three women had direct LUQ and 81 had direct umbilical insertions of the initial laparoscopic cannulas (Table 1). Disposable 12-mm cannulas were used for direct insertion in all cases. Patients with LUQ entry were older but were not of different body mass index. One hundred percent of women in whom the LUQ technique was used had histories of pelvic surgery, compared with 65% of those with umbilical entry. Patients with LUQ insertions also had undergone more procedures (3.1 vs 1.3), including laparotomy in 96% and 21%, respectively. Twelve (14.8%) patients with an umbilical entry had either a midline or Pfannenstiel scar compared with 21 (91.3%) of the LUQ series. Those with prior surgery but not laparotomy had only laparoscopic procedures (Table 2). Those with LUQ direct insertions had a significantly higher proportion of laparotomies. With the LUQ approach, direct cannula insertion was successful with one attempt in 22 cases, compared with a single successful attempt in 78 umbilical insertions. In only one case was three attempts necessary, this with umbilical insertion. In all but one case the
need for more than one insertion attempt was due to preperitoneal placement with full-depth insertion of the cannula. In the remaining case it was due to umbilical insertion of the cannula into omental adhesions that obscured vision. Operating times and postoperative hospital stays were not different in the two groups (Table 3). Periumbilical adhesions were more frequent in the LUQ insertion group. Nine of the 107 women had anterior abdominal wall adhesions that extended to the subumbilical area. Adhesions were dense in all women. Given the retrospective nature of this study, it was not possible to discern whether the adhesions involved omentum or intestine, as this was not consistently recorded in the operative notes. Seven patients with subumbilical adhesions had either a prior midline (1) or Pfannenstiel (6) incision; all had LUQ cannula insertions. Two women with subumbilical adhesions had no prior surgery (Table 4). One had bilateral hydrosalpinges and a history of pelvic inflammatory
597
Laparoscopic Cannula Insertion at the Left Upper Quadrant Howard et al
TABLE 2. Prior Surgical Procedures in Women with LUQ and Umbilical Cannula Insertions Procedures Operative laparoscopies Tubal surgery Bilateral tubal ligation Ovarian surgery Adhesiolysis Endometriosis destruction Diagnostic laparoscopy Other Laparotomies Cesarean section Abdominal hysterectomy + salpingo-oophorectomy Ovarian surgery Tubal surgery Other Vaginal procedures Total vaginal hysterectomy Total number of procedures a % of laparotomies b
LUQ
Umbilical
20 0 1 5 3 4 7 0 53 27 6
82 8 14 10 3 14 27 6 21 5 6
3 1 16 0 0 73 72.6
1 1 8 1 1 104 21.1
TABLE 4. Correlation of Periumbilical Adhesions with Type of Surgical Incision
Type of Incision Pfannenstiel a Midline a Either Pfannenstiel or midline b Neither Pfannenstiel nor midline b
No. (%) of Periumbilical Adhesions
No. of Periumbilical Adhesions
6 (22.2) 1 (16.6) 7 (21.2)
21 5 26
27 6 33
2 (2.8)
69
71
Totals
ap = 1.0. bp = 0.004.
placement of the laparoscope. The rationale for this site is as follows: it is usually an avascular location; it is a relatively thin tissue layer; it allows direction of the Veress needle and cannula into the hollow of the pelvis; it has a low risk of herniation; it affords a midline perspective of the pelvis; and it is cosmetically appealing. Placement of the cannula is often preceded by insertion of a Veress needle through which pneumoperitoneum is established. Although the Veress needle is intended to minimize the risk of major vascular injury, most reported cases of vascular injury appear to have been secondary to the instrument. 44 Direct insertion of the cannula was proposed as a method of avoiding two blind entries into the abdominal cavity, avoiding preperitoneal insuffiation with carbon dioxide, and shortening the time for peritoneal
aTotals add up to more than 104 as some patients had history of more than one procedure. bp = <0.0001.
disease and the other had Crohn disease; both had umbilical cannula insertions. Three complications occurred (Table 5). Two were related to cannula insertions, both in the umbilical insertion group. There were no bowel injuries.
entry. 7
Discussion
Two large case series of direct cannula insertion have been published. In 2000 cases of direct cannula
Gynecologists traditionally have used the umbilicus as the preferred site for introducing a cannula for
TABLE 5. Complications with 104 Laparoscopies TABLE 3. Operating Times, Postoperative Stays, and Periumbilical Adhesions
Operating time (min) Postoperative stay (days) No. (%) periumbilical adhesions
LUQ (n = 23)
Umbilicus (n = 81 )
p Value
156 _+ 64
150 + 61
0.66
1.0 + 1.6
1.3 _+ 1.6
0.53
7 (30.4)
2 (2.5)
0.0003
Laparotomy More than one insertion attempts Bleeding from cannula sitea Pelvic hematoma Trocar injury to mesenteric vessel a Total complications
LUQ (n = 23)
Umbilicus (n = 81)
p Value
1 1
1 3
0.39 0.78
0
1
1 0
0 1
3
6
0.53
aComplications secondary to initial cannula insertions.
598
November 1997, Vol. 4, No. 5
The Journal of the American Association of Gynecologic Laparoscopists
insertion at the umbilicus, no vascular injuries and 3 (0.15%) bowel injuries were reported. 8 Thirty-six percent of patients had prior abdominal surgery. In a series of 937 cases of direct cannula insertion at the umbilicus, no bowel or vascular injuries occurred. 9 Twenty-nine percent of these women had prior abdominal surgery. Survey data from the American Association of Gynecologic Laparoscopists (AAGL) also suggest about a 0.15% rate of bowel injury with closed techniques at the umbilical site.l~ Although open laparoscopy is suggested as a way of avoiding bowel injury, the same survey of AAGL members actually suggested a higher frequency of bowel injury with open laparoscopy, 1.2%, versus 0.15% with closed laparoscopy. This is believed most likely to be due to preference for open laparoscopy in patients with prior surgery and a greater chance of adhesions. However, these data clearly show that open laparoscopy does not completely avoid bowel injury in women with prior pelvic surgery, as injury may still occur at the time of entry into the peritoneal cavity. Adhesions at the umbilicus after pelvic surgery are not uncommon. Bowel or omental adhesions were found posterior to the umbilicus in 27% of 258 patients with a prior Pfannenstiel incision, 55% of 87 with a infraumbilical midline incision, and 67% of 15 with a midline incision extending above the umbilicus, n Similarly in our much smaller series, 22% of 27 patients with a Pfannenstiel scar had umbilical adhesions, as did 17% of 6 with a midline incision. In the larger series, using the umbilical entry site, there were 6 bowel injuries (2 at open laparoscopy) and 15 omental injuries, u Although the L U Q site for laparoscopy was described in 1974, the first detailed description of its use to avoid injury due to umbilical adhesions in patients with prior surgery appears to have been published in 1991.12 The authors suggested using a Veress needle to establish pneumoperitoneum, followed by insertion of a 5-mm cannula and laparoscope to evaluate the subumbilical area for possible adhesions and allow insertion of a 10- or 11-mm cannula and laparoscope at the umbilicus under direct visualization This approach was used in 10 of 41 patients with cancer. 2 In 31 other patients the Veress needle and cannula were inserted at the ninth or tenth intercostal space. All of the patients had prior midline incisions or umbilical hernias. Twenty-eight (68%) had either bowel or omental adhesions at the umbilicus. One patient experienced a small bowel injury during cannula insertion.
599
In that case an 11-mm cannula was used and the injury was ascribed to increased resistance when inserting the larger cannula. A Veress needle was used in 12 patients with left upper quadrant cannula insertion. 3 In three cases the needle was inserted at the umbilicus and in nine at the LUQ. All of the patients had periumbilical adhesions, four involving intestine and eight involving only omentum. No complications were reported. Adhesions at the LUQ are extremely unlikely in women with prior abdominopelvic surgery through a Pfannenstiel or midline incision, so this site theoretically allows avoidance of bowel injuries due to adhesions. We observed no adhesions in the LUQ in any of our 23 patients, similar to earlier reports in 412 and 12 patients. 3 No injuries occurred in 23 cases of direct LUQ cannula insertion, and considering that these women had a mean of 3.1 + 1.5 previous surgeries, this suggests the potential safety of this method. We have found this an easy entry site; however, some precautions and restraint must be stated. The locations of the descending colon and omentum directly below the site suggest the potential for serious injury if care is not taken to avoid excessive force and depth of entry. The enterotomy that occurred in one report 2 may have been the result of excessive force and depth of insertion. Injury is clearly possible to the stomach if it is not empty; thus emptying with an orogastric or nasogastric tube seems mandatory. Also the distance to the aorta is less than in the pelvis, so it seems critical that the cannula not be directed toward the midline. We have continued to have good success with this technique since collecting these data, including in laparoscopic surgery during pregnancy, but this still represents a small series of cases performed by one experienced laparoscopic surgeon. Certainly much more experience is necessary before we can be assured that the method accomplishes its intended aim of totally avoiding bowel or omental injuries due to adhesions in women with prior abdominopelvic surgery.
References
1. Palmer R: Safety in laparoscopy. J Reprod Med 13:1-5, 1974 2. Childers JM, Brzechffa PR, Surwit EA: Laparoscopy using the left upper quadrant as the primary trocar site. Gynecol Oncol 50:221-225, 1993
Laparoscopic Cannula Insertion at the Left Upper Quadrant Howard et al
3. Chang FH, Chou HH, Lee CL, et al: Extraumbilicalinsertion of the operative laparoscope in patients with extensive intraabdominal adhesions. J Am Assoc Gynecol Laparosc 2:335-337, 1995
8. Copeland C, Wing R, Hulka JF: Direct trocar insertion at laparoscopy: An evaluation. Obstet Gynecol 62:655-659, 1983 9. Byron JW, Fujiyoshi CA, Miyazawa K: Evaluation of the direct trocar insertion technique at laparoscopy. Obstet Gynecol 74:423--425, 1989
4. Peterson HB, Greenspan JR, Ory HW: Death following puncture of the aorta during laparoscopic sterilization. Obstet Gynecol 59:133-134, 1982 5. Kurzel RB, Edinger DD Jr: Injury to the great vessels: A hazard of transabdominal endoscopy. South Med J 76:656-657, 1983
10. Levy BS, Hulka JF, Peterson HB, et al: Operative laparoscopy: American Association of Gynecologic Laparoscopists' I993 membership survey. J Am Assoc Gynecol Laparosc 1:301-305, 1994
6. Baadsgaard SE, Bille S, Egebald K: Major vascular injury during gynecologic laparoscopy. Acta Obstet Gynecol Scand 68:283-285, 1989
11. Brill AI, Nezhat FR, Nezhat CH, et al: The incidence of adhesions after prior laparotomy: A laparoscopic appraisal. Obstet Gynecol 85:269-272, 1995
7. Borgatta L, Gruss L, Barad D, et al: Direct trocar insertion vs Veress needle for laparoscopic sterilization. J Reprod Med 35:891-894, 1990
12. Cohen MR, Scoccia B: Double laparoscopy: An alternative two-stage procedure to minimize bowel and blood vessel injury. J Gynecol Surg 7:203-206, 1991
600