approached via intraoperative proctoscopy ± suture place- ment. If the surgeon anticipates intraoperative localization may be difficult, lesions other than rectal or ...
Surg Endosc (1997) 11: 1013–1016
Surgical Endoscopy © Springer-Verlag New York Inc. 1997
Perioperative tumor localization for laparoscopic colorectal surgery S. H. Kim, J. W. Milsom, J. M. Church, K. A. Ludwig, A. Garcia-Ruiz, J. Okuda, V. W. Fazio Department of Colorectal Surgery, A 111, The Cleveland Clinic Foundation, 9500 Euclid Avenue, Cleveland, Ohio 44195, USA Received: 18 July 1996/Accepted: 10 March 1997
Abstract Background: Because of the inability to palpate colonic tumors during laparoscopy, their location must be precisely identified before resection is undertaken. Method: A retrospective study was performed of 58 patients in order to be able to describe our methods of tumor localization for laparoscopic colorectal operations and to review their effectiveness. Results: In all patients, the entire colon was examined preoperatively by colonoscopy. In one patient, preoperative colonoscopic localization was inaccurate. In 31 patients, tumors were easily detectable at surgery. In five patients with the tumor in the right colon, even though the lesion was not detectable at surgery, right colectomy was performed without marking because preoperative colonoscopy reliably identified the lesion adjacent to the ileocecal valve. Twentytwo patients required some type of procedure to localize the tumor. The procedures and their problems were as follows: preoperative tattoo (five)—tattoo not visualized (one); intraoperative colonoscopy alone (six), combined with intraoperative tattoo (four) or clip (three)—poor operative exposure due to bowel distension (nine), hard to see the clip (three), dislodged clip (two), inadequate resection margin (one); intraoperative proctoscopy alone (two), combined with laparoscopic stitch (two)—no problems. In no patient was tumor present at a resection line and in no patient was the wrong segment resected. Conclusions: Reliable preoperative identification of the tumor adjacent to the ileocecal valve can permit right colectomy without marking. Lesions in the upper rectum can be approached via intraoperative proctoscopy ± suture placement. If the surgeon anticipates intraoperative localization may be difficult, lesions other than rectal or cecal ones should probably be marked by preoperative tattooing. Further studies regarding the technique of tattooing are warranted.
Key words: Tumor localization — Laparoscopic surgery — Colon tumors — Rectal tumors
The need for tumor localization in laparoscopic compared to conventional colorectal tumor surgery may be heightened since ‘‘palpation’’ with the hand is not possible. Preoperative colonoscopy alone may be unreliable in delineating the tumor localization with enough accuracy for surgery. This has led to reports of laparoscopic resection of the wrong segment of the colon [6, 12]. Certain techniques must therefore be developed to ensure that the tumor-bearing segment of bowel is removed. The purpose of this retrospective study is to describe the methods we have used to precisely localize tumors where we anticipated intraoperative localization might be difficult at laparoscopic colorectal operations, and to review their effectiveness.
Methods Between December 1992 and July 1996, 58 patients underwent laparoscopic surgery for colorectal tumors. Patients who had familial adenomatous polyposis or underwent stoma creation alone for palliative purposes were excluded from the study. Patients were operated on for both premalignant and malignant diseases of the colon and rectum (14 adenomas, 44 adenocarcinomas). Resection of a segment of bowel was performed in 56 patients (25 right colectomy, two left colectomy, 20 proctosigmoidectomy, seven abdominoperineal resection, and two subtotal colectomy), and a colotomy and removal of polyp in two patients, respectively. All cancer patients who underwent a curative resection were involved in an ongoing prospective randomized study in our department comparing laparoscopic to conventional colorectal cancer surgery.
Results The abstract of this manuscript was selected for poster presentation for the Scientific Session of the SAGES Annual Meeting 19–22 March 1997, San Diego, CA Correspondence to: J. W. Milsom
In all 58 patients, the entire colon was examined preoperatively by colonoscopy. In one patient ([16), preoperative colonoscopic localization was inaccurate. Preoperative colonoscopy reported a 5 × 5 cm adenoma at the splenic
1014 Table 1. Summary of perioperative procedures to precisely localize the lesiona Patient
Location
Pathology
Perioperative procedure
Operation
Comments
Adenoma Cancer Cancer Adenoma Cancer Adenoma Cancer Cancer Cancer Cancer Cancer Cancer Cancer Cancer Cancer Adenoma Adenoma
Intraoperative colonoscopy/clip Intraoperative colonoscopy/clip Intraoperative colonoscopy
C&P PS RC
Multiple adenomas in ASC
[4 [5 [6 [7 [8 [9 [10 [11 [12 [13 [14 [15 [16
ASC SIG CEC ASC SIG SIG SIG ASC ASC REC REC SIG SIG SIG REC TRN TRN
Intraoperative colonoscopy/clip Intraoperative tattoo Intraoperative tattoo Intraoperative colonoscopy Preoperative tattoo Intraoperative proctoscopy Intraoperative proctoscopy Intraoperative colonoscopy Intraoperative colonoscopy Intraoperative tattoo Intraoperative tattoo Intraoperative colonoscopy Intraoperative colonoscopy
PS PS PS RC RC PS PS PS PS PS PS C&P RC
[17
SPF
Cancer
Preoperative tattoo
STC
[18 [19 [20 [21 [22
REC ASC SIG REC ASC
Cancer Cancer Cancer Cancer Cancer
Intraoperative proctoscopy/stitch Preoperative tattoo Preoperative tattoo Intraoperative proctoscopy/stitch Preoperative tattoo
PS RC PS PS RC
[1 [2 [3
3-cm sessile polyp Erroneous localization by preoperative colonoscopy Conversion due to nonvisualization of tattoos
a
CEC: cecum, ASC: ascending colon, TRN: transverse colon, SPF: splenic flexure, SIG: sigmoid colon, REC: rectum, RC: right colectomy, PS: proctosigmoidectomy, STC: subtotal colectomy, C & P: colotomy and polypectomy
flexure but intraoperative colonoscopy revealed the tumor to be in the proximal transverse colon. Laparoscopic right colectomy was performed but inspection of the specimen demonstrated a close distal resection margin (1.5 cm). Because the possibility of cancer could not be excluded, an additional 3 cm of the distal segment was resected through a widened port site. Pathologic examination revealed a villous adenoma. In 31 patients, tumors were easily detectable at surgery. In five patients with the tumor in the cecum or ascending colon, even though the lesion was not detectable at surgery, right colectomy was performed without any marking because preoperative colonoscopy reliably identified the lesion adjacent to the ileocecal valve. Twenty-two patients required some type of procedure (other than preoperative colonoscopy alone) to precisely localize the tumor before or during laparoscopy. In five patients with colon cancer, India ink was injected preoperatively through a colonoscope using a sclerotherapy needle (Flextip Sclerotherapy Needle, Bard Interventional Products, C.R. Bard, Inc., Tewksbury, MA, U.S.A.). The injection was performed more than 24 h preoperatively to avoid distension of the bowel and 2 cm distal to the lesion to avoid direct injection into the tumor. One patient ([17) who had a cancer at the splenic flexure was converted to an open procedure due to a problem associated with tumor localization. Even though India ink was injected in four quadrants 3 days before surgery, the tattoos were not seen at laparoscopic examination. Because the small bowel was slightly distended and the mesentery was too fatty, intraoperative colonoscopy was not performed and the patient was converted to an open procedure and underwent subtotal colectomy with ileorectal anastomosis. At surgery, it was revealed that two tattoos were in the retroperitoneal portion of
the colonic wall, one was within the thick mesentery side of the colonic wall, and one was covered by the omentum. In six patients, the tumor was localized by intraoperative colonoscopy alone. In one case (patient [16), an inadequate resection margin was obtained and an additional resection was needed. In four cases with rectosigmoid cancers, methylene blue was injected through a colonoscope during surgery. In two cases with sigmoid cancer and one case with asending colonic adenoma, the tumor was marked by clips placed laparoscopically during colonoscopy. Clips were hard to see in all three cases and dislodged from the serosa in two cases. Bowel distension by air insufflated during endoscopic examination interfered with operative exposure in nine of the total 13 patients with intraoperative colonoscopy. In four patients with rectal cancer, intraoperative proctoscopy was performed. In two of them, the distal margin of the tumor was marked by a laparoscopically placed stitch after tumor localization using intraoperative proctoscopy. The procedures and their problems are summarized in Table 1 and Table 2, respectively. In no patient was tumor present at a resection line and in no patient was the wrong segment of bowel resected.
Discussion Laparoscopic resection of the wrong segment of the colon, requiring conversion to a standard laparotomy and an additional resection, was first described by Larach et al. [6]. In a recent survey of the members of the American Society of Colon and Rectal Surgeons [12], 18 of 278 responders (6.5%) reported the removal of the wrong segment of the
1015 Table 2. Problems of various procedures for tumor localization Procedure
Problem
Preoperative tattoo (5) Intraoperative colonoscopy (13) Colonoscopy alone (6) Combined with intraoperative tattoo (4) Combined with laparoscopic clipping (3) Intraoperative proctoscopy (4) Proctoscopy alone (2) Combined with laparoscopic stitching (2)
Tattoo not visualized (1) Poor operative exposure due to bowel distention (9) Inadequate resection margin (1) Hard to see the clip (3), dislodged clip (2) No problems
colon during laparoscopic colorectal surgery. Monson et al. [9] reported on a patient who was converted to an open subtotal colectomy with ileorectal anastomosis because a laparoscopically resected right hemicolectomy specimen revealed previously unsuspected multiple adenomatous polyps. Vara-Thorbeck et al. [10] reported on a patient who had a ‘‘relapse’’ 9 months after laparoscopic sigmoidectomy. The second operation by standard operation demonstrated that the relapse was localized 10 cm from the laparoscopic anastomosis, which was healthy. The author presumed this was a case of a synchronous primary tumor which was present in the first operation and suggested simultaneous colonoscopy in determining the localization of the lesion and extension of the resection is important during laparoscopic surgery. McDermott et al. [8] reported a case of nearly obstructing sigmoid colon cancer resected using the laparoscopic technique, in which postoperative bowel obstruction occurred 2 weeks after the initial operation due to an unrecognized synchronous cecal cancer. In our series of 58 patients, we have not experienced any case of either wrong segment removal or missed synchronous cancers. However, in one case (1.8%), a minor problem with tumor localization occurred due to incorrect preoperative localization. The referring physician, an experienced endoscopist who performed preoperative colonoscopy, diagnosed the patient with a 5 × 5 cm adenoma at the splenic flexure, but intraoperative colonoscopy showed the tumor to be located in the proximal transverse colon. This shows that the judgment of even an experienced colonoscopist may sometimes be in error. A series by Vignati et al. [11] reported that endoscopic localization was correct in 86% of 320 colonoscopic examinations. Our experience serves as a warning not to resect bowel based only on preoperative colonoscopy. (This applies to conventional as well as laparoscopic surgery.) If there is any doubt about the precise location of the tumor at surgery, intraoperative colonoscopy is mandatory before resection. Precise tumor localization is not a prerequisite just for laparoscopic surgery. Even in conventional surgery, intraoperative localization of small nonpalpable tumors or polypectomy sites has been reported to be a difficult problem. Frager et al. [3] reported six patients in whom errors of diagnosis and therapy occurred in conventional surgery because of reliance on preoperative colonoscopic tumor localization. Three patients required a second laparotomy for surgical resection of a tumor that was missed at the first exploration. The authors insisted on a preoperative barium enema for precise localization of tumors, and preoperative endoscopic tattoo or intraoperative colonoscopy for cases in which a polyp had already been removed and a segmental
resection was planned because of invasive malignancy found in the resected polyp. Espiner et al. [2] also discussed the hazards of relying only on barium enema in reporting their experience of 12 patients with radiologically proven lesions of the colon. Colonoscopy to the cecum was performed on unopened bowel during laparotomy in all cases. Additional polyps were found in five patients, and in four of these patients the polyp was not really palpable at operation. Several authors [1, 4, 5] advocate the use of colonoscopic tattoo injection for precise tumor localization. Botoman et al. [1] tangentially injected 0.5–1.0 ml of 1:1 diluted India ink. The tattoos were easily seen at laparotomy in 11 of 14 patients. Most of the cases in which no ink was seen occurred early in their series when smaller volumes of ink and more superficial injections were used. The author advocated staining with India ink as permanent and clearly visible even after preoperative radiation. In an experimental study in dogs, Hammond et al. [5] examined the staining characteristics of several tattooing agents. Water-soluble dyes such as methylene blue, toluidine blue, and lymphazurine stained the serosal surface of the bowel quite well; however, by 24 h the dyes had completely diffused away. They observed that India ink and indocyanine green remained visible on the serosal surface of the bowel for up to 7 days. They also showed, in a clinical study of 12 patients [4], that endoscopically injected dye (1% indocyanine green) was easily visualized on the serosal surface of the colon at surgery in all patients. In our series, preoperative colonoscopic tattooing was used in five patients with colon cancer. India ink was injected more than 24 h preoperatively to avoid the problem of distended bowel at surgery. In one case, the patient ([17) was converted to a standard laparotomy partly because the tattoos were not visualized at laparoscopic examination. The operative finding showed the dyes were injected into the retroperitoneal portion of the colonic wall and into the thick mesenteric side of the colonic wall. Additionally, the serosal surface of the bowel successfully injected by the dye was covered by the omentum. Further studies with regard to the technique of injection and the amount of dye are warranted, especially for obese patients. Preoperative colonoscopic mucosal clipping using metallic clips and intraoperative fluoroscopic visualization may be an alternative for these patients [7]. Intraoperative colonoscopy was performed in 13 patients. In nine cases, operative exposure was compromised by bowel distension with air insufflated during colonoscopy. In three patients, laparoscopic clips were applied to the serosal surface of the bowel under the guidance of in-
1016
traoperative colonoscopy. Even though all lesions were successfully resected with this method, clip marking was less attractive than tattooing because clips tended to dislodge from the serosa and were harder to see. In eight patients, the tumor was localized by intraoperative colono- or proctoscopy alone, without dyes or clips. In seven cases, the resection was successful with good margins. In one case (patient [16), an additional resection was needed due to what we believed was an inadequate distal margin (1.5 cm). Recently, intraoperative proctoscopy with a laparoscopic stitch applied to the serosa just distal to the lesion of upper rectal cancer was used. This technique is attractive to localize the tumor and to get a safe distal margin for a rectal lesion in that it may be faster and more precise than endoscopic tattooing. Conclusion Reliable preoperative identification by an endoscopist of a tumor adjacent to the ileocecal valve can permit a laparoscopic right colectomy without marking. Lesions that are in the upper rectum can be approached via intraoperative proctoscopy with or without suture placement distal to the lesion. If the surgeon anticipates intraoperative localization may be difficult, lesions outside of these areas of the colon and rectum should probably be marked by preoperative tattooing. Further studies with regard to the technique of tattooing are warranted.
References 1. Botoman VA, Pietro M, Thirlby RC (1994) Localization of colonic lesions with endoscopic tattoo. Dis Colon Rectum 37: 775–776 2. Espiner HJ, Salmon PR, Teague RH, Read AE (1973) Operative colonoscopy. Br Med J 1: 453–454 3. Frager DH, Frager JD, Wolf EL, Beneventano TC (1987) Problems in the colonoscopic localization of tumors: continued value of the barium enema. Gastrointest Radiol 12: 343–346 4. Hammond DC, Lane FR, Mackeigan JM, Passinault WJ (1993) Endoscopic tattooing of the colon: clinical experience. Am Surg 59: 205–210 5. Hammond DC, Lane FR, Welk RA, Madura MJ, Borreson DK, Passinault WJ (1989) Endoscopic tattooing of the colon: an experimental study. Am Surg 55: 457–461 6. Larach SW, Salomon MC, Williamson PR, Goldstein E (1993) Laparoscopic assisted colectomy: experience during the learning curve. Coloproctology 1: 38–41 7. Lehman GA, Maveety PR, O’Connor KW (1985) Mucosal clippingutility and safety testing in the colon. Gastrointest Endosc 31: 273–276 8. McDermott JP, Devereaux DA, Caushaj PF (1994) Pitfall of laparoscopic colectomy: an unrecognized synchronous cancer. Dis Colon Rectum 37: 602–603 9. Monson JRT, Darzi A, Carey PD, Guillou PJ (1990) Prospective evaluation of laparoscopic-assisted colectomy in an unresected group of patients. Lancet 340: 831–833 10. Vara-Thorbeck C, Garcia-Caballero M, Salvi M, Gutstein D, Toscano R, Gomez A, Vara-Thorbeck R (1994) Indications and advantages of laparoscopy-assisted colon resection for carcinoma in elderly patients. Surg Laparosc Endosc 4: 110–118 11. Vignati P, Welch JP, Cohen JL (1994) Endoscopic localization of colon cancers. Surg Endosc 8: 1085–1087 12. Wexner SD, Cohen SM, Ulrich A, Reissman P (1995) Laparoscopic colorectal surgery—are we being honest with our patients? Dis Colon Rectum 38: 723–727
