Reconstructive

RECONSTRUCTIVE

Motility Differences in Free Colon and Free Jejunum Flaps for Reconstruction of the Cervical Esophagus

Hung-Chi Chen, M.D. Background: Free colon and jejunal flaps have been described as reliable and Antonio Rampazzo, M.D. safe conduits for pharyngoesophageal reconstruction. Compared with free co- Bahar Bassiri Gharb, M.D. lon flaps, free jejunum flaps have a smaller diameter and intrinsic peristaltic Marcus T. C. Wong, M.D. movement, both of which are considered possible causes of dysphagia. In this Samir Mardini, M.D. investigation, the authors evaluated motility differences in free jejunum and Hue-Yong Chen, M.D. colon flaps using radionuclide esophageal scintigraphy. Christopher J. Salgado, M.D. Methods: Patients who received free jejunum flaps (n ϭ 12) or free colon (n ϭ 1) Kaohsiung, Taiwan; Rochester, Minn.; or ileocolon flaps (n ϭ 13) for reconstruction after pharyngoesophagectomy for and Cleveland, Ohio cancer were included. Radionuclide esophageal scintigraphy was performed using technetium-99m–labeled sulfur colloid. Transit rate was evaluated at 1 second (pharyngeal or initial clearance) and 10 seconds (esophageal or clearance throughout). Clinical progression of swallowing was recorded postoperatively. Statistical analysis was performed using the t test. Results: Mean pharyngeal clearance was 61 Ϯ 20 percent for free colon and ileocolon flaps and 70 Ϯ 16 percent for free jejunum flaps. Mean esophageal clearance was 50 Ϯ 27 percent for free colon and ileocolon flaps and 69 Ϯ 17 percent for free jejunum flaps. Esophageal transit rate was significantly shorter in patients who underwent reconstruction with free jejunum flaps (p ϭ 0.04). At 1 year, 10 of 12 free jejunum patients and eight of 14 patients were tolerating solid foods. Conclusions: Although neither flap showed normal swallowing characteristics, free jejunum flaps displayed greater esophageal clearance and should represent the first choice in hypopharyngeal reconstruction. Free colon and ileocolon flaps should be reserved for very proximal oropharyngeal defects and when simultaneous voice reconstruction is desired. (Plast. Reconstr. Surg. 122: 1410, 2008.)

esection of extensive tumors of the hypo- the cervical esophagus following resection for be- pharynx and cervical esophagus represents nign and malignant causes. The choice of flap Ra reconstructive challenge for the plastic sur- seems to be dependent on the surgeon’s training, geon. The ideal reconstructive technique should experience, and the assessment of the defect after be a single-stage procedure with low morbidity extirpation. When comparing, the free jejunum and mortality, and provide rapid restoration of flap to the free colon flap, the small diameter of swallowing function. Both free colon1 and free the jejunum commonly requires end-to-side anas- jejunum2 flaps provide reliable reconstruction of tomosis or some fashioning either to the pharynx or esophagus.3 In addition, the intrinsic cyclic mo- From the Department of Plastic Surgery and the Department of Radiology, E-Da Hospital, I-Shou University; the Division of Plastic Surgery, Mayo Clinic; and the Department of Plastic Surgery, University Hospitals Cleveland, Case West- ern Reserve University. Disclosure: None of the authors has any commer- Received for publication January 29, 2008; accepted March cial associations that might pose or create a conflict 19, 2008. of interest with the information presented in this Presented at the 2008 Meeting of the American Society for article. This includes consultancies, stock ownership Reconstructive Microsurgery, in Los Angeles, California, or other equity interests, patent licensing arrange- January 12 through 15, 2008. ments, and payments for conducting or publicizing Copyright ©2008 by the American Society of Plastic Surgeons the study described in the article. DOI: 10.1097/PRS.0b013e31818820f4

1410 www.PRSJournal.com Volume 122, Number 5 • Free Colon versus Free Jejunum Flap tility of the free jejunum flap may cause spasm on an interrupted suture technique and microscope. swallowing or intussuseption.4,5 Anastomosis of All flaps were inset in an isoperistaltic direction the remnant small bowel in the abdomen, how- and sutured end to end, in two layers, to the phar- ever, reportedly has fewer complications because ynx or distal esophagus. of the better vascularity of the small bowel.6 Free colon and ileocolon flaps, in contrast, conform and shape better, enabling an end-to-end anasto- Radionuclide Esophageal Scintigraphy: Transit mosis with both the pharynx and the esophagus. Scan Study Anastomotic leakage and postoperative ileus are, The study was performed on average after 10 however, more common for large bowel surgery and 11 weeks after the cervical esophageal recon- and may therefore increase donor-site morbidity.5 struction with free jejunum and free colon/ileo- Although manometry has been considered the colon flaps, respectively. All of the patients were standard examination for the evaluation of oe- informed to stop the medications that could have sophageal dysmotility, it gives little information on influenced esophageal motility and were asked to morphology and has low patient acceptance.7 Ra- fast for at least 6 hours before the study. dionuclide oesophageal scintigraphy stands out as Patients were placed in the sitting position with a frequently used alternative to manometry be- their chest and neck against a large-field-view cause it is a safe, noninvasive, and sensitive method gamma camera (Siemens E-CAM double-head cam- that allows quantitative assessment of esophageal era; Siemens Medical Solutions, Malvern, Pa.) to motility.8 The aims of this study were to objectively visualize the mouth, hypopharynx, entire esoph- assess and compare the clinical outcome and agus, and proximal portion of the gastric fundus esophageal motor function of patients who un- in the same field. A radioactive marker was placed derwent cervical esophageal reconstruction with on the cricoid cartilage to show the junction be- free jejunum or free colon or ileocolon flaps using tween the pharynx and the cervical esophagus. radionuclide oesophageal scintigraphy. Patients swallowed a 10-cc liquid bolus of radionuclide labeled with 0.3 to 0.6 mCi of technetium- PATIENTS AND METHODS 99m sulfur colloid (CIS-US, Inc., Bedford, Mass.) A prospective study was carried out evaluating delivered into the mouth using a syringe with a the hypopharyngeal conduit motility in patients blunt plastic tip while sitting in front of the col- undergoing reconstruction of the hypopharynx limator head of the gamma camera (140 KeV, with free intestinal flaps following tumor ablation collimator HiResPar, zoom 1). The transit scan between 2001 and 2004. Patients requiring hypo- evaluation was performed after swallowing the ra- pharyngectomy underwent reconstruction with dioagent and subsequent dry swallowing every 15 free jejunum flaps. Patients with proximal hypo- seconds. The swallowing sequences were recorded pharyngeal defects and thus requiring a better cal- by the microprocessor at 0.5-second intervals and iber match received a free colon flap. Hypophar- then stored on a computer disk. Recordings were yngolaryngectomy defects were treated with free analyzed taking into consideration the cricoid re- ileocolon flaps for simultaneous voice reconstruc- gion and gastroesophageal junction. Initial bolus tion. All of the patients received postoperative transit rate (after 1 second) was recorded as pha- radiotherapy. Transit scan studies were performed ryngeal clearance or pharyngeal transit rate and before postoperative radiotherapy on 26 patients the global esophageal transit rate was recorded who underwent a free jejunum flap (n ϭ 12) or a after 10 seconds to study the entire esophagus free colon (n ϭ 1) or ileocolon flap (n ϭ 13). (esophageal clearance or “throughout” transit There were 24 men and two women. The mean rate). Transit rate was determined based on the amount of residual activity in the segment using age was 55 years. Patient consent was obtained for ϭ Ϫ all studies, and this study was approved by our the following formula: C(t) Emax E(t)/E hospital committee review board. max, where C(t) represents the percentage of pharynx and esophagus emptying at 1 second and 10 seconds, respectively; E max represents the Operative Procedure maximal count rate in the pharynx and esopha- Free colon, ileocolon, and jejunum flaps were gus, respectively; and E(t) represents the segment harvested in a standard fashion. Recipient vessels count rate at 1 second and 10 seconds, respec- were the transverse cervical artery and external tively. Ninety percent was considered the physio- jugular vein (if available) or end-to-side to the logic transit rate in both segments. Progression of internal jugular vein or one of its branches using deglutition training was recorded and compared,

1411 Plastic and Reconstructive Surgery • November 2008 as was the nature of food the patient could tolerate after 4 weeks and 12 months after surgery.

Statistical Analysis Transit rate data with standard deviations were calculated. For the statistical analysis, patients re- ceiving free colon and free ileocolon flaps were treated as one group. The t test (two-sided, 95 percent confidence interval) for samples with different variances was used to compare pharyngeal and esophageal transit rates between the two groups and the differences considered significant for values of p Յ 0.05.

RESULTS There were no free flap losses in either group or reexplorations for thrombosis. In this series, there were no infections, fistulas, or strictures. At a mean follow-up of 18 months, one patient has died as a result of recurrent disease outside of the global period from surgery; however, his evaluation is included. Figures 1 and 2 show the results of a normal and abnormal nuclear medicine transit rate study, respectively. Pharyngeal (initial) and esophageal (throughout) transit rate results for free jejunum and free colon and ileocolon flaps are reported in Table 1. Mean pharyngeal clearance was 61 Ϯ 20 percent for free colon and ileocolon flaps, and 69 Ϯ 16 percent for free jejunum flaps. Mean esophageal clearance was 50 Ϯ 27 percent for free colon and ileocolon flaps, and 69 Ϯ 17 percent for free jejunum flaps (Figs. 3 and 4). Esophageal transit rate was significantly shorter in patients who underwent reconstruction with free jejunum flaps (p ϭ 0.04) compared with free colon or ileocolon flaps. The free jejunum reconstruction patients were all able to swallow liquids and eat semisolid foods after 4 weeks. The free colon and ileocolon patients were all (n ϭ 14) tolerating only liquids at 1 month. These patients experienced a very slow progression of their deglutition; however, after training by occluding the nose to push semisolid foods down their neoesophagus, they were able to tolerate this diet, and half progressed to solid foods. Patients in both groups complained of dysphagia during the initial 2 weeks following reconstruction; however, this was resolved in all patients in the jejunum group by 4 weeks. After 12 Fig. 1. Results from a patient with normal swallowing function. months, 10 of the patients in the jejunum group (Above) Liquid radioagent in the mouth. (Center) After 1 second, had progressed to a solid food diet, whereas two more than 90 percent of the radioagent is cleared by the pharynx could tolerate only semisolids. Six of 14 patients in (arrow shows the upper esophageal sphincter). (Below) After 10 the free colon or ileocolon group continued to seconds,morethan90percentoftheradioagentisclearedbythe complain of dysphagia on their last follow-up and esophagus (arrow shows the lower esophageal sphincter).

1412 Volume 122, Number 5 • Free Colon versus Free Jejunum Flap

Table 1. Pharyngeal and Esophageal Transit Rates Pharyngeal Esophageal Patient Transit Rate (%) Transit Rate (%) Free jejunum flap 13940 24642 35652 46056 56573 67373 77576 87679 97980 10 80 82 11 86 86 12 93 90 Free colon and ileocolon flap 13010 23111 33716 44717 54845 65747 76550 86552 97052 10 73 74 11 76 75 12 82 82 13 82 82 14 91 82

continued on a semisolid diet only, whereas the remainder tolerated solid foods.

DISCUSSION Reestablishment of the alimentary conduit, in- cluding restoration of swallowing and speech, must be achieved in the management of hypo- pharyngectomy and hypopharyngolaryngectomy defects. Among the several procedures that have been described, interposition of jejunal and colonic flaps represents an attractive option in esophageal reconstruction. Since its first description by Seiden- berg and Hurwitt in 1958,9 free jejunal flaps have become a standard option for circumferential reconstruction of the hypopharynx, with a reported success rate between 90 percent and 100 percent.10,11 Jejunal grafts are an ideal hypopharyngeal substi- tute because of their tubed anatomy, mucosal lin- ing, and thin nature. Free ileocolic flap reconstruction, reported for the first time by Kawahara et al.,12 represents a valuable alternative to free Fig. 2. Results from a patient with abnormal swallowing func- jejunum for defects starting high up in the oro- tion. (Above) Liquid radioagent in the mouth. (Center) After 1second, less than 90 percent of the radioagent is cleared by the pharynx. It is possible to see part of the radioagent proximal to agus and multiple peaks of radioactivity are still present in all of the upper esophageal sphincter (arrow). (Below) After 10 sec- the esophageal segments (arrow shows the lower esophageal onds, less than 90 percent of the bolus is cleared by the esoph- sphincter).

1413 Plastic and Reconstructive Surgery • November 2008

Fig.3. Pharyngealclearanceafter1secondfromtheinitialswallow.Thedifference is not statistically significant (p Ͼ 0.05).

Fig. 4. Esophageal clearance after 10 seconds from the initial swallow. The differ- ence between free jejunum and colon is statistically significant (p Ͻ 0.05). pharynx, as it offers a better caliber match and comparison with normal physiology, jejunal graft allows for simultaneous voice reconstruction.13,14 contraction is more prolonged than normal esoph- When compared with adynamic cutaneous and ageal contraction using manometric evaluation.18 musculocutaneous flaps, both jejunum and colon Other radiologic studies have demonstrated a de- display distinct intrinsic motor activity that is pre- lay in orogastric transit in patients above the lower served after microvascular transfer. Postprandial jejunal anastomosis although, clinically, patients jejunal motility consists of rhythmic segmenta- did not display any difficulty in swallowing liquids tions mixing luminal contents, which facilitates or solids.19 Spontaneous contractile activity has absorption. The colon does not demonstrate the been noted in the free jejunal grafts, not inter- same cyclic motor activity but does contract, cor- rupted by propagated water swallows, thus possibly responding to antegrade and retrograde mass contributing to postoperative dysphagia in some movements.15 Because this visceral motor activity individuals.20 is different from normal, physiologic pharyngoe- The motility of the transverse colon used for sophageal motility, the vigorous jejunal peristaltic esophageal replacement was investigated by Dantas activity or the slow, intermittent colonic contrac- and Mamede16 using manometry in 10 patients. tions have both been refuted as possible causes of Most of the time, the colon did not show contrac- occasional dysphagia noted in both types of tions however, there was distention of the interposed reconstruction.16,17 segment with water swallows. If the stimulus was suf- The postoperative function of free jejunal flaps ficient, contractions were evident. Transit time was for pharyngoesophageal reconstruction has been delayed in all patients, causing dysphagia in half, evaluated by manometrics and barium studies. In which was evidenced in our free colon and ileocolon

1414 Volume 122, Number 5 • Free Colon versus Free Jejunum Flap group as well. Other studies using videofluoroscopy of the reconstructed segment. There are few re- have confirmed the absence of peristalsis and tem- ports on the contractile behavior of free colon22,26,27 poral stagnation at the anastomosis site.21,22 and ileocolon transfers,28 but all refer to the absence Smith et al.23 carried out an experimental radio- of peristalsis and its role as a passive conduit, ac- logic assessment of the functional differences be- counting for our findings of a delayed emptying tween jejunum and colon for free esophageal re- pattern. construction in a canine model. They highlighted The differences evidenced by our scinti- that jejunal flaps strongly retained their peristaltic graphic study paralleled the clinical outcomes contractions initiated with swallowing. However, assessed during rehabilitation. Clinically, the this activity was not coordinated with the remain- two groups of patients showed different progress ing esophagus, leading to a “functional stenosis.” through the swallowing rehabilitation process. The colonic flaps did not exhibit peristalsis; how- Patients who underwent reconstruction with jeju- ever, they dilated on filling with a liquid bolus and nal flaps resumed oral semisolid food intake 1 contracted as it emptied. They concluded that month after the operation, and all except two colon represented a more natural functional re- patients progressed to a solid diet after 12 months. placement of the cervical esophagus because it On the contrary, patients who underwent recon- integrated more naturally into overall swallowing struction with free colon and ileocolon flaps were action. very slow to make progress through the rehabili- In this investigation, we used radionuclide tation process, and some of them needed to oc- esophageal scintigraphy to evaluate our series of clude the nose to push the food down through the cervical esophageal reconstructions using free co- neopharyngoesophagus. Their diet was thus ini- lon and free jejunal flaps. Radionuclide esopha- tially limited to liquid foods, with later progress geal scintigraphy has been shown to be a safe, into semisolid and solid foods. These clinical re- noninvasive, quantitative, and sensitive method of sults coincide with previously published clinical evaluating esophageal motility.8 The radionuclide data but are in contrast with the conclusions esophageal transit scan evidenced delayed initial drawn by Smith et al.23 pharyngeal clearance in both groups in comparison with the normal population, as expected. The intrinsic motility of both conduit types failed to CONCLUSIONS restore the normal physiological mechanism of Postoperative radiotherapy has been reported swallowing. Although free jejunal and colonic to affect the peristalsis in intestinal flaps and to flaps display different motility patterns,18–20,22 we positively influence the occurrence of strictures.29 revealed no significant differences in the pharyn- Although this study’s aim was not to evaluate the geal or initial transit rate between the jejunum and effects of postoperative radiotherapy on the func- colon reconstructions. This is perhaps attributable tional outcome of intestinal flaps, we did note that to the base of the tongue, which was preserved in patients demonstrated continuous improvement all cases, being the propulsion force and the most of swallowing even following radiotherapy. Based important factor in the first part of swallowing.24,25 on our clinical evidence and confirmed by the Removal of the larynx should result in loss of nuclear medicine transit study, we conclude that laryngeal elevation and passive opening of the free jejunum flaps have better swallowing charac- upper esophageal sphincter, affecting the oropha- teristics and motility in comparison with free co- ryngeal motility and pharyngoesophageal phase of lon and free ileocolon flaps. Unless simultaneous swallowing.21 Absence of functional differences in voice reconstruction is planned, we prefer to use the first stage of swallowing, however, between our free jejunal flaps for cervical esophageal recon- two groups of patients, suggests that laryngectomy struction. does not play a major role in determining the final Hung-Chi Chen, M.D. function of the hypopharyngeal reconstruction. E-Da Hospital Esophageal clearance was reduced in both I-Shou University groups compared with normal; however, the pa- 1, E-Da Road tients who underwent reconstruction with jejunal Jiau-shu Tsuen, Yan-Chau Shiang Kaohsiung 824, Taiwan flaps exhibited a statistically significantly better [email protected] transit rate compared with colon and ileocolon flaps. Jejunum has been shown to retain its inher- REFERENCES 10 ent contractility after transfer, and this charac- 1. Noguchi, T., Uchida, Y., Hashimoto, T., Wada, S., Suko, T., teristic possibly contributes to the faster emptying and Suzuki, M. Reconstruction of the cervical esophagus by

1415 Plastic and Reconstructive Surgery • November 2008

interposition of the transverse colon (in Japanese). Nippon usage of the ‘funnel flap.’ J. Plast. Reconstr. Aesthet. Surg. 59: Geka Gakkai Zasshi 102: 647, 2001. 1241, 2006. 2. Pavlovics, G., Cseke, L., Papp, A., Tizedes, G., Tabar, B. A., 15. Schwartz’s Principles of Surgery, 8th Ed. New York: McGraw-Hill, and Horvath, P. O. Esophagus reconstruction with free je- 2005. P. 1950. junal transfer. Microsurgery 26: 73, 2006. 16. Dantas, R. O., and Mamede, R. C. Motility of the transverse 3. Talbot, S. G., and Cordeiro, P. G. Inverted, stapled J-pouch colon used for esophageal replacement. J. Clin. Gastroenterol. free jejunal transfer for reconstruction of the pharynx and 34: 225, 2002. esophagus. J. Surg. Oncol. 95: 663, 2007. 17. Kronenberger, M. B., and Meyers, A. D. Dysphagia following 4. Reece, G. P., Schusterman, M. A., Miller, M. J., et al. Mor- head and neck cancer surgery. Dysphagia 9: 236, 1994. bidity and functional outcome of free jejunal transfer recon- 18. Meyers, W. C., Seigler, H. F., Hanks, J. B., et al. Postoperative struction for circumferential defects of the pharynx and cer- function of “free” jejunal transplants for replacement of the vical esophagus. Plast. Reconstr. Surg. 96: 1307, 1995. cervical esophagus. Ann. Surg. 192: 439, 1980. 5. Ozkan, O., Chen, H. C., Cigna, E., et al. Intussusception of 19. Kerlin, P., McCafferty, G. J., Robinson, D. W., and Theile, D. a transferred jejunal flap in cervical esophagus reconstruc- Function of a free jejunal “conduit” graft in the cervical tion. Ann. Plast. Surg. 55: 327, 2005. esophagus. Gastroenterology 90: 1956, 1986. 6. Feng, Y., and Jianjiang, L. Functional outcomes of two types 20. Wilson, J. A., Maran, A. G., Pryde, A., Walker, W. S., and of subtotal colectomy for slow-transit constipation: Ileosig- Heading, R. C. The function of free jejunal autografts in the moidal anastomosis and cecorectal anastomosis. Am. J. Surg. pharyngo-oesophageal segment. J. R. Coll. Surg. Edinb. 40: 195: 73, 2008. 363, 1995. 7. Kinuya, K., Nakajima, K., Kinuya, S., Michigishi, T., Tonami, 21. Moerman, M., Fahimi, H., Ceelen, W., Pattyn, P., and Ver- T., and Takehara, K. Esophageal hypomotility in systemic meersch, H. Functional outcome following colon interposi- sclerosis: Close relationship with pulmonary involvement. tion in total pharyngoesophagectomy with or without laryn- Ann. Nucl. Med. 15: 97, 2001. gectomy. Dysphagia 18: 78, 2003. 8. Mariani, G., Boni, G., Barreca, M., et al. Radionuclide gas- 22. Nakayama, K., Yamamoto, K., Tamiya, T., et al. Experience troesophageal motor studies. J. Nucl. Med. 45: 1004, 2004. with free autografts of the bowel with a new venous anasto- 9. Seidenberg, B., and Hurwitt, E. S. Immediate reconstruction mosis apparatus. Surgery 55: 796, 1964. of the cervical esophagus by a revascularized isolated jejunal 23. Smith, R. W., Garvey, C. J., Dawson, P. M., and Davies, D. M. segment. Surg. Forum 9: 413, 1958. Jejunum versus colon for free oesophageal reconstruction: 10. Giovanoli, P., Frey, M., Schmid, S., and Flury, R. Free jeju- An experimental radiological assessment. Br. J. Plast. Surg. 40: num transfers for functional reconstruction after tumour 181, 1987. resections in the oral cavity and the pharynx: Changes of 24. McConnel, F. M., Hester, T. R., Mendelsohn, M. S., and morphology and function. Microsurgery 17: 535, 1996. Logemann, J. A. Manofluorography of deglutition after total 11. McDonough, J. J., and Gluckman, J. L. Microvascular recon- laryngopharyngectomy. Plast. Reconstr. Surg. 81: 346, 1988. struction of the pharyngoesophagus with free jejunal graft. 25. McConnel, F. M., and O’Connor, A. Dysphagia secondary to Microsurgery 9: 116, 1988. head and neck cancer surgery. Acta Otorhinolaryngol. Belg. 48: 12. Kawahara, H., Shiraishi, T., Yasugawa, H., Okamura, K., and 165, 1994. Shirakusa, T. A new surgical technique for voice restoration 26. Flynn, M. B., and Acland, R. D. Free intestinal autografts for after laryngopharyngoesophagectomy with a free ileocolic reconstruction following pharyngolaryngoesophagectomy. graft: Preliminary report. Surgery 111: 569, 1992. Surg. Gynecol. Obstet. 149: 858, 1979. 13. Chen, H. C., Mardini, S., and Yang, C. W. Voice reconstruc- 27. Wei, F. C., Carver, N., Chen, H. C., Tsai, M. H., and Wang, tion using the free ileocolon flap versus the pneumatic ar- J. H. Free colon transfer for pharyngo-oesophageal recon- tificial larynx: A comparison of patients’ preference and struction. Br. J. Plast. Surg. 53: 12, 2000. experience following laryngectomy. J. Plast. Reconstr. Aesthet. 28. Grage, T. B., and Quick, C. A. The use of revascularized Surg. 59: 1269, 2006. ileocolic autografts for primary repair after pharyngolaryn- 14. Mardini, S., Chen, H. C., Salgado, C. J., Ozkan, O., Cigna, E., goesophagectomy. Am. J. Surg. 136: 477, 1978. and Chung, C. C. Free microvascular transfer of the reverse 29. Biel, M. A., and Maisel, R. H. Gross morphologic and func- ileo-colon flap with ileocaecal valve valvuloplasty for recon- tional effects of postoperative radiation on free jejunal au- struction of a pharyngoesophageal defect: Indication and tografts. Laryngoscope 102: 875, 1992.

1416