ORIGINAL ARTICLE
Upper Digestive Tract Reconstruction for Caustic Injuries Mircea Chirica, MD,∗ Marie-Dominique Brette, MD,† Matthieu Faron, MD,∗ Nicolas Munoz Bongrand, MD,∗ Bruno Halimi, MD,∗ Christine Laborde, RN,∗ Emile Sarfati, MD,∗ and Pierre Cattan, MD, PhD∗
Objective: The aim of the study was to compare the short- and long-term outcomes of colopharyngoplasty and esophagocoloplasty for caustic injuries of the upper digestive tract. Background: Simultaneous esophageal and pharyngeal reconstruction by colopharyngoplasty allows regaining nutritional autonomy in patients with severe pharyngoesophageal caustic injuries. Methods: Patients who underwent upper digestive tract reconstruction for caustic injuries by colopharyngoplasty (n = 116) and esophagocoloplasty (n = 122) between 1993 and 2012 were included. Survival and functional outcomes were analyzed. Success was defined as nutritional autonomy after removal of the jejunostomy and tracheotomy tubes. Quality of life was assessed using the QLQ-OG25 and SF12v2 questionnaires. Results: Overall Kaplan-Meyer survival at 1, 5, and 10 years after colopharyngoplasty and esophagocoloplasty were 92%, 74%, 67% and 92%, 83%, 73%, respectively (P = 0.56). Quality of life and functional results (success: 57% vs 95%, P < 0.0001) were impaired after colopharyngoplasty. On multivariate analysis, older age (odds ratio [OR]: 0.94; confidence interval [CI]: 0.91–0.97 P < 0.0001) and pharyngeal reconstruction (OR: 0.05; CI: 0.02–0.13, P < 0.0001) were associated with failure. The decline in success with age was more pronounced after colopharyngoplasty with only 1 (7%) of 15 patients operated after the age of 55 being self-sufficient for eating and breathing. Laryngeal resection during colopharyngoplasty had no influence on success (54% vs 58%, P = 0.67) Conclusions: The need to associate pharyngeal reconstruction during esophageal reconstruction for caustic injuries has a long-term negative impact on functional outcome. Keywords: caustic ingestion, colopharyngoplasty, esophageal reconstruction, esophagocoloplasty (Ann Surg 2015;261:894–901)
S
evere pharyngeal strictures occur in 0.7% to 6% of patients after caustic ingestion.1 Such injuries are usually the result of massive ingestion of corrosive agents and/or of prolonged pharyngeal storage (hesitation to swallow, provoked vomiting). Under these circumstances, severe esophageal and/or gastric sequels (emergency resection, intractable strictures) are also present, and the restoration of upper digestive tract continuity requires concomitant esophageal and pharyngeal reconstruction. Surgical management of pharyngeal strictures is a difficult challenge as mirrored by the large diversity of surgical procedures described for its treatment.2–10 Throughout time, most of these techniques fell into disuse because of high morbidity and low success rates. In 2007, we have reported the technique of colopharyngoplasty for the treatment of severe pharyngoesophageal caustic injuries, in From the Departments of ∗ General, Endocrine, and Digestive Surgery; and †Otorhinolaryngology, Saint-Louis Hospital, APHP, Universit´e Paris Diderot, Paris, France. Disclosure: The authors declare no conflicts of interest. Reprints: Pierre Cattan, MD, PhD, Centre Hospitalier Universitaire Saint-Louis, 1 avenue Claude Vellefaux, 75475 PARIS Cedex 10, France. E-mail: [email protected]. C 2014 Wolters Kluwer Health, Inc. All rights reserved. Copyright ISSN: 0003-4932/14/26105-0894 DOI: 10.1097/SLA.0000000000000718
894 | www.annalsofsurgery.com
which the same colonic transplant was used for both esophageal and pharyngeal reconstruction.11 The study included 58 patients and showed promising results. Immediate and long-term outcomes of colopharyngoplasty were in the range of those reported after esophagocoloplasty without pharyngeal reconstruction.11 After publication of the initial report, the number of colopharyngoplasty procedures rose steadily every year; currently more than half of the reconstructive procedures for caustic injuries performed at our center include pharyngeal reconstruction. The aim of this study was to report the results of colopharyngoplasty performed in a large cohort of patients at the Saint Louis Hospital in Paris since its first description in 1993. Outcomes were compared with those of esophagocoloplasty without pharyngeal reconstruction performed at our center across the same period.
PATIENTS AND METHODS Patients From January 1993 to December 2012, 267 patients underwent esophageal reconstruction with the colon for caustic injuries at the Saint Louis Hospital in Paris. Patients referred for salvage management after failed reconstruction (n = 29) were excluded. The medical records of 238 patients who underwent first hand reconstruction at our center were retrospectively reviewed and they were the subject of the study. Emergency esophageal resection had been undertaken in 180 patients (76%), whereas esophageal reconstruction was performed because of intractable strictures in 58 patients (24%).
Preoperative Evaluation The management protocol of patients who required esophageal reconstruction has been previously reported.11,12 Briefly, reconstruction was offered to psychologically stable patients after careful psychiatric evaluation. Otolaryngology (ENT) evaluation was the key examination for subsequent reconstruction planning. ENT systematically included fiberoptic nasopharyngoscopy and hypopharyngoscopy with direct laryngoscopy under general anesthesia and was performed in all patients 1 to 3 months after ingestion.11 If the scarring process was still evolving, ENT evaluation was repeated at 6 months. At the beginning of the study period, the minimal delay in reconstruction was 3 months for esophagocoloplasty and 6 months for colopharyngoplasty; starting July 2007, a minimal 6-month delay in reconstruction was respected in all patients, according to the results of a recent publication from our group.12 Preoperative colonoscopy was performed before reconstruction in patients over the age of 55 years and in those with a family history of colon cancer to rule out colonic malignancy.
Technical Considerations Esophagocoloplasty The technique of esophagocoloplasty after caustic ingestion has been described in detail elsewhere.12–14 In brief, a right ileocolic graft was used whenever possible. The cervical esophagus was sewn to the ileum or the colon as suitable. The abdominal intestinal continuity was restored by cologastrostomy on the anterior gastric wall whenever possible; otherwise, reconstruction was performed by Annals of Surgery r Volume 261, Number 5, May 2015
Copyright © 2014 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 261, Number 5, May 2015
either a coloduodenal anastomosis on the second part of the duodenum or a Roux-en-Y loop.
Colopharyngoplasty The technique of pharyngeal reconstruction suffered no major modification since its introduction in 1993.11 Pharyngeal reconstruction was performed by the same surgeon (M.D.B.) across the whole study period. A tracheotomy was constructed at the beginning of the operation. Two distinct cervical operations were performed according to preoperative ENT findings11 : (1) Standard colopharyngoplasty was undertaken to treat hypopharyngeal stenosis (burns involving the esophageal inlet, the piriform sinuses, and the posterior area) and subglottic strictures (adhesions from the retrocricoid or the retroarytenoid area extending to the posterior pharyngeal wall); (2) Colopharyngoplasty with concomitant partial laryngeal resection was performed to treat supraglottic strictures (obliteration of the supraglottic airway with pharyngeal stenosis at epiglottis level or the base of the tongue). Supraglottic laryngectomy was undertaken if ENT showed severe injuries of the epiglottis, whereas suprahyoid pharyngectomy with resection of the base of the tongue was performed if the burns involved the tongue (Fig. 1). The surgical approaches of the larynx were inspired by techniques used in ENT cancer and were guided by the location of burns and the resection/reconstruction required. A lateral approach was used when performing standard colopharyngoplasty. When using this approach a subperichondral resection of the superior/posterior portion of the thyroid cartilage was performed systematically. This manoeuvre offers a large access to the hypopharynx allowing accurate evaluation of burns extent and resection of all scarred tissues and facilitates confection of the pharyngocolonic anastomosis. An anterior approach through the larynx was used whenever partial laryngeal resection was undertaken. In this situation, the pharynx and the larynx were exposed after resection of the upper third of the thyroid cartilage immediately above the insertion of the vocal cords (Fig. 1). A posterior approach was never used as it may hinder access to the supraglottic area. During both operations, the neopharynx was constructed using the cervical end of the coloplasty, which was cut in an oblique fashion to obtain a bevel-edged surface, which covered the remaining raw surface after scar excision. The colonic mucosa was sewn to the mucosa of the posterior pharyngeal wall upward and backward and to the retroarytenoid or retrocricoid mucosa forward. After laryngeal
Pharyngeal Reconstruction for Caustic Injuries
resection, the hyoid bone and the base of the tongue were lowered to the thyroid cartilage by 3 separate stitches. Distinctive features of the abdominal operation during colopharyngoplasty included construction of a longer colonic graft to reach the pharynx and abdominal reconstruction with a Roux-en-Y loop to minimize reflux. In case of initial conservative management, the native esophagus was not systematically resected at the time of reconstruction.12 Since January 2000, thoracic inlet opening by resection of the head of the left clavicle and the sternal manubrium has been performed systematically to avoid graft compression during both esophagoplasty and colopharyngoplasty.12
Postoperative Management In the absence of operative complications, enteral nutrition by the jejunostomy tube was started on the postoperative day (POD) 5 in all patients.
Reestablishment of Oral Feeding After esophagocoloplasty, oral feeding was started on POD 7 in the absence of operative complications; liquids were introduced first and were followed progressively by a semisolid and then solid diet. After colopharyngoplasty, attempts to occlude the tracheotomy tube were initiated on POD 10; oral feeding was started as soon as the patient tolerated the occlusion of the tracheostomy tube for 24 hours; feeding attempts were performed with an obturated canula and a deflated cuff. A semisolid diet was introduced first and liquids were introduced later if no recurrent aspiration occurred. Several principles were respected to optimize this step. Initial feeding attempts were undertaken under uninterrupted surveillance of the health care team. The patient remained seated half an hour before, during, and 1 hour after oral ingestion to avoid reflux-related aspiration. Deglutition was undertaken with the head in an anteflexion position to block the airway and help food passage into the digestive conduit. The morning meal was introduced at first. When this meal was well tolerated for at least 3 days, the following meals (noon, evening) were progressively introduced, one meal at a time. Jelly-like texture food was introduced initially, followed by yoghurts and eventually, by mixed aliments. Changes in food texture were always introduced during the first day meal and then progressively extended to the next meals. The tracheotomy tube was removed after uneventful ingestion of semisolid
FIGURE 1. Surgical approaches to the larynx: resection of the superior/posterior portion of the thyroid cartilage during standard colopharyngoplasty (A); Frontal view showing resection of the thyroid cartilage above insertion of the vocal cords with resection of the epiglottis (B) and of the hyoid bone (C) during supraglottic laryngectomy and suprahyoid pharyngectomy, respectively; Sagittal view of the resection planes during supraglottic laryngectomy (D) and suprahyoid pharyngectomy (E). a indicates arytenoids; e, epiglottis; h, hyoid bone; t, thyroid cartilage. C 2014 Wolters Kluwer Health, Inc. All rights reserved.
www.annalsofsurgery.com | 895
Copyright © 2014 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 261, Number 5, May 2015
Chirica et al
food for 5 consecutive days. Liquids were eventually started using soda water between meals to minimize reflux and aspiration.
Follow-up The whole process of in-hospital reeducation was performed by the ENT surgeon (M.D.B.) and by a dedicated nurse in the digestive unit (C.L.). Management was then continued in facility care structures with orthophony and psychological support until the patient’s condition allowed an outpatient follow-up. The jejunostomy tube was removed during an outpatient visit if weight and psychological conditions were stable and the daily oral intake exceeded 1500 kcal. Clinical evaluation was done every 4 months during the first 2 years and annually thereafter. Psychological follow-up was pursued indefinitely.
Statistical Analysis Postoperative mortality and morbidity were defined as death or complications, respectively, within 90 days after reconstruction. Coloplasty dysfunction was defined as any graft-related condition interfering with the process of oral alimentary intake. Late morbidity included all complication arising after POD 90 and coloplasty dysfunctions occurring at any time. Functional outcomes were evaluated in patients with a follow-up exceeding 1 year. Success was defined as the ability to remove both the tracheotomy and the jejunostomy tubes. Quality of life was assessed using the QLQ-OG2515 and SF12v216 questionnaires. The QLQ-OG25 assesses several items of esophageal and gastric function; scores range from 25 to 100, with lower scores indicating better function. The SF12v2 physical (PCS) and mental (MCS) component summaries measure general aspects of health-related quality of life; scores are calibrated so that 50 is the average score or norm. Quality of life was assessed in patients followed for more than 1 year after reconstruction; questionnaires
were self-administered. The attending surgeon explained the purpose of the questionnaire, indicated that all data would be kept confidential and, if necessary, provided help with reading items or wording the responses. Continuous variables are presented as median [interquartile range(IQR)] and compared with a Wilcoxon test. Categorical variables are presented as count (percentage) and compared with a χ 2 test or Fisher exact test as appropriate. Any variable achieving a P = 0.05 significance in the univariate analysis for success was proposed in a backward logistic regression model. In the logistic regression model, continuous variables were coded with 5-knots restricted cubic splines to relax the linearity assumption. An interaction term was used in the multivariate model seeking for a variation in the effect of age over success according to the type of reconstruction. Survival curves were calculated according to Kaplan-Meier, from the date of reconstruction to the date of death and were compared with a log-rank test. A P < 0.05 was used to denote statistical significance. All analyses were done with R 3.0.2 (The R Core Team, Vienna, Austria).
RESULTS Characteristics of Patients Of 238 patients who underwent esophageal replacement with the colon for caustic injuries, 122 patients (51%) underwent esophagocoloplasty and 116 patients (49%) colopharyngoplasty. The colopharyngoplasty and esophagocoloplasty groups were similar regarding age, sex, psychiatric history, and the nature of ingested agent (Table 1). Ingestion with suicidal intent (96% vs 89%, P = 0.04) and emergency esophagectomy were more frequent (84% vs 68%, P = 0.005) in patients who underwent colopharyngoplasty. During emergency management, patients in the colopharyngoplasty group were also more likely to require tracheotomy construction (66% vs 14%, P < 0.0001) and experienced increased operative morbidity (57% vs 43%, P = 0.03).
TABLE 1. Characteristics of Patients Who Underwent Esophagocoloplasty (n = 122) and Colopharyngoplasty (n = 116) Reconstruction for Caustic Injuries
Age (yrs) Men Psychiatric disease Suicide attempt Acid ingestion Emergency esophagectomy Extended resection∗ Emergency tracheotomy Delay in reconstruction (mo) Graft Right colon Left colon Graft pedicle Right colic Colica media Left colic Abdominal reconstruction Coloduodenal Roux-en-Y Cologastric Thoracic inlet opening Operative time (min)
Esophagocoloplasty (n = 122)
Colopharyngoplasty (n = 116)
40.5 [28–50] 66 (54) 71 (58) 108 (89) 18 (15) 83 (68) 5 (4) 17 (14) 5 [4–7]
39 [29–47] 67 (58) 77 (66) 111 (96) 25 (22) 97 (84) 10 (9) 76 (66) 8 [6–11]
110 (90) 12 (10)
94 (81) 22 (19)
86 (71) 20 (16) 16 (13)
8 (7) 40 (35) 68 (59)
64 (53) 21 (17) 37 (30) 71 (58) 390 [300–480]
33 (28) 74 (64) 9 (8) 102 (88) 420 [360–490]
P 0.31 0.57 0.19 0.04 0.17 0.005 0.15
Upper Digestive Tract Reconstruction for Caustic Injuries Mircea Chirica, MD,∗ Marie-Dominique Brette, MD,† Matthieu Faron, MD,∗ Nicolas Munoz Bongrand, MD,∗ Bruno Halimi, MD,∗ Christine Laborde, RN,∗ Emile Sarfati, MD,∗ and Pierre Cattan, MD, PhD∗
Objective: The aim of the study was to compare the short- and long-term outcomes of colopharyngoplasty and esophagocoloplasty for caustic injuries of the upper digestive tract. Background: Simultaneous esophageal and pharyngeal reconstruction by colopharyngoplasty allows regaining nutritional autonomy in patients with severe pharyngoesophageal caustic injuries. Methods: Patients who underwent upper digestive tract reconstruction for caustic injuries by colopharyngoplasty (n = 116) and esophagocoloplasty (n = 122) between 1993 and 2012 were included. Survival and functional outcomes were analyzed. Success was defined as nutritional autonomy after removal of the jejunostomy and tracheotomy tubes. Quality of life was assessed using the QLQ-OG25 and SF12v2 questionnaires. Results: Overall Kaplan-Meyer survival at 1, 5, and 10 years after colopharyngoplasty and esophagocoloplasty were 92%, 74%, 67% and 92%, 83%, 73%, respectively (P = 0.56). Quality of life and functional results (success: 57% vs 95%, P < 0.0001) were impaired after colopharyngoplasty. On multivariate analysis, older age (odds ratio [OR]: 0.94; confidence interval [CI]: 0.91–0.97 P < 0.0001) and pharyngeal reconstruction (OR: 0.05; CI: 0.02–0.13, P < 0.0001) were associated with failure. The decline in success with age was more pronounced after colopharyngoplasty with only 1 (7%) of 15 patients operated after the age of 55 being self-sufficient for eating and breathing. Laryngeal resection during colopharyngoplasty had no influence on success (54% vs 58%, P = 0.67) Conclusions: The need to associate pharyngeal reconstruction during esophageal reconstruction for caustic injuries has a long-term negative impact on functional outcome. Keywords: caustic ingestion, colopharyngoplasty, esophageal reconstruction, esophagocoloplasty (Ann Surg 2015;261:894–901)
S
evere pharyngeal strictures occur in 0.7% to 6% of patients after caustic ingestion.1 Such injuries are usually the result of massive ingestion of corrosive agents and/or of prolonged pharyngeal storage (hesitation to swallow, provoked vomiting). Under these circumstances, severe esophageal and/or gastric sequels (emergency resection, intractable strictures) are also present, and the restoration of upper digestive tract continuity requires concomitant esophageal and pharyngeal reconstruction. Surgical management of pharyngeal strictures is a difficult challenge as mirrored by the large diversity of surgical procedures described for its treatment.2–10 Throughout time, most of these techniques fell into disuse because of high morbidity and low success rates. In 2007, we have reported the technique of colopharyngoplasty for the treatment of severe pharyngoesophageal caustic injuries, in From the Departments of ∗ General, Endocrine, and Digestive Surgery; and †Otorhinolaryngology, Saint-Louis Hospital, APHP, Universit´e Paris Diderot, Paris, France. Disclosure: The authors declare no conflicts of interest. Reprints: Pierre Cattan, MD, PhD, Centre Hospitalier Universitaire Saint-Louis, 1 avenue Claude Vellefaux, 75475 PARIS Cedex 10, France. E-mail: [email protected]. C 2014 Wolters Kluwer Health, Inc. All rights reserved. Copyright ISSN: 0003-4932/14/26105-0894 DOI: 10.1097/SLA.0000000000000718
894 | www.annalsofsurgery.com
which the same colonic transplant was used for both esophageal and pharyngeal reconstruction.11 The study included 58 patients and showed promising results. Immediate and long-term outcomes of colopharyngoplasty were in the range of those reported after esophagocoloplasty without pharyngeal reconstruction.11 After publication of the initial report, the number of colopharyngoplasty procedures rose steadily every year; currently more than half of the reconstructive procedures for caustic injuries performed at our center include pharyngeal reconstruction. The aim of this study was to report the results of colopharyngoplasty performed in a large cohort of patients at the Saint Louis Hospital in Paris since its first description in 1993. Outcomes were compared with those of esophagocoloplasty without pharyngeal reconstruction performed at our center across the same period.
PATIENTS AND METHODS Patients From January 1993 to December 2012, 267 patients underwent esophageal reconstruction with the colon for caustic injuries at the Saint Louis Hospital in Paris. Patients referred for salvage management after failed reconstruction (n = 29) were excluded. The medical records of 238 patients who underwent first hand reconstruction at our center were retrospectively reviewed and they were the subject of the study. Emergency esophageal resection had been undertaken in 180 patients (76%), whereas esophageal reconstruction was performed because of intractable strictures in 58 patients (24%).
Preoperative Evaluation The management protocol of patients who required esophageal reconstruction has been previously reported.11,12 Briefly, reconstruction was offered to psychologically stable patients after careful psychiatric evaluation. Otolaryngology (ENT) evaluation was the key examination for subsequent reconstruction planning. ENT systematically included fiberoptic nasopharyngoscopy and hypopharyngoscopy with direct laryngoscopy under general anesthesia and was performed in all patients 1 to 3 months after ingestion.11 If the scarring process was still evolving, ENT evaluation was repeated at 6 months. At the beginning of the study period, the minimal delay in reconstruction was 3 months for esophagocoloplasty and 6 months for colopharyngoplasty; starting July 2007, a minimal 6-month delay in reconstruction was respected in all patients, according to the results of a recent publication from our group.12 Preoperative colonoscopy was performed before reconstruction in patients over the age of 55 years and in those with a family history of colon cancer to rule out colonic malignancy.
Technical Considerations Esophagocoloplasty The technique of esophagocoloplasty after caustic ingestion has been described in detail elsewhere.12–14 In brief, a right ileocolic graft was used whenever possible. The cervical esophagus was sewn to the ileum or the colon as suitable. The abdominal intestinal continuity was restored by cologastrostomy on the anterior gastric wall whenever possible; otherwise, reconstruction was performed by Annals of Surgery r Volume 261, Number 5, May 2015
Copyright © 2014 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 261, Number 5, May 2015
either a coloduodenal anastomosis on the second part of the duodenum or a Roux-en-Y loop.
Colopharyngoplasty The technique of pharyngeal reconstruction suffered no major modification since its introduction in 1993.11 Pharyngeal reconstruction was performed by the same surgeon (M.D.B.) across the whole study period. A tracheotomy was constructed at the beginning of the operation. Two distinct cervical operations were performed according to preoperative ENT findings11 : (1) Standard colopharyngoplasty was undertaken to treat hypopharyngeal stenosis (burns involving the esophageal inlet, the piriform sinuses, and the posterior area) and subglottic strictures (adhesions from the retrocricoid or the retroarytenoid area extending to the posterior pharyngeal wall); (2) Colopharyngoplasty with concomitant partial laryngeal resection was performed to treat supraglottic strictures (obliteration of the supraglottic airway with pharyngeal stenosis at epiglottis level or the base of the tongue). Supraglottic laryngectomy was undertaken if ENT showed severe injuries of the epiglottis, whereas suprahyoid pharyngectomy with resection of the base of the tongue was performed if the burns involved the tongue (Fig. 1). The surgical approaches of the larynx were inspired by techniques used in ENT cancer and were guided by the location of burns and the resection/reconstruction required. A lateral approach was used when performing standard colopharyngoplasty. When using this approach a subperichondral resection of the superior/posterior portion of the thyroid cartilage was performed systematically. This manoeuvre offers a large access to the hypopharynx allowing accurate evaluation of burns extent and resection of all scarred tissues and facilitates confection of the pharyngocolonic anastomosis. An anterior approach through the larynx was used whenever partial laryngeal resection was undertaken. In this situation, the pharynx and the larynx were exposed after resection of the upper third of the thyroid cartilage immediately above the insertion of the vocal cords (Fig. 1). A posterior approach was never used as it may hinder access to the supraglottic area. During both operations, the neopharynx was constructed using the cervical end of the coloplasty, which was cut in an oblique fashion to obtain a bevel-edged surface, which covered the remaining raw surface after scar excision. The colonic mucosa was sewn to the mucosa of the posterior pharyngeal wall upward and backward and to the retroarytenoid or retrocricoid mucosa forward. After laryngeal
Pharyngeal Reconstruction for Caustic Injuries
resection, the hyoid bone and the base of the tongue were lowered to the thyroid cartilage by 3 separate stitches. Distinctive features of the abdominal operation during colopharyngoplasty included construction of a longer colonic graft to reach the pharynx and abdominal reconstruction with a Roux-en-Y loop to minimize reflux. In case of initial conservative management, the native esophagus was not systematically resected at the time of reconstruction.12 Since January 2000, thoracic inlet opening by resection of the head of the left clavicle and the sternal manubrium has been performed systematically to avoid graft compression during both esophagoplasty and colopharyngoplasty.12
Postoperative Management In the absence of operative complications, enteral nutrition by the jejunostomy tube was started on the postoperative day (POD) 5 in all patients.
Reestablishment of Oral Feeding After esophagocoloplasty, oral feeding was started on POD 7 in the absence of operative complications; liquids were introduced first and were followed progressively by a semisolid and then solid diet. After colopharyngoplasty, attempts to occlude the tracheotomy tube were initiated on POD 10; oral feeding was started as soon as the patient tolerated the occlusion of the tracheostomy tube for 24 hours; feeding attempts were performed with an obturated canula and a deflated cuff. A semisolid diet was introduced first and liquids were introduced later if no recurrent aspiration occurred. Several principles were respected to optimize this step. Initial feeding attempts were undertaken under uninterrupted surveillance of the health care team. The patient remained seated half an hour before, during, and 1 hour after oral ingestion to avoid reflux-related aspiration. Deglutition was undertaken with the head in an anteflexion position to block the airway and help food passage into the digestive conduit. The morning meal was introduced at first. When this meal was well tolerated for at least 3 days, the following meals (noon, evening) were progressively introduced, one meal at a time. Jelly-like texture food was introduced initially, followed by yoghurts and eventually, by mixed aliments. Changes in food texture were always introduced during the first day meal and then progressively extended to the next meals. The tracheotomy tube was removed after uneventful ingestion of semisolid
FIGURE 1. Surgical approaches to the larynx: resection of the superior/posterior portion of the thyroid cartilage during standard colopharyngoplasty (A); Frontal view showing resection of the thyroid cartilage above insertion of the vocal cords with resection of the epiglottis (B) and of the hyoid bone (C) during supraglottic laryngectomy and suprahyoid pharyngectomy, respectively; Sagittal view of the resection planes during supraglottic laryngectomy (D) and suprahyoid pharyngectomy (E). a indicates arytenoids; e, epiglottis; h, hyoid bone; t, thyroid cartilage. C 2014 Wolters Kluwer Health, Inc. All rights reserved.
www.annalsofsurgery.com | 895
Copyright © 2014 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 261, Number 5, May 2015
Chirica et al
food for 5 consecutive days. Liquids were eventually started using soda water between meals to minimize reflux and aspiration.
Follow-up The whole process of in-hospital reeducation was performed by the ENT surgeon (M.D.B.) and by a dedicated nurse in the digestive unit (C.L.). Management was then continued in facility care structures with orthophony and psychological support until the patient’s condition allowed an outpatient follow-up. The jejunostomy tube was removed during an outpatient visit if weight and psychological conditions were stable and the daily oral intake exceeded 1500 kcal. Clinical evaluation was done every 4 months during the first 2 years and annually thereafter. Psychological follow-up was pursued indefinitely.
Statistical Analysis Postoperative mortality and morbidity were defined as death or complications, respectively, within 90 days after reconstruction. Coloplasty dysfunction was defined as any graft-related condition interfering with the process of oral alimentary intake. Late morbidity included all complication arising after POD 90 and coloplasty dysfunctions occurring at any time. Functional outcomes were evaluated in patients with a follow-up exceeding 1 year. Success was defined as the ability to remove both the tracheotomy and the jejunostomy tubes. Quality of life was assessed using the QLQ-OG2515 and SF12v216 questionnaires. The QLQ-OG25 assesses several items of esophageal and gastric function; scores range from 25 to 100, with lower scores indicating better function. The SF12v2 physical (PCS) and mental (MCS) component summaries measure general aspects of health-related quality of life; scores are calibrated so that 50 is the average score or norm. Quality of life was assessed in patients followed for more than 1 year after reconstruction; questionnaires
were self-administered. The attending surgeon explained the purpose of the questionnaire, indicated that all data would be kept confidential and, if necessary, provided help with reading items or wording the responses. Continuous variables are presented as median [interquartile range(IQR)] and compared with a Wilcoxon test. Categorical variables are presented as count (percentage) and compared with a χ 2 test or Fisher exact test as appropriate. Any variable achieving a P = 0.05 significance in the univariate analysis for success was proposed in a backward logistic regression model. In the logistic regression model, continuous variables were coded with 5-knots restricted cubic splines to relax the linearity assumption. An interaction term was used in the multivariate model seeking for a variation in the effect of age over success according to the type of reconstruction. Survival curves were calculated according to Kaplan-Meier, from the date of reconstruction to the date of death and were compared with a log-rank test. A P < 0.05 was used to denote statistical significance. All analyses were done with R 3.0.2 (The R Core Team, Vienna, Austria).
RESULTS Characteristics of Patients Of 238 patients who underwent esophageal replacement with the colon for caustic injuries, 122 patients (51%) underwent esophagocoloplasty and 116 patients (49%) colopharyngoplasty. The colopharyngoplasty and esophagocoloplasty groups were similar regarding age, sex, psychiatric history, and the nature of ingested agent (Table 1). Ingestion with suicidal intent (96% vs 89%, P = 0.04) and emergency esophagectomy were more frequent (84% vs 68%, P = 0.005) in patients who underwent colopharyngoplasty. During emergency management, patients in the colopharyngoplasty group were also more likely to require tracheotomy construction (66% vs 14%, P < 0.0001) and experienced increased operative morbidity (57% vs 43%, P = 0.03).
TABLE 1. Characteristics of Patients Who Underwent Esophagocoloplasty (n = 122) and Colopharyngoplasty (n = 116) Reconstruction for Caustic Injuries
Age (yrs) Men Psychiatric disease Suicide attempt Acid ingestion Emergency esophagectomy Extended resection∗ Emergency tracheotomy Delay in reconstruction (mo) Graft Right colon Left colon Graft pedicle Right colic Colica media Left colic Abdominal reconstruction Coloduodenal Roux-en-Y Cologastric Thoracic inlet opening Operative time (min)
Esophagocoloplasty (n = 122)
Colopharyngoplasty (n = 116)
40.5 [28–50] 66 (54) 71 (58) 108 (89) 18 (15) 83 (68) 5 (4) 17 (14) 5 [4–7]
39 [29–47] 67 (58) 77 (66) 111 (96) 25 (22) 97 (84) 10 (9) 76 (66) 8 [6–11]
110 (90) 12 (10)
94 (81) 22 (19)
86 (71) 20 (16) 16 (13)
8 (7) 40 (35) 68 (59)
64 (53) 21 (17) 37 (30) 71 (58) 390 [300–480]
33 (28) 74 (64) 9 (8) 102 (88) 420 [360–490]
P 0.31 0.57 0.19 0.04 0.17 0.005 0.15
