- Email: [email protected]
Function of a free jejunal “conduit” graft in the cervical esophagus
GASTROENTEROLOGY 1986;90:1956-63
Function of a Free Jejunal “Conduit” Graft in the Cervical Esophagus PAUL KERLIN, GERALD J. McCAFFERTY, and DAVID THEILE
DAVID W. ROBINSON,
Gastroenterology Department and the Head and Neck Clinic, Brisbane, Queensland, Australia
The function of an autotransplanted jejunal graft used to reconstruct the pharyngoesophagtis was evaluated in 12 patients, 2-40 mo after surgery. On clinical assessment most patients swallowed liquids and solids with minimal difficulty, although several described a need for liquids to “flush” solids to the stomach. Radiologic studies demonstrated a delay in orogastric transit in some patients above the lower anastomosis. The mechanism was apparent on motility studies: swallows generally failed to induce contractions in the graft, although the esophagus below exhibited stripping peristaltic waves. Regular contractile activity, characteristic of phase III of the intestinal migrating motor complex, was identified in 11 grafts. There WQSno temporal association between migrating motor complexes in the graft and those recorded at the jejunal donor site. Instillation of nutrients into the gastric antrum induced a typical ‘fed” pattern of contractions in the intact jejunum but not in the extrinsically denervated graft. In conclusion, the graft provides a useful, though generally passive conduit. The graft maintains its intrinsic motor repertoire, which is asynchronous with that of the donor site. The findings also support the hypothesis that extrinsic nerves are required to induce a ‘fed” pattern of intestinal motility. Received June 17,1985. Accepted December 10, 1985. Address requests for reprints to: Paul Kerlin, M.D., F.R.A.C.P., Gastroenterology Department, Princess Alexandra Hospital, Ipswich Road, Woolloongabba, Queensland, 4102, Australia. This work was supported in part by grants from the National Health and Medical Research Council and the Queensland Cancer Fund. This work was presented, in part, at the Annual Scientific Meeting of the American Gastroenterological Association in New York in May 1985. The authors thank Beverly Harris, R.N., Sandra Kane, R.N., and Allan Sharrock for valuable technical assistance and Mrs. Jan Frizzell for secretarial assistance. 0 1986 by the American Gastroenterological Association 0016-5085/86/$3.50
Princess
Alexandra
Hospital,
After radical surgical excision of cancer of the oropharynx, hypopharynx, and cervical esophagus, reconstruction of the pharynx and upper esophagus is required. One of the options the surgeon may consider is to utilize a free jejunal “conduit” graft to bridge the defect (l-4). In this one-stage operation another member of the surgical team harvests a segment of jejunum of appropriate length supplied by a suitable vascular pedicle. The intestinal segment is placed isoperistaltically and secured to restore pharyngoesophageal continuity. The graft is revascularized by microvascular surgical techniques using recipient vessels of suitable size in the neck (e.g., superior thyroid artery and vein). Patency of the vascular anastomoses is suggested by return of a normal pink color and contractions in the graft. A permanent tracheostomy is fashioned, and in the recovery period, speech rehabilitation is achieved by the use of an electronic larynx or a “valve” prosthesis that permits esophageal speech. In general, for advanced carcinoma, surgery is preceded by cytotoxic therapy and followed by a course of radiotherapy (3). With advances in microvascular techniques, the use of free jejunal transfers for reconstruction after laryngoesophagectomy has become an accepted procedure. Further, the one-stage procedure preserves the distal esophagus and the lower esophageal sphincter. The aims of this study were to evaluate the swallowing ability of a series of patients after this operation and to investigate the motor function of the graft in relation to the adjacent esophagus and to the donor site in the intact jejunum.
Materials and Methods From a group of 56 patients pharyngolaryngectomy and insertion Abbreviation plex.
who had undergone of a free jejunal graft,
used in this paper: MMC, migrating
motor com-
lune 1986
FUNCTION OF A FREE JEJIJNAL “CONI)LJIT” GRAFT
1957
barium and bread soaked in barium (to stimulate solids) were used. The location of any delay or “hold-up” in transit of a bolus was noted. Manometric
Figure
1. Barium swallow examination of the jejunal interposition graft. Arrows indicate the upper and lower margins of tht? graft.
12 patients were selected for investigation. The selection process was dependent on the availability of the patient for study. The patients (11 men, 1 woman) ranged in age from 51 to 75 yr (mean ? SD, 60.4 ? 7.2 yr). Their grafts measured from 5 to 14 cm (8.5 * 2.2 cm) in length, and at the time of study the patients were 2-40 mo postoperative. The research protocol was approved by the Princess Alexandra Hospital Ethics Committee and each patient gave written informed consent. Assessment of graft function included a clinical assessment of swallowing ability, radiologic studies, and motility recordings. CJinicaJ
Assessment
The patient’s ability to swallow liquids and solids was determined by interview. Any changes in the consistency of food chosen and an estimate of times required to complete meals were noted. A history of regurgitation was sought and changes in postoperative weight were recorded. Upper gastrointestinal endoscopy was performed if the patient complained of dysphagia or had a delay on barium studies.
After a 12-h fast, motility was recorded by a lowcompliance infusion system using 0.5..mlimin flow rates (5). Each subject had recordings from the graft and distal native esophagus. In this study a six-lumen polyvinyl tube incorporated perfusion sites at 5, 10, 15, 17.5, 20.0, and 22.5 cm from the distal end. The assembly initially was passed into the stomach and a slow pull-through in 0.5-cm increments was performed to measure the lower esophageal sphincter pressure. The upper sites were then positioned in the graft. Subjects lay in bed at an angle of 45” from the horizontal and received 7-m\ water boluses to swallow every 5-10 min during the 2.5-4-h study. Subjects who tolerated this study well were offered an additional motility study to record simultaneously contractile activity in the jejunal graft and the intact jejunum. In this study a seven-lumen tube (OD = 4.2 mm) incorporated perfusion sites at 0,20,40, 100,105,110 cm from the distal end. The final radiopaque catheter was used to inflate and deflate a balloon device, which allowed quick placement of the tube (6). During these recordings the perfusion system featured an electronic integrator (7) that recorded the motility index (area under the curve) each minute for the six recording sites. After 4 h of recording during fasting, subjects swallowed an additional fine tube through which 200 ml of Ensure-Plus (Ross Laboratories, Columbus, Ohio) was delivered directly into the stomach. This tube was then flushed clear with water and removed. Motor recording continued for 2 h after the test meal. Analysis
Each subject underwent (Figure 1) of deglutition and
shown
swallowed fluoroscopic graft function.
examination Both liquid
Data
Results are
Assessment
of Motility
The response to deglutition was assessed by comparing the timing and nature of motor responses in the graft with contractions in the adjacent esophagus. Recordings from the graft were examined for migrating motor complex (MMC)-like activity. This required continuous contractile activity for more than 3 min from at least one perfusion site located at the level of the graft. Data concerning the frequency of contractions during maximal activity and the interval between MMC-like activities were noted. In recordings from the jejunak graft and intact jejunum the temporal relationship between MMCs in the two locations was analyzed by visual inspection. After the test meal the motor patterns in the autotransplanted (extrinsically denervated) jejunum and the intact innervated (donor) jejunum were compared.
Results Radiologic
Assessment
of clinical in
Table
liquids
and
radiologic
1. Eleven and
solids
culty. Seven patients gained Several subjects described
assessments
of the with
12
patients
minimal
weight postoperatively. occasional episodes
diffiof
1958
GASTROENTEROLOGY Vol. 90, No. 6
KERLIN ET AL.
Table 1. Clinical and Radiologic Assessments Radiologic
Clinical
Patient
Months postoperative
Dysphagia with liquids
+ +++
+++
a
-
+
a
-
EC A.P. MS. A.G. D.W.
33 31 a a 20
10
P.L. E.H.
4 40
10
R.T. W.R. M.H.
Dysphagia with solids
Length of graft (cm1 9
a 9 7
5 6.5 14 a
Liquid “flush”0
t (liquid)
+
-
Regurgitation
+ t + + +
-
Delay in transit of Liquidb
Solid”
Location
+
t
Lower anastomosis Lower anastomosis
t (solid] -
t ++t
+ +++
t (liquid) -
-
+
-
-
+ (liquid] -
+ -
aperis-
Upper and lower anastomosis Lower anastomosis
-
a Patients use liquids to “flush” ingested solids to stomach. b Liquid barium. c Bread soaked in barium. Assessment: + +, intermittent; + + +, severe.
nasal regurgitation with liquids, and 6 patients described a need for liquids to “flush” ingested solids to the stomach. Most subjects ate more slowly than preoperatively, chewed more thoroughly, and favored soft food. Most patients reported a minor improvement in swallowing with time but could not discount the contribution of adaptation to the new conditions and life-style. The barium studies were particularly useful to assess the function of the graft. The liquid barium and bread soaked in barium passed rapidly to the stomach in 7 subjects. Two patients had a “hold-up” of the liquid and the solid boluses at the lower anastomosis. The subjects were unaware of this phenomenon. The obstruction was clearly “functional” inasmuch as the boluses passed either spontaneously, after a few seconds, or following a sip of water which formed a column of fluid above the apparent narrowing. The lumen at the anastomosis opened widely as the bolus passed. A further patient demonstrated a functional obstruction with bread only at the lower anastomosis. Endoscopy with a flexible fiberscope excluded strictures in each of these patients. Patient E.H., who was 3 yr and 4 mo postoperative and described mild dysphagia with solids, had a delay of barium entering and traversing the graft. Fiberoptic endoscopy disclosed the cause to be a fibrotic stricture of the graft. Mucosal biopsy specimens demonstrated radiation damage. After several dilatations with the Eder Peustow technique, the dysphagia was relieved. The patient has remained asymptomatic throughout a 12-mo observation period. Despite initial good function, patient D.W. developed anorexia and dysphagia and lost weight. Dysphagia was unrelated to the
Esophageal talsis
-, none; +, mild;
graft and was secondary to aperistalsis of the native esophagus. He died soon after investigation from complications of a primary lung cancer. Manometric
Studies
The mean lower esophageal sphincter pressure was 13.7 ? 5.8 mmHg. Three patients had a hypotensive lower esophageal sphincter based on our normal range for healthy controls (lo-25 mmHg). Results of recordings from the graft and distal native esophagus are shown in Table 2. As shown in Table 2 and Figure 2 (left panel), the most common response to a 7-ml bolus of water was no response in the graft and peristalsis in the adjacent esophagus. Similarly, most spontaneous “dry” swallows did not elicit contractions in the jejunal graft. The alternative response of the graft to deglutition is shown in the right panel of Figure 2. This burst of contractions in the graft was not always associated with peristalsis in the esophagus. This response was seen less frequently than no response and appeared to be in part artifactual due to forcing a bolus of air into the graft. During the recordings, regular contractile activity, typical of phase III of the MMC, was present in the jejunal graft of 11 patients. Multiple complexes were observed in the majority. No MMC was recorded in patient J.D. 2 ma after surgery. When the study was repeated at 12 mo after surgery, three MMCs were observed. Figure 3 illustrates the independence of contractile activity in the free jejunal graft and the native esophagus. Deglutition did not interrupt MMC-like activity. Between phase III-like activity
FUNCTION
June 1986
Table
2.
Results
of Monometric
OF A FREE JEJUNAL “CONDUIT”
GRAFT
Studies Pattern of motor response of graft to deglutition
(hl
(%I
(%I
4.0 2.5 3.3 3.5 4.0 4.0 4.0 3.7 3.30 3.0b
70 100
30
4 1
42-45 -
50 50 100 80 55 77
50 50 20 45 23
2 4 2 5 4 0
77 35-73 30 16-80 42-130 -
5-10 3 8 5-9 7 6-24 3.5-15 -
54 74
46 26
0 3
57, 60
4-7
10
63 68 75
37 32 25
3 1 3
63, 75 60,110
3-4 4 11-14
9 10 10
Figure
b Twelve
Duration of phase III (mm)
graft
EC. A.P. MS. A.G. D.W. P.L. E.H. D.P. J.D.
postoperative.
Period of cycle (min)
in jejunal
Patient
a Two months
Number of MMCs
activity
No activity
3.3 4.0 4.0
Multinle contractions
MMC-like
Duration of recording
R.T. W.R. M.H.
1959
-
Contractions during phase III (No./min) 10 10 9 9.3-10 10 8-9.5 9 -
months postoperative.
2. Response of jejunal interposition graft to swallowing. Three perfusion sites, G1--G,, are positioned in the graft and three, EI-E3, are positioned in the adjacent esophagus. The arrows indicate 7-ml boluses of water. The left panel represents the usual response-no response in the graft and peristalsis in the native esophagus. The right panel illustrates a less frequent response--a burst of contractions in the graft and peristalsis in the esophagus.
1960
KERLIN ET AL.
GASTROENTEROLOGY
Vol. 90, No. 6
01 I
1
%
Figure
3. Independence
of the jejunal graft and the esophagus.
Three perfusion
I
sites, G,-G,,
L
I
are positioned
1
6Omm Hg
I
1
I
_
in the graft and three, E,-E,,
are positioned in the adjacent esophagus. Regular contractile activity characteristic of phase III of the migrating motor complex is present in the graft in the left panel. The adjacent esophagus is quiescent. In contrast, the right panel from the same recording illustrates motor quiescence in the graft and peristalsis in the esophagus as the patient swallows.
fronts the grafts were usually quiescent, except for any contractions induced by deglutition or coughing via the tracheotomy. Six subjects had prolonged recordings from the jejunal graft and the intact jejunum (the donor site). At least one perfusion site remained within the graft
throughout the recordings (3-4.5 h, fasting). Migrating motor complex-like activity (2 to 4 cycles) was present during fasting in 5 patients [Figure 4). On visual analysis there was no temporal association between complexes present in the graft and those in the intact jejunum. The rate of contractions during
1
9Omm
h_-__
..
.
Hg
,
3-
Figure
4. Recording
of migrating motor complex from autotransplanted motor complexes at the two locations were asynchronous.
jejunal
graft (G1,Z) and intact jejunum
(J1-J3). The migrating
June 1986
FUNCTION
OF A FREE JEJUNAL
“CONDUIT”
GRAFT
1961
mm Hg
J3
1yA
52
1
1
t
I
Minutes Figure
5. Response of jejunal interposition graft (GI-G3) and intact jejunum (J,-J3) to food. A liquid antrum via a separate tube, thus bypassing the graft. A typical “fed” pattern of contractions in the autotransplanted (extrinsically denervated) graft.
aiways
MMC-like activity in the graft was slower (8.5-10.3/min) than the rate in the corresponding donor jejunum (10.5-1 Urnin). Recordings continued for 1.5-2.5 h after the liquid nutrient meal. In 5 subjects the Ensure-Plus was delivered directly into the stomach by a fine feeding tube which was then removed. This maneuver eliminated nutrient contact with the extrinsically denervated jejunal mucosa. The fifth subject drank the test meal. In each patient a typical “fed” pattern of intestinal motility was established in the intact jejunum. In contrast, the jejunal grafts remained relatively quiescent (excluding activity in the graft associated with deglutition) (Figure 5). In 2 patients MMC-like activity was observed in the graft during the 2 h after the test meal (Figure 6).
Discussion Reconstruction of the pharynx and cervical esophagus after ablative surgery is a major surgical challenge. Although the use of a free revascularized jejunal graft was pioneered by Seidenberg et al. (8) in 1959, more widespread application of the procedure
nutrient meal is delivered is present in the jejunum
into the but not
had to await the advances in microvascular surgical techniques of the last decade. The reconstruction offers a single-stage procedure with low morbidity and short hospitalization (1,3). The majority of our patients can swallow liquids and solids without significant difficulty. Nevertheless, 6 (50%) of the patients studied described a need to “flush” ingested solids with a sip of liquid. The symptom appears to represent a functional obstruction at the lower squamocolumnar anastomosis. During fluoroscopic examination the zone of “hold-up” was demonstrated to open in response to the hydrostatic pressure of a water bolus. The absence of contractions of the graft in response to deglutition explains the lack of adequate propulsive forces in the conduit. In theory, the function of longer jejunal segments could therefore be compromised by this “passive” nature of the graft. Further, longer segments could necessitate additional microvascular anastomoses. Indeed, McDonald et al. (9) have used a free jejunal graft to replace the total esophagus in a patient with a lye stricture. Two vessel groups were anastomosed to recipient vessels with apparently good function. The maximal length of jejunum used
1962
Figure
GASTROENTEROLOGY Vol. 90, No. 6
KERLIN ET AL.
6.
Plot of motility index each minute (electronic integrator] for perfusion site in jejunal graft (G) and intact jejunum (J1-J3) during fasting and after a test meal. The orrow and dotted line indicate delivery of Ensure-Plus directly into the gastric antrum via a separate tube. The migrating motor complexes at the two sites are asynchronous. In contrast to the continuous “fed” activity in the jejunum after the meal, migrating motor complexlike activity continues in the autotransplanted (extrinsically denervated) graft (*).
5
7
E”
is i=4 03 =2 , 0 7 6 5 4 3 2 1 0 0
05
1
1.5
2
2.5
3
TIME (h)
by our group has been a 24-cm graft following excision of an upper esophageal carcinoma. Detachment of the mesentery from each end of the conduit allowed it to be pulled out to its full length. During the motility recordings from the autotransplanted jejunal grafts, bursts of regular contractile activity, typical of phase III of the MMC, were observed. This observation helps to clarify the potential mechanisms for control of intestinal motility. Postulated control influences include extrinsic innervation, neuromuscular continuity of the bowel wall, and humoral factors. The MMC is the characteristic interdigestive motor pattern in the human small intestine (5,lO). During harvesting of the graft, the autonomic nerves that accompany the blood vessels are interrupted along with the division of the vascular arcade. Thus the presence of the MMC in the jejunal graft suggests that its presence in the intestine is independent of extrinsic innervation. The MMC is present also in the porcine and the canine models of autotransplantation (11,121. Results of these animal studies suggest that the initiation and propagation of the MMC within the small intestine resides in the intrinsic intramural neural plexi. A further important observation in our patients was the lack of a temporal association between
““A
4
45
5
5.5
I
MMCs in the autotransplanted jejunal graft and the donor jejunum. Sarr and Kelly (12) also noted complete disruption of the temporal association between motor patterns of the innervated bowel and of the isolated autotransplanted segment in their canine model. It appears that the orderly migration of the MMC from one segment of intestine to the next requires continuity to the intrinsic nerves. Indeed, the MMC will propagate across a site of transection and reanastomosis of the small bowel after regeneration of the intrinsic nerves (13). A third important physiologic observation became apparent in the feeding studies. In humans, food interrupts cyclic motor activity at all levels of the small intestine (5). This transition is prompt and the postprandial motility consists of irregular bursts of contractions interspersed with transient quiescence. However, although a typical “fed” motor pattern was established in the small bowel (donor site), the graft remained quiescent. In a previous study Meyers et al. (2) have reported that feeding induced continuous contractions in the graft. They considered that this response must be controlled by hormones. In contrast, feeding does not interrupt the MMC in the extrinsically denervated canine small bowel (12). The occurrence of MMCs in 2 patients and quiescence in the remainder
June 1986
of our patients in the fed state supports the latter view. Thus, our results and those in the canine model suggest that the extrinsic nerves are necessary for the postprandial inhibition of the MMC. In conclusion, a free jejunal graft provides a useful, though generally passive conduit. A zone of functional hold-up may occur at the lower anastomosis, and such patients may take liquid frequently during a meal to “flush” ingested solids to the stomach. The motility studies have also demonstrated that the autotransplanted jejunal graft maintains its intrinsic motor repertoire, which is asynchronous with that of the donor site. Finally, the findings also support the hypothesis that extrinsic nerves are required to induce a “fed” pattern of intestinal motility. References Robinson DW, MacLeod A. Microvascular free jejunum transfer. Br J Plast Surg 1982;35:258-67. Meyers WC, Seigler HF, Hanks JB, et al. Postoperative function of “free” jejunal transplants for replacement of the cervical esophagus. Ann Surg 1980;192:439-48. McCafferty GJ. Free jejunal graft reconstruction of the pharynx and cervical oesophagus. J Otolaryngol Sot Australia 1983/4;5(3):171-3.
FUNCTION
OF A FREE JEJUNAL
“CONDUlT”
GRAFT
1963
4. Hester TR, McConnel FMS, Nahai F, Jurkiewicz MJ, Brown RG. Reconstruction of cervical esophagus, hypopharynx and oral cavity using free jejunal transfer. Am J Surg 1980;140: 487-91. 5. Kerlin P, Phillips S. Variability of motility of the ileum and jejunum in healthy man. Gastroenterology 1982;82:694-700, R, Phillips SF. Rapid intubation of the 6. Kerlin P, Tucker ileo-colonic region of man. Aust NZ J Med 1983;13:591-3. 7. Kerlin P, Zinsmeister A, Phillips S. Motor responses to food of the ileum, proximal colon, and distal colon of healthy humans. Gastroenterology 1983;84:762--70. B, Rosenak SS, Hurwitt ES, Som ML. Immediate 8. Seidenberg reconstruction of the cervical esophagus by a revascularized isolated jejunal segment. Ann Surg 1959;149:162-71. 9. McDonald
JC, Rohr MS, Tucker
autotransplantation
of the
WY. Recent
kidney.
experiences
jejunum,
and
with
pancreas.
Ann Surg 1983;197:678-87. 10. Vantrappen
G, Janssens
digestive
motor
bacterial
overgrowth
J, Hellemans
complex
of normal of the
J. Ghoos subjects
small
Y. The inter-
and patients
intestine.
J Clin
with Invest
1977;.59:1158-66. 11. Aeberhard control
PF,
Am J Physiol 12. Sarr
Magnenat
of migratory
complex
WA.
Nervous
of the small
bowel.
1980;238:GlOZ-8.
MG, Kelly
planted
LD, Zimmermann
myoelectric
canine
KA. Myoelectric jejunoileum.
activity
of the
Gastroenterology
autotrans-
1981;81:303-
10. 13. Sarna
S, Stoddard
control 1981;241
C, Belbeck
of migrating (Gastrointest
L, McWade
myoelectric Liver Physiol
D. Intrinsic
complexes. 4) G16-23.
nervous
Am J Physiol
Function of a Free Jejunal “Conduit” Graft in the Cervical Esophagus PAUL KERLIN, GERALD J. McCAFFERTY, and DAVID THEILE
DAVID W. ROBINSON,
Gastroenterology Department and the Head and Neck Clinic, Brisbane, Queensland, Australia
The function of an autotransplanted jejunal graft used to reconstruct the pharyngoesophagtis was evaluated in 12 patients, 2-40 mo after surgery. On clinical assessment most patients swallowed liquids and solids with minimal difficulty, although several described a need for liquids to “flush” solids to the stomach. Radiologic studies demonstrated a delay in orogastric transit in some patients above the lower anastomosis. The mechanism was apparent on motility studies: swallows generally failed to induce contractions in the graft, although the esophagus below exhibited stripping peristaltic waves. Regular contractile activity, characteristic of phase III of the intestinal migrating motor complex, was identified in 11 grafts. There WQSno temporal association between migrating motor complexes in the graft and those recorded at the jejunal donor site. Instillation of nutrients into the gastric antrum induced a typical ‘fed” pattern of contractions in the intact jejunum but not in the extrinsically denervated graft. In conclusion, the graft provides a useful, though generally passive conduit. The graft maintains its intrinsic motor repertoire, which is asynchronous with that of the donor site. The findings also support the hypothesis that extrinsic nerves are required to induce a ‘fed” pattern of intestinal motility. Received June 17,1985. Accepted December 10, 1985. Address requests for reprints to: Paul Kerlin, M.D., F.R.A.C.P., Gastroenterology Department, Princess Alexandra Hospital, Ipswich Road, Woolloongabba, Queensland, 4102, Australia. This work was supported in part by grants from the National Health and Medical Research Council and the Queensland Cancer Fund. This work was presented, in part, at the Annual Scientific Meeting of the American Gastroenterological Association in New York in May 1985. The authors thank Beverly Harris, R.N., Sandra Kane, R.N., and Allan Sharrock for valuable technical assistance and Mrs. Jan Frizzell for secretarial assistance. 0 1986 by the American Gastroenterological Association 0016-5085/86/$3.50
Princess
Alexandra
Hospital,
After radical surgical excision of cancer of the oropharynx, hypopharynx, and cervical esophagus, reconstruction of the pharynx and upper esophagus is required. One of the options the surgeon may consider is to utilize a free jejunal “conduit” graft to bridge the defect (l-4). In this one-stage operation another member of the surgical team harvests a segment of jejunum of appropriate length supplied by a suitable vascular pedicle. The intestinal segment is placed isoperistaltically and secured to restore pharyngoesophageal continuity. The graft is revascularized by microvascular surgical techniques using recipient vessels of suitable size in the neck (e.g., superior thyroid artery and vein). Patency of the vascular anastomoses is suggested by return of a normal pink color and contractions in the graft. A permanent tracheostomy is fashioned, and in the recovery period, speech rehabilitation is achieved by the use of an electronic larynx or a “valve” prosthesis that permits esophageal speech. In general, for advanced carcinoma, surgery is preceded by cytotoxic therapy and followed by a course of radiotherapy (3). With advances in microvascular techniques, the use of free jejunal transfers for reconstruction after laryngoesophagectomy has become an accepted procedure. Further, the one-stage procedure preserves the distal esophagus and the lower esophageal sphincter. The aims of this study were to evaluate the swallowing ability of a series of patients after this operation and to investigate the motor function of the graft in relation to the adjacent esophagus and to the donor site in the intact jejunum.
Materials and Methods From a group of 56 patients pharyngolaryngectomy and insertion Abbreviation plex.
who had undergone of a free jejunal graft,
used in this paper: MMC, migrating
motor com-
lune 1986
FUNCTION OF A FREE JEJIJNAL “CONI)LJIT” GRAFT
1957
barium and bread soaked in barium (to stimulate solids) were used. The location of any delay or “hold-up” in transit of a bolus was noted. Manometric
Figure
1. Barium swallow examination of the jejunal interposition graft. Arrows indicate the upper and lower margins of tht? graft.
12 patients were selected for investigation. The selection process was dependent on the availability of the patient for study. The patients (11 men, 1 woman) ranged in age from 51 to 75 yr (mean ? SD, 60.4 ? 7.2 yr). Their grafts measured from 5 to 14 cm (8.5 * 2.2 cm) in length, and at the time of study the patients were 2-40 mo postoperative. The research protocol was approved by the Princess Alexandra Hospital Ethics Committee and each patient gave written informed consent. Assessment of graft function included a clinical assessment of swallowing ability, radiologic studies, and motility recordings. CJinicaJ
Assessment
The patient’s ability to swallow liquids and solids was determined by interview. Any changes in the consistency of food chosen and an estimate of times required to complete meals were noted. A history of regurgitation was sought and changes in postoperative weight were recorded. Upper gastrointestinal endoscopy was performed if the patient complained of dysphagia or had a delay on barium studies.
After a 12-h fast, motility was recorded by a lowcompliance infusion system using 0.5..mlimin flow rates (5). Each subject had recordings from the graft and distal native esophagus. In this study a six-lumen polyvinyl tube incorporated perfusion sites at 5, 10, 15, 17.5, 20.0, and 22.5 cm from the distal end. The assembly initially was passed into the stomach and a slow pull-through in 0.5-cm increments was performed to measure the lower esophageal sphincter pressure. The upper sites were then positioned in the graft. Subjects lay in bed at an angle of 45” from the horizontal and received 7-m\ water boluses to swallow every 5-10 min during the 2.5-4-h study. Subjects who tolerated this study well were offered an additional motility study to record simultaneously contractile activity in the jejunal graft and the intact jejunum. In this study a seven-lumen tube (OD = 4.2 mm) incorporated perfusion sites at 0,20,40, 100,105,110 cm from the distal end. The final radiopaque catheter was used to inflate and deflate a balloon device, which allowed quick placement of the tube (6). During these recordings the perfusion system featured an electronic integrator (7) that recorded the motility index (area under the curve) each minute for the six recording sites. After 4 h of recording during fasting, subjects swallowed an additional fine tube through which 200 ml of Ensure-Plus (Ross Laboratories, Columbus, Ohio) was delivered directly into the stomach. This tube was then flushed clear with water and removed. Motor recording continued for 2 h after the test meal. Analysis
Each subject underwent (Figure 1) of deglutition and
shown
swallowed fluoroscopic graft function.
examination Both liquid
Data
Results are
Assessment
of Motility
The response to deglutition was assessed by comparing the timing and nature of motor responses in the graft with contractions in the adjacent esophagus. Recordings from the graft were examined for migrating motor complex (MMC)-like activity. This required continuous contractile activity for more than 3 min from at least one perfusion site located at the level of the graft. Data concerning the frequency of contractions during maximal activity and the interval between MMC-like activities were noted. In recordings from the jejunak graft and intact jejunum the temporal relationship between MMCs in the two locations was analyzed by visual inspection. After the test meal the motor patterns in the autotransplanted (extrinsically denervated) jejunum and the intact innervated (donor) jejunum were compared.
Results Radiologic
Assessment
of clinical in
Table
liquids
and
radiologic
1. Eleven and
solids
culty. Seven patients gained Several subjects described
assessments
of the with
12
patients
minimal
weight postoperatively. occasional episodes
diffiof
1958
GASTROENTEROLOGY Vol. 90, No. 6
KERLIN ET AL.
Table 1. Clinical and Radiologic Assessments Radiologic
Clinical
Patient
Months postoperative
Dysphagia with liquids
+ +++
+++
a
-
+
a
-
EC A.P. MS. A.G. D.W.
33 31 a a 20
10
P.L. E.H.
4 40
10
R.T. W.R. M.H.
Dysphagia with solids
Length of graft (cm1 9
a 9 7
5 6.5 14 a
Liquid “flush”0
t (liquid)
+
-
Regurgitation
+ t + + +
-
Delay in transit of Liquidb
Solid”
Location
+
t
Lower anastomosis Lower anastomosis
t (solid] -
t ++t
+ +++
t (liquid) -
-
+
-
-
+ (liquid] -
+ -
aperis-
Upper and lower anastomosis Lower anastomosis
-
a Patients use liquids to “flush” ingested solids to stomach. b Liquid barium. c Bread soaked in barium. Assessment: + +, intermittent; + + +, severe.
nasal regurgitation with liquids, and 6 patients described a need for liquids to “flush” ingested solids to the stomach. Most subjects ate more slowly than preoperatively, chewed more thoroughly, and favored soft food. Most patients reported a minor improvement in swallowing with time but could not discount the contribution of adaptation to the new conditions and life-style. The barium studies were particularly useful to assess the function of the graft. The liquid barium and bread soaked in barium passed rapidly to the stomach in 7 subjects. Two patients had a “hold-up” of the liquid and the solid boluses at the lower anastomosis. The subjects were unaware of this phenomenon. The obstruction was clearly “functional” inasmuch as the boluses passed either spontaneously, after a few seconds, or following a sip of water which formed a column of fluid above the apparent narrowing. The lumen at the anastomosis opened widely as the bolus passed. A further patient demonstrated a functional obstruction with bread only at the lower anastomosis. Endoscopy with a flexible fiberscope excluded strictures in each of these patients. Patient E.H., who was 3 yr and 4 mo postoperative and described mild dysphagia with solids, had a delay of barium entering and traversing the graft. Fiberoptic endoscopy disclosed the cause to be a fibrotic stricture of the graft. Mucosal biopsy specimens demonstrated radiation damage. After several dilatations with the Eder Peustow technique, the dysphagia was relieved. The patient has remained asymptomatic throughout a 12-mo observation period. Despite initial good function, patient D.W. developed anorexia and dysphagia and lost weight. Dysphagia was unrelated to the
Esophageal talsis
-, none; +, mild;
graft and was secondary to aperistalsis of the native esophagus. He died soon after investigation from complications of a primary lung cancer. Manometric
Studies
The mean lower esophageal sphincter pressure was 13.7 ? 5.8 mmHg. Three patients had a hypotensive lower esophageal sphincter based on our normal range for healthy controls (lo-25 mmHg). Results of recordings from the graft and distal native esophagus are shown in Table 2. As shown in Table 2 and Figure 2 (left panel), the most common response to a 7-ml bolus of water was no response in the graft and peristalsis in the adjacent esophagus. Similarly, most spontaneous “dry” swallows did not elicit contractions in the jejunal graft. The alternative response of the graft to deglutition is shown in the right panel of Figure 2. This burst of contractions in the graft was not always associated with peristalsis in the esophagus. This response was seen less frequently than no response and appeared to be in part artifactual due to forcing a bolus of air into the graft. During the recordings, regular contractile activity, typical of phase III of the MMC, was present in the jejunal graft of 11 patients. Multiple complexes were observed in the majority. No MMC was recorded in patient J.D. 2 ma after surgery. When the study was repeated at 12 mo after surgery, three MMCs were observed. Figure 3 illustrates the independence of contractile activity in the free jejunal graft and the native esophagus. Deglutition did not interrupt MMC-like activity. Between phase III-like activity
FUNCTION
June 1986
Table
2.
Results
of Monometric
OF A FREE JEJUNAL “CONDUIT”
GRAFT
Studies Pattern of motor response of graft to deglutition
(hl
(%I
(%I
4.0 2.5 3.3 3.5 4.0 4.0 4.0 3.7 3.30 3.0b
70 100
30
4 1
42-45 -
50 50 100 80 55 77
50 50 20 45 23
2 4 2 5 4 0
77 35-73 30 16-80 42-130 -
5-10 3 8 5-9 7 6-24 3.5-15 -
54 74
46 26
0 3
57, 60
4-7
10
63 68 75
37 32 25
3 1 3
63, 75 60,110
3-4 4 11-14
9 10 10
Figure
b Twelve
Duration of phase III (mm)
graft
EC. A.P. MS. A.G. D.W. P.L. E.H. D.P. J.D.
postoperative.
Period of cycle (min)
in jejunal
Patient
a Two months
Number of MMCs
activity
No activity
3.3 4.0 4.0
Multinle contractions
MMC-like
Duration of recording
R.T. W.R. M.H.
1959
-
Contractions during phase III (No./min) 10 10 9 9.3-10 10 8-9.5 9 -
months postoperative.
2. Response of jejunal interposition graft to swallowing. Three perfusion sites, G1--G,, are positioned in the graft and three, EI-E3, are positioned in the adjacent esophagus. The arrows indicate 7-ml boluses of water. The left panel represents the usual response-no response in the graft and peristalsis in the native esophagus. The right panel illustrates a less frequent response--a burst of contractions in the graft and peristalsis in the esophagus.
1960
KERLIN ET AL.
GASTROENTEROLOGY
Vol. 90, No. 6
01 I
1
%
Figure
3. Independence
of the jejunal graft and the esophagus.
Three perfusion
I
sites, G,-G,,
L
I
are positioned
1
6Omm Hg
I
1
I
_
in the graft and three, E,-E,,
are positioned in the adjacent esophagus. Regular contractile activity characteristic of phase III of the migrating motor complex is present in the graft in the left panel. The adjacent esophagus is quiescent. In contrast, the right panel from the same recording illustrates motor quiescence in the graft and peristalsis in the esophagus as the patient swallows.
fronts the grafts were usually quiescent, except for any contractions induced by deglutition or coughing via the tracheotomy. Six subjects had prolonged recordings from the jejunal graft and the intact jejunum (the donor site). At least one perfusion site remained within the graft
throughout the recordings (3-4.5 h, fasting). Migrating motor complex-like activity (2 to 4 cycles) was present during fasting in 5 patients [Figure 4). On visual analysis there was no temporal association between complexes present in the graft and those in the intact jejunum. The rate of contractions during
1
9Omm
h_-__
..
.
Hg
,
3-
Figure
4. Recording
of migrating motor complex from autotransplanted motor complexes at the two locations were asynchronous.
jejunal
graft (G1,Z) and intact jejunum
(J1-J3). The migrating
June 1986
FUNCTION
OF A FREE JEJUNAL
“CONDUIT”
GRAFT
1961
mm Hg
J3
1yA
52
1
1
t
I
Minutes Figure
5. Response of jejunal interposition graft (GI-G3) and intact jejunum (J,-J3) to food. A liquid antrum via a separate tube, thus bypassing the graft. A typical “fed” pattern of contractions in the autotransplanted (extrinsically denervated) graft.
aiways
MMC-like activity in the graft was slower (8.5-10.3/min) than the rate in the corresponding donor jejunum (10.5-1 Urnin). Recordings continued for 1.5-2.5 h after the liquid nutrient meal. In 5 subjects the Ensure-Plus was delivered directly into the stomach by a fine feeding tube which was then removed. This maneuver eliminated nutrient contact with the extrinsically denervated jejunal mucosa. The fifth subject drank the test meal. In each patient a typical “fed” pattern of intestinal motility was established in the intact jejunum. In contrast, the jejunal grafts remained relatively quiescent (excluding activity in the graft associated with deglutition) (Figure 5). In 2 patients MMC-like activity was observed in the graft during the 2 h after the test meal (Figure 6).
Discussion Reconstruction of the pharynx and cervical esophagus after ablative surgery is a major surgical challenge. Although the use of a free revascularized jejunal graft was pioneered by Seidenberg et al. (8) in 1959, more widespread application of the procedure
nutrient meal is delivered is present in the jejunum
into the but not
had to await the advances in microvascular surgical techniques of the last decade. The reconstruction offers a single-stage procedure with low morbidity and short hospitalization (1,3). The majority of our patients can swallow liquids and solids without significant difficulty. Nevertheless, 6 (50%) of the patients studied described a need to “flush” ingested solids with a sip of liquid. The symptom appears to represent a functional obstruction at the lower squamocolumnar anastomosis. During fluoroscopic examination the zone of “hold-up” was demonstrated to open in response to the hydrostatic pressure of a water bolus. The absence of contractions of the graft in response to deglutition explains the lack of adequate propulsive forces in the conduit. In theory, the function of longer jejunal segments could therefore be compromised by this “passive” nature of the graft. Further, longer segments could necessitate additional microvascular anastomoses. Indeed, McDonald et al. (9) have used a free jejunal graft to replace the total esophagus in a patient with a lye stricture. Two vessel groups were anastomosed to recipient vessels with apparently good function. The maximal length of jejunum used
1962
Figure
GASTROENTEROLOGY Vol. 90, No. 6
KERLIN ET AL.
6.
Plot of motility index each minute (electronic integrator] for perfusion site in jejunal graft (G) and intact jejunum (J1-J3) during fasting and after a test meal. The orrow and dotted line indicate delivery of Ensure-Plus directly into the gastric antrum via a separate tube. The migrating motor complexes at the two sites are asynchronous. In contrast to the continuous “fed” activity in the jejunum after the meal, migrating motor complexlike activity continues in the autotransplanted (extrinsically denervated) graft (*).
5
7
E”
is i=4 03 =2 , 0 7 6 5 4 3 2 1 0 0
05
1
1.5
2
2.5
3
TIME (h)
by our group has been a 24-cm graft following excision of an upper esophageal carcinoma. Detachment of the mesentery from each end of the conduit allowed it to be pulled out to its full length. During the motility recordings from the autotransplanted jejunal grafts, bursts of regular contractile activity, typical of phase III of the MMC, were observed. This observation helps to clarify the potential mechanisms for control of intestinal motility. Postulated control influences include extrinsic innervation, neuromuscular continuity of the bowel wall, and humoral factors. The MMC is the characteristic interdigestive motor pattern in the human small intestine (5,lO). During harvesting of the graft, the autonomic nerves that accompany the blood vessels are interrupted along with the division of the vascular arcade. Thus the presence of the MMC in the jejunal graft suggests that its presence in the intestine is independent of extrinsic innervation. The MMC is present also in the porcine and the canine models of autotransplantation (11,121. Results of these animal studies suggest that the initiation and propagation of the MMC within the small intestine resides in the intrinsic intramural neural plexi. A further important observation in our patients was the lack of a temporal association between
““A
4
45
5
5.5
I
MMCs in the autotransplanted jejunal graft and the donor jejunum. Sarr and Kelly (12) also noted complete disruption of the temporal association between motor patterns of the innervated bowel and of the isolated autotransplanted segment in their canine model. It appears that the orderly migration of the MMC from one segment of intestine to the next requires continuity to the intrinsic nerves. Indeed, the MMC will propagate across a site of transection and reanastomosis of the small bowel after regeneration of the intrinsic nerves (13). A third important physiologic observation became apparent in the feeding studies. In humans, food interrupts cyclic motor activity at all levels of the small intestine (5). This transition is prompt and the postprandial motility consists of irregular bursts of contractions interspersed with transient quiescence. However, although a typical “fed” motor pattern was established in the small bowel (donor site), the graft remained quiescent. In a previous study Meyers et al. (2) have reported that feeding induced continuous contractions in the graft. They considered that this response must be controlled by hormones. In contrast, feeding does not interrupt the MMC in the extrinsically denervated canine small bowel (12). The occurrence of MMCs in 2 patients and quiescence in the remainder
June 1986
of our patients in the fed state supports the latter view. Thus, our results and those in the canine model suggest that the extrinsic nerves are necessary for the postprandial inhibition of the MMC. In conclusion, a free jejunal graft provides a useful, though generally passive conduit. A zone of functional hold-up may occur at the lower anastomosis, and such patients may take liquid frequently during a meal to “flush” ingested solids to the stomach. The motility studies have also demonstrated that the autotransplanted jejunal graft maintains its intrinsic motor repertoire, which is asynchronous with that of the donor site. Finally, the findings also support the hypothesis that extrinsic nerves are required to induce a “fed” pattern of intestinal motility. References Robinson DW, MacLeod A. Microvascular free jejunum transfer. Br J Plast Surg 1982;35:258-67. Meyers WC, Seigler HF, Hanks JB, et al. Postoperative function of “free” jejunal transplants for replacement of the cervical esophagus. Ann Surg 1980;192:439-48. McCafferty GJ. Free jejunal graft reconstruction of the pharynx and cervical oesophagus. J Otolaryngol Sot Australia 1983/4;5(3):171-3.
FUNCTION
OF A FREE JEJUNAL
“CONDUlT”
GRAFT
1963
4. Hester TR, McConnel FMS, Nahai F, Jurkiewicz MJ, Brown RG. Reconstruction of cervical esophagus, hypopharynx and oral cavity using free jejunal transfer. Am J Surg 1980;140: 487-91. 5. Kerlin P, Phillips S. Variability of motility of the ileum and jejunum in healthy man. Gastroenterology 1982;82:694-700, R, Phillips SF. Rapid intubation of the 6. Kerlin P, Tucker ileo-colonic region of man. Aust NZ J Med 1983;13:591-3. 7. Kerlin P, Zinsmeister A, Phillips S. Motor responses to food of the ileum, proximal colon, and distal colon of healthy humans. Gastroenterology 1983;84:762--70. B, Rosenak SS, Hurwitt ES, Som ML. Immediate 8. Seidenberg reconstruction of the cervical esophagus by a revascularized isolated jejunal segment. Ann Surg 1959;149:162-71. 9. McDonald
JC, Rohr MS, Tucker
autotransplantation
of the
WY. Recent
kidney.
experiences
jejunum,
and
with
pancreas.
Ann Surg 1983;197:678-87. 10. Vantrappen
G, Janssens
digestive
motor
bacterial
overgrowth
J, Hellemans
complex
of normal of the
J. Ghoos subjects
small
Y. The inter-
and patients
intestine.
J Clin
with Invest
1977;.59:1158-66. 11. Aeberhard control
PF,
Am J Physiol 12. Sarr
Magnenat
of migratory
complex
WA.
Nervous
of the small
bowel.
1980;238:GlOZ-8.
MG, Kelly
planted
LD, Zimmermann
myoelectric
canine
KA. Myoelectric jejunoileum.
activity
of the
Gastroenterology
autotrans-
1981;81:303-
10. 13. Sarna
S, Stoddard
control 1981;241
C, Belbeck
of migrating (Gastrointest
L, McWade
myoelectric Liver Physiol
D. Intrinsic
complexes. 4) G16-23.
nervous
Am J Physiol