Keywords hypopharyngeal cancer - total pharyngolaryngectomy - postoperative complication -
stricture of the esophagojejunal anastomosis - background factor
Introduction
The morbidity and mortality rates for hypopharyngeal cancer have recently been on
the rise in Japan. Early-stage hypopharyngeal cancer is not associated with any major
subjective symptoms; however, it is prone to lymph node metastasis. Approximately
60% of the patients initially diagnosed with hypopharyngeal cancer are already at
an advanced stage.[1 ] Patients with advanced hypopharyngeal cancer require extended surgery and reconstructive
procedures. They often experience a significant decline in quality of life (QOL) because
of the adverse effects of surgery on various functions, including speech and swallowing.
In addition, many patients have underlying conditions, such as alcohol dependence,
hepatic dysfunction, renal dysfunction, diabetes mellitus, hypertension, and arteriosclerosis,
and these risk factors increase the incidence of postoperative complications.[2 ] Total pharyngolaryngoesophagectomy (TPLE) followed by free jejunal flap reconstruction
is currently the standard surgical treatment for advanced hypopharyngeal cancer. In
this study, we analyzed a set of advanced hypopharyngeal cancer cases that involved
TPLE, followed by free jejunal flap reconstruction in our department, with a focus
on postoperative complications and a literature review.
Participants and Methods
This study included 23 patients with advanced hypopharyngeal cancer who underwent
TPLE followed by free jejunal flap reconstruction in our department from April 2020
to January 2022. The Department of Head and Neck Surgery was responsible for resectioning
the primary lesion and dissecting the neck. The Department of Gastrointestinal Surgery
collected the jejunum, and the Department of Plastic and Reconstructive Surgery was
responsible for the reconstruction.
As potential background factors associated with postoperative complications, the following
data were retrospectively examined based on available medical records: age, sex, preoperative
performance status (PS), stage, nutritional status, past medical history (multiple
primary cancers, hypertension, and diabetes mellitus), preoperative chemotherapy/radiotherapy
information, preoperative blood tests (hemoglobin level, serum total protein level,
albumin level, and platelet count), surgery duration, and intraoperative blood loss.
Glasgow Prognostic Score (GPS) was used to evaluate nutritional status. The Japan
Clinical Oncology Group postoperative complications criteria (Clavien–Dindo classification)
were used to assess postoperative complications. Grade I and II complications were
considered mild, whereas grade III to V complications were considered moderate to
severe. Differences in the means and proportions were determined by univariate analysis
using the Mann–Whitney U and Fisher's exact tests, respectively, followed by a multivariate
analysis using logistic regression analysis. Differences with p < 0.05 and p < 0.1 were considered statistically significant and near-significant, respectively.
EZR was used for statistical analysis.[3 ]
Results
Of the enrolled participants, 21 were men and 2 were women. The ages ranged from 49
to 83 years (mean 72.3 years). The length of postoperative hospital stay and postoperative
follow-up ranged from 14 to 72 days (mean 26.0 d) and 4 to 138 weeks (mean 69.6 weeks),
respectively. The histopathological type was squamous cell carcinoma for all cases.
The lesions were identified at the following subsites: pyriform sinus in 12 cases
(52%), posterior wall in 8 cases (35%), and the postcricoid area in 3 cases (13%).
Concerning stage classification, there were 3, 3, and 17 cases of stage II, III, and
IV, respectively. Ten patients had multiple primary cancers, including three with
esophageal cancer, three with gastric cancer, one with colorectal cancer, four with
laryngeal cancer, and two with lung cancer. Seven patients had double primary cancer,
whereas three had triple primary cancer. The surgery and intraoperative blood loss
duration ranged from 412 to 587 minutes (mean 493.3 min) and 50 to 800 mL (mean 237.0 mL),
respectively. Neck dissection was performed in all cases, including 17 cases (74%)
in which bilateral neck dissection was performed.
[Fig. 1 ] shows the Kaplan–Meier curve for relapse-free survival. The relapse-free survival
rate at postoperative week 52 was 47.8%. Of the 23 patients, 3 died during the follow-up
period, 2 died from distant metastasis, and 1 died from suffocation because of intratracheal
bleeding during postoperative chemotherapy following mediastinal lymph node metastasis.
Fig. 1 Relapse-free survival curve. The relapse-free survival rate at postoperative week
52 was 47.8%.
Postoperative complications occurred in 12 cases (52%), including mild and moderate-to-severe
complications in five and seven cases, respectively. Mild complications included esophagojejunal
anastomotic stricture in three cases, stenosis of the tracheostoma in one case, and
pneumonia in one case. Moderate-to-severe complications included esophagojejunal anastomotic
stricture in six cases (plus chylorrhea in one case) and stenosis of the tracheostoma
in one case. Free jejunal flap necrosis was not observed in any patient. Of the cases
in which esophagojejunal anastomotic stricture occurred, endoscopic dilation was performed
in five cases, and two patients were incapable of oral ingestion because of esophageal
anastomotic stricture. One of these patients also had stricture resulting from esophageal
cancer, and reestablishment of oral ingestion failed even after radiotherapy. The
other patient also could not reestablish oral ingestion through swallowing rehabilitation,
whereas other interventions were unsuccessful because of marked cognitive impairment
and treatment refusal.
The differences in background factors between patients with and without complications
and the results of univariate analysis are summarized in [Table 1 ]. There tended to be more stage IV cases in the complication-positive group (n = 12) than in the group without (n = 11; 11/12 [92%] vs. 6/11 [55%]; p = 0.069). The platelet count also tended to be lower (22.6 ± 2.2 × 104 /μL vs. 27.6 ± 2.2 × 104 /μL; p = 0.065), and intraoperative blood loss tended to be higher (304.2 ± 63.0 vs. 163.6 ± 15.9 mL;
p = 0.056) than that of the complication-negative group. [Table 2 ] summarizes the differences in background factors between patients with and without
moderate-to-severe complications and the results of univariate analysis. In the moderate-to-severe
complication-positive group (n = 7), hemoglobin levels were significantly higher (14.1 ± 0.3 vs. 12.7 ± 0.5 g/dL;
p = 0.049) and the platelet count was significantly lower (18.8 ± 1.9 × 104 /μL vs. 27.7 ± 1.8 × 104 /μL; p = 0.008) than those in the moderate-to-severe complication-negative group (n = 16). Moreover, intraoperative blood loss tended to be higher (302.1 ± 66.8 vs.
208.4 ± 42.5 mL; p = 0.095). Multivariate analysis revealed a significant difference in platelet count
(p = 0.045). There were no differences between the complication-positive and -negative
groups or between the moderate-to-severe complication-positive and -negative groups
concerning age, sex, PS, proportion of patients with cervical lymph node metastasis,
GPS, proportion of patients with multiple primary cancers, proportion of patients
with hypertension, proportion of patients with diabetes mellitus, history of chemotherapy,
history of radiotherapy, total serum protein levels, and duration of surgery.
Table 1
All complications and background factors
Complication-positive (n = 12)
Complication-negative (n = 11)
p -Value
Age (years)
74.5 ± 2.2
69.9 ± 3.0
0.241
Sex (male/female)
11/1
10/1
1.000
PS ≥1
6/12 (50%)
2/11 (18%)
0.193
Patients with positive lymph nodes
11/12 (92%)
7/11 (64%)
0.155
Stage IV
11/12 (92%)
6/11 (55%)
0.069[a ]
GPS ≥1
4/12 (33%)
4/11 (36%)
1.000
History of multiple primary cancers
5/12 (42%)
5/11 (45%)
1.000
History of Hypertension
4/12 (33%)
6/11 (55%)
0.414
History of diabetes mellitus
3/12 (25%)
2/11 (18%)
1.000
History of chemotherapy
2/12 (17%)
2/11 (18%)
1.000
History of radiotherapy
2/12 (17%)
3/11 (27%)
0.640
Albumin level (g/dL)
3.84 ± 0.4
3.68 ± 0.5
0.686
Hemoglobin level (g/dL)
13.4 ± 0.5
12.8 ± 0.5
0.310
Serum total protein (g/dL)
5.3 ± 0.4
5.4 ± 0.3
0.926
Platelet count (×104 /μL)
22.6 ± 2.2
27.6 ± 2.2
0.065[a ]
Surgery duration (min)
506.5 ± 14.4
479.0 ± 17.0
0.176
Intraoperative blood loss (mL)
304.2 ± 63.0
163.6 ± 15.9
0.056[a ]
Abbreviations: GPS, Glasgow Prognostic Score; PS, preoperative performance status.
a
p < 0.1.
Table 2
Moderate-to-severe complications and background factors
Moderate-to-severe complication-positive (n = 7)
Moderate-to-severe complication-negative (n = 16)
p -Value
Age
75.9 ± 3.1
70.8 ± 2.2
0.132
Sex (male/female)
7/0
14/2
1.000
PS ≥1
2/7 (29%)
6/16 (38%)
1.000
Patients with positive lymph nodes
7/7 (100%)
11/16 (69%)
0.272
Stage IV
6/7 (86%)
11/16 (69%)
0.621
GPS ≥1
2/7 (29%)
6/16 (38%)
1.000
History of multiple primary cancers
3/7 (43%)
7/16 (44%)
1.000
History of Hypertension
2/7 (29%)
8/16 (50%)
0.405
History of diabetes mellitus
1/7 (14%)
4/16 (25%)
1.000
History of chemotherapy
1/7 (14%)
3/16 (19%)
1.000
History of radiotherapy
1/7 (14%)
4/16 (25%)
1.000
Albumin level (g/dL)
3.8 ± 0.4
3.7 ± 0.5
0.710
Hemoglobin level (g/dL)
14.1 ± 0.3
12.7 ± 0.5
0.049[a ]
Serum total protein (g/dL)
5.1 ± 0.5
5.5 ± 0.3
0.504
Platelet count (×104 /μL)
18.8 ± 1.9
27.7 ± 1.8
0.008[a ]
Surgery duration (min)
499.6 ± 23.3
490.6 ± 13.0
0.738
Intraoperative blood loss (mL)
302.1 ± 66.8
208.4 ± 42.5
0.095[b ]
Abbreviations: GPS, Glasgow Prognostic Score; PS, preoperative performance status.
a
p < 0.05.
b
p < 0.1.
Discussion
Head and neck cancer cases account for approximately 5% of all cancer cases. A national
survey of 205 medical institutions conducted by the Japan Society for Head and Neck
Cancer in 2019 revealed that out of 13,658 patients with head and neck cancer, 2,858
(20.9%) were patients with hypopharyngeal cancers, which was second only to oral cancer
(29.2%) in number. Patients with hypopharyngeal cancers have a peak age of onset in
their sixties to seventies, and 91% are men.[5 ] Consistent with this report, the vast majority of the patients included in the present
study were men (21/23, 91%) and the mean age was 72.3 years.
For safety and restoring oral ingestion early, the free jejunal flap is the first-choice
material for reconstructive surgery after TPLE for advanced hypopharyngeal cancer.[4 ]
[5 ]
[6 ] At our hospital, the free jejunal flap was selected as a material for reconstruction
following TPLE in all cases. With respect to intraoperative factors (surgery duration
and blood loss) and the length of postoperative hospital stay, our results were equivalent
or superior to those in other institutions.[4 ]
[7 ]
Typical complications after TPLE and free jejunal flap reconstruction included esophagojejunal
anastomotic stricture, necrosis of the grafted flap resulting from clot formation
in the anastomosed blood vessels, anastomotic suture failure, bowel obstruction, and
stenosis of the tracheostoma.[8 ] In the present study, esophagojejunal anastomotic stricture was the most frequent
complication. The univariate and multivariate analyses of the background factors revealed
risk factors for these postoperative complications. In patients with moderate-to-severe
complications, the univariate analysis showed a significantly lower platelet count
and a near significantly higher hemoglobin level, whereas the multivariate analysis
showed a significantly lower platelet count.
Risk factors previously associated with complications after surgery for advanced head
and neck cancer, as well as free flap reconstruction, included duration of surgery,
excessive fluid infusion, surgeon skill level, history of chemoradiotherapy, and nutritional
status.[7 ]
[8 ]
[9 ]
[10 ] In the present study, low platelet count was a significant background factor for
moderate-to-severe complications ([Table 2 ]). An animal study in esophagectomized rats showed that ischemia at the anastomotic
site is involved in anastomotic stricture,[11 ] which indicates the possibility that low platelet counts contributed to anastomotic
stricture through increased intraoperative blood loss and caused local ischemia. The
intraoperative blood loss tended to be larger among complication-positive patients
([Tables 1 ] and [2 ]).
Identifying high hemoglobin levels as a background complication factor is paradoxical,
and the reason remains unclear. Diluting blood without changing blood volume improved
microcirculation in an animal study, as blood viscosity decreases upon dilution.[12 ] Hemodilution has also been applied in the surgery of head and neck tumors.[13 ] For the cases in the present study, this mechanism may act to decrease complications.
Identification of stage IV disease as a near-significant background factor for complications
is understandable from the standpoint that the invasiveness of surgery is higher than
that for stage III and less advanced diseases. No previous reports have documented
the associations between post-TPLE complications and stage; however, an association
between the total number of lymph nodes removed intraoperatively and postoperative
anastomotic stricture was reported in patients who underwent esophagectomy for esophageal
cancer.[14 ]
The most frequent complication in our cohort was esophagojejunal anastomotic stricture,
which occurred in nine cases (39%), including all severity grades. This incidence
rate is similar to that reported previously.[15 ] Anastomotic stricture is a complication that has major effects on postoperative
treatment. In our department, endoscopic examinations are proactively performed when
symptoms indicative of anastomotic stricture, such as reflux of food and drink, are
evident for early detection and treatment. Endoscopic dilation is the treatment of
choice for anastomotic stricture; however, dilation may be difficult to achieve in
patients with low preoperative PS, decreased cognitive function, or no motivation
to receive treatment. Thus, it is important to check for these background factors
in advance to ensure that patients can complete the treatment. QOL will inevitably
be disrupted if dysphagia cannot be improved because the patient must depend on gastrostomy
or total parenteral nutrition.
Mori et al stated that postoperative radiotherapy is associated with anastomotic stricture
and that frequent use of bougies during and after irradiation is important to prevent
stricture.[16 ] Moreover, mechanical anastomosis is more prone to postoperative stricture than manual
anastomosis[17 ]; however, we used manual anastomosis in all cases, and the incidence of anastomotic
stricture was comparable with previously reported rates. Thus, the anastomotic technique
is unlikely to be a significant risk factor.
Necrosis of the free jejunal graft is the most serious complication of TPLE/free jejunal
reconstruction. The incidence rates of free jejunal flap necrosis were approximately
1 to 5%,[8 ]
[14 ]
[17 ] and free jejunal flap reconstruction is a reconstructive procedure that results
in the highest graft survival rates in the head and neck region. In the present study,
we showed that jejunal flaps survived in all of the 23 cases, which resulted in a
satisfactory surgical outcome. Our department manages patients who have undergone
this procedure in the intensive care unit. Blood flow is checked in the anastomotic
vessels with an ultrasonic Doppler flowmeter, and the graft color is monitored every
hour for early detection of jejunal flap necrosis. As a graft monitoring method, recent
studies have measured local oxygen saturation using near-infrared devices, such as
pulse oximeters.[18 ]
[19 ] This technique may develop into a useful method for the early detection of flap
necrosis.
Conclusion
We analyzed background factors for postoperative complications in 23 patients who
underwent TPLE and free jejunal flap reconstruction for advanced hypopharyngeal cancer.
Postoperative complications occurred in 12 cases (52%), and esophagojejunal anastomotic
stricture, the most frequent complication, was observed in 9 cases (39%). Free jejunal
flap necrosis did not occur in any of the cases. A low platelet count was identified
as a significant background factor for moderate-to-severe complications. Moreover,
the percentage of stage IV cases and the amount of intraoperative blood loss tended
to be higher in patients who experienced complications. Our results suggest that monitoring
these background factors in advance for early detection of complications is important
for completing treatment in patients with advanced hypopharyngeal cancer.
