Keywords
postoperative nausea and vomiting - colorectal cancer
Introduction
Postoperative nausea and vomiting (PONV) is a common adverse effect of anesthesia
and surgery and is also frequently seen after colorectal surgery in between 8 and
44% of the patients.[1]
[2]
[3]
[4]
[5]
[6] Postoperative nausea and vomiting is extremely distressing, as it exacerbates conditions
such as general fatigue and/or postoperative wound pain in the already stressful early
postoperative period. These outcomes were also reported to lead to a prolonged hospital
stay and increased medical costs due to continuing symptoms and additional medication.[5]
[7]
[8]
[9]
Many types of surgical procedures performed under general anesthesia induce PONV.
In particular, colorectal cancer surgery involves the gastrointestinal tract and induces
not only PONV but also gastrointestinal-related complications, such as ileus, bowel
obstruction, and anastomotic leakage.[10]
[11] Gastrointestinal symptoms following colorectal cancer surgery tend to be closely
related to PONV, and the incidence and etiology of PONV in these cases may need to
be considered separately from that in nongastrointestinal surgeries. However, while
several risk factors for PONV following general surgery have been identified, few
studies have discussed the risk factors for PONV associated with colorectal surgery.[5] A risk analysis in patients with colorectal cancer may help to determine the specific
risk factors for PONV following colorectal cancer surgery.
Therefore, the present study investigated the risk factors for PONV in cases of colorectal
cancer surgery.
Materials and Methods
Patients
A retrospective study of 204 consecutive patients who underwent colorectal surgery
in a single institution at the Department of Surgery, Saga Medical Center Koseikan,
between January 2018 and March 2020, was conducted. The medical records of all patients
were reviewed in detail. The inclusion criterion was: primary tumor resection performed
with open or laparoscopic surgery as elective surgery. The exclusion criteria were:
nonresection of the primary tumor, emergency operation, and bowel obstruction at the
time of operation. For the present study, the 204 total patients were divided into
2 groups based on the presence of postoperative nausea (PON) or postoperative vomiting
(POV).
All patients and their families were informed about the surgical procedure and provided
their written consent. Broad consent was obtained for the present study. The medical
ethics committee of the Saga Medical Center Koseikan reviewed and approved the present
study design (permission number: 21-03-01-05).
Approach
Colorectal cancer was confirmed preoperatively by colonoscopy and a pathological examination.
The characteristics of the patients and the preoperative, operative, and postoperative
parameters were analyzed. Data on the following variables were obtained: gender, age,
body mass index (BMI), American Society of Anesthesiologists (ASA) physical status,
smoking and drinking alcohol habit, medical history, bowel obstruction at the initial
diagnosis, preoperative therapy, number of times the patient had undergone fasting
preoperatively, blood test results, tumor location (right-sided: ascending to transverse
colon, or left-sided: descending to the rectosigmoid colon and the rectum), TNM stage,
tumor size, operative procedure, operative time, intraoperative bleeding, blood transfusion,
additional operative procedure (resection of an additional organ due to another disease
during the same operation), stoma creation, kind of anesthesia and analgesia, intraoperative
fluid volume, urine volume during the operation, admission to the intensive care unit,
starting day of drinking or eating solid food, day of first defecation, postoperative
complications, and duration of postoperative hospital stay. Univariate and multivariate
analyses were performed to determine the clinicopathological factors associated with
PONV. Postoperative nausea and vomiting was defined as any nausea and/or vomiting
occurred by postoperative day 2 in the present study.
Surgical Procedure and Patient Management
All patients underwent open or laparoscopic surgery with adequate lymphadenectomy.
The decision to perform postoperative continuous epidural analgesia, continuous intravenous
analgesia or no continuous analgesia was determined based on each the condition of
each patient. The first choice for continuous analgesia was set as epidural analgesia.
However, if patients were being medicated with antiplatelet agents for heart and/or
brain disease, continuous intravenous analgesia or no continuous analgesia at all
was prescribed. The epidural catheter was inserted at Th10-11, Th11-12, or Th12-L1
at the time of surgery. The discontinuation of continuous analgesia was determined
by the doctor in charge, depending on the condition of the patient, such as whether
or not they would be able to get out of bed and how much pain control they required
after the operation. The epidural analgesia solution was composed of between 500 and
1000 μg fentanyl and 500 mg levobupivacaine with or without 5 mg droperidol mixed
with normal saline to a total volume of 300 mL. The bolus dose was 3 ml at a basal
infusion rate of between 2 and 4 ml/h and a lockout interval of 30 minutes. The intravenous
analgesia solution was composed of 1000 μg fentanyl with or without 2.5 mg droperidol
mixed with normal saline brought to a total volume of 50 ml. The bolus dose was 1 mL
at a basal infusion rate of 1 ml/h, with a lockout interval of 10 minutes. The gastric
tube was inserted at the introduction of the anesthesia and was removed in the operation
room immediately after the operation. The pathological tumor stage was classified
according to the seventh edition of the UICC-TNM classification. Prophylactic antiemetic
medication, such as aprepitant, perphenazine, metoclopramide, dexamethasone, ondansetron,
and ramosetron, was not used in any case.
Statistical Analyses
Continuous variables were expressed as the median and interquartile range (IQR), and
categorical variables were expressed as numbers. In the univariate analysis, group
comparisons used the Wilcoxon rank sum test for continuous variables and the Fisher
exact test for binary variables. A multivariate analysis using stepwise multiple logistic
regression was performed using the variables found to be significant (p < 0.1) by a univariate analysis. The data were expressed as odds ratios (ORs) with
95% confidence intervals (CIs). A p-value < 0.05 was considered statistically significant. All analyses were conducted using
IBM SPSS Statistics for Windows, version 25 (IBM Corp., Armonk, NY, USA).
Results
Among the 204 patients (116 male, 88 female) with colorectal cancer, 54 (26.5%) and
25 (12.3%) patients experienced PON and POV, respectively. No perioperative mortality
was encountered.
[Table 1] shows the results of the univariate analysis of the factors associated with PONV.
The univariate analysis showed that female gender (p < 0.001), no current alcohol drinking habit (p = 0.003), and no stoma creation (p = 0.023) were associated with PON. Postoperative vomiting was significantly correlated
with female gender (p = 0.009), a high BMI (p = 0.017), and right-sided colon cancer (p = 0.001). Regarding patient factors, the medical history of the patient, bowel obstruction
at the initial diagnosis, and number of times the patient had undergone fasting preoperatively
were not associated with PONV. Regarding the operative and postoperative factors,
anesthesia factors showed no significant association with PONV. The multivariate logistic
regression analysis revealed that female gender (OR: 4.225; 95%CI: 2.170–8.226; p < 0.001) was an independent risk factor for PON. Body mass index (OR: 1.148; 95%CI:
1.018–1.295; p = 0.025), and right-sided colon cancer (OR: 3.337; 95%CI: 1.287–8.652; p = 0.013) were independent risk factors for POV ([Table 2]).
Table 1
Univariate analysis of the patients
|
Nausea
|
Vomit
|
|
(-) n = 150
|
(+) n = 54
|
p-value
|
(-) n = 179
|
(+) n = 25
|
p-value
|
|
Patient factors
|
|
Gender (male: female)
|
99:51
|
17:37
|
< 0.001
|
108:71
|
8:17
|
0.009
|
|
Age (years old; median [IQR])
|
71 [63–78]
|
73 [66–81]
|
0.157
|
71 [65–78]
|
76 [68–82]
|
0.075
|
|
Body mass index (kg/m2; median [IQR])
|
22.5 [20.2–25.0]
|
22.7 [19.0–26.6]
|
0.241
|
22.3 [19.8–25.1]
|
24.4 [20.0–27.2]
|
0.017
|
|
ASA-PS (PS0,1: PS2,3)
|
33:117
|
12:42
|
1.000
|
40:139
|
5:20
|
1.000
|
|
Currently smoking (yes:no)
|
43:107
|
9:45
|
0.102
|
48:131
|
4:21
|
0.330
|
|
Currently drinking alcohol (yes:no)
|
86:64
|
18:36
|
0.003
|
95:84
|
9:16
|
0.136
|
|
Patient history
|
|
Cardiac disease (yes:no)
|
16:134
|
10:44
|
0.156
|
22:157
|
4:21
|
0.534
|
|
Pulmonary disease (yes:no)
|
3:147
|
2:52
|
0.610
|
4:175
|
1:24
|
0.483
|
|
Cerebrovascular disease (yes:no)
|
12:138
|
3:51
|
0.763
|
13:166
|
2:23
|
1.000
|
|
Hypertension (yes:no)
|
65:85
|
23:31
|
1.000
|
75:104
|
13:12
|
0.391
|
|
Diabetes mellitus (yes:no)
|
33:117
|
13:41
|
0.850
|
40:139
|
6:19
|
0.803
|
|
Dementia (yes:no)
|
4:146
|
3:51
|
0.384
|
5:174
|
2:23
|
0.206
|
|
Abdominal surgery (yes:no)
|
57:93
|
22:32
|
0.746
|
70:109
|
9:16
|
0.830
|
|
Bowel obstruction at the initial diagnosis (yes:no)
|
14:136
|
2:52
|
0.246
|
16:163
|
0:25
|
0.228
|
|
Preoperative therapy (yes:no)
|
9:141
|
0:54
|
0.116
|
9:170
|
0:25
|
0.604
|
|
Number of times the patient had undergone
fasting preoperatively (times; median [IQR])
|
5[5-6]
|
5[5-7]
|
0.707
|
5[5-6]
|
5[5-7]
|
0.365
|
|
Blood test results
|
|
Sodium (mmol/L; median [IQR])
|
141 [139–142]
|
141 [139–142]
|
0.554
|
141 [139–142]
|
141 [139–143]
|
0.445
|
|
Potassium (mmol/L; median [IQR])
|
4.2 [3.9–4.4]
|
4.1 [3.9–4.4]
|
0.293
|
4.2 [3.9–4.4]
|
4.1 [3.8–4.4]
|
0.198
|
|
Hemoglobin (g/dl; median [IQR])
|
12.8 [11.1–14.2]
|
13.0 [11.2–14.1]
|
0.574
|
12.8 [11.3–14.1]
|
13.2 [10.7–14.9]
|
0.534
|
|
Albumin (g/dl; median [IQR])
|
4.1 [3.6–4.3]
|
4.1 [3.8–4.3]
|
0.739
|
4.1 [3.6–4.3]
|
4.1 [3.8–4.4]
|
0.682
|
|
Tumor factor
|
|
Tumor location (right:left)
|
50:100
|
25:29
|
0.101
|
58:121
|
17:8
|
0.001
|
|
T category (- T2: T3, T4)
|
47:103
|
20:34
|
0.500
|
57:122
|
10:15
|
0.496
|
|
N category (negative:positive)
|
85:65
|
34:20
|
0.520
|
103:76
|
16:9
|
0.666
|
|
M category (negative:positive)
|
135:15
|
50:4
|
0.786
|
162:17
|
23:2
|
1.000
|
|
Maximum tumor length (mm; median [IQR])
|
40 [25–55]
|
44 [26–53]
|
0.840
|
40 [28–55]
|
38 [16–56]
|
0.437
|
|
Operative factors
|
|
Operative procedure (laparoscopic:open)
|
130:20
|
47:7
|
1.000
|
156:23
|
21:4
|
0.752
|
|
Operative time (minutes; median [IQR])
|
245 [204–303]
|
258 [222–306]
|
0.788
|
251 [205–303]
|
261 [207–309]
|
0.483
|
|
Intraoperative bleeding (ml; median [IQR])
|
27 [10–65]
|
20 [7–61]
|
0.404
|
25 [8–62]
|
20 [10–106]
|
0.762
|
|
Blood transfusion (yes:no)
|
18:132
|
8:46
|
0.636
|
20:159
|
6:19
|
0.102
|
|
Additional operative procedure (yes:no)
|
10: 140
|
5: 49
|
0.549
|
12:167
|
3:22
|
0.403
|
|
Stoma creation (yes:no)
|
14:136
|
0:54
|
0.023
|
14:165
|
0:25
|
0.225
|
|
Anesthesia factors
|
|
Sevoflurane (yes:no)
|
48:102
|
13:41
|
0.303
|
54:125
|
7:18
|
1.000
|
|
Desflurane (yes:no)
|
100:50
|
40:14
|
0.393
|
122:57
|
18:7
|
0.820
|
|
Continuous intravenous anesthesia (yes:no)
|
147:3
|
53:1
|
1.000
|
175:4
|
25:0
|
1.000
|
|
Maximum intraoperative temperature (°C: median [IQR])
|
37.0 [36.4–37.4]
|
37.0 [36.6–37.2]
|
0.759
|
37.0 [36.4–37.4]
|
37.0 [36.8–37.2]
|
0.963
|
|
Infusion volume (ml:median [IQR])
|
2,510 [2,100–3,000]
|
2,690 [2,203–3,350]
|
0.506
|
2,550 [2,115–3,025]
|
2,540 [2,175–3,813]
|
0.793
|
|
Total dose of fentanyl/body weight
(mg/kg:median [IQR])
|
4.82 [3.56–6.46]
|
4.40 [3.62–6.17]
|
0.447
|
4.86 [3.59–6.54]
|
4.13 [3.38–5.43]
|
0.082
|
|
Urine/body weight/ operation time
(ml/kg/h:median [IQR])
|
2.86 [1.62–5.40]
|
3.48 [1.93–5.66]
|
0.222
|
3.00 [1.65–5.53]
|
3.63 [1.97–4.76]
|
0.668
|
|
Colloidal solution (ml:median [IQR])
|
500 [400–1000]
|
500 [400–1000]
|
0.762
|
500 [438–1,000]
|
500 [88–1,350]
|
0.833
|
|
Postoperative factor
|
|
ICU admission (yes:no)
|
2:148
|
2:52
|
0.286
|
3:176
|
1:24
|
0.410
|
|
Continuous epidural analgesia (yes:no)
|
143:7
|
53:1
|
0.684
|
172:7
|
24:1
|
1.000
|
|
Continuous intravenous analgesia (yes:no)
|
6:144
|
0:54
|
0.344
|
6:173
|
0:25
|
1.000
|
|
Continuous usage of fentanyl as analgesia
(yes:no)
|
148:2
|
53:1
|
1.000
|
177:2
|
24:1
|
0.326
|
|
Continuous usage of droperidol as analgesia
(yes:no)
|
88:62
|
27:27
|
0.337
|
103:76
|
12:13
|
0.395
|
Abbreviations: ASA-PS, American Society of Anesthesiologist physical status; ICU,
intensive care unit; IQR, interquartile range.
Numbers in bold represent statistical significance.
Table 2
Results of the multivariate analysis
|
Multivariate analysis
|
|
Outcome
|
Parameter
|
OR
|
(95% CI)
|
p-value
|
|
Nausea
|
Female gender
|
4.225
|
(2.170–8.226)
|
< 0.001
|
|
Vomit
|
Female gender
|
2.540
|
(0.973–6.632)
|
0.057
|
|
Body mass index
|
1.148
|
(1.018–1.295)
|
0.025
|
|
Right-sided colorectal cancer
|
3.337
|
(1.287–8.652)
|
0.013
|
Abbreviations: CI, confidence interval; OR, odds ratio.
Numbers in bold represent statistical significance.
The postoperative course and complications were also examined. The starting day of
drinking was significantly later in the group with POV than in the group without POV
(p = 0.006). The day of first defecation and starting day of eating solid food were
not correlated with the presence of either PON or POV. The postoperative stay was
significantly shorter in the group with PON than in the group with POV (p = 0.028) ([Table 3]).
Table 3
Postoperative course and complications
|
Nausea
|
Vomit
|
|
(-) n = 150
|
(+) n = 54
|
p-value
|
(-) n = 179
|
(+) n = 25
|
p-value
|
|
Starting day of drinking (POD0-1: ≥2)
|
149:1
|
53:1
|
0.058
|
178:1
|
22:3
|
0.006
|
|
Day of first defecation (POD0-4: ≥POD5)
|
146:4
|
49:5
|
0.057
|
173:6
|
22:3
|
0.083
|
|
Start day of eating solid food (POD; median [IQR])
|
3[3-4]
|
3[3-5]
|
0.329
|
3[3-5]
|
4[3-5]
|
0.249
|
|
Postoperative complication CD≥3 (yes:no)
|
9:141
|
3:51
|
1.000
|
11:168
|
1:24
|
1.000
|
|
Postoperative ileus CD≥3 (yes:no)
|
2:148
|
0:54
|
1.000
|
2:177
|
0:25
|
1.000
|
|
Postoperative stay (days; median [IQR])
|
10 [9–12]
|
9[8-11]
|
0.028
|
10[9-12]
|
9[8-11]
|
0.131
|
Abbreviations: CD, Clavien-Dindo Number in bold represents statistical significance;
IQR, interquartile range; POD, postoperative date.
Discussion
Postoperative nausea and vomiting, once it occurs, is not only uncomfortable for the
patient but also causes problems that prevent the patient from leaving the bed early
and resuming eating. In addition, vomiting may carry a risk of aspiration pneumonia
and dehydration. In the present study, PON and POV occurred in 26.5% (54/204) and
12.3% (25/204) of the patients, respectively. These incidence rates seemed to be comparable
to those observed in a previous report (between 8 and 44%).[1]
[2]
[3]
[4]
[5]
[6] However, further improvements are needed to achieve better surgical results and
an improved quality of life for the patients. Although the issue of PONV is often
discussed in anesthesiology, it is not still widely recognized by many surgeons, who
do not seem to be aware of it.
The risk factors of PONV seemed to vary widely based on the patient-related, anesthesia-related,
and intraoperative and postoperative factors.[8] Postoperative nausea and vomiting can be triggered by several perioperative stimuli,
including opioids, volatile anesthetics, anxiety, adverse drug reactions, and motion.[8] Thus far, numerous evidenced risk factors of PONV have been shown in the guidelines
for the management of PONV, such as female gender, history of PONV or motion sickness,
nonsmoking, younger age, general versus regional anesthesia, use of volatile anesthetics
and nitrous oxide, postoperative opioids, duration of anesthesia, and type of surgery
(cholecystectomy, laparoscopic, gynecological).[12] However, these risk factors may vary depending on the type of surgery and on the
background of the patient concerning the specific disease, like colorectal cancer.
In addition, there may be other unique risk factors depending on the disease and type
of surgery. Therefore, continual improvements and countermeasures against the risk
factors of PONV may help reduce its incidence.
Our study demonstrated that female gender was an independent risk factor for PON in
a multivariate analysis. In addition, female gender was a significant factor for POV
in a univariate analysis and was close to significance in a multivariate analysis.
Female gender was the strongest overall predictor and risk factor for PONV.[8]
[12]
[13] The mechanism underlying the relationship between female gender and an increased
incidence of PONV is unknown at present.[13] However, it was confirmed that female gender was a risk factor for PONV even in
the field of colorectal surgery. In general, younger age is also a strong risk factor
for PONV.[8]
[12] Our study showed no correlation between age and the occurrence of PONV. This may
be because the patients presenting with colorectal cancer often tend to be older than
those presenting with otolaryngological or orthopedic diseases.
Our study demonstrated that a high BMI was an independent risk factor for POV. Although
some previous reports have suggested that the incidence of PONV is increased in obese
patients,[14]
[15] a recent systematic review did not find a relationship between BMI and the incidence
of PONV. Intra-abdominal pressure in obese patients could be more dependant on a direct
mass effect from the intra-abdominal adipose tissue than non-obese patients.[16] Regarding the postoperative period of the gastrointestinal tract, we speculated
that increased visceral fat could lead to PONV from increased abdominal pressure and
gastroesophageal reflux. Body mass index as a risk factor for PONV is also still controversial,
so further studies are needed.
In our study, right-sided colon cancer was an independent risk factor for POV. This
is a novel finding in the present study, and no previous report found a similar result.
We speculated that it was because in cases of right-sided colon cancer, including
the cecum of the transverse colon, the mesentery of the right-sided colon is isolated
and dissected from the ventral side of the duodenum and/or from the pancreatic head.
In addition, there are more opportunities to handle the stomach and the duodenum during
dissection of the mesentery of the ascending and transverse colon, which may cause
temporary upper gastrointestinal paralysis due to the mechanical stimulation, resulting
in POV. Cienfuegos et al. and Masoomi et al. previously reported the clinicopathological
differences between right- and left-sided colon cancer and found that postoperative
ileus was significantly more frequent in cases of cancer in the right colon than in
those in the left colon.[10]
[11] Therefore, it is possible that the occurrence and mechanism of PONV and postoperative
ileus may be closely related. In our study, the occurrence of POV and PON did not
correlate with the incidence of postoperative ileus. This phenomenon may be because
we only had 2 cases (1%) of postoperative ileus. Therefore, it would be interesting
to examine a larger number of patients in the future.
For the prevention of PONV, several prophylactic antiemetic medications, such as metoclopramide,
aprepitant, ramosetron, granisetron, dexamethasone, droperidol, fosaprepitant, and
ondansetron have been reported.[4]
[17]
[18]
[19]
[20] In our study, droperidol was used for either epidural analgesia or continuous intravenous
analgesia since we expect to obtain an antiemetic effect. Continuous usage of droperidol
as analgesia was not correlated with the incidence of PONV in the univariate analysis.
This result might have been because the dosage of droperidol was small, so continuous
usage was not associated with the occurrence of PONV. Prophylactic antiemetic medications
have a large impact on patient care in high-risk populations. However, in a general
surgical population at low to moderate risk of PONV, most patients will not benefit
from routinely administered prophylactic antiemetic medications, since ∼ 70% of the
patients do not suffer from PONV.[20]
[21] In addition, multimodal prophylaxis therapy is recommended for patients with medium
or high risk.[12] For postoperative procedures involving the gastrointestinal tract, such as colorectal
cancer surgery, prophylactic administration of antiemetics may be considered, especially
for high-risk cases. Therefore, it is important to identify patients with high-risk
factors and to consider administering prophylaxis to these patients.
The postoperative course, such as the day on which patients started drinking and eating
solid food, the day of first defecation, and the duration of postoperative stay, were
also analyzed in the present study. Among them, only the starting day of drinking
was significantly later in the group with POV than in the group without POV. This
should be understandable, as patients who have vomited may have difficulty starting
to drink. However, interestingly, PONV did not affect the day of first defecation,
the starting day of eating solid food, or the occurrence of postoperative ileus (CD≥3).
Regarding the postoperative stay, patients with PON tended to have a shorter postoperative
stay than those without PON. We speculated that the reason for this was because the
group without PON tended to have more patients with left-sided colorectal cancer than
the group with PON (67 versus 53%) in our study. Therefore, left-sided colorectal
cancer patients may remain hospitalized for a longer duration due to concerns about
anastomotic leakage in rectal cancer patients. The occurrence of POV was not correlated
with the postoperative stay. Together, these results suggest that PONV after colorectal
cancer surgery seems to be only temporary in the early postoperative period and might
be unlikely to influence the final outcome.
One limitation of the present retrospective study is that it was conducted at a single
institution. Therefore, further studies will be needed in order to confirm the risk
factors of PONV after colorectal cancer surgery.
Conclusion
In conclusion, our findings suggest that female gender is an independent risk factor
for PON after colorectal cancer surgery. In addition, a high BMI and right-sided colon
cancer are independent risk factors for POV. Postoperative nausea and vomiting might
be relatively unlikely to influence the postoperative course concerning the oral intake
of solid food and postoperative stay. However, PONV is still extremely distressful
for patients in the early postoperative period. Therefore, assessments using these
factors might help prevent PONV.
