After an Institutional Review Board approved this prospective randomized double-blinded controlled study (no. 2020-11-022-002), 72 adult patients aged 26-81 years, with an American Society of Anesthesiologists physical status ≤ III, and who were scheduled to undergo elective CRS with HIPEC under general anesthesia provided informed consent to receive the 4Q-TAP block and were enrolled in the study. Patients who had local infections or coagulation abnormalities, could not communicate, were drug-addicts, or were using opioids by various routes before surgery were excluded from the study.

On the day of the surgery, the 72 patients were randomly assigned to either the 4Q-TAP group or the control group using a computer-generated randomization table. We used a block randomization method where block sizes of 2 and 4 were also randomly mixed. We used sealed, opaque, sequentially numbered envelopes that were opened by a single investigator when the patient arrived at the operating room (Fig. 1).

The 4Q-TAP group received a postoperative, ultrasound-guided, four-quadrant TAP block, and the control group received an equivalent volume of normal saline. The same type of syringe was used for administering the study drugs; the color of the syringe and that of the study drugs was identical, and eventually the syringes containing the study drugs were indistinguishable. A designated independent investigator prepared four syringes containing equal volume mixtures of 0.75% ropivacaine and 2% lidocaine (10 ml × 4) or normal saline (10 ml × 4) according to the assignment, and did not participate in any other study processes. All patients, surgeons, anesthesiologists, and follow-up observers were blinded to the assignment.

The patients were taken to the operating room where standard monitors were applied, including pulse oximetry, electrocardiography, blood pressure and temperature monitoring, capnography, and end-tidal gas analysis. Before the induction of anesthesia, all patients were asked about experiencing motion sickness, smoking history, and current smoking habits. Anesthesia was performed by an attending anesthesiologist. None of the patients received any premedication. Total intravenous anesthesia with propofol and remifentanil was administered to the patients, the blood pressure was adjusted to approximately 20% of the preoperative blood pressure, and the bispectral index level was maintained between 40 and 60. Neuromuscular blockage was performed with 0.8 mg/kg of rocuronium iv, and for sustained relaxation, 5 µg/kg/min of rocuronium was infused continuously. The lungs were ventilated with O₂ air to maintain normocapnia with a basal positive end-expiratory pressure of 5-7 cmH2O. All patients underwent goal-directed fluid therapy wherein the total volumes of fluids and transfusion, urinary output, and estimated blood loss were examined. To avoid surgical site infections, antibiotics were administered 1 h before skin incision, and 30 mg of ketorolac and 5 mg of dexamethasone were administered to relieve excess abdominal inflammation, if not contraindicated. Ketorolac was not used in patients with elevated creatinine concentrations to avoid the possibility of acute kidney injury in perioperative patients [9]. Just before HIPEC, 2 g of propacetamol was prophylactically administered to avoid profound systemic hyperthermia. HIPEC was performed for 90 min in all patients. On completion of HIPEC, the abdominal cavity was opened again, and the surgical field was checked for bleeding or injury due to HIPEC. The abdomen was finally closed, and stoma formation was performed as necessary.

After suturing, all patients received ultrasound-guided bilateral TAP block from an experienced anesthesiologist. After the block, the neuromuscular blockade was reversed, and the endotracheal tube was extubated. IV PCA (Autoselector^Ⓡ, ACE Medical Co., Ltd., Korea) containing fentanyl (0.1 µg/kg/ml), nefopam (60 mg), and lidocaine (400 mg) in 100 ml was applied to the patients, and maintenance infusion was started at 2 ml/h in both groups and titrated appropriately to the pain level. Considering the postoperative pain period of 4-5 days, the total amount of IV PCA was prepared at 200 ml for the 4Q-TAP group and 300 ml for the control group. The programmed bolus dose was 1 ml and the lock-out time was 15 min. The routine postoperative use of NSAIDs was prohibited because of the risk of acute kidney injury after CRS [9].

In the PACU, recovery profiles and pain scores of all patients were measured, and 0.1 mg/kg of oxycodone [10] was administered as a rescue analgesic when the numerical rating scale (NRS) was 5 or higher. The number of rescue analgesics and the incidence of complications such as nausea, dizziness, sedation, or hypotension were observed. Patients who experienced hypotension or excessive bleeding during surgery were transferred to the intensive care unit. 4Q-TAP block.

The skin was prepared for sterilization, and the position of the transverse abdominis muscle was confirmed using ultrasound. Due to peritonectomy, the peritoneal line was not well differentiated and the intestine was often in an ileus state with little movement; therefore, the correct block point was carefully confirmed by moving the probe back and forth widely. First, the position of the transversus abdominis muscle was confirmed in the midaxillary line between the costal margin and the iliac crest. Then, the linear probe was moved forward until it met the rectus abdominis muscle. Between the transversus abdominis and rectus abdominis muscles, bilateral subcostal TAP blocks were performed with a 10 ml mixture of 2% lidocaine and 0.75% ropivacaine per site. The operator carefully inserted the needle using the in-plane ultrasound technique to avoid bowel perforation. When the target point was reached, 0.5 ml of the mixture was injected after confirming the negative pressure. If there was a spread of local anesthetics between the target muscles, the remaining 9.5 ml was injected. For the lateral approach, the probe was positioned in the midaxillary line between the costal margin and iliac crest, and the same amount of drug was injected between the transversus abdominis and internal oblique muscles. If the lateral approach in the midaxillary line was challenging due to stoma formation or drainages, the probe was moved slightly to the rear.

The primary endpoint was less than 3 times of the rescue analgesic administration (%) for 5 postoperative days. Secondary endpoints included fentanyl doses of IV PCA every 6 h for 5 days, oxycodone requirement in PACU, total opioid consumption converted to morphine milligram equivalent (MME), local anesthetic toxicity, hospital stay, and postoperative morbidity and mortality.

Maintenance doses of IV PCA were started at 2 ml/h and titrated according to the patients’ NRS and complications, if any. When a patient's NRS was 5 or higher and not controllable with the IV PCA bolus, the patient was administered 25 mg of pethidine or 50 mg of tramadol as rescue analgesics. If a patient was calm during sleep, they were not disturbed and were evaluated at NRS of < 5.

Despite the use of the bolus injection, if the patient’s NRS score was higher than 5 and they frequently requested additional rescue analgesics, we increased the dose of IV PCA by 1 ml/hr. If the patient’s NRS score was less than 3, or they had opioid-related side effects such as sedation, dizziness, nausea, retching, or vomiting, we reduced the maintenance dose of IV PCA to 1 ml/h. If opioid-related side effects persisted without improvement, IV PCA was temporarily stopped, and symptoms were observed. When symptoms subsided and an NRS score higher than 5 was reached, IV PCA infusion was restarted at 1 ml/h.

The total frequency of additional analgesic administration for 5 postoperative days was recorded. The NRS scores and maintenance dose of IV PCA were also recorded every 6 h for 5 postoperative days. The IV PCA duration was checked, and if it was stopped, the reason and the time of stopping were recorded. Any adverse events associated with TAP block or IV PCA, hospital stay, and postoperative morbidity and mortality were recorded. As an index of the recovery profile following surgery, postoperative morbidity scores were evaluated in nine domains including infectious, pulmonary, renal, gastrointestinal, cardiovascular, hematological, wound, neurological, and pain problems at 3, 5, 8, 15, and 30 days postoperatively [11]. Pulmonary complications were evaluated based on the requirement of supplemental oxygen, and the infectious domain was evaluated by the current use of antibiotics or body temperature of > 38°C. Renal complications were evaluated by oliguria, increased serum creatinine (> 30% of baseline), or urinary catheter placement, while the gastrointestinal domain was evaluated by an inability to tolerate an enteral diet, including a feeding tube, for any reason. A case wherein a normal diet could not be processed even on the 5th day after CRS with HIPEC was defined as prolonged postoperative ileus, and its incidence was also compared. Cardiovascular domains were evaluated with the following criteria: de novo myocardial infarction or ischemia, hypotension requiring drugs or fluids, arrhythmia, pulmonary edema, or thromboembolic events. Neurological deficits, including confusion and delirium, wound problems, and bleeding requiring blood transfusion were evaluated in each domain. Surgical pain requiring parenteral opioids or regional blocks was also evaluated in the pain domain.

The primary outcome was the percentage of patients who received less than three rescue analgesics administered for 5 postoperative days. Seventy-two patients were required to have an 80% chance of detecting, as significant at the 5% level, an increase in the primary outcome measure from 20% in the control group to 50% in the 4Q-TAP group.

Data are presented as mean ± SD or number of patients (%). Nonparametric data are presented as median (1Q, 3Q). Quantitative variables were analyzed using the independent t-test or Mann-Whitney U test after normality assumption test with Kolmogorov-Smirnov test. Qualitative variables were analyzed using the chi-square test or Fisher’s exact test. Repeated measure of analysis of variance (RM ANOVA) was performed for repeatedly measured data such as NRS and the amount of fentanyl in IV PCA. All data were analyzed using SPSS^® software (version 26.0, IBM Co., USA) and GraphPad Prism (version 7.05, GraphPad Software, USA). Statistical significance was set at P < 0.05.

A total of 72 patients were approached for the study and randomized into two groups (Fig. 1). Three patients in the control group were excluded from the study. In 2 patients, palliative surgery was performed instead of CRS, and 1 patient had no CRS and no HIPEC due to poor general condition. Consequently, 36 patients in the 4Q-TAP group and 33 patients in the control group participated in this study. There was no significant difference in the demographic characteristics, Apfel’s postoperative nausea and vomiting (PONV) risk scores, and cancer origin (Table 1) [12,13].

Peritoneal carcinomatosis index (PCI) scores and completeness of cell reduction were not different between the two groups. Estimated blood loss was significantly higher in the control group than in the 4Q-TAP group (400 ml vs. 350 ml, median) (P = 0.036). However, the operation time and anesthetic time were similar and other intraoperative variables including fluid management, transfusion, and urinary output were not different (Table 2).

RM ANOVA was performed to compare the two groups of NRS scores measured repeatedly every 6 h for 5 postoperative days. However, the sphericity assumption was not satisfied (P < 0.001, Mauchly's sphericity test); hence, the degrees of freedom were adjusted with the Greenhouse-Geisser sphericity correction (adjustment factor = 0.484) [14,15]. No significant difference was observed in the two groups for the NRS scores measured every 6 h for 5 postoperative days. (RM ANOVA with Greenhouse-Geiser sphericity correction, between-subject effect, P = 0.407). The NRS score was significantly different with the passage of time (within-subject effect, P = 0.008), but there was no interaction effect between time and group (P = 0.686) (Fig. 2).

Rescue analgesic consumption was divided into the PACU and the ward. In the PACU, the 4Q-TAP group had a significantly smaller oxycodone requirement compared to the control group. The dose of oxycodone was mean ± SD, 0.04 ± 0.07 mg/kg in 4Q-TAP group and 0.09 ± 0.08 mg/kg in control group (P = 0.011). The number of patients who did not receive analgesics in the PACU was also significantly higher in the 4Q-TAP group (24 patients, 66.7%) than in the control group (10 patients, 30.3%) (P = 0.003). However, in the ward, the total frequency of rescue analgesic administration < 3 times for 5 postoperative days was not different between the two groups (24 patients [66.7%] in 4Q-TAP group vs. 15 patients [45.5%] in the control group, P = 0.076).

The administration of fentanyl dose in IV PCA was also analyzed using the RM ANOVA. It also did not satisfy the sphericity assumption (P < 0.001, Mauchly's sphericity test), and the degree of freedom was adjusted with the Greenhouse-Geisser sphericity correction (Adjustment factor = 0.203) [14]. The fentanyl dose administered by IV PCA for 5 postoperative days was significantly lower in the 4Q-TAP group than in the control group (RM ANOVA with Greenhouse-Geiser sphericity correction, between-subject effect, P = 0.038). The fentanyl dose in IV PCA was significantly different with the passage of time (within-subject effect, P < 0.001), and there was also a significant interaction effect between the two variables (P = 0.029) (Fig. 3). A significant dose reduction of fentanyl in IV PCA of the 4Q-TAP block was maintained for 36 h after surgery (Fig. 3).

To specify the effects of the 4Q-TAP block, we converted all administered opioid doses, including IV PCA and rescue analgesics to MME per kilogram [16]. Total MME/kg for 5 postoperative days in the 4Q-TAP group was 2.35 (1.63, 3.27) mg/kg and significantly smaller than 3.21 (2.28, 4.31) mg/kg in the control group (Mann-Whitney U test, P = 0.009).

Consequently, the 4Q-TAP group effectively reduced total opioid consumption during the 5 days after CRS with HIPEC.

Hospital stay and mortality rates were not significantly different between the two groups (Table 3).

In the postoperative morbidity score, not all domains of postoperative complications differed between the groups (Table 3). The incidence of pulmonary morbidity was similar between the groups. There was no difference in the duration of supplementary oxygen use (6.3 vs. 7.5 days, P = 0.348), and no patient in either group required ventilator care for 5 postoperative days.

The incidence and duration of gastrointestinal morbidity were not significantly different between the two groups (Table 3). Time of liquid diet and regular diet were similar in both groups and prolonged postoperative ileus also occurred at a similar frequency (4Q-TAP, 33.4% vs. Control, 36.0%) in both groups.

The total period of parenteral opioid administration was not different between the groups (mean, 15.3 vs. 16.9 days). Local anesthetic toxicity, such as seizures, metallic taste, blurred vision, ear fullness, or fatal arrhythmia, did not occur in either group.

During 5 days after the surgery, there was no significant difference between the groups in the incidence of IV PCA-related adverse events, including sedation, PONV, dizziness, headache, and heartburn (Table 4). Sedation developed in 2 patients in both the groups. PONV within 24 h developed in 8 patients (22.2%) in the 4Q-TAP group and 4 patients (12.1%) in the control group (P = 0.269). The overall incidence of nausea for 5 days was higher in the 4Q-TAP group (33.3%, 12 patients) than in the control group (18.2%, n = 6), but the difference was not significant (P = 0.152).

The incidence of temporary clamping of IV PCA was 6.1% in the control group and 22.2% in the 4Q-TAP group (P = 0.057), but most of them used IV PCA again without further side effects. The PCA duration was shorter in the control group (89.8 ± 37.1 h) than in the 4Q-TAP group (97.1 ± 38.1 h), but the difference was not statistically significant.