Comparison of epidural infusion versus intrathecal morphine block as part of enhanced recovery after open pancreatoduodenectomy

The accepted approach to pain management following open pancreatoduodenectomy (PD) remains controversial, with the most recent enhanced recovery after surgery (ERAS) protocols recommending epidural anesthesia (EA). Few studies have investigated intrathecal (IT) morphine, combined with transversus abdominis plane (TAP) blocks. We aim to compare the different approaches to pain management for open PD.


| INTRODUCTION
Postoperative pain following abdominal surgery, including pancreatoduodenectomy (PD), is challenging to manage, yet critical for functional recovery. 1The ideal perioperative analgesic technique for these patients remains under debate.Opioids have long been the mainstay of pain management, but the drastic increase in the prevalence of opioid use disorder and opioid overdose deaths over the past two decades has prompted a paradigm shift to opioidsparing approaches. 2,3hanced recovery after surgery (ERAS) programs have been launched in a variety of surgical fields as a multidisciplinary and evidence-based framework designed to improve clinical outcomes.
A key element of these programs includes multimodal pain management.5][6] The latest ERAS update strongly recommends the use of epidural analgesia (EA) based on review of the literature, concluding that EA offers improved analgesia compared to intravenous opiates, improved return of postoperative intestinal function, and reduced pulmonary complications. 5spite this recommendation, the optimal pain management approach for PD remains controversial.The reported use of EA varies from 11% to 85%, 1 and a recent international survey showed that only 61% of surgeons regularly follow ERAS protocols. 7Other acute pain management options include continuous wound infiltration, spinal or intrathecal (IT) blocks, and regional blocks such as transversus abdominis plane (TAP) blocks.In North America, TAP block was ranked as the most commonly used technique for analgesia, but there is a lack of research on the effectiveness of this strategy for PD. 7 With substantial variability in current clinical practice and the paucity of data, further research is needed to determine the optimal method for pain control after open PD that uses an opioid-sparing approach.The current study aims to evaluate our institutional experience and compare the effectiveness of EA, IT morphine blocks, and regional TAP blocks for open PD.Additional covariates extracted to assess postoperative recovery included fluid boluses within 48 h of surgery or between 48 and 96 h of surgery, as well as postoperative days for solid food intake, Foley catheter removal, clearance for discharge home by physical therapy, and length of stay (LOS).In the epidural group, our practice is to keep the catheter in place until epidural removal.In the IT morphine group, the catheter was removed on postoperative Day 1 unless the catheter was needed for strict output monitoring or other reasons.

| Patient selection
The postoperative day of catheter removal was the final date of removal with spontaneous voiding if the catheter was replaced or not.Postoperative day for epidural removal and epidural complications were recorded for all patients receiving EA.Epidural complications were characterized as none, repositioning required, or general epidural malfunction.

| Clinical management
Preoperatively, patients routinely received acetaminophen, gabapentin, and celecoxib in the preoperative holding area.A perioperative pain management strategy including either EA, IT morphine, and/or TAP blocks was performed for each patient.The selection of pain management modality is based on surgeon and anesthesiologist preference after discussion.
EA was administered via a catheter typically placed at the T8-T10 vertebral level preoperatively.An infusion of 0.125% bupivacaine and 10 μg/mL hydromorphone was typically started after induction and continued postoperatively at a rate of 6 mL/h.Patients typically had the option to administer a bolus of 3 mL up to every 20 min.Dosing and frequency adjustments were made by a dedicated anesthesia acute pain management team postoperatively.IT morphine was administered with a one-time dose by the anesthesia team before induction; all patients who underwent IT morphine also received TAP blocks.Patients who receive IT morphine are admitted with cardiopulmonary monitoring for 24 h, during which time no routine PRN narcotics are ordered, and any narcotic doses must be approved by the surgical team.TAP blocks utilized liposomal bupivacaine (Pacira Biosciences), using 20 mL or 266 mg total, and were administered by the surgeon under direct visualization just superficial to the transversus abdominis muscle bilaterally.
Postoperatively, patients were managed using an enhanced recovery pathway that included intravenous or enteral acetaminophen, and selective use of anticonvulsant pain medication, antispasmodic, and nonsteroidal anti-inflammatory (NSAID) medications.
Generally, hydromorphone was the preferred intravenous narcotic, and tramadol was the preferred oral narcotic.Pain levels were assessed for all patients by nursing staff every 4-8 h using a standardized numeric rating scale where 0 indicated no pain and 10 indicated the patient was experiencing the worst possible pain.A narcotic prescription was provided based on the inpatient requirement before discharge.

| Statistical analysis
Patient characteristics and postoperative pain management elements were stratified by perioperative analgesic modality received (EA, IT morphine with TAP blocks, or TAP blocks alone) and were compared using Pearson's chi-square tests and Fisher's exact tests, as appropriate.A descriptive analysis was performed on a subgroup of patients who received EA as well as those who received TAP blocks only.All analyses were performed using STATA software (version 16/MP; Stata Corp.).p Values less than 0.05 were considered statistically significant.

| Patient demographics and characteristics
There was a total of 50 patients included in the study, which consisted of 18 patients in the EA group, 24 patients in the IT morphine group, and 8 patients in the TAP block only group.Table 1 summarizes patient demographics and characteristics.The median age for the overall cohort was 66 years old [interqartile range [IQR]: 58.0-73.0years] and the median BMI was 26.9 kg/m 2 [23.6-30.2].
The overall cohort consisted of 21 female (42%) and 29 male (58%) patients.The majority of patients underwent surgery for pancreatic adenocarcinoma (56%).A history of pancreatitis was found in 18 patients (36%).Preoperatively, nine patients had a narcotic prescription (18%), eight patients had a benzodiazepines prescription (16%),  and three patients were taking gabapentin or pregabalin (6%).There were no statistically significant differences between baseline characteristics between pain modality groups.

| Postoperative recovery
Table 3 shows the postoperative complications and outcomes.The most common postoperative complication was delayed gastric emptying which developed in eight patients (16%).There was no significant difference in postoperative complications between groups.In the overall cohort, 36%

| Epidural management
A total of 18 patients were in the EA group (Table 4).

| DISCUSSION
][9][10] Our study retrospectively reviewed patients at our institution who underwent open PD to evaluate the differences in clinical outcomes and recovery for epidural infusion compared to IT block with TAP blocks.We demonstrated that these offer similar clinical pain control with earlier Foley removal in the IT morphine patients.TAP blocks only appear to be inferior, with a higher postoperative opioid requirement, suggesting that it is not an adequate strategy on its own to control pain in this patient population.
In a recent survey of surgeons worldwide, the most frequently used analgesic modality for an open PD was EA (50%), followed by intravenous morphine (24%), spinal analgesia (10%), TAP blocks (9%), and continuous wound infiltration (8%). 7When stratified by country, surgeons in North America utilized TAP blocks at a much higher rate, with approximately one-third utilizing TAP blocks, almost equally as popular as EA.
ERAS pathways have been reported and implemented for multiple surgeries with the goal of reducing complications, decreasing costs, and reducing length of stay. 5In 2019, updated recommendations from the ERAS Society were published to guide postoperative care after PD. 5 For pain control, the guidelines recommended EA over intravenous morphine.
EA was favored due to its reduced gastrointestinal and pulmonary complications compared to intravenous opiates. 5In a comparison between EA and PCA in a randomized control trial across nine European pancreatic surgery centers, there were no differences in GI complications and the two procedures were determined to be comparable with regard to effectiveness and safety. 10Although EA is recommended and used by the majority of surgeons in worldwide survey, 7 some studies have noted adverse effects related to EA, including increased postoperative complication rates, intensive care unit (ICU) admissions, and duration of hospital stays in these patients. 1,11,12Given these adverse effects, it is worthwhile to consider alternative modalities that may be associated with fewer complications.
The current ERAS guidelines mention the use of spinal anesthesia modalities and their potential to have a significant downstream opioidsparing effect, 5 although the recommendations were limited by few studies available for pancreatic surgery. 13,14In a study of 233 patients, with PCA, and PCA alone in patients undergoing major hepatobiliary surgery. 13They found that IT morphine with PCA was associated with less postoperative hypotension, reduced intravenous fluid requirements, shorter hospital stays, and lower incidence of respiratory complications compared to thoracic EA.
Our current study, which represents a fairly homogeneous patient population, demonstrated a similar performance of IT morphine plus TAP blocks compared to EA.In our experience, there was no difference in the number of patients requiring fluid boluses or postoperative LOS.Our study showed no difference between analgesic modalities for the postoperative day of solid food initiation.
One notable difference was that IT morphine was associated with earlier Foley removal compared to EA patients.All patients were managed using an enhanced recovery protocol that included early ambulation and evaluation by physical therapy, a similar approach to diet advancement, and a median LOS of 6 days.
Our study has several limitations that should be addressed.First, the study is retrospective and includes consecutive sampling at a single institution, which can be inherently biased due to the decisions that are made in a real-world setting and practice patterns of the individuals involved, as well variability between surgeons and anesthesiologists.
Although an established postoperative ERAS pathway was in place and used by all surgeons to guide postoperative management, there may have been subtle differences or deviations that contributed to the results.For instance, the TAP alone cohort were patients of a single surgeon during a period when IT morphine administration was not feasible.However, we included the group in this study to compare the results and for any readers who may be curious.The balance between the preoperative and postoperative characteristics of the three groups suggests that the unavoidable biases present in a retrospective study did not contribute significantly to the result.Additionally, the study is also relatively small which limits the statistical power, but even with this limitation, it shows the feasibility and safety of utilizing IT morphine to provide similar results to the more established technique of epidural for perioperative pain control after major abdominal surgery.

| CONCLUSIONS
Our study revealed that IT morphine with TAP blocks provides comparable analgesia to EA in patients undergoing open PD and should be considered as a viable alternative for this patient population in the ERAS guidelines.We demonstrated earlier removal of urinary catheters and similar functional recovery and length of stay.Although difficult to quantify, IT morphine is anecdotally less resource intensive since the patients do not need to be seen by the pain management service beyond the additional administration.In the context of an established pancreatic surgery program following ERAS principles, this study presents preliminary evidence that IT morphine with TAP blocks may offer a less resource-intensive approach to pain management.Future prospective randomized studies are needed to confirm these findings.
This retrospective cohort study included all patients who underwent an open PD for neoplastic or pre-neoplastic disease at a single tertiary medical center between July 2020 and July 2022.The study was approved by the Institutional Review Board.To evaluate a surgically homogeneous cohort, we included patients who underwent planned or conversion to open PD via an upper midline incision for neoplastic or pre-neoplastic disease.

2. 2 |
Covariates and outcomes Patient demographics, operative details, and aspects of postoperative management were abstracted from the electronic medical record.Demographic data and characteristics included age, sex, and body mass index (BMI).Histology and postoperative surgical outcomes and complications were compared.Perioperative pain management, including doses of narcotics and blocks, was collected as well as postoperative pain scores and complications.For patients who received opioid medications postoperatively, including hydromorphone, tramadol, and oxycodone, the postoperative day that narcotics were started and the total postoperative amount was recorded.Total morphine milligram equivalents (MME) were calculated for both the hospitalization as well as the discharge prescription if given.The opioid used for IT morphine and EA were not counted toward postoperative opioid use or MME.Postoperative complications within 30 days from operation were captured and reported.

T A B L E 1
Demographics and characteristics for open PD patients.66.0 (58.0-73.0)65.0 (54.0-69.0)68.0 (61.0-74.0)61.0 (43.0-70.5)0 by group.A three-way comparison between EA, IT morphine, and TAP blocks was conducted.A pairwise comparison between EA and IT morphine was also performed.The median IT morphine dose was 200 μg.The median pain score on POD 1 was 3.0 ([1.0-5.0],p = 0.080) for the overall cohort, without any significant difference between groups.The lowest and highest pain scores on POD 3 were not significantly different.Postoperative administration of non-opioid adjunctive pain medication, including acetaminophen, NSAID, pregabalin/gabapentin, and antispasmodic, was comparable between the three groups.All patients received postoperative acetaminophen for at least 24 h.With regard to postoperative intravenous opioid requirement, the TAP block-only group required significantly more total intravenous hydromorphone during the hospital stay (1.1 mg [0.7-23.4mg]) compared to the other two groups (EA: 0.0 mg [0.0-0.4 mg], IT morphine: 0.2 mg [0.0-1.8 mg], p = 0.004), with 63% of TAP block-only patients (n = 5) requiring patient-controlled analgesia (PCA).Oxycodone administration was significantly higher in the TAP block-only group (10.0 mg [0.0-87.5 mg]) compared to the other two groups (EA: 0.0 mg [0.0-0.0 mg], IT morphine: 0.0 mg [0.0-2.5 mg], p = 0.027).Oral narcotics were started at a median of POD 2 [2.0-4.0 days] without any significant difference between the groups (p = 0.240).There was a numeric difference between MME during hospitalization between the three modalities although it did T A B L E 2 Postoperative pain management details for open PD patients.
of patients (N = 18/50) required at least one fluid bolus for hemodynamic resuscitation within 48 h of surgery.Eighteen percent of patients (N = 9/ 50) required at least one fluid bolus 48-96 h after surgery.There was no difference between the three groups in the number of patients requiring fluid bolus therapy.The median POD for solid food intake was 4.0 [4.0-6.0], which was similar between groups.Foley removal was earlier in the IT morphine group at median POD 1.5 [1.0-2.0]compared to the other groups (EA: 2.0 [2.0-3.0],TAP: 2.0 [2.0-2.5],p = 0.007).Foley removal was also earlier in the IT group when compared just to the EA group (p = 0.002).A total of five patients required Foley reinsertion due to failure to spontaneously void, with no significant difference between pain control modalities.There was no significant difference between the T A B L E 3 Postoperative complications and outcomes.
The median time to epidural removal was POD 3.0 [3.0-3.0].Of the patients who received EA, 2 patients had epidural catheter malfunction, 1 patient required epidural repositioning, and 15 patients had no epidural issues or complications.

T A B L E 4
Abbreviations: IQR, interquartile range; POD, postoperative day.

Table 2
displays the postoperative pain management details stratified