Postoperative Pain Relief after Pancreatic Resection: Systematic Review and Meta-Analysis of Analgesic Modalities

Postoperative pain after pancreatic resections is frequent [1, 2]. This may be attributed to a high incidence of preoperative pain resulting in use of analgesics prior to surgery and resection requiring extensive abdominal dissection with big incisions [2]. Inadequate pain control following any surgical procedure increases overall morbidity, hospital stay and recovery time [3, 4].

Epidural analgesia (EDA) is generally the analgesic modality of choice in pancreatic surgery and has been shown to have lower systemic complications such as pneumonia [5, 6], acute coronary syndrome, thromboembolism and renal failure[7], though at the expense of the need for vasopressors [8], excessive fluid administration [5], lengthened intensive care stay [9], and higher rates of complications such as postoperative pancreatic fistula (POPF) [8]. When EDA cannot be used, opioids via patient controlled analgesia (PCA) is often the preferred alternative. Unlike EDA, PCA does not promote hypotension, so undoubtedly is associated with a reduced need for vasopressor therapy and fluid administration [5]. More recently, transabdominal wound catheters (TAWC) have become more common in general abdominal surgery, as it has shown to be comparable to EDA in terms of pain relief with fewer complications, such as block failure and hypotension [10].

Although a variety of pain modalities have been explored for the management of postoperative pain after pancreatic surgery, the literature is generally limited to pair-wise comparisons, small study sizes and heterogeneity in their study population[2, 5, 8, 9, 11‐23] making it difficult to justify routine use of one pain modality over the other. This is reflected in the recently published ERAS guidance [6] which recommends EDA for postoperative pain relief and TAWC as an alternative, however the majority of evidence for this recommendation was extrapolated from non-pancreatic surgery.

The present meta-analysis and systematic review therefore aimed to summarize and compare the efficacy of different local and regional pain relief modalities in the management of postoperative pain following pancreatic resection.

The study protocol was registered on PROSPERO (ID: CRD42020215886).

The original search identified 4912 studies, of which twenty-six full text articles were screened. Following this, twelve studies [5, 8, 13, 15, 17‐21, 35‐37] met the inclusion criteria, of which eight studies [5, 13, 19‐21, 35‐37] were included in the meta-analysis. The remaining four studies [8, 15, 17, 18] were included in the narrative review (Fig. 1).

Overall, 1313 patients were included. This incorporated five RCTs [5, 8, 15, 17, 21] and seven retrospective cohort studies [13, 18‐20, 35‐37], published between 2020 and 2008. Studies were conducted in the USA (n=6), UK (n=1), New Zealand (n=1), Netherlands (n=1), Korea (n=1), Italy (n=1) and other parts of Europe (n=1) (Table 1).

Eight articles compared EDA and PCA [5, 13, 19‐21, 35‐37] and were included in the quantitative analysis. Other articles compared EDA and TAWC [8, 15] (n=2), high dose PCA and low dose PCA (n=1) [17] and functional EDA and prematurely aborted EDA (n=1) [18]. Of these, none of the included studies stated the use of an ERAS pathway in their methods section. Where applicable, two out of four studies defined POPF and one out of four studies defined DGE using the ISGPS definitions.

The present systematic review and meta-analysis of postoperative pain management in pancreatic surgery has demonstrated that EDA provides similar level of postoperative pain relief when compared to PCA on POD2 and POD4 after both PD and DP. Furthermore, there were no significant differences in pain relief or other perioperative outcomes when comparing EDA and TAWC, high dose PCA and low dose PCA or functional EDA and aborted EDA in PD.

EDA is widely accepted as the gold standard for pain relief following major abdominal surgery [6]. However, a recent meta-analysis of RCT’s of EDA in major abdominal surgery has shown that although EDA may provide superior pain control, the perioperative outcomes are comparable to other forms of analgesia such as PCA [38]. Furthermore, patients on EDA require increased perioperative fluid administration due to sympathetic blockade [5], and have an increased incidence of perioperative complications, particularly higher POPF rate in those undergoing PD in several recent studies [8, 39‐41]. In the present review, when EDA was compared with PCA, pain scores were comparable and both groups had similar postoperative complications. Although comparable data was not available on the use of postoperative fluid requirement, Klotz et al. [5] in a RCT comparing EDA with PCA showed significant weight gain and need for vasopressors with EDA, albeit with no significant increase in postoperative complications, in addition to higher failure rate with EDA (18.5%). Similarly, Simpson et al. [42] in a retrospective series, showed 31% of patients developed either hypotension or opioid toxicity after EDA in the postoperative period, albeit with improved pain scores compared to non-EDA. A more recent study using a goal-directed fluid restriction strategy with EDA during pancreaticoduodenectomy has shown lower rates of POPF and DGE [41]. The present evidence regarding the impact of volume of perioperative fluids and postoperative complications in pancreatic surgery is predominantly derived from retrospective studies and larger studies are warranted. ERAS society guidelines for pancreatic surgery suggest a high evidence level for superior pain control with EDA and a low evidence level for recommendation of EDA to reduce overall morbidity [6]. The results from the present meta-analysis suggest EDA and PCA provide similar levels of pain relief and morbidity postoperatively, however further studies are needed with predefined end-points to see the effect of EDA on POPF and morbidity following pancreatic surgery [6].

TAWC are increasingly being used in pancreatic surgery, given the perceived benefits of TAWC in major abdominal surgery [10, 43]. TAWC provides a similar level of pain relief as EDA and is associated with fewer complications [8, 10, 44]. Two studies in the present review compared EDA with TAWC with different primary and secondary outcomes, making interpretation of benefits of one analgesic modality over other difficult. No significant difference was found in OT, POPF, DGE, LOS, significant morbidity, mortality or opioid use. On the contrary, a study by Newhook et al. [12] found EDA resulted in lower opioid requirements compared to TAWC, however the pain scores in the postoperative period were similar between the analgesic modalities and failure rate was higher with EDA when compared to TAWC. Furthermore, on POD3 there was tendency trend for increased need for vasopressors after EDA with a higher proportion of patients with a postoperative rise in creatinine compared to baseline. The postoperative outcomes in all included studies were comparable between EDA and TAWC, findings similar to a recent RCT of EDA and TAWC in HPB surgery [8] which showed comparable pain relief with EDA and TAWC, however TAWC was associated with shorter anaesthetic time, lower mean cumulative vasopressor and opioid consumption. A post-hoc sensitivity analysis including only patients undergoing PD again showed non-inferiority of TAWC over EDA.

There are several limitations to the present review. The postoperative pain scores were assessed by few studies thereby meta-analysis was only possible for EDA versus PCA. The varied primary and secondary outcomes of included studies meant we could not undertake a meta-analysis of perioperative outcomes and was limited to a narrative review. Most the included studies were non-randomised, thus at risk of bias. Furthermore, there is paucity of data on patient related outcomes and a lack of data regarding the perceptions and preferences of patients. However, this is the first comprehensive review of analgesic management in patients undergoing pancreatic surgery comparing relative benefits for each analgesic modality.

For a practicing clinician, the present review summarized the available evidence on postoperative pain management after pancreatic surgery. The majority of evidence is centered around the use of EDA, PCA and TAWC, with comparable pain relief with all three analgesic modalities, in addition to a similar profile of postoperative complications. Depending on the availability of local expertise, all the above analgesic modalities provide adequate pain relief in the postoperative period. Nevertheless, there is still a lack of robust randomised evidence regarding the impact of increased fluid requirements with EDA and postoperative complications such as POPF, when compared to PCA or TAWC, as none of the trials were adequately powered to evaluate this. In addition, is it unknown which analgesic modality provides adequate pain relief when patients develop postoperative complications such as postoperative acute pancreatitis or POPF. Further high-powered RCTs are warranted to assess the relative merits of these analgesic modalities on not only postoperative pain, but postoperative outcomes with emphasis on patient related outcomes and quality of life, particularly in the setting of ERAS pathways. In addition, given the morbidity profile of PD and DP is different, future trials should aim to separate these two patient groups when evaluating postoperative outcomes.