Minimally invasive techniques versus opioids in patients with unresectable pancreatic cancer: a systematic review and meta-analysis of randomised controlled trials

Introduction

Background

Pancreatic cancer (PC) is currently the sixth leading cause of cancer-related death worldwide. With a 5-year survival rate of only 12.8% and a worldwide incidence of more than 510,000 cases in 2022, PC is expected to become the second-leading cause of cancer-related mortality by 2030 (1,2). Even with highly performing diagnostic modalities, the cancer is unresectable at the initial diagnosis in 80–90% of cases (3). A survey analysis from the Pancreatic Cancer Action Network showed that 93% of PC patients had tumour-related pain during the disease course, while for the majority, the pain ranged between moderate to severe (4). The burden of the disease’s physical manifestations significantly deteriorates quality of life (QoL), affecting mental, physical and social well-being. As the course progresses, the focus turns to symptomatic management, with pain control being at the centre of palliative care in PC (5).

Rationale and knowledge gap

The pathophysiology of pain in PC is complex, being the sum of visceral, somatic, and neuropathic factors (6). Even though the three-step treatment ladder established by the World Health Organisation (WHO) has been the criterion formally preferred, opioids are often misused and associated with adverse events (AEs) further decreasing QoL (7). The necessity for an additional step on the treatment ladder has thus arisen, with the celiac plexus (CP) receiving increasing attention in the pain management of PC patients, and various procedures being shown to ablate it and reduce the disease burden. New minimally invasive techniques have been proposed, such as high-intensity focused ultrasound (HIFU), microwave ablation, cryoablation, irreversible electroporation, and stereotactic body radiation therapy. Previous meta-analyses have shown that celiac plexus neurolysis (CPN) is safe and effective with the advantage of fewer AEs compared to standard analgesia (8,9). However, the studies included followed the patients for a maximum of 8 weeks and focused only on one type of intervention. In summary, while some methods have shown promising results in pain control, there is still a gap in the knowledge regarding the best-performing method (10).

Objective

We aimed to compare all literature evidence from randomised controlled trials (RCTs) to find the most effective intervention to treat pain, with minimal risk of AEs, in patients with unresectable PC. This manuscript is written in accordance with the PRISMA reporting checklist (available at https://tgh.amegroups.com/article/view/10.21037/tgh-24-141/rc).

Methods

This systematic review and meta-analysis is reported based on the recommendation of the PRISMA 2020 guideline (11) (Table S1), while the Cochrane Handbook methodology was followed (12). Initially, a network analysis was planned; however, due to the low number of available articles, only a meta-analysis using a 2-level model with direct and a 3-level model with indirect comparison could be performed. Additionally, we could estimate survival curves based on individual patients’ data extracted from reported Kaplan-Meier curves. Besides, the study protocol was fully adhered to and is registered on PROSPERO (CRD42023477094).

Eligibility criteria

Only RCTs were included, irrespective of publication date or language. Studies were eligible for inclusion if they reported exclusively on (P) adult patients (human subjects with a minimum age of 18 years) with a proven diagnosis of unresectable PC and who received cancer-related pain-reducing interventions independent of other tumour characteristics. The main intervention groups accepted (I) were CPN, ablative techniques, and drug administration, compared to (C) pharmacological treatment (opioids, non-opioids, oncological therapies) or placebo. Our outcomes of interest (O) were pain, QoL, survival, AEs, and analgesic drug use.

Information sources

Our systematic search was conducted on October 29, 2023, in five major databases: MEDLINE (via PubMed), Embase, CENTRAL (The Cochrane Central Register of Controlled Trials), Scopus, and Web of Science. No filters or restrictions were used. The Scopus database was searched using the “Article title, abstract, keywords” option. The search key included terms and synonyms for “pancreatic cancer”, “pain”, and “randomised controlled trials”. The entire search key can be found in the Table S1.

Selection process

Two independent review authors (IIR and OCI) performed the selection. All references were imported into Endnote 20 (Clarivate, 2013) to remove duplicates. Title-abstract selection was performed in Rayyan, followed by full-text selection. A third senior independent reviewer (A.R.) resolved any disagreements. Cohen’s kappa coefficient (κ) was calculated to assess inter-rater reliability during the selection. The reference lists of all the included articles were checked to identify additional eligible papers on November 29, 2023, on citationchaser, and title-abstract selection was performed in Rayyan. Five full-text articles could not be retrieved, despite efforts to contact the authors.

Data collection process

Data were collected independently from the eligible articles by three authors (I.I.R., I.C.M., and O.C.I.) in a standardised form, and a third independent reviewer (A.R.) resolved any disagreements. The following data were extracted from the eligible articles: title, first author, DOI code, year of publication, period of data collection, country, single/multicentre study, study design, inclusion and exclusion criteria, the total number of patients, patient demographics, the number of patients in the intervention and control groups, interventions, outcomes, and outcome dynamics over time. The sample size, mean and corresponding standard deviation (SD), and median with the corresponding quartiles were extracted for continuous outcomes. For dichotomous outcomes, the total number of patients and those with the event of interest in each group separately were extracted or calculated from the studies where it was available. The hazard ratio (HR) with its 95% CI was extracted for time-to-event data. However, several articles did not report the desired HR as an outcome measure and used a different outcome measure for time-to-event data. Where a Kaplan-Meier curve was given, an attempt was made to gather or estimate the individual patient’s data based on these plots at the first step of the analysis. The WebPlotDigitizer tool was used to extract the values from the plot. Where possible, the extracted and calculated data (e.g. median time-to-event times, given year survival probabilities, HR based on Kaplan-Meier estimates and Cox-proportional hazard model) were compared with the published data. This difference was found to be neglectable. Where studies provided incomplete data, the corresponding authors were contacted with requests for additional information necessary for analysis.

Study risk of bias (RoB) assessment

Two authors independently performed the RoB assessment with the ROB-2 Tool (I.I.R. and O.C.I.), and an independent third investigator (A.R.) resolved the disagreements.

Statistical analysis

As considerable heterogeneity was assumed among studies’ populations, the random effects models were used in a frequentist framework.

To assess pain, an analysis of Visual Analog Scale (VAS) scores was performed. As more interventions may be available in the same study and create a dependent effect, a “3-level” (multilevel) model was used to include one, two and multi-arm studies together for pooling the VAS means and proportions for the presence or absence of given AEs. The analyses were performed separately for different time points. Details of the models used can be found in Appendix 1.

The survival probabilities at specific time points were estimated and pooled with a 3-level model to indirectly compare the survival in all groups. An estimation for distribution-free pooled survival curves was also performed using the curve estimate method.

Results were considered statistically significant if the pooled CI did not contain the null value. The findings were summarised in tables, on forest plots, and on estimated survival curves. If appropriate, between-study heterogeneity was described by the between-study variance (τ2) and I² statistics based on the 3-level model. Prediction intervals were not reported, as the study number was small. Potential outlier publications were explored using different influence measures and plots.

All statistical analyses were calculated by R software using the meta (v7.0.0) package for basic meta-analysis calculations and plots and dmetar (v0.1.0) package for additional influential analysis calculations. Package metafor (v4.4.0) was used for the multilevel models. Survival (v3.5.8) and survminer (v0.4.9) packages were used for individual patient data-based calculations. MetaSurvival (v0.1.0) package was used to estimate survival curves, which were plotted using the ggplot2 (v3.5.0) package. Publication bias was not assessed because less than 10 studies were included in the meta-analysis.

Results

Search and selection

The systematic literature search yielded 5,491 results. No additional study was found with citationchaser. A summary of the selection process is available in Figure 1. In total, 21 RCTs were included.

Figure 1 The flowchart of the included studies selection according to PRISMA guidelines.

Basic characteristics of included studies

The included studies reported on 1,644 patients from nine countries. Out of these, 1,085 patients were randomised to different minimally invasive treatment options for the management of cancer-related pain. A summary of the baseline characteristics of the included studies is detailed in Table 1. The most used procedure was percutaneous CPN (P-CPN) (ten studies) (13-22), while pharmacological management consisting of opioids was the main control (nine studies) (15,17,20,22-27). Other methods for performing CPN were intraoperatively (I-CPN) and with endoscopic-ultrasound guidance (EUS-CPN). A small part of the included studies analysed different techniques, such as HIFU (27-29), radiofrequency ablation (30), interventional chemotherapy (31) and external-beam radiotherapy (32). Only seven studies were double-blinded (13,15,17,20,25,26,33). Except for one, the rest were randomised 1:1 (23). All studies reported several of our outcomes of interest, at multiple time points from study initiation, most frequently at 4 and 12 weeks. The patient demographics were similar across studies regarding age and gender distribution.

Pain relief

All included studies reported pain scores measured at different time points, most frequently at 4, 8, and 12 weeks from randomisation (range, 1–32 weeks). The most common way to assess pain was on the 11-point VAS or equivalents, with only two studies choosing different scales (Table S2) (26,30). The baseline pain values ranged from moderate (4.5–7.4) to severe (7.5–10) in all studies (34). Six studies with almost 440 patients were available for analysis 4 weeks after the intervention (Figure S1) (13,15,20,22,24,27). VAS values decreased with 3.5 to almost 6 points in all study groups after receiving the intervention compared to the initial assessment. Figure 2 showcases a summary of the dynamic of pooled means for pain scores with each intervention, at each follow-up.

Figure 2 The dynamic of the pain scores. (A) Pooled means for baseline pain scores, (B) summary of pooled means of pain scores at each follow-up time, (C) time dynamic of the pooled means for pain scores with each intervention. CI, confidence interval; EUS-CPN, endoscopic ultrasound-guided celiac plexus neurolysis; HIFU, high-intensity focused ultrasound; I-CPN, intraoperative celiac plexus neurolysis; MRAW, raw mean; P-CPN, percutaneous celiac plexus neurolysis; SD, standard deviation; VAS, Visual Analogue Scale.

P-CPN

By the end of the 4 weeks, the pooled mean scores for pain were 2.27 (95% CI: 1.63–2.91) with P-CPN, and 2.80 (95% CI: 2.17–3.42) with opioids. At eight weeks, these scores started to increase slowly and at the 12-week follow-up, pain scores higher than 3 on the VAS scale were noted (Figures S2,S3). Results of the direct comparison between P-CPN and opioids using the 2-level model are available in Figure S4.

EUS-CPN

Individual studies showed at 4 weeks from randomisation, pain scores of 1.30 (95% CI: 0.68–1.92) with EUS-CPN and at 8 weeks values of 2.50 (95% CI: 0.97–4.03). At 12 weeks from baseline, while P-CPN and opioids showed a similar potency, the pain scores after EUS-CPN were slightly lower, 1.47 (95% CI: −0.44 to 3.38) as shown in individual studies (Figures S2,S3).

I-CPN

One study analysing I-CPN showed, at 8 weeks, pain scores similar to those achieved 4 weeks after EUS-CPN (25).

HIFU

At 4 weeks from randomisation, pain values from individual studies were 1.45 (95% CI: 1.37–1.53) with HIFU.

QoL

QoL was assessed in 754 patients from 11 studies with 18 reporting scales (Table 2) (13,15,19-24,26,30,33). All provided results concerning different types of CPN techniques. In most cases, a rise in the scale value meant an increase in the QoL. A significant difference between study arms was observed in only 4 of the 11 studies included (15,19,23,30), with an additional one (24) reporting a significant improvement for the intervention group compared to its baseline, although only for “communication with the medical staff”. In the other cases, the QoL assessment showed the same tendency for both compared treatments, with an initial increase followed by an eventual decrease.

P-CPN

Compared to opioid treatment, P-CPN was shown to statistically increase the values on the physical domain of the Short Form Health Survey (SFHS) QoL scale, at 4, 8 and 12 weeks from baseline (15). When two different approaches of P-CPN were analysed, the transaortic P-CPN demonstrated a significant superiority concerning the QoL evaluation compared to the bilateral splanchnicectomy (19).

EUS-CPN

EUS-CPN significantly improved the general QoL of patients compared to opioids at 12 weeks from baseline in one of the included studies (23). However, EUS-RFA provided a significantly better result than EUS-CPN for the functional and symptomatic domains of the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire core questionnaire 30 (EORTC C30) scale at 4 weeks from procedures (30).

Survival analysis

The survival analysis included nine RCTs, in which a total of 680 patients were observed (15,19-21,24,25,28,29,31). Based on the estimated survival curve from individual patients’ data (detailed statistical description in Appendix 1), the median survival time in days was 126.22 (95% CI: 87.21–165.31) for opioids. Given the relatively small number of deaths in most of the articles and the short follow-up time, 95% CI could not be calculated for all interventions, and the median number of days survived was 250.61 with chemotherapy. Figure 3 shows a summary of the survival curves after each intervention (for individual calculations, see Figure S5).

Figure 3 The summary of median survival time curves with each intervention analysed. CXT, chemotherapy; EUS-CPN, endoscopic ultrasound-guided celiac plexus neurolysis; HIFU, high-intensity focused ultrasound; P-CPN, percutaneous celiac plexus neurolysis.

P-CPN

The median survival time in days was 86.71 (95% CI: 62.45–136.88) for P-CPN. This value was the lowest among the rest.

EUS-CPN

The median number of days survived after EUS-CPN was 145.38.

HIFU

The median number of days survived was 294.97 after receiving HIFU.

Analgesic drugs use

Fourteen studies involving a total of 1,010 patients reported results concerning analgesic drug use in patients randomised to CPN interventions, CP ablation, HIFU, or opioid treatment (Table 3) (13-15,17,19-24,26,27,30,33). All patients received at least one opioid (in most cases, oral morphine), and in two studies, the consumption of non-steroid anti-inflammatory drugs (NSAIDs) was also reported (17,23). A statistically significant difference between the study groups was observed in seven articles (15,17,19,21-23,27), with a decreased need for opioids after CPN (six articles) (15,17,19,21-23) and HIFU (one article) (27). However, following the same dynamic as the pain relief, analgesic drugs were less consumed at 4 weeks from study initiation, but the consumption showed an increase at 12 weeks.

P-CPN

Three of the included studies (15,17,22) showed a statistically significant decrease in the use of analgesic drugs, at multiple time points after baseline, for patients who received P-CPN. When directly compared to EUS-CPN, one study highlighted a statistically significant difference in analgesics consumption, at 8 weeks from randomisation, favouring P-CPN (21).

EUS-CPN

Only one study reported a significant change in the analgesic drugs use after EUS-CPN. The results were measured at 12 weeks from baseline and analysed the consumption of morphine, oxycodone and NSAIDs separately (23).

I-CPN

Our study included only one article that compared the effects on analgesic drugs use after I-CPN and P-CPN. There was no statistically significant difference at any of the time points given concerning tramadol intake (14).

HIFU

At 2 and 3 weeks from baseline, one study showed that morphine use significantly decrease after HIFU (27).

AEs

The AEs found most were gastrointestinal (Table 4, Figures S6-S10). Results from individual studies show opioids had a high risk for nausea and vomiting. Haematological AEs, such as anaemia, leukopenia, and thrombocytopenia, were found only after treatments that included the use of chemotherapy (31,32). No important neurological damage was reported for any of the minimally-invasive procedures (Figure S10).

P-CPN

Pooled proportions for diarrhoea were 0.23 (95% CI: 0.14–0.35) with P-CPN (Figure S6). Pooled proportions for nausea and vomiting with P-CPN were 0.18 (95% CI: 0.05–0.47) (Figure S7). For cardiovascular AEs, the pooled proportions were 0.26 (95% CI: 0.14–0.45) for P-CPN (Figure S8). Among these, the most frequently cited were hypotension and orthostatic hypotension. Acute pain related to the procedure was rarely mentioned, with the pooled proportions of 0.12 (95% CI: 0.07–0.21) (Figure S9).

EUS-CPN

With EUS-CPN, pooled proportions for diarrhoea were 0.11 (95% CI: 0.05–0.22) and for cardiovascular AEs were 0.13 (95% CI: 0.04–0.39). For acute pain emerged after procedure, the pooled proportions were 0.23 (95% CI: 0.12–0.39), almost double than the ones of P-CPN.

I-CPN

Compared to a sham intervention, I-CPN showed similar surgery-related risks for cholangitis and biliary leaks (25). The incidence of diarrhoea, nausea and vomiting, and cardiovascular events was lower than with the other types of CPN.

HIFU

The AEs cited in the HIFU groups were in strong connection with the chemotherapy that was simultaneously administered. However, the incidence of the haematological AEs was lower than in the groups randomised to receive only chemotherapy.

RoB assessment

The results of the RoB assessment are presented in Figures S11-S15. The results for pain and QoL were similar, as they are patient-reported outcomes. The domains with the highest risk were the measurement of outcomes and the selection of reported outcomes. Most of the studies did not blind the patients who were the assessors of their symptoms. Moreover, as there is no standard protocol for pain and QoL evaluation, multiple outcome surveys and different effect measures were used. The variety of statistical calculations also raised the RoB in the case of analgesic drug use. Other general factors contributing to the RoB for all outcomes were the lack of a registered protocol, scarce information on the randomisation process, and missing patient data. However, the studies presented no risk concerning deviations from the intended interventions.

Heterogeneity

There was a high I² value among the studies included in the meta-analysis of pain scores [total I²=89% (64%–98%)]. The reason behind this may be the wide variety of interventions, with different approaches, types of imaging guidance, and alcohol dilutions used for the CPN.

Discussion

Key findings

Our analysis showed that based on individual studies’ results EUS-CPN provided slightly better pain scores at 4-, 8-, and 12-week measurements, with fewer associated AEs, significantly improved QoL, and median survival time moderately enhanced compared to opioids. Even if they provided good pain relief, P-CPN and opioids were associated with slightly higher pain increases in time. In the short term, promising results were also achieved with HIFU, which significantly decreased the pain scores compared to opioid treatment at 21 days after randomization (27).

Strengths and limitations

One of the main strengths of our analysis is the large number of diverse procedures included. This is the first systematic review and meta-analysis to compare various intervention types. Moreover, conducting analyses at multiple time points is an important factor in our study, as it allowed us to observe the dynamic of patients’ symptoms in both the short and long term. We included only RCTs, so as to provide the highest quality of evidence, and we followed a strict methodology for performing our research.

We acknowledge the limitations of our research, such as not being able to compare the interventions directly due to the low number of available RCTs. Other limitations included the variety of effect measures, types of statistical analyses, and patient-reported scales used in every study, which, together with the lack of clear criterion standard for pain and QoL assessment, meant that we could not evaluate the results from every available trial altogether. Finally, the control groups distributed to pharmacological treatment were followed only for opioid intake in most cases, while palliative oncological care, other types of painkillers, and illicit drugs were disregarded.

Comparison with similar studies

Our results are consistent with a previous meta-analysis of six RCTs on CPN in unresectable PC (8). When comparing the dynamics of pain after each intervention, a similar trend can be observed in all cases. There was an initial decrease in the VAS score, followed by several escalations of pain values at each reported time point. The same pattern applied to analgesic drugs intake. Another meta-analysis of ten RCTs showed improved results at the 12-week follow-up compared to the earlier time points reported, but the authors chose to analyse P-CPN and EUS-CPN altogether (35). In our case, irrespective of the low number of studies, the most promising interventional technique was shown to be EUS-CPN, with better pain scores at 12 weeks from baseline, improved QoL, and fewer AEs. Its superior safety results from the real-time precise anatomical visualisation that helps avoid retrocrural neurological structures. This can be further improved using colour Doppler to identify interposing vessels (36). However, its best efficacy was reported at 4 weeks, with one study showing that EUS-CPN had similar pain scores as oxycodone and fentanyl at longer follow-ups (24).

Explanations of findings

Our analysis did not find a clinically relevant difference between the pain control provided by P-CPN and opioids at any of the measured time intervals. Regarding techniques, three of the included studies presented a posterior approach [two with fluoroscopy guidance (15,20), one by the means of CT (22)], while in another study, a US-guided anterior approach was chosen (13). Although the most frequent is the posterior, the anterior CPN can lead to a lower risk for spinal cord and kidney injuries (36). Other possibilities have been described. The transaortic approach (16,19) provides a more reliable spread of the neurolytic agent, with the cost of a higher risk for retroperitoneal haemorrhage, while the transdiscal (21) can be used if the patient’s anatomy obstructs the para-aortic approach but has a higher risk for disc trauma (36). We did not have sufficient data to investigate the most efficient approach for P-CPN. Some authors consider that the analgesic results are independent of the technique, but others argue that specific methods should be used based on the characteristics of the tumour. For example, Ischia et al. found no difference in pain control with either bilateral splanchnicectomy, retrocrural neurolysis, or transaortic neurolysis, but they preferred the last as it had the lowest incidence of post-intervention orthostatic hypotension with limited spread of the alcohol in the psoas compartment (16). Süleyman et al. showed that the bilateral splanchnicectomy had better results for tumours localised in the body and tail of the pancreas compared to transaortic P-CPN and suggested an explanation could be the inappropriate distribution of alcohol during CPN due to the extension of the tumour (19).

It is crucial to notice that the number of patients available for analysis rapidly decreased between follow-ups. As patients in palliative care have more pain and shorter life expectancy, the full effect of the performed interventions could not be assessed in all the randomised patients. Only one study presented the observed values compared to the carry-forward ones, with small discrepancies (20). EUS-CPN was associated with slightly better survival rates between all CPN modalities, outperforming P-CPN. It should be mentioned that in the studies that analysed CPN, the population comprised palliative terminally ill patients. This was not the case for the patients from the HIFU studies, with Eastern Cooperative Oncology Group (ECOG) scores of 0–2 (27-29). These baseline differences translate into the significant diversity of our survival analysis results. However, when compared to chemotherapy alone, HIFU seemed to prolong survival. The mechanism behind it has not been explained, but possible theories include its effect on the small vessels around the tumour by blocking vasa vasorum, and on the activation of the immune system (27,29,37). The synergistic effect of HIFU and chemotherapy contributed to the best outcome, with improved QoL and prolonged survival, as it changed cell membrane permeability and enabled easier penetration of the drug in the tumour cells (29).

It is essential to note that the AEs of interventional methods usually lasted for up to a maximum of 1 week from the procedure, as they occurred from the transient loss of the sympathetic tone (38). For example, in the case of CPN and HIFU, diarrhoea was reported to last only 3 days (15-17,22,33). Haemodynamic disturbances were commonly described in the first days after P-CPN and EUS-CPN (14-17,21,22,33), with only one study reporting the requirement of inotropic support (19). Procedure-related transient pain was present in 12% of patients with P-CPN compared to 23% with EUS-CPN, although the latter included one study on EUS-CGN with 44% of the patients in pain after the intervention. Opioids can be associated with long-term AEs throughout the treatment period (39). As the follow-up time increases, the opioid use rate and the total doses consumed are expected to grow (40). Considering the impact of prolonged AEs, minimally invasive techniques seem to have more advantages over opioids. Two studies showed a significantly improved QoL at 12 weeks from P-CPN and EUS-CPN, respectively (15,23). However, we did not find any similar differences at longer follow-ups.

Implications and actions needed

Implementing scientific findings in practice is crucial (41,42). We suggest that minimally-invasive procedures should be considered more often and sooner in the management of patients with painful unresectable PC. Our analysis showed that EUS-CPN can control pain for up to 12 weeks. However, the maximal effect of CPN is achieved when given early in the disease course (43). While EUS-CPN can be performed at the time of the diagnosis and staging, results from RCTs emphasise the need to consider it early in the pain management of unresectable PC (26). Additionally, in a case series of abdominal cancers, early bedside CPN proved to be a satisfying first-line therapy and a good alternative to opioids (44).

Neurolytic procedures have been used for more than a century, gradually evolving and becoming more readily available (43,45). Nowadays, they are believed to be the second most used modality for treating PC pain after opioid intake (36). However, the specific uptake rates of these procedures among PC patients are not well-documented in the available literature. Their world-wide distribution is also proven by the studies included in our analysis, with RCTs conducted in many different countries, from Europe, Asia and the United States of America. The utilization of CPN may be influenced by factors such as patient selection criteria, availability of specialized medical professionals trained in the technique, and the clinical judgment of healthcare providers regarding its appropriateness for individual patients. Moreover, the healthcare policies of each country play a crucial role in the access people have to these techniques (36).

Still, some barriers prevent the implementation of early interventional pain management in patients with PC. These procedures depend greatly, on one hand, on the local expertise and skills, as well as on the available treatment modalities. The lack of awareness among practitioners and the low number of trained specialists detain access to early cancer pain management through interventional procedures. Even though CPN has already been proven safe (36), there are still misconceptions about the risks related to the invasiveness of these techniques. On the other hand, precise timing is also crucial, as malignant infiltration of the plexus may prevent achieving optimal results (45). Often, the focus on systemic therapies such as chemotherapy and radiotherapy overshadows early pain management. This results in a delay for interventional treatments until the pain becomes severe and refractory when the success rate of these procedures is significantly decreased. There can also be cases when other associated pathologies must be initially treated, such as coagulopathies, thrombocytopenia and intraabdominal infections, which are contraindications for CPN (36). Hence, even though already established as a potent intervention with a moderate level of recommendation (2A+), CPN is sometimes missing from the routine clinical protocol of PC pain. The reasons that still need to be addressed are the inadequately trained physicians, the scarce healthcare system resources and the lack of multidisciplinary team discussion (46).

PC is considered to be the most expensive type of cancer in terms of overall costs, combining direct costs for the healthcare system and indirect costs from work absence of the diseased (47). Opioids currently represent the standard cancer pain treatment. However, there is a marked inter-individual variability in the response to them, with many patients reporting insufficient pain control or diverse side effects (39). In our analysis, the most common AEs connected to opioid intake were nausea, vomiting, and constipation. Treating these side effects increases the financial cost of the medical treatment. With a total of almost 7,000 US$ during best supportive care, PC has the highest financial burden of both inpatient and outpatient expenses compared to other types of cancers (48). Our results showed that EUS-CPN and I-CPN had the lowest rate of AEs, but the EUS-guided approach had the advantage of being less invasive. Moreover, the studies included in our analysis showed a significant decrease in the number of analgesic drugs used after procedures such as P-CPN and EUS-CPN, with better outcomes after the ladder. This translates to reduced long-term medication costs and a lower burden of opioid-induced side effects. From a financial perspective, EUS-CPN appears superior again, with a total price lower than CT-guided P-CPN (1,100 versus 1,400 US$) (49). Therefore, even if initial procedure costs seem higher, performing EUS-CPN could reduce the overall costs of unresectable PC pain management. Regarding cost-effectiveness, interventional procedures seem to be superior, with durable decreased medication use and improved patient outcomes. However, cost-analysis data in this clinical setting remains limited.

There are more options for pain treatment not yet included in RCTs, such as microwave ablation, cryoablation, irreversible electroporation, or stereotactic body radiation therapy. As technical failures can occur, whether these interventions should be repeated or combined is still unknown (16). One prospective study on 156 patients showed that multiple repetitions of EUS-CPN provided superior pain control compared to the initial treatment (50). The analgesic effect of CPN was seen to diminish in time, with pain scores increasing again after 3 months from intervention. However, since these procedures are usually offered to late-stage patients and only almost a third of them have the chance for repeated CPN, there is no high level of evidence yet to support this practice (45). Hence, more and stricter RCTs must be conducted in this field, with rigorous methodologies comparing interventional procedures to one another, and multiple repetitions to a one-time intervention.

To find the most effective management, patient-reported outcomes (such as pain and QoL assessments) need standardisation. While one meta-analysis tried to identify the minimal clinically significant differences in chronic pain, it was focused on non-cancer pain (51). Future studies are needed to conclude what clinically relevant change is from the baseline, as there is no clear definition in use.

Conclusions

Minimally invasive techniques are potent tools for providing pain reduction with fewer associated adverse effects and reduced opioid use for unresectable PC patients. Future RCTs are needed to define the characteristics of the most appropriate method for patient-tailored management.

Acknowledgments

We wish to acknowledge the assistance provided by Dr. Abdel-Ghaffar’s study group (Department of Anesthesia, Intensive Care and Pain Management, Faculty of Medicine, Suez Canal University) in sending us additional data for analysis.

Footnote

Reporting Checklist: The authors have completed the PRISMA reporting checklist. Available at https://tgh.amegroups.com/article/view/10.21037/tgh-24-141/rc

Peer Review File: Available at https://tgh.amegroups.com/article/view/10.21037/tgh-24-141/prf

Funding: This study was funded by the Centre for Translational Medicine, Semmelweis University. Additional funding was provided by the New National Excellence Program of the Ministry for Innovation and Technology from the source of the National Research, Development, and Innovation Fund (No. ÚNKP-23-3-II-PTE-1996 to BT). Sponsors had no role in the design, data collection, analysis, interpretation, and manuscript preparation.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tgh.amegroups.com/article/view/10.21037/tgh-24-141/coif). All authors report that funding was provided by the Centre for Translational Medicine, Semmelweis University. B.T. reports additional funding provided by the New National Excellence Program of the Ministry for Innovation and Technology from the source of the National Research, Development, and Innovation Fund (No. ÚNKP-23-3-II-PTE-1996). The authors have no other conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Bray F, Laversanne M, Sung H, et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2024;74:229-63. [Crossref] [PubMed]
2. Cancer Stat Facts: Pancreatic Cancer. Available online: https://seer.cancer.gov/statfacts/html/pancreas.html
3. Rawla P, Sunkara T, Gaduputi V. Epidemiology of Pancreatic Cancer: Global Trends, Etiology and Risk Factors. World J Oncol 2019;10:10-27. [Crossref] [PubMed]
4. Westermann A, Matrisian LM, Rahib L. The need for improvement in the management of fatigue, depression and pain in pancreatic cancer. J Clin Oncol 2019;37:429.
5. Hameed M, Hameed H, Erdek M. Pain management in pancreatic cancer. Cancers (Basel) 2010;3:43-60. [Crossref] [PubMed]
6. Lohse I, Brothers SP. Pathogenesis and Treatment of Pancreatic Cancer Related Pain. Anticancer Res 2020;40:1789-96. [Crossref] [PubMed]
7. World Health O. Cancer pain relief. Geneva: World Health Organization; 1986.
8. Arcidiacono PG, Calori G, Carrara S, et al. Celiac plexus block for pancreatic cancer pain in adults. Cochrane Database Syst Rev 2011;2011:CD007519. [Crossref] [PubMed]
9. Asif AA, Walayat SK, Bechtold ML, et al. EUS-guided celiac plexus neurolysis for pain in pancreatic cancer patients - a meta-analysis and systematic review. J Community Hosp Intern Med Perspect 2021;11:536-42. [Crossref] [PubMed]
10. Fergadi MP, Magouliotis DE, Rountas C, et al. A meta-analysis evaluating the role of high-intensity focused ultrasound (HIFU) as a fourth treatment modality for patients with locally advanced pancreatic cancer. Abdom Radiol (NY) 2022;47:254-64. [Crossref] [PubMed]
11. Page MJ, McKenzie JE, Bossuyt PM, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 2021;372: [Crossref] [PubMed]
12. Cochrane Handbook for Systematic Reviews of Interventions. Cochrane; 2023. Available online: www.training.cochrane.org/handbook
13. Abdel-Ghaffar ME, Ismail SA, Ismail RA, et al. Comparison Between Two Volumes of 70% Alcohol in Single Injection Ultrasound-Guided Celiac Plexus Neurolysis: A Randomized Controlled Trial. Pain Physician 2022;25:293-303.
14. Abdelbaser I, Shams T, El-Giedy AA, et al. Direct intraoperative versus percutaneous computed tomographyguided celiac plexus neurolysis in non-resectable pancreatic cancer: A randomized, controlled, non-inferiority study. Rev Esp Anestesiol Reanim (Engl Ed) 2022;69:71-8. [Crossref] [PubMed]
15. Dong D, Zhao M, Zhang J, et al. Neurolytic Splanchnic Nerve Block and Pain Relief, Survival, and Quality of Life in Unresectable Pancreatic Cancer: A Randomized Controlled Trial. Anesthesiology 2021;135:686-98. [Crossref] [PubMed]
16. Ischia S, Ischia A, Polati E, et al. Three posterior percutaneous celiac plexus block techniques. A prospective, randomized study in 61 patients with pancreatic cancer pain. Anesthesiology 1992;76:534-40. [Crossref] [PubMed]
17. Polati E, Finco G, Gottin L, et al. Prospective randomized double-blind trial of neurolytic coeliac plexus block in patients with pancreatic cancer. Br J Surg 1998;85:199-201. [Crossref] [PubMed]
18. Saeed R, Qamar I, Kakepoto I, Shahzad M, Bangush T, Javed A. Comparison of Effectiveness and Complications with Two Strengths of Dilution of Alcohol in Celiac Plexus Block in the Management of Relieving Pain. Pakistan Journal of Medical & Health Sciences 2019;13:1015-7.
19. Süleyman Ozyalçin N, Talu GK, Camlica H, et al. Efficacy of coeliac plexus and splanchnic nerve blockades in body and tail located pancreatic cancer pain. Eur J Pain 2004;8:539-45. [Crossref] [PubMed]
20. Wong GY, Schroeder DR, Carns PE, et al. Effect of neurolytic celiac plexus block on pain relief, quality of life, and survival in patients with unresectable pancreatic cancer: a randomized controlled trial. JAMA 2004;291:1092-9. [Crossref] [PubMed]
21. Yoon WJ, Oh Y, Yoo C, et al. EUS-Guided Versus Percutaneous Celiac Neurolysis for the Management of Intractable Pain Due to Unresectable Pancreatic Cancer: A Randomized Clinical Trial. J Clin Med 2020;9:1666. [Crossref] [PubMed]
22. Zhang CL, Zhang TJ, Guo YN, et al. Effect of neurolytic celiac plexus block guided by computerized tomography on pancreatic cancer pain. Dig Dis Sci 2008;53:856-60. [Crossref] [PubMed]
23. Gao L, Yang YJ, Xu HY, et al. A randomized clinical trial of nerve block to manage end-stage pancreatic cancerous pain. Tumour Biol 2014;35:2297-301. [Crossref] [PubMed]
24. Kanno Y, Koshita S, Masu K, et al. Efficacy of EUS-guided celiac plexus neurolysis compared with medication alone for unresectable pancreatic cancer in the oxycodone/fentanyl era: a prospective randomized control study. Gastrointest Endosc 2020;92:120-30. [Crossref] [PubMed]
25. Lillemoe KD, Cameron JL, Kaufman HS, et al. Chemical splanchnicectomy in patients with unresectable pancreatic cancer. A prospective randomized trial. Ann Surg 1993;217:447-55; discussion 456-7. [Crossref] [PubMed]
26. Wyse JM, Carone M, Paquin SC, et al. Randomized, double-blind, controlled trial of early endoscopic ultrasound-guided celiac plexus neurolysis to prevent pain progression in patients with newly diagnosed, painful, inoperable pancreatic cancer. J Clin Oncol 2011;29:3541-6. [Crossref] [PubMed]
27. Li L, Li HL, Wang XH, et al. Randomized controlled trial of follow-up in pancreatic cancer patients treated by intermediate frequency ultrasound combined with opioid analgesic. Biomedical Research 2017;28:3545-50.
28. Li X, Wang K, Zheng L, et al. Retrospective analysis of high intensity focused ultrasound combined with S-1 in the treatment of metastatic pancreatic cancer after failure of gemcitabine. Am J Cancer Res 2016;6:84-90.
29. Lv W, Yan T, Wang G, et al. High-intensity focused ultrasound therapy in combination with gemcitabine for unresectable pancreatic carcinoma. Ther Clin Risk Manag 2016;12:687-91. [Crossref] [PubMed]
30. Bang JY, Sutton B, Hawes RH, et al. EUS-guided celiac ganglion radiofrequency ablation versus celiac plexus neurolysis for palliation of pain in pancreatic cancer: a randomized controlled trial (with videos). Gastrointest Endosc 2019;89:58-66.e3. [Crossref] [PubMed]
31. Han GH, Yin ZX, Meng XJ, et al. Prospective randomized clinical trial of two drug delivery pathway in the treatment of inoperable advanced pancreatic carcinoma. Chin J Dig Dis 2006;7:45-8. [Crossref] [PubMed]
32. Shinchi H, Takao S, Noma H, et al. Length and quality of survival after external-beam radiotherapy with concurrent continuous 5-fluorouracil infusion for locally unresectable pancreatic cancer. Int J Radiat Oncol Biol Phys 2002;53:146-50. [Crossref] [PubMed]
33. Levy MJ, Gleeson FC, Topazian MD, et al. Combined Celiac Ganglia and Plexus Neurolysis Shortens Survival, Without Benefit, vs Plexus Neurolysis Alone. Clin Gastroenterol Hepatol 2019;17:728-738.e9. [Crossref] [PubMed]
34. Jensen MP, Chen C, Brugger AM. Interpretation of visual analog scale ratings and change scores: a reanalysis of two clinical trials of postoperative pain. J Pain 2003;4:407-14. [Crossref] [PubMed]
35. Okita M, Otani K, Gibo N, et al. Systematic review and meta-analysis of celiac plexus neurolysis for abdominal pain associated with unresectable pancreatic cancer. Pain Pract 2022;22:652-61. [Crossref] [PubMed]
36. Urits I, Jones MR, Orhurhu V, et al. A Comprehensive Review of the Celiac Plexus Block for the Management of Chronic Abdominal Pain. Curr Pain Headache Rep 2020;24:42. [Crossref] [PubMed]
37. Dubinsky TJ, Cuevas C, Dighe MK, et al. High-intensity focused ultrasound: current potential and oncologic applications. AJR Am J Roentgenol 2008;190:191-9. [Crossref] [PubMed]
38. Cornman-Homonoff J, Holzwanger DJ, Lee KS, et al. Celiac Plexus Block and Neurolysis in the Management of Chronic Upper Abdominal Pain. Semin Intervent Radiol 2017;34:376-86. [Crossref] [PubMed]
39. O’Brien T, Christrup LL, Drewes AM, et al. European Pain Federation position paper on appropriate opioid use in chronic pain management. Eur J Pain 2017;21:3-19. [Crossref] [PubMed]
40. Lu Z, Zhang N, Giordano SH, et al. Opioid use and associated factors among pancreatic cancer patients diagnosed between 2007 and 2015. Cancer Med 2022;11:2296-307. [Crossref] [PubMed]
41. Hegyi P, Erőss B, Izbéki F, et al. Accelerating the translational medicine cycle: the Academia Europaea pilot. Nat Med 2021;27:1317-9. [Crossref] [PubMed]
42. Hegyi P, Petersen OH, Holgate S, et al. Academia Europaea Position Paper on Translational Medicine: The Cycle Model for Translating Scientific Results into Community Benefits. J Clin Med 2020;9:1532. [Crossref] [PubMed]
43. Vig S, Bhan S, Bhatnagar S. Celiac Plexus Block - An Old Technique with New Developments. Pain Physician 2021;24:379-98.
44. Bhatnagar S, Khanna S, Roshni S, et al. Early ultrasound-guided neurolysis for pain management in gastrointestinal and pelvic malignancies: an observational study in a tertiary care center of urban India. Pain Pract 2012;12:23-32. [Crossref] [PubMed]
45. Drewes AM, Campbell CM, Ceyhan GO, et al. Pain in pancreatic ductal adenocarcinoma: A multidisciplinary, International guideline for optimized management. Pancreatology 2018;18:446-57. [Crossref] [PubMed]
46. Hochberg U, Ingelmo P, Solé E, et al. Early Interventional Treatments for Patients with Cancer Pain: A Narrative Review. J Pain Res 2023;16:1663-71. [Crossref] [PubMed]
47. Cipora E, Partyka O, Pajewska M, et al. Treatment Costs and Social Burden of Pancreatic Cancer. Cancers (Basel) 2023;15:1911. [Crossref] [PubMed]
48. Kang DW, Shim YB, Lee EK, et al. Healthcare resource utilization and medical costs in patients with terminal cancer during best supportive care. PLoS One 2022;17:e0269565. [Crossref] [PubMed]
49. Gress F, Schmitt C, Sherman S, et al. A prospective randomized comparison of endoscopic ultrasound- and computed tomography-guided celiac plexus block for managing chronic pancreatitis pain. Am J Gastroenterol 1999;94:900-5. [Crossref] [PubMed]
50. Facciorusso A, Del Prete V, Antonino M, et al. Response to repeat echoendoscopic celiac plexus neurolysis in pancreatic cancer patients: A machine learning approach. Pancreatology 2019;19:866-72. [Crossref] [PubMed]
51. Olsen MF, Bjerre E, Hansen MD, et al. Minimum clinically important differences in chronic pain vary considerably by baseline pain and methodological factors: systematic review of empirical studies. J Clin Epidemiol 2018;101:87-106.e2. [Crossref] [PubMed]