Article Text
Abstract
Background Opioids such as morphine are widely used for the management of pain associated with acute pancreatitis. Interestingly, opioids are also known to affect the immune system and modulate inflammatory pathways in non-pancreatic diseases. However, the impact of morphine on the progression of acute pancreatitis has never been evaluated. In the current study, we evaluated the impact of morphine on the progression and severity of acute pancreatitis.
Methods Effect of morphine treatment on acute pancreatitis in caerulein, L-arginine and ethanol–palmitoleic acid models was evaluated after induction of the disease. Inflammatory response, gut permeability and bacterial translocation were compared. Experiments were repeated in mu (µ) opioid receptor knockout mice (MORKO) and in wild-type mice in the presence of opioid receptor antagonist naltrexone to evaluate the role of µ-opioid receptors in morphine’s effect on acute pancreatitis. Effect of morphine treatment on pathways activated during pancreatic regeneration like sonic Hedgehog and activation of embryonic transcription factors like pdx-1 and ptf-1 were measured by immunofluorescence and quantitative PCR.
Results Histological data show that treatment with morphine after induction of acute pancreatitis exacerbates the disease with increased pancreatic neutrophilic infiltration and necrosis in all three models of acute pancreatitis. Morphine also exacerbated acute pancreatitis-induced gut permeabilisation and bacteraemia. These effects were antagonised in the MORKO mice or in the presence of naltrexone suggesting that morphine’s effect on severity of acute pancreatitis are mediated through the µ-opioid receptors. Morphine treatment delayed macrophage infiltration, sonic Hedgehog pathway activation and expression of pdx-1 and ptf-1.
Conclusion Morphine treatment worsens the severity of acute pancreatitis and delays resolution and regeneration. Considering our results, the safety of morphine for analgesia during acute pancreatitis should be re-evaluated in future human studies.
- Morphine
- pancreatitis
- regeneration
- hedgehog
- PDX-1
- PTF-1
- macrophages
- severity
Statistics from Altmetric.com
Significance of this study
What is already known on this subject?
Opioids are widely used for the management of pain in acute pancreatitis (AP) and are generally considered safe.
However, due to a lack of other potent options for analgesia, there is insufficient research on the possible adverse effects of opioids on the disease process itself and it has not been systematically evaluated in pancreatitis.
What are the new findings?
Our observations in experimental mouse models of AP suggest that morphine leads to (1) worsening of AP; (2) augmentation of intestinal permeability, increased bacterial translocation and dissemination into systemic organs; (3) persistence of inflammation and injury in the pancreas and lungs; and (4) delay in regenerative response of the pancreas.
The adverse effect of morphine is mediated through µ-opioid receptors.
How it might impact on clinical practice in the foreseeable future?
This study will stimulate judicial use of morphine, and by extrapolation, its derivatives for the control of pain during AP.
This study will encourage use of alternate non-opioid strategies for control of pain.
This study will inspire comparative clinical evaluation of the safety of various opioids as well as experimental research to modify opioids to attenuate their proinflammatory side effects.
Introduction
Pancreatitis is the most common cause of hospitalisation for gastrointestinal diseases and has significant morbidity and mortality. While most attacks of acute pancreatitis (AP) are mild and self-resolving, patients with severe AP experience high mortality and morbidity owing to early local complications (eg, pancreatic necrosis) and systemic organ failure.1 2 Severe abdominal pain is the hallmark symptom of patients suffering from AP3–5 and is increasingly recognised as a diagnostic and prognostic factor in AP.6–10 Regardless of the cause or severity of the disease, supportive treatment in the form of pain control with analgesics, intravenous fluids and correction of electrolyte and metabolic abnormalities constitutes the backbone of AP management.11 12 Opioid analgesics (morphine, hydromorphone or fentanyl) are commonly used for treatment of AP associated pain.13 Most of the currently available opioid analgesics function, at least in part, via binding to the mu (µ) opioid receptors (MOR).14 Since AP is a painful disease, frequently requiring high doses of narcotic analgesics, the choice of opioids for analgesia in AP has been a topic of much debate over the past decades. Traditionally, morphine was avoided due to the concern that it leads to sphincter of Oddi dysfunction, thus leading to worsening of the disease, and meperidine was preferred as the drug of choice. Although numerous studies have15–17 addressed this issue, the available literature and data have failed to draw a conclusion due to the absence of a head-to-head comparison of the different narcotics. Moreover, there seems to be no data linking this physiological mechanism with the clinical outcome of worsening pain or disease severity.18 19 Furthermore, it is unclear if opioid treatment as such, when compared with non-opioid analgesics, leads to worsening of the disease.
Intriguingly, literature in other diseases suggests that morphine can modulate the immune response and that it disrupts the intestinal barrier function causing increased bacterial translocation predisposing the patient to infections and enhanced morbidity.20 Furthermore, increased intestinal permeability due to breakdown of gut mucosal barrier is also observed in AP and has been associated with increased severity of the disease.21–24 The aim of this study was to evaluate the effect of morphine administration on the severity of AP, resolution of injury, regeneration of the pancreas and systemic injury. In the current study, we evaluated the impact of morphine treatment on various parameters of local and systemic injury during AP using three different mouse models of AP. Interestingly, our study suggests that morphine increases the severity of AP and delays pancreatic regeneration after AP. This work has clinical significance as it suggests that use of opioids for pain control may worsen outcomes in pancreatitis. If these results bear out in future human studies, alternate strategies for pain management may need to be explored.
Materials and methods
Experimental animals
Wild-type (WT) mice (C57BL/6; 25–30 g) were purchased from Charles River Laboratories (Wilmington, Massachusetts, USA). Mu opioid receptor knockout (MORKO) mice (C57BL/6129/Ola genetic background) have been described before.25 Animals were housed and maintained as described previously. All animal experimental protocols were approved by Institutional Animal Care and Use Committee at the University of Minnesota.
Experimental setup
AP was induced using either caerulein, L-arginine or ethanol–palmitoleic acid (POA) as described in the online supplementary section. After induction of AP, either placebo or morphine subcutaneous pellet was implanted as described previously.26 Animals were euthanised at different time points. Severity of AP and regeneration were compared with placebo implanted group.
Supplementary materials
Details of methods for tissue collection, intestinal permeability evaluation, 5-Bromo-2’deoxyUridineincorporation for evaluation of regeneration, myeloperoxidase (MPO) assay, quantification of necrosis, immunohistochemistry, immunofluorescence (IF), RNA isolation and quantitative PCR (qPCR) and materials are described in the online supplementary section.
Statistical analysis
Results are expressed as means±SEM. Unpaired Student’s t-test, Mann-Whitney test or analysis of variance with Tukey-Kramer post hoc test (>2 groups) were used. α=0.05 and two-tailed p values are reported.
Results
Morphine causes persistence of inflammation and necrosis during AP
The effect of morphine treatment on pancreatic necrosis as well as on local and systemic inflammation was evaluated in caerulein model of AP at 48, 72 and 120 hours time points from start of the experiment (see online supplementary figure 1A). In mice with caerulein AP, without morphine treatment (placebo), significant morphological damage in the form of necrosis, inflammatory infiltrate and pancreatic oedema was observed at 48 hours (figure 1A) and, as expected, restoration to normal architecture was observed at 120 hours after induction of AP. However, in the morphine treatment group, greater amount of pancreatic oedema, acinar cell necrosis and inflammation was observed at 48 hours when compared with placebo group and the damage persisted until 120 hours with lesser morphological recovery (figure 1A). Blinded quantification at 48, 72 and 120 hours (figure 1B) revealed significant increase in pancreatic necrosis in the group receiving morphine as compared with placebo.
Caerulein AP led to marked increase in pancreatic (figure 1C) and lung (figure 1D) neutrophil sequestration at 48 hours, as measured by MPO activity, with gradual decrease towards baseline over time. However, with morphine treatment in caerulein AP, an increased neutrophilic sequestration into the pancreas (figure 1C) was observed at both the 48 and 72 hours time point compared with the placebo group. Similarly, in this group lung neutrophil sequestration at 48 hours was also significantly higher when compared with mice with AP but without morphine treatment (figure 1D). In the presence of morphine, there was also a trend towards increased neutrophil sequestration into the lungs at 72 and 120 hours; however, it did not reach statistical significance (figure 1D). Real-time PCR analysis of pancreatic tissue revealed a rapid increase in messenger RNA (mRNA) expression of inflammatory cytokines (interleukin-6 and tumour necrosis factor-alpha) and chemokine (CXCL2) at 48 hours in the presence of morphine, which stayed high at 120 hours, compared with the AP group without morphine (figure 1E).
Morphine delays macrophage infiltration into the pancreas
We evaluated the impact of morphine treatment on macrophage recruitment after caerulein AP by using macrophage marker F4/80 at different time points. In the pancreatitis alone group, macrophage infiltration increased significantly 48–72 hours after injury and then decreased around 120 hours postinduction (figure 1F). Interestingly, morphine treatment resulted in a significant delay in macrophage infiltration during AP, with decreased macrophage infiltration into the pancreas at 48 hours and persistence of infiltration at 120 hours when compared with AP without morphine treatment.
Morphine treatment accentuates intestinal permeability during AP
Morphine and severe AP are known to independently compromise gut barrier function. Since intestinal permeability with bacterial translocation has been demonstrated to propagate systemic injury during AP, we evaluated whether combination of AP with morphine treatment will further augment intestinal permeability and its consequences. As seen in figure 2A, both morphine treatment alone and caerulein pancreatitis lead to increased gut permeability at 48 hours, as measured by increased fluorescein isothiocyanate-conjugated dextran translocation from gut lumen to blood. Intriguingly, mice with AP that were treated with morphine had significantly higher intestinal permeability compared with either morphine treatment or AP alone (figure 2A). Bacterial translocation across the gut barrier is a direct consequence of increased intestinal permeability. Bacterial dissemination into various organs was evaluated at the 48 hours time point by culturing the organ homogenates. Culture of homogenates from liver and mesenteric lymph nodes revealed a significant increase in colony-forming units (CFUs) in the AP group that were morphine treated compared with AP or morphine treatment alone (figure 2B and C). We observed a trend towards an increase in the CFU in the lungs in mice with AP treated with morphine as well (figure 2D); however, this did not reach statistical significance. This data clearly demonstrated a significant increase in bacterial translocation at 48 hours following morphine treatment in AP model.
Morphine treatment increases the severity of AP via MOR
Next, we evaluated the role of MOR in morphine-induced worsening of AP by evaluating the effect of morphine on pancreatic necrosis and neutrophil infiltration in MORKO mice. As seen in figure 3A, caerulein treatment in MORKO mice led to marked pancreatic necrosis at 48 hours. Intriguingly, in contrast to our observation where morphine treatment led to worsening of pancreatic injury in WT mice, morphine treatment of MORKO mice with caerulein AP did not increase pancreatic injury as measured by quantification of necrosis and pancreatic MPO (figure 3A and B). To further confirm the role of MORs, we evaluated whether this phenomenon can be reversed with the use of opioid receptor antagonist naltrexone. As seen in figure 3C and D coadministration of naltrexone (40 mg/kg) and morphine after induction of caerulein AP in WT mice reduced pancreatic necrosis and neutrophil sequestration at 48 hours as compared with morphine treatment alone.
Morphine also increases pancreatic inflammation and injury in L-arginine and ethanol–POA models of AP
To rule out whether morphine induced worsening of AP is model specific, we evaluated the effect of morphine on L-arginine and ethanol–POA induced pancreatitis (see online supplementary figures 2A and 3A). As seen in representative histology pictures (H&E staining) as well as quantification of necrosis (figure 4A and B), morphine treatment of mice with L-arginine (figure 4A) and ethanol–POA (figure 4B) AP led to significant increase (p<0.05) in pancreatic necrosis. Furthermore, morphine treatment of mice with L-arginine and ethanol–POA induced AP led to a statistically significant increase (p<0.05) in MPO activity in the pancreas (figure 4C and D) as compared with mice without morphine treatment. However, only a statistically non-significant trend towards increase in the lung MPO activity was observed (see online supplementary figure 4A and B).
Morphine treatment delays pancreatic epithelial regeneration in response to injury
To elucidate whether morphine treatment during AP leads to delayed recovery by affecting regeneration, we evaluated the impact of morphine on pathways activated during pancreatic regeneration. One such pathway is the Hedgehog pathway (HH). Sonic Hedgehog (Shh) is the major ligand for the HH pathway in vertebrates. As seen in representative figures as well as quantification (figure 5A), induction of caerulein pancreatitis led to increase in Shh staining at 48 hours, which peaked at 72 hours and then decreased back at 120 hours. With addition of morphine, we observed a significant delay in the upregulation of the HH ligand Shh, as evidenced by decreased staining (compared with mice with caerulein pancreatitis without morphine treatment) at 48 hours, and then a persistence of staining at 120 hours. mRNA transcript levels of HH pathway receptor patched (PTCH) and one of the major transcription factors Gli were also studied in pancreatic tissue by real-time PCR at different time points. Both these components are known to be upregulated early during regeneration. Our results show (figure 5B) that in the caerulein AP group (without morphine treatment) there was an early increase in both PTCH and Gli mRNA expression, which returned to normal by 120 hours. However, morphine treatment led to increased expression of both PTCH and Gli at the 120 hours. These results suggest a significant delay or frame shift in the HH pathway response to pancreatic injury in the presence of morphine.
Morphine delays upregulation of embryonic markers during pancreatic regeneration
Progenitor markers of pancreatic embryogenesis are overexpressed during regeneration after injury. The expression levels of these markers (pdx-1 and ptf1) were used to follow the effect of morphine on the pancreatic regenerative response. Even though pdx-1 expression (as measured by IF) in mice with AP, which were treated with morphine, increased with time (figure 5C), it was significantly less than that seen with mice with AP that were treated with placebo. Suggesting that morphine treatment leads to delay in upregulation of pdx-1 expression in mice with AP. We also measured the pdx-1 expression at mRNA level and as seen in figure 5D, morphine treatment of mice with AP showed a trend towards decreased level of pdx-1 mRNA at 48 hours, when compared with mice with AP treated with placebo. However, the pdx-1 levels were significantly more at 120 hours in morphine-treated mice compared with placebo suggesting that morphine treatment leads to delayed but persistent increase in pdx-1 mRNA expression. Ptf-1, another transcription factor, is also known to be upregulated early following pancreatic injury. As observed with pdx-1 expression, there seems to be a clear delay in Ptf-1 mRNA expression upregulation in the pancreas of mice treated with morphine after injury as compared with placebo, (figure 5E) (p<0.0005). This again suggests that morphine leads to delay in the upregulation of embryonic transcription factors in an injured pancreas.
Morphine leads to a delay in the proliferative response in the pancreas after injury
To study proliferation during pancreatic regeneration, tissue sections were stained with Ki-67 (figure 6A). When compared with mice with caerulein pancreatitis but no morphine treatment, there was a significant (p<0.05) delay in upregulation of Ki-67 staining in the morphine treated group. This was confirmed in a dynamic system by analysing BrdU incorporation in pancreatic acini after induction of acute pancreatic injury with and without morphine. As expected, there was a significant decrease in nuclei that stained positive for BrdU incorporation in the group treated with morphine as compared with placebo at the 48 hours time point (figure 6B). Both these sets of data suggest a clear disruption or delay in the regenerative response of the injured pancreas secondary to the presence of morphine.
Discussion
Morphine has been shown to increase intestinal permeability and bacterial translocation by increasing intestinal transit time27 and independently by compromising intestinal barrier function through a specific toll-like receptor (TLR) mediated mechanism.20 Loss of gut barrier function and integrity in AP has been demonstrated in several studies. This has been implicated as one of the key events responsible for the progression from a single-organ injury to multiple-organ failure.21 28–30 Intriguingly, our data suggest that morphine treatment of mice with AP leads to a significantly increased disruption in the gut barrier integrity, which is worse than that observed in AP or with morphine treatment alone. This breakdown and failure is associated with an increased translocation of bacteria, inflammatory and toxic products that can be directly responsible for an enhanced systemic inflammatory response and sepsis. We believe that this can cause a higher number of intestinal microbes gaining access into the systemic circulation and dissemination into distant organs, as evidenced by the increased number of CFU in systemic tissues. In turn, this can contribute to more local and systemic inflammation, thus worsening the severity of the disease.
To mechanistically dissect whether morphine-induced worsening of AP is mediated by specific effects of the drug, we targeted the MOR, a G-protein coupled receptor.31 In the absence of the MOR, morphine is unable to further increase the severity of AP, implicating a direct role of the drug acting via these receptors. It should be noted the MORKO mice used in the current study are in 129S mixed background. Significant variation in severity of caerulein-induced AP has been observed in different strains of mice, with C57BL/6 strain developing relatively mild AP.32 Thus, we observe more inflammation with caerulein-induced AP in these mice when compared with that in C57BL6 mice. However, morphine treatment does not further worsen inflammation in the MORKO animals suggesting the role of the MOR in morphine-mediated effects. Most commonly used opioids in the clinical setting are at least in part agonists at the MOR, which is the major receptor for their analgesic effects. We further confirmed our hypothesis by using an opioid receptor antagonist naltrexone, which has a high affinity for the MOR.
Neutrophil and monocyte infiltration into the pancreas is an early event during AP33 contributing to local injury and disease progression.34 35 This initial response is followed by the migration of other inflammatory cells, importantly the macrophages, a key inflammatory highly plastic cell type. The initial population is the proinflammatory (M1 or classical activation) followed by the reparative (M2 or alternative activation) population, which have an important role in resolving the inflammatory response and are essential in tissue restoration.36–38 Attenuation of macrophage chemotaxis, recruitment and function such as phagocytosis and superoxide production25 39 40 by morphine has been reported. Our data are consistent with these previously reported studies where we see a significant delay in macrophage recruitment into the pancreatic tissue in the mice receiving morphine. However, once recruited, we observe a persistence in inflammation and a delay in its resolution. This observation is consistent with our previous observation of morphine’s abrogation of endotoxin/ lipopolysaccharide (LPS) tolerance.41 LPS tolerance is a phenomenon where persistent presence of LPS creates a proinflammatory state through TLRs, followed by ‘tolerance’ phenomenon where over time the inflammation is suppressed as an innate response to prevent hyperinflammation.42 We recently showed that the presence of morphine in the systemic circulation reverses this LPS tolerance via a microRNA (miRNA)mediated mechanism both in vivo and in vitro monocyte and macrophage cultures, in a time-dependent manner leading to a persistently severe inflammatory response and septicaemia.41 In the context of our experimental pancreatitis setup, with a significant increase in intestinal permeability and hence endotoxaemia, this could definitely be a phenomenon adding to the complexity.
Caerulein-induced pancreatitis in rodents has been recognised to be a fully reversible process.43 44 During regeneration, these studies have elucidated that the pancreatic progenitor markers and signalling pathways instrumental in the embryonic development and differentiation of the pancreas are re-expressed or activated as the pancreas recovers its normal morphology and function.43–45 We observed a delay in restoration to normal architecture in the pancreatitis groups treated with morphine as compared with control. To get a holistic picture, we also evaluated the effect of morphine on this regenerative response using a few of these known embryonic markers and the expression levels of the different components of the HH pathway as a tool to follow the kinetics of tissue restoration. Interestingly, our data clearly suggest that in the presence of morphine, there is a significant delay in the expression levels of the transcription factors (pdx-1 and ptf-1) and the components of the HH pathway (Shh, Gli and PTCH) known to be upregulated and modulating pancreatic epithelial regeneration at various time points. Hence, we show that the use of morphine worsens the severity of AP in three different models of murine AP. It seems to do this by adversely affecting the gut barrier function modulating the immune response and causing a delay in the regenerative response secondary to injury.
Conclusion
Opioids are widely used as analgesics in acute, recurrent and chronic pancreatitis. Our experimental data sheds more light on the effect of morphine on the disease process of pancreatitis, both in terms of the inflammatory and the regenerative response secondary to acute injury as represented schematically in figure 7. We believe this study has important clinical relevance considering the widespread use of opioids worldwide for the management of this disease. These experiments clearly show that opioids have the potential to worsen the morbidity and mortality in the setting of AP. In absence of other potent analgesics, our results suggest extreme caution and judicious use of opioids in patients presenting with AP.
Acknowledgments
This work was initiated at Department of Surgery, University of Minnesota
References
Footnotes
Contributors AKS and SR conceptualised the study. RKD, VD and SuB designed the experiments. UB, RD, HC, JG, AD, AS, ZY, BG and JM conducted the experiments. AKS, RKD, VD, SaB and SuB interpreted the results. AKS, RKD, VD, SR and SuB drafted the manuscript. AKS, RKD, VD, SR and SuB revised the manuscript. AKS is overall contents guarantor.
Funding This work was supported, in whole or in part, by National Institutes of Health DK058694, DK093047, and DK092145 (to AKS), DA034582, K05DA033881 (to SR) and intramural support from Department of Surgery, University of Minnesota.
Competing interests AKS is the co-founder and the Chief Scientific Officer of Minneamrita Therapeutics LLC. SB is a consultant with Minneamrita Therapeutics LLC. AKS is consultant for Sun BioPharma. These relationships have been reviewed and managed by the University of Minnesota and University of Miami in accordance with their conflict of interest policies. Other authors have nothing to disclose.
Provenance and peer review Not commissioned; externally peer reviewed.