Pancreatitis is inflammation of the pancreas; a gland in the central upper abdomen that secretes a number of enzymes involved in digestion and the metabolism of glucose, such as insulin and glucagon. When the pancreas becomes inflamed the digestive enzymes leak out in to the surrounding tissues and start to “digest” these tissues causing further inflammation.
Patients with pancreatitis present with central upper abdominal pain radiating to the back. The diagnosis is confirmed by finding high levels of pancreatic enzymes in the patient’s blood e.g. amylase or lipase, or a CT abdomen that shows an inflamed pancreas.
The body’s response to the pancreatitis can be extreme with a severe systemic inflammatory response which mimics sepsis:
- Fever
- Tachycardia, hypotension, tachypnoea,
- Increased White Blood Cells (WBC) in blood
- Raised C Reactive Protein (CRP)
- Shock and multi-organ failure
But this is a severe response to inflammation, not infection!
What causes pancreatitis?
The two most common causes of pancreatitis in the UK are and gallstones (50%) and excess alcohol consumption (30%). Other causes include: drugs (e.g. steroids), viruses (e.g. mumps), autoimmune damage, post-procedure (e.g. endoscopic retrograde cholangiopancreatography ERCP), and even apparently scorpion stings! The things you learn hey?
How is pancreatitis treated?
When I was a junior doctor, back in the mists of time, the treatment of pancreatitis was supportive. These patients were often so unwell with multi-organ failure that they ended up on a high dependency or intensive care unit. They needed fluids, oxygen, pain killers and nutritional support. These patients with pancreatitis either got better as the inflammation settled down on its own or they died. The main differences in modern management of pancreatitis are: the routine use of procedures like ERCP to remove the underlying gallstones in order to allow the inflammation to settle and the increasing use of antibiotics to treat suspected infections; and that is where the controversy begins.
Antibiotics in pancreatitis
So why do doctors want to give antibiotics to patients with pancreatitis? The answer I get given is infection! Infection, either in necrotic pancreatic tissue or as a complication of pancreatitis (e.g. pneumonia), is stated as a leading cause of morbidity and mortality in acute pancreatitis. So as antibiotics treat infections, antibiotics should be given in pancreatitis. Right? Well maybe not…
The controversy in the use of antibiotics in pancreatitis is when to use them; should they be used to try and prevent infection (prophylaxis) or to treat infection when it is evident. We now know that antibiotics should ONLY be given to treat infection, and that they make no difference as prophylaxis1, but it has taken us a number of years to get to this point.
Just to recap, in pancreatitis the pancreas itself has broken down releasing those destructive digestive enzymes, further breaking down or necrotising the pancreas and its surrounding tissue (remember the patient is essentially digesting themselves!). This doesn’t happen in every patient; less than 20% of patients with pancreatitis develop necrosis. Necrosis is diagnosed using imaging such as CT scans which shows diffuse or focal areas of dead pancreatic tissue >3 cm in size or >30% of the pancreas.
This necrotic tissue is the perfect culture media for many of the upper gastrointestinal microorganisms which leak out of the damaged gut and cause infection in the gooey mess left behind. Infection happens in about 30% of patients with necrotising pancreatitis i.e. only 6% of all patients with pancreatitis. The mortality in these 6% of infected necrotic pancreatitis is as high as 70%; therefore it was proposed back in the late 1990s that antibiotics should be given to all patients with necrotising pancreatitis to prevent these life-threatening infections from occurring. But is this still correct, or is it overkill!?
Antibiotic prophylaxis in necrotising pancreatitis
At first glance it might make sense to try and prevent such a devastating infection from occurring, the problem is you would actually either have to eliminate all of the microorganisms from the upper gastrointestinal tract in order to prevent them getting in to the necrotic tissue, or you would have to get the antibiotic concentration so high in the necrotic tissue that no microorganisms could get established and cause infection.
The original research that suggested a benefit of antibiotics only looked at small numbers of patients with necrotising pancreatitis and these were non-blinded studies. The results were not conclusive.
Subsequent randomised, double-blind, placebo controlled trials have failed to show any benefit from the use of antibiotic prophylaxis in necrotic pancreatitis, including at least one study by authors who originally proposed the potential benefits to prophylaxis! A further review of using prophylaxis in necrotising pancreatitis combining the results of 11 randomised, double-blind, placebo controlled trials showed that to prevent 1 patient with pancreatic necrosis becoming infected 1,429 patients with pancreatic necrosis would need to be given prophylaxis with broad-spectrum antibiotics; or to put it another way for 1 patient to get some benefit 1,428 patients would get no benefit (i.e. they would still either get infected or not irrespective of the prophylaxis) and be exposed to the risk of broad-spectrum antibiotics (e.g. C. difficile, allergic reactions, side-effects, toxicity). From this evidence, I find it impossible to justify giving antibiotic prophylaxis to patients with necrotic pancreatitis.
So who should get antibiotics?
If you look at the UK and USA guidelines (see links at the end of the blog) in this area there is a bit of discrepancy. The UK guideline says there is no consensus and prophylaxis can be given for up to 14 days in necrotic pancreatitis. The USA guideline clearly says not to give prophylaxis. Why the difference? Look at the dates; the UK guideline is from 2005, the USA guideline is from 2013. The USA guideline is more up-to-date and uses 8 years of more evidence to answer the question.
Treatment of infected necrosis in pancreatitis
So it is now known that antibiotics do not prevent infection occurring, but what should be done when infection does occur. There is consensus that once necrotic tissue has become infected it should be treated.
One of the tricky bits with this is in deciding who actually has infected necrosis? It is often very difficult to distinguish severe pancreatitis without infection, from sepsis as a result of infection, this is because they both present in the same way:
- Fever
- Tachycardia, hypotension, tachypnoea,
- Increased White Blood Cells (WBC) in blood
- Raised C Reactive Protein (CRP)
- Shock and multi-organ failure
As a result many patients with pancreatitis but without an infection end up being treated with broad spectrum antibiotics. Wouldn’t it be great to have a way of telling who has infection and who doesn’t?
One way of telling if a patient has infected pancreatic necrosis is to do a fine needle aspiration (FNA) of some of the necrotic tissue, and it is recommended that all patients with necrotising pancreatitis should undergo CT-guided FNA to look for infection. In this procedure a thin needle is passed either through the abdominal wall under CT scan guidance or through the wall of the stomach using an endoscope and some tissue sucked out. This tissue can then be Gram stained to look for bacteria and cultured. A positive Gram stain or culture proves the necrosis is infected (the result can be available within a couple of hours). Antibiotics are often started empirically to cover the normal flora of the upper gastrointestinal tract whilst waiting for the microbiology results. If the Gram stain and culture are negative the antibiotics can be stopped, as no infection is present.
What if there is no CT guided FNA available? If infected necrosis is suspected, then you’ll need to start empirical antibiotics, but remember only 30% of the necrotic pancreatitis’ are actually infected therefore 70% would be given antibiotics unnecessarily.
The current evidence suggests that up to two thirds of patients with infected necrosis can be successfully treated with antibiotics alone without the need for surgical intervention to remove necrotic material (necrosectomy). In the group of conservatively treated patients, mortality is about 12% compared to 40-55% who underwent surgery.
Antibiotics need to be chosen that get into the pancreas and the necrotic tissue. Appropriate choices include Piptazobactam, cephalosporins, carbapenems, fluoroquinolones and metronidazole. My personal preferences would be:
1st Line |
IV Piptazobactam 4.5g QDS |
2nd line (if 1st line contraindicated) |
IV Ciprofloxacin 400mg TDS PLUS IV Metronidazole 500mg TDS |
Treatment duration
Now to a bit of a problem with the current UK and USA guidelines; no one gives any indication what-so-ever about how long to keep the antibiotics going for in patients with infected pancreatic necrosis. My personal practice is to treat these patients as an intra-abdominal abscess and give at least 2 weeks and then review the patient in terms of the clinical response and the WBC and CRP. If there is evidence of ongoing infection then I would consider continuing for a further 2 weeks but as the patient has failed to improve, I would try and persuade my surgical colleagues to think about an early necrosectomy.
A patient who has failed to respond to 2 weeks of antibiotics or deteriorates clinically should be considered for an early necrosectomy where as much of the infected necrotic material as possible is removed. If the patient remains stable a delayed necrosectomy after 4 weeks is preferable as by this time the necrotic material should have liquefied and will have walled off from the rest of the body making it easier to remove.
So what does this mean for your patient? To put it simply, the patient needs a CT scan. If the CT scan shows that the patient has a necrotic pancreas then consider whether you need to undertake a CT-guided FNA to look for infection. If the FNA Gram stain shows microorganisms THEN start antibiotics. If the CT doesn’t show necrosis then only give antibiotics if there is another clear focus of infection e.g. cholecystitis, pneumonia.
Okay put like that it sounds simple, but in reality these patients are often too unwell to have a CT scan, let alone a CT-guided FNA! So in reality a clinical decision has to be made about whether to start antibiotics whilst waiting for the patient to be stable enough for further investigation.
For our patient above, the surgical registrar and the Microbiologist discussed whether the patient was well enough to have a CT scan. The Surgical Registrar was happy that the patient could have a CT scan and this was arranged. The scan did not show any necrosis so a decision was made not to start antibiotics at this stage. The patient slowly improved over the next few days and eventually made a full recovery. The patient underwent ERCP to remove their underlying gallstones.
So is there anything else out there which can help doctors decide whether to give antibiotics to these complex patients when it is difficult to tell if they have an infection or not? One answer may be the infection biomarker “procalcitonin”… but that’s the subject of a future blog… watch this space.
References
- American Guideline – American college of gastroenterology guideline: management of acute pancreatitis, American Journal of Gastroenterology 2013
- British guideline – UK guidelines for the management of acute pancreatitis, Gut 20
Here's another link for you... on Gastroesophageal Reflux Disease (GERD), we have been asked by Very Well Heath to display this link to help them raise awareness of GERD. The article is reviewed by board-certified physicians. NB Nuts and Bolts have not been involved in reviewing this material or benefited from displaying this link.