Why do we use antibiotics for SBP prophylaxis?

You may have noticed that we often start antibiotics for patients with cirrhosis to prevent spontaneous bacterial peritonitis (SBP). Have you ever wondered why?

Let’s start with a refresher on SBP, courtesy of Hersh Shroff.   

Hersh highlights some critical facts about SBP in his Quick Tips post:

1. SBP is common in patients with decompensated cirrhosis and is associated with high morbidity and mortality. SBP can result in a whole host of bad outcomes including acute kidney injury, sepsis, multiorgan failure, and has an estimated in-hospital mortality rate of up to 20%.
2. If you have had SBP once, it is likely to recur. It is estimated that SBP survivors have a 70% risk of recurrence within 1 year!

Why does SBP happen in cirrhosis in the first place?

As many of you know, ascites develops in cirrhosis when pressure is increased across the liver due to portal hypertension causing fluid to build up in the abdomen. Activation of the renin-angiotensin system contributes to increased salt and water retention from the kidneys, which also leads to ascites, as can be seen in the diagram below.

Bacteria from the intestines can make their way into the bloodstream and subsequently into ascitic fluid causing SBP. This is called bacterial translocation. Individuals with cirrhosis are at higher risk of bacterial translocation for several reasons: 1) gut microbiome alterations 2) increased gut mucosal permeability 3) deficiencies in the local and systemic immune response 4) portosystemic collaterals, providing a perfect storm to allow gut bacteria to bypass the liver and enter the systemic circulation.

I highly recommend checking out figure 1 of this NEJM article for an awesome review of the pathophysiology of infections in cirrhosis. Here are some highlights:

• The gut microbiome is altered leading to decreased synthesis of beneficial acids and metabolites that strength the intestinal barrier
• The tight junctions between intestinal epithelial cells are disrupted causing increased gut mucosal permeability
• Patients with liver dysfunction have decreased ability to produce complement, acute-phase proteins, and soluble pattern-recognition receptors, resulting in diminished defense against bacterial infections.
• The lamina propria, which is made up of loose connective tissue beneath the epithelium, has immune defense functions that are decreased in cirrhosis.

Why is there low protein in cirrhotic ascites? The fibrous tissue deposition in the liver makes it difficult for proteins to squeeze past the hepatic sinusoids. These proteins, including complement and immunoglobulins, are critical for bactericidal activity to fight infections. With reduced opsonic activity, there is a higher likelihood of infection.

How do I diagnose SBP?

SBP is diagnosed through direct evaluation of the ascitic fluid with a paracentesis. You have made the diagnosis if there is an absolute polymorphonuclear leukocyte (PMN) count of at least 250 cells/mm3 (calculated by total WBC count X PMN %) in the fluid sample with positive ascitic fluid culture, in the absence of an intra-abdominal surgically treatable source of infection. In order to correct the PMN count for red blood cells, simply subtract 1 PMN per 250 red blood cells found in the fluid sample.

You should plan to repeat a paracentesis at 48 hours for all patients with SBP to ensure the PMN count is decreasing by at least 25% to ensure they are responding appropriately to antibiotics. A recent study found that the percent decline in PMNs may inform prognosis and may help you uncover a resistant infection.

There are two other clinical conditions you should be aware of: culture-negative neutrocytic ascites and non-neutrocytic bacterascites. Culture-negative neutrocytic ascites refers to individuals who have an ascitic fluid PMN count of at least 250 cells/mm3 with a negative bacterial culture. As it is generally considered to be a variant of SBP, you can treat this the same as SBP with a course of antibiotics.

Non-neutrocytic bacterascites refers to a positive bacterial culture with a normal neutrophil count in the fluid of less than 250 cells/mm3. Non-neutrocytic bacterascites may represent colonization of the ascitic fluid and typically does not require treatment unless there are other clinical signs of infection. Some cases of non-neutrocytic bacterascites may represent early spontaneous bacterial peritonitis so follow up paracentesis is recommended to see if the PMN count meets criteria for SBP.  

Who is most at risk of getting SBP?

Hopefully, I have convinced you that individuals with cirrhosis are at higher risk of developing SBP than the general population for the reasons stated above. But what patients with cirrhosis are at highest risk of SBP and might benefit from preventative measures? This really comes down to three main clinical settings:

1. A prior history of SBP (secondary prophylaxis)
2. Upper gastrointestinal bleed
3. Low-protein ascites less than 1.5 g/dL PLUS impaired liver or renal function (primary prophylaxis) **This is a little controversial- more on that soon!

Individuals who have had SBP before are at particularly high risk of developing it again. SBP recurs within a year in about 70% of individuals with a prior episode of SBP and is associated with a low survival rate of only 30 to 50%. The identification of prior SBP as a risk factor for future episodes led to randomized control trials of prophylactic antibiotics.

One of the first randomized control trials published in 1990 from a group in Barcelona, studied the effects of antibiotic prophylaxis with norfloxacin versus placebo in patients who had recently recovered from an episode of SBP. The results were impressive with a significantly lower recurrence of SBP in the patients on norfloxacin versus placebo (20% versus 68%). This, along with other studies, led to the recommendation to use antibiotics for secondary prophylaxis in patients that have a history of SBP. Long-term secondary prophylaxis is currently recommended until liver transplantation or death.

Upper gastrointestinal bleeding is the second clinical scenario where antibiotics are used for SBP prophylaxis. Upper GI bleeding is associated with increased risk of SBP during or immediately after the bleeding episode with an incidence up to 66% in advanced cirrhosis, likely due to increased bacterial translocation at the time of the bleed. Studies looking at antibiotic prophylaxis after GI bleeding have shown a reduction in in-hospital infections, death, and rebleeding. This is best illustrated in the figure below:

Antibiotic prophylaxis is currently recommended to be started as early as possible after upper GI bleeding according to the American Association for the Study of Liver Diseases and to continue for 7 days.  

What about those with low protein ascites? Do they need primary prophylaxis?  

Low protein levels in ascitic fluid have been found to be associated with increased risk of SBP dating back to a 1986 Gastroenterology paper studying protein concentrations in ascitic fluid. Dr. Runyon looked at the protein concentrations in the ascitic fluid of patients with SBP as compared to those with sterile portal hypertension ascites and ascites due to miscellaneous causes. He found that the protein concentration was significantly lower in the group with SBP. He also found that patients with an initial protein concentration of less than 1 g/dL were 10 times more likely to develop SBP during their hospitalization than those with a protein concentration greater than 1 g/dL.

Individuals with low protein in their ascites and no other risk factors have a risk of developing SBP of less than 20% at 1 year. The risk is much higher with low protein ascites and markers of severe liver failure (Child-Pugh score 9 points with serum bilirubin 3 mg/dL) and/or circulatory dysfunction (serum creatinine 1.2 mg/dL, blood urea nitrogen 25 mg/dL, or serum sodium 130 mEq/L). A randomized control trial by Fernandez compared norfloxacin to placebo in patients with low protein ascites plus impaired liver or renal function and found a significantly different 1-year probability of developing SBP (7% versus 61%, P < 0.001) and improved 1-year survival (60% vs 48%, p=0.05).

One major caveat about this study is that it did not include death as a competing risk in the analysis, which may bias the results.

Results from this study helped lead to recommendations from the American Association of Liver Diseases and the European Association for the Study of the Liver to use antibiotic primary prophylaxis in individuals with low-protein ascites less than 1.5 g/dL PLUS impaired liver (Child-Pugh score 9 points with serum bilirubin 3 mg/dL) or renal dysfunction (creatinine 1.2 mg/dL, blood urea nitrogen 25 mg/dL, or serum sodium 130 mEq/L).  

Why is there controversy around giving antibiotics for primary SBP prophylaxis?

Many of the randomized control trials of primary SBP prophylaxis versus placebo had flaws in their methodology. Some studies, for example, included patients who had a prior history of SBP, adding heterogeneity to the study populations and potentially biasing the study results. In some of the other RCTs, there was no significant difference between the treatment and placebo group. A huge shout out to Jeremy Louissaint for his work in this area, who put this wonderful chart together going through the nitty gritty of each study and potential for study bias!

More recent studies also call into question Runyon’s findings from his 1986 Gastroenterology paper. For example, in a 2015 retrospective analysis of 385 hospitalized patients with ascites receiving a paracentesis, there was no significant difference in ascitic fluid protein between patients with and without SBP.

In addition, there was a 2019 paper in the American Journal of Gastroenterology that found that patients on primary SBP prophylaxis had worse outcomes than those on secondary prophylaxis, raising additional questions about the use of antibiotics for primary SBP prophylaxis. There is currently a randomized control trial underway called ASEPTIC that will be looking at primary SBP prophylaxis in patients with liver cirrhosis and ascites persisting for >3 months, who have never had an episode of SBP and are not receiving prophylactic antibiotics. Patients randomized to receive trimethoprim-sulfamethoxazole (TMP/SMX or Bactrim) or placebo will be evaluated over a minimum of 18 months to assess important clinical outcomes such as overall survival and incidence of SBP between the two groups. This study will importantly provide much needed evidence on the safety and efficacy of TMP/SMX for primary SBP prevention. Stay tuned for the results of the study in a few years! 

What antibiotics are used for SBP prophylaxis?

The majority of SBP infections are due to gram negative bacteria like E. Coli and Klebsiella and therefore, antibiotic prophylaxis should have good coverage against gram negative organisms.

You may have noticed that many of the RCTs evaluating antibiotics for SBP prophylaxis utilized norfloxacin, which is not currently available in the United States. Norfloxacin was previously approved in the United States in 1986 but was discontinued by Merck in 2014 for unclear reasons (potentially financially motivated?) but notably not due to safety or efficacy concerns. Norfloxacin has the advantage of being incompletely absorbed by the intestine and is thought to be less likely to cause bacterial resistance, so it is too bad we do not have access to it anymore in the United States.

The current antibiotic regimens recommended for primary and secondary SBP prophylaxis in the United States are ciprofloxacin 500 mg/day PO or TMP/SMX double strength PO daily. However, it is important to note that antibiotic choice should be tailored to patient history, allergies, cost, and availability of the medication as there are several other antibiotics that could be considered in the right clinical scenario.

Antibiotics for infection prophylaxis after a GI bleed is typically with intravenous ceftriaxone and only lasts for a total of 7 days. Once the bleeding has stopped it is also possible to transition to oral antibiotics with ciprofloxacin or TMP/SMX to complete the 7-day antibiotic course.

What are the risks of long-term SBP prophylaxis?

There are several major concerns associated with long-term antibiotic use:

• Antibiotic resistance. This can lead to decreased antibiotic effectiveness and cause changes in the microbiologic profile of SBP infections over time by increasing multi-drug resistant organisms. Unfortunately, there is already evidence that this is happening.
• Antibiotic side effects. Fluoroquinolones have a black box warning for increased risk of tendonitis and tendon rupture and TMP/SMX is associated with a risk of hyperkalemia, increases in creatinine, and rarely Steven-Johnson skin rash.
• Clostridium difficile. SBP prophylaxis is associated with an increased risk of c. diff. The development of c. diff infection is associated with poor outcomes in cirrhosis.
• Polypharmacy. Individuals with advanced liver disease are likely already multiple medications and long-term antibiotics add to their pill burden. Long-term antibiotics also increase healthcare costs and increase the risk of drug-drug interactions.

All and all, there is still work that needs to be done to better understand the patient groups that most benefit from SBP prophylaxis. Hopefully, this review helps summarize some of the major studies and recommendations for SBP prophylaxis and potentially inspires you to do some research of your own to ensure we are using antibiotics judiciously in patients with cirrhosis.