Previously I blogged about the use of antibiotics in pancreatitis and I mentioned procalcitonin as an exciting new marker of infection. I have little experience of this laboratory test but have heard many others talk about how useful it is. But what is it? What is it useful for? Can it help in the diagnosis and management of infection?
What is procalcitonin?
Procalcitonin (PCT) is a protein produced by the thyroid, lungs and intestine in response to inflammation, especially when the inflammation is caused by bacterial infection. In contrast, levels of PCT do not increase much when inflammation is caused by viruses or most non-infectious causes. Levels increase quickly (within 2-4 hours) and have a half-life of 24-36 hours. It is suggested (by the manufacturers of the PCT assay machines) that PCT can therefore be used to identify septic patients and predict who is at risk of developing severe sepsis. The amount of PCT is said to be related to the amount of inflammation therefore in theory the level of PCT can also be used to monitor response to treatment; a decrease in PCT corresponding to a favourable response to treatment.
What is procalcitonin?
Procalcitonin (PCT) is a protein produced by the thyroid, lungs and intestine in response to inflammation, especially when the inflammation is caused by bacterial infection. In contrast, levels of PCT do not increase much when inflammation is caused by viruses or most non-infectious causes. Levels increase quickly (within 2-4 hours) and have a half-life of 24-36 hours. It is suggested (by the manufacturers of the PCT assay machines) that PCT can therefore be used to identify septic patients and predict who is at risk of developing severe sepsis. The amount of PCT is said to be related to the amount of inflammation therefore in theory the level of PCT can also be used to monitor response to treatment; a decrease in PCT corresponding to a favourable response to treatment.
How is procalcitonin measured?
PCT is measured on an analyser. This assay/machine is not currently widely available in the UK but this is changing. The assay is performed on a serum sample taken from a vein and collected in a red or yellow vacutainer. A minimum volume of 0.25mL serum is usually required; which corresponds to 0.5mL blood.
PCT is measured on an analyser. This assay/machine is not currently widely available in the UK but this is changing. The assay is performed on a serum sample taken from a vein and collected in a red or yellow vacutainer. A minimum volume of 0.25mL serum is usually required; which corresponds to 0.5mL blood.
Guess I'm measuring the wrong PCT! ...Wishful thinking.
According to the Association of Clinical Biochemists UK the normal range of PCT in adults is ≤ 0.05 ng/mL and the ranges for diagnosing pneumonia and sepsis are below. Raised levels need to be interpreted in the clinical context of the patient. If the patient does not clinically have an infection then don’t do a PCT, you won’t be able to interpret the result because there are no criteria for interpreting PCT in the absence of an infection e.g. if the PCT is 1.0 ng/mL and the patient has no infection then what does this mean? Also, in order to use the pneumonia criteria below the patient does have to have evidence of pneumonia such as consolidation on a chest x-ray, the PCT values do not apply if the chest x-ray is normal; all the PCT will “help” with is deciding if the pneumonia is bacterial or viral or deciding to give shorter durations of treatment… but which of us will be brave enough to do that? Do you trust a PCT result over the standard treatment of 7 days for pneumonia and decades of clinical experience with this?
If a diagnosis of pneumonia is suspected:
If a diagnosis of pneumonia is suspected:
Absence of infection: |
PCT < 0.1 ng/mL |
Bacterial infection unlikely: |
PCT ≥ 0.1–< 0.25 ng/mL |
Bacterial infection possible: |
PCT ≥ 0.25–< 0.5 ng/mL |
Suggestive of bacterial infection: |
PCT ≥ 0.5 ng/mL |
If a diagnosis of systemic bacterial infection or sepsis is suspected:
Systemic infection unlikely: |
PCT < 0.5 ng/mL |
Systemic infection possible: |
PCT ≥ 0.5– < 2 ng/mL |
Systemic infection likely: |
PCT ≥ 2 ng/mL |
Severe sepsis or septic shock likely: |
PCT ≥ 10 ng/mL |
Is it just me or are there a lot of “unlikelys”, “possibles” and “likelys” in these ranges?! To me, the maybe maybe-not approach is not really very helpful. It also means another set of baffling ranges to remember especially when you factor in the need to remember a different range for each condition!
Another problem with PCT is that is not actually 100% sensitive or specific. Localised infections and most non-infectious inflammatory conditions do not give a raised PCT but high PCTs can occur in some non-infectious conditions such as severe trauma, major surgery or cardiogenic shock; remember PCT is measuring a response to inflammation.
How should procalcitonin be used in acute pancreatitis?
I originally looked at procalcitonin as a potential help in diagnosing infected pancreatic necrosis, so how useful could it be in this condition?
It has been proposed that a cut off of 1.8ng/mL on 2 consecutive days can be used to predict infected necrosis with sensitivity of 95% and specificity of 88%. A few years later the same authors proposed a cut off of >3.5ng/mL on 2 consecutive days with a sensitivity of 93% and specificity of 88%. So which is it; 1.8ng/mL or 3.5ng/mL? There are only a few papers using PCT in diagnosing infected pancreatic necrosis and it is clear there’s still work to be done. A test that fails to spot 5-7 out of 100 patients with infected necrosis is of limited use on its own.
In practice the PCT would not be used in isolation for diagnosing infected pancreatic necrosis but rather within the clinical context. The PCT would be looked at in conjunction with the physiological parameters of the patient such as heart rate, respiratory rate, blood pressure and other blood tests such as C-reactive protein (CRP), the peripheral white blood cell (WBC) count and serum lactate. Finally, if there is any doubt about the diagnosis then either a CT-guided FNA should be done or empirical treatment started.
There is another significant drawback in the measurement of PCT in pancreatitis and that is that the test does not provide accurate assessment in patients with raised bilirubin levels as the colour interferes with the reading of the test, giving false results. This is a real shame as many pancreatitis patients are jaundiced due to gallstone obstruction of the bile ducts.
What about the claims for predicting sepsis and the severely septic patient?
Whilst researching this topic I came across a fantastic video from Dr James Faix, Associate Professor of Pathology at Stamford University in the USA, which was recorded in 2012. Research in this area has probably moved on a bit in the last 4 years but this really is a good summary of the use of PCT in sepsis.
I would add two points to this summary:
Another problem with PCT is that is not actually 100% sensitive or specific. Localised infections and most non-infectious inflammatory conditions do not give a raised PCT but high PCTs can occur in some non-infectious conditions such as severe trauma, major surgery or cardiogenic shock; remember PCT is measuring a response to inflammation.
How should procalcitonin be used in acute pancreatitis?
I originally looked at procalcitonin as a potential help in diagnosing infected pancreatic necrosis, so how useful could it be in this condition?
It has been proposed that a cut off of 1.8ng/mL on 2 consecutive days can be used to predict infected necrosis with sensitivity of 95% and specificity of 88%. A few years later the same authors proposed a cut off of >3.5ng/mL on 2 consecutive days with a sensitivity of 93% and specificity of 88%. So which is it; 1.8ng/mL or 3.5ng/mL? There are only a few papers using PCT in diagnosing infected pancreatic necrosis and it is clear there’s still work to be done. A test that fails to spot 5-7 out of 100 patients with infected necrosis is of limited use on its own.
In practice the PCT would not be used in isolation for diagnosing infected pancreatic necrosis but rather within the clinical context. The PCT would be looked at in conjunction with the physiological parameters of the patient such as heart rate, respiratory rate, blood pressure and other blood tests such as C-reactive protein (CRP), the peripheral white blood cell (WBC) count and serum lactate. Finally, if there is any doubt about the diagnosis then either a CT-guided FNA should be done or empirical treatment started.
There is another significant drawback in the measurement of PCT in pancreatitis and that is that the test does not provide accurate assessment in patients with raised bilirubin levels as the colour interferes with the reading of the test, giving false results. This is a real shame as many pancreatitis patients are jaundiced due to gallstone obstruction of the bile ducts.
What about the claims for predicting sepsis and the severely septic patient?
Whilst researching this topic I came across a fantastic video from Dr James Faix, Associate Professor of Pathology at Stamford University in the USA, which was recorded in 2012. Research in this area has probably moved on a bit in the last 4 years but this really is a good summary of the use of PCT in sepsis.
I would add two points to this summary:
- The value of PCT has been shown in predicting whether someone with sepsis might go on to develop severe sepsis
- There are still a number of patients who will have a negative PCT who still develop severe sepsis (Negative Predictive Value) and some who have a positive PCT who do not develop severe sepsis (Positive Predictive Value); over treating those who don’t go on to develop severe sepsis is not really a problem as it is a safe approach but not treating some who go on to develop severe sepsis through over reliance on a PCT result is a problem (see graph below for NPV and PPV in predicting severe sepsis)
The promotional material from the companies who produce the PCT assays suggest it should be used in the diagnosis of sepsis but this is not what NICE or the Surviving Sepsis Campaign say. NICE state “there was not enough evidence to recommend that these tests are used in the NHS”. The Surviving Sepsis Campaign suggest that “low procalcitonin levels or similar biomarkers [can be used] to assist the clinician in the discontinuation of empiric antibiotics in patients who initially appeared septic, but have no subsequent evidence of infection” but they clarify this by also saying “clinical experience with this strategy is limited and the potential for harm remains a concern”. Draw your own conclusions!
Should I rush out and buy a procalcitonin assay for my laboratory?
Well the answer to this is what do you want to use it for? Gosh, Microbiologists never answer the question posed, always presenting another one of their own! If you are looking for the magic test which tells you if a patient has sepsis or infected pancreatic necrosis then forget it, PCT does not do this with sufficient accuracy to be the means by which you tell if a patient is septic or not (see would you recognise a sick patient, sepsis blog).
Personally I’m not convinced it provides much more information than a carefully taken medical history, examination and more conventional blood tests in terms of telling how sick a patient is, but then I have little experience of the test in the clinical setting. I don’t think we need to rely solely on chemical markers to diagnosis sick patients and I’m not going to be rushing out and ordering a load of PCTs on my patients just yet. In fact The Association for Clinical Biochemistry conclude “CRP is a routinely available non-specific marker of inflammation, whose usefulness is not superseded by PCT. However, the use of PCT and CRP in parallel may improve the specificity and sensitivity of the diagnosis of bacterial infection/sepsis”. I agree.
Do you have any experience of procalcitonin in clinical practice? What do you think of it? How do you use it? Let me know what you think.
Should I rush out and buy a procalcitonin assay for my laboratory?
Well the answer to this is what do you want to use it for? Gosh, Microbiologists never answer the question posed, always presenting another one of their own! If you are looking for the magic test which tells you if a patient has sepsis or infected pancreatic necrosis then forget it, PCT does not do this with sufficient accuracy to be the means by which you tell if a patient is septic or not (see would you recognise a sick patient, sepsis blog).
Personally I’m not convinced it provides much more information than a carefully taken medical history, examination and more conventional blood tests in terms of telling how sick a patient is, but then I have little experience of the test in the clinical setting. I don’t think we need to rely solely on chemical markers to diagnosis sick patients and I’m not going to be rushing out and ordering a load of PCTs on my patients just yet. In fact The Association for Clinical Biochemistry conclude “CRP is a routinely available non-specific marker of inflammation, whose usefulness is not superseded by PCT. However, the use of PCT and CRP in parallel may improve the specificity and sensitivity of the diagnosis of bacterial infection/sepsis”. I agree.
Do you have any experience of procalcitonin in clinical practice? What do you think of it? How do you use it? Let me know what you think.
RSS Feed