By Matthew Light, MD
Peer Reviewed
A 79 year-old female, 20 pack-year ex-smoker with history of mild systolic congestive heart failure is referred to the emergency department (ED) after an outpatient visit with her primary care physician where she complained of fever, sinus congestion, myalgias, and productive cough. On arrival she is febrile, mildly tachypneic and tachycardic, but normotensive and with a normal oxygen saturation on room air. On exam she has bronchial breath sounds and rales in her left chest. The patient has a leukocytosis and her chest x-ray shows a lingular consolidation with air bronchograms. The Pneumonia Severity Index (a composite of patient comorbidities, exam findings, and laboratory and radiographic results, PSI) is IV [1]. The patient is given ceftriaxone and azithromycin and admitted to the medicine service. The question arises, is there a role for checking a procalcitonin (PCT) level in this patient as part of the initial work-up or tracking PCT levels during hospitalization?
Rates of hospitalization for pneumonia (PNA) in the United States, though trending down, are approximately 200 and 350 per 10,000 patients in the 75-84 and 85 years or older age groups, respectively [2]. In addition 50,000 deaths were attributed to PNA (US data only) in 2010 [3]. With such substantial morbidity and mortality, there is always a fear of under-treatment, and as such, patients are kept in the hospital longer and given more days of antibiotics. However, this fear is opposed by other health care goals such as reducing costs, lengths of stay, adverse drug reactions, and over-use of antibiotics.
Clinical practice is changing thanks to new research into the optimal duration of antibiotics and markers of disease etiology and progression. Previous Infectious Disease Society of America/American Thoracic Society (IDSA/ATS) guidelines recommended 7-21 days of antibiotic treatment for PNA. However, this was changed to five days (minimum) [4] in 2007 after a meta-analysis showed non-inferiority for patients treated less than seven days for mild to moderate community acquired pneumonia (CAP) [5]. But with viral lower respiratory tract infections (LRTIs) representing up to one third of cases [6] in adults, and sputum cultures that are often unrevealing, significant diagnostic dilemmas remain. How can we accurately distinguish between bacterial and viral infections and can we reliably track disease progression?
Enter procalcitonin, usually produced by neuro-endocrine C-cells in the thyroid as part of normal calcium homeostasis. In response to endotoxin and inflammatory cytokines released during bacterial infection (i.e. IL-1 beta and TNF-alpha), parenchymal cells in the lung (as well as liver, kidney, muscle, and adipose tissue) have the green light to begin a constitutive secretory pathway, resulting in elevated plasma concentrations [7]. In addition, interferon gamma, associated with viral processes, blocks this constitutive pathway. As a result, procalcitonin has the potential to differentiate bacterial from viral infections (a problem in LRTIs, see above), and in bacterial processes, assess response to antibiotics. With this in mind, a series of trials was designed to evaluate the clinical utility of this convenient biomarker in patients with CAP.
In 2005, Masia et al. published a one year prospective, observational study in Chest looking at the use of procalcitonin in predicting both etiology and prognosis in adult patients with PNA [8]. They found that in patients with a low PSI (I-II), procalcitonin tended to be higher in classic bacterial infections (S. pneumoniae, Haemophilus influenzae, etc.) compared to other infections, including atypical bacteria, viral pathogens, and mixed infections (p=0.08). Further, they report that a PCT cut off of > 0.15 ug/L identified 37.5% of patients with any bacterial infection (typical or atypical) verses just 13.3% of non-bacterial infections (p=0.03). However, PCT levels increased along with PSI scores, such that patients with a score of III or higher had elevated levels regardless of PNA etiology, reducing the test’s diagnostic utility. In looking at the PSI class III-V data, PCT levels above 0.50 ug/L were associated with lower blood pressures, higher rates of empyema, mechanical ventilation, septic shock, and death (all p values < 0.05).
In a subsequent prospective, randomized trial by Christ-Crain et al., 302 patients presenting to the ED with CAP were randomized to a procalcitonin-guided antibiotic algorithm or standard of care [9]. Procalcitonin was checked at baseline and on days four, six, eight, as well as at six weeks (study duration). Antibiotic duration was significantly reduced in the procalcitonin arm (median 5 days vs 12 days, p < 0.001), and there was no difference in adverse outcomes between groups, including persistence of pneumonia, need for ICU stay, empyema, acute respiratory distress syndrome (ARDS), and death. At six weeks of follow-up, cure rates were approximately 83% in both groups. Of note, antibiotic costs were significantly lower in the PCT group ($29,428 vs $59,535, p < 0.001, USD). However adding assay costs to antibiotics negated this savings ($290 per patient in the PCT group vs $190 in the control group). Costs were based on an average of 3.5 PCT assays per patient and the authors estimated that the intervention would be cost-effective if less assays were performed or if the cost per assay fell below $25 (currently the assay costs between $25 and $30) [10].
Finally, in the ProHOSP trial Schuetz et al. randomized 1,381 patients presenting to the ED with any LRTI to procalcitonin-guided antibiotic therapy or standard of care [11]. In the intervention arm, antibiotic initiation or continuation was encouraged for PCT levels > 0.25 ug/L, strongly encouraged for levels > 0.50 ug/L, discouraged for levels < 0.25 ug/L, and strongly discouraged for levels < 0.10 ug/L. Levels were checked on admission and at two-day intervals. Adherence to the study protocol was good, with only a 9.6% rate of protocol non-adherence in the CAP subset. Significant results (CAP data only presented here) included fewer antibiotic days (mean 7.2 days in PCT arm, 10.7 days in control arm, 37% reduction) and no difference in mortality (5.2% in PCT arm vs 5.6% in control arm). Examining the antibiotic data more closely, by day five 35% of patients in the PCT group had been taken off antibiotics vs 8% in the control group. Interestingly, there was no difference in hospital length of stay. Although not specifically addressed by the authors, a potential explanation is that physicians chose to monitor patients longer in whom antibiotics were discontinued per PCT-guided therapy.
Before drawing any hard conclusions, it should be noted there are limitations to the data. First, the trials discussed above are all European (as are most PCT-related clinical investigations), raising the question of applicability to the US patient population. The trials also use varying PCT schedules and cut-off values, making study comparison and a uniform recommendation for the use of PCT difficult. Lastly, as highlighted by Hayashi and Paterson in a 2011 review, the centers involved in these trials had varying degrees of antibiotic stewardship programs, which on their own have the potential to effectively limit antibiotic use [12].
Looking at the data as a whole, the question of whether or not to check procalcitonin levels during hospitalization for CAP is not so simple. The above data suggests a PCT-guided approach to antibiotic use is safe, which was confirmed in a 2012 meta-analysis [13], but when is it useful? In terms of differentiating bacterial and non-bacterial infections, there is little utility when the PSI is above II, as the severity of the infection and associated inflammatory response will elevate the PCT regardless of etiology (of note, PSI II and greater patients are those that are typically admitted to the hospital). With respect to tracking PCT for prognosis, in patients responding clinically within two to three days, data supports a five day course of antibiotics regardless of the PCT level. So the algorithm, while reassuring, does not seem to affect practice in this situation. Judicious use of antibiotics by clinicians and an effective antibiotic stewardship programs come into play in this scenario.
Where PCT is most useful is in sicker patients with less cardiopulmonary reserve. Stopping antibiotics in these patients can be a difficult decision and an elevated PCT (especially > 0.50 ug/L) indicates risk of clinical deterioration and potentially the need for prolonged therapy. Here, comparing PCT to the level on admission taken together with the remainder of the clinical picture can facilitate this decision. At a current assay cost of approximately $25, overall costs can be reduced in CAP if clinicians can limit the test to two PCT levels per patient per hospitalization. While checking PCT may not reduce length of stay, it can reduce antibiotic days and influence bacterial resistance patterns.
Case Conclusion:
In the case presented above, the patient’s initial procalcitonin was 0.8 ug/L. Because of her comorbidities and overall slow clinical response, a repeat level was checked on hospital day five and found to be 0.55 ug/L and intravenous antibiotics were continued. By hospital day seven, the patient had been afebrile for four days and noticeably more alert, sitting up in bed, and able to feed herself. A final procalcitonin level was 0.15 ug/L and antibiotics were then stopped. The patient was monitored for an additional 24 hours and discharged home in good condition. No specific pathogen was identified.
Dr. Matthew Light is a 3rd year resident at NYU Langone Medical Center
Peer reviewed by Melanie Maslow, MD, Infectious Disease Section Editor, Clinical Correlations
Image courtesy of Wikimedia Commons
References
1. Fine MJ et al. A prediction rule to identify low-risk patients with community-acquired pneumonia. N Engl J Med. 1997 Jan 23;336(4):243-50. http://www.ncbi.nlm.nih.gov/pubmed/8995086
2. National Hospital Discharge Survey, United States, 2000-2010. http://www.cdc.gov/mmwr/preview/mmwrhtml/mm6133a4.htm
3. Centers for Disease Control and Prevention FastStats. http://www.cdc.gov/nchs/fastats/pneumonia.htm
4. Infectious Disease Society of America/American Thoracic Society consensus guidelines on the management of community-acquired pneumonia in adults. Clin Infect Dis. 2007 Mar 1;44 Suppl 2:S27-72. http://cid.oxfordjournals.org/content/44/Supplement_2/S27.full.pdf+html+html
5. Li JZ, Winston LG, Moore DH, Bent S. Efficacy of short-course regimens for community-acquired pneumonia: a meta-analysis. Am J Med. 2007 Sep;120(9):783-90.
6. Ruuskanen O, Lahti E, Jennings LC, Murdoch DR. Viral pneumonia. Lancet. 2011 Apr 9;377(9773):1264-75. doi 10.1016/S0140-6736(10)61459-6. http://www.ncbi.nlm.nih.gov/pubmed/21435708
7. Christ-Crain M, Muller B. Procalcitonin in bacterial infections: hype, hope, more or less? Swiss Med Wkly. 2005 Aug 6;135(31-31):451-60. http://www.ncbi.nlm.nih.gov/pubmed/16208582
8. Masia, M et al. Usefulness of Procalcitonin Levels in Community-Acquired Pneumonia According to the Patients Outcome Research Team Pneumonia Severity Index. Chest. 2005; 128(4):2223-2229. doi:10.1378/chest.128.4.2223.
9. Christ-Crain M et al. Procalcitonin guidance of antibiotic therapy in community-acquired pneumonia: a randomized trial. Am J Respir Crit Care Med. 2006 Jul 1;174(1):84-93.
10. Can procalcitonin help guide therapy for suspected pneumonia & other infections? http://pulmccm.org/2013/review-articles/can-procalcitonin-help-guide-antibiotic-therapy-for-suspected-pneumonia-review-chest/
11. Scheutz P. Effect of procalcitonin-based guidelines vs standard guidelines on antibiotic use in lower respiratory tract infections: the ProHOSP randomized controlled trial. JAMA. 2009 Sep 9;302(10):1059-66. doi: 10.1001/jama.2009.1297. http://www.ncbi.nlm.nih.gov/pubmed/19738090
12. Hayashi Y, Paterson DL. Strategies for reduction in duration of antibiotic use in hospitalized patients. Clin Infect Dis. 2011 May;52(10):1232-40. doi: 10.1093/cid/cir063. http://www.ncbi.nlm.nih.gov/pubmed/21507920
13. Scheutz, P et al. Procalcitonin to Guide Initiation and Duration of Antibiotic Treatment in Acute Respiratory Infections: An Individual Patient Data Meta-Analysis. Clin Infect Dis. 2012 Sept 1; 55(5): 651-662. http://www.ncbi.nlm.nih.gov/pubmed/22573847