Calprotectin as a Biomarker for Melioidosis Disease Progression and Management

ABSTRACT

Melioidosis is a neglected tropical disease that is caused by the bacterium Burkholderia pseudomallei and is underreported in many countries where the disease is endemic. A long and costly administration of antibiotics is needed to clear infections, and there is an unmet need for biomarkers to guide antibiotic treatment and increase the number of patients that complete therapy. We identified calprotectin as a lead biomarker of B. pseudomallei infections and examined correlations between this serum protein and the antibiotic treatment outcomes of patients with melioidosis. Serum levels of calprotectin and C-reactive protein were significantly higher in patients with melioidosis and nonmelioidosis sepsis than in healthy controls. Median calprotectin levels were higher in patients with melioidosis than in those with nonmelioidosis sepsis, whereas C-reactive protein levels were similar in both groups. Notably, intensive intravenous antibiotic treatment of patients with melioidosis resulted in lower levels of calprotectin and C-reactive protein (P < 0.0001), coinciding with recovery. The median percent reduction of calprotectin and C-reactive protein was 71% for both biomarkers after antibacterial therapy. In contrast, we found no significant differences in calreticulin levels between the two melioidosis treatment phases. Thus, reductions in serum calprotectin levels were linked to therapeutic responses to antibiotics. Our results suggest that calprotectin may be a sensitive indicator of melioidosis disease activity and illustrate the potential utility of this biomarker in guiding the duration of antibiotic therapy.

INTRODUCTION

The worldwide disease burden of melioidosis is estimated to be 165,000 cases per year, including 89,000 fatalities (1). The causative agent of melioidosis, Burkholderia pseudomallei, is a Gram-negative soil bacterium that is intrinsically resistant to several antibiotics, including penicillin, ampicillin, first- and second-generation cephalosporins, gentamicin, tobramycin, and streptomycin (2). The recommended treatment for melioidosis is a minimum of 10 to 14 days of intravenous antibiotic therapy (acute phase) with meropenem and/or ceftazidime, followed by 3 to 6 months of oral therapy (eradication phase) with co-trimoxazole (sulfamethoxazole and trimethoprim) or co-amoxiclav (amoxicillin and clavulanic acid) (3, 4). The endpoints are preventing mortality for the acute-phase treatment and preventing recurrence for the eradication-phase treatment. Keeping in contact with patients during the eradication phase is difficult in many countries where the disease is endemic, leading to poor treatment compliance. Consequently, the duration of antibiotic administration in the acute phase is often extended beyond 14 days (3), resulting in great variation in the length of treatments. The intense treatment regimen is also unaffordable in many developing countries due to the high costs of antibiotics and hospitalization (5). Patient monitoring is essential, and clinical evaluations along with blood cultures are the only tools currently used in melioidosis patient management. Yet, tailoring specific antibacterial treatments to individual cases would result in better outcomes. Serum biomarkers can be monitored as an additional tool to guide antibiotic therapy, and the clinical benefits of serum biomarkers for patients with bacterial sepsis were demonstrated (6, 7). A multidimensional proteomics approach was previously used to identify molecular signatures of nonhuman primates (NHPs) infected with Burkholderia mallei (8), a bacterium that is closely related to B. pseudomallei. Calprotectin (CALP), a calcium and zinc binding protein, was elevated in serum samples of NHPs infected with B. mallei, while calreticulin (CALR) and additional host biomarkers were also enriched within infected tissues. CALP is a potent activator of innate immunity mediated through Toll-like receptor 4 (9) and is primarily a product of neutrophils. Elevated serum CALP levels are associated with inflammatory diseases (10, 11), while fecal CALP assays are widely used to differentiate inflammatory bowel disease from irritable bowel syndrome and ulcerative colitis (12). In the present study, we confirm that serum CALP levels are elevated in NHPs with respiratory infections caused by B. pseudomallei and examine the possibility of using CALP as a general biomarker for clinical responses during the acute and eradication phases of antibiotic treatment in human cases of melioidosis.

RESULTS

Infections of Macaca mulatta and other NHPs faithfully replicate human melioidosis, and these laboratory disease models can be used to study many aspects of infection (13). We first used an M. mulatta model of melioidosis to confirm that fluctuations in serum CALP were associated with infection, observing that serum levels were significantly elevated in M. mulatta infected for 28 days after challenge with B. pseudomallei compared to levels in preexposure serum samples (see Fig. S1 in the supplemental material). The mean CALP level was 2.01 ± 2.2 μg/ml in preexposure serum samples and 6.2 ± 2.3 μg/ml in postexposure samples. Based on the results from the preclinical animal model of disease, we next examined human cases of infection. The human study cohort consisted of 31 patients with melioidosis confirmed by culture and real-time PCR (RT-PCR), 18 nonmelioidosis sepsis cases, and 15 healthy control subjects from the same geographic areas of Sri Lanka, collected during September 2014 to December 2015. Demographic and clinical characteristics of the study group are presented in Table 1. Out of 31 melioidosis cases, 23 (74%) carried the major risk factor diabetes. The nonmelioidosis sepsis patients included cases with leptospirosis and typhus diagnosed by commercial rapid diagnostic tests or clinical symptoms (data not shown). Melioidosis patients received ceftazidime or meropenem for a minimum of 2 weeks during the acute phase of treatment, followed by co-trimoxazole for a minimum of 3 months. Paired blood samples from melioidosis cases were collected during the acute and eradication phases of the antibiotic treatment to monitor disease progression.

TABLE 1.

Patient and clinical characteristics of the study population

Characteristic	No. (%) of patients in:
	Melioidosis (acute) group	Nonmelioidosis sepsis group	Healthy control group
Age (median [range]) (yr)	50 (32–74)	40 (18–74)	39 (22–66)
Sex
Male	23 (74)	14 (78)	15 (100)
Female	8 (26)	4 (22)	0 (0)
Comorbidity
Diabetes	23 (74)	2 (11)	0 (0)
Alcoholism	3 (10)	2 (11)	6 (40)
Kidney disease	3 (10)	4 (22)	1 (7)
Othera (liver disease, thalassemia, SLE, ITP)	6 (20)	0 (0)	0 (0)
None	2 (6)	10 (56)	8 (53)
Clinical symptom
Sepsis	22 (71)	18 (100)	NAb
No sepsis	9 (29)	0 (0)	NA
Fever	27 (87)	18 (100)	NA
<15 days	17 (55)	18 (100)	NA
>15 days	10 (32)	0 (0)	NA
Unconfirmed	4 (13)	NA	NA
Cough	11 (35)	6 (33)	NA
Arthralgia/myalgia	11 (35)	9 (50)	NA
Abdominal pain	4 (13)	5 (28)	NA
Abscess	4 (13)	0 (0)	NA

^aSLE, systemic lupus erythematosus; ITP, idiopathic thrombocytopenic purpura.

^bNA, not applicable.

We compared serum levels of CALP with serum levels of CALR and C-reactive protein (CRP), which is frequently elevated in other bacterial infections (14). The median CALP levels in the melioidosis acute and eradication phases were 4.39 μg/ml (interquartile range [IQR], 1.58 to 6.53 μg/ml) and 1.39 μg/ml (IQR 0.52 to 1.96 μg/ml), respectively (Fig. 1), and the difference was significant (P < 0.0001). Median CALP levels were higher, but not statistically significant, for melioidosis patients with diabetes (4.92 μg/ml) than for those without diabetes (2.57 μg/ml). There was an increase in CALP levels for patients with <15 days of fever (4.55 μg/ml) compared with those of melioidosis patients who had >15 days of fever (3.68 μg/ml). The levels of CRP also differed significantly between the acute and eradication treatment phases (P < 0.0001). The median CRP level for the acute phase was 9.65 μg/ml (IQR, 6.62 to 18.58 μg/ml) and for the eradication phase was 2.91 μg/ml (IQR, 0.67 to 8.56 μg/ml). There was a significant increase (P < 0.001) in median CRP with >15 days of fever (10.69 μg/ml) compared with that found with <15 days of fever (6.35 μg/ml). No significant differences were observed in median CRP levels between patients with diabetes (8.65 μg/ml) and those without diabetes (19.05 μg/ml). We observed significant differences in the levels of both CALP and CRP between acute melioidosis patients and healthy controls (P < 0.0001) and between nonmelioidosis sepsis patients and healthy controls (P < 0.0001).

FIG 1.

Serum inflammatory marker changes in melioidosis patients during acute and eradication phases of antibiotic treatment. Each symbol represents one melioidosis patient. Horizontal lines represent median concentration of each biomarker. Production of CALP (a) and CRP (b) in the serum shows a decreased trend, whereas CALR levels (c) show none. Data were analyzed using the Mann-Whitney U or Wilcoxon signed-rank test. ****, Significant at P < 0.0001; NS, not significant.

In contrast to the serum levels of CALP and CRP, serum levels of CALR did not differ significantly across the two treatment phases of melioidosis, with median levels of 4.00 ng/ml (IQR, 3.06 to 8.25 ng/ml) for the acute phase and 4.14 ng/ml (IQR, 2.53 to 6.35 ng/ml) for the eradication phase. Differences in biomarker levels between study groups are summarized in Table S1 in the supplemental material. The percent reduction of CALP, CRP, and CALR after antibiotic treatment was 71.17%, 71.15%, and 15.14%, respectively; this is another measure that can be used for prognosis. No correlations were seen among the levels of all three biomarkers (data not shown). Receiver operating curve analyses were performed to assess the prognostic utility of CALP and CRP. For acute melioidosis compared to sepsis caused by other infections, the area under the curve (AUC) was 0.75 for CALP and 0.57 for CRP (see Fig. S2 in the supplemental material), indicating that CALP is better than CRP as a prognostic marker for melioidosis.

DISCUSSION

This is the first report, to our knowledge, describing CALP as a potential biomarker for responses to antibiotic treatment for melioidosis. We examined correlations of three inflammation biomarkers with the antibacterial treatment phases in disease management. Although serum levels of CALP and CRP dropped by >70% in the eradication phase of antibiotic treatment, CALP was a better prognostic marker for melioidosis. In agreement with earlier reports showing elevated levels in acute melioidosis cases (15, 16), CRP levels were also elevated in the acute treatment phase and were significantly reduced in the eradication phase. Because type 2 diabetes is the most significant risk factor associated with melioidosis, we noted that the diabetic subgroup of melioidosis patients had higher CALP levels than found in the nondiabetic group. In contrast, CRP levels were higher in nondiabetic than in diabetic melioidosis patients. The distribution of patients with diabetes (74%) among the Sri Lankan study cohort of melioidosis patients that we examined was similar to that of cases reported in India (76%) (17), Thailand (60%) (18), and Australia (42%) (19). In addition, the AUC values for CALP were better than for CRP (0.75 versus 0.57) in melioidosis compared with nonmelioidosis sepsis cases, suggesting that CALP may provide better specificity than CRP. A detailed time-dependent analysis of CALP levels through complete disease resolution may further narrow the diagnostic window. Other biomarkers that were previously explored for melioidosis include the cytokines interleukin 6 and 10 and tumor necrosis factor-α for predicting mortality (20), CRP for diagnosis (15, 16), and endothelial modulator proteins C and S along with antithrombin for replacement therapy (21). This is the first study to examine the relationship between patient responses to antibiotic therapy for melioidosis and disease biomarkers, most specifically CALP.

Although more cases of melioidosis are being discovered as a result of active surveillance (22), melioidosis is a neglected disease that is underreported in many countries in which it is endemic. Treatment requires long and costly administration of antibiotics, and few antimicrobials are effective due to the intrinsic resistance of B. pseudomallei (2, 23). The potential spread of acquired resistance (24, 25) and the lack of any new promising drugs in development are ominous circumstances in regard to the future of melioidosis disease management. As generally only clinical symptoms are used for disease management, there is an unmet need for biomarkers that can help to monitor disease progression and guide antibiotic treatment. For example, procalcitonin, a sepsis biomarker, is useful for implementation of antibiotic stewardship programs (26). A recent trial showed a 7% survival benefit from use of calcitonin monitoring as guidance in managing critically ill patients (27). Procalcitonin also correlated with severity of B. pseudomallei infection (28).

Our observations with CALP have implications for the management of melioidosis. CALP is a blood-based biomarker that may be used to evaluate antibiotic responses, since blood cultures are likely to be negative, and it may help determine whether to stop or modify antibiotic treatments. Commercial FDA-approved assays are available for fecal CALP measurement, although none are labeled for serum determinations. It may also be possible to develop low-cost assays for CALP measurement in serum that could be used in developing countries with the highest melioidosis disease burden.

MATERIALS AND METHODS

Patient enrollment.

Melioidosis patients >18 years old were enrolled in this study. Matched blood serum samples were collected during the acute and eradication phases of the antibiotic treatment. Control serum samples were collected from nonmelioidosis sepsis patients and healthy volunteers. Ethics approval was obtained from the Ethics Review Committee, Faculty of Medicine, University of Colombo, Sri Lanka; Office of Human Use and Ethics (OHU&E) of the U.S. Army Medical Research Institute of Infectious Diseases (USAMRIID); and the U.S. Army Medical Research and Material Command, Office of Research Protection, Human Research Protection Office (USAMRMC-ORP-HRPO).

Confirmation of B. pseudomallei infections.

Patients with presumptive melioidosis clinical presentations (fever, pneumonia, sepsis, and abscess) were selected for B. pseudomallei screening. Conventional culture techniques were used to isolate and identify B. pseudomallei from blood and other patient specimens as previously described (22). An RT-PCR assay was used to confirm B. pseudomallei isolates. The RT-PCR probes targeted a B. pseudomallei homologue of the Salmonella enterica serovar Typhimurium lpxO gene, Yersinia-like fimbrial, and Burkholderia thailandensis-like flagellum and chemotaxis gene clusters, as previously described (22).

Animal studies.

Adult rhesus macaques (Macaca mulatta) were infected with B. pseudomallei strain HBPUB 10134a (29) by an estimated inhaled dosage of ∼357 CFU (range, 261 to 531 CFU). Blood samples were collected before and after exposure at various time points. Research was conducted under a protocol approved by the Institutional Animal Care and Use Committees in compliance with the Animal Welfare Act, Public Health Service Policy, and other federal statutes and regulations relating to animals and experiments involving animals. The facility (USAMRIID) where this research was conducted is accredited by the Association for Assessment and Accreditation of Laboratory Animal Care International and adheres to principles stated in the Guide for the Care and Use of Laboratory Animals (30). Animal care was provided in accordance with these established guidelines.

Biomarker assays.

Serum levels of CALP, CALR, and CRP were examined by using commercially available enzyme-linked immunosorbent assay (ELISA) kits from DRG International, Inc. (Springfield, NJ, USA), LifeSpan BioSciences, Inc. (Seattle, WA, USA), and Invitrogen (Frederick, MD, USA), respectively. The assays were performed in duplicate according to the manufacturer's instructions, using serum dilutions of 1:50 (CALP), 1:5 (CALR), and 1:5,000 (CRP) in test kit buffers. The ELISA absorbance was read on a microplate reader (Infinite 200 Pro; Tecan, Morrisville, NC, USA). Quantitation of biomarkers in serum was calculated by generating a standard curve with recombinant human CALP, CALR, and CRP provided by the manufacturers. Serum concentrations of biomarkers were estimated with GraphPad Prism software (GraphPad Software, La Jolla, CA, USA), by interpolation from the 4-parameter logistic regression algorithm of the standard curve.

Statistical analysis.

The nonparametric Wilcoxon signed-rank test, Student's t test for paired samples, and the Mann-Whitney U test for unpaired samples were performed to compare groups (GraphPad Prism version 7.0). Outliers were identified using the robust regression and outlier removal method (31), with Q set to 1%, and statistical significance was established at P < 0.05.