Changes of C-reactive protein and Procalcitonin after four weeks of treatment in patients with pulmonary TB

Abstract

Objective

Tuberculosis (TB) remains a public health concern worldwide, affecting millions of people every year. Detailed characterization of disease pathophysiology is key to proper diagnosis, disease progression, or treatment follow-up and evaluation. The present study investigated C-reactive protein and Procalcitonin (PCT) as candidate markers of early treatment response and disease activity.

Methods

From September to December 2019, 21 HIV-negative consecutive TB patients were recruited, within the setting of the Gabonese TB specialized hospital and the National Laboratory of Public Health, in a prospective study. CRP and PCT levels were measured by chemiluminescence at diagnosis and 4 weeks following the initiation of anti-TB treatment.

Results

The mean concentration of CRP in TB patients was 114.7 mg/L (95 % CI: [83.8–145.6]) at diagnosis and 20.2 mg/L (95 % CI: [14.1–26.4]) 4 weeks following anti-TB treatment. The drop in CRP concentrations between diagnosis, and week 4 following anti-TB treatment showed was significant (p < 0.0001). The average concentration of PCT at the time of diagnosis was 0.3 ng/mL (95 % CI: [0.19–0.41]). PCT Concentration dropped below 0.05 ng/mL 4 weeks following the start of anti-TB treatment (p < 0.01).

Conclusion

CRP and PCT are potential TB biomarkers, each, carrying important keys. If the drop in both proteins may indicate a significant reduction of the Mtb burden, the maintenance of CRP above the inflammation threshold could indicate the presence of residual bacilli. However, the clinical translation of the present finding will require more investigation.

1. Introduction

Tuberculosis (TB) continues to be a public health concern worldwide. An estimated 10 million people got active TB worldwide.[1] For 20 years the disease kills more than one (1) million people per year.[2] 1.5 million people died from TB in 2020.[1] Numerous studies have contributed to our current understanding of the complex biology of pulmonary tuberculosis and have subsequently provided evidence that immune or biochemical parameters involved in TB response could serve as diagnostic, disease activity, and progression biomarkers.[3], [4], [5], [6], [7] The search for TB biomarkers is ongoing.[8], [9] In the present study we investigated C-reactive protein (CRP) and Procalcitonin (PCT) as candidate markers for early TB disease activity.

C-reactive protein and Procalcitonin have been used as diagnostic markers of bacterial sepsis.[10], [11], [12] Although CRP has been previously investigated in TB, this is the first study looking at the potential value of PCT in tuberculosis disease.

2. Methods

We conducted a longitudinal prospective study, evaluating the change in CRP and PCT between diagnosis and the first month of the anti-TB standard treatment intensive phase. During this intensive phase all patients received isoniazid (INH), rifampin (RIF), pyrazinamide (PZA), and ethambutol (EMB) as recommended.[13].

Between September and December 2019, consecutive HIV-negative TB patients were recruited within the setting of the Gabonese TB specialized hospital and the National Laboratory of Public Health. TB disease was diagnosed by sputum smear and Xpert® MTB/RIF assay (GeneXpert, Cepheid – Europe).

For each patient, sputum smear tests were done and four (4) ml of blood was collected at diagnosis and one (1) month following the initiation of anti-TB therapy. Collected blood samples were transported for serum processing and analysis at the laboratory of the Libreville Mother and Child University Hospital. CRP levels were measured using the Cobas CRP test (on Cobas C111 analyzer, Roche - France) and PCT levels were measured using the VIDAS® B.R.A.H.M.S. PCT assay (Biomerieux - France). The data obtained were then analyzed using Graph Pad Prism Software.

2.1. Ethics

The research was done following Gabonese ethical guidelines and regulations, and approval was obtained from the Gabonese national ethics committee and registered under the number: PROT-N°0089/2019/PR/SG/CNER. Written informed consent, was obtained for all participants before they were enrolled in the study.

3. Results and discussion

A total of 21 HIV-negative TB patients were recruited (81 % men and 18 % women) aged between 28 and 57 years old (median age: 32 years). The Xpert® MTB/RIF assay showed that all patients had drug sensitive Mtb strains. 11 patients were lost during follow-up, and 10 were seen at month 1. Seven (7) patients were sputum negative after one (1) month of treatment, whereas, three (3) remained positive (Table 1).

Table 1.

Mycobacterium Tuberculosis (Mtb) load, CRP, and PCT concentration at diagnosis and four (4) weeks after the initiation of anti-tuberculosis (anti-TB) standard treatment.

			Diagnostic (pre-treatment)			After 4 Weeks of anti-TB treatment
ID	Sex	Age	Bacilli load	CRP (mg/L)	PCT (ng/mL)	Bacilli load	CRP (mg/L)	PCT (ng/mL)
P1	M	30	2+	115,42	0,17	0	7,22	<0,05
P2	M	34	3+	82,94	0,3	1+	19,1	<0,05
P3	M	30	2+	180,09	0,12	0	19,23	<0,05
P4	M	27	2+	98,34	0,32	0	17,4	<0,05
P5	M	48	1+	134	0,5	0	14,34	<0,05
P6	M	57	1+	176	0,27	0	34,18	<0,05
P7	F	41	3+	43,78	0,27	1+	17,75	<0,05
P8	M	28	2+	90,14	0,5	0	14,89	<0,05
P9	M	28	2+	85,83	0,5	1+	34,45	<0,05
P10	M	51	2+	140,57	0,07	0	23,96	<0,05
P11	M	41	1+	142,5	1,15	No follow-up
P12	F	35	1+	99,92	0,05
P13	F	28	2+	70,08	0,05
P14	M	32	3+	107,7	0,07
P15	M	27	2+	132,73	0,05
P16	F	28	3+	184	0,17
P17	M	34	2+	57,23	0,05
P18	M	41	3+	189,94	0,07
P19	M	38	2+	85,31	0,12
P20	M	28	3+	14,34	0,06
P21	M	28	3+	44,08	0,06

The median and the mean levels of CRP in TB patients at diagnosis were respectively 99.92 mg/L (IQR: 71) and 108.3 mg/L (95 % CI: [83.8–130.9]). This level corresponded to the levels observed in bacterial infection.[14], [15] 4 weeks after the initiation of treatment, CRP median and mean concentration dropped respectively to 18.43 mg/L (IQR: 11.77) and 20.25 mg/L ((95 % CI: [14.1–26.4]). Both non-parametric and parametric comparisons of the circulating CRP concentrations between diagnosis, and 4 weeks following anti-TB treatment showed that the drop in CRP was significant (p < 0.0001) (Fig. 1). Further analysis showed a significant correlation between bacilli count or presence and the circulating concentration of CRP in patients (p = 0.007) (Fig. 2a). Bacilli-positive patients had a significantly higher level of CRP (p < 0.0001) (Fig. 2b). The receiving operator curve (ROC) analysis confirmed that CRP might be useful in discriminating bacilli-positive patients (AUC: 0.93; p = 0.0006) (Fig. 2c). Although the levels of CRP decreased drastically after one (1) month of treatment, the levels remained above the inflammation threshold, [16], [17] even in patients with negative sputum smears. This may be the reflection of the underlining activity of non-sterilized bacilli, as an optimal Mycobacterium Tuberculosis (Mtb) sterilization requires a 6 months regimen.

Fig. 1.

The concentration of CRP at diagnosis and 4 weeks after the initiation of treatment. (a) Unpaired (Mann Whitney) t-test; (b) illustration of the paired t-test. The statistical significance was defined as p < 0.05.

Fig. 2.

(a) Correlation between bacilli count and the circulating concentration of CRP in patients (r = 0.47; p = 0.007). (b) Bacilli-positive patients had a significantly higher level of CRP (p < 0.0005). (c) The receiving operator curve (ROC) analysis of CRP as indicator of bacilli-positivity (AUC: 0.93; p = 0.0006).

Also, CRP being an opsonizing factor, its production should reflect bacilli presence. It has been reported that a high level of CRP after anti-TB drug therapy was associated with a delay in bacilli negativation.[18] Although not investigated here, others have shown the return of CRP to its normal levels after 6 months. [19] Moreover, studies have shown CRP to have a good screening performance for active pulmonary tuberculosis in both HIV-positive and negative patients. [4], [20], [21], [6] This comfort CRP as an indicator of disease activity.

PCT has been investigated for different conditions and diseases. [10], [11], [22], [23], [24], [25], [26], [27], [28], [29], [30], [31], [32] However, this is one of the two studies investigating the level of PCT in tuberculosis. At diagnosis, all TB patients had their procalcitonin (PCT) concentrations between 0.05 and 1.15 ng/ml (median level: 0.12 ng/ml (IQR: 0.25)), which is above the normal range. 95 % (20 out of 21) of patients had a level of PCT corresponding to localized infection (0.05–0.5 ng/ml). Only one patient had a level of PCT in the sepsis range (1.15 ng/ml). The average concentration of PCT at the time of diagnosis was 0.3 ng/mL (95 % CI: [0.19–0.41]). In all patients seen after 4 weeks of anti-TB treatment, PCT concentration dropped below 0.05 ng/mL (which is the normal range). Both non-parametric and parametric analysis showed that the level of PCT at diagnosis is significantly higher than its level 4 weeks following anti-TB treatment (p < 0.0001 and p = 0.003 respectively) (Fig. 3). Correlation analysis showed that the circulating concentration of PCT correlates significantly with bacilli count (p = 0.0007) (Fig. 4a). Bacilli-positive patients had a significantly higher level of PCT (p < 0.0001) (Fig. 4b). The ROC analysis also showed that PCT might be useful in discriminating bacilli-positive patients (AUC: 0.93; p < 0.0001) (Fig. 4c).

Fig. 3.

The concentration of Procalcitonin at diagnosis and 4 weeks after the initiation of treatment. (a) Unpaired (Mann Whitney) t-test; (b) illustration of the paired t-test. The statistical significance was defined as p < 0.05.

Fig. 4.

(a) Correlation between bacilli count and the circulating concentration of PCT in patients (r = 0.57; p = 0.0007). (b) Bacilli-positive patients had a significantly higher level of PCT (p < 0.0001). (c) The receiving operator curve (ROC) analysis of PCT as indicator of bacilli-positivity (AUC: 0.93; p < 0.0001).

It has been suggested that the concept of predictive models using combinations of non-specific host markers may be valid. [4], [9], [33] The challenge is to identify the strongest markers that will contribute to such models. [8], [34] Reports showed that PCT levels correlate with pneumonia severity and prognosis, improving the prognostic power of severity scoring systems. [35], [36], [37] Also, PCT is being used not only as a marker of bacterial sepsis but also as an indicator to guide the decision to initiate and stop antibiotic treatment. [38], [39] Furthermore, here we showed that PCT and CRP dropped with Mycobacterium Tuberculosis (Mtb) load (Table 1) and represent candidate markers for active TB disease and treatment response.

The low number of patients and the high rate of patients lost to follow-up characterized the present study. Therefore, despite the observed statistically, significant correlations between the selected markers (PCT and CRP) and bacilli load, the translation of these results into clinical use will require more investigations.

In conclusion, CRP and PCT are potential active TB biomarkers, that may carry important information on inflammation and bacilli load. However, further investigation is needed to establish PCT as a marker of TB treatment response and disease prognosis. Also, the value of associating PCT and CRP for the diagnosis of TB severity and prognosis needs to be investigated.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.