Effect of Pulmonary Tuberculosis on Protein C, S, and Antithrombin-III among Therapy-naïve Ghanaian Adults; A Comparative Cross-Sectional Study

Background. Tuberculosis (TB) constitutes a global emergency as it affects one-third of the world’s inhabitants. Although pulmonary Tuberculosis (PTB) is curable, immunological responses to the infection induce several haematological derangements. This study evaluated the effect of PTB on Protein C, Protein S, Antithrombin-III, and blood count parameters. Methods. Ninety adults with ages ≥ 18 years were purposively recruited: 60 PTB patients and 30 non-TB controls. All patients were diagnosed with sputum GeneXpert MTB/Rif assay. Blood specimens were collected from each participant for Protein C, S, Antithrombin-III and complete blood count. Results. Pulmonary TB was associated with signicantly reduced Protein C activity (101.46 [87.61-128.3] vs 121.44 [99.50-149.8] IU/L, p=0.038), RBC (3.88±0.91 vs 4.80±0.55, p<0.0001), HgB (10.24±2.42 vs 11.78±1.42, p=0.0019), HCT (32.21±7.79 vs 42.05±4.97, p<0.0001), MCV (83.80 [79.33-90.08] vs 89.00 [83.75-92.00], p=0.0133) and PDW (12.95 [10.73-15.00] vs 15.30 [14.18-15.93], p<0.0001) compared to controls. Conversely, PTB patients were associated with signicantly increased MCH (26.83±4.33 vs 24.59±1.99, p=0.0086), TWBC (7.76 [6.06-9.78] vs 6.50 [4.85-7.50], p=0.0047), Abs. GRAN (5.27 [3.30-6.71] vs 3.75 [2.48-4.75], p=0.0226), RDW-CV (13.70 [13.20-15.43] vs 12.95 [12.50-13.65], p<0.0001), MCHC (32.10 [28.70-35.63] vs 27.85 [27.40-28.53], p<0.0001) and MPV (8.3 [6.7-9.7] vs 7.0 [6.4-7.5], p=0.0027) compared to controls. The PTB patients were disproportionately affected with anaemia (91.7%, p=0.001), erythrocytopenia (75.0%; p ≤ 0.001) and reduced HCT (80.0%, p ≤ 0.001). The frequency of thrombocytosis, leucocytosis, and granulocytosis (50.0%, p=0.013; 23.3%, p=0.013; 18.3%, p=0.025; respectively) in PTB patients were signicantly higher than in controls. Conclusion. Our ndings suggest that PTB predisposes patients to hypercoagulability, with a signicant reduction in Protein C activity but not Protein S and antithrombin-III. The condition causes derangements in erythrocytes, leucocytes, and thrombocytes, and disproportionately causes anaemia. Protein C activity and complete blood count are useful in the management of PTB and should be included in the routine workup for patients.


Introduction
Tuberculosis (TB) is a chronic and infectious disease [1] caused by a group of bacteria that belong to the Mycobacterium tuberculosis complex (MTBC) [2,3]. Mycobacterium tuberculosis is the preeminent species of the MTBC known to cause TB among humans [4]. In humans, TB disproportionately (70%) infects the lungs resulting in pulmonary tuberculosis (PTB) [5], and is acquired when aerosols containing tubercle bacilli from an infected person are inhaled by a susceptible host [6]. Worldwide, the estimated tuberculosis burden is 1.86 billion cases, which approximates to 32% of the global population, with nearly nine million new infections recorded each year [7]. Tuberculosis burden in Africa is second after Asia and accounts for about 26% of the world's cases [8]. Also, it is reported to be the second most signi cant course of death in countries with poor and low economies [9] like Ghana. Such countries are known to harbour the greatest burden of cases and deaths (about 95% and 98%, respectively) associated with TB [10]. In Ghana, TB infects approximately 46,000 people every year [11] and more than 50% of them are adults [12].
Pulmonary TB induces haemostatic changes which are ascribed to the upregulation of procoagulant factors, diminished levels of anticoagulants and brinolytic factors [13]. Consequently, these altered processes culminate in increased coagulability [14] which may, in turn, lead to the formation of clots in deep veins (DVT) and obstruction of the blood ow in blood vessels [15]. A blood clot in deep veins is more frequent in older adults, and its rate of occurrence has a positive correlation with age [16]. It is estimated that 1.5% to 3.4% of TB infected patients suffer from such vascular complications [17] and could even exceed 10% since the exact incidence is not known [18].
Although some studies have explored haematological derangements among PTB patients in other parts of the world, most of these studies only highlighted derangements in blood clotting pro les and or platelet count, while they failed to include tests to measure activities of natural anticoagulants like protein C, S, and antithrombin-III and also, the full complement of the complete blood count (CBC) parameters. For instance, studies by Eteudo et al. [13] and Kutiyal et al. [19] assessed coagulation pro les but failed to measure anticoagulant activities. In Ethiopia, Kassa et al. [20] and Kahase et al. [21] measured only blood count parameters. Despite the thrombotic complications associated with PTB management of patients with the disease in Ghana does not include haematological assays to assess protein C, S and antithrombin-III activities, and CBC which could provide information on their risk of thrombosis and associated sequelae. Thus, understanding the effect of PTB infection on natural anticoagulant activities among Ghanaian adults remains elusive.
This study, therefore, sought to determine the effect of PTB on protein C, S, and Antithrombin-III activities and CBC parameters among Ghanaian adults. This was to determine the usefulness of natural anticoagulant activities and CBC in the management of PTB.

Study design/Study site
This was a hospital-based comparative cross-sectional study conducted at the Nkawie-Toase Government Hospital, Ghana between November 2019 and May 2020. The hospital is among a selected few within Ashanti Region that operates a GeneXpert molecular-based rapid diagnostic analyser and has been designated a tuberculosis diagnostic referral centre by the National Tuberculosis Control Programme (NTP), Ghana. In addition to analysing sputum specimens collected at the hospital, the hospital's laboratory is also responsible for diagnosing TB in sputum specimens received from within the Atwima-Nwabiagya municipality as well as those referred from adjoining districts. The municipality is located on coordinates 6°40'N 1°49'W.

Study population
Ninety (90) participants with ages ≥18 years were enrolled in the study using purposive sampling. This comprised 60 cases selected from newly diagnosed pulmonary TB patients who were naïve to antituberculosis therapy and 30 healthy control subjects. Each participant was assessed for eligibility using a pre-tested closed-ended questionnaire. Among information captured with the questionnaire were sociodemographic data such as gender, age, residence, education, occupation, and history of smoking and alcohol intake. Diagnosis of all cases was done using chest radiograph and participants whose results suggested an abnormality were made to produce sputum for con rmation and semi-quanti cation with GeneXpert molecular-based assay (MTB/Rif assay). Individuals with malaria, HIV, viral hepatitis, coagulopathies, and history of contraceptives and/or anti-in ammatory drug usage were excluded.

Sputum analysis
Freshly expectorated sputum specimen was diluted three folds (1 in 3 dilutions) by adding two volumes of the sample reagent to one volume of sputum. The sputum-sample reagent mixture was vigorously shaken for 10 seconds using the Vortex-Genie 2 mixer (Scienti c Industries Inc., U.S.A) and incubated at room temperature for 10 minutes. The mixture was shaken a second time and further incubated for an additional 5 minutes. Once liquefaction of the sputum was complete the Xpert MTB/Rif cartridge was labelled and carefully opened without touching the rear side. A plastic bulb pipette was used to transfer 2ml of the lique ed sputum into the specimen chamber of the cartridge while avoiding the formation of air bubbles. The cartridge was then placed in and analysed with a four-channel GeneXpert rapid point of care molecular-based analyser (Cepheid, Sunnyvale, CA, U.S.A).

Blood specimen collection and analysis
Approximately 4.0 millilitres (mL) of whole blood was collected aseptically by venipuncture from each participant using evacuated blood collection tubes containing 3.8% sodium citrate and tri-potassium ethylenediaminetetraacetic acid (K 3 EDTA) anticoagulant (BD, Becton, Dickinson and Company, USA). The blood-containing tubes were gently inverted about 8 times to allow adequate mixing of anticoagulants and blood. The K 3 EDTA blood was immediately used to measured complete blood count using the Horiba ABX Micros ES 60 automated haematological analyser (Horiba ABX, France). Citrated blood was centrifuged at 3000 revolutions per minute (rpm) for 15 minutes, and the plasma used to measure protein C, S, and antithrombin-III activities by solid-phase sandwich enzyme-linked immunosorbent assay

Statistical data analysis
The data was inputted into Microsoft Excel 2016 and transferred into IBM SPSS Statistics 23 for recoding and analysis. Visualization of the data was performed using GraphPad Prism 8.4.3. All data except for age, red blood cell, haemoglobin, haematocrit, mean cell haemoglobin, and absolute lymphocyte count were signi cantly skewed and non-parametrically distributed (p≤0.05). The non-parametric data were presented as median and interquartile ranges (IQR), and parametric as mean and standard deviation (SD). Means of PTB and controls were compared using Levene's test for equality of variance and t-test for two independent groups. The Mann-Whitney's U test was used to compare non-parametric distributions between PTB cases and controls. Chi-square and Fisher's Exact tests were used appropriately to test association and differences in proportions between categorical data.

Ethical consideration
The study was approved by the Committee on Human Research, Publication and Ethics (CHRPE) of the School of Medicine and Dentistry at the Kwame Nkrumah University of Science and Technology, Kumasi (reference number: CHRPE/AP/245/20). Also, written permission was obtained from the management of Nkawie-Toase Government Hospital, and all participants consented by signing or thumb printing a consent form before enrolment.

Results
Sociodemographic characteristics of the participants Table 1 shows frequency and differences in proportions of socio-demographic variables strati ed by controls and pulmonary TB cases.
Of the 60 PTB patients, 30   Levels of protein C, S, and antithrombin-III activities of participants Haematological pro le of participants   Frequency of haematological derangements in the study participants Figure 2 shows the frequency of various haematological derangements in the study participants strati ed by PTB cases and controls. Increased percentage of the cases were signi cantly associated with leucocytosis (high TWBC) compared to controls ( Thrombocytosis was present in 50% (30/60) of the cases, whereas 30% (9/30) of the controls showed the derangement. Also, thrombocytopenia was present in 15.0% (9/60) and 3.3% (1/30) of cases and controls respectively; the association was signi cant (p=0.013). Mean platelet volume was mostly normal, with 31.7% (19/60) and 48.7% (14/30) of the cases and controls respectively, showing low values. Low PCT was observed in 21.7% (13/60) and 6.7% (2/30) of cases and controls respectively, with only 1.7% showing high values. Pulmonary TB was signi cantly associated with PDW (p=0.005), but not MPV and PCT (p=0.317 and p=0.144, respectively) ( Figure 2).

Discussion
Tuberculosis (TB) remains the most frequent cause of mortality among adults presenting with transmissible diseases, and although treatment exists, affected individuals succumb to complications like thromboembolism and formation of clots in deep veins [22]. Consequently, we determined the effect of PTB on the level of activities of protein C (PC), protein S (PS), and antithrombin-III (AT-III) in therapynaive Ghanaian adults. Also, we reported the haematological pro le in PTB patients. The cases were signi cantly older than controls with half of the infections within the 38-57-year group, which was consistent with ndings of Wangdi et al. [23]. This association between old age and TB is ascribed to the vulnerability that comes with waning immunity in older people [24,25]. Also, ageing causes immune depletion through the diminished function of the thymus and decreased propensity to synthesize cytokines and replace senescent T lymphocytes [25]. Also, more males were infected than females, and this increased susceptibility of the male gender to TB in the current study corroborates ndings of Mukherjee et al., [26]. This is because, oestradiol in females induces a cellular immune response by stimulating interferon-gamma to activate macrophages in response to the infection [27,28] whereas, testosterone which is higher in males, inhibits the body's immunity [28].
This study showed a signi cant reduction in the activity of PC but not PS and AT-III among PTB cases compared to controls, and the magnitude of the effect of PTB on PC activity was medium. This was consistent with ndings of Turken et al. [29] and Kager et al. [15]. Marlar and Gausman [30] posit that although estimating the level of these biomarkers does not always suggest venous thromboembolic tendencies (phenotype), correctly determining the level of even a single biomarker in plasma is pathognomic of the gene expression (genotype). Our ndings, therefore, a rms the early body of knowledge which suggests that PTB predisposes patients to increased coagulability and associated sequelae.
The study by Turken et al. [29] reported a signi cant decline in PC activity in PTB patients who were naïve to antituberculosis chemotherapy. Similarly, Kager et al. [15], reported signi cantly decreased concentrations of PC, antithrombin, and unbound PS in different groups of TB patients, and attributed these derangements to reduced anticoagulant processes that accompany TB infection. For instance, the reduction in PC activity observed in the present study is thought to result from negative regulation of the processes that culminate in PC synthesis which occur mainly via the inability to copy endothelial cell protein C receptor and thrombomodulin, thus, compromising the potential to activate PC [31]. Also, low production, high consumption and leakage that result from endothelial cell damage can cause a decline in PC levels [32].
Furthermore, Kager et al. [15] explained the anomaly in PS as been because, the majority of PS in the systemic circulation is bound to C4b, a complement factor that increases greatly in TB and therefore the reduced concentration of unbound protein S is justi ed to be the result of a rise in systemic C4b that occurs in TB patients. Similar to PC, PS is a vitamin K-dependent protein synthesized primarily within liver cells [33,34], as a result, altered function of the liver will impair its production. Also, more than half (53.3%) of our cases had recent or past history of alcoholism, a situation which can cause alcoholinduced liver in ammation and damage and consequently de ciency in protein S. This may have contributed to the low activity of protein S observed among PTB cases in this study.
Antithrombin, on the other hand, becomes used up and deactivated by in ammation-mediated degradation [31]. Also, immune mechanisms to in ammation cause the liberation of cells like neutrophils and associated cytokines to additionally diminish the activity of antithrombin [31].
The majority of blood count parameters investigated in this study were signi cantly deranged among PTB patients compared to controls. We observed signi cantly reduced RBC count, HgB, and HCT among the PTB patients than controls, which corroborate ndings of studies from Nigeria [35], Ethiopia [21] and India [36]. However, this was in contrast to ndings from a longitudinal prospective study in Ethiopia by Kassa et al. [20] who reported rather high HgB and HCT values in treatment-naïve TB patients. The most plausible explanation is that Kassa et al. [20] included both pulmonary and extra-pulmonary TB patients and also compared values of untreated patients with those receiving treatment, but not with healthy controls as found in our study.
Anaemia has been identi ed as a frequent occurrence and known determinant for loss of lives in TB patients [36] and was by far the most distinctive haematological derangement observed among the patients in the present study. Approximately 92% (91.7%) of the PTB patients were anaemic, 75.0% had erythrocytopenia and 80.0% had low red cell mass. These corroborate ndings of Olaniyi et al. [37] and Padhariya and Shah [38] in which anaemia was observed among 90.3% and 88% of TB patients respectively. Also, the reduced HgB in PTB has been reported in other studies around the globe [39,40]. This PTB-associated anaemia is attributed to the involvement of cytokines, which interfere with the uptake, release and storage of iron culminating in a reduction in the amount of iron in the blood, and consequently reduces its accessibility to the erythrocytes for the production of haemoglobin [20,35,41]. During in ammation, there is an upsurge in the synthesis of cytokines like IFN-gamma, TNF-alpha, and interleukins (IL)-1, IL-6 [20] and IL-10 [42]. The interplay of these cytokines in the pathogenesis of anaemia in chronic disease like PTB has been widely reported in other studies. Interleukin 6, liberated by cells of the innate immunity, induces the synthesis of hepcidin which hampers uptake of iron from the duodenum [42] and favours the synthesis of ferritin, and together these events retain more iron in storage than is required [41]. Furthermore, the upsurge in TNF-alpha and IL-1 directly inhibits renal synthesis and release of erythropoietin, which is a growth-stimulating hormone for the production of erythrocytes; production of erythroid progenitor cells in the bone marrow is thus inhibited leading to impaired synthesis of red cells and haemoglobin, with the net effect of anaemia [42].
Also, TNF-alpha and gamma-interferons induce the synthesis of unstable superoxide and nitric oxide species which can cause unmediated injury to progenitor erythroid cells [41,43]. This hampers the expansion of progenitor population of the erythroid lineage and further result in abnormal synthesis and function of the red cells' stimulatory hormone (erythropoietin) which cause the bone marrow to react poorly [20] leading to reduced red cell mass, haemoglobin level and anaemia as the overall effect. Furthermore, the anaemia in TB has been linked to the longevity of Mycobacterium tuberculosis through its ability to directly metabolize heme to generate iron for its utilization and as a result causes a reduction in haemoglobin [21]. The erythrocytopenia and reduced red cell mass may also be ascribed to phagocytosis of red blood cells (erythrophagocytosis) which results in reduced survival of the cells and the inability for the bone marrow to replenish senescent cells due to the depleted serum iron [41].
In the initial phase of anaemia in chronic diseases like TB recycling of iron is modestly challenging and may manifest in most cases as moderate anaemia with normocytic erythrocytes, but as the disease progresses there is diminished uptake of iron from the intestines which results in a marked depletion of iron and a characteristic microcytosis of the erythrocytes [41]. This was evident in the present study by the signi cantly increased proportion of the cases that had microcytic erythrocytes, and also, increased variability in red cell size (RDW) which could be pathognomic of iron de ciency [20] among the TB cases.
These ndings were consistent with some recent studies [21,35], but contrasting to ndings of other studies from Nigeria by Akpan et al. [44] and Okafor et al. [45], plausibly because their studies recruited subjects with heterogeneous ages that encompassed both children and adults whereas we considered entirely adults.
The blood of PTB cases showed normal haemoglobinization (normochromasia) which was evident by the overall increased MCH and MCHC observed, and the increased proportion of the patients that recorded normal-to-high MCH and MCHC estimates. This corroborates a study by Kassa et al. [20], and was also consistent with a review study by Weiss and Goodnough [42], which posits that anaemia associated with chronic diseases presents with normocytic normochromic blood ndings.
The nonspeci c innate and cellular immunity remains the foremost preventive mechanism against the spread of M. tuberculosis [46]. This leads to the formation of a granuloma, the hallmark of the infection which causes an immune equilibrium between the bacillus and affected host [47]; it is characterised by increased leucocyte populations like macrophages and lymphocytes that encapsulate the microorganisms within the lungs. This natural response of leucocytosis explains the signi cantly increased TWBC count and the increased proportion of the cases that showed leucocytosis. Furthermore, as the disease becomes chronic, increased synthesis of cytokines by T lymphocytes and macrophages occur leading to increased maturation of lymphocytes [41]. This was evident by the increased absolute lymphocyte count and the predominantly normal values observed among the cases, although not signi cant. However, signi cantly reduced absolute monocyte counts were observed among the cases, which may plausibly be the result of increased differentiation of monocytes to macrophages in response to the increased need to contain the tubercle bacilli and subsequently, their entrapment in granuloma shells. This may in turn lead to the formation of more monocyte-derived macrophages leading to reduced circulating monocytes. Neutrophils form the bulk of the granulocytes and are also usually the rst cell type to arrive at the affected area [21]. This early immune response, therefore, justi es the signi cantly elevated granulocytes in PTB observed in this study.
Platelet count and PCT were increased among the PTB patients than controls although not signi cant, and was consistent with ndings of Eteudo et al. [13]. Fifty per cent of the patients showed thrombocytosis (p=0.013), which corroborate ndings from other studies [40,48,49]. Most in ammatory diseases including PTB favour reactive thrombocytosis through acute phase response [48]. During this phase of the infection, IL-6 and TNF-alpha are among several cytokines that are synthesized and released in large quantities, and these proteins induce the development of thrombocyte-producing precursors and consequently, increased synthesis of platelets in the blood [50]. However, Renshaw and Gould [51] posit that reactive thrombocytosis may not be apparent in all TB patients and that reporting on the increased thrombocyte count in TB patients is more relevant in already developed countries where exacerbation of the infection by other comorbidities is unlikely.
Mean platelet volume provides information on platelet size and function, and in most cases posit that bserved patients were unknown. has been useful in evaluating the reaction to therapy. Unlike PLT and PCT, MPV was signi cantly elevated in PTB patients and was consistent with ndings of Tozkoparan et al. [39] who ascribed the derangement to elevated thrombopoietic activity. Conversely, PDW was signi cantly decreased in the cases (p<0.0001) and the proportion of cases that had low PDW was signi cantly decreased than in the control group (p=0.005). This suggests that there existed some form of platelet anisocytosis in the PTB population. Together, PDW, PCT, and MPV provide information on platelet activation/function; the elevated PDW, therefore, could represent changes in the surface area of the platelets following their activation. This is due to the immune involvement of platelets in active TB disease [52].

Conclusions
Our ndings suggest that PTB predisposes patients to hypercoagulability, with a signi cant reduction in Protein C activity but not Protein S and antithrombin-III. The condition causes derangements in erythrocytes, leucocytes, and thrombocytes, and disproportionately causes anaemia which may culminate in hypoxia and subsequent death of patients. Protein C activity and complete blood count are useful in the management of PTB and should be included in the routine workup for patients.

Data Availability
The datasets generated and or analyzed during the current study are available in the Harvard Dataverse repository, doi: 10.7910/DVN/6MMBSA.

Con icts of Interest
The authors declare that there is no con ict of interest regarding the publication of this paper.

Funding Statement
The work did not receive any speci c grant from funding agencies in public, commercial, or non-pro tsectors.

Supplementary Materials
All relevant data are within the article.