Hyponatremia Due to Pulmonary Tuberculosis: Review of 200 Cases

Background Pulmonary Tuberculosis (PTB) is one of the common diseases with high prevalence of mortality and morbidity in developing countries. Various complications have been reported along with PTB. The subclinical electrolyte imbalances are customary in cases with PTB. Objectives The aim of this study was the evaluation of patients with PTB and hyponatremia. Patients and Methods We evaluated patients with diagnosis of secondary PTB who have been admitted to Baqiyatallah hospital, Tehran, Iran from 2005 till 2010. The diagnosis of PTB was based on the appearance of acid fast bacilli in sputum smears or sputum cultures, without any evidence of miliary TB. Demographic and laboratory characteristics relative to electrolytes were recorded according inclusion and exclusion criteria. Results The mean age was 59.22 ± 20.57 years and 91 (45.5%) patients were male. The mean serum sodium concentration was 134.54 ± 4.95 mmol/L and more than half of subjects (51%) have shown hyponatremia. The mean age difference between hyponatremic and eunatremic groups was statistically significant (61.95 versus 56.02 years) (P = 0.047). No significant relationship was found between hyponatremia and gender, anti-TB medications and co-morbidity conditions. Conclusions In this study, an older age was suggested as an important predisposing factor for hyponatremia in patients with PTB which had been observed as less of a determinant. We recommend further evaluations for hyponatremia in patients presenting with PTB, particularly for those who are older.


Background
hyponatremia is considered as one of the most com mon and important electrolyte abnormalities. hypona tremia must be considered in all seriously ill hospitalized patients (1). hyponatremia is defined as depletion in the serum sodium (Na) concentration to a level below 136 mmol/l and severe hyponatremia defined as serum sodi um concentration lesser than 115 mmol/l which it can be considered as life-threatening condition (2,3). The preva lence of severe hyponatremia and its non-severe form are estimated 1-4% and 15-30% of inpatients, respectively (4).
Usually, hyponatremia results from water retention secondary to an inability to match water excretion with intravenous or oral water absorption. Effective circulat ing volume depletion causing non-osmotic release of an tidiuretic hormone (ADh) and the syndrome of inappro priate ADh secretion (SIADh) are disorders in which ADh secretion is not suppressed despite decrease in plasma osmolality. These are the two most common causes of hy ponatremia (5).
The diagnosis of SIADh is established upon the exclu sion of other hyponatremia etiology. This syndrome has been reported in a number of clinical conditions, such as malignancies (pulmonary, mediastinal and extrathoracic tumors), central nervous system disorders (inflammato ry or demyelinating diseases, stroke and trauma), drugs (desmopressin, prostaglandin-synthesis inhibitors, phe nothiazines, tricyclics and serotonin-reuptake inhibi tors) and pulmonary diseases (acute respiratory failure, positive-pressure ventilation and infections) (2,5). PTB is one of the rare pulmonary infections which can induce hyponatremia. Tuberculosis (TB) is considered as one of the common illnesses in developing countries such as Iran which can present with various clinical manifesta tions. TB can induce hyponatremia via several mecha nisms containing local invasion to the adrenal glands (adrenal insufficiency) (6,7), local invasion to hypothala mus or pituitary gland (8,9), Tubercular meningitis (10 12) and inappropriate ADh secretion via pulmonary in fection (13)(14)(15)).

Objectives
The aim of this study was to evaluate the prevalence of hyponatremia in a large number of Iranian patients with PTB.

Design and Participants
We have prospectively evaluated patients with a diagno sis of secondary PTB who were admitted to the infectious wards of Baqiyatallah general hospital, Tehran, Iran from march 2005 till march 2010. Demographic and laborato ry characteristics were recorded. The confirmation of sec ondary PTB was based on appearance of acid fast bacilli on a sputum smear or Mycobacterium tuberculosis on a sputum culture, in the absence of radiological features of miliary TB. The patients with abnormal mental status, any evidence of tubercular meningitis, edema-forming conditions, uncontrolled hyperglycemia, renal insuffi ciency or failure, hyperlipidemia, receiving diuretics or any medications related to SIADh or induced vasopressin release were excluded.

Assessment and Treatment
In addition, hIV-positive patients with diagnosis of hIV by enzyme-linked immuneabsorbent assay (ElISA) were excluded, following counselling and securing written informed consent. Any sodium or calcium (Ca) wast ing condition such as renal diseases was overruled. All patients received a regular hospital diet. All patients received a typical daily regimen composed of Rifampin (R), Isoniazid (h), Pyrazinamide (Z) and Ethambutol (E) for the firsttwo months, followed by Rh for the following 4-10 months (depending on the progress of the disease and treatment response based on Who guideline). Si multaneously, blood and urine samples were gathered to determine measures of electrolytes and osmolality. Base line blood samples and urine specimens were acquired 2 to 3 hours after breakfast. The samples were collected before any prescription intravenous fluids.

Statistics
Data were collected to compare the profile and labora tory characteristics of PTB patients with or without hypo natremia. The data were analyzed by using SPSS software (17th edition) and P value less than 0.05 was considered significant. Quantitative and qualitative data were re ported using mean ± standard deviation (SD) and infre quency (percentage). After checking normal distribution of quantitative data, the parametric or non-parametric tests were used. For the analysis of qualitative data with normal distribution, student t-test, ANoVA, and Pearson correlation and for the abnormal distributed variables, Mann-Whitney U, Kruskal-Wallis and spearman correla tion tests were used.

Results
Two hundred patients that were diagnosed and treated for active PTB have been enrolled. The mean age was 59 ± 20 years (in range of 13-102 years) and 91 (45.5%) patients were males. Females had higher mean age than males (60 ± 21 versus 57 ± 19) but this difference wasn't statistically significant. The mean serum Na concentration was 134 ± 4 mmol/l. The female's mean concentration of serum Na was 134 ± 4.9 vs. the male's mean concentration of serum Na was 134 ± 5 mmol/l (P = 0.513). of the group, 96 (48%) of the patients had normal serum Na, whereas 102 (51%) patients had hyponatremia (47.1% male vs. 52.9% female) and two (1%) patients developed hypernatremia.
There is no significant relationship between gender and hyponatremia (P = 0.670). The mean age difference be tween hyponatremic and eunatremic groups was statisti cally significant (61 versus 56 years, respectively) (P.value = 0.047). Moreover, there is no significant correlation be tween anti-tuberculosis medications and hyponatremia (P = 0.369). Serum Na concentration was enumerated for all different anti-TB drugs in Table 1 . The mean corrected serum Ca concentration was 8.8 ± 0.7 mmol/l. The females mean serum Ca concentration was 8.9 ± 0.8 vs. the males mean Ca which was 8.8 ± 0.6 (P = 0.441). one hundred thirty seven (68.5%) patients had normal serum Ca concentration, whereas 59 (29.5%) pa tients had hypocalcaemia and 4 (2%) patients had hyper calcemia. hTN was the most common co-morbid condition pres ent in 46 patients (23%), while 42 patients (21%) had con trolled diabetes mellitus, 8 (4%) patients had migraine headache, 1 patient had Parkinson's disease and another patient had epilepsy. There was no significant relation ship between co-morbid conditions and hyponatremia (P = 0.102). Simultaneously, pneumonia was documented in 18 patients (9%), while no patients had a diagnosis of sepsis. hospital mortality was documented in 13 patients (6.5%). No significant correlation was found between hy ponatremia and mortality (P = 0.218). All of the patients with hyponatremia became eunatremic after anti-tuber culosis therapy.

Discussion
Given the findings, the prevalence of hyponatremia and hypocalcaemia obtained as 51% and 29% which is com patible with the former reports. Age was the sole vari able which was different between PTB patients with and without hyponatremia. We found that hyponatremia was higher than expected in non-tubercular patients accord ing to the previous reports. In previous studies, the preva lence of hyponatremia among inpatients, especially the ones who admitted in respiratory wards, was reported in a various range (2.48%-40%) (16,17). This issue can be relat ed to the type of diseases and age. As reported in several studies, patients who were admitted in children or geri atric wards, admitted in the intensive care unit or cardiac care unit, and those who admitted in the emergency de partment were more susceptible for hyponatremia. Fur thermore hyponatremia is more to be observed in neo plastic, brain, endocrine and pulmonary diseases. The incidence of severe hyponatremia has been estimated as 1.1% in hospitalized patients whereas in that report, PTB was the most common underlying disease (24%) which is in keeping with our findings (18).
In 1969, Chung and hubbard have noted that nearly 11% of patients with active TB (pulmonary or non-pulmonary) are affected with hyponatremia, and it is apparent that the main cause of serum sodium depletion in these pa tients is SIADh (19). Vorherr et al. has reported a case with PTB and hyponatremia and found antidiuretic agents in tuberculous lung tissues (14). Bryant et al. has suggested the syndrome of inappropriate secretion of antidiuretic hormone for patients with an infectious pulmonary dis ease such as PTB (20). Schorn et al. reported two cases of PTB and an abnormal inappropriate antidiuretic hor mone level as a justifier mechanism (15). Cockcroft et al., reported a 74-year-old woman with miliary tuberculosis which had complicated by severe hyponatremia due to SIADh (21). Usalan et al. reported a case of TB who initially presented with lethargy due to hyponatremia evidently resulting from SIADh (22). Finally, lee reported a case of PTB with refractory hyponatremia due to SIADh (13).
Although in this study we did not evaluate patients for etiology of hyponatremia, it can be beneficial if we con cisely review causes of hyponatremia in patients with PTB; SIADh is a considerable complication of pulmo nary infection, inflammatory and neoplastic disorders, although its prevalence and mechanism are poorly re garded. SIADh has been displayed in infectious situations such as TB. In one of the first reports, Weiss et al., reported hyponatremia in resulting from SIADh in patients with PTB (23). Then it was declared that an increased ADh level in the presence of hyponatremia in PTB cases is an indi cator for ectopic ADh production. Few studies demon strated that the ADh level was not detectable following full anti-TB therapy (13,24).
SIADh was usually demonstrated in patients with TB and there are various causative factors for SIADh in tu berculosis. SIADh may occur following PTB, as well as tubercular meningitis. There are many reports of SIADh associated with pulmonary, miliary and central nervous system-related TB. More than 60% of the patients with tubercular meningitis may present with hyponatre mia or SIADh at first presentation (25). SIADh must be considered in every case with hyponatremia with low serum osmolality condition, a normal acid-base state, urine osmolality over 100 mosm/kg, and urine sodium concentration more than 40 meq/l. Also, generalized or local infections are important and unregarded causes of SIADh. Multiple infectious diseases are associated with this syndrome (26).
In addition to the disease itself, some anti-TB drugs can also be round anemia. Nakashita et al. has reported a case of SIADh caused by ethionamide in a patient with PTB also they suggested that anti-TB drugs should be consid ered as the possible cause of SIADh but the result of this study was that the incidence of hyponatremia in patients who received ethionamide with maximum dose was not higher than those who received a lesser one (27).
Besides, non-tuberculoses pneumonia is a very rare cause of SIADh induced hyponatremia. Charles et al. de scribed a patient who had lobar pneumonia presenting with confusion due to severe hyponatremia (28). Also, Rivers et al., expressed a child with excessive secretion of ADh in association with common pulmonary infections (29).
Endocrine system involvements by TB, as the other mechanism, can induce hyponatremia which is impor tant to consider in patients with PTB. TB was revealed to involve adrenal glands directly (30) and this involvement lead to overt or subclinical adrenal insufficiency and hyponatremia (6). Pituitary gland may also be involved by the tuberculosis bacilli. hypopituitarism has been re ported in 20% of cases years after the treatment of tuber cular meningitis in childhood. The reason seemed to be tuberculosis lesions impressing the hypothalamus, pitu itary stalk and indirectly or directly, the pituitary gland itself (8).
The hyponatremia due to PTB is usually mild to moder ate, asymptomatic, and self-limited. SIADh is commonly reversible with effective PTB treatment in most cases (25). Therefore, it can be overlooked if the physician does not give sufficient attention. on the other hand, patients who affected by hyponatremia were more likely to have higher mortality. Sharma et al. suggested hyponatremia as predictors of development and outcome in patients with acute respiratory distress syndrome due to tubercu losis (31).
Moreover, other electrolyte disturbances such as hy percalcemia have been reported as one of the most com mon electrolyte imbalance in 25.7% of patients with PTB although hyponatremia developed lower prevalence (22.15%) (32). Furthermore, the incidence of hyponatre mia in patients with AIDS complicating with TB is higher. Smith et al. revealed that hyponatremia was discovered in 60% of AIDS patients with a diagnosis of generalized tuberculosis, however half of these patients of dissemi nated tuberculosis were only diagnosed after death. In our study, we overruled the hIV positive patients and it can be the reason of the differences between our findings and others.
In the present study, the higher age among the demo graphic characteristics was observed in patients with PTB and hyponatremia than has previously been noted. Few studies evaluated the role of age on presenting hypona tremia in patients with neural defects (33). The age range in mentioned reports was documented widespread and generally it is impossible to clearly the role of age on hy ponatremia in patients with PTB. In addition, it should be suggested that hyponatremic patients should be evalu ated for PTB when an initial investigation failed to test. We would recommend that patients with PTB (especially older age patients) should be closely observed for electro lyte imbalance. We also recommend further studies with a greater sample size mainly focused on the predispos ing factors of electrolyte imbalance in patients suffering from PTB.