Results
A 45 years old male presented in chest OPD with history of fever for 8 weeks duration accompanied with hiccoughs and chest pain for 6 weeks. Fever was of mild to moderate grade and sometime associated with chills. There was no history of breathlessness or cough. Two weeks later patient developed right sided chest pain along with hiccoughs.

These symptoms were persistent and did not respond to antibiotics and symptomatic treatment. There was no past history suggestive of Tuberculosis, other respiratory infection and any chronic medical illness. On chest examination there was diminished breath sounds on infrascapular region Rt. side. Patient was carrying a X-RAY (Figure 1) & CECT chest (Figure 2) which showed a lobulated well defined (bilobed) homogenous shadow on right side posterior aspect, but not obscuring the right heart border or the right upper mediastinal structures. The right hilum was clearly seen through the shadow. The upper limit was seen reaching up to the medial right hemi-diaphragm. Left hilum appeared prominent and bilateral Costo-phrenic angles were clea, malignancy, tuberculosis. After clinical assessment patient was advised admission for further workup. On investigation the reports were Hb-13 gm./dl, TLC-5 600/ml, DLC-P₅₆, L₄₀, M₂ E₂, Blood Urea-38 mg/dl, Sr. Createnine-0.8 mg/dl, ESR-12 mm, Sr Na-140 mg/dl, Sr K- 4.1 mg/dl.

 

Figure 1 Patient was carrying a X-RAY (date on film is the capturing date)

Figure 2 CECT chest (date on film is the capturing date)

Aspiration of loculated collection was done on the clinical suspicion of infective pathology and sent for ADA (adenosine deaminase), cytology, and biochemical tests. The fluid was negative for malignant cells, but ADA level was raised 73.6 U/L.
The clinical presentation of prolonged fever supported by the raised level of ADA in the loculated pleural fluid suggested a diagnosis of Tuberculosis. The patient was started on HREZ. The doses were Isoniazid (H)-5 mg/kg, Rifampicin(R)-10 mg/kg, Pyrazinamide(Z)-25 mg/kg, Ethambutol(E)-15 mg/kg. The drugs were started on daily basis and regime was initial two months of intensive phase followed by four months of continuation phase.

After one week of initiation of  ATT patient developed a sudden bout of vomiting and vomited copious yellowish pus like fluid around 100~150 mL with tinge of blood at the beginning. To evaluate these sudden onset new symptoms he was reassessed and diagnosis of impaired liver function was made and investigated accordingly, the liver functions turned out to be within normal range. From the very next day pt. was symptomatically better and on chest examination breath sounds were normal on both sides. A follow-up chest radiograph was done which showed normal both lung fields except for mild bi-hilar prominence (Figure 3).
 

Figure 3 Normal both lung fields except for mild bi-hilar prominence (date on film is the capturing date)

Such a sudden disappearance of the lung shadows is very unusual and prompted us a search to explain the discharge of collection through esophagus.

A barium meal continuation was carried out which revealed no evidence of any abnormal tract/fistula. There was focal thickening of the oblique foramen in the upper part therefore CECT examination of the chest was done to rule out any subtle abnormality inside in plain radiograph examination in barium swallow (Figure 4). It showed a small patch of consolidation in right side lower zone.
 

Figure 4 Plain radiograph examination in barium swallow

He took ATT for six months and in the follow up of one year he remained asymptomatic.

 Discussion
The presence of Pleural effusion in tuberculosis can range from 4% to 25 % cases depending of endemicity of disease and many other factors (Qiu et al., 2006; Valdes et al., 1996). In India the rate of pleural effusion is about 20% cases of tuberculosis (Sharma and Mohan, 2004). Typically it occurs 4~7 months following initial infection. It is caused by rupture of small sub-pleural focus (Berger and Mejia, 1973) and leads to obstruction of the lymphatic pores in the parietal pleura which causes accumulation of protein in pleural cavity. In the natural history, tuberculous pleuritis usually resolves spontaneously, but the patient frequently develops active TB at a later date. The mean duration for complete resorption of the pleural fluid is approximately 6 weeks, but it can be as long as 12 weeks (Tani et al., 1964), however, if untreated, 65% of patients will develop pulmonary or extrapulmonary tuberculosis within 5 years (Berger and Mejia, 1973). Residual pleural thickening is common. In treated case of tuberculosis average patient becomes afebrile within 2 week, complete reabsorption of fluid takes place in 6~12 weeks and the incidence of pleural thickening at 6~12 months is 50%. Size of original effusion and the presence or absence of small radiologic residual pleural disease do not correlate with subsequent appearance of active tuberculosis (Roper and Waring, 1955).

After extensive search of literature the average duration of resolution of pleural effusion in tuberculosis come to be is at least one month; we could not find any case in which it had been disappeared in less than two weeks. This feature makes case perplexing. 

 After getting the new radiological finding, patient was interrogated for any bout of excessive cough with copious amount of secretion and any other symptom suggestive of expulsion of fluid through airway or GI tract. There was no history of any new respiratory symptom since the start of ATT. 

 There was a history of vomiting after one week of start of treatment, the contents being muco-purulent with tinges of blood at the beginning. We were surprised how the loculated collection was vanished after an episode of vomiting, without any radiological evidence of trachea-esophageal fistula, or esophageal pleural fistula.

So again review of admission CECT scan film was done, we could make out the two areas of fluid collection were in communication with each other and presence of a necrotic lymph node in between pleural fluid and esophagus. From this we hypothesized this node has probably given a track for fluid and it was vomited out as evident in history (as in esophageal-pleural fistula), after it closed spontaneously. Follow-up CECT scan was unable to explain the tract of fluid.

Esophageal –pleural fistula (EPF) is caused by iatrogenic trauma (esophageal instrumentation or external trauma), diseases of the esophagus such as corrosive esophagitis, esophageal ulcer and neoplasm, and rarely spontaneously (van Den Bosch and Laros, 1980; Wechsler and Laufer, 1982). EPF is an uncommon complication of iatrogenic trauma (endoscopic instrumentation) or post-pneumonectomy.

 In our case there is no such history of trauma, surgery of other cause as described. The expulsion of fluid through esophagus was sudden and was not preceded by typical symptoms of esophagitis. The radiological signs of the EPF depend upon site, duration, and severity of perforation; and more importantly, the integrity of pleura.  The diagnosis of esophageal rupture/EPF is made clinically; however, for confirmation, the imaging is required. The imaging modalities include chest radiograph, ultrasound, barium swallow, contrast-enhanced CT, and MRI with each modality having its advantages, and chest CT is very useful modality. In the index case there were clinical evidences of EPF but X-ray, barium meal & CECT chest were unable to explain EPF.

In all cases described earlier there was history suggestive of EPF and many times tract could also be localized. Here there was no history suggestive of EPF and during brief period of hospitalization there were only hiccoughs as symptoms of GI system. To make scene more interesting the tract obliterated after the expulsion of fluid with minimal residual changes. After vast search of literature we were unable to find such unique type of presentation of tuberculosis as a vanishing lung tumor making this case worth to be shared with medical community as eye opener.

References
Berger H.W., and Mejia E., 1973, Tuberculous pleurisy, Chest, 63(1): 88-92 http://dx.doi.org/10.1378/chest.63.1.88  PMid: 4630686

Qiu L., Teeter L.D., Liu Z., Ma X., Musser J.M., and Graviss E.A., 2006, Diagnostic associations between pleural and pulmonary tuberculosis, J. Infect., 53(6): 377-386 http://dx.doi.org/10.1016/j.jinf.2005.12.023 PMid: 16466663
 
Roper W.H., and Waring J.J., 1955, Primary serofibrinous pleural effusion in military personnel, Am. Rev. Tuberc., 71: 616-34 PMid: 14361976 

Sharma S.K., and Mohan A., 2004, Extrapulmonary tuberculosis, Indian J. Med. Res., 120(4): 316-353 PMid: 15520485

Tani P., Poppius H., and Makipaja J., 1964, Cortisone therapy for exudative tuberculous pleurisy in the light of the follow-up study, Acta Tuberc. Pneumol. Scand., 44: 303-309 PMid: 14168051 

Valdes L., Alvarez D., Valle J.M., Pose A., and San José E., 1996, The etiology of pleural effusions in an area with high incidence of tuberculosis, Chest, 109(1): 158-162 http://dx.doi.org/10.1378/chest.109.1.158  PMid: 8549179

van Den Bosch J.M., Swierenga J., Gelissen H.J., and Laros C.D., 1980, Postpneumonectomy oesophagopleural fistula, Thorax, 35: 865-868 http://dx.doi.org/10.1136/thx.35.11.865 PMid: 7221985 PMCid: 471399

Wechsler R.J., Steiner R.M., Goodman L.R., Teplick S.K., Mapp E., and Laufer I., 1982, Iatrogenic esophageal-pleural fistula: Subtlety of diagnosis in the absence of mediastinitis, Radiology, 144: 239-243 PMid: 7089274