Pleuropulmonary amoebiasis has been cited with evidence as far back as 1903.¹ Since then, thoracic complications of amoebiasis are being diagnosed more often and reported with impressive frequency.
Pleuropulmonary amoebiasis is the third most common manifestation of amoebiasis in the body after amoebic colitis and liver abscess. It is probably the most important complication of amoebic liver abscess. The data on the incidence according to some authors is as follows (before hepatic scintigraphy and serological tests were available).

TABLE I

The probable explanation for the marked difference in observations by different authors may be because:

1. the cases selected were from clinical series only, post-mortem series only or both.
2. the criteria for diagnosis of pleuropulmonary amoebiasis were different in all series.

Pathology and pathogenesis
Pleuropulmonary amoebiasis may be primary or secondary.
At present, the term 'primary' lung involvement simply denotes the absence of liver amoebiasis. This could result from haematogenous^7-9 or Iymphatic spread directly from the storehouse of intestinal ulceration.
'Secondary' pleuropulmonary amoebiasis always follows manifest or occult amoebic liver abscess.
Strictly speaking, a 'primary' lung lesion should imply infection of the lung by E. Histolytica in the absence of either hepatic or even an intestinal lesion caused by the same strain of the organism.⁸ This is a matter of dispute and according to DeBakey¹⁰ and Wilmot⁸ such a concept is difficult to prove. (If it does exist, then the term 'primary' used currently should be labelled as 'secondary' and the term 'secondary' as 'tertiary' pleuropulmonary amoebiasis.)
Since this monograph is restricted to amoebic liver abscess and its various manifestations, the discussion which follows is almost exclusive of secondary pleuropulmonary amoebiasis.

Mode of spread
How does E. Histolytica travel from a liver abscess to involve the lung and the pleura? An embolic origin has been suggested.^11,12 This would explain patients with bilateral lung lesions. Takaro¹³ has suggested invasion of the lung by Iymphatics from a focus beneath the diaphragm. These methods of spread, if they do occur, must be rare. The most common mode of spread appears to be direct by rupture through the diaphragm.¹⁴ 'Erosion' with migration of E. Histolytica along the adhesions has also been described.⁷
Rogers¹⁴ suggested rupture of amoebic Iiver abscess as a probable method of spread (Fig. 38). Whether an abscess ruptures into the pleural space or directly into the lung depends on the extent of adhesions between the lung, pleura and diaphragm.
Although a subphrenic amoebic abscess rupturing through the diaphragm to involve the lung and pleura has been described,¹⁵ this must be very rare as usually extension of an abscess through the capsule of the liver is preceded by the formation of adhesions between the liver (Fig. 39) and the diaphragm. Thus, local or general peritonitis is avoided and the chances of a superior surface abscess rupturing into the subphrenic space to cause a subphrenic abscess become remote.
In other cases frank perforation may riot be present⁷ and the diaphragm is firmly adherent to the under surface of the lung. The abscess may slowly erode these adhesions. Here, by virtue of the cytolytic and histolytic properties amoebae may set up new lesions-giving rise to the 'collar stud'⁷ or 'collar button' abscesses in the lung. These may occasionally occur at a surprisingly long distance from the diaphragm. Figure 40 shows a lung abscess at a considerable distance from the hepatic shadow an attached to it by a linear strand.¹⁶
A superior surface liver abscess is one, which is most prone to cause pleuropulmonary involvement. As the abscess extends- upwards the inflammatory process involves the diaphragm and later the overIying pleura to cause pleuritis. This is followed by a sympathetic pleural effusion.^17,18 Adhesions between the two pleural surfaces may form later. The underlying amoebic liver abscess may rupture into the pleural space to cause an empyema.^18,19 This is especially true if rupture takes place before extensive adhesions between the two pleural surfaces have occurred (Fig. 41). Since some adhesions have invariably formed, the empyema is often localised. It extends more vertically and laterally than along the base¹⁶ (Figs. 42 a,b). This empyema may secondarily involve the underlying lung or empty into a bronchus forming a broncho-pleuro-hepatic fistula.^2,20,21 Pyopneumothorax, Pneumothorax and even haemothorax may also rarely occur.
On the other hand, when the two surfaces of the diaphragmatic pleura are extensively adherent, the liver abscess may either rupture directly into the lung (Fig. 43) or a bronchus,^7.9.12.13.22 or E. Histolytica may migrate along the adhesions as described earlier. Invasion of lung-parenchyma by E. Histolytica leads to the development of interstitial pneumonitis. With the formation of inflammatory exudate and migration of E. Histolytica into the spaces, consolidation of that part of the lung soon follows. The organisms, exerting their histolytic action cause liquefaction resulting in the formation of a lung abscess.^7-9,23 This abscess may rupture into a bronchus thus establishing a broncho-hepatic fistula.
A broncho-hepatic fistula may, therefore, occur in one of the following ways:

1. Rupture of an amoebic empyema into a bronchus.
2. Rupture of a liver abscess directly into a bronchus.
3. Rupture of a resultant amoebic lung abscess into a bronchus.

Often the abscess and/or empyema have been emptied of their contents by coughing ^2,18,20,21 Bilious fluid is sometimes expectorated. This is due to formation of a rare bilio-bronchial fistula where a connection between the broncho-hepatic fistula and the biliary system of the liver is established.

The fate of pleuropulmonary amoebiasis
With prompt and adequate treatment resolution of the lesion is the rule. In other cases sequelae may occur. Formation of diaphragmatic adhesions has been described. Residual persistent fibrosis, cavity formation etc., which at a later stage may result in bronchiectasis in the affected segment^8,12 have also been documented. A granulomatous mass may form-known as amoeboma of the lung.²³ It may present as a coin shadow. Thickened pleura often follows pleural involvement.
To sum up, pleuropulmonary amoebiasis can be classified as follows:²⁴

1. Haematogenous pulmonary abscess without liver involvement;
2. Haematogenous pulmonary abscess and independent liver involvement (Figs. 44, 45a,b,c);
3. Pulmonary abscess from liver abscess;
4. Broncho-hepatic fistula with minimal pulmonary involvement; and
5. Empyema extending from a liver abscess.

Clinical features of 'right' sided pleuropulmonary amoebiasis
In the following discussion only those criteria by which one suspects an amoebic etiology and the method of establishing the diagnosis are presented. No attempt is made to give detailed accounts of the clinical manifestations of pleural effusion, lung abscess or any of the numerous forms of amoebic involvement of the lung and pleura. Left sided pleuropulmonary amoebiasis is discussed in a later chapter.
The patients with pleuropulmonary amoebiasis are predominantly in the third or fourth decade although other ages are not exempt.^12,20 In keeping with the sex incidence of amoebic liver abscess, this condition also occurs predominantly in males who account for about 85-95% of the patients.¹⁸
It is easy to suspect and diagnose pleuropulmonary involvement in known cases of amoebic liver abscess when they complain of pain in the right lower chest or right shoulder. The same holds true for such patients if they start getting a dry cough, or associated expectoration of chocolate coloured sputum with or without preceding haemoptysis. The sputum may sometimes be greenish yellow and foul smelling.¹⁶ These patients usually show other symptoms of infection such as a septic type of fever with or without chills, anorexia, malaise, weakness and loss of weight. On examination the right side of the chest may show signs of fluid in the pleural space. Tracheal shift, however, is minimal, probably because of adhesions. In others, signs of consolidation with or without colIapse may be present.
The difficulty in establishing amoebic etiology, however, arises in patients who neither have symptoms suggesting amoebic liver abscess nor a tender palpable hepatomegaly. The most important prerequisite in the diagnosis of such cases is a high index of suspicion. This is especially true in countries where amoebiasis is endemic. The physician should always be aware of a possible amoebic etiology in a wide spectrum of respiratory illness. In fact pleural effusion or lung abscess at the right lung base should be looked upon with suspicion in our country.
Based on my experience, the criteria which aid in establishing a possible amoebic etiology, are as follows:

1. Past history of dysentery or diarrhoea should make one highly suspicious of a probable amoebic etiology. However, absence of such a history does not exclude pleuropulmonary amoebiasis. ¹⁶
2. Localised pain and tenderness over the liver area, however minimal or insignificant, is also highly suggestive. If the patient also complains of right shoulder pain^25,26 or hiccough suggesting diaphragmatic irritation, one should keep in mind the possibility of amoebiasis.
3. Brownish sputum (Fig. 46) especially in a patient with clinical evidence of pathology in the right lung base should always suggest the diagnosis.⁶ Initial haemoptysis often precedes expectoration of dark reddish brown sputum in which E. Histolytica may sometimes be demonstrated.⁸ Sometimes sputum becomes greenish yellow and foul smelling because of secondary infection by aerobic or anaerobic organisms, as well as admixture with bile (biloptysis). D'Abrew²⁷ has observed typhoid bacilli to be the most common organism causing secondary infection in pleuropulmonary amoebiasis.
4. In pleural effusion, especially of the right side, an aspirate may show high protein content with innumerable pus cells, but without organisms. Aspiration if repeated after a few days may often be rewarded with appearance of brown or greenishyellow pus, diagnostic of a liver abscess rupture resulting in an empyema.¹⁶ Examination of the purulent material may show vegetative forms of E. Histolytica.
5. Presence of liver cells in the pleural aspirate or in the sputum is a valuable indirect evidence of amoebic liver abscess.
6. In a chest X-ray of a case with right sided pleural effusion or other lung pathology, if the right dome of the diaphragm is raised or has any of the characteristics described in Chapter 3, Section IV, it is highly suggestive of amoebic liver abscess.
7. Lastly, if a case with effusion on the right side diagnosed as tuberculous, does not respond to the usual anti-tuberculous drugs, amoebiasis should always be ruled out.^16,28

However, in spite of all this, sometimes the diagnosis is so difficult that the real pathology is revealed only on the operation table. This happened in four of our cases i n the days when liver scan and serological tests were not available.¹⁶

Investigations
As most investigations done are the same as in amoebic liver abscess only a few pertinent points need to be mentioned below.

Demonstration of E. Histolytica
This, although diagnostic, is demonstrable only in one fourth of the patients.⁹ The sputum, aspirated fluid and also the stools should be examined. Often the aspirate does not show the parasite. In the sputum, E. Histolytica has to be distinguished from E. Gingivalis, the latter being an oral commensal. For positive identification, iron haematoxylin stained preparations should be examined especially for erythrophagocytosis. ²³
In a lung abscess bronchoscopic aspirates should be examined. Rarely a lung biopsy may be done to look for E. Histolytica in tissue sections.¹³

Radiological findings
Since the lung is adherent to the diaphragm the collection of pleural exudate may be more lateral than basal (Fig. 42a & 42b) or the fluid may be loculated only in the supradiaphragmatic region (Fig. 47). A massive pleural effusion may be present without a significant tracheal shift¹⁶ (Fig. 48).
A lung abscess will show as a typical shadow (Figs. 49 a, b) with or without a fluid level depending on whether or not the abscess has ruptured into the bronchus.
Schorr²⁹ has also described a string like shadow proceeding from a localised bulge of the diaphragm to a pneumonic shadow in the lung. This was seen in one of our cases (Fig. 50).
Obliteration of the costophrenic angle due to pleural reaction (Fig. 51) and a triangular shadow from the hilum of the lung^8,13 extending upwards (Figs. 52a, b, c) are quite common.¹⁶
Pneumoperitoneum may be useful; at times the track from the liver to the lung can be demonstrated (Fig. 53)¹⁶
Rarely a bronchogram, if done, may show a supradiaphragmatic abscess (Fig. 54). Pyopneumothorax may be seen if the lung abscess has burst into the pleural space'6 (Fig. 55).

Liver scan and serology
Now-a-days the diagnosis of pleuropulmonary amoebiasis has been made easy by investigations such as liver scan and serological tests for amoebiasis. If liver scan shows a cold area and the serological tests are positive, the lung or pleural pathology under investigation is most likely to be amoebic in origin.

Therapeutic test
Lastly, if any lung or pleural pathology responds rapidly to anti-amoebic treatment, it must in all probability be pleuropulmonary amoebiasis.

Prognosis and mortality
The prognosis, at present, for a patient with pleuropulmonary amoebiasis is very good. With prompt diagnosis and speedy institution of adequate therapy, no patient should be allowed to succumb to the disease. Recovery is rapid in patients with a bronchohepatic fistula as parenchymal involvement is minimal.¹² However, when the lung is involved, recovery may be protracted. Secondary infection, by causing persistent irreversible pulmonary lesion may worsen the prognosis.
In 1939, Oshsner and DeBakey²⁵ reported a mortality of 77.7% in amoebic empyema, 43.2% in lung amoebiasis and 10% when a bronchopleurohepatic fistula occurred. Years later in 1958, Takaro and Bond⁹ found the overall mortality to be 19%. Thus, it is obvious that with a better understanding of the disease on the part of clinicians and advent of better facilities for diagnosis and treatment, the mortality has decreased considerably.
The mortality rate in cases requiring surgery varies in different series. Takaro and Bond^9,13 in 1976 found no further reduction in the mortality rate of 33% as observed in 1958. However, Bautista O'Farrill reporting a large series found that the mortality had fallen from 14% to 8.7% after 1970.³⁰
The treatment is discussed in a later section.

References

1. Futcher, T B. I. Am. Med. Ass., 1903, 42, 480. 2 Ochsner, A, and DeBakey, M E, J. Thoracic Surg, 1936, 5. 285.
2. Ochsner, A, and DeBakeyf M E, Surgerv, 1945, 13, 460.
3. Oshsner, A and Debakey ME Surgery, 1943, 13, 460
4. Hughes, F A, and Westphal, K F. Arch Surg., 1947, 55, 304.
5. Webster, B H. Dis. Chest, 1956, 30, 315
6. Webster, B H. Am J. Resp. Dis., 1960, 81, 683.
7. Rodreguez, G. General Thoracic Surgery, Ed. by T W Shields Lea and Febiger, Philad., 1972.
8. Wilmot, A J, Clinical Amoebiasis, Blackwell Scientific Publications, Oxford, 1962.
9. Takaro, T. and Bond, W M, Surg Gyn. Obst. (I.A.S.), 1958, 107, 209
10. DeBakey, M E, New Orleans Med. 1.,1938, 87, 32
11. Cantile, I, J. Trop. Med., 1904, 7, 263.
12. DeBakey, M E, and Ochsner, A, Surg. Gyn Obst (I.A.S.), 1951, 92, 209.
13. Takaro. T (IN) Gibbon's Surgery Of the Chest. Ed. by Sabiston, D c, and Spencer, F c, w B Saunders & Co, Philad., 1976, 523
14. Rogers, L, Practitioner, 1933, 13, 117.
15. Langston, H T. and Fox, R T. Arch Surg, 1947, 55, 618.
16. Kapoor, O P. Shah, N A, et al, J. Ind. Med. Ass., 1961, 36, 506.
17. Kleeton, R W, and Hood, H. Med. Clin N. Am., 1938, 22, 27.
18. Vyas, P N. Bhargave, N M, et al, Ind. J Surg., 1963, 25, 141.
19. Rasaretnam, R. and Paul, A T. et al, Brit J. Surg., 1974, 61, 713
20. Craig C F. The Etiology, Diagnosis and Treatment ot Amoebiasis, Williams & Wilkins Co, Baltimore, USA, 1944
21. Lamont, N M, and Pooler, N R. Quart J. Med. 1958, 27, 390.
22. Ramachandran, S, Brit J. Surg., 1976, 63, 220.
23. Subramaniam, R. and Madangopalan, N. Amoebiasis, Sandoz Monograph, 1970.
24. Ochsner, A, and DeBakev, M E, Surgery, 1943, 13, 460
25. Ochsner, A and DeBakey, M E, Surg. Gyn. Obst (l A S. )~ 1939, 69, 392.
26. Ramachandran, 5, Post-Grad Med. J., 1972, 47, 685.
27. D'Abrew,, A L, (IN) Practice of Cardiothoracic Surgery, 4th Edition, Edward Arnold Publications, London, 1976, 667.
28. Chaudhary, R N. and Raichaudhari, M N. Ind. Med. Gaz., 1946, 81, 66
29. Schorr, S. and Schwartz, A, Am. J Roentgen,, 1951, 66, 546.
30. Bautista O' Farrill, J. Proc. Internat. Conf. on Amoebiasis, 1975, 888, Ed. by Sepulveda, B. and Diamond, L s. Instituto Mexicano Del Seguro Social Mexico 1976