Infective endocarditis : an enigma of diagnosis and treatment.
Almost a century has gone by since the first Gulstonian lecture by Sir William Osler in 1885 regarding clinical, microbiologic and therapeutic features of infective endocarditis. Yet during the past three decades, dramatic changes in the clinical features of endocarditis are noteworthy and demand a fresh look at the problem. A physician of today relying on the criteria laid down decades ago would miserably fail to suspect such a sinister infection in almost 90% of the patients as seen at present. The disturbing feature of overall mortality of 30% is well documented and has not altered during the past thirty years despite development of many diagnostic and therapeutic measures like improved microbiological techniques, newer antibiotics, measurement of antibiotic levels and advanced surgical intervention. The undeclined mortality rate producing therapeutic paradox is attributable to delayed diagnosis due to the changing clinical picture, inadequate antibiotic therapy prior to confirmation of diagnosis, infection by uncommon organisms including fungus, culture negative endocarditis and prosthetic valve infections.
The uncommon features in six cases of infective endocarditis involving the aortic valve are presented to highlight the problems in diagnosis and subsequent therapy. A case of prosthetic valve endo carditis requiring further replacement of the infected prosthesis, ultimately succumbing to uncontrollable infection is included in this study with the autopsy data.
Six cases of aortic valve endocarditis presenting with atypical clinical features seen during a short period of six months are included in this study.
All patients had detailed clinical, radiological and serial echocardiographic evaluation. Serial total and differential counts, sedimentation rate, haemoglobin, total proteins and albumin: globulin ratio were done in all. Urine for microscopic red cells was checked daily. All patients had three venous and three arterial sample cultures prior to initiation of antibiotics. Whenever suspected, anaerobic media and fungal cultures were done. Delayed cultures (three arterial and three venous samples) were done if initial cultures did not yield any organisms. Standard regime of therapy with intravenous crystalline penicillin and intramuscular streptomycin in adequate dosage was started. If cultures grew any organisms, depending on sensitivity, appropriate antibiotics were started. In cases with culture negative endocarditis, gentamicin and cloxacillin were added to crystalline penicillin.
The observations are shown in [Table 1]- [Table 3] and [Fig. 1] and [Fig. 2]
Six patients (5 males and 1 female) in the age group of 20-27 years (mean age: 23 years) had involvement of aortic valve. The etiology of aortic regurgitation (A.R.) was rheumatic in 4, while 2 had essentially normal aortic valve with aortic regurgitation secondary to endocarditis. One case who had aortic valve replacement for rheumatic aortic regurgitation had endocarditis on the prosthetic valve.
Symptoms [Table 1]
Intertestingly enough, 3 patients were afebrile and all of them had a course of antibiotics prior to admission. Rapid onset of clubbing was striking in 3 patients. In one case, the presenting symptom was a lethal cerebral embolus and another presented with Osler's nodes.
Signs [Table 2]
While anemia, splenomegaly and clubbing were present in majority, only one had Osler's nodes. Changing murmurs were not a feature but in one patient a new murmur appeared suggesting periprosthetic aortic regurgitation. Premature closure of mitral valve and other findings suggestive of acute aortic regurgitation were present in two cases.
Infecting organisms and mortality [Table 3]
Only surviving patient had infection with Streptococcus viridans. Four cases had negative cultures and one had polymicrobial infection. Among the patients with negative cultures, at autopsy only one had evidence of cryptococcal fungal infection. In the remaining 3 cases, intracellular bacteria and extensive vegetations were seen. Cultures were not performed in these.
Causes of death
Cerebral embolism was responsible for death in one case whereas multiple systemic emboli were the terminal events in another case. In four cases the deaths were related to uncontrolled infection and progressive haemodynamic deterioration.
The persisting problem of infective endocarditis has emerged as a challenge. Although the changing picture of infective endocarditis is very well reported in the literature, the clinical and bacteriological manifestations differ from report to report., , , , 
There has been a remarkable decrease in the incidence of infective endocarditis in the younger age group in the West. This is probably related to a marked decline in the incidence of rheumatic heart disease. In the present study, all the patients belonged to a younger age group (average age 23 years). Average age of thirty years was reported by Patel. All the patients except one were males, thus corroborating with earlier studies reporting a male preponderance.
Infective endocarditis affecting a diseased valve is well-known; however, there is enough evidence now that a normal valve can also be involved. In some of the series,  the affection of normal valves in as much as 53% is noteworthy. In all the patients aortic valve was involved though 2 patients had no aortic valve pathology prior to infection. Predominant mitral valve ivolvement is well documented. The same was corroborated from a necropsy report of 45 patients at our centre, during the period 1967-72, where mitral valve was involved in 44.5% against 22.2% involvement of the aortic valve. This disparity could be due to a small number of patients in this series.
One patient had endocarditis of the prosthetic valve. The incidence of infection of prosthetic valve varies from 1-47%.2 Experience at our centre from a review of 205 cases of valve replacement during the period 1969-1975 revealed an incidence of 9.5%.9 Out of 205 patients 15 had fungal infection, three patients had mixed and one had pure bacterial infection.
The triad of fever, changing murmurs and embolic phenomenon suggesting infective endocarditis have changed considerably over the years, ,  The frequent use of bacteriostatic and inadequate dosage of bacteriocidal antibiotics often leads to delay and modification of the presenting symptoms and signs. Three patients had a week's course of antibiotics prior to admission and all these patients presented without fever. The diagnosis in these cases was suspected on the basis of other clinical features. Changing murmurs perhaps have been overemphasised as feature of infective endocarditis. None of our patients had this feature. On the contrary, Oakleyhas stressed that the development of new murmurs characterize the diagnosis as in one of our cases with prosthetic aortic valve where regurgitant murmur appeared ten weeks after surgery. There is evidence that many of the symptoms and signs formerly attributed to microembolism are due to deposition of immune complexes. Nevertheless, Osler's nodes, Roth's spot, Janeway's lesion and mycotic aneurysm are considered to be embolic. In our study, one patient had Osler's nodes as the presenting symptom. Other had primary neurological presentation. Mycotic aneurysms were not seen in our series.
Rapid onset of clubbing was a striking feature and was noted in three patients. In two of them, clubbing appeared within six weeks of onset of illness. None of the reported series, ,  has emphasised this manifestation. One patient presented with a complaint of acute onset of clubbing without any other symptoms and had evidence of aortic valve disease with persistently elevated E.S.R. The first manifestation of infective endocarditis was a fatal cerebral embolism.
Infective endocarditis causing destruction of valve leading to acute incompetence, though reported by some authors is not a common feature. Damage to the aortic valve may be associated with either fenestration, inversion of a cusp or rupture of a mycotic aneurysm. Two patients presented with signs and symptoms of acute aortic regurgitation, including clinical evidence of premature closure of the mitral valve.
Echocardiography is helpful in establishing the diagnosis of infective endocarditis. Recently numerous reports., ,  have described the accurate localization of valvular vegetations. The echocardiographic findings consist of a mass of fuzzy echoes adjacent to the valve structure. In one patient, vegetations could be recorded in the aortic valve and also had evidence of premature closure of mitral valve [Fig. 1]. All patients had evidence of left ventricular volume over-load and fluttering of anterior leaflet of the mitral valve.
The changing picture of infective endocarditis has highlighted the shift or organisms from the usual Streptococcus viridans to other streptococci, staphylococci, Gram negative bacteria, fungus and various exotic organisms. Out of six patients, four had negative blood cultures constituting the largest group; in the remaining. One had Streptococcus viridans infection and the other had polymicrobial infection. This observation emphasizes the changing picture of the disease. Hampton et al reported negative blood cultures varying from 31-66%. The possible causes of negative culture include unusual organisms with specialised growth requirements, previous antibiotic therapy, prolonged infection with signs mainly due to immune process, non-bacterial endocarditis, and fungus or rickettsia infection. Out of four culture negative patients, three had history of previous antibiotic therapy. Polymicrobial infective endocarditis has been recently recognised specially in operated cases. The patient who had prosthetic valve surgery had polymicrobial infection and the cultures grew Streptococcus viridans, E. coli and Pseudomonas pyocyaneus.
Many series have reported best survival in infections with Streptococcus viridans. Patney has, reported 64% survival. The only surviving patient in our study had Streptococcus viridans endocarditis. Those with negative blood cultures succumbed to uncontrolled infection and progressive deterioration of cardiac status. Fatality rate is higher with negative blood cultures against those with positive cultures. This difference may be attributed to the difficulty encountered in selecting and exhibiting optimally effective therapy. Polymicrobial infections too are associated with very high mortality rate. Herman et al have reported 3'7-81% mortality in cases following surgery. One patient (N.M.) had infection of aortic prosthetic valve with progressive regurgitation requiring second replacement but ultimately succumbed to uncontrolled infection and multiple systemic emboli. At autopsy, massive infarct was present in the parietal lobe of the brain and in kidneys. Heart was large and there was left ventricular and left atrial hypertrophy. The Starr-Edwards prosthesis in the aortic position was occluded to a considerable extent by vegetations which were seen on the inferior surface of the valve [Fig. 2]
Our series is small, yet its careful analysis shows that despite advancement in diagnostic procedures including microbial culture techniques, infective endocarditis poses a problem and confirmation of diagnosis is difficult. Identification of the causative organisms remains a major problem. Consequently, effective therapy is eluded and prognosis remains unchanged.