Infective endocarditis- (a survey of the past 50 years)Ajita P Mehta, Kalyani M Dave, Suman G Kinare
Department of Pathology and. Microbiology, Seth G.S. Medical College. Parel, Bombay-400 012., India
Autopsy incidence of infective endocarditis during the past 50 years was analysed to review the changes in the clinico-pathologic pattern if any. 185 cases were recorded in a total of 39931 autopsies giving an average incidence of 0.46%. The cases were classified into groups I, II & III depending upon whether the endocarditis supervened on normal hearts, diseased hearts or followed surgery, respectively. The fifty year period could be divided into three phases on incidence. First phase (1927-41) represented pre-antibiotic era in which group I cases predominated and causative organisms in 801" o f cases were virulent cocci such as staphylococci, haemolytic streptococci and pneumococci. In the second phase (1948-66) the general incidence was reduced by 50%. Group II cases predominated and the most common etiologic agent was streptococcus viridans (40%). In the third phase (1967-76) the general incidence has reached again to preantibiotic level with coagulase positive staphylococci, Gram negative bacilli and fungi accounting for 90% of cases. Group III contributed significantly in this phase.
In the one and half century since Bouillaud  described it first, infective endocarditis has retained the interest of clinicians, pathologists and microbiologists alike. Horder,  Thayer  and Libman and Cellar  studied extensively and catalogued precisely the natural history of the disease with predictable signs and symptoms as well as the course in pre antibiotic era. Introduction of antibiotics brought about marked reduction in the incidence of the disease though not complete disappearance. The interest in the disease however was kept alive not merely because of its persistence in the anti biotic era, but because of the striking change in the clinical picture on account of the emergence of resistant strains. The changed and changing pattern of the disease has been analysed by several workers in developed countries. ,,,,,,,,,
In a developing country like India with higher incidence of infections, as well as rheumatic disease, combined with inadequate facilities of detection and poor availability of higher antibiotics, the pattern of infective endocarditis may not be the same as in the developed countries.  Hence it was thought relevant to review the pattern of this disease through last 50 years as reflected in the autopsy data.
Data was collected from the autopsies performed in the Department of Pathology of the King Edward VII Memorial Hospital during the years 1927 to 1976. The number of autopsies constituted, on an average 37% of deaths but after 1962 the number always exceeded 40%. The information about the causative organisms was based on microbiological studies carried out at autopsy as well as on their detection in paraffin sections.
Infective endocarditis was found in 185 cases out of 39931 autopsies carried out during the period 1927 to 1976. The percentage varied over the years from 0% to 1.8% giving an average of 0.46%. The graphic representation of yearly incidence as percentage of autopsied cases shows rather irregular distribution with sudden high peaks, [Figure 1]. However, the overall distribution if studied closely showed persistent decline after 1947 with complete absence of cases during 1951-54. There was irregular rise between 1955-1966. From 1967 onwards the rise is persistent and uniform. These periods can be considered as representing three phases of the disease. In further analysis therefore, the data is compared taking these phases into consideration. The first phase (1927-47) corresponds to the preantibiotic era in our country as antibiotics became freely available after 1947. In the 3rd phase (1967-76), apart from antibiotics there was one more modifying factor viz. increase in the number of cardiovascular operations, particularly the open heart. The overall incidence of infective endocarditis in the pre-antibiotic period was 0.58% of autopsies. It showed more than 50% reduction in the second phase being 0.26%. The incidence again rose to 0.55% in the last decade. To separate out the significance of important contributary factors the cases were grouped in three categories.
Group I: Endocarditis supervening on normal hearts-78 cases (42%).
Group II: Endocarditis supervening on diseased hearts-66 cases(35.8%).
Group III: Endocarditis following cardiac surgery-41 cases (22%).
The change in the pattern of relative incidence of three groups through various phases is depicted in [Figure 2].
Group I which reflects the sepsis in general, was predominating the preantibiotic era. It showed marked reduction in the second phase but assumed its position again in the last decade. The emergence of group III in the last decade corresponded with the increase in open heart surgery. Out of 41 cases, endocarditis was a complication of open heart surgery in 37. In 34 of these, endocarditis supervened on replaced prosthetic valves. Group II cases predominated in post-antibiotic era but showed marked decline after 1967. This however, is not true in absolute terms. If absolute incidence in autopsies is considered, the group II cases showed a steady decline after 1947. Group II cases formed 0.24% of autopsies in the first phase, 0.20% in the second phase and 0.12% in the last phase.
Group II was further analysed to find out the incidence of predisposing heart disease in various periods. Rheumatic heart disease formed the bulk of these cases. Rest of the cases comprised of congenital heart disease. Syphilitic heart disease was a rare predisposing cause and only one case each was noted in the first and second periods. The incidence of rheumatic heart disease in this group was 87.5%, 85.6%, and 63.6% respectively in the three phases. Thus the incidence of rheumatic heart disease showed distinct relative decrease only in the last decade with corresponding rise in congenital heart disease.
Age and Sex distribution
Nearly 75% of cases belonged to 2nd, 3rd and the 4th decades with peak incidence in the 3rd decade [Table 1]. This pattern of incidence is not essentially changed over the years and is seen typically in group II cases. Group I cases showed higher incidence in the 4th decade in pre-antibiotic era. In the last decade also number of cases are seen in the 5th, 6th and 7th decades though the peak is in the 3rd decade. Group III cases show higher incidence in the 2nd decade but this is a highly selective group and does not compare with naturally occurring disease.
There is overall male predominance in sex distribution. There is marked difference in sex distribution in various groups in different periods. Group II shows the highest ratio in each period ranging from 7:1 to 3: 1, while group I shows higher ratio in the pre-antibiotic era mainly. It is difficult to interprete these results as various extraneous factors influence the sex ratio. These will be discussed later.
Sites of Endocarditis
The lesions in the majority of cases are confined to the valves. Mural endocardium was involved only in 10 cases and that too in group I and III [Table 2]. Mitral valve was the commonest site in all the groups. Aortic valve involvement is comparatively more common in group III but in this group the site of involvement is directly related to the operative procedure. Multivalvular involvement particularly the combined involvement of mitral and aortic valves is much more common in group II. No significant difference was observed in the sites of involvement through various phases but isolated involvement of right sided valves was rarely seen in earlier years and 4 out of 5 cases observed in group I occurred in the last decade. In these cases there was evidence of prolonged intravenous administration.
Information about organisms responsible for endocarditis was available in 124 cases [Table 3].
In nearly 50% of group I cases coagulase positive staphylococcus was the etiologic organism. In group II, on the other hand, streptococcus viridans was responsible for nearly 40%; cases. In group III fungus was responsible for 60% of cases. In the pre-antibiotic era, in 21 out of 26 cases the endocarditis was caused by pyogenie cocci. Gram negative bacilli were responsible for one case only while no fungal endocarditis was noted [Table 3].
In the post-antibiotic era infection with streptococcus viridans and less virulent organisms were as common as virulent orgardsms. In the last decade though virulent cocci were the commonest agents, there is remarkable increase in infections due to gram negative bacilli and fungi. The fungal infection in all but one case followed prosthetic valve involvement.
Embolic lesions in major organs are common in all the three groups [Table 4].
Embolic infarcts septic or aseptic are seen in myocardium in 34% of cases and are important from the point of view of cardiac failure. In group I cases, kidney is the most frequently involved organ while in group II spleen appears to be the most frequent target. Involvement of central nervous system is significantly higher in group I and III cases than in group II cases. Lung involvement is as expected more common in group I patients.
Introduction of antibiotics and chemotherapeutic agents in the management of infectious diseases changed dramatically the natural history and prognosis of many infective disease. With emergence of resistant strains however, these are returning with a changed clinical profile. This is amply confirmed in case of infective endocarditis by several reported series. ,,,,,,,,,
The mortality of infective endocarditis was 100% before the advent of antibiotics.  In developed countries it is now reduced to around 30% , There is no data available in India but mortality is not likely to be less than 50%. The study of autopsies alone is likely to give a rather biased data especially if the mortality is not very high, However, in the absence of proper laboratory data particularly the micro-biological one, autopsy data becomes more dependable and hence such a study was thought worthwhile. Moreover a ratio of total deaths and the number of autopsies performed is fairly satisfactory and hence the data obtained can be considered representative.
The antibiotics did bring down the total incidence of endocarditis in our series by more than 50% . In the last decade there is again a notable increase and the incidence has almost returned to pre-antibiotic level. The important change noted in the clinical profile in developed countries is the change in age of patients. This change is attributed to the marked decrease in the incidence of rheumatic heart disease which is a disease of young. ,,, In our data there is no significant difference in the age groups over the last 50 years. Most of the cases belong to 2nd, 3rd and the 4th decades and the peak incidence is still in the 3rd decade.
Another feature of the changed clinical profile is increasing male preponderance; which is more marked in older age group. ,, In the present series the male preponderance is marked in the preantibiotic era. The ratio decreased markedly in the post-antibiotic period but in the last decade there is a rise again. However, it is difficult to draw any definite conclusions from this data. There are several factors which influence the male and female ratio. The ratio of beds available in the hospital for males and females is 2:1. In earlier years due to socio-economic conditions there was reluctance on the part of females to enter a general hospital as well as there was deliberate neglect by family in providing medical relief. This may account for high ratio observed especially in pre-antibiotic era. It is however, noteworthy that male preponderance is seen also in cases of bacterial endocarditis associated with rheumatic heart disease. Autopsy data from the same source shows equal incidence of rheumatic heart disease in males and females in the post-antibiotic era,  indicating increased succeptibility of male rheumatic patients to bacterial endocarditis.
The distinction between acute and subacute endocarditis which was mainly dependent on the clinical duration of the illness is not of much clinical use now as it depends mainly on the causative organism and its succeptibility to antibiotics. Instead, it is better to classify the cases into those with previously normal heart and those with previously diseased hearts. In the first group, the primary condition is marked bacteriaemia or septicaemia by virulent organisms. In the second group bacteriaemia with less virulent organisms which may have no effect on normal valves affects the diseased valves. Because of known susceptibility of this group, endocarditis is generally diagnosed and treated earlier. Moreover, the organisms affecting these hearts are usually susceptible to antibiotics. In the first group, the clinical diagnosis is often difficult and majority of cases are diagnosed only at autopsy. , This group is the one which was curbed in early post-antibiotic era but is on the increase in recent years. ,,,,, In the present series also this group was dominant in the pre-antibiotic era and is dominating again in the last decade, thus confirming the change observed in developed countries. The second group dominated in the early post-antibiotic period mainly due to the decrease in the first group. The absolute figures show that there is a gradual decrease in this group throughout the years though the decrease is not as much as in developed countries, mainly because the rheumatic heart disease is not on wane in this country as yet. The incidence of congenital heart disease and with bacterial endocarditis is significantly increased during the last decade chiefly on account of the increased facilities for diagnosis and treatment available for these patients which are attracting them to hospital. Post-operative bacterial endocarditis constitutes one of the serious complications of cardiac surgery particularly valve surgery. The incidence of this complication is rather high and the factors responsible for this are discussed in a separate communication.
The etiologic organisms which play the most crucial role in the course and prognosis of endocarditis are showing the same change in the pattern as has been observed by other workers. ,,,,,,, The increasing role of staphylococci, gram negative bacilli and fungi is clearly seen in the present series.
As far as the site of endocarditis is concerned mitral valve is the site of predilection irrespective of the etiologic agent in the present series. This is in conformity of some of the reported series.  Recently however, some workers have reported more frequent involvement of aortic valve than mitral particularly in cases in whom endocarditis supervened on normal valves. , The tricuspid involvement is rare but is on the increase in the recent years. It is particularly common in drug addicts. In the present series the increase in the tricuspid involvement in recent years is probably the result of prolonged intravenous therapy. Involvement of mural endocardium is seen in virulent infections and is seen in the present series mostly in the last decade.
Embolic lesions which are so characteristic of endocarditis are common in all types of endocarditis. Kidney involvement is more common in group I cases, while splenic in group II cases. Notable is the involvement of myocardium which is higher than in the reported series. , This is particularly common in post-operative endocarditis.
In brief, it can be said that the pattern of endocarditis has changed over last 50 years. The pattern now resembles to that reported from developed countries as far as the general incidence and etiologic organisms are concerned. The decrease in cases of endocarditis associated with rheumatic heart disease is gradual and is not of the degree reported in developed countries. The increasing incidence of endocarditis particularly the one supervening on normal hearts in the last decade calls for vigilance.
We are indebted to the Dean, King Edward VII Memorial Hospital and Seth G. S. Medical College, Bombay, and also to Dr. (Smt.) S. M. Sant, Prof. of Pathology and Microbiology, K.E.M. Hospital and Seth G.S. Medical College, for the permission to use the records.
We also thank Dr. A. K. Murthy, Resident Pathologist, King Edward Memorial Hospital, for the help in collecting the data.
[Figure 1], [Figure 2]
[Table 1], [Table 2], [Table 3], [Table 4]