The Spectrum of Severe Rheumatic Mitral Valve Disease in a Developing Country: Correlations among Clinical Presentation, Surgical Pathologic Findings, and Hemodynamic Sequelae

Methods

We retrospectively studied 748 consecutive patients with rheumatic heart disease who had mitral valve surgery between 1983 and 1986 at Baragwanath Hospital in South Africa. All patients were blacks living either in townships in the environs of the hospital (equivalent to inner-city areas in the metropolitan United States) or in rural areas. Eleven patients in whom concurrent infective endocarditis or primary degenerative leaflet disease (1 patient) confounded evaluation of the hemodynamic severity of the rheumatic lesion itself were excluded from the analysis. Of the remainder, 520 were female and 217, male. Patients ranged in age from 4 to 73 years (mean, 27 ±13 years [±SD]; median, 25 years); 271 (37%) were 20 years of age or younger. All were in New York Heart Association (NYHA) class 3 or 4 and receiving maximal medical treatment.

Preoperative Classification of Valve Lesions

Before surgery, all patients were examined independently by at least two cardiologists. Mitral valve lesions were classified as purely regurgitant, purely stenotic, or mixed according to recognized clinical, radiologic, and echocardiographic criteria [9]. Mitral regurgitation was considered to be pure when associated with unrestricted valve leaflet excursion and a normal mitral orifice area as assessed by two-dimensional echocardiography. Pure mitral stenosis was diagnosed when no clinical or echocardiographic evidence for regurgitation was found. Mixed mitral valve disease was diagnosed when features of both regurgitation and stenosis were present. Cases in which discordance existed among clinical, hemodynamic, and surgical pathologic assessments were excluded from the analysis.

Surgical Evaluation

Hemodynamic Assessment

The hemodynamic severity of valve lesions was confirmed by measurements of left atrial, left ventricular, and right ventricular pressures using fluid-filled catheters connected through strain gauge transducers (Statham, Oxnard, California) to a multichannel recorder (Honeywell Meddars, Lenexa, Kansas). The transmitral end-diastolic gradient (timed according to the R wave on the electrocardiogram) was determined by analysis of simultaneous left atrial and left ventricular pressure recordings. The ratio of left atrial “V” wave to mean left atrial pressure was calculated as an index of the hemodynamic severity of mitral regurgitation [10]. In 74 patients (10%), pressure measurements were not obtained for technical reasons or because of hemodynamic instability.

Surgical Anatomy

After cardiotomy, mitral valves were examined independently by two experienced surgeons according to a standard protocol, implemented in all patients having rheumatic valve surgery at Baragwanath Hospital between 1983 and 1986 at the instigation of one of the surgeons who was then doing research in this area. Valve leaflets were assessed for their pliability as well as for evidence of retraction (scarring) or calcification. Valve commissures were evaluated for evidence of fusion. Mitral valve prolapse was diagnosed at operation if the free edge of one or more scallops of a leaflet, almost invariably the anterior, could be retracted toward the left atrium beyond the free margin of the complementary leaflet and above the plane of the mitral annulus without applying tension to the chordae tendineae [5, 11]. The size of the mitral annulus and the length of the chordae tendineae were assessed. The size of the mitral annulus was graded by inspection using a 4-point scale from 0 (normal size) to 3+ (markedly dilated). In addition, the annular diameter was measured using valve sizers in all patients who had insertion of mechanical prostheses or valve rings (Carpentier, Santa Ana, California). The chordae tendineae were assessed for length (elongated, shortened, or normal), fusion, and evidence of rupture. Among patients who had mitral valvuloplasty, the requirement for chordal shortening procedures was used as ancillary evidence for the presence of chordal elongation.

Surgical Classification of Valve Lesions

The hemodynamic and anatomic features that were considered diagnostic of pure mitral regurgitation included absent or minimal end-diastolic gradient across the mitral valve; pliable, freely mobile mitral valve leaflets; and absence of commissural fusion or subvalvular disease. Pure mitral stenosis was confirmed at surgery in patients with no clinical or echocardiographic evidence for mitral regurgitation when there was marked commissural fusion and a substantial transmitral end-diastolic gradient.

Assessment of Rheumatic Activity

Clinical Evaluation

Rheumatic activity was diagnosed preoperatively when there was serologic evidence for antecedent group A β-hemolytic streptococcal infection in addition to at least two major (or one major plus two minor) criteria of acute rheumatic fever (revised Jones criteria) [12].

Macroscopic Evaluation

Valves were examined during surgery for the macroscopic features of active rheumatic carditis that have been described [5, 13]: fibrinous pericarditis with epicardial involvement; pinhead vegetations on the free edges of the valve leaflets in the absence of infective endocarditis; and nonspecific signs of acute inflammation including edema, erythema, and hemorrhage within leaflet tissue.

Histologic Evaluation

Where sufficient cardiac tissue was available for adequate histologic assessment, the following light microscopic findings were used as criteria of rheumatic activity [13, 14]: fibrinoid necrosis in valve leaflet or annular tissue; polymorphonuclear or histiocyte infiltration; edema; and neovascularization.

Analysis of Data

For continuous variables, such as pressure data, analysis of variance with the Tukey allowance for multiple comparisons was used to compare three or more groups, and two-sample t-tests were used in instances when only two groups were compared. For categorical data, such as the prevalence of mitral valve prolapse, chi-square tests were used. Three x two contingency tables were constructed to establish the presence of intergroup differences; individual groups were then compared using the Bonferroni correction for multiple comparisons. Statistical differences were recorded using two-tailed P values.

Results

Correlation between Preoperative and Surgical Classification of Valve Lesions

Surgical and preoperative classifications of valve lesions were in agreement in 714 of 737 (97%) cases. The 23 patients in whom discordance existed between the preoperative and surgical assessment of the mitral valve lesion were excluded from the analysis.

Relative Prevalence of Valve Lesions

When the entire patient population was considered, the three types of mitral valve lesion were documented with similar frequency: Two hundred nineteen patients had pure mitral regurgitation (31%), 275 had pure stenosis (38%), and 220 had mixed lesions (31%). However, 36 of 46 patients (78%) who had surgery in the first 10 years of life had pure mitral regurgitation; pure regurgitation was the most common lesion in patients 20 years of age and younger, accounting for 58% of surgically treated rheumatic mitral valve disease (158 of 271 cases) in this age group. Eighty-nine percent (194 of 219) of patients with pure mitral regurgitation were 30 years of age or younger (Figure 1). In contrast, the prevalence of mitral stenosis increased with age. Only 20% of patients with pure stenosis (55 of 275) were less than 20 years of age. Mixed mitral valve disease also increased in frequency until the fourth decade, after which its prevalence declined slightly.

Hemodynamic Data

Average values for mean left atrial pressure were similarly increased in all three groups (24 mm Hg for pure regurgitation and pure stenosis and 25 mm Hg for mixed disease). Pure mitral regurgitation was also characterized by a markedly raised left atrial V wave (46 ±18 mm Hg) and an average V wave:mean left atrial pressure ratio of 1.9:1, indicating severe hemodynamic compromise [10]. In addition, left ventricular end-diastolic pressure was elevated, whereas the end-diastolic gradient across the mitral valve was trivial. Pure mitral stenosis was hemodynamically differentiated from pure regurgitation by a high transmitral end-diastolic gradient (15 ±6 mm Hg, P < 0.001 compared with pure regurgitation), an unremarkable left atrial V wave magnitude relative to the mean left atrial pressure, and a normal left ventricular end-diastolic pressure (9 ±15 mm Hg, P < 0.001 compared with pure regurgitation). Mixed lesions had hemodynamic features of both regurgitation and stenosis: The magnitude of the left atrial V wave was intermediate between values for pure regurgitation and stenosis, but the left ventricular end-diastolic pressure was not increased (10 ±5 mm Hg), owing to restricted left ventricular filling across the stenotic mitral orifice. Hemodynamic data are summarized in Table 1.

Diagnosis of Rheumatic Activity

Immediately before surgery, 107 patients (14%) were considered to have acute rheumatic fever with carditis according to the revised Jones criteria; this was used as presumptive clinical evidence of ongoing rheumatic activity. At surgery, macroscopic features of active carditis were found in 128 of 714 patients (18%). There was agreement between the preoperative clinical and macroscopic surgical diagnosis of rheumatic activity in 98 patients (Table 2). In 9 patients, the macroscopic findings did not confirm the preoperative clinical diagnosis of activity, whereas in 30 patients who did not fulfill the Jones criteria, cardiac tissues appeared to be acutely inflamed at surgery in the absence of clinical evidence for infective endocarditis. Of these 39 patients in whom there was discordance between preoperative clinical and macroscopic surgical diagnosis of ongoing rheumatic activity Table 2, histologic data were available in 18. The histologic findings confirmed the negative preoperative clinical diagnosis in 8 patients, the positive macroscopic surgical diagnosis in 6 patients, and the negative macroscopic surgical diagnosis in 4 patients. Pathologic specimens were obtained in a total of 304 patients, of whom 39 (13%) had histologic evidence for rheumatic activity. Twenty of these patients also tested positive according to clinical and macroscopic criteria, whereas 6 had macroscopic but not clinical features of active carditis. The remaining 13 patients had histologic features of active carditis in the absence of preoperative or surgical criteria for rheumatic activity. Conversely, clinical or macroscopic criteria were positive in 24 patients with no histologic evidence for disease activity.

Relation of Rheumatic Activity and Valve Lesion

Of 128 patients with surgically diagnosed rheumatic activity, 106 had pure mitral regurgitation (83%); 17, mixed mitral valve disease (13%); and 5, pure stenosis (4%) (Figure 2). Rheumatic activity was diagnosed at surgery in 47% of patients with pure regurgitation, 7% with mixed lesions (P < 0.001 compared with regurgitation), and only 2% with pure stenosis (P < 0.001 compared with regurgitation).

Relation of Rheumatic Activity and Age

Active carditis was not detected in any patient older than 30 years. Macroscopic evidence of ongoing rheumatic activity was found in 28 of 46 having surgery in the first decade of life (60%), 93 of 225 in the second decade (41%), and 7 of 185 in the third decade (4%) (see Figure 2).

Surgical Anatomy of Valve Lesions

Mitral Annulus

The dimensions of the mitral annulus were evaluated at surgery in 314 patients; 257 had dilated annuli (Table 3). Varying degrees of dilatation were recorded in 177 of 186 patients with pure regurgitation (95%) and in 65 of 78 with mixed lesions (83%) but in only 15 of 50 patients with pure stenosis (30%). Mitral prostheses were inserted in 546 patients. The average external diameter of these prostheses was 30 ±2 mm. Despite the young age of patients with pure mitral regurgitation (19 ±11 years), the mean prosthesis external diameter for this group (30 ±2 mm; n = 219) was similar to that for patients with mixed lesions (30 ±2 mm; age, 30 ±12 years; n = 217) and greater (P < 0.001) than that for patients with pure stenosis (29 ±2 mm; age, 31 ±12 years; n = 110).

Chordae Tendineae

Chordal length at surgery was documented in 340 patients. Chordae were elongated in 92% (159 of 173) of patients with pure regurgitation, in 45% (48 of 106) with mixed mitral valve disease, and in only 1 patient with mitral stenosis (Table 3). One hundred eleven patients in whom chordae appeared elongated at surgery had mitral valve repair. All required chordal shortening procedures to prevent anterior leaflet prolapse and to restore valve competency. In contrast, chordae appeared to be shortened in 90% (55 of 61) of patients with pure stenosis, 49% (52 of 106) with mixed mitral valve disease, and only 3% (5 of 173) with pure regurgitation. Chordal rupture was diagnosed at surgery in 57 patients, 44 of whom had pure mitral regurgitation; the remaining 13 patients with chordal rupture had mixed mitral valve disease. Rupture was confined to anterior leaflet chordae in 47 patients. In 1 patient, chordae to anterior and posterior leaflets were involved; another patient had isolated posterior leaflet chordal rupture; and a single patient with severe rheumatic activity had a torn anterolateral papillary muscle. In the remaining 7 patients, the site of rupture was not documented.

Valve Leaflets

Leaflet anatomy was described in 130 patients with pure rheumatic mitral regurgitation (see Table 3). One hundred twenty-three (95%) had pliable mitral leaflets without macroscopic evidence for scarring or commissural fusion. In contrast, leaflets were scarred and rigid (calcified) in 56% of patients with mixed lesions (105 of 187) and in 38% with pure stenosis (104 of 273). All had varying degrees of commissural fusion. The prevalence of leaflet rigidity increased with age. Leaflets were pliable in 72% of patients 30 years of age or younger and rigid in 63% of patients older than 40 years.

Prolapse of the anterior mitral leaflet was confirmed at surgery in 183 of 219 patients (84%) with pure regurgitation. In contrast, prolapse was found in only 22% of patients with mixed lesions and was not documented in any patients with pure stenosis (P < 0.001 compared with pure regurgitation). The prevalence of prolapse among patients with mixed mitral valve disease declined progressively with age, from 60% in the first decade to less than 10% in patients older than 40 years.

Discussion

In developed countries, hemodynamically severe rheumatic mitral valve disease generally presents in the fourth decade or later, with thickened scarred valve leaflets usually manifesting as mitral stenosis with or without concurrent regurgitation [15]. These lesions are the degenerative sequelae of previous rheumatic inflammation and represent the “burned out” end stages of the disease. In our study, the spectrum of hemodynamically severe rheumatic mitral valve disease differed entirely, in that one third of patients presented with pure mitral regurgitation, more often than not in association with uncontrolled ongoing rheumatic activity. Mild mitral regurgitation, manifested by an isolated systolic murmur, is commonly diagnosed in young patients with acute rheumatic fever and carditis but frequently disappears with effective treatment of the acute illness and continued prophylactic penicillin therapy [3, 16, 17]. Our data suggest that in developing countries, this lesion, if left untreated, may progress rapidly to a form of pure mitral regurgitation so severe that early surgery becomes mandatory. Neither the high relative prevalence of pure mitral regurgitation among hemodynamically significant rheumatic mitral valve lesions nor the close association between severe mitral regurgitation and uncontrolled disease activity has, to our knowledge, been described previously. The only published report of a demographically similar subgroup of rheumatic patients found acute carditis with verrucous rheumatic vegetations in 174 of 796 rheumatic patients (22%) studied at autopsy [4]. Sixty-two percent of those patients died before the age of 30 years, but the functional nature of their mitral lesions was not documented, and stenotic and regurgitant lesions were not differentiated. It is conceivable that severe pure regurgitation may have accounted for the unexplained high early mortality rate among the young rheumatic patients reported by Bland and Jones [1] in their benchmark longitudinal study. Many of their patients' deaths were ascribed to an enigmatic “myocardial factor” [1, 18] rarely recognized in rheumatic patients today [7].

Several explanations are possible for the apparent underestimation of the frequency of pure, severe regurgitation among hemodynamically unstable patients with rheumatic mitral valve disease. Jhaveri and colleagues [19] commented that pure mitral regurgitation may have been inappropriately excluded in patients with a mitral mid-diastolic murmur in their own and other studies that relied solely on clinical data. Several large series on pure mitral regurgitation did not include young patients; hence, this important subgroup may have been overlooked [19, 20]. Finally, the high prevalence of hemodynamically significant pure mitral regurgitation reported in this study may be a function of host, environmental, or streptococcal factors that are specific to the geographic area and population studied [21].

The hemodynamically significant regurgitant rheumatic mitral valve lesion has long been considered merely an anatomic variant of its stenotic counterpart, in which retraction of scarred valve leaflets has disrupted the integrity of the mitral seal [15, 20, 22]. Because this chronic degenerative process is usually accompanied by at least some fusion of mitral commissures, the relative prevalence of pure regurgitation among hemodynamically severe mitral valve lesions has consistently been reported to be low [1, 15, 20, 22]. Most of our patients with pure regurgitation had thin, pliable leaflets with none of the scarring, rigidity, calcification, commissural fusion, or chordal shortening to which the regurgitant rheumatic lesion has been traditionally attributed [15, 23]. Most had annular dilatation, chordal elongation, and anterior leaflet prolapse [5]. Annular dilatation has previously been documented in patients with rheumatic mitral regurgitation [7, 13, 22, 24-26], but its role in the pathogenesis of this lesion has not been emphasized. More recently, mitral valve prolapse has been noted in patients with mitral regurgitation after rheumatic fever [27]. Tomaru and coworkers [28] found histologic evidence of previous inflammation (which they assumed to be of rheumatic origin) in valves excised from a group of patients with severe mitral regurgitation caused by mitral valve prolapse. We recently reported 73 patients with pure, severe rheumatic mitral regurgitation, most of whom had marked annular dilatation, chordal elongation, and anterior leaflet prolapse [5]. Concurrent myxomatous involvement of the valve leaflets was observed in only one patient. Thus, the surgical pathologic findings and, presumably, the mechanism of pure severe mitral regurgitation appear to differ from those of the longer-term, slowly progressive stenotic and mixed lesions.

The most striking demographic feature of the hemodynamically significant pure regurgitant lesion reported in our study was its high prevalence among young patients. The mean age of patients with pure mitral regurgitation was 19 ±11 years. The high incidence of concurrent refractory active rheumatic carditis among patients with pure severe mitral regurgitation suggests that uncontrolled rheumatic activity is responsible for clinical expression of this form of rheumatic mitral valve disease that is rarely seen in developed countries. Indeed, histologic examination of valves removed from patients with severe pure rheumatic mitral regurgitation has shown a high incidence of acute plus chronic inflammatory changes, implying that clinically significant regurgitation usually develops after recurrent or persistent rheumatic activity and is virtually never the result of a single episode of carditis [5, 29]. In developing countries, where living conditions are unfavorable, medical facilities are less accessible, and rheumatic fever prophylaxis is suboptimal, the initial acute illness may go unchecked, and more severe disease patterns may be encountered.

Although patients with the more familiar degenerative forms of rheumatic mitral valve disease (mitral stenosis and mixed lesions) were older and demographically similar to those reported in developed countries [20, 22, 30, 31], 20% of our patients with mitral stenosis were 20 years of age or younger. Similar findings have been reported previously from developing countries [6, 33-37]. In contrast, only 1% of patients who have surgical treatment of mitral stenosis in developed countries are under 20 years of age [32]. This may reflect differences in the virulence of the primary rheumatic insult. In more developed communities, older patients with critical stenosis frequently lack a definite history of active rheumatic carditis, reflecting a milder subclinical primary illness with a lesser degree of postinflammatory commissural adhesion [38]. Subsequent progression of the stenotic lesion is thought to be caused by either undetected, recurrent, low-grade rheumatic activity with scarring of the valve [14] or mechanical trauma resulting from altered flow patterns across the deformed mitral orifice [38, 39]. The end point of this process is a deformed, rigid, calcified valve. In contrast, most of our young patients with mitral stenosis had thin, pliable leaflets with no macroscopic evidence of leaflet involvement other than commissural adhesions. These findings could be explained by earlier and more extensive adherence of the leaflet margins after more severe or prolonged acute exudative inflammation, resulting in hemodynamically significant stenosis even before the onset of chronic degenerative sequelae.

In the group of patients with mixed mitral valve disease, demographic and surgical pathologic characteristics of pure regurgitation and pure stenosis overlapped. In most patients, annuli were dilated, but chordae were either elongated with mitral valve prolapse (as in pure regurgitation) or shortened with fibrosed retracted leaflets, as evidenced by the 170 patients in our study with mixed mitral valve disease and no mitral valve prolapse.

Klibanoff and colleagues [40] found frequent dissociation of pathologic from clinical or laboratory evidence for “rheumatic activity”. In our study, the presence of acute rheumatic fever according to the Jones criteria was used as presumptive clinical evidence of ongoing “rheumatic activity”. The clinical diagnosis of activity correlated well with the macroscopic features of rheumatic activity at surgery, but preoperative predictors failed to detect rheumatic activity in 23% of cases with macroscopic evidence of active disease.

Our data provide new insights into the natural history and pathogenesis of rheumatic mitral valve disease. Differences in the time course, surgical anatomy, and relation to disease activity of pure regurgitant and stenotic lesions suggest different mechanisms for these two end points of a common primary pathologic insult. Despite the cross-sectional nature of this study and the highly selected patient population (hemodynamically severe rheumatic mitral valve disease requiring surgery), these findings may explain the high early morbidity and mortality of rheumatic heart disease in developing countries.