Nitish Naik, VK Bahl

Department of Cardiology, All India Institute of Medical Sciences, New Delhi

Rheumatic fever and rheumatic heart disease continues to remain a major drain on the health resources of our country. Despite its near disappearance in the West nearly three decades ago, this disease still affects a large part of our population.¹ The exact incidence/prevalence of the disease is presently not accurately known. However, some recent data acknowledge the lack of decline of this disease.^2,3 This is particularly disheartening as very little has been achieved in the past century that has significantly altered the clinical course of rheumatic fever. Therapeutic options have been limited as most randomized trials raised doubts regarding the efficacy of even steroids, thought to be the mainstay of treatment of acute rheumatic fever (ARF). Numerous editorials, reviews and scientific statements have attested to the lack of benefit of steroids in preventing the sequelae of rheumatic fever.^{4 –9} In the absence of any useful therapeutic tool, it has been difficult to modify the natural history of ARF. Very few large, controlled, randomized trials have been conducted for evaluating therapeutic modalities for treatment of rheumatic fever. Many factors have contributed to this dampened enthusiasm. A poor understanding of the pathogenesis of this disorder along with the lack of an appropriate animal model of rheumatic fever has greatly hampered clinical progress. Also, the near disappearance of the disease in the West has also shifted focus from this disease.^1,10 However, the disease continues to be rampant in the developing world and hardly any work worthy of mention has been conducted from the Indian subcontinent.  This article reviews the therapeutic options in rheumatic fever and emphasizes the immense need for further research in this field.

The role of steroids for the treatment of ARF has been a controversial issue. The support for such therapy emanated from clinical reasoning supporting the use of strong anti-inflammatory agents to suppress the acute inflammatory response in ARF. As the process of valvular involvement progresses through three distinct stages, viz. an acute inflammatory stage, a second phase with granulomatous inflammation and a final phase of fibrosis, it seemed logical to limit valvular inflammation by the use of steroids. Initial case reports and nonrandomized observational studies supported the use of steroids and found a clinically favorable response following their use. However, randomized trials refuted these claims, as most trials did not find any significant benefit of steroids. From the large number of randomized trials conducted on the use of steroids in rheumatic fever, five trials have been identified that appropriately identified patients as per the Jones criteria, were properly conducted and involved clinical follow-up at one year.^11–15 However, all these trials, except the one conducted by the Rheumatic Fever Working Party (RFWP), were small, enrolled a small number of patients and employed varying doses of steroids for different periods. The largest of these trials was a joint study conducted from the United States, United Kingdom and Canada.¹¹ The results of this trial did not demonstrate any superiority of steroids over salicylates in preventing the development of subsequent rheumatic heart disease in treated children. Only one trial conducted by Dorfman et al.¹² demonstrated a favorable effect with steroids. One other trial by Stolzer et al.¹⁵ did show some benefit with steroids although it did not reach statistical significance. A meta-analysis of these trials was published in 1995—this also did not find any significant benefit of steroid therapy.¹⁶ However, the results of this meta-analysis were dominated by the negative results of RFWP. Hence, although the authors acknowledged that steroids were not useful as per analyzed data, one observation was apparent: a treatment benefit with steroids in the range of 10% was still possible.

 These trials have been briefly reviewed below to ascertain the quantum of benefit with steroids (Table 1). The first large, randomized, double-blind, placebo-controlled trial evaluating the role of steroids in ARF was published in 1955 by the RFWP.¹¹ This has been the largest trial conducted till date on therapy for ARF. The study enrolled 505 children less than 16 years of age in the United States, United Kingdom and Canada with ARF, fulfilling the Jones criteria. Patients were randomized to one of the three arms: ACTH 80–120 units and then tapered (n=162), or cortisone 300 mg/day (n=167) and then tapered, or aspirin 60 mg/lb/day and then tapered (n=168). Therapy was administered for a period of 6 weeks. Definite or questionable evidence of pre-existing heart disease was present in 128 patients. Clinical outcomes that were assessed included resolution of carditis, congestive heart failure (CHF), effect on heart size, pericarditis and chorea. Follow-up data at 1 year as well as at the end of 5 and 10 years have been published by this group.^17,18 The authors did not find any significant differences in all these outcomes except for an earlier disappearance of soft apical systolic murmurs. However, by 1 year, even this difference in clinical outcome was lost. Clinical evidence of carditis was present in 74, 89, and 89 patients in the ACTH, cortisone, and aspirin groups. Of these 35, 36, and 49 children, respectively, in the ACTH, cortisone and aspirin groups without previous rheumatic heart disease had grade II murmurs at enrollment. At the end of 1 year, the murmur persisted in the same grade in 10/35 (28%), 7/36 (19%), and 9/49 (18%) of these patients, respectively. It progressed to grade III in 3/35 (8.5%) in the ACTH group, 2/36 (5.5%) in the cortisone group, and 6/49 (12.2%) in the aspirin group. However, as the subgroups were small, these differences in the three groups did not reach statistical significance. The authors concluded that steroids did not have any beneficial effect in the treatment of ARF. The major criticism of this study has been its choice of a lower dose of steroids along with a shorter duration of therapy. Also, this study did not have adequate power to detect differences of 10% or less. Despite its negative results, some investigators felt that corticosteroids were helpful in sick patients.

Dorfman et al.¹² used hydrocortisone as an anti-inflammatory agent and reported the effects of this drug in 150 patients with ARF. This trial demonstrated a statistically significant benefit of steroids over salicylates. Trial patients comprised those with an initial episode of ARF. The interval between starting steroid therapy and the presumed onset of ARF was 18 days. Patients were randomized to one of the four groups: hydrocortisone, aspirin, hydrocortisone with aspirin, and no therapy.  Eighteen patients were dropped from the study due to various reasons. Hydrocortisone was given in a dose of 250 mg/day (in four divided doses) for 4 days to patients weighing more than 80 lb (200 mg to those weighing less than 80 lb). The dose was then decreased to 100 mg/day for 8 weeks. The drug was subsequently tapered off over 4 weeks to complete 12 weeks of steroid therapy. All patients on this dose developed cushingoid features. In the aspirin group, the dose of aspirin was monitored according to serum levels to maintain a target level of 20–30 mg/dl.Grade II and III murmurs were present in ^{16, 14, 13,} and 13 patients in the hydrocortisone, aspirin, hydrocortisone with aspirin, and no therapy groups, respectively. At 15 weeks, ^{5, 10, 3,} and 9 patients still had an auscultable murmur. At the end of 1 year, these figures changed to ^{2, 7, 2,} and 9 patients, respectively. If one combines the steroid groups and compares them with those on aspirin and on no specific anti-inflammatory agents, then 29 of 64 patients on steroids had grade II or III murmur while 28 of 67 patients in the other group had grade II/ III murmurs. 

At 1 year, 4 of the 29 patients in the steroid group and 16 of the 28 patients in the other group had persistent grade II/III murmurs (p=0.01). None of the patients on steroids developed new murmurs while 1 patient in the other group had a new murmur at 1 year. Similarly 1 patient on steroids and 5 on either aspirin or no therapy developed a new basal diastolic murmur at 1 year. However, meaningful analysis was precluded due to the small number of patients.

 Two separate studies were conducted by the RheumaticFever Study Group (RFSG) comparing the role of steroids and salicylates in ARF.¹³ The first study, published in 1960, enrolled 57 children, 12 years or older with clinical evidence of moderate-to-severe carditis within 4 weeks of onset of ARF. Twenty-nine patients were randomized to prednisone 60 mg/day for three weeks which was subsequently tapered gradually over a period of 9 weeks. In the salicylate group, 28 patients received aspirin in a dose of 50 mg/lb/day for 9 weeks which was then tapered off over 3 weeks. The authors concluded that steroid therapy was not better than salicylates in preventing the development of rheumatic heart disease. The effect of steroids in controlling CHF could not be assessed as 4 patients in the aspirin group who had CHF were switched to steroids when they fared poorly. Thus, although steroids did not improve the resolution of murmurs, the authors believed that steroids were life-saving in the presence of CHF.

A second study was published by this group in 1965 which reported the effects of a higher dose of prednisone on clinical outcomes.¹⁴ Seventy-three children with their first attack of ARF and clinical evidence of moderate-to-severe carditis were randomized to either of the two groups: oral prednisone in a dose of 3 mg/lb/day (to a maximum dose of 200 mg) for a week and then tapered or aspirin 50 mg/lb/day for 6 weeks. Therapy was initiated within 3 weeks of the initial clinical manifestation. After a 7-day course, 15 of 34 patients still had active disease; a repeat course of steroids could control activity in an additional 7 patients; 8 of 15 patients continued to have active disease despite an additional week of high-dose steroids. At 1 year, 55.8% of patients in the steroid group had residual heart disease while 66.6% in the aspirin group had residual damage. The higher proportion of patients with residual damage in the aspirin group was felt to be due to a larger number of patients with more severe carditis in this group and not due to a beneficial effect of steroid therapy. The authors concluded that although steroids did not significantly prevent residual damage at 1 year, they were more effective in controlling acute clinical manifestations.

Another randomized controlled trial was published by Stolzer et al. in 1955. ¹⁵ This study recruited 152 airmen with ARF from an Airbase hospital in the United States. The patients were randomized to three groups: 41 patients received cortisone, 53 received aspirin and a third group of 41 patients received ACTH. Therapy was initiated on an average on day 8 of illness and was continued for a period of 6 weeks. Cortisone was given intramuscularly in a dose of 300 mg on the first day followed by 200 mg/day for 4 days and then gradually tapered off over a period of 6 weeks. Salicylates were given in a dose of 60 mg/lb on days 1 and 2, 40 mg/lb on days 3 to 7, followed by 30 mg/lb from the second to the sixth week. Corticotrophin was started at a dose of 120 mg administered intramuscularly on days 1 to 4 which was tapered off over a period of 6 weeks. Data of 128 patients who completed follow-up were analyzed at 14 months. Eighteen percent of patients in the cortisone group had clinical evidence of carditis at onset of therapy as against 11.5% in the aspirin group, and 10.3% in the corticotrophin group. In patients with a significant murmur at the onset of therapy, the apical systolic murmur persisted in 29% of cortisone-treated patients, 44% of ACTH-treated patients, and 50% of aspirin-treated patients. Similarly, among patients who had a murmur at the end of therapy, the smallest number in whom a significant murmur could still be found at 14 months were in the cortisone-treated group. In patients without a murmur at the start of therapy, a new murmur developed in 13% of cortisone-treated patients, 27% of those in the ACTH group and 25% of the aspirin group. Three patients in the aspirin arm and one in the corticotrophin arm had evidence of prolonged rheumatic activity; none of the patients in the cortisone arm had signs of continued rheumatic activity. Thus, cortisone was best in preventing apical systolic murmurs and also in aiding their resolution. However, the number of patients in each subgroup was small and no definite conclusions could be drawn.

A meta-analysis of these five trials was performed by Albert et al.¹⁶ Based on this analysis, the authors concluded that corticosteroids were not more effective than salicylates in preventing the development of rheumatic heart disease (estimated odds ratio 0.88, 95% confidence interval: 0.53–1.46). The results of this meta-analysis were dominated by the negative result of the RFWP. Significant heterogeneity existed in results from individual studies and thus clubbing these trials together in a meta-analysis is also suspect. Due to the wide confidence intervals, especially when data are analyzed using patients with carditis at admission, the authors concluded that the question whether steroids are marginally superior to salicylates still remains to be answered. A 10% treatment effect in favor of steroids could not be ruled out from these studies. Some authors have attempted to use more potent steroids such as methylprednisolone as anti-inflammatory therapy in ARF.

However, such reports are few, involving a small number of patients and with a limited follow-up.^19,20 No systematic studies have been conducted using this drug to date.

The lack of beneficial effects with corticosteroids prompted a search for alternative agents in ARF. An important, albeit negative prospective, double-blind, randomized placebo-controlled study was conducted from Auckland based on the premise that altering the immunological status with intravenous immunoglobulin (IVIG) would reduce the extent and severity of carditis in ARF.²¹ Patients were randomly allocated to IVIG (1 g/kg on days 1 and 2 and repeated again on days 14 and 28 in a dose of 0.4 g/kg) or placebo. Twenty-nine patients were enrolled in the IVIG arm and 32 received placebo; of these, 17 and 22, respectively, had evidence of carditis. Eight patients without carditis at admission developed carditis (subclinical in 4) 2 to 6 weeks later. Four patients in the IVIG group had evidence of moderate-to-severe mitral regurgitation at admission. At the end of 1 year, 2 still continued to have moderate-to-severe mitral regurgitation. In the placebo group the corresponding figures were 6/32 and 5/32, respectively. However, due to the small numbers in these subgroups, meaningful subgroup analysis is not feasible.Overall, 35% of patients in the IVIG arm and 27% in the placebo arm did not have evidence of carditis at the end of 1 year (p=NS). This study was designed to detect a 60% improvement in the IVIG arm and 25% improvement in the control arm at 1 year. This would have shown whether IVIG was useful in ARF, as it was found to be in Kawasaki disease. As only a 4%–5% improvement was seen in the IVIG arm, this study was underpowered to detect a smaller yet significant benefit with IVIG.

Whither Steroids?: Conventional teaching thus still does not lend credence to the role of steroids in ARF. Largely, three distinct groups of patients with ARF and carditis can be identified: those with evidence of subclinical or mild carditis, those with severe carditis and evidence of CHF, and those with carditis and marked valvular involvement without CHF. Steroids are clearly of no significant benefit in the first group as the natural history of rheumatic heart disease is very favorable in these patients even without any therapy. In the second group, steroids are useful for ameliorating CHF and have been life-saving in some instances. It is in the third group that the role of steroids remains controversial. Even a 10% absolute risk reduction in the resolution of carditis would be significant in the absence of any other useful therapeutic agent. Limited data from some South African countries have shown lower valve replacement rates in centers such as Capetown where steroids are used routinely in patients with rheumatic carditis.²² In contrast, valve replacement is performed more frequently in Johannesburg where steroids are not given to patients with carditis. However, disappointingly, no large studies have been conducted from the Indian subcontinent that have documented/refuted these contentions. Large randomized trials are clearly required to resolve this dilemma for the developing world. Assuming a 10% absolute risk reduction in the treatment arm and 35% resolution rate in the control arm, for a study to achieve a statistical significance of 0.05 and a power of 90%, the study sample would need to include 1046 patients with moderate-to-severe rheumatic carditis. If a 10% relative reduction is seen in the treatment arm, over 5000 patients in each arm would be required to demonstrate a statistically significant benefit. It would be difficult for a single center to recruit this number of patients and a multicenter study is required.

Role of Immunomodulators: Steroids have not been shown to modify the natural history of most chronic inflammatory disorders. Thus the role of other immunomodulatory agents needs to be explored. Immunological abnormalities involving both humoral and cell-mediated immunity have been demonstrated in patients with rheumatic heart disease.²³ A variety of streptococcal antigens have been identified along with a host of cross-reactive antibodies. A chronic autoimmune cell-mediated injury may possibly be involved in producing fibrosis and valvular destruction. Indeed, the valvular endothelium has been shown to react with various cell adhesion molecules on CD4+ and CD8+ lymphocytes.^24,25In such a case, the role of various anticytotoxic drugs such as anti-tumor necrosis factor (TNF) in the prevention of rheumatic heart disease needs to be explored. Whether other immunosuppressants or immunomodulators such as cyclosporine or methotrexate could possibly be of any use is speculative. The duration of such a therapy would also be difficult to decide as current markers of disease activity are nonspecific. Certain manifestations of rheumatic heart disease such as mitral stenosis, the commonest manifestation of chronic rheumatic heart disease, develop insidiously over a period of many years. It is possible that subclinical occult "rheumatic" activity continues unabated in some individuals in which case some form of chronic immunosuppressive therapy may prove to be beneficial. However, these suppositions are speculative although some of them may hold some promise in the future. 

Supportive Therapy: All patients with rheumatic carditis are advised bed rest and receive penicillin to eradicate pharyngeal streptococci. In patients with severe mitral regurgitation, arterial vasodilators including intravenous nitroprusside have been shown to be useful in controlling CHF. In patients with intractable heart failure not responsive to drugs, valve replacement surgery has been shown to be of benefit. In such patients amelioration of regurgitation controls heart failure. This demonstrates that CHF is not due to myocardial dysfunction but due to primary valvular dysfunction.^26,27

Rheumatic fever continues to plague the developing world and poses a serious public health problem. Many children, adolescents and young adults still fall prey to this formidable disease which has defied therapy over the past century. The role of steroids remains to be clarified. It is clearly not required in patients with mild carditis and has been shown to be life-saving in patients with severe carditis and CHF. Whether steroids can prevent the development of subsequent rheumatic heart disease in some patients is not clear, although most published randomized trials do not lend support to this contention. New, more effective therapeutic measures are urgently required to battle the ravages of streptococcal throat infections.

Correspondence:
Professor VK Bahl, 
Department of Cardiology, 
All India Institute of Medical Sciences, 
New Delhi 110029.
e-mail: vkbahl2002@yahoo.com

References

1. Gordis L. The virtual disappearance of rheumatic fever in the United States: lessons in the risk and fall of disease. Circulation 1985; 72: 1155–1162

2. Padmavati S. Rheumatic fever and rheumatic heart disease in India at the turn of the century. Indian Heart J 2002; 53: 35–37 

3. Lalchandani A, Kumar HRP, Alam SM, Dwivedi RN, Krishna G, Srivastava JP, et al. Prevalence of rheumatic fever and rheumatic heart disease in rural and urban school children in district Kanpur [Abstr]. Indian Heart J 2000; 52: 672

4. Stollerman GH. Rheumatic fever. Lancet 1997; 349: 935–942 

5. Dajani AS. Rheumatic fever. In: Braunwald E, Zipes DP, Libby P (eds). Heart disease. 6th ed. Philadelphia: WB Saunders; 2001, pp. 2192–2198

6. Chakko S, Bisno AL. Acute rheumatic fever. In: Fuster V, Alexander RW, O’Rourke RA (eds). The heart. 10th edition. New York: McGraw Hill; 2001, pp. 1657–1665

7. da Silva NA, de Faria-Pereira BA. Acute rheumatic fever: still a challenge. Rheumatic Dis Clin North Am 1997; 23: 545–568 

8. Thatai D, Turi ZG. Current guidelines for the treatment of patients with rheumatic fever. Drugs 1999; 57: 545–555 

9. Stollerman GH. Rheumatic fever in the 21st century. Clin Infect Dis 2001; 33: 806–814

10. Kaplan EJ. Global assessment of rheumatic fever and rheumatic heart disease at the close of the century. Influence and dynamics of populations and pathogens: a failure to realize prevention? Circulation 1993; 88: 1964–1972

11. Rheumatic Fever Working Party (RFWP) of the Medical Research Council of Great Britain, and the Subcommittee of the Principal Investigators of the American Council on Rheumatic Fever and Congenital Heart Disease, American Heart Association. The treatment of acute rheumatic fever in children: a cooperative clinical trial of ACTH, cortisone and aspirin. Circulation 1955; 2: 343–371

12. Dorfman A, Gross JI, Lorincz AE. The treatment of acute rheumatic fever. Pediatrics 1961; 27: 692–706 

13. Combined Rheumatic Fever Study Group (RFSG). A comparison of the effect of prednisone and acetylsalicylic acid on the incidence of residual rheumatic carditis. N Engl J Med 1960; 262: 895–902 

14. Combined Rheumatic Fever Study Group (RFSG). A comparison of short-term intensive prednisone and acetylsalicylic acid therapy in the treatment of acute rheumatic fever. N Engl J Med 1965; 272: 63–70

15. Stolzer BL, Houser HB, Clark EJ. Therapeutic agents in rheumatic carditis. Arch Intern Med 1955; 95: 677–688 

16. Albert DA, Harel L, Karrison T. The treatment of rheumatic carditis: a review and meta-analysis. Medicine 1955; 74: 1–12

17. Rheumatic Fever Working Party of the Medical Research Council of Great Britain and the Subcommittee of the Principal Investigators of the American Council on Rheumatic Fever and Congenital Heart Disease, American Heart Association. The natural history of rheumatic fever and rheumatic heart disease. Ten-year report of a cooperative clinical trial of ACTH, cortisone and aspirin. Circulation 1965; 32: 457–476

18. Rheumatic Fever Working Party of the Medical Research Council of Great Britain and the Subcommittee of the Principal Investigators of the American Council on Rheumatic Fever and Congenital Heart Disease, American Heart Association. The evolution of rheumatic heart disease in children. Five-year report of a cooperative clinical trial of ACTH, cortisone and aspirin. Circulation 1965; 32: 457–476 

19. Akcoral A, Oran B, Tavli V, Unal N, Cevik N. Effects of high-dose intravenous methylprednisolone in children with acute rheumatic carditis. Acta Pediatr Jpn 1996; 38: 28–31

20. Camara EJ, Braga JC, Alves-Silva LS, Camara GF, da Silva Lopes AA. Comparison of an intravenous pulse of methylprednisolone versus oral corticosteroids in severe acute rheumatic carditis: a randomized clinical trial. Cardiol Young 2002; 12: 119–124

21. Voss LM, Wilson NJ, Neutze JM, Whitlock RML, Ameratunga RV, Cairns LM, et al. Intravenous immunoglobulin in acute rheumatic fever. A randomized controlled trial. Circulation 2001; 103: 401–406

22. Visvanathan K, Manjarez RC, Zabriskie JB. Rheumatic fever. Curr Treat Options Cardiovasc Med 1999; 1: 253–258 

23. Veasy LG, Hill HR. Immunologic and clinical correlations in rheumatic fever and rheumatic heart disease. Pediatr Infect Dis J 1997; 16: 403–407

24. Roberts S, Kosanke S, Terence DS, Jankelow D, Duran CM, Cunningham MW. Pathogenic mechanisms in rheumatic carditis: focus on the valvular endothelium. J Infect Dis 2001; 183: 507–511 

25. Galvin JE, Hemric ME, Kosanke SD, Factor SM, Quinn A, Cunningham MW. Induction of myocarditis and valvulitis in lewis rats by different epitopes of cardiac myosin and its implications in rheumatic carditis. Am J Pathol 2002; 160: 297–306

26. Marcus RH, Sareli P, Pocock WA, Meyer TE, Magalhaes MP, Griuvc T, et al. Functional anatomy of severe mitral regurgitation in active rheumatic carditis. Am J Cardiol 1989; 63: 577–584 

27. Essop MR, Wisenbaugh T, Sareli P. Evidence against a myocardial factor as the cause of left ventricular dilation in active rheumatic carditis. J Am Coll Cardiol 1993; 22: 826–829