090795 Studying the Classics

648 THE NEW ENGLAND JOURNAL OF MEDICINE Sept. 7, 1995

CLINICAL PROBLEM-SOLVING

In this Journal feature, information about a real patient also be caused by reactivation of the old tuberculosis in- is presented in stages (boldface type) to an expert clinician, who fection. She could also be having pulmonary emboli, responds to the information, sharing his or her reasoning with since there is an increased incidence of pulmonary em- the reader (regular type). The author’s commentary follows. boli in pregnancy. I would also worry that hemoptysis and breathless- STUDYING THE CLASSICS ness could be due to left heart failure. In heart failure hemoptysis is caused by the rupture of fragile anasto- GEORGE E. THIBAULT, M.D. moses between the pulmonary and bronchial veins when they are subjected to high pressure. The patient A 25-year-old Hispanic woman in the 27th week is Hispanic; we have not been told where she spent her of pregnancy came to the emergency room com- childhood. Acute rheumatic fever and rheumatic heart plaining of cough and shortness of breath. She re- disease are still endemic in many developing countries. ported that she had been well until two days ear- This could be the clinical presentation of mitral steno- lier. In the past day she had also been nauseated, sis in a previously asymptomatic woman who became and she had vomited once. In the emergency room symptomatic only in the third trimester of pregnancy, she was found to have a temperature of 38°C when the cardiac output is at its maximum. The failure (100.5°F) and was thought to be dehydrated. In- to hear a murmur on cardiac examination does not ex- travenous fluids, ampicillin, and gentamicin were clude mitral stenosis, particularly in the presence of administered. She was given a prescription for tachycardia and a lot of respiratory noise. She could erythromycin and sent home. She returned the also have heart failure due to other causes, such as car- next day because she was more dyspneic and ex- diomyopathy. It is a bit early in the pregnancy for peri- plained that she had coughed up a half-cup of partum cardiomyopathy. blood. She also reported that she had been treated with antibiotics one month earlier for a pre- The serum sodium and potassium, blood urea sumed upper respiratory infection. Since then nitrogen, and serum creatinine concentrations she had been tired and short of breath. Four pre- were normal. The hematocrit was 35 percent; the vious pregnancies and deliveries had been unre- white-cell count was 9900 per cubic millimeter markable. Four years earlier she had been treated with 82 percent neutrophils and 4 percent band for tuberculosis. On examination her tempera- forms; the platelet count was 230,000 per cubic ture was 38.6°C (101.4°F), her pulse was 120 to 130 millimeter. The urinalysis was unremarkable. Ar- beats per minute, her blood pressure was 94/50 terial-blood gas values measured while the pa- mm Hg, and the respiratory rate was 30 per tient was breathing 2 liters of supplemental oxy- minute. The cardiac examination was considered gen per minute were as follows: partial pressure to be unremarkable, but the chest examination of oxygen, 61 mm Hg; partial pressure of carbon revealed bibasilar rhonchi and wheezes. Her arte- dioxide, 26 mm Hg; pH 7.44; and calculated bicar- rial oxygen saturation was 94 percent while she bonate concentration, 17 mmol per liter. Her chest was breathing oxygen at a rate of 2 liters per film is shown in Figure 1. minute. The patient is mildly anemic and has no leukocyto- At the time of her first visit to the emergency room, sis, but there is a slight leftward shift. She has a sub- the patient’s fever and symptoms were most suggestive stantial arterial–alveolar pressure gradient. The blood of a respiratory infection, but there also seemed to be a gas values are consistent with chronic respiratory alka- gastrointestinal component. A community-acquired fe- losis but she could also have a mild degree of metabolic brile illness with respiratory and gastrointestinal symp- acidosis. toms would be consistent with legionella infection; The chest film is quite interesting. The hilar vessels erythromycin may have been an appropriate drug. I do are plump, the upper-lobe vessels also appear to be en- not know how sick she looked at the time of the first larged, and there are increased interstitial markings. visit, so I cannot say whether further investigation This could be interstitial inflammation or interstitial should have been undertaken then. Because hemoptysis pneumonia, but it looks more like heart failure to me. is not a common feature of community-acquired pneu- There is also straightening of the left heart border, monias, I would want to consider either a more aggres- which is suggestive of left atrial enlargement, and some sive, necrotizing bacterial pneumonia or another proc- enlargement of her central pulmonary arteries. The ess responsible for the hemoptysis. Hemoptysis could x-ray film makes me suspect heart failure more strongly; the findings would be quite consistent with mitral From Brigham and Women’s Hospital, 75 Francis St., Boston, MA 02115, stenosis. I still cannot exclude the diagnosis of intersti- where reprint requests should be addressed to Dr. Thibault. tial pneumonitis. The chest film does not show evi-

Vol.333 No.10 CLINICAL PROBLEM-SOLVING 649

perform Gram’s staining of the fluid to look for bacte- ria and white cells. The fact that her jugular venous pressure is elevated means that she has right atrial hypertension and, by definition, right heart failure. That does not prove that she has left heart failure, although that is the most common cause of right heart failure. However, patients with severe pneumonia or recurrent pulmonary emboli can present with acute cor pulmo- nale and isolated right heart failure. Given her earlier x-ray film, these findings are most consistent with left heart failure leading to or coincident with right heart failure. We are told no gallops were heard. If this were a case of severe, dilated cardiomyopathy caused by acute myocarditis or of peripartum cardiomyopathy, I would expect to hear a loud third heart sound. In a patient Figure 1. Chest Film Obtained in the Emergency Room. who has a rapid heartbeat with a lot of respiratory noise, murmurs of mitral stenosis and the opening snap may not be recognized. There was only slight cardio- dence of lobar pneumonia, active tuberculosis, or find- megaly on the chest film. That would still be consistent ings suggestive of pulmonary emboli, though chest with early acute myocarditis or mitral stenosis. I would films are both insensitive and nonspecific with respect do an echocardiogram to evaluate the valves and myo- to the last diagnosis. cardial function. On the basis of the chest film, I believe tuberculosis, necrotizing bacterial pneumonia, and pulmonary em- The electrocardiogram is shown in Figure 2 and boli are unlikely. The probability of heart failure has the chest film obtained after the patient was intu- gone up, and I am particularly concerned about the bated in Figure 3. possibility that she has rheumatic heart disease. I would want to repeat the cardiac examination in a quiet room The electrocardiogram shows sinus tachycardia and and obtain an echocardiogram. I also would want to a vertical axis. This axis could be normal, but it could obtain expectorated sputum for examination, because I also be indicative of right heart strain. There is also ev- have not entirely excluded the possibility that she has idence of left atrial enlargement. The vertical axis and an atypical pulmonary infection. left atrial enlargement are consistent with mitral stenosis, though neither of these abnormalities is specific. The patient was treated with fluids and intrave- She now has progressive opacification of both lung nous erythromycin (500 mg every six hours), but fields. What was previously an interstitial process is she became increasingly dyspneic and was airlift- now an alveolar process. With this degree of radio- ed to a tertiary care center. On arrival she had an graphic abnormality, one cannot be specific about the arterial oxygen saturation of 40 percent, and she cause. It is consistent with cardiogenic pulmonary ede- was immediately intubated. The fluid suctioned ma, but it is also consistent with noncardiogenic pulmo- from the endotracheal tube was pink and frothy. nary edema or any progressive lung injury leading to The physical examination revealed jugular venous the adult respiratory distress syndrome. distention to the angle of the jaw. The cardiac examination disclosed tachycardia (120 beats per minute and regular); S1 and S2 were audible; no gallops or murmurs were heard. Examination of I aVR V1 V4 the chest revealed diffuse rhonchi. She had no cyanosis, clubbing, or edema of the arms and legs.

II aVL V2 V5 The patient did not get better on erythromycin and got worse when she received a large volume load. If she did have rapidly progressive pneumonia, the failure to III aVF V V improve in the ﬁrst 24 hours does not necessarily mean 3 6 that the wrong antibiotic was chosen. But I am even more suspicious that she has something other than infection. Pink frothy ﬂuid coming from the endotracheal II tube is characteristic of pulmonary edema and points Figure 2. Electrocardiogram and Lead II Rhythm Strip Obtained strongly to severe heart failure. I would still want to at the Time of Transfer to a Tertiary Care Center.

650 THE NEW ENGLAND JOURNAL OF MEDICINE Sept. 7, 1995

Neelsen stains of three sputum samples were negative for acid-fast bacilli. Sputum cultures were negative. The results of thyroid-function tests were normal. Viral studies were negative. The patient was treated with digoxin and furosemide. She delivered a healthy child and later underwent balloon mitral valvuloplasty, with an excellent result.

Mitral stenosis masquerading as pneumonia! I think the main problem was that no one thought of the diagnosis. In any event, it was a happy ending.

COMMENTARY This is, in fact, a classic case of a previously common, and now relatively uncommon, disease. The doc- tors caring for this patient probably never thought of Figure 3. Chest Film Obtained at the Time of Transfer. the diagnosis of mitral stenosis. It may be that none of them had ever actually seen a patient with mitral ste- The patient was treated with ampicillin, genta- nosis; certainly pulmonary infection is far more com- micin, erythromycin, and isoniazid. A Swan–Ganz mon. On the other hand, even without consideration of catheter was inserted and the following initial mitral stenosis, the evidence for heart failure should findings were obtained: pulmonary-artery pres- have led them to the differential diagnosis of heart fail- sure, 58/39 mm Hg; pulmonary-capillary wedge ure in a young pregnant woman. It shows once again pressure, 35 mm Hg; central venous pressure, 20 that when parts of the patient’s clinical course and mm Hg; cardiac output, 4.75 liters per minute; and some of the findings are incongruent with the assigned 5 systemic vascular resistance, 779 dyn •sec•cm . diagnosis, we must broaden the differential diagnosis The patient was treated with furosemide; her and reexamine the primary data. systolic pressure fell to 60 mm Hg and metabolic The natural history of mitral stenosis in the era be- acidosis developed. Dopamine was given, but the fore mechanical intervention has been well described, fetus was apparently still viable. and because the disease has become so uncommon in the United States, the description is worth reiterat- This patient was treated with four antibiotics for pos- ing.1-3 Mitral stenosis is virtually always a sequela of sible bacterial pneumonia, atypical pneumonia, and tu- acute rheumatic fever, although there are rare congen- berculosis. She is clearly very sick, and it is hard to ital cases that usually present in infancy or childhood, fault this decision, but I fear that the preoccupation and even rarer cases of obstruction of the mitral valve with finding an infectious cause for her illness may have due to other inflammatory or infiltrative processes.4 Mi- distracted her physicians from making the logical di- tral stenosis primarily affects women, and as many as agnosis of heart failure. The hemodynamic measure- 40 percent of the patients have no history of acute ments confirm my impression that she is in severe rheumatic fever. After an episode of rheumatic fever, heart failure, and I suspect that she has been in heart the physical findings of mitral stenosis evolve over the failure all along. In addition to severe myocardial dis- next 5 to 15 years.5 This natural history may be con- ease and mitral stenosis, aortic stenosis can cause pul- densed into a much shorter period in certain geograph- monary edema that is refractory to therapy. At this ic regions, such as Asia, Africa, and Central America.6 point, we need to focus our attention on the cause of the This acceleration may be due to climatic conditions, severe heart failure, and I am even more eager to see socioeconomic factors, or repeated streptococcal infec- her echocardiogram. I still think she may have severe tions. mitral stenosis. The clinical presentation of mitral stenosis may take one of several forms.7,8 Patients may be entirely asymp- An echocardiogram showed severe rheumatic tomatic until an acute stress causes tachycardia and in- mitral stenosis with a mean mitral-valve gradient creased cardiac output, leading to an abrupt elevation of 17 mm Hg and a mean valve area of 0.8 cm2. of left atrial pressure and acute pulmonary edema. The Mild mitral regurgitation was present. The left precipitating factor may be a febrile illness, the onset atrium was enlarged, but there were no thrombi. of rapid atrial fibrillation, intense physical activity, or The pulmonary-artery systolic pressure was esti- pregnancy. In fact, the onset of symptoms at the begin- mated to be 75 to 85 mm Hg. The right atrium ning of the third trimester of pregnancy is common was enlarged. The left ventricle and aortic valve because it is a time of maximal hemodynamic stress were normal. Blood cultures were negative. Ziehl– caused by the increased blood volume, cardiac output,

Vol.333 No.10 CLINICAL PROBLEM-SOLVING 651 and heart rate.9 The increased heart rate is poorly tol- and the opening snap. The optimal situation for hear- erated because the diastolic time is shortened and the ing the low-frequency rumble is a quiet room, with the left atrium cannot empty adequately. Similarly, in- patient lying in a left lateral decubitus position and the creased flow across the valve in any high-output state bell of the stethoscope lightly applied at the cardiac leads to increased pressures, because the valve orifice apex. In truth, “silent” mitral stenosis is usually the re- is fixed. One large study of pregnancy and rheumatic sult of inadequate auscultation.12 There are patients, heart disease showed that in 75 percent of the patients however, in whom the cardiac output is so low that the who had heart failure it developed after the 26th week murmur may become inaudible. of pregnancy.10 The clinical presentation in this case In addition to the murmur, there are other helpful was classic. The patient was at the beginning of her findings on auscultation. A loud first heart sound may third trimester and had tachycardia when she first pre- first suggest the diagnosis, but this sound may diminish sented. She may have had an upper respiratory infec- as the valve becomes more fixed and immobile. The tion that initially contributed to her increased hemo- opening snap of the stenotic valve may be mistaken for dynamic load, but her fever may also have been due a third heart sound, but it occurs earlier in diastole and to pulmonary edema. A vicious circle was thus estab- has a higher frequency. The length of time between the lished: increased left atrial pressure led to tachycardia, opening snap and the second heart sound correlates in- hypoxemia, and fever, which led to further increases in versely with the left atrial pressure.13 The higher the left atrial pressure. left atrial pressure, the closer the opening snap will be As the mitral valve narrows, exercise intolerance and to the second heart sound; this finding may be helpful dyspnea on exertion due to gradually rising left atrial in serial physical examinations of a patient with known pressure slowly develop in many patients. Once the mitral stenosis. An accentuated pulmonary component valve is damaged, further narrowing does not require of the second sound, indicative of pulmonary hyperten- new episodes of rheumatic fever or other acute insults. sion, may first suggest mitral stenosis by leading to an In some patients cardiac output decreases in response exploration of the causes of pulmonary hypertension. to the valve stenosis, and thus the left atrial pressure None of these findings were recognized in this patient, remains nearly normal. These patients have fewer partly because of her tachycardia and loud lung sounds, symptoms of pulmonary congestion and complain of fa- but the fact that they were not detected is also a re- tigue and exercise intolerance caused by the limitation minder that physical-examination skills have atrophied of cardiac output. A small number of patients present in our era of sophisticated imaging studies. with symptoms and signs of pulmonary hypertension Ultrasonography was first used in cardiology in the leading to right heart failure. Such patients appear to diagnosis of mitral stenosis,14 and one of the first uses have disproportionate reactivity of the pulmonary vas- of two-dimensional echocardiography was the estima- culature for any given degree of left atrial hypertension. tion of mitral-valve area in mitral stenosis.15 Doppler They rapidly go through the phase of left atrial hyper- flow studies are helpful for assessing coexisting mitral tension and first come to medical attention with pulmo- regurgitation and for estimating the degree of pulmonary hypertension and right ventricular failure. Ac- nary hypertension, an important prognostic factor in cordingly, mitral stenosis should always be part of the the natural history of the disease. Cardiac catheteriza- differential diagnosis of “idiopathic” pulmonary hyper- tion has been the “gold standard” for quantifying the tension and unexplained right heart failure. severity of mitral stenosis. A calculated mitral-valve Before interventions were available to relieve mitral area of 1.0 cm2 or less per square meter of body-sur- stenosis, approximately 40 percent of patients were face area, according to the formula of Gorlin and Gor- dead 10 years after the onset of symptoms, and at 20 lin,16 is evidence of “critical” mitral stenosis and con- years only 20 percent were still alive. Sixty percent of stitutes an indication for mechanical intervention in a the deaths were attributable to heart failure, but 30 symptomatic patient. Though cardiac catheterization percent were caused by systemic or pulmonary embo- is still required in patients who are suspected of having li.11 Atrial fibrillation is an almost inevitable conse- coronary artery disease, other valvular lesions, or my- quence of the long-standing left atrial hypertension, ocardial disease, young patients with “pure” mitral and systemic embolization, which affects 30 to 40 per- stenosis (such as this patient) can be treated with me- cent of patients over the course of the disease if they do chanical interventions on the basis of noninvasive not receive anticoagulant therapy, is often a devastat- studies alone. ing complication. The medical treatment of severe mitral stenosis is It should be possible to make the diagnosis of mitral difficult, as this case illustrates. Most of the interven- stenosis on physical examination, but it can be difficult tions that are used to treat other causes of heart failure to do so. This patient’s story is a familiar one in that either are ineffective or may have adverse consequenc- the murmur of mitral stenosis was not recognized by es. Digoxin is helpful in controlling the ventricular re- several observers. As the discussant noted, when a pa- sponse in a patient with atrial fibrillation, but it has no tient with mitral stenosis has tachycardia and heart role in lowering the left atrial pressure in patients in failure, it is particularly difficult to hear the murmur normal sinus rhythm, because depressed left ventricu-

652 THE NEW ENGLAND JOURNAL OF MEDICINE Sept. 7, 1995

lar function is not the cause of the left atrial hyperten- REFERENCES sion. Vasodilators, which are effective in treating heart 1. Bland EF, Jones TD. Rheumatic fever and rheumatic heart disease: a twenty failure due to decreased left ventricular systolic func- year report on 1000 patients followed since childhood. Circulation 1951;4: tion or regurgitant valvular lesions, may exacerbate 836-43. symptoms in patients with mitral stenosis. Such pa- 2. Wilson JK, Greenwood WF. The natural history of mitral stenosis. Can Med Assoc J 1954;71:323-31. tients are unable to increase cardiac output in response 3. Olesen KH. The natural history of 271 patients with mitral stenosis under to vasodilation, and the decrease in systemic blood medical treatment. Br Heart J 1962;24:349-57. pressure leads to reflex tachycardia. Diuretics lower 4. Braunwald E. Mitral stenosis. In: Braunwald E, ed. Heart disease: a textbook of cardiovascular medicine. 4th ed. Philadelphia: W.B. Saunders, 1992: left atrial pressure and relieve symptoms of congestion, 1007-18. but at the risk of markedly depressing cardiac output. 5. Walsh BJ, Bland EF, Jones TD. Pure mitral stenosis in young persons. Arch Intern Med 1940;65:321-7. Hence, mechanical intervention is the only effective 6. Joswig BC, Glover MU, Handler JB, et al. Contrasting progression of mitral means to improve hemodynamic function in a patient stenosis in Malayans versus American-born caucasians. Am Heart J 1982; with severe symptoms. 104:1400-3. 7. Wood P. An appreciation of mitral stenosis. I. Clinical features. BMJ 1954; Patients in whom pulmonary hypertension has devel- 1:1051-63. oped as a result of mitral stenosis are almost always 8. Hugenholtz PG, Ryan TJ, Stein SW, Abelmann WH. The spectrum of pure symptomatic, and they are candidates for mechanical mitral stenosis: hemodynamic studies in relation to clinical disability. Am J Cardiol 1962;10:773-84. intervention to lower the left atrial pressure and im- 9. Elkayam U, Gleicher N. Hemodynamics and cardiac function during normal prove flow across the mitral valve. Pulmonary hyper- pregnancy and the puerperium. In: Elkayam U, Gleicher N, eds. Cardiac problems in pregnancy: diagnosis and management of maternal and fetal tension increases the risks associated with all interven- disease. 2nd ed. New York: Alan R. Liss, 1990:5-24. tions, but it is not a contraindication to intervention to 10. Szekely P, Turner R, Snaith L. Pregnancy and the changing pattern of rheu- alter the natural history of the disease. Unlike pulmo- matic heart disease. Br Heart J 1973;35:1293-303. 11. Rowe JC, Bland EF, Sprague HB, White PD. The course of mitral stenosis nary hypertension due to increased arterial flow from without surgery: ten- and twenty-year perspectives. Ann Intern Med 1960; left-to-right cardiac shunts (Eisenmenger’s syndrome), 52:741-9. the pulmonary hypertension due to mitral stenosis is 12. Harvey WP. Silent valvular heart disease. In: Likoff W, ed. Valvular heart disease. Vol. 5. No. 2 of Cardiovascular clinics. Philadelphia: F.A. Davis, reversible. If the left atrial pressure is normalized, the 1973:77-95. pulmonary-artery pressures will fall.17 13. Ebringer R, Pitt A, Anderson ST. Haemodynamic factors influencing opening snap interval in mitral stenosis. Br Heart J 1970;32:350-4. Closed mitral-valve commissurotomy was one of the 14. Zaky A, Nasser WK, Feigenbaum H. A study of mitral valve action recorded first successful surgical interventions for cardiac dis- by reflected ultrasound and its application in the diagnosis of mitral stenosis. ease.18 Open surgical valvotomy or repair and mitral- Circulation 1968;37:789-99. 15. Henry WL, Griffith JM, Michaelis LL, McIntosh CL, Morrow AG, Epstein valve replacement are now common for severe mitral SE. Measurement of mitral valve orifice area in patients with mitral valve stenosis. In properly selected patients, however, per- disease by real-time, two-dimensional echocardiography. Circulation 1975; cutaneous balloon valvotomy successfully relieves mi- 51:827-31. 19-21 16. Gorlin R, Gorlin SG. Hydraulic formula for calculation of the area of the tral stenosis. Balloon valvotomy is likely to be suc- stenotic mitral valve, other cardiac valves, and central circulatory shunts. cessful when the valve is not severely thickened, when Am Heart J 1951;41:1-29. 17. Braunwald E, Braunwald NS, Ross J Jr, Morrow AG. Effects of mitral-valve the valve is not heavily calcified, when the mobility of replacement on the pulmonary vascular dynamics of patients with pulmo- the leaflets is preserved, and when there is little sub- nary hypertension. N Engl J Med 1965;273:509-14. valvular fusion.22 These factors are most likely to be 18. Ellis LB, Abelman WH, Harken DE. Selection of patients for mitral and aortic valvuloplasty. Circulation 1957;15:924-35. present in younger patients who are still in normal si- 19. Palacios I, Block PC, Brandi S, et al. Percutaneous balloon valvotomy for nus rhythm. In this case the patient appeared to be an patients with severe mitral stenosis. Circulation 1987;75:778-84. ideal candidate for this less invasive intervention. Bal- 20. Reyes VP, Soma Rajn B, Wynn J, et al. Percutaneous balloon valvuloplasty compared with open surgical commissurotomy for mitral stenosis. N Engl J loon valvotomy has been successfully used during preg- Med 1994;331:961-7. nancy to stabilize a patient until delivery at term.23 21. Carabello BA, Crawford FA. Therapy for mitral stenosis comes full circle. N Engl J Med 1994;331:1014-5. Despite a classic clinical presentation, the diagnosis 22. Abascal VM, Wilkins GT, O’Shea JP, et al. Prediction of successful outcome in the case of this patient was missed by many of those in 130 patients undergoing percutaneous balloon mitral valvotomy. Circula- who were caring for her. In medicine, as in art and lit- tion 1990;82:448-56. 23. Palacios IF, Block PC, Wilkins GT, Rediker DE, Daggett WM. Percutaneous erature, the study of the classics may provide guidance mitral balloon valvotomy during pregnancy in a patient with severe mitral and knowledge that are still relevant today. stenosis. Cathet Cardiovasc Diagn 1988;15:109-11.