Breathlessness in microvascular angina

Respiratory Medicine (1994) 88, 731-736

Original Articles

Breathlessness in microvascular angina A. TWEDDEL*, R. CARTER'~, S. W . BANHAM~ AND I. HUTTON*

*Department of Medical Cardiology and ~fDepartment of Respiratory Medicine, Royal Infirmary, Glasgow, U.K.

In patients with microvascular angina (MA), there is some evidence from studies of plethysmography, that there are widespread microvascular abnormalities. In addition to exertional chest pain, all these patients complain of breathlessness, with no evidence of airways obstruction or resting left ventricular dysfunction. Progressive exercise testing was performed in 12 age and sex matched controls and 12 patients (three males), in whom the diagnosis of M A was established on the basis of exertional chest pain, abnormal thallium scans, and an attenuated myocardial flow response to a vasodilator challenge, with angiographically entirely normal epicardial vessels. Symptom limited exercise was performed with on line ventilation and expired gas analysis, measuring minute ventilation, oxygen consumption and carbon dioxide production and arterial blood gas values using a transcutaneous system. Anaerobic threshold was calculated by curve fitting a plot of oxygen consumption against carbon dioxide production. Compared to controls (49-7 4- 7.3 SD % predicted maximum VO2) in patients with MA, the anaerobic threshold was reduced (41-6 4- 5'82; P<0'02) although still within accepted normal limits. Maximal (symptom limited) oxygen consumption, as a percentage of predicted, was reduced 60"73 4- 16"51 compared to 87-21 4- 5"2 (P<0-003). The ventilatory response (VE/VCO 21 1- 1 CO a output) was significantly increased in the M A patients compared to controls (35'9 4-8-01 and 27.5 4- 3"08, respectively; P<0.02). This increased ventilatory response on exertion was associated with a raised physiological dead space at rest compared to controls (MA VD/VT 0"40 4- 0"03 vs. 0"35 4- 0"01; P<0.003) which persisted on exertion (at maximum exercise M A VD/Vx 0-33-4-0"05 vs. control 0"22 4- 0"05; P<0'002). There was no significant change in the mean transcutaneous PCO 2 at rest (36-2 4- 2.7) compared to maximum exercise (37'8 4- 3'8) in the M A patients suggesting that the increased ventilatory response in this group was not associated with inappropriate hyperventilation in this group. These results suggest that in patients with M A there is a significant degree of 'wasted ventilation' on exercise testing, consistent with areas of the lung being underperfused compared to their ventilation. The increased ventilatory requirement on exertion may contribute directly to the sensation of breathlessness experienced in these patients.

Introduction

vascular level (4,5). With the uncertainty as to the pathogenetic mechanisms and the absence of directly Since the advent of coronary angiography, there documented vascular abnormalities, there is no have been numbers of reports of patients with chest universal acceptance of this syndrome. Due to this pain and normal coronary arteries. Several published uncertainty many other explanations have been series, including the N I H sponsored coronary artery proffered, ranging from psychiatric illness (6), to surgery study (CASS) have reported an incidence of ascribing the symptoms to hyperventilation (7). 10-30% (1-3). In most patients, the pain is not of Hyperventilation is known to induce electrocardiocardiac but of musculoskeletal or oesophageal origin. graphic changes and the alkalosis induced by hyperIn a proportion of patients, however, the pain is of ventilation can produce coronary vasoconstriction cardiac origin, which can be inferred from an (8-10). impaired coronary dilator response at the microIt has been proposed that the underlying pathogenetic mechanism in this syndrome, termed 'microReceived 10 March 1993 and acceptedin revisedform 9 September vascular angina', is dysfunction of small intramural 1993. :~Towhom all correspondenceshould be addressedat: Department pre-arterial coronary arteries (11), although morphoRespiratory Medicine,Queen ElizabethBuilding,Royal Infirmary, logically these vessels appear normal (12,13). There is evidence that this may be a generalized disorder of AlexandraParade, GlasgowG31 2ES, U.K. 0954-6111/94/100731+06 $08.00/0

9 1994W. B. SaundersCompany Ltd

732

A. Tweddel et al.

vascular and nonvascular smooth muscle function. Sax et al. (14) have demonstrated coexisting abnormal forearm hyperaemic responses to transient ischaemia, Cannon et al. (15) oesophageal motility disorders and bronchoconstrictor responses to methacholine inhalation (16). A recent review by Cannon (17) suggests that the weight of evidence implicates endothelial dysfunction within the myocardial microvasculature. The aim of this study was to measure cardiorespiratory responses to exercise in patients considered to have microvascular angina and to assess involvement of the respiratory system. Methods PATIENT POPULATION

Twelve patients, of whom three were males aged 29-62 years were studied, a diagnosis of microvascular angina (MA) was established on the basis of exertional chest pain, thought to be anginal and present for more than 12 months, abnormal thallium scan (18) and an attenuated myocardial flow response to atrial pacing stress (19), with angiographically normal epicardial coronary vessels. No patient had major electrocardiographic abnormalities such as left bundle branch block, hypertension or abnormal echocardiography. Resting pulmonary function was normal as was the chest X-ray. At no time had there been any clinical suspicion of thrombo-embolic disease in any patient. Findings were compared with age and sex matched controls recruited as volunteers from the general population in whom there was no cardiorespiratory disease. PROGRESSIVE EXERCISE T E S T I N G

Symptom limited exercise tests were performed using an electrically braked bicycle ergometer, with the patient breathing through a low dead space, low resistance valve box. The valve box incorporates a turbine ventilometer on the inspired limb for the measurement of inspired minute ventilation (PK Morgan Ltd, Rainham, England). The expired limb is fed through a mixing chamber from which samples of expired air are analysed for the fractional concentrations of carbon dioxide and oxygen by an infra red spectrometer and paramagnetic analyser, respectively (PK Morgan, Ltd, Rainham, England). Throughout each test, minute ventilation (VE), oxygen consumption (VOe), and carbon dioxide production (VCO2) were measured by on line ventilation and expired gas analysis (PK Morgan Ltd, Rainham England) using standard equations (20).

Arterial blood gas values were monitored throughout exercise testing using a transcutaneous system (TCM3, Radiometer Ltd, Copenhagen) heated to 45~ Transcutaneous values of oxygen (PtcCO2) and carbon dioxide (PtcCO2) tension were continuously monitored during the exercise test. The TCM3 monitoring system incorporates an automatic correction equation, developed by Sigaard-Andersen (21), to correct the transcutaneous PCO 2 back to the value that would be obtained at 37~ Using this automatic correction procedure we have previously shown good agreement between PtcO 2 and PaCO 2 (22). In a group of 42 patients the regression of PtcCO 2 on PaCO 2 was: PtcCO 2 (corrected to 37~ (PaCO 2 - 0.05 (R=0"98, 95% CI 5: 2.4). In a previous study of 25 patients, in whom indwelling arterial cannulae were inserted, 140 measurements were obtained during incremental exercise testing following an in vivo calibration routine (23). Using the analysis of Bland and Altman (24), the mean difference between arterial PCO 2 and PtcCO 2 was 0'22 mmHg with limits of agreement of 2.2 and - 1-70 mmHg. The analysis for PaCO 2 and PtcO 2 showed a mean difference of 0.62 mmHg; the limits of agreement were 4.42 and - 3.38 mmHg. The good agreement between arterial and transcutaneous values of oxygen and carbon dioxide demonstrated in the previous validation studies allows the calculation of standard indices of gas exchange using transcutaneous and mixed expired gas concentrations. The indices calculated were the alveolar-arterial oxygen gradient (A-aO2) and dead space/tidal volume ratio (VD/VT) using standard equations (20). The patient was initially monitored for 2 min at rest whilst seated on the bicycle ergometer. The patient was then instructed to cycle with no additional load for 2 min; thereafter the work load was increased by increments of 25 W every 2 rain until symptom limited maximum. The maximal exercise values were compared to the normal values of Jones and Campbell (25). The non-invasive anaerobic threshold on exertion was calculated by the curve fitting method of Beaver et al. (26), using a plot of oxygen consumption against carbon dioxide production. STATISTICAL A N A L Y S I S

Data from normal volunteers and patients are presented as means :t= standard deviation. Data were compared using Student's t-test. A P value of less than 0"05 was considered significant. Results Individual data for symptom limited maximal exercise parameters are presented in Table 1. In patients

Microvascular angina

Table 1

733

Maximal (symptom limited) exercise responses Microvascular angina

Pat.

Age

VO 2

1 2 3 4 5 6 7 8 9 10 11 12

53 29 54 51 41 44 62 37 56 58 47 46

0.84 (49) 1.98 (72) 1.04 (52) 0-98 (52) 0.65 (26) 0-98 (57) 1.44 (65) 1'12 (61) 2.80 (93) 1.33 (66) 1-24 (64) 1.02 (59)

Controls VEmax

Pat.

Age

VO2

VEmax

30.9 57.7 39.9 43.2 20.9 41.5 57"5 50"9 102"3 51.3 43.7 43.8

1 2 3 4 5 6 7 8 9 10 11 12

53 32 59 44 39 40 45 48 54 64 56 48

1-44 (83) 1.94 (87) 1.53 (90) 1.41 (82) 1.93 (97) 1.41 (83) 1"55 (83) 2"09 (93) 1.79 (84) 1.39 (83) 1.88 (87) 1.71 (86)

39-9 47.1 42.7 47.9 65.1 50.7 53-9 59"9 56-9 47.2 60-0 50.7

Maximal exercise values in individual patients (V~, maximum minute ventilation, 1 min 1; VO2 maximal oxygen uptake, 1min ~ with percentage predicted). 70-

100 P < 0-003 80

e-1

,~

6050-

60 4O4O

30-

20

20-

I

I

control MA Mean maximal oxygen consumption, corrected for body weight (% predicted ml k g - t min-~) in controls (n--12) compared to patients with MA (n= 12).

Fig. 1

with M A physical work capacity was reduced compared to normal controls. Exercise was limited in all patients by a combination of pain and breathlessness. Maximal (symptom limited) oxygen consumption, corrected for body weight (% predicted ml kg 1 and min 1), was reduced in patients with M A compared to controls (60"73 • 16"51 vs. 87.21 • 5.2; P<0.003) (Fig. 1). Anaerobic threshold was calculated by curve fitting a plot of oxygen consumption against carbon dioxide production, expressed as a percentage of m a x i m u m predicted VO2 for age and sex. This was reduced in patients with M A compared to controls, but was still within accepted normal limits (41.55 • 5.82 vs. 49-73 • 7.3; P<0.02) (Fig. 2). Resting ventilation was higher in the M A patients

P < 0.02

T T

100-

control MA Mean anaerobic threshold expressed as percentage of predicted maximum VO21 min-1 in controls (n=12) compared to patients with MA (n= 12).

Fig. 2

compared to controls, although this did not reach statistical significance (9.68 1 min 1 -4- 2'09 vs. 9.141 rain ~ • 1.51). The slope of the ventilatory response on exertion (VEIVC02 1- 1 C O 2 output) was increased in the M A patients compared to controls (35.9 • 8.01 vs. 27"5 • 3'08; P<0"02) (Fig. 3). The increased ventilatory response was not associated with a significant fall in the transcutaneous PCO2 in the M A patients (at rest PtcCO2 36'2 -4- 2.7; at maxim u m exercise PtcCO2 37"8 • 3"8) suggesting that it was not due to hyperventilation (Fig. 4). In the control group there was a similar increase in P t c C O 2 from 37-8 • 2.9 at rest to 3 9 . 4 • at m a x i m u m exercise. The increased ventilatory response in the M A group was, however, associated with a raised

734

A. T w e d d e l et al.

exercise (12-4 -4-3" 1 vs. 11.1 4- 5.0 mmHg) between the MA and control patients, respectively.

60 P < 0.02

50

T

40 30 ~J

v

20 10 0

control MA Ventilatory response on exertion (VE/VC02; 1 1CO2 output) in controls (n= 12) compared to patients with MA (n= 12).

Fig. 3

degree of 'wasted ventilation' compared to controls at rest ( V D / V T 0"40 4- 0"03 vs. 0"35 -4-0"01; P<0"003) which persisted on exertion (at maximum exercise; VD/VT microvascular 0.33 -4-0-05 vs. control 0.22 4- 0.05; P<0'002) (Fig. 5). There was no significant difference between the A-aO2 gradient at rest (11"7 4- 3-4 vs. 11'0 + 3"4 mmHg) or at maximum

Discussion

Patients with exertional chest pain, breathlessness and normal coronary arteriograms pose a difficult problem for the clinician, both from the point of view of diagnosis and management. There is little consensus as to how to identify patients with MA, or indeed whether such patients exist (27,28). It has previously been suggested that all the symptoms in such patients can be explained on the basis of hyperventilation (29,30). However we found no evidence of inappropriate ventilation in our patients. This data is supported by a previous study from Lewis et al. (31), who, using intra-arterial blood gas sampling, similarly found no inappropriate ventilation and only a small reduction in arterial P C O 2 a t maximal exercise, compensating for the onset of anaerobic metabolism. Again in keeping with the findings of Lewis et al. (31), the ventilatory cost of excretion of CO 2 was increased, which is characteristic of patients with heart failure (32). This increased ventilatory response, with no accompanying drop in PCO2, is consistent with ventilation/perfusion inequality. The normal alveolar-arterial oxygen difference, however,

50:

40

3O hi0

2O

10

Fig. 4

(n= 12).

Rest Exercise Transcutaneous PC02 at rest and at maximum exercise in MA patients

Microvascular angina

0.5P < 0.003 0,4

P < 0.002

D

0.3

0.2

0.1

0.0

N

I

I

Rest Exercise Fig. 5 Mean physiological dead space (VD/VT)a t rest and maximum exercise in controls ([~, n= 12) and patients with MA ([], n= 12).

suggests that most of the inequality is caused by an increase in lung units with abnormally high ventilation/perfusion ratios suggesting that areas of the lung are being underperfused compared to their ventilation. This increase in the physiological dead space is reflected in the raised VD/VT in the MA patients. A similar finding of an increased cost of carbon dioxide excretion due to a change in the distribution of ventilation perfusion ratios producing an increased physiological dead space or 'wasted ventilation' but with a normal A-aO 2 gradient has been shown in patients with chronic heart failure (33,34). The reduced cardiac output in these patients contributes to their dyspnoea by limiting perfusion to the lungs as well as exercising muscles. A reduced aerobic capacity on exercise testing and ventilation/ perfusion inequality may be due to cardiopulmonary deconditioning. That this would not appear to be a major factor in the MA patients is the finding of an anaerobic threshold which although lower than that in the control group, was within accepted normal limits during the exercise test (25), an anaerobic threshold below normal being considered to be an indication of cardiopulmonary deconditioning. Resting measurements of lung function were normal in all patients and there was no evidence of loss of lung volume. No patient gave a history of pulmonary disease and there was no reason to suspect previous thromboembolic disease. It might, therefore, be hypothesized that the pathophysiological mechanisms underlying this 'wasted ventilation' are similar to those occurring in the myocardium, namely a functional abnormality of small vessels. This may

735

produce underperfusion of areas of the lung leading to the increase in lung units with a high ventilation/ perfusion ratio, and hence, the raised VDIVT. It has been suggested that the measurement of end tidal partial pressure of CO 2 is of particular value in this group of patients (35). In the study by Lewis et al. (31), however, PCO 2 measured by direct arterial sampling correlated with end-tidal CO 2 at rest. This relationship was lost on exercise consistent with the increase in physiological dead space. Lewis concluded that end-tidal CO 2 cannot be used as an indication of arterial PCO 2 in this group of patients. The transcutaneous monitoring system, however, does correlate with arterial values at rest and on exertion and has the advantage that it is still noninvasive. In summary, using an entirely non-invasive technique we have found that in patients with MA, there is a significant degree of 'wasted ventilation', the increased cost of carbon dioxide excretion in the MA patients contributing to the sensation of breathlessness. This effect has been described in other patient groups with reduced cardiac reserve. This may well indicate involvement of the pulmonary circulation in a similar pathophysiological process to that which affects the myocardium. Non invasive progressive exercise testing, with transcutaneous measurement of arterial blood gases would appear to be helpful in the identification of this group of patients with MA. References

1. Kemp HG, Kronmal RA, Vietstona RE, Frye RL and the Coronery Artery Surgery Study (CASS) participants. Seven year survival of patients with normal or near normal coronary arteriograms. A CASS registry study. J Am Coll Cardiol 1986; 7: 479483. 2. Dart AM, Davies HA, Dalal J, Ruttley M, Henderson AH. 'Angina' and normal coronery arteriograms: A follow-up study. Eur Heart J 1980; 1:97 100. 3. Proudfit WL, Shirley EK, Sones FM. Selective cine coronary arteriography. Correlation with clinical findings in 1,000 patients. Circulation 1966; 33: 901410. 4. Cannon RO, Epstein SE. Microvascular angina as a cause of chest pain with angiographically normal coronery arteries. Am J Cardiol 1988; 61: 1338-1343. 5. Opherk D, Zebe H, Weihe E et al. Reduced coronary dilatory capacity and ultra-structural changes of the myocardium in patients with angina pectoris but normal coronary arteriograms. Circulation 1987; 63: 817-825. 6. Bass C, Wade C, Gardener WN et al. Unexplained breathlessness and psychiatric morbidity in patients with normal and abnormal coronary arteries. Lancet 1983; 19:605 609. 7. Freeman LJ, Nixon PGF. Chest pain and the hyperventilation syndrome: Some aetiological considerations. Postgrad Med J 1985; 63:957 961.

736

A. Tweddel et al.

8. Rasmussen K, Bagger JP, Bottzauw J, Henningsen P. Prevalence of vasospatic ischaemia induced by the cold pressor test or hyperventilation in patients with severe angina. Eur Heart J 1984; 5: 354-361. 9. Bass C, Chambers JB, Kiff P e t al. Panic anxiety and hyperventilation in patients with chest pain: a controlled study. Q J Med 1988; 260: 949-959. 10. Chauhan A, Mullins PA, Taylor G, Petch MC, Schofield PM. Effect of hyperventilation and mental stress on coronary blood flow in Syndrome X. Br Heart J 1993; 69:516 524. 11. Epstein SE, Cannon RD. Site of increased resistance to coronary flow in patients with angina pectoris and normal epicardial coronary arteries. JACC 1986; 8: 459-461. 12. Tweddel AC, Davies MJ, Martin W, Hutton I. Small vessels in myocardial biopsies in microvascular angina. Eur Heart J 1991; 12: 709. 13. Richardson PJ, Livesley B, Oram S. Angina pectoris with normal coronary arteries. Transvenous myocardial biopsy in diagnosis. Lancet 1974; 2: 677-680. 14. Sax FL, Cannon RD, Hanson C, Epstein SE. Impaired forearm vasodilator reserve in patients with microvascular angina. Evidence of a generalised disorder of vascular function? N Engl J Med 1987; 317: 1366-1370. 15. Cannon RD, Cattau EI, Yakshe PN et al. Coronary flow reserve, esophageal motility and chest pain in patients with angiographically normal coronary arteries. Am J Med 1990; 88: 217-222. 16. Cannon RD, Peden DB, Berkebite C, Schenke WH, Kaliner MA, Epstein SE. Airways hyper-responsiveness in patients with microvascular angina: Evidence for a diffuse disorder of smooth muscle responsiveness. Circulation 1990; 82:2011-2017. 17. Cannon RD, Camici PG, Epstein SE. Pathophysiological dilemma of Syndrome X. Circulation 1992; 85: 883 892. 18. Tweddel AC, Martin W, Hutton I. Thallium scans in Syndrome X. Br Heart J 1992; 68: 48-50. 19. Tweddel AC, Martin W, Hutton I. Do patients with microvascular angina have ischaemia? JACC 1992; 19: 79~791. 20. Cotes JE. Lung Function Fourth edn. Oxford: Blackwell, 1978. 21. Siggard-Andersen O. The Acid-Base Status of Blood. Copenhagen: Munksgaard, 1976; p. 89.

22. Carter R. The measurement of transcutaneous oxygen and carbon dioxide tensions during exercise testing. Breath 1989; 38: 2-6. 23. Sridhar MK, Carter R, Moran F, Banham SW. Use of a combined oxygen and carbon dioxide transcutaneous electrode in the estimation of gas exchange during exercise. Thorax 1993; 48:643 647. 24. Bland M J, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; i: 306-310. 25. Jones NL, Campbell EJM. Clinical Exercise Testing. Second edn. Toronto: Saunders, 1982. 26. Beaver WL, Wasserman K, Whipps BJA. New method for detecting anaerobic threshold by gas exchange. J Appl Physiol 1986; 60: 2020~027. 27. Cannon RD, Camici PG, Epstein SE. Pathophysiological dilemma of Syndrome X. Circulation 1992; 85: 883-892. 28. Hutchison SJ, Poole-Wilson PA, Henderson AH. Angina with normal coronary arteries: A review. J Med 1988; 72: 677-688. 29. Evans DW, Lum LC. Hyperventilation: An important cause of pseudoangina. Lancet 1977; 22: 155-157. 30. Nixon PGF. Hyperventilation and cardiac symptoms. Int Med 1989; 10:67 84. 31. Lewis NP, Hutchison SJ, Willis N, Henderson AH. Syndrome X and hyperventilation. Br Heart J 1991; 65: 94-96. 32. Frank LI, Wilson JL, Ferraro N. Exercise ventilation and pulmonary artery wedge pressure in chronic stable congestive heart failure. Am J Cardiol 1986; 57: 249-253. 33. Rubin SA, Brown HV. Ventilation and gas exchange during exercise in severe chronic heart failure. Am Rev Respir Dis 1984; 129 (Suppl.): 563-564. 34. Lewis NP, MacDougall IC, Willis N, Henderson AH. The ventilatory cost of exercise compared in chronic heart failure and chronic renal anaemia. Q J Med 1992; 83:523 531. 35. Chambers JC, Kift PJ, Gardener WN. Value of measuring end tidal partial pressure of carbon dioxide as an adjunct to treadmill exercise testing. B M J 1988; 296: 1281-1285.