Iemura et al's hypothesis that endothelial dysfunction is associated with coronary aneurysms which have regressed after Kawasaki disease (KD) is interesting[1]; however, there are two issues with respect to their study design that need to be considered, in addition to the issues reported before.[2, 3]

(1) Is endothelial dysfunction associated with regressed coronary aneurysms per se late after KD?

Iemura et al showed that acetylcholine-induced vasodilatation is impaired in regressed coronary aneurysms (groups 1 and 2) but there is preserved vasodilatation in patients without coronary involvement during the acute illness of KD (group 3). This could be explained either by the direct effect of the regressed aneurysm or the difference among the study groups (groups 1 and 2 versus group 3). The Kawasaki vasculitis was severe enough to cause coronary aneurysms in some segments during the acute illness.

Although in Fig 1 Iemura et al showed that there is acetylcholine-induced vasoconstriction in the regressed aneurysm, readers might note that the rest of the segments in the right coronary artery without coronary lesions from the onset, including segments remote from the regressed aneurysms, also seem to constrict in response to acetylcholine. This might suggest that the regressed aneurysm per se does not impair endothelial function after KD.

Iemura et al did not present any data analysing the acetylcholine-elicited responses in normal segments from disease onset in groups 1 and 2; therefore the alternative hypothesis remains possible i.e. that coronary segments, regardless of their association with regressed aneurysms, constrict in response to acetylcholine in patients with a history of severe Kawasaki vasculitis. Consistent with this hypothesis, previous reports concluded that endothelial function is impaired in unaffected coronary segments and brachial arteries (usually uninvolved with aneurysms during the acute illness) in patients with severe acute Kawasaki vasculitis in the long term follow up period:

1.5-12.5 years (mean (SE) 6.5 (1.1)) after acute KD in Mitani et al[2]
5.3-17.1 years (median 11.3) after KD in Dhillon et al[4]
16-28 years old (mean 22) at the time of examination in Kamiya et al[5]
1-14 years old (median: 3.5) at the time of examination in Sano et al[6]

(2) Is endothelial function entirely preserved in KD patients without coronary involvement during the acute illness (group 3)?

This is an issue that we did not specifically address in our previous study.[2] Iemura et al showed that coronary arteries normally dilate in response to acetylcholine in patients without coronary lesions during the acute illness (group 3) on the basis of analysing multiple coronary segments.[1] However, in KD patients some coronary segments (ie, proximal segments) are characteristically more susceptible to severe abnormalities than others in terms of morphological, angiographic findings and vasomotor function.[2, 7, 8] Abnormal vasomotor function might be unmasked by specifically focusing on some "susceptible" segments in group 3, as we did in uninvolved proximal coronary segments in patients with coronary aneurysms in other segments.[2] An alternative hypothesis that some susceptible segments are associated with impaired acetylcholine-induced vasodilatation even in group 3 remains to be tested.

Iemura et al's conclusion might not be consistent with a previous report showing that endothelial function is impaired in KD patients without aneurysms in any segment during the acute illness.[4] This inconsistency remains to be addressed.

References

1. Iemura M, Ishii M, Sugimura T, et al. Long term consequences of regressed coronary aneurysms after Kawasaki disease: vascular wall morphology and function. Heart 2000;83:307-11.

2. Mitani Y, Okuda Y, Shimpo H, et al. Impaired endothelial function in epicardial coronary arteries after Kawasaki disease. Circulation 1997;96:454-61.

3. Mitani Y. Coronary endothelial dysfunction after Kawasaki disease. J Am Coll Cardiol 2000;35:821-3

4. Dhillon R, Clarkson P, Donald AE, et al. Endothelial dysfunction late after Kawasaki disease. Circulation 1996;94:2103-6.

5. Kamiya Y, Onouchi Z, Hamaoka K. Arteriosclerosis long after Kawasaki disease: endothelial function. Acta Cardiol Paediat Jpn 1997;13:252 (Japanese).

6. Sano T, Tagawa T, Itagaki Y, et al. Endothelial dysfunction after Kawasaki vasculitis. Jpn Circ J 1999;63:338 (Japanese).

7. Naoe S, Takahashi K, Masuda H, et al. Kawasaki disease: with the particular emphasis on arterial lesions. Acta Pathol Jpn 1991;41:785-97.

8. Kato H, Sugimura T, Akagi T, et al. Long-term consequences of Kawasaki disease: a 10-21 year follow-up study of 594 patients. Circulation 1996;94:1379-85

Dear Editor,

We share many of the reservations that Morley-Davies and Byrne have regarding the North American Vasovagal Pacemaker Study (VPS).[1] However, this study should be considered in the context of the other available data supporting the role of pacing in selected patients. We also are interested in the study by Di Girolamo et al[2] suggesting a benefit for orthostatic training in those with neurocardiogenic (vasovagal) syncope. However, they do not classify the tilt test induced haemodynamic responses seen in their population. In those with vasovagal syncope the different haemodynamic collapse patterns seen on tilt testing[3] may have a very useful role in identifying not only those who may benefit from pacing but also those in whom pacing may be inappropriate even as a 'last resort'. The consensus is that pacing is of the greatest benefit in those with a predominantly bradycardiac (cardioinhibitory) response on tilt testing.

Furthermore, the population in Di Girolamo et al's study[2] is not comparable to that in VPS and most other studies of pacing in vasovagal syncope. The former's population were all adolescents (mean age 16 years), whereas much older patients were recruited into VPS (mean age 43 years). Our experience suggests that many adolescents will over time 'grow out of' their fainting episodes. Hence, in the majority of cases our advice to this group is simply reassurance.

Since the beginning of 1997 we have run a prospective register of the all tilt tests that we have undertaken including the management suggested. Up until December 1999 over 600 patients had been tilted. Of these 186 were under 35 years of age. Only 5 (2.7%) of these young patients have been paced. In each cases multiple non-pacing strategies (though not orthostatic training) were attempted prior to offering pacing.

In selected patients with vasovagal syncope pacing offers excellent symptomatic relief, enabling a return to normal life.

Dr A S Kurbaan
Department of Cardiology, London Chest Hospital,
London E2 9SX, UK

Dr R Sutton
Department of Cardiology, Royal Brompton Hospital,
London SW3 6NP, UK

References

1. Connolly SJ, Sheldon R, Roberts RS, Gent M on behalf of the Vasovagal Pacemaker Study investigators. The North American Vasovagal Pacemaker Study (VPS). A randomised trial of permanent cardiac pacing for the prevention of vasovagal syncope. J Am Coll Cardiol 1999;33:16-20.

2. Di Girolamo E, Di Iorio C, Leonzio L, Sabatini P, Barsotti A. Usefulness of a tilt training program for the prevention of refractory neurocardiogenic syncope in adolescents: a controlled study. Circulation 1999;100:1798-1801

3. Kurbaan AS, Franzén A-C, Bowker TJ, Williams TR, Kaddoura S, Petersen MEV, Sutton R. Usefulness of tilt test induced patterns of heart rate and blood pressure using a two stage protocol with glyceryl trinitrate provocation in patients with syncope of unknown origin. Am J Cardiol 1999;84:695-670.

We read the case report by Elliott et al (1) with great interest and would like to add to their findings. In three of five consecutive patients undergoing percutaneous septal ablation (PTSMA) for hypertrophic cardiomyopathy (HCM) they observed left ventricular cavity opacification during selective injection of the first septal perforator with the contrast agent Optison.

We treated a 61 year old man with a long history of HCM and limiting breathlessness depsite maximal medical therapy with PTSMA. His peak resting intraventricular gradient, measured using a double -lumen pitail catheter, was 134 mmHg with a post-ectopic gradient of 165 mmHg. With a 2.0 by 20 mm balloon inflated, initial injection of the echocardiographic contrast agent Levovist (Schering AG, Germany) selectively into the first septal perforator showed opacification of the basal septum as demonstrated on transoesophageal echocardiography. Using appropriate technical considerations 2.5 ml of alcohol was then injected into this artery. The ballon was left inflated in the septal perforator for 15 minutes and subsequently deflated. Following this the resting gradient was negligible but a peak provokable gradient of 105 mmHg was obtained. In view of this further ablation was planned and the adjacent second septal perforator was selectively instrumented. Levovist injection into this artery however resulted in echocardiographic enhancement of the mid septum distal to the site of flow acceleration.It was therefore decided not to proceed further with this artery and we returned to the original septal perforator. Repeat echocardiographic contrast injection through a shorter balloon (2.0 by 15mm) produced clear opacification of the left ventricular cavity.

In view of the safety considerations raised by Elliott el al (1)regarding the sytemic effects of alcohol we elected not to proceed further with the procedure. This case adds to the findings documented previously (1). It demonstrates that this phenomenon occurs with both Optison and Levovist contrast agents. The mechanism however is not clear although a number of possibilities are raised by Elliott et al (1). Interestingly in our patient left ventricular opacification occurred only after the first alcohol injection whereas in those cases discussed previously opacification occurred prior to alcohol injection. This could relate to our subsequent choice of a shorter balloon (the same size as used in the cases described by Elliott et al (1)) which may increase the number of septal sub-branches exposed to the contrast agent. Alternatively it may relate to the toxic effects of alcohol on the vasculature.

Mechanisms aside, our findings and those documented previously suggest that caution may be appropriate for the injection of alcohol where this phenomenon is observed.

1 Elliott PM, Brecker SJ, McKenna WJ. Left ventricular opacification during selective intracoronary injection of echocardiographic contrast in patients with hypertrophic cardiomyopathy. Heart 2000;83:e7

Dear Editor:

We read with great interest the review "Cardiocutaneous fistula" published by Dr. Danias and co-workers. At our Institution, we encountered a similar case, concerning a 72-year-old female, who developed this complication eleven months after coronary bypass sugery and linear resection of an anteroapical left ventricular aneurysm; the postoperative course had been complicated by mediastinitis necessitating surgical debridement and irrigation with 0.5% povidone-iodine until removal of chest tubes. Initially, the patient presented mimicking subacute endocarditis and subsequently developed the fistulous tract from the cardiac apex through the chest wall. As stated by others [1] [2] [3] there was no tendency to spontaneous healing, and bleeding was only mild and transient, resembling chronic fistula secondary to chronic sternal osteomyelitis; cultures were negative. Elective operation was not planned because of poor general state and relative paucity of symptoms. Conditions gradually declined until, during hospitalization, massive hemorrhage ensued; no hemodynamic instability was observed during compressive maneuvers. Emergency operation was accomplished through re-sternotomy after institution of moderately hypothermic (28 °C) cardiopulmonary bypass through the femoral vessels. Cooling was continued during dissection of extremely dense adhesions and fragile tissues to obtain complete exposure the left ventricle, avoiding free dissection of the great vessels; the defect was repaired during circulatory arrest (20 °C) removing all foreign material and infected tissue along with two chronically eroded costal segments. The patient died on the fourth postoperative day in multi-organ failure.

It is difficult to give guidelines, but some points should be stressed:

High-risk surgery is probably the only therapeutic option and should be planned electively avoiding life-saving emergency operations.

In our case, infection was not secondary to chest wall erosion due to prosthetic material (a theoretical possibility previously stated by others [3]). Mediastinitis may represent the inital cause, but also a consequence of implantation of contaminated prosthetic material during aneurysmectomy.

The role of extensive use of prosthetic material (in particular, Teflon felt) outside the heart is unclear, but may be important. This specifically applies to patients undergoing left ventricular aneurysm repair, who often present additional risk factors for infection (advanced age, associated heart failure, longer cardiopulmonary bypass times, etc).

In case of postoperative mediastinitis should extracardiac Teflon be replaced with autologous pericardium? This is a very difficult issue, due to the different extensiveness of surgical options, but may be important in the development of fistulae/pseudaneurysms of the heart/great vessels. In case of extensive use of prosthetic material, removal should probably be considered.

Re-sternotomy [1] may not be the optimal approach for repair. If operation involves the left ventricle alone, an anterior thoracotomy through the 5th or 6th intercostal space [4] is more appealing and less cumbersome, despite the necessity of extrathoracic cannulation, deep hypothermia and low-flow bypass/circulatory arrest. Sternotomy may prolong surgery rather than represent a more expeditious technique. We reccomend to avoid re-sternotomy whenever possible, especially in case of previous mediastinitis.

Marco Pocar, MD - Francesco Donatelli, MD - Adalberto Grossi, MD
Istituto di Malattie dell'apparato cardiovascolare e respiratorio - Università degli Studi di Milano
Divisione di Cardiochirurgia - IRCCS, Ospedale Maggiore Policlinico - Via Francesco Sforza, 35 - 20122 Milano, Italy

References

1 McHenry MC, Longworth DL, Rehm SJ, et al. Infection of the cardiac suture line after left ventricular surgery. Am J Med 1988;85:292-300.

2 Stanford W, Reuben CF, Flemma RJ, et al. Management of long-standing cardiocutaneous fistulas after resection of left ventricular aneurysms. J Thorac Cardiovasc Surg 1980;79:789-792.

3 Deuvaert FE, Wellens F, De Paepe J, at al. Cardiocutaneous fistula after left ventricular anuerysm repair. Case report and review of the literature. J Cardiovasc Surg (Torino) 1984;25:560-562.

4 Wellens F, Vanermen H. Treatment of the infected cardiac suture line. J Cardiac Surg 1988;3:109-118.