HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Personal Folders Download to citation manager Author home page(s): Gösta Pettersson Frederic S. Joyce Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Pettersson, G. Articles by Joyce, F. S. Search for Related Content PubMed PubMed Citation Articles by Pettersson, G. Articles by Joyce, F. S.

Eur J Cardiothorac Surg 1998;13:678-684
© 1998 Elsevier Science NL

Treatment of aortic valve endocarditis with the Ross operation¹

^a Department of Cardiothoracic Surgery, Copenhagen University Hospital-Rigshospitalet, Copenhagen, Denmark
^b Division of Cardiothoracic Surgery, Albany Medical College, Albany, NY, USA

Received 29 September 1997; received in revised form 11 February 1998; accepted 16 February 1998.

Corresponding author. Department of Cardiothoracic Surgery, The Heart Center, Copenhagen University Hospital-Rigshospitalet, 2100 Copenhagen , Denmark. Tel.: +45 35452100; fax +45 35452548.

	Abstract

Top Abstract Introduction Material and methods Surgical findings and pathology... Results Discussion Appendix A. Conference... References

 
Objective: Standard treatment of patients with infective endocarditis is radical debridement and valve replacement, in cases with advanced pathology the treatment is usually root replacement with either a composite graft or a homograft. Enthusiasm for the use of the Ross operation in non-infective aortic valve disease is increasing, but use of the pulmonary autograft in the treatment of aortic valve endocarditis has been limited. The objective of this prospective study is to present the technique and results of our experience with aortic valve endocarditis treated with the Ross operation. Materials and methods: Since 1992 we have treated 35 patients (median age 41 years, range 6–71 years) having aortic valve endocarditis with a Ross operation. Twenty-four patients had advanced disease defined as pathology due to endocarditis extending beyond the valve cusps (13 patients) or prosthetic valve endocarditis (11 patients). Twenty-two patients had active disease at the time of surgery, and 12 had undergone one to four previous heart operations. Results: There were two operative deaths (5.8%), both related to severe disease with very advanced pathology and heart failure. Intraoperative echocardiography demonstrated no or trivial autograft insufficiency in all patients. There have been no late deaths. There has been one (probable) recurrent right-sided endocarditis in a drug addict during a follow-up period of 3–56 months. One patient has been reoperated on for homograft stenosis. Conclusions: We are enthusiastic about the use of the Ross operation in aortic valve endocarditis and in younger patients with advanced pathology, it is our preferred treatment modality. Following removal of the autograft, unparalleled exposure of the left ventricular outflow tract is obtained. Even in patients with very advanced pathology the left ventricular outflow tract is usually intact, allowing autograft implantation in the standard fashion. For selected patients with simple endocarditis, the Ross operation is an attractive option on its usual merits.

Key Words: Aortic valve • Prosthetic valve • Endocarditis • Pathology • Pulmonary autograft • Ross operation

	Introduction

Top Abstract Introduction Material and methods Surgical findings and pathology... Results Discussion Appendix A. Conference... References

 
The original main indications for a Ross operation were aortic valve replacement in children and young women/potential mothers because of the autograft's ability to grow, expected long-term durability and freedom from anticoagulation [1]. With increasing experience and development of the surgical technique, autograft implantation as free standing root [2], the indications and the applications of the operation have been expanded [3].

Standard treatment of patients with infective endocarditis is radical debridement and valve replacement, in cases with advanced pathology the treatment is usually root replacement with either a composite graft [4] or a homograft [5] [6] [7] [8] [9] [10]. Many patients with endocarditis are young, have advanced pathology and are unable or unwilling to accept life long anticoagulation. Homograft aortic root replacement, although an excellent immediate solution because of its tissue quality and resistance to reinfection, will require repeated difficult reoperations for homograft replacement in the future because of its inevitable degeneration.

Use of the Ross operation in patients with active endocarditis was pioneered in Austin, TX in 1990 [11] and in Copenhagen in 1992 [12] [13]. In addition to the qualities of the homograft, the autograft was expected to provide growth potential and better long-term durability. It was presumed that the living autograft valve would be resistant to persistent and recurrent infection and that this resistance would not change over time. The difficulty procuring an adequate supply of aortic homografts and the relative abundance of pulmonary homografts were additional arguments in favor of the Ross operation.

Our early experience with the Ross operation in patients with endocarditis lived up to our expectations and disclosed new qualities of the operation and the autograft which we found particularly valuable in patients with advanced pathology. The objective of this prospective study is to present the technique and results of our experience with 35 cases with aortic valve endocarditis treated with the Ross operation since 1992.

	Material and methods

Top Abstract Introduction Material and methods Surgical findings and pathology... Results Discussion Appendix A. Conference... References

 
Patients
Between December 1992 and June 1997, 35 patients, 29 males and 6 females, with a median age of 41 years (6–71 years), underwent a Ross operation for ongoing native or prosthetic aortic valve endocarditis or complications related to previous aortic valve endocarditis. Three patients were intravenous drug abusers. Eleven patients had undergone one to four previous heart operations (seven patients had one operation, two patients had two operations, one patient had three operations and one patient had four operations). One patient had a pacemaker because of a previous heart block. Eight patients were in New York Heart Association functional class IV, 13 were in class III, seven were in class II and seven were in class I at the time of surgery. Two patients had end-stage renal disease.

The diagnosis was based on clinical findings, echocardiography, blood cultures, operative findings, and specimen microscopy and culture. The infection was classified as active at the time of surgery in 22 patients and remote (previous, healed) in 13 patients. Eleven patients had simple endocarditis (infectious pathology limited to the valve cusps/leaflets) and 24 patients had advanced disease (infectious pathology extending beyond the valve cusps/leaflets in 13 patients; prosthetic valve endocarditis in 11 patients).

Fourteen patients required one or more concomitant procedures (five patients required mitral valve repair, two patients required tricuspid valve repair, two patients required coronary artery bypass grafting, one patient required VSD closure, four patients required fistula repair, two patients required ascending aorta replacement).

Methods
Surgical principles
Our surgical principles for the treatment of endocarditis are based on the belief that complete debridement of infected and necrotic tissue is a prerequisite to success. The distinction between simple and advanced endocarditis is considered important, and usually this distinction is made preoperatively by transesophageal echocardiography (TEE). This enables the surgeon to be better prepared and to inform the patient of his options (and risk) before surgery. In simple endocarditis, cusp excision accomplishes the goal of complete debridement, and the choice of treatment modality is based on the usual patient – and surgeon-related considerations in these cases, from insertion of a standard valve prosthesis to a Ross operation. Advanced endocarditis requires more extensive debridement and more difficult reconstructive efforts. In these cases, aortic root replacement is required. Dismantling of the aortic root as well as division and removal of the pulmonary artery facilitates complete and safe debridement. Our preferred conduit for reconstruction of the aortic root in the setting of advanced endocarditis is the autograft followed by the aortic homograft and lastly the composite graft. A composite graft is mainly considered in cases of remote disease.

Intraoperative echocardiography is performed routinely both before and after cardiopulmonary bypass. The preoperative echocardiogram is used to evaluate the aortic pathology, left ventricular function and size, pulmonary valve and other pathology. The postoperative echocardiogram is used to evaluate autograft function, the success of concomitant procedures, ventricular function and thoroughness of air evacuation.

Myocardial protection
Myocardial protection is achieved with intermittent cold antegrade and retrograde crystalloid cardioplegia supplemented with ice slush in the pericardium.

Surgical technique
All operations are performed through a median sternotomy, on cardiopulmonary bypass with moderate systemic hypothermia. Two-stage atrial and aortic arch cannulation is standard, but for special indications femoral and/or subclavian artery and bicaval cannulation are used. Before starting cardiopulmonary bypass, the external diameter of the pulmonary artery sinotubular junction is measured and correlated with the echocardiographic findings. The aorta is opened with a transverse incision about 1 cm above the right coronary artery. The pathology is inspected and the valve leaflets are excised. Debridement is started and continued until the surgeon can evaluate the extent of the pathology. The aortic annulus is sized and if the diameters of the pulmonary artery sino-tubular junction and the aortic annulus are compatible and the infectious pathology does not include severe proximal destruction and tissue loss in the left ventricular outflow tract (LVOT) or involvement of the pulmonary artery or valve, a Ross operation is considered feasible/possible. Localizing the central fibrous body and the lateral trigone is important in regard to this judgment. In cases of simple endocarditis, the aortic root is not damaged until a satisfactory autograft has been harvested. In cases of advanced endocarditis, it is essentially predetermined that some form of root replacement is necessary, and we may proceed immediately with the dismantling of the infected aortic root. The aortic root is opened widely by excising the coronary arteries and making a vertical incision in the non-coronary sinus down to the level of the annulus. This combined with autograft removal provides excellent exposure of the base of the heart and ventricular outflow tracts. Thorough and complete debridement of all infected and non-vital tissue (or prosthetic material from previous valve surgery) is done. Cavities and fistulas are opened and adjacent tissue is removed as necessary. VSDs, fistulas and mitral or tricuspid valve lesions are repaired. Additional exposure can easily be obtained by various transatrial techniques. Autologous pericardium is the material of choice for repairs.

Even in cases of very advanced endocarditis with total aorto-ventricular separation, large pseudoaneurysms and fistulas, the proximal left ventricular outflow tract is usually intact and autograft insertion can be performed in standard fashion to healthy tissue at the annular or subannular level proximal to the debrided areas. In this way these areas are exteriorized from the circulation and endothelial integrity of the left ventricular outflow tract can be reestablished. All anastomoses are performed with running monofilament polypropylene suture. The right ventricular outflow tract is reconstructed with a pulmonary homograft. The right ventricular to homograft anastomosis has in some cases been performed after removal of the aortic crossclamp to allow perfection of the hemostasis in the autograft dissection bed. The standard autograft implantation technique has been described in detail previously [14].

Antibiotics
Our cardiologists' initial routine management of patients suspected to have endocarditis is to start antibiotic treatment immediately after adequate blood cultures have been secured and to adjust the antibiotic regimen to the sensitivity pattern of grown microorganisms as soon as possible. Surgery, when indicated, is not postponed to allow any specific period of preoperative treatment, but all patients in this series had been treated for at least 1 week before surgery. Patients with active, culture-positive endocarditis are routinely treated for a total of 6 weeks with microorganism-specific intravenous antibiotics. Patients with remote endocarditis are given standard prophylactic antibiotics consisting of 1.5 g cefuroxim and 80 mg gentamycin every 8 h for 48 h. The ultimate length of antibiotic treatment is determined by the surgical findings, operative culture and microscopy results, clinical response and patient compliance.

Follow-up
The patients were seen in the clinic and examined by echocardigraphy after 1 and 12 months and then once a year.

	Surgical findings and pathology

Top Abstract Introduction Material and methods Surgical findings and pathology... Results Discussion Appendix A. Conference... References

 
The pathology found at surgery is depicted in Table 1.

View larger version (93K): [in this window] [in a new window]   Fig. 1. (a) Horseshoe pseudoaneurysm (marked with arrows) with extension to the left under the pulmonary artery perfectly exposed following removal of the pulmonary artery (autograft). After debridement, the LVOT appeared intact and of normal size. (b) Standard autograft implantation with running 4–0 monofilament suture exteriorizing the pathology. AG, autograft; CFB, central fibrous body; LA, left atrium; LCA, left coronary artery; LT, lateral trigone; MV, mitral valve; RA, right atrium; RCA, right coronary artery; RVOT, right ventricular outflow tract.

View larger version (88K): [in this window] [in a new window]   Fig. 2. (a) Advanced acute endocarditis with invasion of the aortic annulus, the aortic wall and the wall of the right atrium causing extensive tissue necrosis and fistulation to the right atrium. The infectious process is invading the triangle of Koch and the tricuspid valve annulus. The extension of the infection is marked with arrows. (b) The appearance following radical debridement including resection of part of the right atrial wall. The right coronary artery is visible in the wall of the debrided area (*). A pericardial patch was used to reconstruct the atrial wall. Notice the intact normal sized LVOT. CFB, central fibrous body; LT, lateral trigone; LVOT, left ventricular outflow tract; MV, mitral valve; PP, pericardial patch; RA, right atrium; RCA, right coronary artery.

Affection of the pulmonary artery or valve precluding a Ross operation has been observed in three patients during the study period, in two cases as involvement of the pulmonary artery wall and in the third as pulmonary valve endocarditis, in which case the pulmonary valve was replaced with a homograft. The aortic valves were replaced with an aortic homograft in these three cases.

	Results

Top Abstract Introduction Material and methods Surgical findings and pathology... Results Discussion Appendix A. Conference... References

 
Median cardiopulmonary bypass and cross clamp times were 183 min (range 87–302 min) and 142 min (range 42–224 min), respectively.

Post-bypass echocardiography revealed no or trace (<1+) autograft insufficiency in 22 patients and mild (1+) insufficiency in 13 patients.

There were two operative deaths (5.8%). The first operative death was due to post-bypass heart failure in a 45-year-old man who weighed 240 kg and had sepsis, heart failure, severe aortic insufficiency and fistulas to the right ventricle and left atrium. The second operative death occurred in a 50-year-old man with advanced endocarditis with extensive invasion and proximal destruction of the myocardium as mentioned above. Reconstruction required pericardial repair of the LVOT and he succumbed to bleeding after rupture of the LVOT–pericardial patch suture line.

All patients were cured of their endocarditis. There has been one probable recurrent endocarditis on the right side in a drug addict 1 year after the Ross operation. The autograft was well functioning and without pathology.

One patient developed pulmonary homograft stenosis which required reoperation 9 months after his Ross operation. Other complications included: late tamponade (n=1), transient but dialysis-requiring renal failure (n=1) and sternal pseudoarthrosis (n=1). There was no new case of heart block.

Follow-up is complete. The 33 surviving patients were evaluated clinically and studied echocardiographically the last time at a median follow-up length of 25 months (range 3–56 months). At the last examination 18 patients had no autograft insufficiency, 14 had <1+ and one had 2+ autograft insufficiency. The New York Heart Association functional class at the latest follow-up was: class I 26 patients and class II seven patients.

	Discussion

Top Abstract Introduction Material and methods Surgical findings and pathology... Results Discussion Appendix A. Conference... References

 
The results of this series demonstrate that the Ross operation can be performed successfully in patients with aortic valve endocarditis, even in the severely ill patient with advanced disease and serious preoperative complications. Since our two deaths occurred in two very sick patients with very advanced pathology, the wisdom of always performing a very complicated procedure when less demanding ones are available may be questioned. The primary motivation for using a Ross operation for patients with endocarditis is to provide them with an aortic valve replacement that in addition to not requiring anticoagulation, will have a low risk of re-infection and will have the greatest chance of being a permanent solution to their aortic valve problem. However, these arguments are valid only if the patient survives the primary operation.

Although endocarditis in children is rare (two patients in this series), it does occur, and the autograft's growth potential is a factor in its favor in this subgroup.

In several cases in this series, the autograft's more supple tissue characteristics compared with the homograft or composite graft was considered clearly advantageous. Autograft tissue quality allowed direct implantation of coronary arteries devoid of aortic wall cuff in two cases and this was the main argument for choice of a Ross procedure in the oldest patient in the series. Autograft healing is with minimal scar tissue and without risk of future calcification and reoperation for late autograft insufficiency (if and when it occurs) has proven to be a simple operation comparable with a primary valve replacement.

We believe the key to success in endocarditis is the complete removal of all bacteria-harboring and non-vital tissue/foreign material, and in many cases this is more safely and effectively accomplished with the added exposure provided by removal of both the pulmonary and aortic roots. An important observation in this series of Ross operations is that even in cases with very advanced pathology, proximal destruction and loss of tissue did occur but was uncommon. Following autograft harvesting, debridement and preparation of the aortic root and identification of the two trigones, the LVOT was in most cases intact and of normal size and allowed standard autograft implantation [14]. This observation is equally important, whatever method of reconstruction is to be used, composite graft, aortic homograft or autograft. Occasionally with left-sided extension of the infectious process, division of the pulmonary artery alone may facilitate debridement and increase safety. The total exposure of the pathology prevents the possibility of leaving closed off cavities behind and secures drainage to the pericardium.

Involvement of the pulmonary artery wall or valve in the infectious process to such a degree that it precludes a Ross operation is uncommon, but does occur.

Our technique for the Ross operation was very much developed to be suitable also for endocarditis, and minimal use of foreign material was an important point. Since according to our standard implantation technique the proximal autograft–LVOT suture line is not supported by a pericardial or prosthetic material strip, autograft implantation deep into the annulus/LVOT is important to provide support for the distensible autograft.

There is no reason to expect that the autograft is more reinfection-resistant than the homograft initially, but as homograft endothelium and hemodynamics inevitably deteriorate, an increased propensity for endocarditis can be expected in the homograft. The incidence of endocarditis after autograft root replacement has been very low. Early postoperatively there seems to be a small risk of infection and pseudoaneurysm formation in the proximal LVOT–autograft suture line, a complication that has been experienced not only after autograft but also after aortic homograft implantation.

Calcification of the homograft aortic wall will essentially always commit the surgeon to a re-root replacement in the failing homograft, a difficult undertaking even in most cases. Studies are not available that specifically address the issue of frequency and risk of reoperation for homograft failure in this patient population, although Yacoub reports a total operative mortality of 5.1% for a large series of patients undergoing reoperation with a homograft, 101 of whom had previous homograft insertion [15].

Some surgeons favor a simple operation in the setting of active infection and sepsis. In our experience the sepsis syndrome has been quickly terminated by the combination of complete surgical removal of the infection and appropriate antibiotics, and sepsis is not considered a contra-indication to advanced surgery.

The preoperative differentiation between simple and advanced endocarditis has important ramifications for both patient and surgeon. In simple endocarditis, root replacement is not necessary and thus, the full spectrum of treatment options is available, including the Ross operation. In advanced endocarditis, however, root replacement is indicated/necessary in the majority of cases. In these cases there is often a need for concomitant procedures to repair fistulas or other valve lesions and these operations are much more demanding. Accurate preoperative identification of the pathology allows triage of the patient to surgeons who are experienced in dealing with these problems and allows the surgeon to plan his surgical strategy in advance.

This material includes patients with both active and remote endocarditis. Although there are important differences between these two groups, they have in common the pathology and a propensity for one reason or other for recurrent endocarditis, and the study of the autograft's fate in these patients is of interest. Routine antibiotic prophylaxis has not been recommended to Ross operated patients but could be considered in case of autograft insufficiency.

Conclusions
We are enthusiastic about the use of the Ross operation in aortic valve endocarditis and in the difficult group of younger patients with advanced pathology, it is our preferred treatment modality. Following removal of the autograft, unparalleled exposure of the pathology and LVOT is obtained facilitating radical debridement.

Even in patients with very advanced pathology the LVOT is usually intact allowing standard autograft implantation. For patients with simple endocarditis, the Ross operation is an attractive option on its usual merits. The ultimate value of the Ross operation in patients with endocarditis will be determined by the long-term results.

	Footnotes

 
Presented at the 11th Annual meeting of the European Association for Cardio-thoracic Surgery, Copenhagen, Denmark, September 28 – October 1, 1997.

	Appendix A. Conference discussion

Top Abstract Introduction Material and methods Surgical findings and pathology... Results Discussion Appendix A. Conference... References

 
Dr C. Knott-Craig (Oklahoma, OK, USA): I would like to take issue with perhaps one or two things: in your indications for a Ross operation rather than a homograft, you mentioned (i) that there was a limited supply of aortic homografts as opposed to pulmonary homografts; that is difficult to conceptualize and (ii) that there is no growth in the homograft as opposed to the autograft; that, of course, pertains to children who are still growing and not to the adult patient, which constitutes the majority of your patients.

I would just like to mention that, in those patients who have active endocarditis, the aortic root is often very friable. These are probably not a good group of patients on whom to start doing Ross procedures, even though that may be a very exciting prospect. In these cases, using a homograft may be safer.

Dr Pettersson: I can only say that I agree with this. I think the primary object is to have the patient survive. I would not advise anybody to start doing the Ross operation in cases with endocarditis although it happened to be that way in my case. My first case, a 10-year-old boy from Greenland, was really not attractive for any alternative. Also my second patient was really challenging, a 17-year-old boy who had had four previous operations. Those difficult early operations revealed special properties with the autograft. I think you have to be very experienced with left ventricular outflow tract surgery, and I also have the advantage of having the combination of congenital and adult surgery experience. Saving the life of the patient has a higher priority than doing another Ross operation.

	References

Top Abstract Introduction Material and methods Surgical findings and pathology... Results Discussion Appendix A. Conference... References

 

1. Ross D., Jackson M., Davies J. Pulmonary autograft aortic valve replacement: long-term results. J Card Surg 1991;6:529-533.[Medline]
2. Stelzer P., Jones D.J., Elkins R.C. Aortic root replacement with pulmonary autograft. Circulation 1989;80(Suppl III):209-213.
3. Joyce F., Tingleff J., Pettersson G. Changing indications for the Ross operation. Sem Thorac Card Surg 1996;8:336-344.
4. David T.E., Komeda M., Brofman P.R. Surgical treatment of aortic root abscess. Circulation 1989;80(Suppl I):269-274.
5. Glazier J.J., Verwilghen J., Donaldson R.M., Ross D.N. Treatment of complicated prosthetic valve endocarditis with annular abscess formation by homograft aortic root replacement. J Am Coll Cardiol 1991;17:1177-1182.[Abstract]
6. Haddock D., Barrtt-Boyes B., Macedo T., Kirklin J.W., Blackstone E. Aortic valve replacement for active infectious endocarditis in 108 patients. J Thorac Cardiovasc Surg 1992;103:130-139.[Abstract]
7. McGiffin D.C., Kirklin J.K. The impact of aortic valve homografts on the treatment of aortic prosthetic valve endocarditis. Sem Thorac Card Surg 1995;7:25-31.
8. Ross D. Allograft root replacement for prosthetic endocarditis. J Cardiac Surg 1990;5:68-72.[Medline]
9. Tuna I.C., Orszulak T.A., Schaff H.V., Danielson G.K. Results of homograft aortic valve replacement for active endocarditis. Ann Thorac Surg 1990;49:619-624.[Abstract]
10. Watanabe G., Haverich A., Speier R., Dresler C., Borst H.G. Surgical treatment of active infective endocarditis with paravalvular involvement. J Thorac Cardiovasc Surg 1994;107:171-177.[Abstract/Free Full Text]
11. Oswalt J.D., Dewan S.J. Aortic infective endocarditis managed by the Ross operation. J Heart Valve Dis 1993;2:380-384.[Medline]
12. Joyce F.S., Tingleff J., Pettersson G. The Ross operation: results of early experience including treatment of endocarditis. Eur J Cardio-thorac Surg 1995;9:384-392.[Abstract]
13. Joyce F.S., Tingleff J., Pettersson G. The Ross operation in the treatment of prosthetic aortic valve endocarditis. Sem Thorac Card Surg 1995;7:38-46.
14. Joyce F.S., Tingleff J., Pettersson G. A systematic technical approach to the Ross operation. J Heart Valve Dis 1996;5:391-403.[Medline]
15. Yacoub M. The use of unstented homograft valves for aortic valve reoperations. J Thorac Cardiovasc Surg 1994;107:152-161.[Abstract/Free Full Text]

This article has been cited by other articles:

				A.C. Yankah, H. Klose, R. Petzina, M. Musci, H. Siniawski, and R. Hetzer Surgical management of acute aortic root endocarditis with viable homograft: 13-year experience Eur. J. Cardiothorac. Surg., February 1, 2002; 21(2): 260 - 267. [Abstract] [Full Text] [PDF]

				A. Prat, J. I. Saez de Ibarra, A. Vincentelli, C. Decoene, O. H. Fabre, B. Jegou, and C. Savoye Ross operation for active culture-positive aortic valve endocarditis with extensive paravalvular involvement Ann. Thorac. Surg., November 1, 2001; 72(5): 1492 - 1495. [Abstract] [Full Text] [PDF]

				J M Guerra, M P Tornos, G Permanyer-Miralda, B Almirante, M Murtra, and J Soler-Soler Long term results of mechanical prostheses for treatment of active infective endocarditis Heart, July 1, 2001; 86(1): 63 - 68. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Personal Folders Download to citation manager Author home page(s): Gösta Pettersson Frederic S. Joyce Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Pettersson, G. Articles by Joyce, F. S. Search for Related Content PubMed PubMed Citation Articles by Pettersson, G. Articles by Joyce, F. S.