European Journal of Cardiovascular Medicine

The development of mechanical heart valve prostheses transformed the surgical treatment of advanced valvular heart disease. The Björk-Shiley valve, introduced by Viking Olov Björk and Donald Shiley in the late 1960s, was among the earliest successful tilting-disc mechanical prostheses and represented an important transition from caged-ball prostheses to lower-profile valves with improved haemodynamic characteristics [1].

The Björk-Shiley design subsequently underwent several modifications involving disc material, opening angle and strut configuration [1,5]. Long-term studies demonstrated considerable durability of several Björk-Shiley designs [3-5], although particular attention was directed towards outlet-strut fracture associated predominantly with certain convexo-concave models [6-8].

Historical studies demonstrated satisfactory long-term performance of Björk-Shiley prostheses. In the randomized comparison reported by Oxenham et al., approximately 20 years of follow-up showed substantially greater freedom from reoperation with the Björk-Shiley mechanical prosthesis than with porcine bioprostheses [3]. At 20 years, survival with the original prosthesis intact was reported as 23.5% for Björk-Shiley recipients compared with 6.7% for recipients of porcine prostheses [3]. Earlier follow-up from the same randomized experience had also demonstrated durable performance of the mechanical prosthesis [2].

Other large observational series confirmed the long-term reliability of Björk-Shiley valves [4,5]. More than 20-year follow-up after aortic valve replacement with the standard Björk-Shiley prosthesis demonstrated sustained long-term performance [4], while other long-term analyses described the evolution and valve-related outcomes of Björk-Shiley prostheses over extended follow-up [5]. Long-term experience with the Björk-Shiley Monostrut valve has likewise supported durable performance [14].

However, reports of prostheses functioning beyond 30-40 years remain uncommon and have increasingly appeared as isolated case reports. A functioning first-generation Björk-Shiley aortic prosthesis at 42 years was reported by Soofi et al. [9]. Other reports have documented 31-year, 37-year and 40-year durability of early Björk-Shiley prostheses [10-12]. More recently, functioning Björk-Shiley prostheses in both the aortic and mitral positions were reported 48 years after implantation [13].

Consequently, a cohort containing multiple patients with three to four decades of documented follow-up provides an opportunity to examine the real-world longevity of these historical mechanical prostheses. The objective of the present study was therefore to evaluate long-term follow-up of Björk-Shiley mechanical valves implanted in the aortic and mitral positions, with particular emphasis on prostheses remaining under documented follow-up beyond 20, 30 and 40 years.

Study design

This was a retrospective, observational, single-centre study of patients with a documented history of Björk-Shiley mechanical heart valve implantation. Institutional clinical records were reviewed to identify patients with Björk-Shiley valves implanted in the aortic and/or mitral positions.

Patient identification

Patients were eligible when the operative record clearly documented implantation of a Björk-Shiley prosthesis. Patients were categorized according to prosthetic position as Björk-Shiley aortic valve replacement, Björk-Shiley mitral valve replacement, or Björk-Shiley prostheses in both aortic and mitral positions. Patients undergoing concomitant cardiac procedures were retained because the principal objective was assessment of long-term follow-up of the implanted Björk-Shiley prosthesis.

Follow-up

Follow-up duration was calculated from the documented date of valve implantation to the date of the last recorded follow-up. For patients in whom either the implantation date or last follow-up date was unavailable, duration was considered unavailable rather than imputed. Aortic and mitral prostheses were analysed separately because patients with double-valve replacement contributed a prosthesis to both position-specific analyses.

Long-term follow-up was categorized as ≥20 years, ≥25 years, ≥30 years, ≥35 years and ≥40 years.

Outcomes

The primary outcome was the duration of documented follow-up following implantation of a Björk-Shiley mechanical prosthesis. Secondary descriptive outcomes included the proportion of valves with documented follow-up exceeding 20, 25, 30, 35 and 40 years and comparison of long-term follow-up according to aortic versus mitral position.

Where reliable information is available from the original records, subsequent analysis should additionally include valve-related mortality, thromboembolism, major bleeding, prosthetic valve thrombosis, infective endocarditis, structural valve dysfunction and valve reoperation.

Statistical analysis

The unit of analysis for position-specific summaries was the prosthesis. The proportion reaching each threshold was calculated as the number with documented follow-up of at least that duration divided by the number with calculable follow-up in that position (aortic, n=26; mitral, n=30). These proportions describe recorded follow-up duration and are not estimates of patient survival, valve survival or event-free survival. Seven patients contributed prostheses in both positions, so position groups are not statistically independent.

Continuous follow-up duration was summarized using the mean, median and maximum as available from the study records. Counts and percentages were reported for categorical follow-up thresholds. No hypothesis testing or between-group survival comparison was performed because individual-level time-to-event outcomes and censoring indicators were not available. Kaplan–Meier estimation, log-rank testing and hazard ratios were therefore not appropriate.

Descriptive calculations were checked against the reported prosthesis denominators. No inferential P values were calculated for the follow-up curves. 

Study population

Fifty-two individual patients with Björk-Shiley prostheses were identified. The cohort contained predominantly male patients (43/52; 82.7%), while 9 patients (17.3%) were female. Twenty-six patients had a Björk-Shiley prosthesis in the aortic position and 33 had a Björk-Shiley prosthesis in the mitral position. Seven patients had Björk-Shiley prostheses in both the aortic and mitral positions. Accordingly, valve-position analyses were performed at the prosthesis/position level rather than by simply adding the two groups as independent patients.

Aortic Björk-Shiley prostheses

Twenty-six Björk-Shiley prostheses were identified in the aortic position. Complete dates for calculation of follow-up were available in all 26 patients. The cumulative documented aortic follow-up was approximately 700 patient-years. Mean follow-up was 26.9 years, while median follow-up was 25.6 years. The maximum documented duration was 40.4 years. Twenty-two of 26 aortic prostheses (84.6%) had ≥20 years of documented follow-up. Thirteen (50.0%) had ≥25 years, 11 (42.3%) had ≥30 years, 5 (19.2%) had ≥35 years and 1 (3.8%) had ≥40 years of documented follow-up.

Table 1. Long-term follow-up according to prosthetic position

Mitral Björk-Shiley prostheses

Thirty-three Björk-Shiley prostheses were identified in the mitral position. Complete implantation and follow-up dates were available for 30 patients. Three patients had insufficient dates for calculation and were excluded only from analyses requiring duration of follow-up. The 30 evaluable patients contributed approximately 871.9 patient-years of documented follow-up. Mean follow-up was 29.1 years, and median follow-up was 29.7 years. Maximum documented follow-up was 43.4 years. Twenty-five patients had ≥20 years, 19 had ≥25 years and 15 had ≥30 years of documented follow-up. Eight patients had ≥35 years and 4 patients had ≥40 years of documented mitral Björk-Shiley follow-up.

Very-long-term prosthesis follow-up

At ≥20 years, 22/26 aortic (84.6%) and 25/30 mitral (83.3%) prostheses had documented follow-up. At ≥25 years, the corresponding figures were 13/26 (50.0%) and 19/30 (63.3%); at ≥30 years, 11/26 (42.3%) and 15/30 (50.0%); at ≥35 years, 5/26 (19.2%) and 8/30 (26.7%); and at ≥40 years, 1/26 (3.8%) and 4/30 (13.3%). Figure 1 depicts the proportions with documented follow-up reaching each threshold. These descriptive curves must not be interpreted as Kaplan–Meier survival estimates.

A particularly notable feature of the cohort was the number of patients reaching the third and fourth decades following implantation. Overall, multiple aortic and mitral prostheses remained under documented follow-up beyond 30 years. The longest aortic follow-up was approximately 40.4 years, whereas the longest mitral follow-up was approximately 43.4 years. Four mitral Björk-Shiley prostheses had documented follow-up exceeding 40 years.

Table 2. Distribution of exceptionally long-term follow-up

Principal findings

The principal observation of this study is the exceptional duration of documented follow-up achievable after Björk-Shiley mechanical valve implantation. The aortic cohort accumulated approximately 700 patient-years and the mitral cohort approximately 872 patient-years of documented follow-up. More than 40% of evaluable aortic prostheses and 50% of evaluable mitral prostheses had been followed for at least 30 years. Five prostheses had documented follow-up reaching or exceeding approximately 40 years. These findings extend the previously reported long-term durability of Björk-Shiley mechanical valves [3-5,9-13] and provide cohort-level evidence of observed longevity into the fourth postoperative decade.

Comparison with historical Björk-Shiley series

Previous long-term investigations established the durability of Björk-Shiley mechanical prostheses [2-5]. In the randomized comparison with porcine bioprostheses extending to approximately 20 years, Björk-Shiley prostheses demonstrated substantially lower reoperation rates and superior survival with the original prosthesis intact [3]. At 20 years, survival with the original prosthesis intact was 23.5% in Björk-Shiley recipients compared with 6.7% in recipients of porcine prostheses [3].

Long-term experience with standard Björk-Shiley aortic prostheses similarly demonstrated sustained performance beyond two decades [4], and another major series described the long-term evolution of the Björk-Shiley prosthesis [5]. Experience with the Björk-Shiley Monostrut prosthesis also demonstrated favourable long-term performance [14]. The current series extends the clinical perspective from conventional 15-20-year follow-up into the 30-40-year period.

Exceptional durability beyond 40 years

Individual reports have demonstrated that Björk-Shiley prostheses may remain functional for extraordinarily prolonged periods [9-13]. Soofi et al. reported a first-generation Björk-Shiley aortic valve functioning satisfactorily 42 years after implantation [9]. Other authors reported Björk-Shiley prostheses functioning at approximately 31, 37 and 40 years [10-12]. More recently, a patient with Björk-Shiley valves in both the aortic and mitral positions was reported with both prostheses functioning after 48 years [13].

The significance of the present study differs from that of these individual reports. Rather than describing a single exceptional survivor, the present cohort contains multiple patients with more than three decades of documented follow-up and several prostheses reaching four decades. This clustering of very-long-term observations supports the concept that extreme Björk-Shiley longevity is not necessarily restricted to isolated anecdotal cases, although the present data should be interpreted as observed follow-up rather than actuarial survival.

Aortic versus mitral position

A numerically greater proportion of evaluable mitral prostheses reached follow-up thresholds of 25–40 years than did aortic prostheses. This is a comparison of recorded observation duration, not of actuarial survival, prosthesis durability or freedom from adverse events. Implantation era, differential documentation, patient selection and overlap between position groups may account for the difference.

Mean documented follow-up was 26.9 years in the aortic group and 29.1 years in the mitral group. Fifteen mitral prostheses had ≥30 years of follow-up compared with 11 aortic prostheses. Four mitral prostheses exceeded 40 years of follow-up, with the longest reaching 43.4 years. These observations should be interpreted descriptively rather than as evidence that the Björk-Shiley prosthesis has greater durability in the mitral position. The relatively small sample, retrospective selection and different follow-up patterns prevent such a causal conclusion.

Historical concern regarding outlet-strut fracture

Any contemporary discussion of Björk-Shiley valves must distinguish the overall family of Björk-Shiley prostheses from the specific convexo-concave models associated with outlet-strut fracture [6-8]. Long-term surveillance of Björk-Shiley convexo-concave valves identified outlet-strut fracture as an important although model-dependent complication [7,8]. Risk varied according to prosthesis characteristics, including opening angle, valve size and manufacturing factors [7,8]. Mechanical failure and its clinical presentation were also described in earlier Björk-Shiley experience [6].

Consequently, the exceptionally long follow-up observed in the present cohort should not be interpreted as evidence that all Björk-Shiley designs have identical risk profiles. Where possible, the precise Björk-Shiley model - standard spherical disc, plano-convex, convexo-concave or Monostrut - should therefore be retrieved from operative records and incorporated into the final analysis [6-8,14].

Clinical implications

Many patients with historical mechanical prostheses continue to present to contemporary cardiac services. The present findings demonstrate that Björk-Shiley prostheses may remain in situ for several decades, consistent with earlier long-term series and exceptional case reports [3-5,9-14]. The presence of an old mechanical prosthesis alone should therefore not be equated with structural failure; clinical decisions should be based on the specific prosthesis model, documented valve function, symptoms, thromboembolic or bleeding history, anticoagulation status, ventricular function and the risks of redo cardiac surgery.

The findings also reinforce the importance of preserving historical operative information. Identification of the exact prosthesis model and size is particularly relevant decades after surgery because the long-term complication profile differs among Björk-Shiley designs, especially in relation to convexo-concave valves and outlet-strut fracture [6-8].

LIMITATIONS

A central statistical limitation is that the available data describe time to last documented follow-up rather than time to death, reoperation or other clinical events. Event dates, censoring indicators and verified outcomes are not available; Kaplan–Meier curves, log-rank tests, hazard ratios and event-free survival probabilities therefore cannot be derived from the current dataset.

This study has several important limitations. First, it is a retrospective observational study based on historical clinical records and is therefore susceptible to incomplete documentation. Second, the cohort appears to represent patients available for long-term follow-up rather than an inception cohort consisting of all consecutive Björk-Shiley implants performed during the original operative period. The present data therefore establish observed longevity among followed patients but cannot, in their current form, establish 30- or 40-year actuarial survival for all implanted patients.

Third, complete follow-up duration could not be calculated in 3 mitral patients because of missing dates. Fourth, the precise Björk-Shiley model is not documented consistently in the available dataset. This is particularly important because standard, plano-convex, convexo-concave and Monostrut designs have different historical risk profiles [6-8,14]. Fifth, detailed longitudinal information regarding thromboembolism, major bleeding, prosthetic valve thrombosis, endocarditis, reoperation and echocardiographic haemodynamics is not consistently represented in the present dataset.

Finally, survival status alone without a documented date of death is insufficient for conventional Kaplan-Meier analysis. Further verification of mortality and event dates would considerably strengthen the study.

This retrospective series describes exceptionally long documented follow-up after Björk–Shiley mechanical valve implantation in aortic and mitral positions. Among evaluable prostheses, 11 aortic and 15 mitral prostheses had at least 30 years of recorded follow-up, and one aortic and four mitral prostheses had at least 40 years. These observations should not be interpreted as actuarial survival rates.

These observations demonstrate the remarkable potential longevity of early-generation tilting-disc mechanical heart valves and extend previous long-term series and case reports into the fourth postoperative decade [3-5,9-14]. The findings should be interpreted as evidence of observed prosthesis longevity rather than actuarial survival until complete mortality and valve-related event data for the original implantation cohort are obtained.

Long-term surveillance, appropriate anticoagulation and accurate identification of the specific Björk-Shiley model remain essential in patients living with these historical prostheses.

PROPOSED TABLES AND FIGURES

Figure 1. Observed distribution of documented follow-up duration for Björk–Shiley prostheses by valve position (aortic n=26; mitral n=30). Points indicate the percentage of evaluable prostheses with records extending to at least each threshold. This is a descriptive follow-up distribution, NOT a Kaplan–Meier survival curve; dates of clinical events and censoring were not available.

1. Björk VO. A New Tilting Disc Valve Prosthesis. Scandinavian Journal of Thoracic and Cardiovascular Surgery. 1969 Jan 1;3(1):1–10. https://doi.org/10.3109/14017436909131912
2. Álvarez L, Escudero C, Figuera D, Castillo-Olivares JL. The Björk-Shiley valve prosthesis. Journal of Thoracic and Cardiovascular Surgery. 1992 Nov 1;104(5):1249–58. https://doi.org/10.1016/s0022-5223(19)34613-6
3. Oxenham H, Bloomfield P, Wheatley DJ, Lee RJ, Cunningham J, Prescott RJ, et al. Twenty year comparison of a Bjork-Shiley mechanical heart valve with porcine bioprostheses. Heart . 2003 June 13;89(7):715–21. https://doi.org/10.1136/heart.89.7.715
4. Valla J, Langanay T, Corbineau H, Sevray B, Chaperon J, Ménestret P, et al. [Long-term results over 20 years of aortic valve replacement with the standard Björk-Shiley prosthesis]. 89, PubMed. National Institutes of Health; 1996. p. 1137–43. https://pubmed.ncbi.nlm.nih.gov/8952837
5. Lindblom D, Björk VO, Semb BKH. Mechanical failure of the Björk-Shiley valve. Journal of Thoracic and Cardiovascular Surgery. 1986 Nov 1;92(5):894–907. https://doi.org/10.1016/s0022-5223(19)35850-7
6. Blot WJ, Ibrahim MA, Ivey TD, Acheson DE, Brookmeyer R, Weyman AE, et al. Twenty-Five–Year Experience With the Bjork-Shiley Convexoconcave Heart Valve. Circulation. 2005 May 31;111(21):2850–7. https://doi.org/10.1161/circulationaha.104.511659
7. Ericsson AB, Lindblom D, Semb G, Huysmans HA, Thulin LI, Scully HE, et al. Strut fracture with Björk-Shiley 70U convexo-concave valve An international multi-institutional follow-up study. European Journal of Cardio-Thoracic Surgery. 1992 Jan 1;6(7):339–46. https://doi.org/10.1016/1010-7940(92)90169-x
8. Bloomfield PM, Wheatley DJ, Prescott RJ, Miller H. Twelve-Year Comparison of a Bjork–Shiley Mechanical Heart Valve with Porcine Bioprostheses. New England Journal of Medicine. 1991 Feb 28;324(9):573–9. https://doi.org/10.1056/nejm199102283240901
9. Ahn H, Kim KH, Kim DJ, Jeong DS. Long-term Experience with the Björk-Shiley Monostrut Tilting Disc Valve. Journal of Korean Medical Science. 2007 Jan 1;22(6):1060. https://doi.org/10.3346/jkms.2007.22.6.1060
10. Soofi M, Ignaszewski M, Ashton T, Miyagishima R. Longest follow-up of in situ working Björk-Shiley valve: 42-year follow-up. Interact Cardiovasc Thorac Surg. 2016;22:375–7.
11. Decicco G, Coletti G, Fucci C, Lorusso R. Björk-Shiley Delrin prosthesis: 31 years of durability (and beyond). European Journal of Cardio-Thoracic Surgery. 2006 Jan 5;29(3):411.https://doi.org/10.1016/j.ejcts.2005.11.034
12. Brizzio ME, Zapolanski A, Pantazopoulos JS, Mentakis M. Unexpected durability of a Björk-Shiley first-generation aortic valve after 40 years of implantation. PubMed. 2010 Jan 1;19(1):149–51. https://pubmed.ncbi.nlm.nih.gov/20329501
13. Sansone F, Zingarelli EM, Dato GMA, Punta G, Flocco R, Ponte S del, et al. The 37-year durability of a Björk-Shiley Delrin-disc aortic valve prosthesis. PubMed. 2012 Jan 1 ;39(2):284–5.https://pubmed.ncbi.nlm.nih.gov/22740755

McManus T, Zinn PD. Forty-Eight-Year Functionality of Björk-Shiley Aortic and Mitral Valves in One Patient. Cureus. 2023 Sept 10;15(9). https://doi.org/10.7759/cureus.44976