A little background is helpful in understanding the regulatory environment in which the Björk-Shiley strut fracture issues were being evaluated and acted upon. These observations come from my recollection of events during work at the FDA from 1976 to 1983 and from documents publicly available through the FDA and other sources. I was the Supervisor Surgical Devices Branch (September 1980 – November 1983) Division of Cardiovascular Devices, Center for Devices and Radiological Health, U.S. Food and Drug Administration (FDA).

The original medical device amendments to the Food, Drug & Cosmetic Act were enacted in May 1976 and the first tasks the FDA had to achieve were the establishment of regulations for the major provisions of the act, including device classifications, Investigational Device Exemption rules (IDE), section 510(k) rules and premarket approval rules (PMA). 510K regulations require device manufacturers, who must register, to notify the FDA of their intent to market a medical device at least 90 days in advance. Classification was crucial as the eventual level of scrutiny for a class of products is based on this.  The law required any implant, life-supporting, or life-sustaining device to be placed in Class III (requiring premarket approval) unless there was sufficient data to down-classify the device while maintaining its reasonable safety and effectiveness.  A significant struggle took place over many implant devices with lots of effort to down-classify applied by industry and sometimes professional groups.  There was never any serious discussion, however, for using any other class than Class III for heart valve replacements.  So that’s the section where this device type landed and remains today, and it is true for any heart valve implant regardless of design.

With that in place and subsequent issuance of regulations for IDE and PMA around early 1978, the FDA was structured to begin review of new devices coming to the market.  The structure was there but the experience of conducting reviews was not.  By law, Class III products that were on the market as of May 1976 were allowed to remain on the market until the FDA called for PMA applications for those products.  This was not done for heart valves until about 1986, but in the meantime, new products were reviewed as PMAs by the FDA; no heart valve was allowed to advance through the less stringent 510(k) process.  The first of these products was the Björk-Shiley 60-degree Convexo-concave (BSCC) valve, whose application was submitted in 1978 (approval circa 1979).  (Note the spherical disc version of the Björk-Shiley valve was first marketed in 1971, thus labeled a pre-amendment device, and did not suffer strut fractures.)  The second valve reviewed through PMA was the Medtronic Hall valve, submitted in 1979 and approved in 1981. The third valve to undergo PMA review was the St. Jude Medical Valve submitted in 1981 and approved in 1982.  This timing is relevant to the strut fracture episode in that mechanical valve alternatives were available on the US market during the times of concern.

Not long after market approval in the early 1980, strut fractures of the BSCC valve began to be reported.  These reports were sporadic at first but increased in frequency as the valve became more popular peaking in the period of 1983-1987.  FDA and Shiley had a series of meetings during that time frame to discuss the scientific findings of what Shiley was researching.  Various subjects came up regarding design, manufacturing, and use, and from my view, the presented data were mixed and confusing, at least up to the point where I left the agency in late 1983. What was happening was obvious: outlet strut welds were failing in service, and that failure was catastrophic to the patient, with roughly 2 of 3 events leading to death.  The cause or causes (?) were not made clear to the FDA.  

What was apparent — the reluctance of the manufacturer to conduct a full recall.  The intent seemed to be to limit the damage to specific lots of products and recall only those.  This approach did not stop or reduce the occurrence of events.  Eventually, the FDA withdrew approval of the product and forced the total recall in 1986.  That represented half a decade of exposure of patients to a risky product.  There have been many postmortems on this issue, none of which are complementary to either the FDA or Shiley, and for good reasons in retrospect.  The most comprehensive and maybe least inflammatory review was published in Circulation 2005; 111 (21): 2850-2857):

www.ahajournals.org/doi/full/10.1161/circulationaha.105.540518

As I left the agency, the issue was only beginning to boil over.  For me, to speculate about the specifics of the Björk Shiley episode after that time would be inaccurate, but I believe it’s reasonable for me to provide observations about the early regulatory climate at the FDA.  

From the start of those regulations in 1976, the FDA’s preference for their relationship with the industry was that compliance with rules and regulations should be voluntary in nature and that despite the legal adversarial nature of regulations, it was best for all parties, including patients. That business should be conducted based on science over law.  This was always a precarious balancing act and remains so today in my experience.  In considering the events surrounding Björk Shiley strut fractures, one should be aware that this occurred relatively early in the FDA’s regulation of medical devices, and even the major controversy surrounding breast implants had yet to occur.   So basically, this was the first major problem to face the FDA as it was determining how best to exercise its legal authority and the first such problem for the industry to address as well.  Did either side fare well?  In retrospect, probably not, but in the midst of the storm, I think the folks intimately involved would also admit to errors and lessons learned in the process. 

I am sure it was not a comfortable place to be, having personally experienced a similar event later in my career with a totally different outcome based largely on the Shiley experience (that would be the Duromedics mechanical valve recall in 1988.) Duromedics original prosthesis: what do we really know about diagnosis and mechanism of leaflet escape? - PubMed (nih.gov).

It was suggested that part of the FDA’s reluctance to force a full recall was based on the market dominance of Shiley at the time and concern that removal of the product would put patients at risk of not being able to get needed medical care due to supply problems.  While I’m sure that would come up in discussions, I’m doubtful the FDA would bring that up since other already approved alternatives existed or even give the concern serious consideration based on available competition.  Certainly, this was not a concern by 1986, when the approval was withdrawn by the FDA. 

We all need to be aware that when there is resistance or doubt within our regulatory and legal systems in the US, things move very slowly.  I believe that is why the FDA prefers voluntary compliance to forced action.  It is really “coerced compliance” versus the legal due process route.

The aftermath of the Björk Shiley events within the heart valve world caused a major tightening of scrutiny of heart valve products by all parties involved.  Everyone has become ever more sophisticated about the design, testing, manufacture, and use of valve devices, even as those devices have become more complicated over time.  It has been more than 35 years since the PMA withdrawal, and in that time, international standards have been developed and revised multiple times. Plus, a medical society devoted strictly to heart valves with participation from all parties has evolved and matured.  

Today, it is likely that one or two failures of a mechanical nature for a valve product would result in either never getting approval or an immediate recall of all such products. I believe the market would demand it and force it even if not done by regulators or industry. I believe this because these days, valves that are removed from the market when ‘fixed’ never successfully return to the market.  It is interesting to note that both the 70-degree BSCC valve and the Shiley Monostrut valve (which eliminated welding altogether) were never approved for the US market.  I suspect a soured relationship is the root cause of that fact; plus, Shiley was sold by Pfizer to the Italian company Sorin in 1992. 

Pfizer completes sale of Shiley product lines - UPI Archives