To The Editor:

In their recent letter to the editor, Dr. Dua et al. have raised an important issue regarding cold welding with wrongly placed screws, which is common with normal non locking plates. Cold welding occurs during excessive tightening of a wrongly placed screw, when the opposite surfaces lose their oxide layer and reactive metal surfaces come into contact with each other. Cold welding can further lead to jamming of the screw, and another process known as freezing. This whole process occurs because of excessive torque applied through the screw head.

In the manual describing the correct use of this particular implant, it states that the screw driver for the locking screw has a limited torque mechanism; thus, excessive torque, which causes cold welding, can not be applied. Secondly, Titanium is a softer material than stainless steel (both materials have been associated with cold welding), so if a surgeon mistakenly uses another screw driver, the head of the locking screw will be deformed by the excessive applied torque and not susceptible to cold welding.

The author(s) of this letter to the editor did not receive payment or other benefits or a commitment or agreement to provide such benefits from a commercial entity. No commercial entity paid or directed, or agreed to pay or direct, any benefits to any research fund, foundation, educational institution, or other charitable or nonprofit organization with which the author(s) are affiliated or associated.

To The Editor:

First, we would like to thank Dr. Malhotra who in his recent letter to the editor described a useful and practical tip regarding proper placement of a locking head screw.

Second, we would respectfully disagree with Vallier,et al.(1) who concluded that backing out of the locking screw was a failure by the surgeon to properly lock the screw head.

If a screw head is not properly locked into the threads of the plate, a phenomenon called ‘Cold welding’ occurs;a cold welded screw head is more, rather than less, difficult to remove from the plate than a normally locked screw. In our view, a normally locked screw can back out from the plate only if the screw has lost its hold in bone.

References

1. Heather A. Vallier, Theresa A. Hennessey, John K. Sontich, and Brendan M. Patterson Failure of LCP Condylar Plate Fixation in the Distal Part of the Femur. A Report of Six Cases J Bone Joint Surg Am 2006; 88: 846-853.

The author(s) of this letter to the editor did not receive payment or other benefits or a commitment or agreement to provide such benefits from a commercial entity. No commercial entity paid or directed, or agreed to pay or direct, any benefits to any research fund, foundation, educational institution, or other charitable or nonprofit organization with which the author(s) are affiliated or associated.

To The Editor:

We read with great interest the article titled "Failure of LCP Condylar Plate Fixation in the Distal Part of the Femur. A Report of Six Cases"(1). In their report, the authors have shown some x-rays with backing out of locking screws from the distal part of femur. This could happen only if these screws were not locked into the plate hole properly. If locked properly, a locking screw cannot back out but can fail only due to breakage usually at the head-thread junction. Therefore, we believe this complication represents a technical error by the surgeon rather than implant failure.

Based on our experience, we suggest a very helpful tip to prevent this very common but infrequently recognized complication, which usually occurs because of improper seating of the locking zig into plate hole. While seating the zig into plate hole, the zig should be rotated first in the reverse direction to be in conformity with the threads in plate hole and then it should be tightened in the opposite direction. This technique allows perfect locking of locking screw in the plate hole.

The author(s) of this letter to the editor did not receive payment or other benefits or a commitment or agreement to provide such benefits from a commercial entity. No commercial entity paid or directed, or agreed to pay or direct, any benefits to any research fund, foundation, educational institution, or other charitable or nonprofit organization with which the author(s) are affiliated or associated.

Reference:

1. Vallier HA, Hennessey TA, Sontich JK, Patterson BM. Failure of LCP condylar plate fixation in the distal part of the femur. A report of six cases. J Bone Joint Surg Am. 2006; 88: 846-853.

To The Editor:

In addition to the technical errors reported by the authors in their excellent paper, we would like to draw attention to two additional errors that increase the chance of LCP failure.

Use of conventional compression screws in the proximal portion of an LCP after a locking screw has been inserted is not recommanded, as it creates large stresses on thescrew's locking head and thread junction which may lead to screw failure.

Use of too many locking screws in the proximal portion of the plate leads to a very rigid construct, which also creates stress risers on screws, as well as on the plate, and can lead to failure.

The authors recommend the use of non-locking bicortical compressive screws in the proximal portion of the plate as their pull out strength is superior to unicortical locking screws. We believe that bicortical locking screws create a construct with more pull out strength than bicortical compressive screws. In our opinion, they should be given preference over compressive screws, when possible.

The author(s) of this letter to the editor did not receive payment or other benefits or a commitment or agreement to provide such benefits from a commercial entity. No commercial entity paid or directed, or agreed to pay or direct, any benefits to any research fund, foundation, educational institution, or other charitable or nonprofit organization with which the author(s) are affiliated or associated.

We appreciate the interest of Dr. Johnson in our recent article, “Failure of LCP Condylar Plate Fixation in the Distal Part of the Femur”(1). Dr. Johnson has raised several concerns, which we hope to address. It was our intention to provide a description and critical assessment of these instances in which failure of the LCP was encountered. We hope to contribute to the development of further understanding of this implant, and of locked plating as a concept, in order to define the indications and limitations of these techniques, and to promote optimal treatment of our patients.

We agree with Dr. Johnson that features of the injury (e.g. severe soft tissue damage, open fracture, comminution), as well as medical comorbidities such as advanced age, osteoporosis, obesity, diabetes mellitus, and tobacco use, may have a detrimental effect on fracture healing. This is stated in our discussion(1). At our level I trauma center, the majority of our patients have high energy injuries, and many also have underlying medical problems. The six patients presented in this series are typical of our patient population. Further detail is provided in Table I,(1) which compares the patients with LCP failures versus all LCP cases during our first 36 months of use of this implant.

Allograft bone grafting was performed primarily in one of the two closed fractures we described (Case 3). This was an 88 year-old woman with poor bone quality. We do not have experience with primary grafting of open fractures, but respect Dr. Johnson’s opinion in this regard. One of the patients in this series (Case 2) underwent a scheduled bone graft with autogenous iliac crest and allograft seven weeks after injury. Because this series of patients represents a retrospective assessment of their clinical course to date, there was no specific protocol, regarding radiographs, weight-bearing, or bone grafting. As is stated in the manuscript, patients were instructed to remain non-weight-bearing for a minimum of twelve weeks postoperatively, after which time progressive weight-bearing was allowed. Five of the six patients in this series began weight-bearing after 12 weeks, and Case 2 began weight bearing at 16 weeks. Recommended intervals of follow-up ranged between 6-10 weeks; however, in some cases patients missed follow-up appointments for reasons unrelated to their femur fracture (Case 6).

In all patients some fracture healing was noted on biplanar radiographs obtained prior to weight-bearing, for example Figs. 2-A and 2- B, which demonstrate metaphyseal bone formation, especially along the medial and posterior aspects of the distal femur. Our initial submission actually contained 21 figures, several of which depicted early healing in the other patients. These figures were reduced to a more reasonable number for publication at the request of the reviewers and editor. In our experience, high energy fractures in this location generally take several months for healing to be mature. Perhaps a more aggressive strategy, as Dr. Johnson proposed, to exchange implants and perform autogenous grafting during this period, would have prevented some of these failures. Because the function of our patients was not limited by pain and/or deformity, a less aggressive strategy was undertaken. It is an interesting question to consider how many patients successfully treated for distal femur fractures with an LCP, would have had unnecessary secondary operations for bone grafting, and/or implant exchange, by taking a more aggressive strategy.

We have encountered two types of implant failure with this device, now in 9 of approximately seventy patients we have treated over the past 5 years with the LCP, and in two cases sent to us from other hospitals (total of 11 cases). Plate breakage has occurred between 16 weeks and 14 months in these patients (a total of 5 plate failures). Locked plates and conventional plates have very different mechanical principles of fixation, thus providing different environments for healing, and different modes of failure. Locked screws in the distal fracture segment create a “single- beam” construct where the stiffness of the fixation is the sum of all the locked screws versus that of nonlocked screws, which function individually by providing compressive force at the plate-bone interface(2). When axial loading exceeds the stiffness of the nonlocked screw farthest from the fracture, motion will occur at the screw plate interface with screw loosening. We believe that catastrophic failure with locking plate breakage, or with shear breakage of screws at the locked screw-plate interface, is due to stiffness created because all of the locked screws in a fractured segment are a single-beam construct(2,3).

We have also observed failure with broken screws at the screw-plate interface, and in some cases loosening of locked screw heads, resulting in varus collapse (cases 3 through 6), as well as 2 additional cases since this report was written. Screw breakage and varus malalignment has occurred between three and six months after injury in all cases, shortly after weight-bearing was initiated. We have not seen failure in the distal femur with conventional fixed angle implants, like the 95º angled blade plate, which we routinely use in this location, unless coronal plane fractures or distal comminution preclude its placement. We believe that further comparative assessment of the biomechanics and cost-effectiveness of conventional fixed angle devices versus locking plates in the distal femur will be of great value in defining the optimal indications and clinical performance for each of these implants.

1. Vallier HA, Hennessey TA, Sontich JK, Patterson BM. Failure of LCP Condylar Plate Fixation in the Distal Part of the Femur. A report of six cases. J Bone Joint Surg. 2006: 88A: 846-853.

2. Egol KA, Kubiak EN, Fulkerson E, Kummer FJ, Koval KJ. Biomechanics of locked plates and screws. J Orthop Trauma. 2004; 18:488-493.

3. Fulkerson E, Koval K, Preston CF, Iesaka K, Kummer FJ, Egol KA. Fixation of periprosthetic femoral shaft fractures associated with cemented femoral stems: a biomechanical comparison of locked plating and conventional cable plates. J Orthop Trauma 2006; 20:89-93.

Appendix: The following appendix represents more specific responses to Dr. Johnson’s comments.

Case 1: We did not publish any radiographs for this patient, so we were uncertain whether Dr. Johnson was referring to Case 2 in his remarks. Patient 1 began progressive weight-bearing at 3 months and was seen twice on an outpatient basis with continued healing on plain radiographs before he returned at 9 months with sudden pain and deformity due to plate breakage.

Case 2: We agree with Dr. Johnson that if the fractures were healed the plate would not have broken at 14 months. However, this patient was followed clinically and radiographically at 6-12 week intervals after he started weight-bearing. Plain radiographs demonstrated increased bone formation over time. He had no pain, minimal swelling, and had returned to work on his farm, so a nonunion was not diagnosed earlier.

Case 6: Because of healing, particularly noticeable on the medial and posterior aspects of the distal femur at 3 months, progressive weight- bearing was initiated, despite loosening of one or more screws in the distal femur. This patient had initial fixation consisting of one 7.3mm locking screw (which became loose), three 5.0mm locking screws (two which broke and one which became loose), a 6.5mm partially threaded cancellous screw outside of the plate, and two 5.0mm conical screws (which became loose). She followed up one month later with a stable radiograph and minimal pain. She missed an appointment 6 weeks later and returned 7 months after the injury with the radiographs in Fig. 2-C.

To The Editor:

I read with great concern the article entitled, "Failure of LCP Condylar Plate Fixation in the Distal Part of the Femur"(1). The authors have attempted to provide a detailed analysis of the reasons for implant failure, but I believe this article falls far short of the intended goal. By definition, these fractures occurred in the setting of severe soft tissue injuries and whether or not one uses “minimally invasive” operative technique probably will not affect healing rates under the circumstances described. In addition, four of the six failures were type IIIA open fractures. I believe there are some details not specifically addressed in this small series of plate failures that need both recognition and explanation.

All patients with failures in this series had co-morbidities associated with their femoral fractures. These co-morbidity factors, acknowledged by the authors, had a high probability of adversely affecting the healing outcome. Several were severe polytrauma patients; three had initial spanning of the open fractures with external fixation, or their traumatic condition precluded immediate internal fixation; several were obese; and one had bilateral fractures or had diabetes and other co-morbidities.

Allograft bone was used for primary bone grafting in three patients at the index procedure. This, I have found, is of absolutely no benefit in augmenting a measurable callus response in patients with this degree of fracture trauma and find it inconsequential as a method of augmentation of healing response in grafting acute open fractures.

In these six failures, there was no standard method of determining if sufficient fracture healing had occurred before loading the implant. There was no protocol mentioned on how progressive weight bearing was begun, and there was no standard periodic follow-up with clinical or radiographic evaluation after weight bearing was initiated. The authors state that at a certain period of time postoperatively and with radiographic evidence of “some healing,” the patients were allowed progressive weight bearing. In every case, these patients were not followed up at routine intervals to assess the effect of this “progressive weight bearing” and whether or not weight bearing was detrimental to fracture healing. Follow up intervals after initial weight bearing was begun varied between 1.5 and ten months and seemed to occur only when patients returned with symptoms. Implant loosening, deformity, or plate fatigue was then diagnosed by the operating surgeon. Although failures were attributed by the authors to the implant and the technology, they may in fact be more directly related to either a lack of appropriate management of weight bearing status, failure of early detection of implant loosening, or absence of aggressive intervention by the operating surgeon to modify or alter the healing process.

Of note, only two of these six “implant failures” required revision surgery, four other “plate failures” healed. Both implant fatigue failures were after prolonged periods of what one can only assume was full weight bearing on ununited fractures with failures at nine and fourteen months postoperatively. I have reviewed the figures presented by the authors and do not believe that these failures can be attributed to failure of locked plate fixation (appendix).

There are major problems with this kind of report and it is of concern that this report was accepted by the Journal without further scrutiny as to causes of failure. The radiographs of Case 6 were misinterpreted (see appendix) suggesting potential errors in the interpretation of the rest of the data. Why were these patients allowed progressive weight bearing when radiographs revealed only “some healing” and why were they then not followed more closely to asses the presence or absence of further healing. If impending failure had been detected earlier in these patients, would the authors have intervened surgically at that time?

I believe all implant failures in this series were due to progressive weight bearing on ununited fractures before adequate healing was achieved. I do not find evidence of implant failure as a sole cause of these complications. The authors reference two series using locking plate technology with implant loosening or screw breakage of between 5 and 18%. Difficult fractures managed in less than optimal situations allowing weight bearing on plate implants before sufficient healing occurs is not a technique or implant design error as it is a management error. Lack of more closely monitored follow- up after initialization of weight bearing is a major reason of failure in these patients and lack of pursuing an aggressive implant exchange or autogenous grafting appears to have contributed to these failures. Locking Condylar Plate (LCP) technology is an attempt to bridge our capability to stabilize more extreme fractures. It has great potential but is not a 100% successful implant and as with any other implant, requires the surgeon to use the art of surgery to manage the postoperative environment surrounding the plate to maximize the healing potential. Finally, I would note that management of these fractures requires attention to detail and experience that seem to matter more than whether or not one has achieved the status of a fellowship trained orthopaedist in a Level I Trauma Center.

The author(s) of this letter to the editor did not receive payment or other benefits or a commitment or agreement to provide such benefits from a commercial entity. No commercial entity paid or directed, or agreed to pay or direct, any benefits to any research fund, foundation, educational institution, or other charitable or nonprofit organization with which the author(s) are affiliated or associated.

Reference:

1. Valliers H, Hennessey T, Sontich JK, Patterson BM. Failure of LCP condylar plate fixation in the distal part of the femur. J Bone Joint Surg Am. 2006;88A:846-853.

Appendix:

Case 1 radiographs at nine months showed a dystrophic nonunion, plate failure, no mention of weight bearing status, no mention of suspected nonunion or of any bone grafting techniques performed at time of plating. The reader has no information regarding when weight bearing was begun, and at what point full weight bearing was allowed by surgeon.

The radiographs of Case 2 show segmental bone loss. The patient underwent bone grafting at 7 weeks post injury with autogenous bone combined with allograft croutons. Progressive weight bearing was allowed at four months. The question arises if this fracture was healed or the surgeon failed to diagnose the nonunion. If the fracture was healed, the plate would not have broken. The first reduction of the fracture is anatomic, the second reduction is in valgus.

Case 3 was allowed to weight bear at 3 months after “some healing present on radiographs”. The next follow-up is three months later and revealed varus deformity and loosening.

Case 4 had “some fracture healing was noted on plain radiographs” and was allowed to weight bear. At eighteen weeks postoperative, two broken screws noted but went on to heal with 16 month radiographs showing 5 degrees of varus.

Case 5 had radiographs at 12 weeks revealing “some healing”. Patient was started on progressive weight bearing and not followed again for another 3 months when she reported mild pain and stiffness. Five degrees was noted callus and eventual healing ensued without revision.

Case 6 had twelve week radiographs revealing a shortened fracture with malreduction and implant loosening. Specifically, the authors need to answer concerns about this case. There are at least 4 screws loose in the distal articular segment in Figures 2A and 2B, page 850, at twelve week postoperative, not “one single screw” as the authors stated in their article. Even with “one loose screw,” the presence of malreduction and displacement does not constitute “some healing”. Figure 2B, if enlarged, clearly shows multiple loose screw fixations and posterior displacement deformity. It is also clear from the radiographs that many of the distal fixation screws were not all locking screws and perhaps only one of six screws was actually placed in a locking configuration. Of the screws in the distal articular fragment, two 5.0 mm screws are loose and have conical screw heads making them non-locking screws. One screw is a partially threaded 6.5mm cancellous screw (only 32 mm of screw has thread and the screw head has no locking possibility), a 7.3mm screw of unknown type, and one 5.0 mm screw which is conflicting with a freehand 5.0 mm conical screw and cannot be locked under these conditions. The fracture, at this point, is unstable, there are at least four loose screws in the distal fragment, the fracture is malreduced into valgus with posterior displacement in the lateral view. The valgus reduction is evident by the lack of parallelism between the lower 5.0 mm locking screw and a parallel line drawn through the joint space on the AP radiograph. The authors state “some healing” was evident so progressive weight bearing allowed. There is no “healing” evident at twelve weeks in this case, the fracture is unstable and the patient is then allowed to bear weight without follow- up for four more months when an “implant failure” is noticed and symptoms of pain worsen. It is clear on Figure 2C that there are multiple conical 5.0 mm screws (non-locking screws) that are loose, several broken screws and a 6.5 mm lag screw (non-locking) which is also loose. There is also medial displacement of the distal articular fragment. A four month period of ambulation on a clearly unstable and malreduced fracture is not an error of technology, but an error of diagnosis and management. The revision of the loose fixation to a 95 degree blade plate, (which is probably intra-articular), results in another valgus malreduction, (no lateral radiograph provided). How is this case a failure of locked plate fixation? The fixation was essentially non-locked and loose at twelve weeks, the fracture was malreduced and unstable. The misdiagnosis of “some healing” allowed further unprotected weight bearing until complete failure of the implant occurred and then this is reported as an implant failure?