Purpose: The purpose was to test the outcome of a custom computer as- ... TMJ reconstruction system seems to be useful in the treatment of the multiply.
J Oral MaxillofacSurg 53:106-115, 1995
Custom CAD~CAM Total Temporomandibular Joint Reconstruction System: Preliminary Multicenter Report LOUIS G. MERCURI, DDS, MS,* LARRY M. WOLFORD, DDS,t BRUCE SANDERS, DDS,:~ R. DEAN WHITE, DDS,§ ANITA HURDER, MS, II AND WILLIAM HENDERSON, PHD II Purpose: The purpose was to test the outcome of a custom computer assisted design/computer assisted manufactured (CAD/CAM) total temporomandibular joint (rMJ) reconstruction system. Patients and Methods: There were 215 patients (13 males and 202 females); the average age at reconstruction was 40.9 +_ 10.3 years (range, 15 to 77 years). There were 363 joints placed, 296 bilateral and 67 unilateral. The patients had TMJ problems for an average of 10.3 _+ 7.0 years (range, 1 to 44 years), and had undergone a mean of 5.4 _+ 4.8 (range, 0 to 28) prior unsuccessful surgeries. Preoperative and postoperative data were collected for up to 48 months using a standardized data collection format. Subjective data related to pain, function of the lower jaw, and diet, were obtained using a visual analogue scale. Objective measures of mandibular range of motion were made directly on the patient preoperatively and postoperatively. Results: Preliminary analysis of these data reveals a statistically significant decrease in pain, an increase in function, and improvement in diet (P < .0001) from the preoperative measurements to 1 and 2 years postoperatively. There was also improvement in mandibular vertical range of motion. The number of previous surgeries was a strong predictor of postoperative pain, function, and diet scores, as well as of maximal interincisal opening. A life table analysis of failures indicates good durability of the prosthesis over time. Conclusion: These preliminary data indicate that this custom CAD/CAM total TMJ reconstruction system seems to be useful in the treatment of the multiply operated, and/or anatomically mutilated TMJ. Total alloplastic joint reconstruction is a biomechanical rather than a biologic solution for the treatment of severe joint disease. 1'2 It provides an efficient and effective means of dealing with distorted articular anatomy, while avoiding the issue of the vascularity of the periarticular tissues so essential in autogenous reconstruction, and it eliminates the need for a second surgical site with all the morbidity involved. However, there is a need for a custom total alloplastic temporomandibular joint (TMJ) reconstruction prosthesis because of the increasing number of functionless, multiply operated TMJs in which a stock prosthesis is unable to correct the mutilated or variable anatomy. These patients present with one of the following situations: 1) ankylosed, degenerated, or resorbed joints; 2) failed autogenous bone or soft tissue grafts resulting in a joint area that is poorly vascularized; 3) destruction of an autogenous graft as the result of previous pathology;
* Professor of Surgery and Chief of the Division of Oral and Maxillofacial Surgery and Dental Medicine, Department of Surgery, Stritch School of Medicine, Loyola University Medical Center, Maywood, IL. t Clinical Professor of Oral and Maxillofacial Surgery, Baylor College of Dentistry; in private practice at Baylor University Medical Center, Dallas, TX. :~ Adjunct Professor of Oral and Maxillofacial Surgery, University of California at Los Angeles School of Dentistry; Active Staff, St John's Hospital and Health Center, Santa Monica, CA. § In private practice, Bedford, TX. It Biostatistician, Hines Veterans Affairs Cooperative Studies Program Coordinating Center, Hines, IL. This study was partially funded by grants from the Michael Reese Dental Medicine Association and Techmedica. Address correspondence and reprint requests to Dr Mercuri: Depaltment of Surgery, Division of Oral and Maxillofacial Surgery, Loyola University Medical Center, 2160 S First Ave, Maywood, IL 60153, © 1995 American Association of Oral and Maxillofacial Surgeons 0278-2391/95/5302-000253.00/0
106
107
MERCURI ET AL
Materials and Methods
FIGURE 1. The Techmedica Custom C A D / C A M fossa component. The bone interface is CP titanium mesh. The articulating surface is ultra-high molecular weight polyethylene.
4) resorbed fossas and condyles with severe functional and anatomic discrepancies due to failed Vitek Proplast-Teflon interpositional implants (Houston, TX), where the long-term character of the surrounding soft tissue and its ability to support an autogenous bone graft is in doubt as the result of having been affected by the foreign body giant cell reaction; 5) failed or failing Vitek fossas and/or condyles, or other alloplasts (ie, Silastic [Don Coming, Midland, MI] other total or partial joint prostheses); and 6) severe inflammatory disease where autogenous tissue reconstruction would be unpredictable (ie, rheumatoid arthritis, psoriatic arthritis, etc). The concept of the Techmedica (Camarillo, CA) custom computer assisted design/computer assisted makeup (CAD/CAM) Total TMJ reconstruction prosthesis was developed in 1988 based on the need for an alloplastic prosthesis to deal with the aforementioned clinical situations. Techmedica had been using CAD/ CAM technology for years in dealing with similar situations in orthopedic surgery. Therefore, using the same criteria for its use, the same technology found useful in orthopedics, and the same materials proven successful over many years in total orthopedic joint replacement, the TMJ CAD/CAM system evolved. The fossa component of the CAD/CAM system is constructed of commercially pure (CP) titanium customformed sheet with welded mesh that interfaces with the bone of the residual fossa, and that is mechanically bonded to a dense, ultra-high molecular weight polyethylene (UHMWPE) articulating surface (Fig 1). The mandibular component is composed of a chrome-cobaltmolybdenum condyle attached mechanically to a CP titanium or 6A1/4V ELI alloy ramus component (Fig 2). The biomechanical characteristics, functional load capacities, fatigue resistance, and wear testing data over more than l0 million cycles show that the components of this system function well within the loading forces common to the TMJ (MMED Systems, LaCanada, CA). The purpose of this study was to evaluate the outcome of the use of the Techmedica Custom CAD/CAM Total TMJ Reconstruction Prosthesis in 215 patients (363 joints) followed up to 48 months postoperatively.
The 215 patients who required total TMJ reconstruction (363 joints) for the aforementioned reasons were treated with the same protocol by one of the eight oral and maxillofacial surgeons involved in this limited prospective clinical consortium study. All of the patients were fully informed of the nature of their problem and the fact that the surgical treatment recommended was the total alloplastic custom CAD/CAM TMJ reconstruction system. After obtaining informed consent, the patients were all treated in the following manner: The surgeon determined the need for the prosthesis. 2. The patient was fully informed as to the nature of the prosthesis. 3. The surgeon completed the custom TMJ prescription and preoperative evaluation form and returned both to Techmedica. . The patient underwent a prescribed computerized tomographic (CT) scan with the teeth in occlusion. In cases where there were metallic prostheses or screws in the fossa and/or ramus, or, if the surgeon preferred, in cases of severe degeneration due to Proplast-Teflon use, a failed autogenous bone graft, or inability to achieve a functional occlusion, the procedure was staged as follows: stage I, surgery to remove all metallic or other foreign materials, thorough debridement of the joint, achievement of the best functional occlusion, and placement of maxillomandibular fixation (MMF); or stage II, CT scan with the jaws in fixation. .
FIGURE 2. The Techmedica Custom C A D / C A M ramal/condylar component. The ramal component is 6AI/4V ELI titanium alloy or CP titanium. The condylar head is chrome-cobalt-molybdenum.
108
CUSTOM CAD/CAM TOTAL TMJ RECONSTRUCTION
FIGURE 3. The custom patient model that is milled using the data obtained from the preoperative protocol CT scan.
FIGURE 5. templates.
The wax fossa, and ramal and condylar component
FIGURE 6. model.
The completed custom prostheses on the custom
5. The CT tape was sent directly to Techmedica. 6. A custom model was fabricated (Fig 3). 7. A model was forwarded to the surgeon for modification (ie, model surgery such as condylectomy, coronoidectomy, or jaw repositioning) and for custom design of the final prosthesis (Fig 4). 8. The modified model and the doctor's prescri p tion for custom design were returned to Techmedica where they were used by the biomechanical engineers to develop preliminary computer-generated blueprints and templates. 9. The custom design blueprint, templates, and modified model were returned to the surgeon for written approval (Fig 5). All of this material was then returned to Techmedica so that the manufacturing process could commence. 10. The custom TMJ prosthesis components were fabricated and shipped to the surgeon's hospital for sterilization (Fig 6). 11. The prosthesis was placed: the fossa component via the standard preauricular incision and the
FIGURE 4. model,
The surgeon's custom design traced on the custom
MERCURI ET AL
"109
FIGURE 7. The custom prostheses implanted through standard preauricular (A) and retromandibular (B) approaches.
condyle/ramus component via a retromandibular/submandibular approach (Fig 7). 12. The patient was examined on a scheduled basis (every 2 months postoperatively for the first 8 months, at 12 months, and yearly thereafter) and the postoperative data sheets were returned to Techmedica at regular intervals. The pre- and postoperative data were collected using a standardized format. Subjective data regarding pain, function of the lower jaw, and diet, were collected using a visual analogue scale of 0 to 56 mm. Objective evaluation of mandibular range of motion, measurable as interincisal opening and left and fight lateral excursion, were directly measured in millimeters in a standard fashion. The effects of the prosthesis on the three subjective and three objective measures were evaluated by examining the mean values preoperatively and postoperatively over time. A multivariate repeated measures analysis was not attempted because of the large dropoff in the number of patients followed over time. In place of this, the paired t test was used to compare preoperative means to means at 1 and 2 years postoperatively. The tests were done only at limited time points
to avoid the problem of a large number of statistical tests. Multiple regression analysis was used to identify patient characteristics that might be predictive of outcomes at 1 year postoperatively. One year was chosen because the sample size at that point was relatively large. A life table analysis 3 was done to estimate the proportion of "surviving" prostheses at 6-month intervals postoperatively up to 48 months. The probability of failure in an interval was calculated assuming that withdrawals in the interval happened, on the average, midway in that interval. Greenwood's formula 4 was used to calculate the 95% confidence intervals for the cumulative probability of survival. A Cox regression analysis 5 was performed to determine if there were any patient characteristics significantly related to time of prosthesis failure. All programming was done using SAS (SAS Institute, Inc, Cary, NC).
Results Two hundred fifteen patients were implanted with 363 (296 bilateral, 67 unilateral) Techmedica Custom CAD/CAM Total TMJ Reconstruction Prostheses. Two hundred two were females, with a mean age at surgery of 40.9 _+ 10.3 (range, 15 to 77) years. Thirteen
1 10
Table 1.
CUSTOM CAD/CAM TOTAL TMJ RECONSTRUCTION
Changes in Subjective Measures Over Time Pain
Function
Diet
Time
n
Mean
SD
n
Mean
SD
n
Mean
SD
Preoperative Month 2 Month 4 Month 6 Month 8 1 Year 2 Years 3 Years 4 Years
205 104 70 69 41 85 64 27 6
42.2 19.5 16.3 15.9 19.3 17.9 21.5 20.9 23.5
11.6 13.2 13.1 15.2 16.3 14.7 16.0 16.1 21.2
206 99 67 68 40 85 64 27 6
39.5 20.1 16.2 17.4 17.0 19.4 22.4 19.2 22.0
12.2 12.8 12.7 13.1 14.9 14.1 16.1 13.8 15.4
203 100 65 67 39 85 64 27 6
37.3 22.4 15.2 17.3 15.1 16.7 18.7 19.9 20.0
12.8 14.4 12.9 13.9 16.0 14.3 15.5 16.6 14.7
of the patients were men with a mean age at surgery of 40.3 ___9.6 (range, 25 to 61) years. The mean followup period for all patients was 13.6 (range, 0 to 48) months. The patients had TMJ problems for an average of 10.3 + 7.0 (range, 1 to 44) years, and had undergone a mean of 5.4 + 4.8 (range, 0 to 28) unsuccessful prior TMJ surgeries. Table 1 shows the effect of the prostheses on the subjective measurements over time. A higher score on these measurements reflects more dysfunction. All three measurements improved significantly over time. At 1 year postoperatively, there was a 58% reduction in the mean pain score, a 51% improvement in the mean function score, and a 55% improvement in the mean diet score. Paired t tests showed that all improvements were significant at the P < .0001 level. At 2 years postoperatively, these improvements were 49%, 43%, and 50%, respectively (all had P < .0001). The improvements created by the prostheses appeared in the second month postoperatively and continued throughout the 4-year follow-up period. Table 2 shows the effect of the prostheses on the objective data over time. A higher score on these measurements reflects improved function. These results are more mixed than the results for the subjective data. At 1 year postoperatively, there was a 27% improvement in the mean maximum interincisal opening. Using a
Table 2.
paired t test, this improvement was statistically significant (P = .0001). This improvement in mean maximum interincisal opening was sustained at the end of year 2 (31% improvement, P = .0012). On the other hand, there was a 21% decline in mean left lateral excursion (P = .02) and 15% decline in mean right lateral excursion (P = .06) from the preoperative evaluation to 1 year postoperatively. These results were sustained at the end of year 2 (31% decline, P = .0004, for mean left lateral excursion; 19% decline, P = .0037, for mean right lateral excursion). These reductions in lateral excursion results were expected because the lateral pterygoid muscle was detached in all cases of condylar amputation to place the prosthesis. A multiple regression analysis was performed to identify important predictors of the 1-year scores for the three subjective outcome variables and the three objective outcome variables. The predictor variables examined were age of the patient, years duration of the TMJ problem, number of previous surgeries, gender of the patient, etiology of the TMJ problem (trauma vs other), left side prosthesis, right side prosthesis, and preoperative value of the outcome variable. The results of the multiple regression analyses are presented in Table 3. The number of previous surgeries was an important predictor for the I-year postoperative score with all three subjective variables and for maximum
Changes in Objective Data Over Time Maximum Incisal Opening
Right Lateral Excursion
Left Lateral Excursion
Time
n
Mean
SD
Preoperative Month 2 Month 4 Month 6 Month 8 1 Year 2 Years 3 Years 4 Years
198 92 71 69 37 80 60 27 5
24.2 26.6 29.6 30.1 31.0 30.7 31.6 32.4 31.2
10.6 7.0 7.1 8.2 7.4 8.2 9.1 8.7 5.7
190 83 62 65 35 74 58 27 5
Mean
SD
n
Mean
SD
2.9 2.7 2.8 2.6 2.4 2.3 2.0 1.9 1.8
3.0 2.8 2.5 2.3 1.9 1.8 1.9 1.1 0.8
189 83 62 65 35 74 58 27 5
2.7 2.4 2.9 2.6 3.0 2.3 2.2 2.0 4.0
2.8 2.4 2.5 2.5 2.6 2.0 2.2 1.9 4.1
111
MERCURI ET AL
Table 3. Predictors of 1-Year Postoperative Scores on the Three Subjective and Three Objective Outcome Variables (P Values*) 1-Year Function Score
1-Year Diet Score
l-Year Maximal Incisal Opening
l-Year Left Lateral Excursion
.78 .40
.51 .81
.18 .13
.06 .13
.34 .91
.24 .78
.002 .82 .04 .56 .59
.01 .67 .47 .53 .60
.01 .53 .26 .62 .45
.03 .25 .69 .51 .75
.99 .71 .20 .0001 .79
.81 .74 .10 .07 .0001
.81 .30 63
.11 .25 64
.04 .28 64
.02 .44 57
.27 .51 54
.25 .56 54
Pain Score
1-Year
Predictor Age TMJ duration No. previous surgeries Sex Trauma (Y, N) Left side prosthesis Right side prosthesis Preoperative value of outcome variable R2
n
1-Year Right Lateral Excursion
* Multiple regression analyses.
interincisal opening. On average, there was a 1.6-unit worsening in pain score, a 1.3-unit worsening in function score, a 1.3-unit worsening in diet score, and a 0.7-mm decrease in m a x i m u m interincisal opening for each previous operation that the patient underwent. The number o f previous operations was not predictive o f the 1 year postoperative score on the left or right lateral excursion tests. An etiology o f trauma was margina!ly predictive o f 1 year postoperative pain score. Patients with a traumatic etiology had an average 8point increase in pain score compared with those patients with other etiologies. An etiology o f trauma was not related to any o f the other outcome variables. The preoperative scores for diet and m a x i m u m interincisal opening were significantly related to the 1-year postoperative values o f those variables. To further describe the effect of multiple prior unsuccessful T M J surgical interventions on the subjective and objective outcome variables, the patient sample
was divided into three groups based on the number of such procedures they had undergone. In group 1, patients had undergone zero to four prior surgeries, in group 2, five to nine prior surgeries, and in group 3, 10 or more prior surgeries. These divisions were arbitrarily chosen before data analysis began based on the cumulative experience of the surgeons involved in the study in dealing with such failed cases in the past. There were 111 patients in group 1, 61 patients in group 2, and 31 patients in group 3 for w h o m some outcome data were available. Figures 8 through 13 show the mean levels o f the three subjective and three objective outcome variables for groups 1 through 3 separately. Although there was some improvement in pain in all three groups (Fig 8), it is clear that the pain was least in group 1, moderate in group 2, and greatest in group 3 patients throughout the follow-up period. These same trends were also noted for function (Fig 9), diet (Fig 10), and m a x i m u m
3o
VAS
VA$
5O
50
40i
3O
20
20
1(?
10
0
4
8
12
16
20
24
28
32
36
40
44
48
Time in Months
FIGURE 8. Subjective pain data compared with the number of prior surgeries. - - [ ] - - , 0 to 4 surgeries; +, 5 to 9; ~, 10 or more surgeries,
0 0
4
8
12
16
20
24
28
32
36
40
44
48
Time in Months
FIGURE 9. Subjective function data compared with the number of prior surgeries. --IS]--, 0 to 4 surgeries; +, 5 to 9; ~, 10 or more surgeries.
1 12
CUSTOM CAD/CAM TOTAL TMJ RECONSTRUCTION
VA$
MIO
50
40
30 20
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
. . . . . . . . . . .
20
10
10
. . . . . . . . . . . . . . .
0
4
8
12
.
.
.
.
.
16
.
.
.
.
.
20
.
.
.
.
24
.
.
.
.
.
28
.
.
.
.
.
32
.
.
.
.
.
36
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
40
44
0
48
4
8
12
16
20
24
28
32
36
40
44
48
Time in Months
Time in Months
FIGURE 10. Subjective diet data compared with the number of prior surgeries. - - [ 3 - - , 0 to 4 surgeries; +, 5 to 9; ~, 10 or more surgeries.
FIGURE 11. Objective maximum interincisal opening data compared with the number of prior surgeries. - - [ ~ - - , 0 to 4 surgeries; +, 5 to 9; ~, 10 or more surgeries.
interincisal opening (Fig 11). Groups 1 and 2 seem to be similar and better than group 3 for left lateral excursion (Fig 12). No difference between groups 1, 2, and 3 are apparent for right lateral excursion (Fig 13). Failures were divided into four categories for evaluation: 1) design failures, the prosthetic components did not fit, or had loosened or become dislodged because of a design fault; 2) biologic failures, infection or immunologic response; 3) material failures, any failure directly attributable to breakage or particulation of the materials; and 4) patient failures, patients requesting removal of the prostheses without any biologic indication. There were 19 patients (8.8%) with som e type of failure. Nine patients (4.2%) had design failures, three (1.4%) had biologic failures, one (0.5%) had both design and biologic failures, one (0.5%) had a material failure, and five (2.3%) requested removal of the prosthesis. There were 28 patients (13%) lost to follow-up and two patients (0.9%) who died from causes unrelated to the prostheses. Table 4 presents a life table analysis of the failure times for all categories. The next to the last column shows the proportion of patients with "surviving"
prostheses at 6-month intervals after implantation up to 4 years. The last column gives the 95% confidence intervals for these proportions. From Table 4, it is seen that 92% of patients survived 1 year, 92% survived 2 years, 90% survived 3 years, and 88% survived 4 years. If design failures, which might be expected to occur early in any new device, are removed, the corresponding percentages were 96%, 96%, 94%, and 92%, respectively. If both design and patient failures are removed, which gives a truer picture of the long-term safety and effectiveness of the prosthesis, the corresponding percentages were all 97%. A Cox regression analysis was performed to determine if there were any patient characteristics significantly predictive of prosthesis failure. The same independent variables used in the multiple regression analysis were used in this analysis. None of the independent variables was a significant predictor of failure. Discussion The prior use of, and indications for, a total alloplastic TMJ reconstruction prosthesis have been chroni-
LLE
RLE 10
iI .................................................. 6
4
4
8
12
16
20
24
28
32
36
40
44
48
0
4
8
12
16
20
24
28
:32
36
40
44
48
Time in Months
Time in Months
FIGURE 12. Objective left lateral excursion data compared with the number of prior surgeries. - - ~ - - , 0 to 4 surgeries; +, 5 to 9; • , 10 or more surgeries.
FIGURE 13. Objective fight lateral excursion data compared with the number of prior surgeries. --[S]--, 0 to 4 surgeries; +, 5 to 9; ~, 10 or more surgeries.
MERCURI ET AL
Table 4.
1 13
Follow-Up Life Table of "Survival" of the Prostheses (All Categories of Failures)
Years After Surgery
No. of Patients at Beginning of Interval
No. of Patients Withdrawn in Interval*
No. of Patients With Prosthesis Faitures in Interval
0 0.50 1.00 1.50 2.00 2.50 3.00 3,50 4,00
211 190 178 145 119 74 51 18 4
6 11 33 26 43 23 32 14 4
15 1 0 0 2 0 I 0 0
Probability of Failure in Interval .0721 .0054 .0000 .0000 .0206 .0000 .0286 .0000 .0000
Probability of Survival in Interval
Cumulative Probability of Surviving to End of Interval
95% Confidence Interval for Cumulative Probability
.9279 .9946 1.0000 1~0000 .9794 1,0000 ,9714 1.0000 1,0000
1.0000 .9279 .9229 .9229 .9229 .9039 .9039 .8780 .8780
-(.8921,.9637) (.8859, .9599) (.8859,.9599) (.8859,.9599) (.8602, .9489) (.8602,.9489) (.8097,.9450) (.8097,.9450)
* Patients were withdrawn because of the shortness of the follow-up time for the patient or due to losses to follow-up.
cled. 6~° In all of the literature cited, concerns about stability, biocompatability of materials, and longevity are discussed. The return of the patient's stomatognathic apparatus to function was appreciated by the authors, but seldom documented objectively or in a standardized fashion as has been done in this study. Most of the patients encountered in the cited literature were candidates for alloplastic TMJ reconstruction because of either developmental or traumatic conditions, or following resection for neoplasia. Today, we find ourselves with the same stability, biocompatability, functional, and longevity concerns as our predecessors, and a similar patient population, but we are also confronted with a new and growing population presenting a multifaceted d i l e m m a - - t h e multiply operated, functionless, in chronic pain, TMJ patient who presents with scarred and mutilated TMJ anatomy requiring alloplastic reconstruction. Recent literature indicates there are also patients who have
FIGURE 14. Soft tissue removed from between the articulating surfaces of a Techmedica custom C A D / C A M total TMJ reconstruction prosthesis in function for 2 years. Note the absence of any inflammatory or giant cells. There is no evidence of particulation of the alloplastic materials (hematoxylin-cosin stain, original magnification x 10).
had Proplast-Teflon, Silastic, and acrylic resin placed in the TMJ where the failure rate of subsequent autogenous reconstruction of the TMJ is high.l~ On taking a history from these patients, the signs and symptoms present before any of their prior surgery often showed that the original etiology of the problem was not intraarticular and, therefore, not likely to have been resolved by surgery on the TMJ. The data from this study show that the signs and symptoms of myofascial pain dysfunction (MPD) syndrome or headache; with no jaw dysfunction, were the original complaint in 5.58% and 1.86%, respectively, of all cases. MPD patients and all of the headache patients were part of group 3 (10 or more prior unsuccessful TMJ surgeries). An increase in TMJ surgery has occurred over the past two decades in an attempt to surgically correct many of the problems not manageable in a nonsurgical fashion. Although many patients are helped with surgical intervention, surgical failures have added to the growing list of multiply operated, and anatomically mutilated, chronic pain TMJ patients. As can be seen from the data, group 3 patients subjectively and objectively do the poorest after alloplastic TMJ reconstruction (Figs 10 through 13). This is directly attributable to the misdiagnoses discussed above, as well as to multiple operations resulting in scarring and/or atrophy of the surrounding masticatory musculature, development of neurologic and/or vascular pathology in the area, inappropriate use of alloplasts in the TMJ, as well as patients developing systemic diseases. Patients such as those in group 3 who have undergone multiple procedures and are still functionless and in pain that is either the same or worse than when they began, present the TMJ reconstructive surgeon with the biggest challenge. These patients, as the data reveal, are likely to improve only slightly, subjectively or objectively. As further indication of the labile and desperate nature of these group 3 patients,
114 all five who requested their CAD/CAM prostheses be removed were in this group, as well as either in the MPD or headache subgroups previously mentioned. There was no clinical or radiographic evidence of any problems associated with the prostheses in these cases. However, the patients all believed that if their prostheses were removed their chronic TMJ/facial pain would be resolved. As might be expected, this did not occur. The etiologic data in this study reveal that ProplastTeflon, either as the Vitek interpositional implant or as part of the Vitek-Kent hemi or total TMJ replacement system, was involved in 17.21% of all cases. Again the majority of these patients (33.3%) were in group 3 (10 or more prior unsuccessful TMJ surgeries). The presence of such implants, and the foreign body giant cell reaction involved, had led to massive condylar and fossa destruction, complicating the reconstruction. Attempts at autogenous TMJ reconstruction (ie, temporalis fascia, dermal, auricular cartilage, costochondral, or sternoclavicular joint grafts) have met with little success because the scarred tissue bed into which these autogenous materials must be placed cannot provide an adequate blood supply for these free autogenous tissue grafts to survive, u Also, it seems to be impossible to remove all remnants of Proplast-Teflon impregnated into adjacent tissue and/or bone. The particles left in the tissues have the potential of continuing the foreign body giant cell reaction, which affects the viability of any autogenous graft material placed.1 ~Therefore, an alloplastic total TMJ prosthesis offers the best solution for such cases. The demographic data from this study reveal that the need for total TMJ reconstruction involves a much younger patient population than those who need a total hip or knee reconstruction. This population averages 40.9 years of age, which means that the longevity of the prosthesis is an important variable. Longevity of
FIGURE 15. Soft tissue removed from between the articulating surfaces of a Techmedica custom C A D / C A M total TMJ reconstruction prosthesis in function for 3 years. Note the absence of any inflammatory or giant cells. There is no evidence of particulation of the alloptastic materials (hematoxylin-eosin stain, original magnification x25).
CUSTOM CAD/CAM T O T A L TMJ RECONSTRUCTION
any implanted material is based on the proper indications for its use, correct placement and maintenance of the prosthesis, the biocompatability of the materials involved, the implant's stability in situ, and the ability of the host to understand the limitations involved with having a prosthesis in place. The Techmedica Custom CAD/CAM Total TMJ Reconstruction Prosthesis was designed with all of these factors in mind. The materials that comprise the prostheses have been used as components of orthopedic total joint systems for decades and they have been the subject of much research over that time. Stress and wear studies under loads much greater than those that can be developed in the TMJ have shown particulation of the UHMWPE leading to loss of bone, with resultant loosening and ultimate failure of some orthopedic prostheses. ~ The biocompatability of CP titanium, titanium alloy, UHMWPE, and chrome-cobalt-molybdenum are well documented in the orthopedic literature. 12 Microscopic examination (Figs 14, 15) of the intraarticular soft tissue taken from between the condylar and fossa elements of CAD/CAM prostheses functioning for 2 and 3 years, respectively, showed no evidence of foreign body, inflammatory, or giant cell reactions typical of those seen in response to alloplasts that are failing or have failed. The custom nature of this prosthesis makes it attractive on two counts. First, because it is customized to the specific anatomy of each case, there is seldom a need to alter either the anatomy or the prosthesis at surgery. This results in less surgical time and less exposure of the prosthesis to multiple try-ins and possible contamination. Secondly, and more importantly, the custom nature of these prostheses means a closer fit to the available bone and they are therefore likely to be much more stable in situ than stock prostheses that must be bent or shaped at surgery to approximate a fit. Customization to the specific anatomy of the patient and the resultant secure fit enhance longevity of the prosthesis. Data presented in the life table analysis reveal good interval success rates and acceptable cumulative success rates. If the design failures that occurred early in development are removed, the cumulative success rate after 4 years increases to 92%. If both the design and patient failures are removed, because the patient failures were not the fault of the implants, the 4-year cumulative success rate becomes 97%. This study showed that a custom CAD/CAM total TMJ reconstruction system resulted in a significant improvement in pain, function, and diet subjective scores as reported by patients, and an objective increase in maximal interincisal opening. Left and right lateral excursions were decreased due to the detachment of the lateral pterygoid muscle. The number of patients evaluated at specific followup intervals was smaller than the total number of patients entered into the study for two reasons: 1) the
CATO T. LAURENCIN
115
p a t i e n t s were accrued into the study over several years and those entered in the later stages had less followup than did the patients entered earlier; and 2) patients m i s s e d some of the f o l l o w - u p visits. A n analysis was done c o m p a r i n g the subjective and objective measurements b e t w e e n the patients' preoperative visit and their last postoperative visit to increase the sample size for the follow-up analysis. This analysis included 141 to 166 of the 215 patients in the study and p r o d u c e d the s a m e conclusions as shown in Tables 2 and 3. Furthermore, to check for bias, 12 baseline characteristics were c o m p a r e d b e t w e e n the patients having and not having evaluations at 1 y e a r postoperatively. O f the 12 characteristics, only one s h o w e d statistical significance: the patients with bilateral reconstructions tended to have 1-year evaluations while those with unilateral reconstructions did not (80.5% vs 60.9%, P = .008). The other 11 baseline characteristics were similar b e t w e e n the two groups, indicating that the patients with the 1-year evaluations were a representative s a m p l e o f all o f the study patients. T h e concept of this device was d e v e l o p e d for use in this limited prospective clinical trial based on the F e d e r a l F o o d and C o s m e t i c A c t section 21§360j(b) as it relates to c u s t o m devices. This device is not c o m m e r - cially available.
Acknowledgment The authors wish to thank Drs Demarcus Smith, David Hoffman, John Lytle, Ray Reid, and Robert Lewis for their assistance in collecting the data.
References
I. Harris WH, Sledge CB: Total hip and knee replacement. Part I. N Engl J Med 323:725, 1990 2. Harris WH, Sledge CB: Total hip and knee replacement. Part II. N Engl J Med 323:801, 1990 3. Remmington RD, Schork MA: Statistics with Applications to the Biological and Health Sciences. Englewood Cliffs, NJ, Prentice-Hall, 1970, section 13-4-3 4. Greenwood M: Reports on Public Health and Statistical Subjects, No. 13. A Report on the Natural Duration of Cancer. Appendix 1. The Errors of sampling of the survivorship tables. London, England, His Majesty's Stationery Office, 1926 5. Cox DR: Regression models and life-tables (with discussion). Series B. J R Star Soc, 34:187, 1972 6. Christensen WR: Arthroplastic Implantation of the TMJ, Oral Implantology. Springfield, IL, Thomas, 1970, pp 284298 7. Morgan DH: Surgical correction of temporomandibular joint arthritis. J Oral Surg 33:766, 1975 8. Kiehn CL, DesPrez JD, Converse CF: Total prosthetic replacement of the temporomandibular joint. Ann Plast Surg 2:5, 1979 9. Kent JN, Misiek DJ, Akin RK, et al: Temporomandibular joint condylar prosthesis: A ten year report. J Oral Maxillofac Surg 41:245, 1983 10. Sonnenburg S, Sonnenburg M: Total condylar prosthesis for alloplastic jaw articulation replacement. J Craniomaxillofac Surg 13:131, 1985 11. Henry CH, Wolford LM: Treatment outcomes for temporomandibular joint reconstruction after Proplast-Teflon implant failure. J Oral Maxillofac Surg 51:352, 1993 12. Gallante JO, Lemons J, Spector M, et al: The biologic effects of implant materials, J Orthoped Res 9:760, 1991
J Oral Maxillofac Surg 53:115-116, 1995
Discussion Custom CAD/CAM Total Temporomandibular Joint Reconstruction System: Preliminary Multicenter Report Cato T. Laurencin, MD, PhD Medical College of Pennsylvania, Philadelphia, PA Few studies are currently available in the refereed literature concerning the fate of patients who have undergone temporomandibular joint (TMJ) alloplastie reconstruction. This fact was made evident at a recent meeting sponsored by the-National Institute of Dental Research and the Food and Drug Administration. Primary questions such as how many TMJ alloplastic reconstructions have been performed, how many are currently being performed each year, and what the consensus indications are for primary and revision surgery remain unanswered. This study by Mercuri et al attempts to provide outcome information on multiply operated patients who have undergone TMJ alloplastic reconstruction. Their study, funded partly by the prosthesis
manufacturer, prospectively examined 363 total joint reconstructions performed in 215 patients. This article provides no information as to the years in which these implants were placed. Follow-up for these patients was at a mean of 13.6 months, and remarkably the authors included at least one patient in the study who had 0 months follow-up. No significant discussion of the limitations of their short-term follow-up was presented in the report. The authors found their demographic data to reveal that the patient population undergoing TMJ arthroplasty was younger than those needing total hip and knee arthroplasty. Indeed, this implies an even greater importance in obtaining longer term follow-up data. For most of the orthopedic surgery liter,atu~e, at least ~a 2-year minimum patient follow-up is required ~for i-ep'orts relating" information on alloplastic joint implants. Some minimum level of follow-up is needed to produce data that may predict patient outcome and allow for comparisons to clinical alternatives. The authors' "survival" table (Table 4) leaves one with many questions. While the table is labeled as being for prosthesis survival, data are given in reference to patients (many
