731

Permanent Proplast
Temporomandibular Joint
Implants: MR Imaging of Destructive
Complications

Kurt P. Schellhas1 We studied the radiologic and pathologic changes in 30 patients (34 joints) In which
Clyde H. Wilkes2 there were locally destructive bone and soft-tissue complications associated with
Mohammad El Deeb3 previously inserted permanent temporomandibular joint (TMJ) Proplast-Teflon Implants.
The cases were selected as representative examples of patients with failed Proplast
Lawrence B. Lagrotteria4
interpositional arthroplasty, in whom images of the TMJ were obtained with conventional
Mark R. Omlie5
radiography, tomography, and MR, and in whom both surgical and histologic findings
were available. Clinical indications for imaging includedjoint pain, restricted joint motion,
crepitus, preauricular swelling, regional lymphadenopathy, malocclusion either acquired
or changed since implant surgery, and facial deformity. Surgery was then performed for
the purposes of implant retrieval and joint debridement because of destructive soft-
tissue and osseous changes observed from the imaging analyses in conjunction with
significant clinical signs and symptoms. The pathologic changes, observed 4-54 months
after implant surgery, included a destructive foreign-body-type granuboma and avascular
necrosis of the mandibular condyle and condylar neck.
Our findings suggest that MR is useful in the detection and evaluation of destructive
complications that may accompany failed Proplast-Teflon implants in the TMJ. MR is
superior to conventional radiography and tomography in detecting soft-tissue lesions
and avascular necrosis of bone. Tomography more accurately delineates soft-tissue
calcifications and cortical margins of osseous structures.

The surgical management of internal derangements involving the temporoman-

dibular joint (TMJ) includes simple meniscectomy, meniscus repair and reposition-
ing, and the insertion of either alloplastic or autogenous implants designed to
function as disk prostheses [1 -1 0]. Alloplastic implants may be either permanent
or retrievable [4-1 0]. Alloplastic implants composed of a laminate of Proplast (Vitek,
Houston, TX) with nonporous Teflon (DuPont, Wilmington, DE), silicone, and Silastic
(Dow Corning, Midland, Ml) have been used for permanent TMJ implants. These
implants have been widely used and continue to be used in this country. We
Received January 7, 1988; accepted after revi-
sion May 23, 1988. analyzed the results of imaging (conventional radiography, tomography, CT, and
Presented at the annual meeting of the American MR) in 30 patients (34 joints that were operated on later) with failed, permanent
Society of Temporomandibular Joint Surgeons, Proplast implants, and we correlated the findings with pathologic data in each case.
Miami, FL, February 1988.
1 Center for Diagnostic Imaging, 5775 Wayzata
Blvd., Ste. 190, St. Louis Park, MN 5541 6. Address Materials and Methods
reprint requests to K. P. Schellhas.
2 Park Place Center, 5775 Wayzata Blvd., Ste. Thirty patients (34 joints) who had undergone TMJ meniscectomy and interpositional
990, St. Louis, Park, MN 55416. arthroplasty with permanent Proplast-Teflon implants were selected for study. The implants
3 7-174 Moos Tower, 515 Delaware St., SE., had been in place for 4-54 months; 21 (62%) had been in place 24-54 months. Clinical
Minneapolis, MN 55455. indications for postoperative imaging included restricted joint motion (30 patients), joint pain

4 801 N. Plaza Dr., Ste. 106, Schaumberg, IL (1 4 patients), bothersomejoint noises(crepitus) during movement(eight patients), preauricular
60195. swelling(six patients), regionallymphadenopathy(three patients), malocclusion either acquired
5 250 Central Ave. N., Wayzata, MN 55391. or changed since implant surgery (18 patients), and clinically evident facial deformity (seven
patients). MR was performed with a 1 .5-T magnet (General Electric, Milwaukee, WI), and a
AJR 151:731-735,
October 1988
0361 -803x/88/i 51 4-0731 single, commercially available surface coil. Closed-mouth sagittal MR projections, 500-800/
C American Roentgen Ray Society 20/2 (TR/TE/excitations), were made with a 3-mm slice thickness, 1 -mm interspace gap, 256
732 SCHELLHAS ET AL. AJR:151, October 1988

x 256 matrix, and 12-cm field of view. Coronal images (TR = 500- irregularities (including disruption of the implant articular sur-
800) were obtained when intraarticular soft-tissue masses and bony face; implant delamination and increased signal, suggesting
destruction or findings suggesting avascular necrosis of the mandib- soft-tissue ingrowth; and fragmentation) were observed in 22
ular condyle were observed on sagittal views (22 cases) [11-13].
patients (25 joints) (Figs. 2-5). Erosion of the articular surface
Two patients (three joints) were studied with unenhanced, thin-
of the mandibular condyle was suggested in 22 joints (Figs.
section (1 .5 x 1 .5 mm thick), axial CT followed by generation of
1 -5). Areas of decreased signal within the condylar marrow
three-dimensional images and mandibular disarticulation for viewing
of the condylar changes [1 4]. The clinical indications for imaging were were observed in 24 joints (Figs. 3-5). Decreased condylar
made available to the radiologist interpreting the imaging studies. marrow signal was interpreted to represent granulomatous
Surgical implant retrieval, including joint debridement and resection marrow involvement when such signal was observed in con-
of granulatomous tissue and devitalized bone, was performed in all junction with focal erosion and/or destruction of the adjacent
cases. The mandibular condyles were inspected for areas of articular bony cortex by a soft-tissue mass (Fig. 3). Avascular necrosis
cartilage erosion, osseous destruction, granulomatous ingrowth into was suggested by the loss of marrow signal combined with
the marrow space, and articular surface collapse. Devitalized bone osseous deformity and/or collapse without adjacent cortical
and granulomatous tissue were removed and submitted for routine
destruction, or by the extension of the degree of signal loss,
histology. Attempts were made to resect the margins of granuloma-
osseous collapse, and deformity beyond the margins of the
tous tissue when direct condylar and/or temporal bone infiltration by
the granuboma was encountered. Enlarged regional lymph nodes were soft-tissue mass within the articular space (Figs. 3-5). In six
biopsied and submitted for histology. joints, a distinction between granulomatous involvement of
the marrow and avascular necrosis could not be made before
surgery (Fig. 3).
Surgical results and histology were available from all 34
Results
joints. lntraarticular foreign-body granulomas were present in
Radiographic findings included side-to-side asymmetry in all. Implant erosion, delamination, and granulomatous in-
mandibular condyle size and shape (30 patients), condylar growth and/or fragmentation were found in 30 of the joints.
deformity in previously operated joints (30 patients), and Granulomatous condylar destruction and either focal or gen-
deviation of the chin toward an implant-containing joint (nine eralized granulomatous ingrowth into the condylar marrow
patients). Tomography confirmed abnormal condylar mor- were encountered in six joints (Figs. 2 and 3). Focal or
phology in all 34 joints containing Proplast implants. Condylar generalized osseous collapse or resorption of the condyle
findings, observed in various combinations within the same without surgical evidence of granulomatous ingrowth into the
joints, included focal articular surface erosions (26 joints), marrow was encountered in 1 2 joints and interpreted to
increased radiographic density or sclerosis (1 0 joints), articu- represent avascular necrosis (Figs. 3-5). Four condyles ex-
lar surface collapse (1 6 joints), and fragmentation (four joints). hibited distinctly separate areas of both granulomatous in-
lntraarticular calcifications were observed in eight joints growth and avascular necrosis (Fig. 3). Articular bearing-
(Fig. 1). surface cartilage was either focally or extensively eroded from
Hypointense, intraarticular soft-tissue masses were ob- all 24 condyles exhibiting alterations of marrow signal. Focal
served in 26 patietns (30 joints) studied with MR. Implant glenoid fossa erosion and perforation by the granulomas were

Fig. 1.-29-year-old woman with joint pain, crepitus, and restricted motion 18 months after Fig. 2.-Sagittal image, 600/20, in 37-year-old
insertion of Proplast implant. woman with malocclusion and bilaterial joint do-
A, Closed-mouth lateral tomogram shows surgically confirmed soft-tissue calcification (short struction. Large intraarticular granuboma (G) with
arrows). Implant lies in joint space. Irregularity in surface of temporal bone at implant interface (long condylar destruction (straIght black arrow) and
arrow). penetration (white arrow) of temporal bone. Lat-
B, Sagittal, midcondyle image, 600/20, reveals focal articular surface erosion (black arrow) with eral attachment of tendon from inferior belly of
intact implant (curved arrows). StraIght whIte arrow points to dystrophic calcification. lateral pterygoid muscle (curved arrow).
AJR:151 , October1988 COMPLICATIONS FROM PROPLAST TMJ IMPLANTS 733

‘,, p

A B C
Fig. 3.-Bilateral condylar destruction, focal avascular necrosis, and granubomatous reaction In 27-year-old
woman 24 months after insertion of implants.
A, Coronal Image, 700/20, shows nonspecific focus of decreased marrow signal (straight black arrow) adjacent to area of condylar destruction by large
granuboma (curved arrows). Focal temporal bone perforation (whIte arrow) confirmed at surgery.
B and C, Sagittal (B) and coronal (C) Images show large granuboma (large curved arrows) surrounding disintegrated Implant. Cortical collapse (small
straight arrows), avascular necrosis (large straight arrows), and condylar destruction (small curved arrow) found at surgery. No surgical evidence of
marrow invasion by granuloma.

Fig. 4.-Large granuloma, avascular necrosis, Fig. 5.-A and B, Avascular necrosis and new large granuloma in 36-year-old woman with severe,
and severe condylar deformity In 31-year-old recently acquired crossbite toward collapsed joint 4 months after removal of 36-month-old failed
woman with severe open bite 36 months after Proplast-Il Implant and replacement with new Proplast-Ill implant. Sagfttal(A)and coronal (B) Images,
bilateral Proplast Insertion. Condyle (solid straight 600/20, show decreased marrow signalfrom condylar neck (stralghtarrows) beneath Intact 4-month-
black arrow) severely deformed and shortened. old implant (curved arrows). Large, highly vascular granuboma (arrowheads) encountered at surgery
Lateral ptetygoid (inferior belly) tendon (open ar- contains Isolated particles of caner Implant. No evidence of condylar involvement by granuloma at
row) Inserts onto free perbosteum above condylar surgery.
neck because of profound bone resorption. Gran-
uboma (curved arrows) dlstends joint capsule.
Extension of granuloma Into parotid gland en-
countered at surgery; no evIdence of condylar
Invasion.

encountered in eight joints and in all six joints in which the Complications associated with alloplastic TMJ implant surgery
diagnosis was suggested by either preoperative tomography have been reported in recent years [4-1 2, 15, 1 6]. Studies
or MR. Two granulomas had eroded through the temporal with conventional radiography, tomography, and CT describe
bone to the dura of the middle cranialfossa, without infiltrating complications consisting of profound bony remodeling and
the dura. destruction but do not identify avascular necrosis specifically
[7-1 1 15-1 7]. Avascular necrosis of the condyle is a frequent
,

Discussion complication offailed Proplast implant surgery and often leads

Proplast-Teflon laminate, silicone, Silastic, and other ma- to the severe structural deformities observed in these cases
terials for TMJ implants were first used in 1981 [3-10]. (Figs. 3-5) [1 2]. Our study showed that the diagnosis of
734 SCHELLHAS ET AL. AJR:151, October1988

A B
Fig. 6.-Granulomatous (G) lngrowth Into corti- Fig. 7.-Granulomatous changes in 31-year-old woman.
cal bone (B) and marrow (M) of condyle (x40). A, Destroyed joint.
Giant cells (arrows) within granuloma removed from joint space (x200).
Sagittal, midcondyle specimen from rhesus mon- B, Regional lymph node. Carbon fragments (straight arrows) surrounded by giant cells (curved
key 6 months after Proplast t.mporomandibular arrows) (x200).
joint arthroplasty. Note perlosteal disruption and
complete absence of articular cartilage. Foci of
granulomatous Invasion (straight arrows). No cvi-
dance of avascular necrosis. Normal marrow cap-
Illary (curved arrow).

avascular necrosis has important implications for manage- are crucial when local MR signal is lost because of magnetic
ment of patients, because facial deformity and unstable mal- field interference from metallic anchoring screws [1 7]. CT also
occlusion were commonly associated complications of avas- may provide osseous detail that may escape detection with
cular necrosis in our series. Avascular necrosis may result in MR [8, 1 4]. Either conventional radiography and tomography
collapse of the mandibular condyle and condylar neck (proxi- or CT provides adequate ancillary osseous detail in most
mal mandible) leading to open bite with bilateral disease or cases, and only rarely would both techniques be required.
crossbite and chin deviation toward the collapsed side (uni- The basic pathophysiology associated with failed Proplast
lateral disease) as the jaw is pulled posteriorly and upward implants appears to be a foreign-body response derived from
by the muscles of mastication. mechanical breakdown of the Proplast-Teflon laminate (Figs.
MR is an accepted imaging procedure for diagnosis of 2-5) [9, 1 0, 1 6]. Primate research indicates that the foreign-
avascular necrosis of bone and internal derangements of the body reaction rapdily destroys articular structures (Fig. 6)
TMJ [1 1 -1 3, 17-25]. The high soft-tissue contrast resolution [9]. Direct granulomatous ingrowth into the condylar marrow
of MR makes this technique ideal for studying the soft-tissue may be observed and, in our experience, appears to precede
and marrow-space complications associated with failed TMJ the development of avascular necrosis (Figs. 1-4 and 6) [9,
surgery [1 2, 1 7-20]. Decreased condylar marrow signal may 1 2]. The granuloma may extend directly into adjacent soft
be interpreted confidently to represent avascular necrosis tissues, including lymphatics, and result in regional lymphade-
when marrow signal changes are accompanied by osseous nopathy (Figs. 2 and 7) [6].
deformity, articular surface collapse without either adjacent We recommend screening patients with retained Proplast
cortical destruction or extension of an intraarticular mass into TMJ implants with submentovertex and anteroposterior jaw-
the marrow space (Figs. 4 and 5). In addition to avascular protruded radiographs, closed- and open-mouth lateral TMJ
necrosis, the differential diagnosis of decreased marrow sig- tomograms, and surface-coil MR images. The complications
nal on Ti -weighted images includes marrow infiltration (gran- associated with permanent implants warrant the clinical and
ulomatous and neoplastic), marrow fibrosis, and osseous radiologic evaluation of these patients.
sclerosis [23-25]. T2-weighted images may be helpful in
differentiating early stages of avascular necrosis from other
entities [25]. ACKNOWLEDGMENTS
Submentovertex and anteroposterior jaw-protruded skull We thank Hollis M. Fritts, Kenneth B. Heithoff, Steven D. Johnson,
radiographs, along with TMJ tomograms, provide important and Becky Borgerson for contributions.
ancillary information regarding soft-tissue calcification and
subtle cortical infractions involving the temporal bone and
mandible, commonly observed in conjunction with these im- REFERENCES
plant complications (Fig. 1) [1 0-1 4, 21]. Radiographs show
1 . Bowman K. Temporomandibular joint arthrosis and its treatment by extir-
important changes in side-to-side facial and mandibular con- potion of the disc. Acta Chir Scand [Suppl] 1947[suppl. 1 1 8];95 :156-162
dyle symmetry. Tomograms and conventional radiographs 2. Ericksson L, Westesson PL. Long term evaluation of meniscectomy of the
AJR:151, October1988 COMPLICATIONS FROM PROPLAST TMJ IMPLANTS 735

temporomandibularjoint. J Oral Maxillofac Surg 1985;43:263-269 14. Schellhas KP, El Deeb M, Wilkes CH, et al. Three dimensional computed
3. Merrill AG. Historical perspectives and comparisons of TMJ surgery for tomography in maxillofacial surgery planning. Arch Oto!aryngo!Head Neck
Internal disk derangements and arthropathy. J Craniomandib Pract Surg 1988;144:438-442
1986;4:74-85 15. Halstad A, Jones CW, Rawlings RD. A study of the reaction of human
4. Dolwick MF, Aufdemorte TB. Silicono-inducod foreign body reaction and tissue to proplast. J Biomed Mater Res 1979;13:121-134
lymphadenopathy after temporomandibular joint arthroplasty. Oral Surg 16. Rooney TP, Haug RH, Toor AH, Indresano AT. Rapid condylar degenera-
Oral Med Oral Pathol 1985;59:449-452 tion after glenoid fossa prosthesis insertion: repeat of three cases. J Oral
5. Timmis DP, Aragon SB, VanSickols JE. Aufdemorte TB. Comparative Maxillofac Surg 1988;48:240-246
study of alloplastic materials for temporomandibularjoint disc replacement 17. Kneeland JB, Ryan DE, Carrera GF, Jesmanowicz A, Fronclsz W, Hyde
in rabbits. J Oral Maxiiofac Surg 1986;44:541-554 JS. Falled temporomandibular pnt prosthesis: MR imaging. Radiology
6. Lagrotteria L, Scapino A, Granston AS, Felgonhauor D. Patient with 1987;165: 179-181
lymphadenopathy following temporomandibularjoint arthroplasty with pro- 18. Harms SE, Wilk AM, Wolford LM, Chiles DG, Mam SB. The temporoman-
pleat. J CraniomandPract 1988;4(2):172-178 dibular joint magnetic resonance imaging using surface coils. Radiology
7. Bronstoin SL Retained alloplastic temporomandibularjoint disk implants: 1985;157: 133-136
a retrospective study. Oral Surg Oral Med Oral Patho! 1987;64:135-145 19. Katzberg RW, Bessette RW, TalIentS RH, et al. Normal abnormal tempo-
8. Heffez L, MaFee M, Rosenberg H, Langer B. CT evaluation of TMJ disc romandibular joint: MR imaging with surface coil. Radiology
replacement with a Proplast-Teflon laminate. J Oral Maxi!!ofac Surg 1986;158:183-189
1987;45:657-665 20. Harms SE, Wilk AM, Magnetic resonance imaging of the temporomandib-
9. El Deeb M, Gatto DJ, Edmunds AS. Temporomandibularjoint disc replace- ular joint. RadioGraphlcs 1987;7(4):521-542
mont materials: a comparative study (abstr). J Dent Rca 1987;66: 119 21 . Schellhas KP, Wilkes CH, Omlie MR, Block JC, Larsen JW, Idolkope BI.
10. Roilne BL, Gatto DJ, Wade M, Waite DE. Tomographac evaluation of Temporomandibular joint imaging: practical applICatiOn of available tech-
temporomandibularjoints following discoplasty or placement of polytetra- nology. Arch Oto!aryngol Head Neck Surg 1987;1 13:744-748
fluoroethylene implants. J Oral Maxiiofac Surg 1988;48: 183-188 22. Aeaskin AB. Aseptic necrosis ofthe mandibularcondyle: common problem?
11 . Schellhas KP, Wilkes CH, Hoithoff KB, Omlie MA, Block JC. Temporoman- Quintessence international 1979;2:87-90
dibularjoint diagnosis of internal derangement using magnetic resonance 23. TottyWG, MurphyWA, Ganz WI, Kumar B, Daum WJ, Siegel MA. Magnetic
imaging. Minn Med 1986;69:516-519 resonance imaging of the normal and ischemic femoral head. AiR
12. Schellhas KP, Wilkes CH, Fritts HM, Omlie MA, Heithoff KB, Jahn JA. 1984;143:1273-1280
Temporomandibularjoint: MR imaging of Internal derangements and post- 24. Gillespy T Ill, Genant HK, Helms CA. Magnetic resonance imaging of
operative changes. A/NA 1987;8:1093-1101,AJR 1988;150(2):381-389 osteonecrosis. Radio.! Clln North Am 1986;24:2:193-208
13. Schellhas KP, Wilkes CH, Omlio MA, et al. The diagnosis of tomporoman- 25. Mitchell DG, Rau VM, Dalinka MK, et al. Femoral head avascular necrosis:
dibular joint disease: two-compartment arthrography and MR. AJNR correlation of MR imaging, radiographic staging, radiOnUclide imaging, and
1988;9:579-588, AJR 1988;151 :341-350 clinical findings. Radiology 1987;162:709-715

The reader’s attention is directed to the commentary on this article, which appears on the following pages.