Journal Logo

Reviews: Orthopaedic Advances

Current Evidence-based Indications for Modern Noncemented Total Knee Arthroplasty

Christensen, David D. MD; Klement, Mitchell R. MD; Moschetti, Wayne E. MD, MS; Fillingham, Yale A. MD

Author Information
Journal of the American Academy of Orthopaedic Surgeons: October 15, 2020 - Volume 28 - Issue 20 - p 823-829
doi: 10.5435/JAAOS-D-20-00435
  • Free


Total knee arthroplasty (TKA) is one of the most common orthopaedic procedures and has traditionally been performed with polymethyl methacrylate bone cement to fix the femoral, tibial, and patellar implants to bone. However, patients are placing increased demands on their prosthesis and pursuing activities that typically were avoided after TKA. In addition, patients are increasingly younger, with patients less than 55 years old comprising the fastest growing age segment in most joint registries.1,2 Patients are also more frequently obese, placing increased stress on the implant cement bone interface which may lead to early failure.3,4 Thus, more durable implants are desirable because of this changing patient cohort.5 Noncemented designs were pursued beginning in the 1980s and have multiple proposed benefits. If osseointegration can be achieved, decreased stress shielding and less delayed aseptic loosening may be an advantage over traditional cemented fixation.3,6,7 Furthermore, overall costs are equivalent because noncemented TKA decreases operating room time and cement supply costs.6

Unfortunately, early noncemented designs that used screw fixation were associated with unacceptably high failure rates, particularly involving the tibial implant. In addition, registry studies have not demonstrated improved survivorship despite the proposed benefits.2 Screw tracks became a conduit into the tibial canal, increasing the functional joint space, thereby leading to tibial tray loosening (Figure 1).8 Current designs have improved porous surfaces with enhanced implant stability that allow for improved boney ingrowth and avoid screw fixation (Figure 2).1,9 This stability requires precisely engineered implants, exact bone cuts, and excellent bone stock.1

Figure 1
Figure 1:
AP radiograph of the knee demonstrating medial screw track osteolysis with extension into the intramedullary canal (osteolysis outlined with dashed line). (Reproduced with permission from Klutzny M, Singh G, Hameister R, et al: Screw track osteolysis in the cementless total knee replacement design. J Arthroplasty 2019;34:965-973.)8
Figure 2
Figure 2:
Photographs of the noncemeted porous titanium tibial baseplate (right) and cemented tibial baseplate (left) of the same implant company. The additional spikes are present on the noncemented tibial baseplate to enhance initial fixation.

We provide an evaluation of the current implant designs and assessment of the available evidence regarding the indications and contraindications to the use of noncemented TKA.

Implant Designs

Noncemented TKA implants are most often composed of cobalt-chromium alloy on the articulating surfaces with a porous coating on the backside to enhance ingrowth. Sintered beads, plasma metal deposition, and electron beam melting 3-D printing (Table 1) are technologies used to create porous surfaces with a mean pore size ranging from 425 to 450 μm and a porosity of 35% to 40% for boney ingrowth.1,3,5,10

Table 1 - Current Noncemented TKA Implants and Materials
Implant Company Nonvise TKA Noncemented Femur Noncemented Tibia Noncemente Patella
 Attune® Cementless CoCr (Porocoat®) - SB CoCr (Porocoat®) - SB
 PFC Sigma CoCr (Porocoat®) - SB CoCr (Porocoat®) - SB
DJO Global
 Empowr ® Porous™ CoCr (P2 ™) - SB Ti+ (P2 ™) - SB
 Truliant® Porous CoCr - SB Ti+ - PD
 Cinetique®a HA, CoCr - SB HA, CoCr - SB HA, CoCr - SB
 Evolution ® CoCr - SB Ti+ (Biofoam®) - 3DP
 Advance® CoCr - SB Ti+ (Biofoam®) - 3DP
 Triathlon® HA (Peri-apatite™) CoCr - SB Ti+ (Tritanium™) - 3DP Ti+ (Tritanium™) - 3DP
 Persona® Ta+ (Trabecular Metal™) - PD Ta+ (Trabecular Metal™) - PD Ta+ (Trabecular Metal™) - PD
 NexGen® Ti+ (Fiber Metal) Ta+ (Trabecular Metal™) - PD Ta+ (Trabecular Metal™) - PD
aCurrently in use outside of the US.
HA = hydroxyapetite, PD = plasma metal deposition, PFC = press fit condylar, SB = sintered beads, TKA = total knee arthroplasty, 3DP = electron beam melting 3-D printing

Cobalt-chromium alloys have widely been used in both noncemented and cemented TKA for its high ultimate strength, biocompatibility, and ability to be formed into textured fixation surfaces and polished for articulation surfaces.11 However, the modulus of elasticity of cobalt-chromium alloy is markedly higher than titanium (210 to 230 versus 110 to 116 Gigapascals) which increases the possibility of stress shielding of the underlying bone.11 However, it is important to note that both metals' modulus of elasticity is markedly higher than cortical human bone (18.0 Gigapascals).11 The fixation surface on cobalt-chromium implants may be composed of sintered cobalt-chromium beads or alternatively porous metal fixed to a cobalt-chromium base (most commonly titanium or tantalum). Noncemented cobalt-chromium TKA implants have demonstrated excellent results with minimal migration on radiosteriometeric analysis and 99.1% to 100% survival for aseptic loosening at 5 to 8 year follow-up.4,7,12,13

Titanium (Ti, atomic number 22) is widely used throughout orthopaedics because of its biocompatibility and lower modulus of elasticity than cobalt-chromium. Unfortunately, it lacks hardness and is susceptible to surface damage, making it unsuitable as an articulating surface.11 Several companies' tibial implants are composed of titanium (Stryker and Microport tibial components), whereas others have titanium porous surfaces affixed onto implants otherwise composed of cobalt-chromium (Exatech and DJO). Several studies of tibial trays composed of titanium have demonstrated excellent survivalship, minimal aseptic loosening compared with cemented designs, and improved survivorship in morbidly obese patients compared with cemented cobalt-chromium designs.3,14,15

Tantalum (Ta, atomic number 73) is an additional metal used to construct noncemented TKA implants. Named after the Greek villain Tantalus, tantalum is an inert metal that is widely used in electronic capacitors. Tantalum has a similar modulus of elasticity to traditional cobalt-chromium alloys, but it is able to be engineered to have similar mechanical properties and porosity of native bone with reliable ingrowth.1,16 Porous tantalum coatings are constructed using thermal deposition on the fixation surface of implants. Porous tantalum has been used for several applications in arthroplasty with excellent results including noncemented TKA, one series reporting 98.9% survival for aseptic loosening at the 10-year follow-up.16

Hydroxyapetite (HA) is a calcium phosphate crystal that forms the inorganic component of human bone and is osteoconductive.11 It is used as a biologically active coating applied to the fixation surface of implants with the intent of improving osteointegration. Using aseptic loosening as an end point, the largest case series reporting on 1,000 patients estimates the 10-year survival of 98.14%.1 It has been established that there is a lower incidence of radiolucent lines in HA-coated implants. However, fewer radiolucent lines does not directly translate to fewer implants requiring revision because implants can be securely ingrown with areas of spot welding.1,17 In addition, HA has been shown to increase the initial stability of implants with fewer implants showing decreased initial micromotion to promote bone ingrowth.1,4,7,18

It was previously proposed that augmented screw fixation of the tibial baseplate would improve initial stability to improve ingrowth. Regrettably, screw tracks increased the effective joint space and became a path for polyethylene and titanium particles. When screws connected with the medullary canal, this could result in catastrophic osteolysis (Figure 1).8 Fortunately, current noncemented designs do not incorporate screw fixation. Rather than relying on the stability afforded by screw fixation, contemporary designs require precise bone cuts without ridges to maximize boney contact and improve stability by press-fit means. Most designs include additional pegs or spikes, particularly in the tibial baseplate for more stability and to resist rotational forces.

Clinical Outcomes

Multiple series have demonstrated excellent outcomes in various patient-reported outcome measures but little benefit in comparative studies over cemented TKAs (Table 2). In three prospective randomized controlled trials (RCTs) comparing cemented and noncemented TKA, no notable difference was seen in the Oxford Knee Score, Knee Society Score (KSS), visual analog score, Forgotten Joint Score, and Knee Injury and Osteoarthritis and Outcome Score at the 2- and 5-year follow-up.3,7,17 Multiple case series have demonstrated similar excellent postoperative patient-reported outcome measures including challenging patient populations such as young patients (KSS: mean pain 92, range 80 to 100; mean function 84, range 0 to 100)5 and patients with rheumatoid arthritis (KSS: mean pain 92, range 80 to 95; mean function 84, range 70 to 90) at a minimum 2-year follow-up.14

Table 2 - Publication of PROMs After noncemented TKA
Publication/Study Design/Implants Follow-up Mean years (Range) Score Noncemented Mean (SD) Noncemented Mean (SD)
Fricka et al17 JOA 2019
Prospective RCT, level of evidence - I
Ta+, CR, FB
KSS cliinical
KSS functional
97.0 (3.2)
94.1 (8.6)
44.8 (3.5)
89.3 (14.2)
92.6 (10.2)
44.0 (3.8)
Van Hamersveld et al7 BJS 2017
Prospective RCT, level of evidence - I
CoCr, HA, CR, FB
5.0 KSS clinical
KSS functional
KOOS psymptoms
KOOS pain
KOOS sports
94.3 (11.7)
90.0 (12.8)
82.1 (14.5)
84.3 (15.1)
80.5 (17.2)
37.9 (28.7)
71.0 (22.8)
91.2 (13.6)
86.4 (20.9)
86.6 (13.2)
86.1 (17.9)
82.9 (17.2)
38.5 (25.6)
74.0 (19.2)
Nam et al3 JBJS 2019
Prospective RCT, level of evidence - I
Ti+ and CoCr, HA, CR, FB
2.08 (2.0-2.43) OKS
24.5 (8.9)
41.7 (18.0)
24.1 (22.5)
23.8 (8.9)
41.4 (17.3)
24.1 (26.0)
McMahon et al19 JOA 2019
Prospective cohort, level of evidence - II
Stainless steel, RP
18.1 (17.0-21.8) OKS
KSS clinial
KSS functional
32.14 (10.4)
23.77 (16.6)
83 (20)
48 (29)

Harwin et al10 JOA 2018
Prospective cohort, level of evidence - II
CoCr, HA, tibial screws, PS, FB
8.0 (7.0-9.0) KSS Clinical
KSS functional

Sultan et al15 JOA 2018
Retrospective case series, level of evidence - IV
Ti+ and CoCr, HA, PS, RP
3.7 (3.0-8.0) KSS clinical
KSS functional

Patel et al14 Ortho 2018
Retrospective case series, level of evidence - IV
Ti+ and CoCr, HA, PS, RP
4 (2-8) KSS clinical
KSS functional

DeFrancesco et al16 JBJS 2018 Retrospective case series, level of evidence - IV
Ta+, CR, monoblock tibia
∼10.0 KSS functional 68
ADL = activities of daily living, CoCr = cobalt chromium, CR = cruciate retaining polyethylene insert, FB = fixed bearing, HA = hydroxyapatite coated, FJS = Forgotten Joint Score, KOOS = Knee Injury and Osteoarthritis and Outcome Score, KSS = Knee Society Score, OKS = Oxford Knee Score, PS = posterior stabilizing polyethylene insert, QOL = quality of life, RCT = randomized controlled trial, RP = rotating platform, Ta+ = tantalum, Ti+ = titanium, TKA = total knee arthroplasty

These results seem to be preserved at mid-term follow-up as shown by Harwin et al10 demonstrating excellent KSS at a minimum 7-year follow-up (mean pain 93, range 80 to 100; mean function 78, range 68 to 95). Although limited because of patients being lost to follow-up, good long-term results have been demonstrated at 17 to 21 year follow-up regarding the Oxford Knee Score (mean 32.1), the Bartlett Patella Score (mean 21.77), and American Knee Society evaluation and function scores, (mean 83 and mean 48, respectively).19 Overall, excellent clinical results have been obtained from noncemented TKA, and the results are comparable to traditional cemented TKA.

Radiographic Outcomes

Radiosteriometric tanalysis is a sensitive and accurate method of precisely measuring implant movement using radiopaque beads placed into the metaphyseal bone and around the polyethylene implant at the time of surgery. Early implant migration maximum total point motion (MTPM) has been found to correlate with long-term implant stability of both cemented and noncemented TKA, specifically less than 0.5 mm MTPM at 1 year and less than 0.2 mm of MTPM per year after the second postoperative year.7,12 Although notable variation was noted, titanium HA-coated noncemented tibial base plates demonstrated 0.92 mm MTPM (range 0.11 to 4.17 mm) at 1 year postoperatively, which is concerning for implant loosening.12 However, after 1 year, no significant implant migration was seen (mean 0.02 mm MTPM, SD 0.40 mm), indicating a stable implant.12 In a RCT comparing noncemented and cemented TKA, noncemented implants on average demonstrated increased total migration after 5 year follow-up compared with cemented implants (MTPM 0.97 versus 0.62 mm, P = 0.003).7 However, most noncemented implants migration occurred in the first 3 months (MTPM 0.90 mm in noncemented TKA versus 0.34 mm in cemented TKA).7 When using images obtained at 3 months postoperatively as a baseline, cemented implants had on average statistically more motion than noncemented implants (MTPM 0.16 to 0.27 mm in cemented TKA versus MTPM 0.09 to 0.13 mm in noncemented TKA, P < 0.05), suggesting that early limited subsidence may be a normal part of the ingrowth process.7


Noncemented TKA implant survivorship data and revision rates have shown comparable or superior results to cemented TKA (Table 3). Three prospective RCTs of cemented and noncemented TKA demonstrated no notable difference in revision for aseptic loosening or for all-cause revision up to 5 years postoperatively (Table 3, cemented TKA survivorship 95.9% to 100% versus noncemented TKA survivorship 95.3% to 100%).3,7,17 Long-term case series have also demonstrated excellent survivorship (Table 3). Noncemented tibial monoblock tantalum implants with posterior stabilized polyethylene tibial implants demonstrated 100% tibial implant survival at 10 years in one series and 97.3% overall survivorship and 99.0% survival related to implant aseptic loosening in a rotating platform design in another series.16,19 Importantly, improved outcomes have been demonstrated in certain high-risk populations, morbidly obese patients (body mass index >40) in particular have shown to have lower rates of revisions (94.6% versus 74.1%, 99.3% vs 87% P < 0.001, respectively) principally by lowering aseptic loosening (99.1% versus 88.2%, 100% versus 94.2% P < 0.01, respectively) with a noncemented implant.4,20 In addition, several cases of patients with supposed contraindications have excellent mid-term outcomes including patients aged older than 65 years (96.7% survivorship at 2.6 years [1.5 to 4]),9 rheumatoid arthritis (99.2% survivorship at 4.0 years [2 to 8]),14 and avascular necrosis 95.9% survivorship at 3.7 years (3 to 8).15 Despite these promising results, registry data have not demonstrated improved long-term survival, suggesting that noncemented fixation may not be beneficial in all patients. In the Australian National registry, noncemented fixation has demonstrated markedly increased rates of major aseptic revisions compared with cemented fixation at all time points up to 5.5 years postoperatively.2

Table 3 - Publication of Survivorship After Noncemented TKA (Level I, II, and III Evidence Only)
Publication/Study Design/Implants Follow-up Mean yr (Range) Survivorship Aseptic Loosening Survivorship All Cause Reoperation
Noncemented Cemented Noncemented Cemented
Fricka et al17 JOA 2019
Prospective RCT, level of evidence - I
Ta+, CR, FB
5.0 97.5% (40/41) 100% (44/44) 95.3% (39/41) 95.9% (42/44)
Van Hamersveld et al7 BJS 2017
Prospective RCT, level of evidence - I
CoCr, HA, CR, FB
5.0 100% (30/30) 100% (30/30) 100% (30/30) 96.7% (29/30)
Nam et al3 JBJS 2019
Prospective RCT, level of evidence - I
Ti+ and CoCr, HA, CR, FB
2.08 (2.0-2.43) 100% (76/76) 100% (65/65) 100% (76/76) 100% (65/65)
McMahon19 JOA 2019
Prospective cohort, level of evidence - II
Stainless steel, RP
18.1 (17.0-21.8) 99.1% (463/467) 97.4% (453/467)
Harwin et al10 JOA 2018
Prospective cohort, level of evidence - II
CoCr, HA, tibial screws, PS, FB
8.0 (7.0-9.0) 100% (104/104) 97.1% (101/104)
Sinicrope et al4 JOA 2019
Retrospective case-control, level of evidence - III
CoCr, HA, tibial screws, PS, FB
Final Follow-up (4.6-14.3)
99.1% (107/108)
99.1% (107/108)
93% (79/85)
88.2% (75/85)

94.6% (103/108)

74.1% (63/85)
Nam et al18 JOA 2017
Retrospective case-control, level of evidence - III
Ti+ and CoCr, HA coated, CR, FB
1.4 (1.0-1.9) 100% (66/66) 100% (62/62) 100% (66/66) 100% (62/62)
Bagsby et al20 JOA 2016
Retrospective case-control, level of evidence - III
Ti+ and CoCr, HA and uncoated, CR & PS, FB
6.1 (3.0-9.2)
3.6 (2.4-4.8)

100% (145/145)
94.2% (145/154)

99.3% (144/145)
87.0% (134/154)
Jorgensen et al2 JBJS 2019
National Joint Registry, level of evidence – III
Cementless TKA
15.0 96.3% 96.9%
CoCr = cobalt chromium, CR = cruciate retaining polyethylene insert, FB = fixed bearing, HA = Hydroxyapetite coated, PS = posterior stabilizing polyethylene insert, RCT = randomized controlled trial, RP = rotating platform, Ta+ = tantalum, Ti+ = titanium, TKA, total knee arthroplasty
Significantly different results (P < 0.05) bolded.


Noncemented fixation in TKA has demonstrated reliable long-term results and has several specific advantages over cemented TKA including shorter operating times and decreased rates of delayed aseptic loosening.4,6 Although registry data may differ, three high-quality prospective studies of noncemented TKA have demonstrated equivalent clinical and radiographic outcomes when compared with the gold standard of cemented techniques.2,3,7,17 Modern designs have corrected previous faults by eliminating screw fixation, improving implant surface fixation, and incorporating highly cross-linked polyethylene.8

Improved survivorship has been clearly demonstrated in morbidly obese patients (body mass index >40, mean 45.3, range 40.0 to 66.1), particularly regarding aseptic loosening and may be considered an appropriate implant in this patient cohort.4,20 Despite the high risk of failure in young patients undergoing TKA regardless of the design or fixation, noncemented fixation has demonstrated successful, sustained results, and decreased the risk of long-term failure mechanisms such as aseptic loosening and late cement failure.5,16

Although excellent results in noncemented TKA have been obtained, this technology is not applicable in all clinical scenarios. Inadequate bone stock with “soft indentable bone” and uneven boney cuts resulting in uneven seating of the final prosthesis may lead to unstable implants and noncemented fixation should be aborted.9 Theoretical contraindications including advanced age (>65 years old), rheumatoid arthritis, and osteonecrosis exist because of their association with poor bone stock and surgeons should critically examine the metaphyseal bone intraoperatively in these patients (Table 4).1 However, concerns about these patient populations have not been substantiated, and successful results have been demonstrated in recent studies.9,14,15

Table 4 - Relative Indications and Contraindications of Noncemented TKA
Relative Indications Supposed Contraindications
Morbid obesity Advanced age (>65 years old)
Robust bone stock Osteoporosis
Even bone cuts Osteonecrosis
Inflammatory arthritis
TKA, total knee arthroplasty


Because TKA is being performed in more obese patients, younger patients who place increased demands on their implants and patients who are living longer; modern noncemented TKA seems to be a viable long-term option. Noncemented TKA continues to be an emerging technology with comparable clinical outcomes and survivorship to the gold standard cement fixation. New implant designs that improve initial stability without the use of screw fixation seem to have mitigated the pitfalls of past designs at least in the short term. Although further high-level studies are needed with long-term follow-up, noncemented TKA may be considered in younger and/or obese patients. Theoretical contraindications such as older patients, patients with inflammatory arthropathies, osteonecrosis, or osteoporosis require critical intraoperative bone stock evaluation before proceeding with noncemented TKA.


References printed in bold type are those published within the past 5 years.

1. Dalury DF: Cementless total knee arthroplasty: Current concepts review. Bone Joint J 2016;98-B:867-873.
2. Jorgensen NB, McAuliffe M, Orschulok T, Lorimer MF, De Steiger R: Major aseptic revision following total knee replacement: A study of 478,081 total knee replacements from the Australian Orthopaedic Association National Joint Replacement Registry. J Bone Joint Surg Am 2019;101:302-310.
3. Nam D, Lawrie CM, Salih R, Nahhas CR, Barrack RL, Nunley RM: Cemented versus cementless total knee arthroplasty of the same modern design: A prospective, randomized trial. J Bone Joint Surg Am 2019;101:1185-1192.
4. Sinicrope BJ, Feher AW, Bhimani SJ, et al.: Increased survivorship of cementless versus cemented TKA in the morbidly obese. A minimum 5-year follow-up. J Arthroplasty 2019;34:309-314.
5. Mont MA, Gwam C, Newman JM, et al.: Outcomes of a newer-generation cementless total knee arthroplasty design in patients less than 50 years of age. Ann Transl Med 2017;5:S24.
6. Yayac M, Harrer S, Hozack WJ, Parvizi J, Courtney PM: The use of cementless components does not significantly increase procedural costs in total knee arthroplasty. J Arthroplasty 2020;35:407-412.
7. Van Hamersveld KT, Marang-Van De Mheen PJ, Tsonaka R, Valstar ER, Toksvig-Larsen S: Fixation and clinical outcome of uncemented peri-apatite-coated versus cemented total knee arthroplasty: Five-year follow-up of a randomised controlled trial, using radiostereometric analysis (RSA). Bone Joint J 2017;99-B:1467-1476.
8. Klutzny M, Singh G, Hameister R, et al.: Screw track osteolysis in the cementless total knee replacement design. J Arthroplasty 2019;34:965-973.
9. Hofmann AA, Wyatt RWB, Beck SW, et al.: Cementless total knee arthroplasty in patients over 65 Years old. Clin Orthop Relat Res 1991;271:28-34.
10. Harwin SF, Levin JM, Khlopas A, et al.: Cementless posteriorly stabilized total knee arthroplasty: Seven-year minimum follow-up report. J Arthroplasty 2018;33:1399-1403.
11. O'Keefe R, Jacobs JJ, Chu CR, Einhorn TA, eds: Orthopaedic Basic Science: Foundations of Clinical Practice, ed 4. Rosemont, IL, American Academy of Orthopaedic Surgeons, 2012.
12. Dunbar MJ, Laende EK, Collopy D, Richardson CG: Stable migration of peri-apatite-coated uncemented tibial components in a multicentre study. Bone Joint J 2017;99-B:1596-1602.
13. Harwin SF, Patel NK, Chughtai M, et al.: Outcomes of newer generation cementless total knee arthroplasty: Beaded periapatite-coated vs highly porous titanium-coated implants. J Arthroplasty 2017;32:2156-2160.
14. Patel N, Gwam CU, Khlopas A, et al.: Outcomes of cementless total knee arthroplasty in patients with rheumatoid arthritis. Orthopedics 2018;41:103-106.
15. Sultan AA, Khlopas A, Sodhi N, et al.: Cementless total knee arthroplasty in knee osteonecrosis demonstrated excellent survivorship and outcomes at three-year minimum follow-up. J Arthroplasty 2018;33:761-765.
16. DeFrancesco CJ, Canseco JA, Nelson CL, Israelite CL, Kamath AF: Uncemented tantalum monoblock tibial fixation for total knee arthroplasty in patients less than 60 years of age mean 10-year follow-up. J Bone Joint Surg Am 2018;100:865-870.
17. Fricka KB, McAsey CJ, Sritulanondha S: To cement or not? Five-year results of a prospective, randomized study comparing cemented vs cementless total knee arthroplasty. J Arthroplasty 2019;34:S183-S187.
18. Nam D, Kopinski JE, Meyer Z, Rames RD, Nunley RM, Barrack RL: Perioperative and early postoperative comparison of a modern cemented and cementless total knee arthroplasty of the same design. J Arthroplasty 2017;32:2151-2155.
19. McMahon SE, Doran E, O'Brien S, Cassidy RS, Boldt JG, Beverland DE: Seventeen to twenty years of follow-up of the low contact stress rotating-platform total knee arthroplasty with a cementless tibia in all cases. J Arthroplasty 2019;34:508-512.
20. Bagsby DT, Issa K, Smith LS, et al.: Cemented vs cementless total knee arthroplasty in morbidly obese patients. J Arthroplasty 2016;31:1727-1731.
Copyright 2020 by the American Academy of Orthopaedic Surgeons.