Few studies have examined the effect of surface, size, or length to ensure better osseointegration in cancer patients.

In 2007, Nelson et al. [7] used four-implant systems in 93 cancer patients. Two had a smooth machined surface (Bränemark^®, Steri-Oss^®) and two had a rough surface (Camlog^®, Straumann^®). No significant differences were found in the success rate of osseointegration.

In 2010, Heberer et al. [8] studied 20 cancer patients who received radiation therapy with adjuvant chemotherapy. Six months after adjuvant therapy, a total of 120 modified and conventional sandblasted acid-etched implants were fitted in the mandible of all 20 patients. The mean success rate at 1 year was 96 % for the conventional implants and 100 % for the modified implants. Both surfaces presented bone loss at the implant shoulder (0.3–0.4 mm). The authors concluded that both surfaces were highly successful in irradiated bone and that the results for the modified surface were better because they were more hydrophilic and had high surface energy.

In 2011, Buddula et al. [9] performed a 21-year retrospective study of 48 patients with 271 implants fitted in areas that had received a minimum of 50 Gy and concluded that the smooth machined surface implants failed 2.9 times more than those with a rough surface. In addition, the machined surface implants failed more frequently in the maxilla than in the mandible and more in posterior than in anterior sectors. Smooth implants with a diameter of ≤3.75 mm failed more frequently than those with a diameter of ≥4 mm. However, in the rough implants, no association was found between failure and location (maxilla-mandible, anterior-posterior), bone quality, or implant diameter and length.

In 2012, Fenlon et al. [10] analyzed 41 patients with 145 implants fitted on 47 flaps and concluded that the survival of implants with a microrough surface is better that that of implants with a macrorough or smooth surface.

In 2014, Gander et al. [11] reported results for 136 microrough Astra® implants placed in 33 patients, 21 of whom had received radiation therapy at least 1 year before the implants were fitted. In this subgroup, 12 of 84 implants failed. The differences between these patients and those who did not receive radiation therapy were not significant.

The initial results of our experience with hydroxyapatite-coated impacted implants (until 1999) were very good in terms of osseointegration, even in irradiated bone. However, we did observe a progressive loss of load-bearing implants, as described by Albrektssön [12], mainly owing to treatment-refractory peri-implantitis. We believe that once this implant becomes contaminated, the surface roughness hampers treatment considerably and cannot prevent peri-implant bone loss.

Therefore, in 2006 we decided to switch to screwed models with a rough titanium surface. The new RBM-treated surfaces, which have a mean roughness of 1.4 Ra, have led to osseointegration rates close to those of noncancer patients, even in patients who have received radiation therapy. No progressive loss of load-bearing implants was observed during follow-up. Furthermore, peri-implantitis is much less frequent.

No consensus has been reached on which is the best prosthesis for rehabilitation of cancer patients. In fact, several authors report on surgical outcome and osseointegration of implants, although few report successful functional rehabilitation. In 2008, Smölka et al. [13] reported that successful osseointegration of implants is one thing, whereas the ability of implants to support a genuinely useful and functional prosthesis is another. The authors found that a 92 % success rate fell to 42.9 % owing to factors such as recurrence and lack of patient cooperation.

The ideal prosthesis in cancer patients is one that prevents lesions of the soft tissue while simultaneously providing appropriate support, retention, and stability. Therefore, in our opinion, prostheses should be implant retained and implant supported. In 1999, Weischer and Mohr [14] described how they modified their rehabilitation protocol in cancer patients during the 1990s. The authors switched from combined mucosa-implant-supported prostheses to exclusively implant-supported prostheses owing to mucosal lesions associated with the prosthesis in patients who had undergone radiation therapy, mainly osteoradionecrosis that began through infection of traumatic mucosal ulcers.

In 2006, Garrett et al. [15] defended their use of overdentures in preference to implants in cancer patients. They gave the following explanations for their preference:

In 2007, Nelson et al. [7] reported results from the rehabilitation of 93 cancer patients (68 with rigid bar-retained overdentures and 25 with fixed implant-supported prostheses). Problems with the overdenture retention clip in 11 of the 68 patients required regular management. Moreover, in two patients, movement of the prosthesis led to mucosal ulcers. No maintenance complications were detected in the patients with fixed implant-supported prostheses; therefore, the authors recommended using fixed prostheses where possible.

By contrast, Kovacs [16] and Linsen et al. [17] (in 2000 and 2009, respectively) found no differences between overdentures and fixed prostheses in terms of which was the best prosthetic solution.

Mancha de la Plata et al. [18] found a better cosmetic and functional outcome with fixed screwed prosthesis in their series of 30 patients who had undergone radiation therapy. In 2014, Fang et al. [19], on the other hand, found a better functional and cosmetic outcome with removable prostheses, since these better compensated for soft tissue defects, although they also reported several advantages of fixed rehabilitation, stating that the ideal prosthetic solution should be carefully designed on an individual basis.

We currently favor implant-supported implant-retained fixed rehabilitation, with neither contact nor support from the mucosa. This is the approach of choice in 90 % of the patients we treat. We only use overdentures when we have fewer than four implants per dental arch, in cases of large soft tissue defect (poor tongue mobility, defective lip support), or when there is a considerable increase in the interarch space that forces us to use excessively bulky fixed structures.

The occlusion of choice in fixed prosthesis is that which is mutually protected with minimum contact in the cantilevered areas; in the case of removable prosthesis, we prefer balanced occlusion.

The consensus in the literature is that to avoid gum lesions and ensure an adequate fit of implant-supported and implant-retained prostheses, it is necessary to fit 4–6 implants per arch [20, 21].

In fibula-based reconstructions, Smölka et al. [13] recommend at least five implants for fixed rehabilitation with “Bränemark-like” ceramic or resin structures. With fewer implants, it is possible to use overdentures with bar or ball attachments (O-ring) or telescopic overdentures. Magnetic abutments are not recommended.

The ideal number of implants for overdentures is the same as for noncancer patients and has been reported to be between two and four [47, 22].

The anatomic location of the implant seems to be important in patients undergoing radiation therapy. Thus, Visch et al. [23] conclude that location is the dominant variable when assessing the success of osseointegration. Better outcomes have been reported in the mandible than in the maxilla and in anterior rather than posterior sectors. Nooh et al [24] offer an overall success rate of 78.9 % in the maxilla and 93.3 % in the mandible. This better outcome in the mandible is generally attributed to the greater bone density and better blood supply in this area [25]. Several studies report 100 % success with implants fitted in the maxilla [8] or similar results for both the maxilla and the mandible [7].

In our department, therefore, the ideal minimum number of implants to ensure fixed rehabilitation is 4–6; when there are fewer, we recommend overdentures. As we generally try to use fixed prostheses, we fit between four and six implants in the edentulous mandible and six in the edentulous maxilla.

We generally fit the implants in the anterior sector; in the posterior sector, we usually fit the last implant in the area of the first and second premolars. Therefore, our rehabilitation procedures usually involve a single premolar and a molar. Although implants can be fitted more distally, such an approach is not useful in prosthetic terms, since these areas are difficult to access owing to limited mouth opening. In addition, hygiene and maintenance are somewhat complicated in the posterior areas for the same reason. Therefore, we generally try to avoid these areas.

In the initial era of implant surgery, radiation therapy was considered an absolute contraindication when fitting implants [26].

Ionizing radiation produces energy that damages or destroys cells by impairing nuclear DNA or altering the molecular characteristics of isolated cells. Radiation therapy in the oral cavity leads to changes in the bone and soft tissue. It affects bone cells and vascularization and reduces remodeling ability through damage to osteoblasts, osteocytes, and osteoclasts. In the case of osteoblasts and osteocytes, the ability of new bone to divide and synthesize is diminished. However, the osteoclasts continue to reabsorb to the extent that bone loss is faster than new bone formation. Radiation damages vessels first through hyperemia, then by endarteritis, thrombosis, and progressive occlusion until the capillaries are obliterated. Consequently, the number of bone marrow cells and bone vascularization are diminished, leading to fibrosis and fatty degeneration of bone marrow.

In the case of soft tissue, oral mucosa, gums, and glands, radiation therapy leads to inflammation, altered salivary composition, reduced amount of saliva, gum detachment, and changes in oral flora. Consequently, function is worsened and the risk of osteonecrosis increases.

Recent systematic reviews show that the risk of osteoradionecrosis is low (2 %) and that this complication is more likely during the first 2 years after radiation therapy. The risk increases with the number of procedures affecting the bone, such as tooth extraction, and in patients with poor oral health and hygiene [27, 28].

In experimental animal studies, Arnold et al. [29] used rat femur and Brasseur et al. [30] used dog mandible to show that the capacity for remodeling and osseointegration was not affected in animals that first receive implants and then radiation therapy. The results were in fact poorer in animals that first received radiation therapy and then implants. Furthermore, mineral distribution and porosity were more heterogeneous in irradiated bone. Despite these effects in dogs, osteons form correctly and osseointegration is achieved, even in bone that is irradiated before implants are fitted, provided that the implants are fitted at a later period (6 months after radiation therapy is complete according to Brasseur et al.). Schweiger [31] found similar results with dog mandibles and concluded that osseointegration was possible. By contrast, Larsen et al. [32] recorded a poorer performance with rabbit tibia and a reduction in the osseointegration surface of the implant in irradiated bone not treated with hyperbaric oxygen.

In a human postmortem histopathology study, Bolind et al. [33] examined implants fitted in irradiated bone and showed that osseointegration had occurred. These authors calculated that the mean contact surface was only 40 %: the normal percentage in nonirradiated bone was 80 % for the mandible and 60 % for the maxilla. A positive correlation was found between the figure for bone–implant contact and time between radiation therapy and placement of the implants. The authors concluded that if sufficient time is left after radiation therapy, human bone can recover part of its capacity for remodeling and achieve osseointegration.

In 2009, Ihde et al. [34] carried out a literature review of all experimental animal studies and studies of patients who had received radiation therapy and implants. They concluded that patients who receive radiation therapy have a two- to threefold greater relative risk of losing implants than nonirradiated patients.

The osseointegration surface (peri-implant trabecular bone) was analyzed by Marxand Morales [35] after 4 months in transplanted bone, normal mandible, and irradiated bone. Seventy-two percent of the implant–bone interface in transplanted bone was occupied by new bone tissue. New bone tissue accounted for 50 % of normal bone and 40 % of irradiated bone; in both cases, the tissue was sufficient to support occlusal loading. Based on Marx [36], hyperbaric oxygen was used to induce neoangiogenesis and increase fibroblast activity, thus ensuring increased partial pressure of oxygen in previously hypoxemic and ischemic irradiated areas. The standard protocol comprises 30 sessions (20 before implantation and 10 after surgery) with hyperbaric oxygen (100 %, at 2.4 atm, 90 min per session).

The series presented by Taylor and Worthington [37] in 1993 comprised 21 implants in previously irradiated mandibular bone. No implants were lost at the end of follow-up (3–7 years).

The series reported in 1998 by Niimi et al. [38] comprised 228 implants placed in irradiated maxillas in the United States and Japan. The success rate was 98 % in the case of irradiated mandibles that underwent hyperbaric oxygen therapy. The corresponding success rate in the upper jaw was 72 %. Ali et al. [39] reported a 60 % success rate in the upper jaw and a 100 % success rate in the lower jaw. In the study by Eckert et al. [40], the survival rate was 64 % in the maxilla and 99 % in the mandible.

In 2007, Nelson et al. [7] studied 93 patients, 29 of whom had undergone radiation therapy. Implant survival in the radiation therapy group was 84 % at 46 months and 54 % at 13 years. No statistically significant differences in osseointegration were found between patients who had received radiation therapy and those who had not.

Linsen et al. [17] found a 94 % implant survival rate in irradiated bone after 10 years. The authors stated that failure occurred mainly during the osseointegration phase (some months immediately after the implants were fitted) and that osteoradionecrosis associated with implants was exceptional.

In their retrospective study of 48 patients who received implants in previously irradiated bone (≥50 Gy), Buddula et al. [41] recorded survival rates of 98.9, 89.9, and 72.3 % at 1, 5, and 10 years, suggesting that osseointegration was progressively lost over the years. Furthermore, failure was much more common in the maxilla than in the mandible (success rate of 80.5 % in the maxilla vs. 93.6 % in the mandible at 5 years). No differences were found between implants fitted on irradiated remnant bone or irradiated flap bone.

In their series of 335 implants and 30 patients, Mancha de la Plata et al. [18] recorded osseointegration in 92.6 % of cases at 5 years in patients who had previously received radiation therapy. In the subgroup of patients with osteoradionecrosis, the success rate fell to 48.3 %.

In 2013, Chambrone et al. [42] carried out a systematic review of implants fitted in irradiated native bone (not flaps) and concluded that radiation therapy has a clearly negative effect on implant survival. They found a 174 % greater possibility of failure when the implants were fitted in irradiated bone. Furthermore, implants fitted in the maxilla had a 496 % greater risk of failure than those fitted in the mandible. Treatment with hyperbaric oxygen provided no survival benefit for implants in cancer patients.

In 2013, Tanaka et al. [25] reviewed and updated findings for implants in cancer patients who had received radiation therapy and found a success rate that increased from 74.4 to 98.9 % and was greater than 84 % in most studies. Both implants and rehabilitation with prostheses were successful.

Zheng et al. [43] proposed three different approaches to improve the condition of irradiated bone: hyperbaric oxygen, bone morphogenetic protein (BMP-2), and osteogenic growth peptide. The authors warned that there was no evidence in favor of these approaches and that they should be used on an individual basis.

An interesting observation was made by Gander et al. [11] in their study of 21 patients who had received radiation therapy that the radiation technique can affect osseointegration, since six implants failed in the only two patients who received conventional radiotherapy. Only six failures of osseointegration were recorded in the remaining 19 patients, who received intensity-modulated radiation therapy.

In our series, the overall success of osseointegration for implants fitted in irradiated bone was very high, close to 95 %, consistent with results in the literature for patients who did not receive hyperbaric oxygen. Most failures in patients who have received radiation therapy are due to locoregional recurrence, distant metastasis, second primary cancer, or diseases other than cancer.

The risk of osteoradionecrosis in the mandible after placement of implants was minimal in our series. We detected mandibular necrosis associated with an implant in only two patients. In both cases, pain and an orocervical fistula were observed in the parasymphysis. In one of these cases, hyperbaric oxygen was used before surgery to treat the complication, which involved withdrawal of the implant and curettage of the area affected accompanied by appropriate long-term antibiotic therapy. The complication resolved with no further sequelae.

Prosthodontic rehabilitation of cancer patients should be planned as a normal part of treatment and therefore included in the treatment schedule before ablative surgery.

The cancer team choose the optimal therapeutic approach based on the following:

The prosthodontics team can estimate the effects of these factors on oral function and the potential difficulties involved in fitting a prosthesis (loss of the neutral zone, i.e., the available dynamic space for fitting a prosthesis between the lip, cheek, and tongue, as well as alterations in tongue mobility, lip closure, and mandibular mobility) [3].

Once the anatomic alteration expected after ablative surgery is known and reconstruction of contour and function is planned, the dental prosthesis must be chosen. Ideally, this is implant supported, and the optimal position for the implants depends on the prosthesis to be fitted.