Statement of problem
Oral rehabilitation after maxillectomy can be performed by prosthetic obturation or with a free fibula flap. Successful prosthetic obturation of large maxillectomy defects can be difficult, and masticatory function is at risk in these patients. Surgical reconstruction might provide adequate masticatory function, but the literature is lacking evidence regarding this topic.
The purpose of this pilot clinical study was to assess masticatory functions and health-related quality of life (HR-QoL) outcomes in patients after maxillectomy reconstructed by using the Rohner or the Alberta Reconstructive Technique and to compare outcomes with patients rehabilitated with an obturator prosthesis.
Material and methods
Mixing ability, maximum occlusal force, maximum mouth opening, and HR-QoL were assessed. Differences between the 2 groups were analyzed by using the Kruskal-Wallis tests for continuous variables and chi-squared tests for categorical variables.
The reconstructed patients (n=11) showed better mixing ability, occlusal force (nonoperated side), and overall mean HR-QoL. The nonreconstructed group (n=13) did not differ from the reconstructed groups in terms of maximum mouth opening, overall mean occlusal force, occlusal force on the operated side, and most HR-QoL questionnaire domains.
Maxillary reconstruction might be beneficial for masticatory performance in patients undergoing maxillectomy. A larger study is justified to support the possible benefit of the reconstruction of maxillary defects regarding mixing ability, occlusal force (nonoperated side), and HR-QoL.
This low-powered pilot study indicated that surgical reconstruction might be functionally favorable over prosthetic obturation of defects. However, other factors including donor-site morbidity, cost, and patient preference may dictate the choice of treatment more than functional outcomes.
Head and neck surgery for both benign and malignant tumors often leads to deformities of head and neck structures and to impairment of functions such as mastication, speech, and swallowing, depending on the extent of the surgical procedure. Thus, adequate oral rehabilitation and reconstruction to recover these oral functions is necessary.
The choice of rehabilitation method in patients who underwent maxillectomy is often based on the extent of the primary surgical defect, comorbidity, and patient preference. However, approaches and treatment protocols may differ from one institution to another. Rehabilitation can be done by using an implant-supported obturator prosthesis or by surgical reconstruction with local flaps, (osteo)fasciocutaneous free flaps, or (osteo)myocutaneous free flaps, mostly followed by implant surgery. Prosthetic obturation has been successfully performed in smaller defects that cannot be closed primarily or with a local flap and enables direct inspection of the wound bed for local recurrence; the value in terms of survival of this direct inspection, however, remains unclear. Nevertheless, problems with retention with a reduced dentition, stability with loss of the lateral scar band, support of soft tissues, occlusal force, trismus, and masticatory performance have been reported.
If dental implant placement is possible, it can improve the performance of obturator prostheses. The results of surgical reconstruction are reported to be favorable in terms of esthetics, oronasal separation, and functional outcomes, even with larger defects. The functional and health-related quality of life (HR-QoL) results of reconstructed patients are better than those of patients with an obturator prosthesis. However, unfavorable positioning of the bone struts can hamper implant placement in the new alveolar bone of reconstructed maxillae. Correct positioning of bone flaps and dental implants can be aided by preoperative surgical planning of the resection and reconstruction. Rohner et al developed a method for surgical reconstruction of defects by using preplanned prefabricated vascularized bone containing free flaps. Both implant placement and prosthesis fabrication are preplanned. Then, the prefabricated fibula, including dental implants, is transplanted to the receptor site, with the occlusion of the definitive prosthesis acting as a guide for bone positioning.
Because the prefabrication of the fibula requires 4 to 6 months of osseointegration, this method is unsuitable for primary reconstruction after oncological resection. In Edmonton, Alberta, the Rohner protocol was adjusted for primary oncological reconstructions and subsequently named the Alberta Reconstructive Technique (ART). The main difference with the Rohner technique is that the tumor ablation, implant placement, and fibula reconstruction are all performed during the first phase of the surgery after careful digital planning of both the resection and the reconstruction.
Compared with reconstructions completed without surgical simulation and design, the ART protocol has been reported to improve anatomic outcomes and speech, increase the proportion of placed implants that are actually loaded, decrease the overall number of implants that are used, and shorten the rehabilitation period in comparison with other reconstructive protocols. However, reports about oral functioning of patients treated with this relatively novel technique are lacking. Therefore, the purpose of this pilot clinical study was to compare masticatory performance, MMO, maximum occlusal force, and HR-QoL in participants with a reconstructed maxilla with those of participants with an obturator. The null hypothesis was that reconstructed participants would show equal performance compared with patients with an obturator.
Material and methods
Patients with a maxilla reconstructed at the University of Alberta Hospital, Edmonton, Canada, and rehabilitated at the Institute for Reconstructive Sciences in Medicine (iRSM) formed the surgical reconstruction group. Matched patients from a Dutch cohort study, operated and rehabilitated in either the University Medical Center Utrecht (UMCU) or at the Radboud University Medical Center (Radboudumc) Nijmegen, The Netherlands, were included in the obturator prosthesis group.
Patients were eligible for inclusion if they had been surgically treated for benign tumors or with a curative intent for malignant tumors. Eligible tumor locations were upper alveolar process, tuber maxillae, palate, and maxillary sinus, respectively, C03.0, C06.1, C05, C31.0 of the World Health Organization International Classification of Diseases Oncology third edition. Furthermore, when patients underwent reconstruction, the protocol was either an ART for malignant tumors or Rohner technique for benign tumors. Second-stage surgery needed to be completed. Patients with an obturator from The Netherlands were included. Exclusion criteria were cognitive impairment or the inability to understand English for the Canadian participants and an inability to understand Dutch for the Dutch participants.
All patients signed an informed consent form. The experimental protocol was approved by the Health Research Ethics Board of Alberta Cancer Committee and the Research Ethics Committee of the University of Utrecht (study ID: HREBA.CC-17-0167 and NL1200604106).
Clinical patient charts were examined for age, sex, duration since surgery, (p)TNM stage of tumor, type and location of tumor, type of treatment (surgery or surgery plus radiotherapy), radiotherapy dosage, and surgical time (minutes), as well as number and position of dental implants. The extent of the initial defect was recorded according to the Brown classification of maxillary defects. Finally, details of the reconstruction were recorded.
Dental status was examined and scored by present natural dentition, dental implants, and prostheses in both the jaws. Also, the occluding pairs were scored as premolar equivalents. Occluding fixed dental prostheses were considered in the count of occluding pairs, but third molars and tissue-supported prostheses were not.
To measure masticatory performance, the patient performed 20 mastication strokes on a wax tablet, in accordance with a previously published protocol. The tablet had a 20-mm diameter, consisting of a 3-mm-thick red and blue layer (plasticine modeling wax, nontoxic DIN EN-71, art. nos. crimson 52801 and blue 52809; Hans Stockmar GmbH & Co). The masticated wax was flattened, photographed from both sides, and then digitally assessed by using a specially developed software program. The mixing ability index was between 0 (perfect mixing ability) and 30 (poor mixing ability) and depicted the spread of the color intensities in the combined image of both sides.
Maximum occlusal force was measured by using an occlusal force transducer, as in a previously published protocol. A unilateral strain gauge mounted on a mouthpiece was placed in the first molar region to measure interocclusal forces. Participants were asked to clench as hard as possible, and measurements were made twice on the left and twice on the right side. The mean of 2 measurements on 1 side was presented as the unilateral occlusal force to differentiate between the operated and nonoperated side. The mean of the highest occlusal force on the left and right sides was presented as the total maximum occlusal force.
MMO was measured extraorally following a previously published protocol. The open and resting positions were measured twice by using digital calipers, and the average of the 2 resting positions was subtracted from the highest value of the 2 maximum opening positions; this difference was defined as the MMO.
HR-QoL was measured by using the EORTC-QLQ-H&N35 questionnaire, a symptom scale, meaning a higher score represents more symptoms. It consisted of 30 items with a 4-point Likert scale (“Not at all,” “A little,” “Quite a bit,” and “Very much”), followed by 5 “yes/no” items. In total, there were 7 multi-item and 11 single-item domains involving aspects of daily functioning. The raw score of each multi-item domain and the score of single-item domains were transformed into a 0-to-100 scale.
A statistical analysis was performed by calculating mean and standard deviation (SD) for continuous variables. Cross-tabulations were made for categorical variables. A chi-squared test was used for categorical outcomes, followed by a post hoc analysis using the adjusted residuals (Z scores) to calculate P values to see which cell accounts for the difference between the groups. Given the small number of patients available for contact, no sample size calculation was performed. Nonnormal distributions were assumed, and differences between groups in continuous outcomes were calculated by using the Kruskal-Wallis test (α=.05). All tests were performed by using a statistical software program (IBM SPSS Statistics, v25; IBM Corp).
Eleven patients (6 ART, 5 Rohner technique) with a reconstructed maxilla were included. All participants had their tumor ablation between April 2010 and January 2016 and were assessed between September 25, 2017, and October 25, 2017. The mean follow-up time was 914 ±662 days, and consequently the closest assessment time for the obturator-wearing participants was 1 year after treatment. Thirteen patients with obturator prostheses (obturator group) met the inclusion criteria at this assessment moment, with a mean follow-up time of 363 ±11 days.
Demographic and treatment details of the reconstructed group are presented in Table 1 . No significant differences were found between the obturator and either of the reconstructed groups regarding sex, TNM stage, or horizontal aspect Brown classification. No differences in radiotherapy administration and mean applied dosage were found between the obturator and ART groups. The follow-up time ( P <.001) and mean surgical time (minutes, P =.001) of the obturator group were significantly shorter than those of the reconstructed groups. Also, the mean age was higher in the obturator group ( P <.001). The number of neck dissections on the ipsilateral side in the obturator group was significantly less ( P <.001).
|Total (N=11)||ART (N=6)||Rohner (N=5)||Obturator (N=13)||NR-R||NR-ART||NR-Rohner||ART-Rohner|
|Age a (mean ±SD)||45.0 ±14.3||47.8 ±16.6||41.6 ±11.9||70.8 ±13.6||<.001||.010||.002||1.000|
|Male (female)||3 (8)||1 (5)||2 (3)||7 (6)||.341||—||—||—|
|T of TNM stage b|
|Days since tumor resection a (mean ±SD)||914 ±662||885 ±187||949 ±1024||363 ±11||<.001||.003||.146||.955|
|Length of surgery a (mean ±SD)||1007.3 ±199.9||1045.7 ±240.8||949.8 ±125.8||197.2 ±103.5||.001||<.001||.005||.849|
|Dose a (mean ±SD)||59.7 ±3.7||59.7 ±3.7||n/a||62.8 ±4.6||.250||—||—||—|
|Ipsilateral neck dissection (%) b||11||6||5||1||<.001||—||—||—|
|Contralateral neck dissection (%) b||2||2||0||0||.108||—||—||—|
|Brown class 39, b|
Functional outcomes are depicted in Table 2 . The dental state of both maxilla and mandible in the obturator group differed significantly from the reconstructed groups ( P <.001 and P =.012). Obturators in the maxillae (of which 2 were implant-supported) and the number of dentate mandibles explained this difference. The number of occlusal units was significantly smaller in the obturator group than that in the reconstructed group ( P =.001). The mean mixing ability index was significantly higher (thus worse) in the obturator group than that in the reconstructed group ( P =.038). The mean MMO in the obturator group was 38.3 ±14.6 mm, 45.7 ±14.6 mm in the ART group, and 47.8 ±3.2 mm in the Rohner group; however, the outcomes among the 3 groups were not significantly different ( P =.315). The mean occlusal force on the nonoperated side of the obturator group was significantly lower than that of the reconstructed group ( P <.001).
|Variable||Nonreconstructed||Reconstructed||P a||P||Post hoc Analysis|
|Dental State||Obturator, N=13||Total, N=11||ART, N=6||Rohner, N=5||NR-R||NR-ART-Rohner||Obturator||ART||Rohner|
|Completely dentate||0||2||2||0||<.001 a||<.001 a||.108||.011||.449|
|Fixed dental prosthesis on implants||0*||8||3||5 ∗||—||—||<.001||.317||<.001|
|Completely dentate||3*||10||5||5||.012 a||.082 a||—||—||—|
|Fixed dental prosthesis on implants||0||1||1||0||—||—||—||—||—|
|Functional outcomes||Mean ±SD||NR-R||KW-test||Obturator-ART||Obturator-Rohner||ART-Rohner|
|Occlusal units||1.62 ±2.8||7.5 ±3.8||5.7 ±4.1||9.6 ±2.2||.001 b||.002 b||.146||.002||.523|
|Mixing Ability Index||27.3 ±5.5||18.2 ±2.4||18.0 ±1.7||18.5 ±3.2||.001 b||.006 b||.025||.038||1.000|
|MMO||38.3 ±14.6||46.7 ±10.6||45.7 ±14.6||47.8 ±3.2||.119 b||.315 b||—||—||—|
|Ipsilateral||126.6 ±135.5||145.4 ±65.0||103.9 ±89.7||162.0 ±55.9||.475 b||.79 b||—||—||—|
|Contralateral||55.5 ±98.4||280.9 ±136.0||283.7 ±193.1||278.7 ±95.0||<.001 b||.005 b||.026||.023||1.000|
|Total||96.8 ±111.7||221.2 ±111.2||204.6 ±137.0||241.1 ±80.8||.020 b||.060 b||—||—||—|
The obturator group reported significantly worse scores than the reconstructed group for the domains “sexuality” ( P =.028), “feeding tube” (11/13 obturator group versus 2/11 in the reconstructed group, P =.001), “weight gain” ( P <.001), “weight loss” ( P =.035), and the overall mean score of the EORTC-QLQ-H&N35 ( P <.001) ( Table 3 ).
|Pain||14.7 ±11.9||12.1 ±11.4||12.5 ±10.2||11.7 ±13.9||.856||—||—||—||—|
|Swallowing||16.7 ±22.8||3.0 ±4.2||4.2 ±4.6||1.7 ±3.7||.273||—||—||—||—|
|Senses||17.9 ±30.0||24.2 ±30.2||38.9 ±34.4||6.7 ±9.1||.128||—||—||—||—|
|Speech||12.8 ±13.5||12.1 ±12.6||13.0 ±10.9||11.1 ±15.7||.915||—||—||—||—|
|Social contact||6.7 ±14.1||4.8 ±6.7||7.8 ±7.8||1.3 ±3.0||.255||—||—||—||—|
|Social eating||20.5 ±24.0||8.3 ±11.2||12.5 ±12.6||3.3 ±7.5||.174||—||—||—||—|
|Sexuality||44.8 ±40.5||3.0 ±6.7||5.6 ±8.6||0.0 ±0.0||.035||.028||.287||.046||1.000|
|Teeth||20.5 ±25.6||12.1 ±22.5||16.7 ±27.9||6.7 ±14.9||.558||—||—||—||—|
|Mouth opening||35.9 ±34.6||18.2 ±31.1||27.8 ±39.0||6.7 ±14.9||.230||—||—||—||—|
|Dry mouth||30.8 ±28.7||24.2 ±26.2||38.9 ±25.1||6.7 ±14.9||.115||—||—||—||—|
|Sticky saliva||30.8 ±28.7||24.2 ±26.2||38.9 ±25.1||6.7 ±14.9||.115||—||—||—||—|
|Coughing||17.9 ±22.0||12.1 ±16.8||11.1 ±17.2||13.3 ±18.3||.817||—||—||—||—|
|Felt ill||12.8 ±21.7||6.1 ±13.5||5.6 ±13.6||6.7 ±14.9||.740||—||—||—||—|
|Pain killers||53.8 ±51.9||27.3 ±46.7||16.7 ±40.8||40.0 ±54.8||.326||—||—||—||—|
|Nutritional supplements||46.2 ±51.9||27.3 ±46.7||50.0 ±54.8||0.0 ±0.0||.161||—||—||—||—|
|Feeding tube||84.6 ±37.6||9.1 ±30.2||0.0 ±0.0||20.0 ±44.7||.001||.001||.002||.049||1.000|
|Weight gain||84.6 ±37.6||0.0 ±0.0||0.0 ±0.0||0.0 ±0.0||.049||<.001||.109||.198||1.000|
|Weight loss||69.2 ±48.0||18.2 ±40.5||16.7 ±40.8||20.0 ±44.7||<.001||.035||.002||.005||1.000|
|Mean overall EORTC H&N35 score||34.5 ±9.8||13.7 ±9.2||17.6 ±6.9||9.0 ±10.2||<.001||<.001||.004||<.001||.430|