To date, studies have not agreed on the effects of irradiation on bone mineral density. The aim of our study was to investigate the changes in mandibular bone mineral density after irradiation at various doses with and without surgery.
Materials and Methods
We implemented a descriptive animal experiment. The sample included 16 female Göttingen Minipigs, randomly assigned to 4 groups and irradiated with equivalent doses of 0, 25, 50, and 70 Gy to the mandibular region. At 3 months after irradiation, the mandibular left premolars and molars were removed, and dental implants were placed. Computed tomography scans were taken before and 6 months after irradiation. The measured bone density was related to a bone phantom to calculate the bone mineral density quotient (BMDQ). The outcome variable was the BMDQ. Other study variables were the radiation dose and surgery. Descriptive and univariate analyses were computed, and significance was set at P ≤ .05.
In the left hemimandible, compared with the control group, a significant decrease in BMDQ was observed: 0.01 at 0 Gy, −0.01 at 25 Gy, −0.06 at 50 Gy, and −0.11 at 70 Gy ( P = .023). The right hemimandible compared with the control group also showed a significant decrease in BMDQ: −0.02 at 0 Gy, −0.08 at 25 Gy, −0.09 at 50 Gy, and −0.11 at 70 Gy ( P = .007).
The present study used a large animal model to simulate the tissue reactions induced by various radiation doses in the mandible. We found a significant decrease in the BMDQ after irradiation, but no significant correlation could be found between the irradiation dose and a decrease in the BMDQ.
Radiotherapy (RT) is one of the treatment options for malignancies of the head and neck region. However, RT results in side effects on healthy tissue. The most serious complication in the head and neck region is the development of osteoradionecrosis (ORN). The definition most widely used for ORN has been “irradiated bone, which becomes devitalized and is exposed through the overlying skin or mucosa and does not heal within a period of three months, without tumor recurrence.” Patient factors that will increase the risk of the development of ORN include age, gender, health, nutritional status, and the use of tobacco and alcohol. The treatment of ORN is challenging, because it often involves major surgery and causes high morbidity.
RT-induced ORN of the mandible was described as early as 1930. However, when modern photon therapy with megavolt energy levels is applied in RT, the high-energy doses are absorbed differently by the bone compared with the kilovolt energy doses used previously. Before modern technology, the incidence of ORN in the mandible was reported as high as 22%. Current data have reported an incidence of mandibular ORN after RT of approximately 5%, most likely because of more precise and refined irradiation protocols.
Various hypotheses have been proposed about the pathophysiology of ORN. According to Marx and Marx and Johnson, ORN is caused by a combination of vascular and cellular changes in the mandibular bone. Histologic evaluation has shown acellularity and avascularity of the bone marrow and a nonviable periosteum. This condition develops in irradiated areas, leading to tissue hypoxia. Hypoxic conditions cause direct damage to the bone cells, the osteoclasts and osteoblasts.
The cell damage is initiated by 2 different pathways: the direct effect of radiation on the bone and tissue cells and the effect of free radicals damaging the healthy cells. Tissue repair due to iatrogenic or noniatrogenic trauma will no longer be possible. Tissue breakdown, however, can also be seen spontaneously, without any trauma.
According to Delanian et al and Delanian and Lefaix, 3 phases can be discerned in the tissue reaction to irradiation. This reaction has been termed “radiation-induced fibroatrophy” (RIF). The initial prefibrotic phase includes damage to the endothelial cells. The loss of the endothelial cells results in exposure of the connective tissue cells to normally foreign stimuli, which could trigger fibroblastic activation. After the initial phase, RIF induces tissue growth composed of active fibroblasts (myofibroblasts), senescent fibroblasts (fibrocytes), and extracellular matrix. The combined damage to the endothelial cells and the connective tissue cells, amplified by the action of cytokines, leads to a permanent state of RIF. Finally, remodeling of the extracellular matrix progressively increases the tissue density, leaving friable tissue with poor vascularization and cellularized fibroatrophy. These areas are prone to developing ORN.
How these pathophysiologic processes affect bone quality and density is not clear. To date, studies have not agreed on the changes in bone mineral density (BMD) after RT. It is still unclear whether BMD measurements will aid in assessing the severity of bone damage.
The purpose of the present study was twofold. First, we intended to introduce a large animal model to investigate the BMD changes in the mandibles of Göttingen Minipigs irradiated with various radiation doses. Second, we wished to investigate a possible correlation between BMD changes and the radiation dose, with and without surgical intervention. Our hypothesis was, according to the pathophysiology and findings of Verdonck et al, that we would expect a small increase in the BMD quotient (BMDQ) after irradiation. The specific aim of our study was to develop a tool to measure the quality of the irradiated bone. We tried to more closely emulate the clinical human patient in the animal model by using a split mouth model to be able to compare both sides of the mandible, performing surgery on the left side only. The surgical intervention consisted of extractions and implant placement to simulate clinical reality. The bone density was measured adjacent to the implants to prove the clinical concept of implant-based prostheses in the animal model. Quantitative computed tomography (CT) was used to evaluate the BMD changes.
Materials and Methods
The experiment was performed in accordance with the European Community Guidelines for the Protection of (Laboratory) Animals. The local animal ethical committee approved the study.
The study design was an animal experiment. For the present study, 16 healthy adult 18-month-old female Göttingen Minipigs were used. There were no additional inclusion or exclusion criteria.
The predictor variable in the present study was the irradiation dose group. Four pigs were randomly assigned to each of the 4 research groups stratified by the radiation dose received (0, 25, 50, and 70 Gy, respectively). The primary outcome variable was the BMDQ. The other study variables were radiation dose and surgery (yes or no). The primary predictor variables were radiation dose and surgery. Other variables, such as age and gender, were the same for all the pigs. No other variables were related to the outcome.
Data Collection Methods
For RT planning purposes, a preparatory planning CT scan was taken with the pigs under general anesthesia with endotracheal intubation. A thermoplastic immobilization mask was used for optimal fixation of the minipig's head. A Siemens Sensation Open CT scanner was used for radiologic documentation (Siemens, Erlangen, Germany). Axial slices of 1 mm were taken. The CT scans were performed with a calibrated bone phantom placed in the mouth of the minipig to calculate the relative BMDQ and absolute bone density. At 26 weeks after irradiation, another native CT scan was taken. The procedure was the same as that before irradiation.
The minipigs underwent irradiation under general anesthesia with endotracheal intubation. They were positioned in an individualized thermoplastics mold for reproducible and steady positioning. Irradiation was delivered with a 6-megavolt linear accelerator with 2 opposing lateral beams (Siemens). The entire body of the mandible was irradiated in 2 fractions with an equivalent dose of 0, 25, 50, or 70 Gy (2 fractions of 0, 6.5, 9.7, and 11.8 Gy). To calculate the equivalent dose, an α/β ratio for late-responding tissues was used. We assumed that the α/β ratio was 3. The second fraction was given 48 hours after the first fraction.
At 13 weeks after irradiation, surgery was performed with the minipigs under general anesthesia with endotracheal intubation and mechanical ventilation. Established protocols for animal care were used to guarantee maximum safety.
To simulate a translational clinical situation, the surgical procedure on the mandible involved extraction of all premolars and molars on the left side. Next, 4 submerged BioComp dental implants (diameter 4.0 mm and length 12 mm) were placed in each pig (BioComp Industries BV, Vught, The Netherlands). The pigs received intramuscular antibiotics (amoxicillin) and intramuscular pain medication (buprenorphine), which were continued for 3 days postoperatively.
Using SimPlant 15 (Materialize, Leuven, Belgium), the data from the 2 CT scans were analyzed. To measure the BMD, we used a region of interest (ROI) of 5 to 10 mm 2 . The ROI tool for measuring bone density was placed on 5 different predetermined locations of the mandibular cortex on both sides of the mandible ( Fig 1 ). Cortical bone was deliberately chosen to eliminate measurement errors caused by dental or implant material. Because the cortical bone contains less collagenous tissue than bone marrow, its measurements of BMD have been considered to be more accurate.
Two of us (L.P., G.B.) performed the measurements independently from one another. The calibrated bone phantom had a predetermined density given in g/mL. The BMDQ was calculated by dividing the measurement by the average BMD of the phantom ( Fig 2 ). The BMDQ in the present study was used to quantitatively compare the changes in the BMD. The BMD values represented the degree of mineralization of bone. The BDM is known to be relatively low in immature newly formed bone compared with the mature bone matrix.