INTRODUCTION

Osteonecrosis of the jaw (ONJ) is a disease described fairly recently. After the reported findings by Marx¹, bisphosphonates were considered the etiological agent responsible for the disease, even being called osteonecrosis due to bisphosphonates²-⁵, which is wrong since many factors in addition to these drugs may be implicated in the etiopathogenesis of ONJ¹,⁶,⁷.

One of the hypothesis about ONJ’s development would be the existence of an excess suppression of bone remodeling, which can be produced by bisphosphonates or by other potent anti-resorptives, such as denosumab, a drug that is also involved in ONJ⁸,⁹. Since these drugs act on the entire skeleton, if there is such an excess of oversupression of bone remodelling, one could expect the existence of alterations in both the amount of BMD and bone quality in other locations. Although there are many descriptions of isolated cases or series of this disease in the literature, outlining its clinical characteristics and possible association with different diseases and risk factors¹,^{3 7},¹⁰, we have not found publications that analyze the possible quantitative alterations and/or qualitative bone in patients with ONJ.

Bone mass measurement by dual radiological absorptiometry (DXA) has been sufficiently validated and is accepted as a reliable bone quantification technique by measuring bone mineral density (BMD)¹¹-¹⁴. However, non-invasive bone quality measurement techniques have not been as successful, due to the many aspects that the concept of bone quality encompasses. Despite this, there are currently two techniques that can estimate some aspects of bone quality. On the one hand, trabecular bone score (TBS), associated with DXA, which offers information on bone microstructure¹⁵-²¹ ; and on the other, quantitative ultrasound (QUS), which although it is not known exactly what bone properties it reflects, its measurements have also been related to bone microarchitecture and some mechanical parameters²²-²⁴.

Thus, our research objective has been to study the possible alterations in the amount of bone tissue in locations other than the jaws and that serve as a reference, measured as BMD by bone densitometry (dual radiological absorptiometry, DXA) in the lumbar spine and proximal femur; as well as in bone quality, estimated, on the one hand, by means of TBS and, on the other, using the parameters obtained by QUS, in a population of patients suffering from ONJ and estimating the presence of certain diseases and treatment that affect the bone. We highlight bisphosphonates as a secondary objective, which could participate in its etiopathogenesis. For this, we take as a reference a control group of patients who had received bisphosphonates for at least 5 consecutive years and who continued to take them at the time of the study.

MATERIAL AND METHOD

RESULTS

Table 1 shows the baseline characteristics and lifestyles of the patients included in the study. The patients were similar in age and the proportion of men and women was similar in both groups.

Table 1 Baseline characteristics and lifestyles of patients with ONJ and controls 

Continuous variables are summarized as mean ± standard deviation or as medians with their interquartile intervals (IQR). Categorical variables are expressed as frequency (%). BMI: body mass index: = (weight/size²).

The patients with ONJ was shorter in height, presented a higher BMI and a lower consumption of tobacco and alcohol than controls. We do not observe statistically significant differences in coffee consumption or physical activity in leisure time.

Table 2 shows the comorbidity of patients with ONJ and controls. Patients who suffered ONJ had a greater comorbidity than controls: they showed a higher prevalence of diabetes mellitus, cancer, rheumatoid arthritis, hyperthyroidism, heart disease, arrhythmias, heart failure and hypercholesterolemia. The prevalence of fragility fractures was similar in both groups.

Table 2 Comorbidity of patients with ONJ and controls 

The results are expressed as frequencies (%).

In table 3, we show the consumption of drugs of both patient groups. In line with the existence of greater comorbidity, patients with ONJ had a significantly higher consumption of oral cortico-steroids, oral anticoagulants and hypnotics than controls and, similarly, a greater number of them had received chemotherapy. Inhaled steroid consumption also showed a trend that was very close to reaching the level of significance (p=0.05).

Table 3 Drugs used by patients with ONJ and controls 

The results are expressed as frequencies (%).

In patients with ONJ, the use of bisphosphonates was mostly via i.v. (75%) and much less orally (8.3%). Furthermore, 16.7% of these patients had never taken or received bisphosphonates. Obviously, since it was an inclusion criterion, in the control group 100% had received or taken bisphosphonates, being mostly oral (92.1%).

In table 4, we present the densitometric, ultrasound parameters and TBS values. BMD showed higher values in patients with ONJ in all anatomical locations where it was determined; both in the lumbar spine and in the proximal limb of the femur, the T-score was also higher. We did not obtain statistically significant differences in the values of TBS and Broadband Ultrasound Attenuation (BUA), while patients with ONJ showed higher values of QUI and SOS than controls.

Table 4 Bone parameters related to qualitative and quantitative aspects: BMD measured in the lumbar spine and proximal limb of the femur, TBS measured in the lumbar spine and ultrasound parameters obtained in the calcaneus 

The results are expressed as medians and interquartile intervals (IQR).

According to the densitometric values ((DXA) observed at the time of the study, we obtained that 28% of patients with ONJ had osteoporosis criteria (T-score ≤-2.5 in any of the following locations: L2-L4 , femoral neck or total hip), while these criteria were appreciated in 47.6% of the control patients, without the difference being significant (p=0.06) (Table 5).

Table 5 Percentage of densitometric diagnosis of osteoporosis at the time of the study in both groups 

DISCUSSION

ONJ is a relatively recent disease, the first series having been described about 15 years ago¹,³⁰-³². Its causal mechanism is unknown³,⁶,⁷,³²-³⁵ and many possible risk factors have been related, but without establishing an unequivocal cause-effect with any of them²,³,⁵-⁷,³⁶,³⁷, which can be considered multifactorial. For a long time, treatment with bisphosphonates has been pointed out as a primary etiologic agent of ONJ, to the point that for some time ONJ was called bisphosphonate-induced osteonecrosis¹-⁴,³¹,³²,³⁵,³⁸. This is still considered by many dental specialists who indicate their withdrawal before orodental surgery.

An etiopathogenic hypothesis that was accepted for a long time indicated that bisphosphonates, administered over several years and at high doses, would produce an excessive suppression of remodeling, which would imply bone development with a smaller amount of bone mass and a severe alteration of quality, what came to termed "frozen" bone³⁹. In support of this hypothesis, it has been observed that the vast majority of patients with ONJ, more than 95%, are patients who have suffered from cancer and who in addition to the basic treatment of the process (surgery, radiotherapy) have received polychemotherapy and high-dose intravenous bisphosphonates, usually zoledronate¹,⁶,⁷,³⁰ at the oncological dose, which is 4 mg i.v. every 28 days (52 mg per year), while in the treatment of osteoporosis, the dose used of the same drug is 5 mg i.v. once a year⁴⁰.

However, there are also some disagreements. First, a considerable percentage of patients with ONJ, 16.7%, had never received bisphosphonates.

On the other hand, the greater BMD measured by DXA in all the locations of these patients with respect to the control patients points to a greater general bone quantity of the former versus the latter. We have not found in the main databases similar studies to ours comparing BMD in patients with ONJ with controls in treatment with bisphosphonates. So we do not know whether or not this finding has been corroborated by other authors. We want to highlight the fact that there was no statistically significant difference in the densitometric diagnosis of osteoporosis between both groups. It may seem logical that there should be a higher percentage of osteoporosis diagnosis among the control group, since treatment with oral bisphosphonates (drug of choice for osteoporosis) was the majority. However, we must bear in mind that long-term treatment with this drug increases BMD, and therefore the T-score, causing its values to depart from the densitometric criteria of osteoporosis.

We also wanted to consider bone quality, a much more controversial aspect, since there is no single definitive and non-invasive method that has been accepted as the “gold standard” for estimating bone quality, unlike what happens with densitometry, which is the universally accepted reference for quantity¹¹-¹⁴. One of the recently described methods for estimating bone quality is the socalled trabecular bone score or TBS¹⁶, which basically evaluates the integrity of the vertebral bone trabeculae, reanalyzing the DXA images¹⁷-²¹. The parameters obtained with QUS have also been proposed as possible indicators of bone quality²²-²⁴. In our series, patients with ONJ showed similar values of TBS and BUA to those of control patients, and the SOS and QUI rates were slightly but significantly higher in the former. This leads us to believe the qualitative bone aspects in both groups were similar, and, if anything, never worse in patients with ONJ than in those taking bisphosphonates. Some authors have described low values of ultrasound parameters in patients with ONJ⁴¹, but as with densitometry in ONJ, there are very few studies similar to ours with which to make comparisons. If we take into account that the control patients took bisphosphonates for a long time and considering that bisphosphonates improve bone quantity and quality⁴²,⁴³, the higher values of BMD, TBS, SOS and QUI in patients with ONJ could indicate that the general bone health of these patients is adequate.

Finally, if we add to the previous results the fact that the prevalence of fragility fractures was also similar in both groups (and considering that bisphosphonates decrease the risk of fracture), we have indirect evidence that the general bone structure, both quantitatively and qualitatively, it is the less similar (if not better) in patients with ONJ and patients under treatment with bisphosphonates.

We totally agree that the cause and development of ONJ is multifactorial, as has been published in multiple studies and agreed by consensus¹-³,⁷,¹⁰,³⁶. As we have observed in our series, patients with this disease have greater comorbidity. Therefore, drug use is also significantly higher in this group of patients⁴⁴ than among those studied here: oral corticosteroids, inhaled corticosteroids, oral anticoagulants, hypnotics and having received polychemotherapy. However, if we focus on bisphosphonates, patients with ONJ have a greater use of potent bisphosphonates via the i.v. (75%), which was closely related to the higher incidence of cancer, but not oral bisphosphonates.

One of the limitations of this study is the sample size of the cases, with only 24 patients. This is due to the low incidence of this disease and the difficulty of getting participation in a study of these characteristics of some patients, due to its delicate clinical situation. On the other hand, the choice of the control group could be discussed. We have chosen patients who had been receiving bisphosphonates for at least 5 years, given that it is precisely this fact that is considered a risk factor associated with the appearance of ONJ⁴⁵-⁴⁷, but which have a beneficial effect on bone in general. Finally, we are aware that unvalued local circumstances, such as oral hygiene, or the presence of dentogingival diseases or dental interventions, have a relevant and decisive specific weight in ONJ pathogenesis, although we do not include them because they are localized circumstances that do not affect the bone in general.

CONCLUSION

Our study results indicate that patients who have suffered ONJ do not appear to have worse bone health (in terms of quantity and quality) in general compared to patients who have been receiving bisphosphonates continuously for at least 5 years. In addition, oral bisphosphonates were not the most used drugs among these patients, so we have to pay more attention to more prevalent ones such as corticosteroids, intravenous bisphosphonates, chemotherapy, hypnotics and oral anticoagulants; while taking into account comorbidities, such as cancer, diabetes, rheumatoid arthritis, hypercholesterolemia, heart disease and thyroid conditions.