If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
OO can be symptomatically managed conservatively with anti-inflammatories, minimally invasive therapies and surgical resection. CT-guided radiofrequency ablation (RFA) was introduced in the early 90s, becoming the most common OO management, with a high success rate.
OO is diagnosed based on clinical symptoms and imaging findings. On radiographs, the “nidus” of OO is visualised as an oval or round radiolucent lesion surrounded by dense osteosclerosis or periosteal reaction. The presence of nidus may be seen histologically containing alternating regions of osteoid and woven bone trabeculae along with osteoclasts along the borders. The interwoven connective tissue also has intense vascularity, with the nidus being clearly demarcated from the lamellar bone.
Thin-sliced CT with a bone window algorithm is the imaging modality of choice to delineate a small nidus, especially in areas where the lesion may be obscured by intense periosteal reaction or reactive sclerosis. It is also an excellent modality to guide percutaneous therapeutic interventions.
Differentiation of OO from other benign skeletal lesions is based on the characteristic radiological features as well as location, size, symptoms, and pathology.
Despite the high success rate of RFA in treating OO, several factors have been described as contributing to osteoid osteoma recurrence or symptom persistence following CT-guided radiofrequency ablation. For instance, Baal et al. reported female gender, young age (≤13 years), and eccentricity index (EI) ≥3 to be significant predictors of symptomatic recurrence.
Nevertheless, to be best of our knowledge, there is no literature currently available assessing the outcomes of OO following RFA, based on the histological presence of a nidus observed on the biopsy sample taken at the time of RFA. Therefore, we aimed to compare the clinical outcomes of osteoid osteoma between the group of patients where a nidus was seen on biopsy at RFA, with those without a visualised nidus on histology. Additionally, we investigated other contributing factors that may affect the outcomes of OO, to reflect our experience as a tertiary orthopaedic oncology centre in managing this condition.
2. Patients and methods
A retrospective review of 88 consecutive patients treated for OO at our tertiary bone tumour centre was conducted. OO was diagnosed with classical radiological findings of a smooth periosteal reaction associated with a nidus visualised on plain X-ray or focal CT scan followed by confirmation with histological samples obtained during radiofrequency ablation. The study period was between November 2005 and March 2015. All cases were discussed at the multidisciplinary meeting (MDT) consisting of orthopaedic oncological surgeons, interventional radiologists, specialist nurses and physicians. Patients were identified from prospectively collated oncology and radiology databases. Patients with histological and radiological findings consistent with OO who were treated with radiofrequency ablation (RFA) were included in this study. Patients with a diagnosis of osteoblastoma (defined as aggressive lesions with a size ≥20 mm with histological confirmation) and those with insufficient clinical records were excluded (Fig. 1). The median follow-up duration was 12.5 months (6–20.8). Patient demographics and lesion characteristics are summarised in Table 1.
Table 1Patient demographics and lesion characteristics.
The outcomes of this study were obtained from the prospectively collected oncology database and clinic letters. Follow-up duration was defined as the time of diagnosis to the last clinical review. Treatment success was defined as complete or partial pain alleviation, while treatment failure was defined as persistent or worsening pain compared to the pain level at presentation. Follow-up examinations and assessments were conducted by orthopaedic surgeons or interventional radiologists six to eight weeks after the procedure. Patients were discharged if they were asymptomatic. Further follow-up clinics were only arranged if patients were still symptomatic or when a new referral was made due to pain recurrence.
All patients had CT at presentation and during percutaneous therapeutic interventions. The histological samples were obtained during the radiofrequency ablation (RFA) therapy prior to the placement of the ablation electrode during the same procedure. All imaging data were available on the picture achieving and communication system (PACS, GE Centricity EnterpriseTM, Version 220.127.116.11, General Electric Healthcare Pty Ltd., Piscataway, New Jersey, USA) and analysed by an experienced musculoskeletal radiologist. Any discrepancy was discussed and resolved in the MDT meeting.
3. Statistical analysis
Statistical analyses were conducted with SPSS Statistics 24 (IBM, Armonk, New York, USA). Median and mean values with ranges were included for continuous variables. Chi-squared test and Fisher's Exact Test were used to test statistical significance for categorical variables. Regression analyses were also performed. A p-value of <0.05 was set to be statistically significant in this study.
4.1 Patient demographics and lesion characteristics
Eighty-eight patients were included in this study. Patients’ mean age was 17.6 years (4-53). There were 63 males (72%) and 25 females (28%). All patients complained of pain during their initial presentation. The median follow-up duration was 12.5 months (IQ 6–20.8).
The mean size of the lesion by the longest dimension was 6.5 mm (2–17 mm). Most lesions were in the diaphyseal region of long bones (58 patients (66%)), followed by the metaphyseal (13 patients (14.8%)), epiphyseal regions (8 patients (9%)) and others (9 patients (10.2%)). Lesions were cortically-based in 71 patients (81%) and intramedullary-based in 17 patients (19%). In 79 patients (90%), the OO was located within the appendicular skeleton, while nine patients (10%) had lesions within the axial skeleton. 67 patients (76%) had lesions in the lower limb, 12 patients (14%) had lesions in the upper limb, seven patients (8%) had spinal lesions and two patients (2%) had a lesion within the pelvis. These are summarised in Table 1.
4.2 Primary outcome
All patients underwent radiofrequency ablation (RFA) as the primary procedure to treat OO. There was better alleviation of pain in the patient group with nidus present on histology compared to those without nidus (p = 0.021, OR 7.4, CI 1.35–41.4). The outcomes of both groups are summarised in the flow chart in Fig. 2. The images of the histological samples with and without nidus are shown in Fig. 3, Fig. 4 respectively.
Sixty-six patients (75%) patients had a nidus on their first biopsy sample. Fifty-five patients (83%) had complete pain resolution and therefore were discharged from follow-up. Of the remaining 11 patients, six (9%) reported no pain alleviation, and five (8%) reported partial pain alleviation after initial RFA. All the six patients without pain alleviation had either a repeat RFA or curettage. Of these, three patients who had curettage or excision had lesions located in the vertebrae, whilst the other three who had repeat RFA had lesions located in the ulnar, metatarsal and proximal femur. Of the patient group with partial pain alleviation, only one had repeat RFA, followed by curettage for a lesion located in the L1 vertebra. The other four patients with partial pain improvement underwent no further interventions. All patients who underwent repeat procedures had nidus in their second biopsy or curettage samples.
In the patient group with no histological nidus (22 patients) demonstrated from the first RFA, 13 patients (59%) demonstrated complete resolution of pain and were discharged after the first follow-up. Of the remaining nine patients, seven (32%) reported no pain improvement, and two (9%) reported partial improvement. Amongst the seven patients who had no pain improvement, six of them underwent a second RFA with complete resolution of pain, while the remaining one patient had to undergo a third RFA before successfully eliminating the pain. Of the seven patients who underwent repeat procedures, only 2 of them had nidus in their second histology. None of the patient group with partial pain alleviation underwent a repeat procedure.
Comparing the group with and without nidus on the first RFA biopsy, the patient group with nidus on biopsy demonstrated less likelihood of requiring a second RFA, with 11% (7 of 66 patients with nidus) compared to 32% (7 of 22 patients without nidus having a second procedure) (p = 0.008, OR 0.092, CI 0.02–1.95)(Table 3). The median duration of pain recurrence following the first procedure with classical OO radiological findings was 34.7 weeks (IQ 25.7–69.9). The median duration between the first and repeat RFA was 52.0 weeks, IQ 33.9–69.0).
Table 2Comparison between the patient cohort with and without pain improvement.
Pain alleviation was significantly better in appendicular lesions compared to axially located lesions (OR = 31.7, CI 2.8–359, p = 0.005). There was no difference in outcomes in term of the gender (p = 0.131). The size of the lesion also did not appear to affect the clinical outcomes in our study (p = 0.38). There was no difference in clinical improvement comparing cortical or intramedullary lesion (p = 0.776) and location of lesions (epiphysis, metaphysis or diaphysis). These are summarised in Table 2.
OO is a common, benign, osteogenic tumour with the potential to resolve spontaneously after several years. Patients usually undergo conservative medical treatment initially. Nevertheless, the vast majority of patients may continue to report ongoing pain with the need for long-term analgesia.
Therefore, other treatment modalities have evolved to treat this condition. OO treatment involves removing or destroying the nidus, traditionally with surgical interventions such as curettage, en-bloc excision, or percutaneous modalities.
However, to the best of our knowledge, there is currently no literature available describing the outcomes of radiofrequency ablation to treat OO based on the presence of nidus on the histological samples taken during RFA or biopsy. Therefore, we aimed to conduct this retrospective study to compare the clinical outcomes of OO between the patient cohort with the presence of nidus on biopsy samples with those without nidus following RFA.
In this study, we achieved a high overall success rate in the management of OO, with 85% of patients reporting pain improvement following the first RFA. These findings are similar to previously reported literature with a success rate of 76%–100%.
The patient group with nidus on histology sample demonstrated better pain resolution compared to the group without nidus (p = 0.021). Understandably, patients with nidus on histology are less likely to require a repeat intervention compared to the group without nidus (p = 0.008). In terms of radiological assessment of nidus, Rehnitz et al. had previously reported no correlation between the CT or MRI patterns of OO morphology with the clinical outcomes.
Radiographs or CT scans were also used to identify of the nidus before and after en-bloc removal, but this had the disadvantages of sizeable surgical excision of a large part of the sclerotic bone. The procedure may involve bone grafting or internal fixation due to the size of the bone defect, often associated with a more extended hospital stay, delayed recovery and high morbidity. Hence, RFA is now the mainstay of treatment for OO replacing surgical excision. However, it is not uncommon for surgical interventions to be performed on spinal lesions due to the close proximity of neural structures and the potential risk of thermal injury from a percutaneous RFA (9,10).
The patient demographics in the present study is similar to previous literature with incidence in young male patients, with a mean age of 17.6 years (4-53).
The majority of OO lesions in our study were found in the diaphyseal region (66%), followed by metaphyseal (15%), epiphyseal regions (9%) and others (10%). 81% of lesions involved the cortical region and 19% intramedullary. There was no significant difference in the outcomes between cortical vs intramedullary lesions in our study (p = 0.776). The outcomes of this study are consistent with a study by Vanderschueren et al. reporting that tumour location within the bone regions (intracortical or extracortical) demonstrated no significant association with RFA outcomes.
Our study reported better outcomes in appendicular compared to axial locations OO (p = 0.005). Additionally, our subgroup analysis showed superior pain improvement when lesions were located within the extremity or pelvis compared to spinal lesions (p = 0.001). It has been reported that RFA of spinal lesions remains a challenge due to the complex anatomical site with inconsistent results.
Four of the seven spinal cases in our study had to undergo open excision and curettage following the initial RFA due to persistent or recurrent symptoms.
This study had several limitations. Firstly, the study was done retrospectively by collecting data and clinical details from pre-existing oncology and radiology databases. As the outcomes of this study were assessed according to clinic letters and clinical entry from follow-up appointments focusing on the level of pain and symptom improvement, we were not able to use an objective scoring system such as the VAS scoring system to evaluate patients’ pain level, which may be helpful to quantify the level of pain improvement. The quality of data was therefore limited by the data available in the clinical records. Additionally, the biopsy technique was not standardised and there was no control group, with several interventionists performing the procedure possibly introducing biases in technical factors. Additionally, due to marked perilesional sclerosis in OO, special needles (Bonopty) are commonly used, which are of thinner gauge (13G and 15G) than the widely used bone biopsy needles (8G and 11G). This could be one of the causes for the biopsy specimen to be crushed, making the identification of nidus difficult for pathologists in some cases. However, despite the differences in the size of RFA needles being used, satisfactory ablation can still be achieved if OO is within the ablation field. Another possible explanation for better results could be that in patients where biopsy has shown nidus, majority of the nidus was removed as part of the biopsy prior to ablation. Furthermore, due to its retrospective nature, the time to symptom recurrence may also be overestimated as patients may not present immediately to the outpatient clinic appointment, hence over-estimating their symptoms-free period. However, designing a prospective study would take some time to include enough patients to achieve statistical significance due to the rarity of the condition. Having emphasised the importance of nidus in biopsy samples, we did not include a radiologically based analysis of success following the first RFA in this study due to the lack of data. Nevertheless, we believe that the main principle of positioning the RFA tip at or adjacent to the nidus to obtain biopsy has been emphasised by this study to achieve better pain alleviation.
Based on the results of this study, we recommend interventionists and orthopaedic oncologists to target nidus during CT-guided biopsy to improve clinical outcomes. We have observed that the histological presence of nidus also correlates with better pain alleviation, consistent with the advice to interventionists to target the nidus under radiological guidance. Further, patients should be counselled about the higher chance of persistent or recurrent symptoms in axial and, in particular, spinally located lesions, compared to appendicular lesions, with the possible inherent need for repeat interventions.
The patient group with nidus on histology samples following RFA demonstrate better pain relief than those without nidus. The presence of nidus in the biopsy significantly decreases the need for repeat interventions. Patients and clinicians may be advised that the alleviation of pain following RFA is better in patients with appendicular lesions than spinal or axially located lesions.
A Adlan: Writing – original draft, formal analysis. M Alaqeel: Writing - Review & Editing. S Evans: Supervision, Writing - Review & Editing, Validation. V Sumathi: Resources. M Davies: Supervision. R Botchu: Methodology, Conceptualization, Supervision, Validation, Writing - Review & Editing.
Declaration of competing interest
The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Co-author also serving as the editorial board member for JCOT, Dr RB.