Introduction

Osteoid osteomas are benign, painful osteoblastic tumors that arise most commonly in young men [1]. Severe pain typically appears at night at the site of the lesion; however, this may be constant at any time during the day [1,2,3]. Radiographically, osteoid osteomas appear as a small (typically < 20 mm) cortically based radiolucent nidus with surrounding dense sclerosis although less commonly may be intra-medullary [1,2,3]. The standard of care for the treatment of osteoid osteomas is image-guided ablation. The technique was first described in 1992 by Rosenthal, and since, numerous prospective and retrospective studies have demonstrated its superiority to other invasive methods [4, 5]. Surgical excision is an alternative with perhaps greater success rates, but is less commonly performed given its invasive nature [4,5,6].

While there have been numerous prospective and retrospective studies on percutaneous ablation for osteoid osteomas, there has yet to be a comparative analysis between pediatric and adult patients in a single study [2, 4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21]. Outcomes for adults and youth are favorable, though questions remain regarding differences between these two cohorts.

This study compares the technical success, complications, and primary and secondary clinical successes in pediatric and adult patients following radiofrequency ablation of osteoid osteomas.

Materials and Methods

Patient Selection

This study was conducted with institutional review board approval and complied with the Health Insurance Portability and Accountability Act. Informed consent was not required for this retrospective study. A retrospective search of the Electronic Medical Record Search (EMERSE; Ann Arbor, MI) database was performed from October 2002 to July 2017 (177 months) using the terms “osteoid osteoma” and “ablation.” This yielded 177 patients.

Inclusion/Exclusion Criteria

All patients suspected to have osteoid osteomas were considered for inclusion (N = 177). Medical records were reviewed to reveal 92 (52.0%) patients who underwent percutaneous radiofrequency ablation of an osteoid osteoma, and were therefore included in the study cohort. Patients who underwent cryoablation of an osteoid osteoma (n = 3) were not included given small sample size.

Patient Demographics

Patient characteristics are given in Table 1. Of the 92 patients, there were 54 (58.7%) pediatric and 38 (41.3%) adult patients. There were 28 (51.9%) males and 26 (48.1%) female pediatric patients with a mean age of 12.9 ± 3.3 (range 4–17 years). As for the adult cohort, there were 31 (81.6%) males and 7 (18.4%) females with a mean age of 24.1 ± 6.5 (range 18–42 years). There were significantly more males than females in the adult group (P = 0.004). Nearly all patients presented with chronic pain at the lesion site, and 43 (75.4%) pediatric and 34 (89.5%) adult patients complained of night pain (P = 0.251). Oral analgesics were effective in 94.4% of pediatric and 92.1% of adult patients (P = 0.688).

Table 1 Patient characteristics
Full size table

Imaging

Advanced imaging after initial evaluation with plain film radiograph was completed via computed tomography, magnetic resonance imaging, or bone scintigraphy. Plain films were acquired prior to ablation in all (100%) patients. Additional imaging modalities, including computed tomography (pediatric [n = 49, 90.7%]; adult [n = 34, 89.5%]), magnetic resonance imaging (pediatric [n = 15, 27.8%]; adult [n = 18, 47.4%]), and bone scans (pediatric [n = 2, 3.7%]; adult [n = 1, 2.6%])were not statistically different between pediatric and adult patients (P = 0.711).

Medical records and imaging were reviewed for size of the radiolucent nidus, location, and whether the lesion was intra-cortical or trabecular and intra- or extra-articular. Follow-up imaging modality was recorded and reviewed to determine whether the ablation was complete. Time to first follow-up imaging was recorded. If a patient underwent reintervention during the study period, repeat imaging was reviewed for evidence of new or recurrent lesion.

Pain

Visual analog scale (VAS) for numeric pain scores (0–10) were recorded at presentation and initial follow-up. Whether patients responded to home analgesics was also recorded. Visual analog scale pain score at reintervention was also recorded.

Technique

For all patients, general anesthesia was utilized in addition to local lidocaine at the biopsy and ablation sites. An 11-gauge bone biopsy needle was advanced into the osteoid osteoma under image guidance. In cases with dense perilesional sclerosis, the tumor was reached through a drill advanced over an appropriate guidewire. Then, a 14-gauge electrode was introduced into the lesion nidus through a coaxial system. The electrode was connected to the radiofrequency generator, and the tip temperature was increased to 90 °C. One cycle of radiofrequency ablation was defined as ablation at 90 °C for a minimum of 5 min followed by passive cooling to approximately 40 °C.

Number of ablation cycles during the same procedure, overall ablation time (min), number of times the needle was repositioned, and biopsy results were also recorded. The size of the ablation area was chosen to encompass the radiolucent nidus and a 5-mm margin. Thermocouples and spinal cord protection techniques were utilized for ablations around the spinal cord [18, 20, 22].

Variables and Outcomes

Technical success, reintervention rate, primary and secondary clinical successes, and total follow-up time (days) were recorded. Technical success was defined as successful placement of the ablation probe tip into the radiolucent nidus followed by ablation with intra-procedural imaging showing adequate coverage of the nidus by the ablation zone. Primary clinical success was defined as visual analog scale of 1 or less after a single ablation. Secondary clinical success was defined as visual analog scale of 1 or less following repeated ablations. Reintervention was further delineated between disease recurrence and incomplete treatment. Disease recurrence was defined as significant increase in visual analog scale after complete ablation or as determined by imaging. Incomplete treatment was defined as a significant increase in visual analog scale and evidence of incomplete ablation on follow-up imaging.

Statistical Analysis

Mean, median, standard deviation, standard error, and interquartile ranges were calculated using spreadsheet software (Excel, Microsoft; Redmond, WA). P values were calculated using the Student’s t test (or Mann–Whitney) for continuous variables, and Chi-square test (or Fischer’s exact test) for categorical variables. P values < 0.05 were considered statistically significant. Kaplan–Meier curve and box plots were generated using SPSS 20.0 software (IBM; North Castle, NY).

Results

Imaging

Pre-procedural imaging is given in Table 1. Pre-procedural imaging was acquired in all patients with computed tomography (pediatric [n = 49, 90.7%]; adult [n = 34, 89.5%]), magnetic resonance imaging (pediatric [n = 15, 27.8%]; adult [n = 18, 47.4%]), and nuclear imaging (pediatric [n = 2, 3.7%]; adult [n = 1, 2.6%]). Some patients had multiple imaging modalities preoperatively. There were no statistical differences in the utilization of imaging modalities between pediatric and adult patients in the pre-procedural setting (P = 0.711).

Post-ablation imaging was acquired in 36 (66.7%) pediatric and 25 (65.8%) adult patients (P = 0.466). Of the pediatric patients, 31 (57.4%) had follow-up plain film radiography and 5 (9.3%) pediatric patients had follow-up computed tomography. For adult follow-up imaging, 22 (57.9%) had plain film radiography, 1 (2.6%) underwent computed tomography, and 2 (5.3%) underwent magnetic resonance imaging.

Lesion Characteristics

Lesion metrics and locations are presented in Table 1 and Fig. 1. Mean lesion size for pediatric and adult patients was 9.6 ± 5.8 mm (range 3–32 mm) and 9.0 ± 4.3 mm (range 3–64 mm), respectively. No statistical difference in lesion size was noted between pediatric and adult patients (P = 0.700). Osteoid osteomas were localized to the femur (pediatric [n = 22, 40.7%]; adult [n = 13, 34.2%]), tibia (pediatric [n = 21, 38.9%]; adult [n = 10, 26.3%]), humerus (pediatric [n = 5, 9.3%]; adult [n = 2, 5.3%]), foot (pediatric [n = 3; 5.6%]; adult [n = 4, 10.5%]), fibula (pediatric [n = 1, 1.9%]; adult [n = 4, 10.5%]), acetabulum (adult [n = 3, 7.9%]), spine (adult [n = 1, 2.6%]), scapula (pediatric [n = 1, 1.9%]), radius (adult [n = 1, 2.6%]), and ulna (pediatric [n = 1, 1.9%]). There was no statistical significance in the trend of lesion locations between pediatric and adult patients; however, there was a trend toward a broader distribution of lesions in the adult cases (P = 0.147). Nearly all lesions were intra-cortical, aside from 3 pediatric patients with intra-medullary osteoid osteomas, two located in the proximal femur and one in the mid-femur. Three osteoid osteomas were intra-articular, all located at the cortex of the femoral head within the hip joint.

Fig. 1
figure 1

Distribution of pediatric (P) and adult (A) osteoid osteomas. Lesion sites include: femur, tibia, fibula, foot, humerus, radius, ulna, acetabulum, scapula, and T4 vertebrae. Note diagram does not reflect sidedness

Full size image

Ablation Techniques

Ablation techniques are given in Table 2. The mean number of ablation cycles for pediatric patients was 1.2 ± 0.5, and 1.1 ± 0.3 for adults. Mean total ablation time for pediatric patients was 6.8 ± 3.3, and 6.5 ± 1.6 min for adults. The number of needle positions was similar between pediatric and adult patients, with 1.2 ± 0.53 and 1.1 ± 0.23 mean positions per case, respectively.

Table 2 Percutaneous osteoid osteoma radiofrequency ablation techniques
Full size table

Biopsy

Biopsy results are given in Table 3. Forty-eight (84.2%) pediatric and 36 (94.7%) adult patients underwent pre-procedural biopsy. Of those biopsy specimens, 28 (58.3%) pediatric and 15 (41.7%) adult were confirmed cases of osteoid osteomas, while 20 (41.7%) pediatric and 21 (58.3%) adult cases were non-diagnostic or negative for osteoid osteoma. Reintervention rates for patients with pathologic evidence of osteoid osteoma and non-diagnostic or negative samples were 6.7 and 14.3%, respectively (P = 0.304).

Table 3 Percutaneous osteoid osteoma ablation follow-up and outcomes
Full size table

Outcomes

Outcomes are presented in Table 3 and Fig. 2. Overall, there were no significant differences between pediatric and adult outcomes. Mean time between ablation and first follow-up imaging was 23.5 ± 17.5 days for pediatric patients and 26.1 ± 18.4 days for adult patients. Total mean follow-up was 95.2 ± 58.7 months for pediatric and 90.0 ± 61.6 months for adult patients. Overall technical success was achieved in 91 cases (98.9%), 54 (100%) pediatric, and 37 (97.4%) adults. The one technical failure in an adult patient was due to ablation probe misplacement.

Fig. 2
figure 2

Primary success rate following initial percutaneous ablation of osteoid osteoma. No significant difference was observed at any point at a 95% confidence interval

Full size image

In total, there were 8 reinterventions during the study period: 5 (9.3%) pediatric and 3 (7.9%) adult patients (P = 0.668). All reinterventions were performed for pain recurrence. Of the reinterventions, 2 (1 pediatric, 1 adult) were due to incomplete treatment and 6 (4 pediatric, 2 adult) were classified as recurrent osteoid osteoma. Median (interquartile range) VAS pain score pre- and post-ablation was 8 [8, 9] and 0 [0–1] for pediatric cases and 6 [6,7,8] and 0 [0–1] for adult cases, respectively. 91.3% of patients had primary clinical success at initial follow-up, with similar rates in pediatric (90.7%) and adult (92.1%) patients (P = 0.880). The majority of pediatric (75.9%) and adult (78.9%) patients experienced complete resolution of symptoms at initial follow-up. Primary clinical success was not statistically different at any point between groups with an overall mean of 31.9 ± 53.1 months (Fig. 2). Additionally, there was a trend toward earlier reintervention in pediatric patients; however, this did not reach statistical significance at a 95% confidence interval (Fig. 2). Secondary success was achieved in all 8 patients requiring reintervention (5 pediatric and 3 adult cases).

Complications

There were two Society of Interventional Radiology Class D complications, including a man who sustained a thigh abscess following radiofrequency ablation of a proximal femoral osteoid osteoma 3 days following the procedure [23]. He fully recovered after incision and drainage followed by a course of antibiotics. Additionally, one patient had rapid onset of flash pulmonary edema following the procedure. The patient survived following 13 days of ventilation and intensive care.

Discussion

Our results confirm that percutaneous ablation is the treatment of choice for osteoid osteomas, and suggests that few differences exist between pediatric and adult patients with respect to clinical presentation, follow-up, and outcomes. Our overall technical and primary clinical success was 98.9 and 91.3%, respectively. These rates are in line with prior publications including a recent review of 1772 patients [2, 5,6,7,8,9,10,11,12,13,14, 16,17,18,19,20,21, 24]. Additionally, our rates of complete pain relief following first radiofrequency ablation (77.2%) are also consistent with prior studies [8, 11]. Long-term outcomes are also favorable with sustained primary clinical success above 90% after 18 months (Fig. 2) and 100% secondary clinical success (Table 3). There was a trend toward earlier reintervention in children; however, this observation did not reach statistical significance at a 95% confidence interval.

Of the 8 (8.7%) patients (5 pediatric, 3 adult) who required reintervention for recurrence of symptoms, the majority (6/8) were recurrent osteoid osteoma rather than incomplete treatment. This suggests that most recurrences were not due to incomplete treatment and that the procedure was likely technically satisfactory in these cases. Moreover, this supports the need for at least one follow-up computed tomography or magnetic resonance imaging at 12 months to monitor for recurrence as suggested by some authors [8].

While nearly all patients (91.6%) underwent percutaneous biopsy prior to ablation, only 51.7% of samples were diagnostic for osteoid osteoma consistent with prior publications [8, 11, 12]. While there were relatively few reinterventions, there was no significant difference between reintervention rates of patients without and with pathologic evidence of osteoid osteoma. As such, it has been argued that biopsy should only be mandated in cases where radiographic and clinical evidences are not diagnostic for osteoid osteoma.

Interestingly, there were marginally higher rates for the use of computed tomography in children (9.3%) compared to adults (2.6%) for advanced diagnostic follow-up imaging. Given the susceptibility of this cohort to long-term radiation exposure, magnetic resonance imaging may be underutilized. It is critical to educate patients and their caretakers about the risks and benefits of computed tomography in the evaluation of bone lesions that are clinically suspicious for osteoid osteoma.

Though we aimed to adhere to proposed osteoid osteoma reporting guidelines, there were several limitations to this study including its retrospective nature [8]. Additionally, different cohorts of operators generally performed adult ablation procedures versus pediatric ablation procedures, so variability in imaging and ablation parameters may be in part related to individual operator preferences. Also, pre- and postoperative pain scores in younger pediatric patients were sometimes reflective of parental assessment of pain score due to age.

Conclusion

Radiofrequency ablation of osteoid osteomas is safe and viable with excellent technical and clinical success rates regardless of age.