1 Introduction

Osteoid Osteomas (OO) are benign, osteoblastic bone tumours with a distinctive osteoid nidus.¹ Their incidence is not well reported in tarsal bones with most studies available reporting on small case series.^2,3 Typically, they present in males with a three to one preponderance and are often diagnosed in patient's twenties or thirties.^1,3–5 Despite their small dimensions, OO presents with disproportionate functional impairment. Classical features include severe nocturnal pain, responsive to non-steroidal anti-inflammatories (NSAIDs).^1,6 In addition, OO are often associated with swelling, and local tenderness, hypothesised to be associated with high vascularity, high prostaglandin concentration (PG3) and presence of nerve fibres within the tumour itself.^6–8

The natural progression of the lesions can vary depending on size and location. In some instances, the tumour may regress spontaneously over 6–15 years. Alternatively, it can grow and cause damage to the surrounding bone.⁸

OO have a pathognomic radiographic presentation, best depicted on CT. An intracortical nidus is described with variable degrees of calcification, as well as a reactive zone with cortical thickening, sclerosis, and bone marrow edema.⁹ Depending on their location in the bone itself (cancellous, intracortical or subperiosteal), they can be associated with periosteal reaction. The latter is rare in foot and ankle OO, given their predisposition for intracortical locations.^1,2

The conundrum can develop due to the edema's signal intensity on MRI, which can obscure the nidus, and deflect from an accurate diagnosis. Most clinicians use multimodal imaging approaches, to narrow their diagnosis excluding inflammatory, infectious, malignant, or traumatic diagnoses.^1–3,6,9

Conservative or medical management may improve symptoms, but the current definitive treatment choices include resection or percutaneous ablation with radiofrequency (RFA) or laser.^1,2

The aim of this case series is to highlight OO as a relevant, yet challenging differential diagnosis for foot and ankle pain, in both traumatic and atraumatic circumstances.

2 Methods

This was a single centre retrospective review of osteoid osteoma in the foot, following a three-year review of a single surgeon's practice. We collected epidemiological data, imaging modalities, time to diagnosis, time to treatment (weeks), and time to final follow up. Imaging modalities were recorded, specifically: the number and types of imaging such as computed tomography (CT) or magnetic resonance imaging (MRI) performed. We documented the number of physicians the patient encountered before a diagnosis was made. We also recorded the number of alternative diagnoses and any complications relating to treatment.

We collected subjective outcome measures, using the FAOS and VAS score. The FAOS (0–100) is a validated outcome measure with five subdomains including symptoms, pain, activity, sports and quality of life. A high score indicates good function, whereas a low score indicates poor function.¹⁰

The inclusion criteria consisted of patients with OOs in the foot; patients who had OOs in the tibia or other long bones were excluded. Informed consent was obtained following the clinical research ethics committee guidelines. Statistical analysis with respect to the subjective outcomes and the epidemiological data was carried out however the case series was underpowered to detect any significant results.

3 Results

3.3

3.3 Imaging

All patients presented with pathognomonic central sclerotic nidus and reactive zone on CT (Table 2). Fig. 1 illustrates a classical talus OO.

Table 2 Imaging modalities and findings.

	Imaging modalities	Bone	Nidus on CT	Periosteal Reaction	Location	Proposed Differential	Size (mm)
1.	XR, CT, MR, MRI with contrast	Talus head at talonavicular joint	Yes	No	Subperiosteal	Osteoblastoma	15
2.	XR, MRI, CT	Talus neck	Yes	No	Intracortical	Calcaneal Fracture	6
3.	XR, MRI, MRI, CT	Intermediate Cuneiform	Yes	No	Subperiosteal	AVN, Osteomyelitis	6
4.	XR, MRI, MRI, CT, Bone scan	Talus neck	Yes	No	Subperiosteal	Stress fracture, Osteomyelitis, PVNS	9

Close

Fig. 1 Patient 1. Axial and sagittal CT views of an osteoid osteoma, on the head of the talus. Note the sclerotic nidus and radiolucent surrounding area.

Export to PPT

Although MRI was the most used imaging modality in our series, it was not the most accurate with three out of the four lesions not diagnosed on initial MRI. Fig. 2. Illustrates patient 1. OO on contrast MRI.

Close

Fig. 2 Sagittal and Axial cuts of T1 weighted MRI with Gadolinium contrast. Note the hypointense nidus and an enhanced peripheral rim, as well as peripheral bone edema.

Export to PPT

Patient 4 due to unusual demographic, despite having a classical radiological presentation, underwent a triple phase bone scan (Fig. 3) to rule out sinister pathology. It revealed marked increased radioisotope uptake in the early dynamic phase, blood pool phase and delayed phase imaging, highly suspicious for osteomyelitis. However, the patient's clinical presentation, laboratory results and history were not consistent with infectious process, which resulted in prompt surgical resection, for both diagnostic and therapeutic purposes.

Close

Fig. 3 Triple phase bone scan images.

Export to PPT

3.4

3.4 Management and follow up

Definitive diagnosis was established on imaging for three out of the four patients, with one being established histologically. The average time to diagnosis was 67 weeks (±60). The patients were managed as per Table 3. One patient is awaiting radiofrequency ablation (RFA).

Table 3 Management and follow up. Assessment of patients complains with FAOS and VAS scores.

	Management	Histology	Period from symptoms to procedure (Weeks)	Complication	Follow up period (Months)	FAOS Pre Intervention	FAOS Post Intervention	VASPre Intervention	VASPost Intervention
1.	Resection	Yes	84	No	37	39%	100%	8/10	1/10
2.	Awaiting RFA	–	–	No		30%	–	9/10
3.	1) RFA2) Resection	Yes	40	Incomplete ablation or Recurrence	7	49%	96%	7/10	1/10
4.	Resection	Yes	14	No	9	61%	94%	8/10	1/10

The mean time from symptom onset to treatment was 46 weeks (±35). The lesions were resected en bloc and sent for histopathology analysis and microbiology. Fig. 4 depicts an intraoperative picture, and Fig. 5 the histological sample.

Close

Fig. 4 Intra-operative picture of a talar head osteoid osteoma. The lesion was removed en bloc, and the bony defect was filled with bone graft.

Export to PPT

Close

Fig. 5 Osteoid Osteoma sample a) x2 magnification. Green: Rim of mature bony trabeculae Yellow: Centre of the lesion with anastomosing trabeculae of woven bone, embedded in a vascular stroma; b) x10 magnification. Yellow: Central nidus with anastomosing trabeculae of woven bone; Blue: osteoblasts set in a vascular stroma; Green: giant cells. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Export to PPT

The symptoms resolved following the initial therapy, with exception of Patient 3, which referred persistent nocturnal pain. A CT conducted post-RFA, showed incomplete ablation or recurrent OO adjacent to the ablated area. The lesion was resected, with subsequent resolution of symptoms.

The mean follow-up time was 17.6 (±16) months. FAOS and VAS scores were recorded. Mean pre intervention FAOS of 45%, with a mean post FAOS of 97%. The mean VAS pre intervention improved from eight to a mean of one.

4 Discussion

Osteoid osteoma of the foot is a challenging diagnosis, which often results in delayed diagnosis. Due to similarities to other pathologies and a broad differential, it requires several imaging modalities, with CT being the most specific. Subsequent histopathology examination is also warranted.

Epidemiologically, OO presents in young, male patients, with 80–95% occurring under the age of 30 and 40 respectively.^3,5 Our sample had an equal sex distribution, but a dispersed age distribution, from 14 to 62 years of age. OO is a rarity in the latter demographic, which contributed to the prolonged investigation period, despite pathognomic imaging.

Osteoid osteomas are generally small, with a subcentimetric nidus, and although in theory OO can affect any bone, long bones are the most affected (>50%).^5,11 Nevertheless, tarsal bone OO, mainly the talus, has been described in small case series.^3,12,13 Similarly, our sample describes three talar lesions and one in the intermediate cuneiform. Patients were referred from GPs, or orthopaedic surgeons. Before the diagnosis was achieved, two to four professionals had assessed these patients. This underlines the significance of an accurate history, as classic symptoms are often the first indication towards the correct diagnosis.

The clinical hallmark of OO is intense night pain responsive to NSAIDS.^1,6 Night pain, swelling and local tenderness was present in 100% of our cases, however only 50% were responsive to NSAIDs. The remaining patients required opioid derivatives. Two of the cases reported a trauma-inducing event. Although no clear reported connection of these tumours to trauma, Vancamp et al. hypothesised that the nidus can be formed reactively post trauma.¹⁴

The average symptom duration at presentation was 41,5 weeks (±47), which is below the described average.^3,9 Our average time to diagnosis from the symptom onset of 67 weeks (±65), is consistent with the literature's 56–144 weeks.¹ The differential diagnoses encountered amongst our cases, prior to treatment were osteomyelitis, osteoblastoma, stress fracture, calcaneal fracture, AVN, and PVNS.

Foot OO poses a particular challenge to diagnose based on a single imaging modality, due to minute anatomy. Even with a multimodal approach, it can be challenging, hence the involvement of a radiology musculoskeletal (MSK) specialist can streamline the diagnosis workup and therapeutic planning.

CT is the gold standard modality for detecting OO, with the depiction of the pathognomic low density lesion with a central dense nidus.^1,8 A CT radiological nidus was detected on 100% of cases. MRI sensitivity is multifactorial (use of contrast, axial cuts, tumour location, and inflammatory response),¹⁵ and it can often be misleading due to perifocal edema obscuring the typical nidus.¹⁶ Based on an initial MRI, only one of four OO was diagnosed. Upon repeat imaging and comparison to previous studies three out of four were diagnosed. Bone scan is 100% sensitive to OO but it lacks specificity. In an early phase it can be useful, due to early uptake in the blood pool phase with subsequent double density appearance. It is also relevant when investigating recurrences or incomplete resections/RFA.

The current treatments of choice are percutaneous, with RFA, laser ablation; or resection, either open or arthroscopic. While a growing trend towards RFA is noted,⁵ in the foot due to neurovascular bundle and chondral proximity, open resection either en bloc or curettage is often preferred.^2,6 Open approach avoids the technical difficulties associated with arthroscopic debridement, provides optimal visualisation, and an opportunity for histopathological analysis.¹⁷ In 75% of our cases, it was ultimately surgical intervention that eliminated the patient's pain.

Our cohort was followed post-operatively on average 17,6 months (±16). There was one case of either recurrence or incomplete RFA, which was noted due to persistence of pain post intervention. The lesion was later resected. Recurrence can happen up to 15% following RFA, and its success is said to be dependant on the accuracy of the imaging.² The improvement of the patient's symptoms was quantified by comparing the subjective pre and post-operative FAOS and VAS. The mean pre-intervention FAOS score was 45% and the mean post- FAOS score was 97%. The mean pre-intervention VAS was 8 and post-intervention 1.¹⁰

An inherent limitation of case series is the generalizability of their findings. One of our limitations is that a patient is still awaiting RFA treatment. Another limitation is the small number in the series, although other series have comparable numbers. Despite several of our discussion points being echoed in previous research, novel findings from this series may not be transferable.

5 Conclusion

Osteoid osteomas can be challenging to diagnose as depicted in this series. Although multiple imaging modalities are often required to achieve a diagnosis, CT is the gold standard. MRI as a first line can be misleading. Treatment options can include radiological/minimally invasive or surgical resection. It is important to keep this differential in mind when evaluating atraumatic foot pain or foot pain without a serious history of trauma.

Data statement

Due to sensitive nature, the raw data is confidential, and unavailable to post.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Institutional ethical committee approval

Permission from the ethical committee to use information from University Hospital is not requires, as per institutional policy, for this research has a limited sample. Data was anonymized and protected under GDPR guidelines of the University. Informed consent was obtained for each participant.

Authors contribution

1) Original draft and editing, Data Collection and curation.

2) Data collection, Picture curation, Writing: editing.

3) Writing: review and editing.

4) Data Collection.

5) Radiological validation and curation.

6) Supervision, writing review