Osteoid osteoma of the foot

Osteoid osteomas constitute benign bone tumors with males being affected three times more often than females [1]. The typical clinical presentation consists of pain—especially at night—that is promptly but only for a short time relieved by salicylates and other nonsteroidal anti-inflammatory drugs [2]. Moreover, local tenderness and swelling may be present, potentially caused by the hypervascularity of the tumour itself and the production of prostaglandins [3]. At radiography, an intracortical nidus with a strongly mineralized centre and accompanying cortical thickening and sclerosis may be found [4]. Depending on the location of the nidus on radiograph, three types can be distinguished: cortical, medullary/cancellous and subperiosteal [5]. Their clinical presentation may be different from conventional osteoid osteomas, including symptoms resembling arthritis, such as joint swelling, stiffness, tenderness and effusion [3, 6]. Osteoid osteomas of the foot often cause a diagnostic dilemma, due to their rarity, varying and atypical symptoms and uncommon radiological appearance [7, 8]. They account for 19.4% of all benign bone tumours in this region, with a particular predilection for talus and calcaneus [9].

Due to their rarity, there are only few case series on osteoid osteomas of the foot region, with limited patient numbers [10]. Over a 20-year period, Houdek et al. were able to include 13 patients with foot and ankle osteoid osteomas [11]. In the most recent study on osteoid osteomas of the foot and ankle region, Gurkan et al. reported on 9 patients [7]. In a systematic review by Jordan et al., 68.1% of included publications were case reports [12]. In this field, no multicentre study has been published yet on a group of over 30 patients with a uniform observational study protocol.

The aim of the present study was to 1) analyze the epidemiological, clinical and radiological findings of 37 patients with osteoid osteomas of the foot, and to 2) propose a diagnostic algorithm based on our observations. We hypothesized that specific symptoms, imaging modalities and clinical tests may be indicative of osteoid osteomas of the foot.

Altogether, 37 patients with osteoid osteoma of the foot, treated between 2000 and 2014 at 6 participating tertiary tumour centres, were retrospectively included. Patients were selected from prospectively maintained tumour databases at the respective centres, using either histological or (in the case no histological analysis had been performed) unambiguous radiological diagnosis of osteoid osteoma of the foot region [13]. Thus, patients were excluded if preoperative imaging had been inconclusive and the suspected diagnosis had not been confirmed histologically. Of all patients, 25 were male (67.6%) and 12 female (32.4%). The mean age of patients at diagnosis of disease was 23.9 years (8–57 years). Images (i.e. X‑ray, computed tomography, CT, magnetic resonance imaging, MRI, bone scan) performed during the diagnostic process were analyzed for the presence of a typical nidus, being a radiolucent area and with or without surrounding sclerosis [12]. Furthermore, symptom duration prior to diagnosis, patients’ minor and major complaints and primarily assumed differential diagnoses were ascertained using medical reports. Anatomical location of the lesion was subdivided into hindfoot (i.e. calcaneus, talus), midfoot (i.e. metatarsals, cuboid, cuneiform, navicular), and forefoot (i.e. phalanges). Ethical approval for this retrospective study had been obtained at the involved centres. The study was performed according to the Declaration of Helsinki.

According to the present study, the often atypical symptoms of foot osteoid osteomas cause a diagnostic delay. The aspirin-test—if performed—is positive in almost all cases and should be initiated early in the diagnostic process. Furthermore, CT scans are superior to MRI in detecting the typical nidus, thus aiding early diagnosis.

In our study, the median delay from onset of symptoms to diagnosis was 1 year, which is less than the median of 1.8 years observed in a systematic review on foot and ankle osteoid osteomas [12]. This may be explained by the fact that our study covered a relatively recent time period where imaging modalities were already quite accurate and that we included only patients treated at tertiary orthopedic tumor centres with a high expertise.

Yet, a mean of 4 imaging studies were necessary to reach the correct diagnosis. A nidus could be detected equally frequent on x‑rays irrespective of the location of osteoid osteomas within bone (i.e. cortical vs. non-cortical). Therefore, even at highly specialized centres, osteoid osteomas of the foot are difficult to diagnose. As reported previously, CT scans were superior to MRI in order to reach the diagnosis of osteoid osteoma [12]. In our study, the typical nidus was visible in 86.2% of cases, which is comparable to the 96.4% reported by Jordan et al. in their systematic review including 223 cases of osteoid osteomas located in the foot and ankle region [12]. On the other hand, MRI revealed a nidus in half of our cases only, due to the fact that in small lesions the nidus and surrounding cortex often show identical signal intensity [17]. Thus, unsurprisingly, a misdiagnosis rate of 35% has been previously reported when MRI was used as the primary imaging modality [12, 18]. In patients with non-cortical osteoid osteomas, significantly more X‑rays were performed during the diagnostic pathway, which could be related to their atypical location and configuration, complicating detection and diagnosis. In our cohort, the most common differential diagnoses were malignant tumors other than osteoid osteoma and osteomyelitis, probably caused by the edematous zones visible on MRI and symptoms as swelling and local tenderness [8, 14]. For example, in the case report by Thiermann et al., an osteoid osteoma of the distal phalanx of the great toe was initially wrongly diagnosed and treated as an ingrown nail with concomitant osteomyelitis, due to bone and soft tissue edema as well as contrast enhancement detected on MRI [8]. These features may also wrongly lead to the differential diagnosis of trauma, which was the case in 10.8% of our patients. It was not assessed to which extent the differential diagnoses were of help in reaching the final diagnosis, although it is more likely that especially the diagnosis of trauma may have resulted in an unnecessary diagnostic delay. The aspirin test, i.e. administration of aspirin with subsequent pain relief, was positive in 90% of our patients prior to invasive treatment, with the number being relatively higher than previously reported 72% [12].

The majority of patients underwent surgical resection of the osteoid osteoma, which is in line with previous observations [7, 12]; however, thermal ablation using, e.g. RFA, is becoming more common [19]. Nevertheless, due to critical structures in close vicinity to bone in the foot, as well as the risk for skin burns and necrosis following RFA [20, 21], open surgery may be favored. Consistent with other reports, in over 90% of our patients symptom relief could be achieved by either RFA or open surgery [22, 23]. Thereafter, as recommended by ESSR, at least one follow-up appointment should be scheduled until patients are free of complaints [14].

Based on our results, the following diagnostic approach for patients with unclear foot pain is suggested: radiographs should be followed by an aspirin test. Upon pain relief, a CT scan of the affected region should be performed in order to visualize the pathognomic nidus.

Limitations of the present study include the non-uniform assessment of patients as they had been treated at different tertiary tumor centres. Furthermore, due to the retrospective design of the study, some data—and in particular regarding follow-up time—was impossible to ascertain; however, with the main aim being to perform a thorough analysis of clinical, diagnostic and treatment-related parameters in patients with osteoid osteoma of the foot, rather than analyzing any outcome variables in the long term, the lack of follow-up information has not affected the study’s research question.

The present retrospective multicentre study confirmed previously published results on osteoid osteomas of the foot region. Potentially due to treatment in highly specialized tertiary tumor centres, the diagnostic delay was relatively shorter than previously described. CT scans proved to be highly sensitive for detection of foot osteoid osteomas, whilst MRI—at least in the present study—was less reliable. An aspirin test is recommended for clinicians in practice prior to CT or MRI examinations in order to strengthen the suspected diagnosis.