A partir de células de osteossarcoma humano (MG-63) que tem características de osteoblastos imaturos produzimos anticorpos monoclonais (Mabs) nomeados PSP 4-5, PSP 42-22 e PSP 85-9. Esses anticorpos reconhecem antígenos de 100, 26 e 20 kDa respectivamente. Avaliamos a especificidade dos mesmos, testando suas expressões em cortes congelados de tecidos oriundos da mesma célula mesenquimal, ou seja, osso, cartilagem, músculo cardíaco e tecido adiposo. Os anticorpos marcaram o periósteo, com distintos padrões de expressão, porém o PSP 42-22 foi o mais específico marcando a camada osteogênica do periósteo. Os anticorpos marcaram também células ósseas. Diante desses resultados nosso objetivo, no presente estudo, foi identificar os antígenos reconhecidos por esses anticorpos usando clonagem e sequenciamento dos genes em biblioteca de cDNA com capacidade de expressão proteica. Outro objetivo foi testar os anticorpos, pela técnica de imunohistoquímica (IHQ), em tumores ósseos primários visando estabelecer os padrões de marcação e se diferenciavam os diferentes tipos de tumores. Os antígenos reconhecidos pelos anticorpos PSP 42-22 e PSP 85-9 foram isolados, purificados e sequenciados. Devido ao elevado peso molecular do PSP 4-5 necessitaremos de outras técnicas para isolar o clone que codifica seu antígeno. O sequenciamento do clone isolado pelo PSP 42-22, mostrou uma sequência com 99% de homologia descrita no \"Homo sapiens\" como sendo \"serologically defined colon cancer antigen 3 (SDCCAG3), transcript variant 3\". A proteína SDCCAG3 foi descrita pela primeira vez em 1998 como um antígeno de câncer de cólon reconhecido por anticorpo autólogo. Vale lembrar que antígeno reconhecido pelo nosso anticorpo tem um peso molecular de aproximadamente 26 kDa, inferior ao da proteína SDCCAG3. Essa diferença poderia ocorrer devido a uma diferença no local da tradução do mRNA das células MG-63, produzindo uma proteína menor (isoforma), ou no clone isolado (3B2-3-1), poderia ocorrer uma mutação produzindo uma proteína mais curta que se expressaria na membrana celular ao contrário de uma proteína mais longa. Já o alinhamento das sequências obtidas dos clones isolados utilizando o PSP 85-9, mostrou uma sequência com 100% de concordância descrita com a porção não codante da proteína \"Homo sapiens microtubule associated protein, RP/EB family, member 1 (MAPRE1)\" portanto um falso positivo; já o clone 1A5-1-3, mostrou uma sequência de 99% de homologia com a proteína \"Homo sapiens Yip1 interacting factor homolog B (S. cerevisiae) (YIF1B), transcript variant 8\". Trata-se de uma proteína de membrana, com 6 isoformas diferentes que pode interagir com o receptor de serotonina. O resultado da IHQ nos tumores ósseos testados mostrou que o PSP 4-5 se expressou predominantemente no citoplasma de osteossarcoma, condrossarcoma e leiomiossarcoma e no núcleo de osteoblastoma. O anticorpo PSP 42-22 não reconheceu nenhum antígeno nos tumores avaliados. Quanto ao anticorpo PSP 85-9 nos condrossarcomas e leiomiossarcomas a expressão foi predominante citoplasmática e nos osteossarcomas e osteoblastoma foi nuclear. Em conclusão, identificamos os antígenos de dois dos anticorpos estudados que apresentam potencial diagnóstico diferencial em tumores ósseos primários / From human osteosarcoma cells (MG-63), which have characteristics of immature osteoblasts, we produced monoclonal antibodies (Mabs) named PSP 4-5, PSP 42-22 and PSP 85-9. These antibodies recognize antigens with 100, 26 and 20 kDa, respectively. We evaluated their specificity by testing their expression in frozen tissue sections from the same mesenchymal cells, that is, bone, cartilage, cardiac muscle and adipose tissue. The antibodies stained the periosteum, with distinct patterns of expression, but PSP 42-22 was the most specific, scoring the osteogenic layer of the periosteum. The antibodies also marked bone cells. With these results, our goal in this study was to identify the antigens recognized by these antibodies using cloning and sequencing of the genes in the cDNA library with capacity of protein expression. Another objective was to test the antibodies by immunohistochemistry (IHC) in primary bone tumors to establish standards for marking and whether they set apart different types of tumors. The antigens recognized by the PSP 42-22 and PSP 85-9 antibodies were isolated, purified and sequenced. Due to the high molecular weight of PSP 4-5 we will need other techniques to recognize its antigen. Sequencing of the clone isolated using the PSP 42-22 showed a sequence with 99% homology described in \"Homo sapiens\" as being \"serologically defined colon cancer antigen 3 (SDCCAG3), transcript variant 3\". The SDCCAG3 protein was first described in 1998 as a colon cancer antigen recognized by autologous antibody. It is worth remembering that the antigen recognized by our antibody has a molecular weight of approximately 26 kDa, lower than the SDCCAG3 protein. This difference could be due to a difference in mRNA translation site of MG-63 cells producing a lower protein; or on the isolated clone (3B2-3-1) a mutation could occur producing a shorter protein that expresses in the cell membrane instead of a longer protein. The alignment of the sequences obtained from isolated clones using the PSP 85-9 showed a sequence with 100% of homology described with the non-coding protein portion \"Homo sapiens microtubule associated protein, RP/EB family, member 1 (MAPRE1), mRNA\" therefore, a false positive; 1A5-1-3 clone showed a 99% homology sequence with the protein \"Homo sapiens Yip1 interacting factor homolog B (S. cerevisiae) (YIF1B), transcript variant 8, mRNA\". It is a membrane protein with 6 different isoforms which can interact with the serotonin receptor. The result of IHC in bone tumors tested showed that PSP 4-5 expressed predominantly in the cytoplasm of osteosarcoma, chondrosarcoma, leiomyosarcoma and osteoblastoma in was in nucleus. The PSP 42-22 antibody did not recognize any antigen in the tumors examined. As for the PSP 85-9 antibody in chondrosarcoma and leiomyosarcoma, the expression was predominantly cytoplasmic and the osteoblastoma was nuclear in osteosarcomas. In conclusion, we have identified the antigens of two of the antibodies studied which have potential differential diagnosis in primary bone tumors

A partir de células de osteossarcoma humano (MG-63) que tem características de osteoblastos imaturos produzimos anticorpos monoclonais (Mabs) nomeados PSP 4-5, PSP 42-22 e PSP 85-9. Esses anticorpos reconhecem antígenos de 100, 26 e 20 kDa respectivamente. Avaliamos a especificidade dos mesmos, testando suas expressões em cortes congelados de tecidos oriundos da mesma célula mesenquimal, ou seja, osso, cartilagem, músculo cardíaco e tecido adiposo. Os anticorpos marcaram o periósteo, com distintos padrões de expressão, porém o PSP 42-22 foi o mais específico marcando a camada osteogênica do periósteo. Os anticorpos marcaram também células ósseas. Diante desses resultados nosso objetivo, no presente estudo, foi identificar os antígenos reconhecidos por esses anticorpos usando clonagem e sequenciamento dos genes em biblioteca de cDNA com capacidade de expressão proteica. Outro objetivo foi testar os anticorpos, pela técnica de imunohistoquímica (IHQ), em tumores ósseos primários visando estabelecer os padrões de marcação e se diferenciavam os diferentes tipos de tumores. Os antígenos reconhecidos pelos anticorpos PSP 42-22 e PSP 85-9 foram isolados, purificados e sequenciados. Devido ao elevado peso molecular do PSP 4-5 necessitaremos de outras técnicas para isolar o clone que codifica seu antígeno. O sequenciamento do clone isolado pelo PSP 42-22, mostrou uma sequência com 99% de homologia descrita no \"Homo sapiens\" como sendo \"serologically defined colon cancer antigen 3 (SDCCAG3), transcript variant 3\". A proteína SDCCAG3 foi descrita pela primeira vez em 1998 como um antígeno de câncer de cólon reconhecido por anticorpo autólogo. Vale lembrar que antígeno reconhecido pelo nosso anticorpo tem um peso molecular de aproximadamente 26 kDa, inferior ao da proteína SDCCAG3. Essa diferença poderia ocorrer devido a uma diferença no local da tradução do mRNA das células MG-63, produzindo uma proteína menor (isoforma), ou no clone isolado (3B2-3-1), poderia ocorrer uma mutação produzindo uma proteína mais curta que se expressaria na membrana celular ao contrário de uma proteína mais longa. Já o alinhamento das sequências obtidas dos clones isolados utilizando o PSP 85-9, mostrou uma sequência com 100% de concordância descrita com a porção não codante da proteína \"Homo sapiens microtubule associated protein, RP/EB family, member 1 (MAPRE1)\" portanto um falso positivo; já o clone 1A5-1-3, mostrou uma sequência de 99% de homologia com a proteína \"Homo sapiens Yip1 interacting factor homolog B (S. cerevisiae) (YIF1B), transcript variant 8\". Trata-se de uma proteína de membrana, com 6 isoformas diferentes que pode interagir com o receptor de serotonina. O resultado da IHQ nos tumores ósseos testados mostrou que o PSP 4-5 se expressou predominantemente no citoplasma de osteossarcoma, condrossarcoma e leiomiossarcoma e no núcleo de osteoblastoma. O anticorpo PSP 42-22 não reconheceu nenhum antígeno nos tumores avaliados. Quanto ao anticorpo PSP 85-9 nos condrossarcomas e leiomiossarcomas a expressão foi predominante citoplasmática e nos osteossarcomas e osteoblastoma foi nuclear. Em conclusão, identificamos os antígenos de dois dos anticorpos estudados que apresentam potencial diagnóstico diferencial em tumores ósseos primários / From human osteosarcoma cells (MG-63), which have characteristics of immature osteoblasts, we produced monoclonal antibodies (Mabs) named PSP 4-5, PSP 42-22 and PSP 85-9. These antibodies recognize antigens with 100, 26 and 20 kDa, respectively. We evaluated their specificity by testing their expression in frozen tissue sections from the same mesenchymal cells, that is, bone, cartilage, cardiac muscle and adipose tissue. The antibodies stained the periosteum, with distinct patterns of expression, but PSP 42-22 was the most specific, scoring the osteogenic layer of the periosteum. The antibodies also marked bone cells. With these results, our goal in this study was to identify the antigens recognized by these antibodies using cloning and sequencing of the genes in the cDNA library with capacity of protein expression. Another objective was to test the antibodies by immunohistochemistry (IHC) in primary bone tumors to establish standards for marking and whether they set apart different types of tumors. The antigens recognized by the PSP 42-22 and PSP 85-9 antibodies were isolated, purified and sequenced. Due to the high molecular weight of PSP 4-5 we will need other techniques to recognize its antigen. Sequencing of the clone isolated using the PSP 42-22 showed a sequence with 99% homology described in \"Homo sapiens\" as being \"serologically defined colon cancer antigen 3 (SDCCAG3), transcript variant 3\". The SDCCAG3 protein was first described in 1998 as a colon cancer antigen recognized by autologous antibody. It is worth remembering that the antigen recognized by our antibody has a molecular weight of approximately 26 kDa, lower than the SDCCAG3 protein. This difference could be due to a difference in mRNA translation site of MG-63 cells producing a lower protein; or on the isolated clone (3B2-3-1) a mutation could occur producing a shorter protein that expresses in the cell membrane instead of a longer protein. The alignment of the sequences obtained from isolated clones using the PSP 85-9 showed a sequence with 100% of homology described with the non-coding protein portion \"Homo sapiens microtubule associated protein, RP/EB family, member 1 (MAPRE1), mRNA\" therefore, a false positive; 1A5-1-3 clone showed a 99% homology sequence with the protein \"Homo sapiens Yip1 interacting factor homolog B (S. cerevisiae) (YIF1B), transcript variant 8, mRNA\". It is a membrane protein with 6 different isoforms which can interact with the serotonin receptor. The result of IHC in bone tumors tested showed that PSP 4-5 expressed predominantly in the cytoplasm of osteosarcoma, chondrosarcoma, leiomyosarcoma and osteoblastoma in was in nucleus. The PSP 42-22 antibody did not recognize any antigen in the tumors examined. As for the PSP 85-9 antibody in chondrosarcoma and leiomyosarcoma, the expression was predominantly cytoplasmic and the osteoblastoma was nuclear in osteosarcomas. In conclusion, we have identified the antigens of two of the antibodies studied which have potential differential diagnosis in primary bone tumors

Η θεραπεία των καλοήθων οστικών όγκων εξαρτάται από την ανατομική εντόπιση, τα συμπτώματα, τη φυσική ιστορία του όγκου και τη θνησιμότητα της θεραπείας η οποία στις περισσότερες περιπτώσεις περιλαμβάνει ή την εκτομή ή την απόξεση αν και όχι σπάνια είναι απαραίτητο να πραγματοποιηθεί μία ευρύτερη εξαίρεση, χρησιμοποιώντας τις ίδιες αρχές όπως στους κακοήθεις οστικούς όγκους. Οι νεότερες διαδερμικές μέθοδοι, με κυριότερο εκπρόσωπο τη διαδερμική θερμοκαυτηρίαση (RFA) έρχονται να δώσουν τη λύση, εξασφαλίζοντας την πλήρη καταστροφή του όγκου αλλά και τη διατήρηση της λειτουργικής κινητικότητας του ασθενούς. Η παρούσα μελέτη εξέτασε την αποτελεσματικότητα και την ασφάλεια της εν λόγω θεραπευτικής μεθόδου στην αντιμετώπιση κατά κύριο λόγο των οστεοειδών οστεωμάτων αλλά και άλλων καλοήθων οστικών όγκων, όπως τα χονδροβλαστώματα και τα οστεοβλαστώματα. Επίσης, μελετήθηκαν τα απεικονιστικά πρότυπα των όγκων πριν και μετά RFA και έγινε συσχέτιση των απεικονιστικών αποτελεσμάτων με άλλες παραμέτρους των υπό μελέτη οστικών όγκων. ΥΛΙΚΟ ΚΑΙ ΜΕΘΟΔΟΣ Από τον Δεκέμβριο του 2003 έως και τον Δεκέμβριο του 2009 συνολικά 33 ασθενείς με συνολικά 33 καλοήθεις οστικούς όγκους υποβλήθηκαν σε διαδερμική θερμοκαυτηρίαση με ραδιοσυχνότητες (κατόπιν ενυπόγραφης πληροφορημένης συναίνεσης). Από τους 33 καλοήθεις όγκους, οι 29 ήταν οστεοειδή οστεώματα, οι τρεις χονδροβλαστώματα και ο ένας οστεοβλάστωμα. Από τους 33 ασθενείς, οι 23 ήταν άντρες και οι 10 γυναίκες (♂/♀: ~2/1), ηλικίας από 11 έως 39 ετών (μέση ηλικία : 22,5 έτη). Η διάμετρος των οστεοειδών οστεωμάτων κυμαίνονταν από 4 έως 12 χιλ. (μέση τιμή:7 χιλ.), των χονδροβλαστωμάτων από 26 έως 32 χιλ. (μέση τιμή:29 χιλ.) και του οστεοβλαστώματος 32 χιλ. Συνολικά 27 (82%) όγκοι εντοπίζονταν στα κάτω άκρα, τέσσερις (12%) στα άνω άκρα και δύο (6%) στην σπονδυλική στήλη. Οι έντεκα (33%) από τις 33 βλάβες είχαν ενδαρθρική εντόπιση. Η διάγνωση των οστεοειδών οστεωμάτων στηρίχθηκε αποκλειστικά σε ακτινολογικά και κλινικά κριτήρια, ενώ για τα χονδροβαλστώματα και το οστεοβλάστωμα πραγματοποιήθηκε βιοψία. Σε όλους τους όγκους η θερμοκαυτηρίαση πραγματοποιήθηκε με τη χρήση ενός άκαμπτου RF ηλεκτροδίου σε σχήμα «ράβδου» εκτός από δύο περιπτώσεις χονδροβλαστωμάτων της μηριαίας κεφαλής όπου χρησιμοποίηθηκε εκπτυσσόμενο RF ηλεκτρόδιο δίκην «ομπρέλας». Καταγράφηκαν και μελετήθηκαν τα ποσοστά επιτυχίας, οι υποτροπές, οι επιπλοκές καθώς και τα αποτελέσματα της στατιστικής ανάλυσης. ΑΠΟΤΕΛΕΣΜΑΤΑ Η μέθοδος ήταν τεχνικά επιτυχής και στους 33 όγκους (100%). Υποτροπή των συμπτωμάτων παρουσιάστηκε σε τρεις περιπτώσεις (9%, 3/33) οστεοειδών οστεωμάτων (ένα ενδαρθρικό στο σπογγώδες οστό και δύο εξω-αρθρικά στην φλοιώδη μοίρα), αντίστοιχα στους δύο, έξι και τέσσερις μήνες μετά την «θερμοκαυτηρίαση» (μέση τιμή : 4 μήνες). Η μη σωστή τοποθέτηση του RF ηλεκτροδίου μέσα στον όγκο και ο τραυματισμός του αρθρικού χόνδρου ήταν οι λόγοι αποτυχίας της μεθόδου. Ο ένας ασθενής υπεβλήθη σε δεύτερο επιτυχή «θερμοκαυτηριασμό» (εξω-αθρικό, φλοιώδες οστεοειδές οστέωμα), ο δεύτερος ακολούθησε τη χειρουργική οδό (εξω-αθρικό, φλοιώδες οστεοειδές οστέωμα) και ο τρίτος συνέχισε τη φαρμακευτική αγωγή (ενδαρθρικό, σπογγώδες οστεοειδές οστέωμα). Έτσι, τα ποσοστά κλινικής επιτυχίας μετά την 1η RF συνεδρία ανέρχονται στο 91% (30/33) και συνολικά μετά και την 2η RF συνεδρία ανέρχονται τελικά στο 94% (31/33). Η περίοδος κλινικής παρακολούθησης κυμάνθηκε συνολικά για όλους τους όγκους από 6 έως 70 μήνες (μέση τιμή follow-up : 28 μήνες). Στις επιπλοκές περιλαμβάνονται μία περίπτωση θερμικής δερματικής κάκωσης 1ου βαθμού (ελάσσονα επιπλοκή), μία περίπτωση ιατρογενούς- εκφυλιστικής οστεοαρθρίτιδας (3η υποτροπή) και μία περίπτωση σηπτικής αρθρίτιδας της αριστερής κατά γόνυ αρθρώσεως με σχηματισμό δερματικού συριγγίου δύο μήνες μετά RFA που αντιμετωπίστηκε χειρουργικά. Για τις 25 (86%, 25/29) συνολικά περιπτώσεις οστεοειδών οστεωμάτων με διαθέσιμο CT follow-up πριν και μετά τη θερμοκαυτηρίαση (6, 12 και 24 μήνες), πραγματοποιήθηκε στατιστική συσχέτιση πολλαπλών μεταβλητών και προέκυψαν τα κάτωθι συμπεράσματα (Kendall’s t-test): 1) Η ελάχιστη ή και μηδαμινή «οστεοποίηση» της «φωλεάς» δεν υποδεικνύει απαραίτητα την κλινική αποτυχία (P=0.14). 2) Η ανίχνευση εσωτερικής «οστεοποίησης» της «φωλεάς» παρουσίασε έντονη θετική συσχέτιση με την μεγάλη (≥12 μήνες) διάρκεια του CT follow-up (P=0.014). 3) Το μεγάλο μέγεθος των οστεοειδών οστεωμάτων (>7 χιλ.) συσχετίζεται έντονα αρνητικά με την φλοιώδη (P=0.001), την εξωαρθρική (P=0.003) και την διαφυσιακή εντόπιση (P=0.001). 4) Επίσης και το μεγάλο (≥ 7χιλ.) μέγεθος (P=0.086), δηλαδή και το μεγάλο μέγεθος τείνει να συσχετίζεται με την «ωριμότητα» του όγκου (παρουσία εσωτερικών αποτιτανώσεων προ RFA) . Στοιχεία εσωτερικής «οστεοποίησης» (μετά RFA) ανεδείχθησαν ήδη με την συμπλήρωση των 12 μηνών από την θερμοκαυτηρίαση για δύο χονδροβλαστώματα ενώ το τρίτο εμφάνισε σχεδόν πλήρη «οστεοποίηση» στους 48 μήνες απεικονιστικού follow-up. Απ’ την άλλη, δεν παρατηρήθηκε αλλαγή στην απεικόνιση του οστεοβλαστώματος της κερκίδας στο εξάμηνο ακτινολογικό follow-up. Μέχρι σήμερα όλα τα περιστατικά των χονδροβλαστωμάτων και του οστεοβλαστώματος παραμένουν –από κλινικής απόψεως- ελεύθερα άλγους ή κινητικής δυσχέρειας. ΣΥΜΠΕΡΑΣΜΑ Η διαδερμική θερμοκαυτηρίαση με ραδιοσυχνότητες (RF Ablation) είναι μία ελάχιστα επεμβατική θεραπευτική επιλογή για τους ασθενείς με οστεοειδή οστεώματα, η οποία παρέχει άμεση ανακούφιση από το άλγος με χαμηλά ποσοστά επιπλοκών και υποτροπών. Η RF θερμοκαυτηρίαση θεωρείται πλέον ως η θεραπεία εκλογής για τα οστεοειδή οστεώματα του περιφερικού σκελετού και της πυέλου αλλά και για τις χειρουργικές υποτροπές. Μειώνει την ενδονοσοκομειακή νοσηλεία και τη διάρκεια της επαναφοράς και αποκατάστασης του ασθενούς. Η οστική βιοψία πριν την RF θεραπεία δεν είναι απαραίτητη εφ’ όσον η διάγνωση μπορεί με ασφάλεια να βασιστεί στα κλινικά και απεικονιστικά ευρήματα. Η αξιολόγηση της κλινικής πορείας και όχι η απεικόνιση είναι το καλύτερο κριτήριο για να γίνει διάκριση μεταξύ των ασθενών με καλή ή όχι κιλινική έκβαση. Η καταγραφή και η συσχέτιση παραμέτρων, όπως το μέγεθος, η θέση και η παρουσία εσωτερικών αποτιτανώσεων των οστεοειδών οστεωμάτων δύναται να βοηθήσουν στην κατανόηση της παθογένεσής τους. Η μέθοδος, σε επιλεγμένες περιπτώσεις (θέση-μέγεθος) και υπό κατάλληλες συνθήκες (πολλαπλές «αλληλοεπικαλυπτόμενες» RF συνεδρείες μικρότερης «ενεργού» ακτίνας), μπορεί να εφαρμοστεί επιτυχώς και σε άλλους καλοήθεις οστικούς όγκους όπως τα χονδροβλαστώματα και τα οστεοβλαστώματα. / Treatment of benign primary bone tumors depends on the anatomical location, symptoms, the natural history of the tumor and the morbidity of treatment and in most cases involves either simple excision or curettage although occasionally it is necessary to perform a complete excision using the same principles as for malignant tumors. CT-guided radiofrequency ablation (RFA), has emerged as minimally invasive alternative to destroy the tumor, overcome surgical difficulties and potential hazards and preserve the functional ability of the patient. The present study demonstrates the healing effect of RFA and evaluates its efficacy and safety in the treatment of osteoid osteomas, chondroblastomas and osteoblastomas. Additionally, this series compare the imaging pattern of the bone tumors prior and post RFA and correlate the results with other selected tumor parameters. MATERIAL AND METHODS From December 2003 to December 2009 a total number of 33 patients (23 male, 10 female, 11-39 years, mean: 22, 5 years) with 33 benign bone tumors were treated with RFA. Informed consent and institutional board approval were obtained. The tumors consisted of 29 osteoid osteomas, three biopsy-proved chondroblastomas and one biopsy-proved osteoblastoma. The mean maximum diameter was 7mm (range: 4-12 mm) for osteoid osteomas, 29 mm for chondroblastomas (range: 26-32 mm) and 32mm for the osteoblastoma. Lesions were located in the limbs (n: 27, 82%), the upper arm (n: 4, 12%) and two in the spine (n: 2, 6%). Intra-articular location was detected in 11(33%) tumors. Diagnosis of osteoid osteomas was base on imaging and clinical criteria. Ablation was performed using a straight rigid RF electrode in 31 tumors while a multi-tined expandable RF electrode was used in two cases of femoral chondroblastomas. Primary success rate, total secondary success rate, recurrences, complications, follow-up and statistical analysis results were assessed. RESULTS Technical success was achieved in 33 patients (33/33, 100%). Recurrence occurred in three osteoid osteomas (3/33, 9%); one intra-articular medullar lesion and two extra-articular cortical lesions at two, six and four months post RFA respectively (mean: 4 months). Failure was attributed to inadequate RF electrode positioning in the cortical lesions whilst articular damage was the main reason for pain relapse in the third intra-articular case. Primary success rate was 91% and total secondary success rate was 94%. Mean clinical follow-up period was 28 months (range: 6-70 months) for all lesions. Complications comprised of one mild thermal skin injury, one hip joint degenerative arthritis (third intra-articular failure case) and a case of septic arthritis with bony changes and cutaneous fistula, due to wound infection, which required surgical debridement. Statistical analysis of bone tumor parameters regarding 25 cases of osteoid osteomaswith available CT-follow-up, like complete, partial or absent ossification of the treated nidus, patient age and sex, tumor size and location, pre-existing calcifications, clinical outcome and CT follow-up, reached the following results (Kendall’s t-test): 1) Absence and/or minimal of post RFA ossification does not necessarily indicate clinical failure (P=0.14). 2) Detection of post RFA ossification showed an intense positive correlation with a long-lasting CT follow-up (≥ 12 months) (P= 0.014). 3) The “big” size of osteoid osteomas (>7mm) showed an intense negative correlation with the cortical (P=0.001), extra-articular (P= 0.003) and diaphyseal location (P=0.001). 4) Also, the “big” size of osteoid osteomas (>7mm) tends to correlate with the presence of calcifications (prior RFA), which represents a “maturity” marker of the tumor (P= 0.086). All three cases of chondroblastomas showed signs of internal ossification post RFA on regular imaging follow-up while the osteoblastoma did not show any imaging changes on the 6-month follow-up. On the other hand, the osteoblastoma and the remaining three cases of chondroblastomas presented an excellent post RFA clinical course without any signs of relapse. CONCLUSIONS Percutaneous RFA is a minimally invasive therapeutic option for osteoid osteomas which provides immediate pain relief and low rates of complications and recurrences. It is considered as the treatment of choice for appendicular and pelvic osteoid osteomas and for surgical recurrences as well. Biopsy is not mandatory provided that the diagnosis can be safely based on clinical and imaging grounds. The determination of an adverse clinical outcome should be based on clinical evaluation and not on the imaging pattern. The study and correlation of tumor parameters like size, location and pre-existing calcifications of osteoid osteomas can help in understanding their pathogenesis. The present study also suggests that RFA, when correctly performed, should be included in the treatment algorithm of selected cases of other benign bone tumors like chondroblastomas and osteoblastomas.