The GREFFI (Groupe d'Etudes des Fibrosarcomes Félins) was founded following the controversy relating to the development of feline vaccine-associated fibrosarcomas, in order to provide rational aetiological information. GREFFI activities, within a wider perspective, have been oriented towards pathogenesis, diagnosis improvement, and design of suitable therapeutic protocols. Faced with such a complex issue, the GREFFI has adopted a multidisciplinary and global approach, involving veterinary practitioners in targeted investigations. They have also established collaborative relationships with the North American VAFSF (Vaccine Associated Feline Sarcoma Task Force) dealing with the same topic.
Origin of "Feline Fibrosarcoma Complex"
Before the 1990s, sarcomas with fusiform cells with cutaneous or subcutaneous locations were known to be rare in cats. They now form part of the most common, if not the most frequent tumors in cats both in France and abroad since, according to investigations, they amount to 12-41% of all feline cutaneous tumors. Frequency estimations in Northern America have reported 20 cases out of 100,000 cats. In Europe, a retrospective investigation covering a period of 5 years reported 0.4 pharmacovigilance cases out of 100,000 cats under care. In France, the mean age of cats developing a fibrosarcoma is 9.6 years within a large interval of 3.2-16 years (95% confidence interval), which is little typical of spontaneous tumors. Distribution seems to have evolved from a single peak at 10-12 years towards bimodal distribution with a first peak at 6-7 years and a second at 10-11 years. Moreover, tumors are significantly more often located at classical injections sites on cats.
In Northern America, epidemiological data have served as a basis for the assumption made by Hendrik and then by other authors, according to which vaccine injections would increase the relative risk of fibrosarcoma development. Other elements support this assumption:
Clinical findings: presence of focal granulomatous panniculitis of post-vaccinal type,
Anatomopathological findings: morphological differences between fibrosarcomas at the site of injection and presence of aluminium hydroxide (constituent of numerous oil-adjuvanted vaccines) in the macrophages infiltrating fibrosarcomas.
In addition, carcinogenesis related to chronic inflammatory phenomena has already been described in another context i.e., chemo- or radio-induced tumors, oesophageal osteosarcoma caused by Spirocerca lupi infestation, appendicular osteosarcoma in the presence of osteosynthesis material, and spinocellular epithelioma of the white cat.
The assumption of vaccine injection-associated tumors is likely; however, it is impossible to determine whether they are attributable to the injection of irritating or allergenic substances all the less so since therapeutic injections of slow release products have also been incriminated. More precisely, the iatrogenic cause appears quite likely but it is not currently possible to define the original substance.
The determination of carcinogenesis is complex and it has not yet been possible to establish a pathogenic pattern. Pathogenesis is even more complex since, in this particular case, the simplest assumptions of FeLV, FeSV, papillomavirus and polyomavirus induction have been excluded. In such a context, further investigations dealing with protooncogens (e.g., c-myc, c-jun) and measurement of free radicals in the microenvironment would be necessary.
Classically, feline fibrosarcoma develops in the form of firm, nodular or multinodular cutaneous lesions, which are neither painful nor ulcerated. Tumors are preferentially located at the injection site:
Interscapular area, dorsal side of the neck: 40-49.5%,
Thorax, flanks: 25-29%,
Loins and back: 13-14%.
Tumors at limb extremities, as they used to be described, have now become rare. FeSV-induced fibrosarcomas, systematically multicentric, are quite rare, too, and mainly affect young kittens.
Feline fibrosarcomas are located subcutaneously and therefore rarely ulcerated, except during the terminal stage. They are rarely painful except when they are very large or infiltrated with deep structures.
The speed of evolution of fibrosarcomas is greatly variable. Small-size nodules may persist as such for quite a long time; inversely, other nodules may double in size within a very short period, which would show that tumoral growth accelerates alongside consecutive excisions.
Although lesions have the same appearance from a macroscopic viewpoint, conventional microscopic examinations show that feline fibrosarcomas are heterogeneous. Their common characteristic is the presence of neoplasic fusiform cells, hence providing evidence of their mesenchymatous origin, but whose differentiation may vary (fibroblasts, osteoblasts, chondroblasts, etc.).
Up until lately, the know classification of feline fibrosarcomas was based on three major histological types:
Purely tumoral type, including sensu stricto fibrosarcomas that represent 30% of the tumors observed. They contain dense so called 'fishbone-shaped" fusiform cells within a collagenic stroma. Mitoses are numerous, and necrotic or haemorrhagic foci are not unusual.
Malignant fibrohistiocytomas, representing 60% of the tumors observed. These tumors are polymorphous and behave like tumoral cells of mesenchymatous appearance, associated with other tumoral cells of histiocytic appearance, all of which contained in a fairly abundant collagenic stroma. Osseous metaplasias, giant multinucleated cells, or myxoid stroma form part of their polymorphism.
Sarcomas (osteosarcomas, chondrosarcomas, hemangiosarcomas, rhabdomysosarcomas, neurofibrosarcomas, etc.) are sometimes associated with the feline fibrosarcoma complex.
Fibromastoses, whose incidence is much lower. They are characterized by a highly differentiated fibroblastic tumoral proliferation, associated with dense collagenic stroma.
The mitotic index is low. Examination of fasciae located close to the lesion often reveals the presence of infraclinical tumoral foci, which may explain post-surgical relapses observed with this type of tumors.
Granulomatous panniculitis are a specific entity since they are not tumoral but inflammatory with the infiltrated necrotic central area surrounded with mononucleated cells often associated with polynucleated neutrophils and eosinophils. This phenomenon could correspond to the materialization of a specific sensitivity of the subcutaneous tissue of cats in which it develops. It could therefore be the source of a future fibrosarcoma. However, the correlation between this inflammatory lesion and fibrosarcoma tumoral lesions has not been established yet.
Data recently obtained by the GREFFI and based on immunohistochemical procedures however question the above classification. More precisely, two elements of observation disturb the differentiation between fibrosarcoma and malignant fibrohistiosarcoma:
According to conventional histology, sensu stricto fibrosarcomas are tumors with fusiform cells that do not contain any other type of cells. In fact, all fibrosarcomas prove to contain a variable quantity of macrophagic cells when the latter are labelled with suitable antibodies.
According to conventional histology, both malignant fibrohistiosarcomas and fibrosarcomas contain enough macrophagic cells to be visible. These macrophages are not tumoral but merely consist of reactional stromal cells.
Overall, the nature of both types of lesions mentioned above is identical, except for the quantity of cells of stromal macrophagic origin which is the only way to differentiate them.
There are two major types of tumoral behaviour:
Tumors with local (relapse) and systemic (metastasis) aggressivity.
They consist of fibrosarcomas (sensu stricto fibrosarcoma and malignant fibrohistiocytoma) and related sarcomas with significant local aggressivity and presenting a serious risk of post-surgical relapses. Metastasic capacity is low (10-15% approximately). Metastases are located in lymph nodes and the lungs, and usually develop quite a long time following onset of the initial tumor.
Lesions with local aggressivity (relapse) only.
They consist of fibromatoses which do not metastize but frequently relapse or develop at another site. Relapses are related to peripheral and infraclinical tumoral foci which are difficult to remove, even during wide surgical excision.
Granulomatous panniculitis develops within a few weeks following an inflammatory stimulus, whichever it is (e.g., injection of vaccine, slow release antibiotics, or injectable antiflea preparations, etc.). However, the detection of panniculitis presents a double advantage:
Revealing the sensitivity of subcutaneous connective tissue in some cats during a chemical or physical aggression,
Their specific evolution would explain the evolution of granulomatous panniculitis to fibrosarcoma.
Treatment of feline fibrosarcoma
For most the above tumors, the therapeutic procedure to be implemented to treat feline fibrosarcomas is complex. The presumed biological behaviour of each tumor and of each individual must be taken into account.
The therapeutic strategy is based on wide surgical excision, whether associated or not with other local (radiotherapy) or general (chemotherapy) treatments. The surgery of tumors with serious local aggressivity must be associated with a complementary treatment so as to reduce local relapses, which is the role of radiotherapy. Chemotherapy aims at limiting the development of distant metastases. Immonogenotherapy is currently used experimentally, but the preliminary results obtained lead us to believe that it will develop rapidly.
Like most tumors, surgery remains the main treatment for feline fibrosarcomas. Both the precocity and the extent of excision are conditions for the success of surgery. In effect, if the operation is carried out when the tumor has already reach quite a large size, excision is difficult to perform and, therefore, probably incomplete. The best results are obtained with tumors smaller than 1cm3: any subcutaneous neoformation must be excised and analysed in the shortest possible time.
In addition, aggressive surgery is largely responsible for the frequency of local relapses since fibrosarcomas are conjunctival tumors and, therefore, encapsulated and frequently infiltrated in muscles and adjacent fascia. It is advisable to remove the whole muscle of fascia in whose contact the tumor has developed (external oblique muscle and fascia in case of abdominal localization, cutaneous muscle in the case of thoracic localization, etc.) and even remove all adjacent muscles, where possible. In the event of interscapular fibrosarcoma, it is sometimes necessary to perform a scapulectomy together with the excision of spinous processes which often are the seat of local relapses.
With a view to reducing the risk of carcinogenesis induced by post-surgical inflammation, reconstruction will be minimal and non-inflammatory stitching material shall be used (in particular, catgut must not be used).
Complementary radiotherapy to surgery is particularly suited for tumors with a potentially high risk of local relapses. Sensu stricto fibrosarcomas, malignant fibrohistiocytomas, and all soft tissue sarcomas with a high mitotic index are good indications for adjuvant radiotherapy. On the other hand, radiotherapy is not a good indication for fibromatoses and sarcomas.
Interstitial radiotherapy or curietherapy allow addressing many of the problems posed by the feline fibrosarcoma.
There are so many parameters involved in the objective evaluation of adjuvant curietherapy that they are difficult to control. Practically speaking, we have observed that the efficiency of adjuvant curietherapy on feline fibrosarcomas depends on several factors:
Tumor size before excision. The tumor size conditions the quality of excision, which is the reason why the curietherapy of large-sized tumors achieves poorer results;
Histological nature of the tumor. Fibrohistiocytomas, sensu stricto fibrosarcomas, and sarcomas with high mitotic index are more radiosensitive than fibromatoses or low-grade fibrosarcomas.
Number of previous excisions. Radiotherapy following a first excision is much more efficient than if the tumor has already relapsed several times.
Extent and quality of surgical excision. The lesser tumoral cells on the excision site, the greater the efficiency of adjuvant radiotherapy.
Early radiotherapy following surgery. The surgical excision region must be radiated as soon as skin cicatrisation is sufficient (usually 1-2 weeks following surgery) and not when the tumor has started relapsing.
Precise localization of the area to be radiated. The radiotherapist must accurately locate the initial tumor in order to determine the area to be radiated. This requires the surgeon to draw precise diagrams or even use pictures.
As a conclusion, the best long-term results can be obtained during adjuvant curietherapy of small-sized tumors (less than 1 cm in diameter) with high mitotic index removed for the first time. Excision must be wide and curietherapy carried out within three weeks following excision.
Feline fibrosarcoma tumors do not metastize greatly (less than 15%), usually quite late and often when satisfactory local examination has been obtained with adjuvant radiotherapy. Moreover, soft tissue sarcomas are considered chemosensitive in both human and veterinary medicine. The advantage of chemotherapy therefore is disputable for such tumors.
Non-specific immunomodulators, which were widely used twenty years ago, have been abandoned in veterinary medicine due to their low efficiency and little specificity. In the United States, only Acenaman® - an Aloe vera extract with immunomodulating properties-has been authorized for marketing for the treatment of numerous solid tumors. The benefit of its use however remains debatable.
In human medicine, the use of cytokines, such as interleukine2 or gamma interferon, are particularly promising but their systemic toxicity makes them delicate to use. This toxicity problem has been solved thanks to the use of genetically modified cytokine producing cells.
Clinical trials were carried out on cats having developed a fibrosarcoma. VERO cells (monkey kidney cell line) that were genetically modified in order to produce interleukine2 (hIL2), were used. When injected into the surgical excision site, the VERO-hIL2 cells behaved as "micropumps" producing human interleukine2. The interleukine2 released is active and stimulates the multiplication of immunocompetent cells on the tumoral site. A number of these cells are specifically directed against the tumoral antigens carried by the operated tumor. This investigation gave evidence of a statistically significant improvement of relapse-free survival and even global survival of the cats subjected to surgery, curietherapy and VERO-hIL2 cells, as opposed to those cats that were only subjected to surgery and curietherapy. This treatment is therefore a promising procedure for the treatment of tumors in both man and the animal. Several protocols are currently implemented with a view to further improving the results obtained.
The prognosis of such tumors continues to be poor although the therapeutic arsenal available allows intervening relatively efficiently. Early and wide excision completed with an adjuvant treatment (curietherapy or immunotherapy) improves the prognosis. Lastly, diagnostic improvement using immunochemical procedures would allow intervening in the best suited manner therapeutically.