Barbara E. Kitchell, DVM, PhD, DACVIM
Feline vaccine-associated sarcomas (VAS) have received a great deal of attention in the veterinary literature over the last 10 years. An increase in observed incidence of feline soft tissue fibrosarcomas was first noted by veterinary pathologists Hendrick and Goldschmidt in the late 1980's. This increased incidence paralleled two temporally- related events. First, in 1985, was the introduction and widespread use of two killed adjuvanted vaccines that had not been previously approved for use in cats. These vaccines were a subcutaneously administered, aluminum-adjuvanted killed rabies virus vaccine and a aluminum-adjuvanted killed feline leukemia virus vaccine. Second, in 1987 the State of Pennsylvania enacted a law requiring rabies vaccination for cats. This effectively increased the number of cats at risk for development of VAS due to an increase in number of cats vaccinated.
Public concern and rumors of lawsuits prompted the California Veterinary Medical Association to bring together a panel of experts to address the issue of feline VAS in August 1996. In November 1996, an organization was formed to promote research and educational efforts for this emerging disease problem. Charter organizations which supported the formation of what is now known as the Vaccine-Associated Feline Sarcoma Task Force (VAFSTF) include the American Veterinary Medical Association, the American Animal Hospital Association, the American Association of Feline Practitioners, Cornell Feline Health Center, and the Veterinary Cancer Society. The VAFSTF has since sponsored many research projects to explore the etiology, pathogenesis, diagnosis and treatment of this disease. The VAFSTF has worked towards development of recommendations regarding vaccination protocols and education of veterinarians and the public with regard to vaccine-site reactions and tumors that form at sites of vaccination.
Investigations have led to a better understanding of the physical, biological, and histological characteristics of these tumors. Estimation of the frequency of feline VAS varies between retrospective studies, and ranges from 1 case/10,000 vaccinates to 1.3cases/1,000 vaccinates. The interval between vaccination and the development of tumors is highly variable. Tumor latency intervals have been reported to be as short as one month and as long as 3.5 years from vaccination (Meyer, EK, personal communication, VCS 18th Annual Conference, 1998). Kass et al. showed a causal and temporal association between feline sarcomas and use of rabies and feline leukemia vaccines. Kass' study also revealed an increased risk of fibrosarcoma development with increased number of vaccines administered. In fact, the risk of developing a fibrosarcoma from a single injection in the cervical/intrascapular region was close to 50% higher than that of nonvaccinates. Risk escalated to 127% when 2 vaccines were administered and climbed to 175% when 3 or 4 concurrent vaccinations were given in the same anatomic site. Kass also observed that vaccines with adjuvants other than aluminum as well as vaccines without adjuvants, were associated with fibrosarcoma development. Epidemiologic studies have implicated an association with feline leukemia (FeLV), rabies, and feline viral rhinotracheitis/calicivirus/panleukopenia virus (FVRCP) vaccines, with monovalent and polyvalent vaccines, and with non-adjuvanted as well as adjuvanted types (Meyer, EK, personal communication, VCS 18th Annual Conference, 1998). A fibrosarcoma has also been reported in association with a lufenuron injection.
Hendrick et al. followed with a study that compared fibrosarcomas that developed at vaccination sites with those arising at nonvaccination sites. This retrospective analysis revealed an association of sarcomas at rabies, FVRCP, and FeLV vaccination sites. These authors hypothesize that normal resident feline fibroblasts and myofibroblasts are induced to proliferate in response to injected adjuvants or other vaccine components. In some cats, these cells ultimately undergo neoplastic transformation.
Histologic and Biologic Behavior
Through histologic evaluation of feline VAS tumor specimens, Hendrick recognized a characteristic inflammatory component associated with these sarcomas that was similar to the granulomatous and lymphoid infiltrates seen in non-malignant injection-site reactions. Many VAS tumors contain spindle cells and multinucleated giant cells with a high degree of nuclear pleomorphism and cellular atypia. Pathology studies comparing vaccine-associated sarcomas (VAS) and non-vaccine associated sarcomas (NVAS) reveal that VAS lesions exhibit histological features consistent with more aggressive behavior than do their NVAS counterparts. Characteristics such as intratumoral necrosis, mitotic activity, and cellular pleomorphism have been shown to be more pronounced in VAS than in NVAS.
Retrospective histologic evaluation of a number of tumor samples revealed a shiny, amorphous, grey-brown material present in the central necrotic zones and within macrophage cytoplasm in 40% of specimens examined. Electron-probe analysis identified this foreign material to be aluminum, a common adjuvant of vaccines.
Immunohistochemical analysis is generally positive for immunoreactive vimentin, the mesenchymal cell intermediate filament, and smooth muscle actin. This immunohistochemical profile supports a fibroblastic or myofibroblastic origin for VAS. Histologically, VAS tumor types include fibrosarcomas, malignant fibrous histiocytomas, rhabdomyosarcomas, soft tissue osteosarcomas, and chondrosarcomas.
These VAS tumors are grossly characterized as well-demarcated, pseudo-encapsulated masses. The biologic behavior of VAS is consistent with local aggressiveness and a high incidence of local recurrence. Furthermore, metastasis is seen in 10-25% of patients. It is observed that VAS is more likely to become metastatic than NVAS.
Research studies supported by the VAFSTF and also by the Morris Animal Foundation have been directed toward understanding the etiology of this form of cancer. Retroviral elements, either endogenous or exogenous, do not appear to be associated with VAS. Ellis et al. investigated 130 suspected VAS from cats. In this population, all 130 suspected VAS were negative for the intratumoral FeLV gp 70 antigen on the basis of immunohistochemistry. One hundred VAS were also examined using polymerase chain reaction and were negative for the FeLV long terminal repeat region.This study decreased the concerns of a retroviral etiology for VAS.
Many groups have explored possible growth factors and oncogenes involved in the pathogenesis of VAS tumor development. Nambier et al. evaluated 40 VAS for expression of nuclear p53 protein by immunohistochemistry (IHC) to detect any correlative association of the p53 tumor suppressor gene in VAS. In 42.5% (17/40), tumor cell nuclei were stained darkly; in 20.0% (8/40), tumor nuclei were stained palely; and in 37.5% (15/40), tumor nuclei were unstained by IHC. This suggests mutations of the p53 gene may play a role in the pathogenesis of these tumors. Further investigation by the University of Minnesota found point mutations in p53 to be relatively rare in VAS and only one of the 20 tumors examined actually harbored the mutation. When these researchers compared matched sets of tumors and blood samples by automated DNA sequence analysis, they demonstrated a loss of heterozygosity at p53 in 39% of cases. This loss had a significant association with increased tumor size (Kanjilal, S, personal communication, VCS 19th Annual Conference, 1999). Research at the University of Pennsylvania revealed that VAS exhibit a mild to strong positive staining for platelet-derived growth factor (PDGF), whereas NVAS have a negative or faint positive reaction. Further, lymphocytes in VAS were positive for PDGF but lymphocytes in normal lymph nodes and Peyer's patches were negative. Macrophages in the area stained positive for PDGF receptor. Hendrick et al. surmised that lymphocytes in vaccine-associated lesions may secrete PDGF to recruit macrophages, and thereby cause fibroblast proliferation as a collateral effect. Researchers at the University of Pennsylvania also identified overexpression of c-jun, a protooncogene. This gene codes for the transcriptional protein AP-1, which is critical to cellular proliferation. Further investigation is ongoing to characterize the resident intratumoral and peritumoral leukocyte population including lineage (T or B) and subset (CD4 vs. CD8, TH1 vs. TH2).
Other investigators have referred to established animal models to gain insight into the mechanisms and pathogenesis of VAS. Four important animal models for sarcoma development include the v-src model in chickens, the bovine papilloma model, the T-cell lymphotrophic retrovirus model, and the v-jun transgenic mouse model. The v-src infected chicken develops tumors at wound sites 10-15 days after injury. Eight to nine month old transgenic mice infected with bovine papilloma virus develop skin tumors in areas prone to scratching. Similarly, the tat gene of human T-cell lymphotrophic retrovirus type I induces mesenchymal tumors in three month old transgenic mice in areas prone to scratching. Finally, v-jun transgenic mice develop tumors at two to three months of age in areas where wounds are inflicted. In each of these models, inflammation plays a role in tumor development.
The concept of tumor development secondary to inflammation and wounding is not new. In fact, many such examples can be found in the human literature including tumors associated with metallic implants, soft tissue sarcomas associated with aluminum oxide hip implants, and angiosarcomas associated with foreign body material. The veterinary literature also provides examples, including esophageal sarcomas associated with Spirocerca lupi in dogs, post-traumatic ocular sarcomas in cats which are histologically similar to VAS, fracture associated sarcomas, and radiation-induced osteosarcomas. Solitary case reports include a liposarcoma associated with a glass foreign body and an osteosarcoma associated with total hip arthroplasty.
Another focus of research has been in the prevention of VAS tumor development. Work done in chickens infected with the Rous Sarcoma virus indicated that if the post-wounding inflammation was inhibited, tumor development was also inhibited. Investigators concluded that inflammatory mediators play a critical role in providing the ideal environment for oncogene integration and activation that leads to tumor development.
Groups at the University of Wisconsin-Madison and MD Anderson Cancer Center have isolated and established thirteen VAS cell lines. These cell lines are being employed to identify and test responses to various growth factors in vitro (Carew, JS, personal communication, VCS 19th Annual Conference, 1999). Investigators hope that expression of certain growth factors and their receptors may provide potential targets for therapeutic intervention.
Diagnostic and Therapeutic Considerations
Many investigators have focused on the diagnostic and therapeutic aspects of this tumor. Veterinarians at the University of California-Davis set out to assess the use of advanced imaging in diagnosis and treatment planning. This group used computed tomography (CT) to evaluate cats with presumed VAS prior to treatment. These CT scans revealed a larger area of involvement detected by contrast-enhanced CT as compared to the pre-contrast images. These investigators surmised that VAS are highly aggressive from the onset, are not impeded by fascial planes, and affect multiple muscle groups at the time of diagnosis (McEntee, M, personal communication, VCS 19th Annual Conference, 1999). This study supports the importance of advanced imaging to determine the extent of surgery and/or the size of the radiation field needed for optimal treatment.
Optimum treatment for VAS is still under investigation. We now know that aggressive first surgical excision provides the best chance for a cure. Hershey et al examined 61 VAS cases treated with surgical excision alone and learned that radical first excision yielded significantly longer median time to first recurrence (325 days) than did marginal first excision (79 days). This study also concluded that cats with tumors located on the limbs had a longer median time to first recurrence (325 days) than cats with tumors located at other sites (66 days). While it is important to note these differences, it is also vital to realize that only 13.8% of these cats had longer than 2-year survival. This study therefore emphasizes the need for effective adjunctive therapy.
The importance of adjunctive therapy such as radiation and chemotherapy has also been evaluated. Cronin et al examined 33 cases treated with radiation therapy followed by surgery. This retrospective analysis found the only variable that influenced treatment success was the presence of tumor cells at the margin of resected tissue. Cytotoxic chemotherapy agents, including carboplatin, doxorubicin, mitoxantrone, cyclophosphamide, and vincristine, have been evaluated by various groups. Although sarcomas in cats are not very chemoresponsive in general, both partial and complete responses have been seen in VAS. Unfortunately, complete cures remain elusive even with the use of aggressive multiple modality treatment. It seems that the best method of treatment of VAS is prevention by avoiding unnecessary vaccination, and prompt aggressive surgical resection of early lesions.
Prevention and Early Intervention
Current recommendations for vaccination may necessitate a change in our thought processes and practice habits. Veterinarians have come to appreciate the need for more rational vaccination guidelines and practices. The VAFSTF and other organizations such as The American Association of Feline Practitioners and Academy of Feline Medicine Advisory Panel on Feline Vaccines have advocated proactive changes in standard vaccination protocols. First, no vaccination should be given in the interscapular space. Instead, rabies vaccines should be administered in the distal aspect of the right hind limb, the FeLV vaccine should be administered in the distal aspect of the left hind limb, and all other vaccines should be administered distally in the right shoulder region. These recommendations were made to increase the potential for complete resection by limb amputation. All vaccines should be administered subcutaneously, as this allows earlier detection of these growths.
Another critical question is the appropriate approach to postvaccination masses, which are most likely benign. Some types of rabies and FeLV vaccines have been associated with postvaccinal reaction masses in 100% of vaccinated cases.Fortunately, most of these masses resolve by two to three months postvaccination. The current recommendation, developed by the VAFSTF, for dealing with vaccination site reactions is to record the anatomic location, shape, and size of all masses that occur at the site of an injection. The group also recommends that all masses that develop in an injection site be managed as if malignant until proven otherwise. A diagnostic biopsy should be performed if the lesion persists for longer than 3 months post-injection, is larger than 2 cm in diameter, or is increasing in size beyond one month post-injection. All biopsies should be performed in such a way that subsequent surgical removal of the biopsy site and tract will not be hindered. Once a VAS has been confirmed, a consultation with an oncologist for current treatment recommendations is critical to optimize treatment planning. Adequate diagnostic assessment combined with multiple modality therapy currently provides optimum chance for cure or long-term remission of these difficult and life threatening tumors (Table 1 and Table 2).
Table 1. Current guidelines for the diagnosis of suspected sarcomas.a
1. Accurately record location, shape, and size of all masses that occur at injection sites.
2. Assume all masses that occur at injection sites are malignant until proven otherwise. Further diagnostics and management are indicated if:
a. The mass persists for more than 3 months following injection.
b. The mass is greater than 2cm in diameter.
c. The mass is increasing in size beyond one month post-injection.
3. If a mass displays any one of the criteria listed above, a diagnostic biopsy should be performed without compromising future definitive therapy. Fine needle aspiration cytology is considered unreliable in the diagnosis of VAS.
a This information is adapted from the current recommendations of the Vaccine Associated Feline Sarcoma Task Force Guidelines.
Table 2. Current guidelines for the treatment of suspected sarcomas.b
1. Complete staging including pre-operative lab tests and thoracic radiographs should be performed.
2. For optimal outcome, VAS should have advanced imaging such as computed tomography (CT) or magnetic resonance imaging (MRI) performed for therapeutic planning.
3. Never "shell out" a sarcoma. Incomplete surgical removal can result in treatment failure.
4. Consultation with an oncologist will aid in the initiation of optimal treatment planning, which may include multi-modality therapy.
5. Submit the entire specimen for histopathology.
6. Report all histologically confirmed VAS to the manufacturer and to U.S. Pharmacopeia Veterinary Practitioners' Reporting Program. (1-800-487-7776)
b This information is adapted from the current recommendations of the Vaccine Associated Feline Sarcoma Task Force Guidelines.