What We Can Say-and Not Say-About the Epidemiology of Vaccine-Associated Sarcomas

It has been a little more than a decade since Drs. Hendrick and Goldschmidt1 submitted their query to the profession, "Do injection-site reactions induce fibrosarcomas in cats?"

Since then the Vaccine-Associated Feline Sarcoma Task Force (VASTF) was formed and there have been many investigations into feline vaccine site-associated sarcoma incidence, prevalence, epidemiology and treatment. This condition is still an emerging epidemic problem, although most vaccinated cats remain unaffected.

The only way to completely stop vaccine-associated sarcomas is to avoid vaccinations, but vaccine avoidance is dangerous, because the diseases they are designed to prevent are still capable of causing epidemics in animal populations. The public and the veterinary profession are faced with a quandary caused by the juxtaposition of the need to vaccinate and the desire to avoid iatrogenic tumors. Recent research on the incidence of feline vaccine site-associated sarcomas2 and on other risk factors that contribute to sarcoma development3 have given the veterinary profession valuable information to guide clients on immunization protocol. This information was needed because the estimates of incidence that practitioners report to their clients varies from 1 in 250 vaccinated cats to 1 in 10,000, and there are many risk factors that have been postulated without evidence.

There are several reasons for the variation of reporting incidence rates. Determination of precise annual incidence of this disease is hampered by the fact that there is no accurate way to determine if a vaccine caused a sarcoma, because there are no pathognomic histological changes.

Although aluminum may provide a marker to indicate a sarcoma occurred at a vaccination site, no marker confirms causality. In addition, many suspected sarcomas are not histologically confirmed. Reports that originate from referral hospitals start with a population bias that skews the incidence rate higher. Granted, incidence is a particularly difficult figure to nail down when dealing with a relatively rare disease, but even with that in mind, the variation noted above should be decreased.

A 1993 epidemiological study4 found an incidence of one- to two-cases/10,000 cats/year. A 1997 study5, sponsored by the American Association of Feline Practitioners, reported the 1992 prevalence of one case/5,000 cats. Note that prevalence refers to the number of cats with vaccine site-associated sarcomas/total population and incidence refers to the number of new cases/the population at risk in a given time period. A recent study2 used a World Wide Web-based survey to gather data from 93 veterinarians on 31,361 vaccinated cats. Data were gathered from Jan. 1, 1998 to Dec. 31, 2000 and allowed for a one- to three-year follow-up of vaccinated cats. To determine incidence, the following criteria were met for the case to be considered epidemiologically related to the vaccination: The vaccine had to have been given during the study period by the veterinarian participating in the study. Also, the reaction had to be reported during the two-year study or the one-year follow-up. The results indicate that postvaccinal reactions were reported on 73 of these cats, yielding a reaction rate of about 12 reactions/10,000 vaccine doses. Of course the actual number of reactions was probably higher because clients do not always report reactions.

Measuring Vaccine-Associated Sarcoma Incidence

Follow-up information to reaction resolution was available on 54 of the cases with benign reactions. More than half resolved in less than a month and 90 percent of the reactions were 3 centimeters or less. There is no reason to remove these masses unless their size is increasing over time, because most of them come up quickly, slowly get smaller and resolve on their own.

One of the 73 reactions progressed to sarcoma. It was first noticed as a 1 cm by 1 cm swelling five days after vaccination and progressed to a 5 cm by 8 cm fibrosarcoma one month after vaccination. Another reaction diagnosed as a sarcoma was not reported by the client, but was noticed by the veterinarian a year later when the patient was presented for annual examination.

Study participants submitted 21 sarcoma case reports during the two-year study period and 14 sarcoma case reports during the one-year follow-up period, but only the two reports mentioned previously met the criteria necessary to contribute to statistical incidence for vaccine site-associated sarcomas. Twelve cases involved sarcomas not located at the site of vaccination. Ten cases were censored because they involved vaccinations that were not administered during the study period. One was censored because the participating veterinarian did not administer the vaccine, and one was censored because the sarcoma developed prior to the study period. Nine case reports from one veterinarian were censored because of failure to submit a biopsy report and vaccination history documentation. Such necessary information about the exclusions underscores how difficult it is to obtain accurate incidence information.

The incidence of postvaccinal injection-site sarcomas during the one- to three-year period of observation was 0.63 sarcomas/10,000 cats vaccinated and 0.32 sarcomas/10,000 doses of vaccine administered. That is about one tumor per 16,000 cats vaccinated. The actual incidence rate is, of course, not precisely predicted by this one study. Incidence rates in the range of one sarcoma per 10,000 to 30,000 cats are compatible with the reported results. It is unlikely that the actual incidence rate is more common than one sarcoma per 5,000 cats regardless of the sensitivity analysis used to evaluate the data.

In addition to vaccination, what factors might influence the development of sarcomas? People have speculated that a wide variety of factors could contribute to tumor formation. As an example, one logical theory is that increasing inflammation could predispose the patient to cancer. Since a large bore needle would logically cause more inflammation than a small bore needle, then use of a large bore needle could be associated with an increased sarcoma rate. To study this speculation and others, a large, prospective multicenter case-controlled study was performed. The complete procedure and results of this study have been published in the Nov. 1, 2003 issue of the Journal of the American Veterinary Medical Association.

The Multicenter Study Conclusions

The study case selection included cats with a vaccination-site sarcoma diagnosed between Jan. 1, 1998 and June 15, 1999. Although most cats had fibrosarcomas, other sarcomas that fulfilled the criteria included malignant fibrous histiocytomas, myxosarcomas, myofibrosarcomas, liposarcomas, leiomyosarcomas, chondrosarcomas and osteosarcomas. The study included two groups of control cats: one group had basal cell tumors and the other had sarcomas in locations (such as head, ears, digits, ventral abdomen) not commonly used for injection. The findings of the study are based on the assumptions that 1) tumors of the control cats are unrelated to vaccination and 2) tumors of the case cats are possibly related to vaccination and other factors.

Cats were divided into groups depending on several factors, including if there had been a vaccination given at the site of the sarcoma and if so, how long it had been since the vaccination was given when the sarcoma was diagnosed. These groups were evaluated in several different timeframes. The following were found to be significantly different in case cats when compared to control cats:

Sex – the case cat group had more females (57 percent) and the control cat group had more males (53 percent).

Traumatic injury – 16 percent of case cats had a history of trauma at the tumor site compared with 9 percent of the control cats.

Long-acting penicillin and methyl prednisolone – Looking back two to three years before tumor diagnosis revealed that some of the case cats and none of the control cats had received these injections.

Temperature of vaccine at time of administration – Administration of cold vaccine was associated with a higher risk of sarcoma development than room temperature vaccine, with an odds ratio of about two.

The following were not found to be significantly different in case cats as compared to control cats: neutering status, concurrent FIP, FIV or FeLV infection, exposure to other cats in the same household, exposure to other treatments (fluids, ketamine, acepromazine, diazepam, butorphanol, praziquantel, lidocaine, acepromazine, dexamethasone, prednisone, prednisolone, ampicillin, amoxicillin, enrofloxacin, non-long-acting penicillin), vaccine brand, vaccine adjuvant, multi-dose vials – shaken or not, mixing vaccines in the same syringe, reuse of syringes, syringe brand, needle gauge and massage of vaccination site shortly after injection.

Additional population-based studies will be needed to confirm these findings and to evaluate new vaccine products purporting to be safer. Until these products get significant market penetration, they cannot be effectively evaluated because of the low incidence of the disease. The performance of large, controlled epidemiologic studies will be the only way to determine if these products are doing more good than harm.

Dr. Kass received his veterinary medical degree in 1983 from the University of California (UC-Davis) School of Veterinary Medicine. He then earned a master's degree in statistics and Ph.D. in epidemiology from UC-Davis. He became a diplomate of the American College of Veterinary Preventive Medicine in 1988. His post-doctoral work was in environmental epidemiology, which he completed in 1990 at the University of California, Los Angeles.

Dr. Kass is an associate professor at UC Davis School of Veterinary Medicine's department of population health and reproduction. His general research area is quantitative epidemiology. Special interest areas are non-experimental inference, epidemiologic methodology and analysis, epidemiology of environmental hazards of animals and humans, adverse health events from vaccinations and companion animal epidemiology.