Comparison of the efficacy of local treatment of equine sarcoids with IL-2 or cisplatin/IL-2

Animals and sarcoids

A total of 46 sarcoids in 35 horses and 1 donkey presented over a 6-year period (1995–2000) at the Faculty of Veterinary Medicine, Utrecht University. The mean age of the animals was 5 years (1 to 14 years). There were 14 mares, 17 geldings and 5 stallions. The breed distribution was 21 Warmbloods, 3 Thoroughbreds, 1 Quarter horse, 2 Friesians, 5 Fjord ponies, 2 Welsh ponies, 1 Haflinger pony, and 1 donkey. The donkey and 27 of the horses had only a single sarcoid. The remaining 7 horses had multiple (up to four) tumours. The tumours were clinically classified into five categories: verrucous, fibroblastic, mixed verrucous and fibroblastic, occult, and nodular. The diagnosis was histologically confirmed. The base of the tumour was measured in two perpendicular directions. The location of the tumours and their clinical type are summarized in Tables 1 and 2. The mean surface area of the tumours (n=46) was 7.9 cm². The smallest tumour had a surface area of 0.36 cm² (0.6×0.6 cm), the largest a surface area of 30 cm².

Treatments

The 36 animals presented over the 6-year period were divided into three groups. The first 11 horses referred formed the first group. The 12 sarcoids of these 11 horses were treated intratumorally daily for five consecutive days with 200,000 IU IL-2 (a kind gift from Chiron, Amsterdam, The Netherlands) dissolved in haemaccel 2% and 0.9% NaCl (total dose 1 MIU of IL-2). The second group consisted of the next ten horses referred. The 12 sarcoids in these ten horses were treated intratumorally with 200,000 IU of IL-2 for ten working days over 2 weeks (total dose 2 MIU IL-2). The next 14 horses and 1 donkey formed the third group. On day 0 the 22 tumours in this group were injected with 1 mg cisplatin (Platinol; Bristol-Myers Squibb, Woerden, The Netherlands) per square centimetre of tumour. The cisplatin was dissolved in water at a concentration of 2 mg/ml. On day 10, 2 ml IL-2 (4.5×10⁶ IU/ml ) was injected intratumorally (total dose 9 MIU IL-2). The 10-day interval was based on previously described procedures [2, 3, 14, 15]. In the three groups fixed IL-2 doses (5 or 10×200,000 or 4.5×10⁶ IU) were used independent of the size of the tumour, as similar locally administered doses of IL-2 have been shown to lead to similar curative effects against small and large tumours [9, 10, 23]. The mean sizes of the tumours in the three treatment groups were similar (7.75, 8.25 and 7.67 cm², respectively). This was expected as the three groups comprised all eligible animals in three successive periods. The sarcoids were treated without surgical excision except in the donkey. This animal presented with a periocular sarcoid that contacted the cornea causing ulceration. A large part of the tumour was removed because the expected swelling after intratumoral IL-2 injection would have worsened the ulceration of the cornea. During surgery (day 0) the remnant of the tumour was injected with cisplatin (1 mg/cm² of tumour), and on day 10 with 2 ml IL-2 (4.5×10⁶ IU/ml ).

Depending on tumour shape and location the field-block technique or the parallel-row technique was used for injection [39]. Special procedures for proper handling and disposal of cisplatin were adhered to. Surgical gowns, caps, masks, gloves and safety glasses were used.

Biopsies and histology

Before treatment a biopsy was taken. A second biopsy was taken in both IL-2 groups at 10–15 days after the last IL-2 injection. In the cisplatin/IL-2 group, biopsies were taken on day 15. The biopsies were fixed in formalin and 10-μm sections were stained with haematoxylin eosin. Biopsies before and after treatment were compared. The relationships between the histological type of the sarcoid and sensitivity to treatment, between the size of the tumour and the effect of treatment and between leucocyte infiltration and the effect of treatment were investigated, and the possible mechanism of tumour regression was also investigated. The follow-up period varied from 1 to 2 years.

Statistics

The results were classified according to the degree of the therapeutic effect. Four categories of degree of regression are defined in Table 3. Different treatment groups were statistically analysed using Student's t-test. P values <0.05 were considered statistically significant. Any short- or long-term side effects were noted. Scoring of side effects was based on clinical observation.

Histology of untreated tumours

Histopathological study of the primary tumours confirmed in all animals the diagnosis of equine sarcoid. In all tumours there was a characteristic fibroblastic proliferation. In some tumours the fibroblastic tissue showed a more mature development with large, sometimes hyalinized, collagenous bundles. At the other end of the spectrum, tumours showed a more cellular aspect with more rounded, dysplastic fibroblasts with coarse irregular nuclei and prominent nucleoli; occasionally mitotic figures were observed. The fibroblastic proliferation showed a trabecular pattern. In some sections whorl formations and even regimentation of nuclei in a linear array were present as can be observed in neurofibroma. The epithelial covering of the tumours changed from ulceration of the atrophic hyperkeratotic epithelial lining to various patterns of mixed variants with enlarged rete pegs, verrucous-like papillary formations and pseudoepitheliomatous hyperplasia. A moderate round-cell infiltrate was present in almost all sections.

Therapeutic effects

Therapeutic effects (complete regression, CR, plus partial regression, PR) are summarized in Table 4. Intratumoral administration of IL-2 with and without cisplatin caused tumour regression. The results in the three different treatment groups at follow-up were compared. Optimal therapeutic results were found 12 months after treatment. Therapeutic effects were observed in 36%, 50% and 80% of the patients after treatment with low-dose IL-2 for 5 and 10 days, and with cisplatin/single high-dose IL-2, respectively. The differences in the CR rate were even more pronounced: 18% and 10% of tumours treated with IL-2 for 5 and 10 days, respectively, and 53% of tumours treated with cisplatin/IL-2. The results in terms of both the therapeutic effect (P=0.02) and CR rate (P=0.02) were better after cisplatin/IL-2 than after the low-dose IL-2 regimens. The therapeutic effects of cisplatin/IL-2 were significantly greater than those of low-dose IL-2 for 5 days (P=0.03), and the CR rate following cisplatin/IL-2 treatment was significantly higher than following the 10-day IL-2 treatment (P=0.001).

Tumour size had no influence on therapeutic effect. Indeed, the CR rates seemed even higher in larger tumours (>10 cm²) than in small or medium ones (50% versus 29–31%; difference not significant). Ineffective earlier surgery and/or cryosurgery performed in primary clinics did not influence treatment outcome, as the differences in results were only marginal (<5% difference in therapeutic effect).

Table 5 indicates that both the CR rate (50% versus 27%) and the therapeutic effect (67% versus 57%) were better in horses with two or more tumours than in those with a single tumour. However, these effects were not significant. All tumour types listed in Table 2 appeared to be sensitive to therapy. Two of two occult tumours (100%) showed a CR, but this was not significantly different from other tumours. There were no statistical indications that a particular tumour type was relatively more sensitive than other tumour types.

Side effects

Short- and long-term side effects were noted. The side effects of the 5-day and 10-day IL-2 treatments were minimal. The side effects varied from tenderness, through erythema to slight oedema. Localized swelling lasted for less than 3 days after treatment. The side effects in the cisplatin/IL-2 group were more serious and lasted longer. The side effects varied from tenderness, through bright erythema, to slight to moderate oedema with patchy crusting. In 64% (14 out of 22 ) of the treated tumours, symptoms disappeared within 30 days. In the remaining 36% these side effects gradually vanished over 60 days.

Histology of the IL-2- and cisplatin-treated tumours

In untreated sarcoids there was often an inflammatory response particularly if the tumour was ulcerating. This inflammation was slightly increased after IL-2 treatment, but was never abundant. The limited changes that were observed included a more pronounced diffuse spread of lymphocytes, occasional clusters of eosinophils, and focal oedema in the tumour with fragmentation of tumour tissue and apoptosis of tumour cells.

After treatment with cisplatin/IL-2, the inflammatory features were far more pronounced. In 3 out of 15 animals a massive eosinophilia spread through the tumour, especially near blood vessels (Fig. 1). Also the perivascular aggregation of lymphocytes was more pronounced when compared with nontreated or IL-2-treated sarcoids. Even moderate plasma cellular infiltration was occasionally observed, often in combination with eosinophilia. Angiogenesis was more obvious and in some sections there was clear endothelial proliferation. Focal oedematous areas were observed, but also more widespread oedema in the tumour, with splitting up of tumour tissue and tumour degradation. In addition apoptosis was present.

Therapeutic effects

This study showed that combined local treatment of equine sarcoids with cisplatin and a single high dose of IL-2 resulted in a more effective antitumour response than five or ten low doses of IL-2. One might question these therapeutic effects, as spontaneous regression of sarcoid has been reported [37]. However, these spontaneous regressions always occur in the first 6 months after the initial diagnosis. This was also the case in the references cited in reference 37, and this was predicted by Vanselow et al. [42]. All these data are in line with our experience in our second-line veterinary health institute: we do not see spontaneous CR or PR. So, the CR and PR we observed were really due to our therapeutic interventions.

The CR rates and therapeutic effects were non-significantly better in horses with two or more tumours than in those with one tumour (73% versus 48% and 100% versus 73%, respectively). Nevertheless, these results suggest that rejection of different tumours in one animal may not be independent, and thus therapeutic results should be analysed preferably per horse, and not per tumour. Better therapeutic results in horses with more tumours could be due to concomitant rejection of tumours [5] or the generation of systemic immunity after IL-2 therapy [24].

There seemed to be a correlation between the amount of cells infiltrating the tumour and the severity of side effects (oedema and erythema). This is not surprising, as cellular infiltration, erythema and oedema are all due to the induction of an inflammatory reaction.

Comparison of therapies

A variety of therapeutic strategies have been used to treat sarcoids. These include surgical methods (excision, CO₂ laser and cryosurgery), chemotherapy (cisplatin) and immunotherapy (Baypamun P, different forms of BCG therapy) [22, 25, 37, 39].

Cisplatin/IL-2 therapy gave significantly better results than placebo or Baypamun P used as an immunostimulant [37]. The therapeutic effects of cisplatin/IL-2 did not differ significantly from most treatments in the cited studies. The exceptions seem to be cisplatin therapy in the study reported by Théon et al. [39] and cryosurgery in the study reported by Klein et al. [22]. Cisplatin (administered four times) may have given significantly better CR rates than our milder regimen of cisplatin/IL-2 treatment, but the former was more toxic. Klein et al. [22] preselected patients prior to cryosurgery based on a prediction as to whether cryosurgery would be effective, but in our present study we did not make such a preselection. In our cisplatin/IL-2 group we had 28 tumours, of which 16 had not been treated previously and 12 had been treated surgically without success. CR rates in these 16 and 12 animals were 42% and 44% and therapeutic effects occurred in 67% and 69%, respectively, indicating that cisplatin/IL-2 therapy was equally effective in both tumour groups. This shows that cisplatin/IL-2 can be effective, even when surgery is not effective.

In contrast to BCG immunotherapy [22], we did not find a worse response in large tumours (>10 cm²). Actually, CR rates appeared to be even higher in these larger tumours than in small (<5 cm²) or medium (5–10 cm²) ones (50% versus 29–31%), although this difference was not significant. Cisplatin/IL-2 had less-severe side effects than usually seen after treatment with BCG.

Histology of sarcoid regression

Various patterns of IL-2-induced tumour regression have been described. These are an immediate regression in which the tumour disappears within a few days due to stasis of the circulation in the tumour, causing extensive necrosis [7, 23], and a more tardy regression which takes months up to a year, which has been observed in bovine ocular squamous cell carcinoma [10, 35]. The histological changes after IL-2 treatment of sarcoid in the present study were of the slow reaction pattern. The histological changes in these tumours 10–15 days after the last IL-2 injection were not spectacular. The lymphocytic-plasma cellular response never dramatically increased compared with the response in untreated sarcoids. This would be expected because lymphocyte accumulation is always a quite late phenomenon after stimulation. The histopathology after cisplatin/IL-2 treatment was far more pronounced, showing a moderate up to sometimes a very impressive (3 out of 15 animals) eosinophilic infiltrate and also more plasma cells. It is known that IL-2 immunotherapy for neoplastic disease induces peripheral eosinophilia [31, 36, 41]. Eosinophils isolated from IL-2-treated cancer patients exert in vitro cytolytic activity against tumour targets [31]. Hypodense eosinophils induce tumour cell apoptosis by a granzyme B-dependent mechanism [6]. We do not have an indication that this eosinophilia stimulated tumour regression, but several clinical reports have suggested a correlation between tumour-associated eosinophilia and increased survival time [19, 30].

Conclusions

In the study reported here, we showed the applicability and effectiveness of local multiple low doses of IL-2 in equine sarcoids. Local cisplatin/single high-dose IL-2 treatment was more effective than the low-dose IL-2 regimens tested. Clinically, this treatment showed less-severe side effects than BCG or repeated local cisplatin treatments [39]. This treatment was at least equally applicable for large tumours as for small tumours, and can be used for equine sarcoids that cannot be removed surgically. Histological changes were pronounced after cisplatin/IL-2 treatment and sometimes showed an enormous eosinophilic infiltrate.