Vaccination-induced autoimmune vitiligo is a consequence of secondary trauma to the skin.

A major concern for cancer vaccines targeting self-tumor antigens is the risk of autoimmune sequelae. Although antitumor immunity correlates with autoimmune disease in some preclinical models, the mechanism(s) linking antitumor immunity and subsequent autoimmune pathology remain(s) to be determined. In the current study, we demonstrated that intradermal (i.d.) immunization with a recombinant adenovirus (Ad) expressing the murine melanoma antigen tyrosinase-related protein 2 (AdmTrp-2) results in a moderate level of tumor protection against the B16F10 murine melanoma without any vitiligo. Similar immunization with an Ad encoding human Trp-2 (AdhTrp-2) resulted in 50-fold greater protective immunity and produced vitiligo in all of the mice, suggesting that the development of autoimmunity may reflect the potency of the vaccine. Interestingly, delivery of AdhTrp-2 by i.m. injection generated protective immunity comparable with that seen in mice that received the vaccine by the i.d. route, but none of the recipients in the i.m. group developed vitiligo. The cellular and humoral responses in the i.m. immunized mice were greater than in the i.d. group; therefore, the lack of vitiligo was not caused by reduced efficacy of the vaccine. These results led us to hypothesize that vaccine-induced vitiligo was associated with local inflammatory responses. Mice immunized i.m. with AdhTrp-2 did develop vitiligo when they subsequently were injected i.d. with either a control Ad vector or complete Freund's adjuvant, suggesting that vitiligo is initiated by some form of trauma within the skin. Our data demonstrated that autoimmune pathology is not an unavoidable outcome of effective cancer vaccines directed against self-tumor antigens.

CTL-dependent and -independent antitumor immunity is determined by the tumor not the vaccine.

Previously, we compared the efficiency of direct injection with an adenovirus (Ad) expressing human gp100 (hgp100) to immunization with dendritic cells (DC) loaded with the same vector ex vivo. The DC vaccine provided the greatest protection against challenge with B16F10 melanoma, and antitumor immunity was found to be CD8(+) T cell-independent. In the current study, we sought to determine whether lack of CD8(+) T cell-mediated antitumor immunity was a function of the vaccine platform or the tumor line. Both Ad and DC/Ad vaccines elicited CD8(+) CTL reactive against hgp100 and provided protection against B16F10 engineered to express hgp100 demonstrating that both vaccination platforms can effectively generate protective CD8(+) T cell-mediated immunity. The hgp100-induced CTL cross-reacted with murine gp100 (mgp100) and lysed B16F10 cells pulsed with mgp100 peptide indicating that the resistance of B16F10 cells to CTL elicited by hgp100 vaccination may be due to a defect in processing of the endogenous mgp100. Indeed, introduction of the TAP-1 cDNA into B16F10 rendered the cells sensitive to lysis by gp100-specific CTL. Furthermore, gp100-immunized mice were protected from challenge with B16F10-TAP1 cells through a mechanism dependent upon CD8(+) T cells. These results demonstrate that tumor phenotype, not the vaccination platform, ultimately determines CD8(+) or CD4(+) T cell-mediated tumor clearance.