Infiltration of Innate Immune Cells into Tumors By Polatuzumab Vedotin Treatment Contributes to Its Antitumor Effect in Xenograft and Syngeneic Mouse Models

Background

Polatuzumab vedotin (Pola) is an antibody–drug conjugate (ADC) comprising an anti-CD79b antibody linked with monomethyl auristatin E (MMAE). It is approved in several countries for first-line and relapsed/refractory treatment of diffuse large B-cell lymphoma (DLBCL). Vedotin-based ADCs, including Pola, might modify the tumor microenvironment due to MMAE-induced immunogenic cell death (Heiser RA, et al., Mol Cancer Ther. 2024; Gray E, et al., J Immunother Cancer. 2023). Also, a recent study suggests that the lymphoma microenvironment, including both innate and adaptive immune cells, could impact patient prognosis (Kotlov N, et al., Cancer Discov. 2021). However, the interplay between the anticancer effect of Pola and the tumor immune microenvironment in DLBCL remains to be elucidated.

In this study, we focused on innate immune cells in the tumor microenvironment, specifically natural killer cells (NK) and macrophages (MΦ), and explored the relationship between the antitumor effect of Pola and immune status.

Methods

Scid mice and NOG mice were subcutaneously inoculated with human DLBCL cells (DB or #5-1 cells). #5-1 cells show low in vivo sensitivity to Pola; they were developed from DB tumors via continuous Pola administration to a scid mouse. DBA/2 mice were subcutaneously inoculated with L1210-hCD79b cells (mouse L1210 cells expressing chimeric CD79b in which the Pola-recognizing site is that of human CD79b). In these models, human IgG or Pola (2–5 mg/kg) was injected intravenously on day 1, and clodronate liposomes or anti-asialo GM1 (reagents that deplete MΦ and NK cells, respectively) was injected regularly from day 0.

Results

In the scid mouse model, significant antitumor effect was observed by TUNEL assay on day 4 and regression of DB tumors (6 of 6 mice) was observed on day 22 in the 2 mg/kg Pola-treated group. Also, numbers of infiltrating MΦ and NK cells assessed by immunohistochemistry (IHC) were higher in the Pola-treated group than in the control IgG group on day 4. However, Pola’s antitumor effect was diminished in mice administered clodronate liposomes or anti-asialo GM1.

Ex vivo MΦ isolated from tumors induced apoptosis in co-cultured DB cells, and Granzyme B expression of NK cells in vivo was detected by IHC. However, the ex vivo MΦ co-culture assay, in vivo evaluation of NK activation, and scRNA-seq analysis showed no discernible upregulation in Pola-treated tumors compared to control IgG-treated tumors in the activation status of MΦ or NK cells. Whereas, in the in vitro studies, Pola induced DB cells to release damage-associated molecular patterns (DAMPs), such as eATP, HMGB1, and calreticulin, characteristic of immunogenic cell death. Also, Pola-treated DB cells enhanced the migration ability of CD45⁺ cells isolated from tumors.

Next, infiltration of MΦ and NK cells into DB tumors after Pola treatment was compared with infiltration into Pola-insensitive #5-1 tumors, in which 2 mg/kg Pola treatment elicited no tumor regression on day 22. Infiltration of MΦ and NK cells into #5-1 tumors after 2 mg/kg Pola administration was significantly less than in DB tumors on day 4. Gene Ontology enrichment analysis based on scRNA-seq data showed that movement-related pathways of MΦ and NK cells were significantly suppressed in #5-1 tumors compared to in DB tumors (FDR = 0.1, BH correction). Interestingly, in the NOG mouse model, which is deficient in mature NK cells and has compromised MΦ function, there was no difference between DB and #5-1 in tumor regression rate with 2, 3, or 5 mg/kg Pola treatment (e.g. tumor regression was observed in 0 of 6 tumors treated by 2 mg/kg Pola in both DB and #5-1 models). This suggests that the different antitumor effects of Pola seen in DB and #5-1 tumors might reflect differences in Pola-induced immune cell infiltration.

Furthermore, administration of clodronate liposomes and anti-asialo GM1 decreased the antitumor effect of Pola (4 mg/kg) in the L1210-hCD79b syngeneic mouse model.

Conclusion

In the DB xenograft model, innate immune cells showed the potential to induce apoptosis of tumor cells and could contribute to the antitumor effect of Pola through their Pola-induced infiltration associated with the release of DAMPs. Moreover, innate immune cells also contributed to the antitumor effect of Pola in the syngeneic mouse model.

Differences in the status of innate immune cells in the tumor microenvironment could affect the efficacy of Pola.