Today's Special on the Anticancer Menu: Immunomodulation by Antifolates

In this issue of Clinical Cancer Research, Schaer and colleagues unveil unsuspected immunomodulatory effects of the multitargeted antifolate pemetrexed (1).

Folate metabolism is crucial for both purine and pyrimidine synthesis, hence representing a key process for DNA and RNA metabolism. Accordingly, chemical inhibitors of the enzymes that convert dietary folate to bioactive tetrahydrofolate are particularly toxic for highly proliferating cells, including (at least some) malignant cells (2). Pemetrexed is particularly efficacious in this sense, at least in part owing to its capacity to inhibit the enzymatic activity of three distinct enzymes involved in folate-dependent purine and pyrimidine synthesis, that is, thymidylate synthetase, dihydrofolate reductase, and phosphoribosylglycinamide formyltransferase (GART; ref. 2). Consistent with such a potent biological activity, pemetrexed mediates good clinical activity in patients with malignant pleural mesothelioma and locally advanced or metastatic non–small cell lung carcinoma (NSCLC), and is approved by various regulatory agencies worldwide for these indications. Recently, the immune checkpoint inhibitor (ICI), pembrolizumab, has been shown to ameliorate the therapeutic effects of pemetrexed-based chemotherapy as first-line intervention against chemotherapy-naive, stage IIIB or IV, nonsquamous NSCLC (3, 4), suggesting that pemetrexed may exert some degree of immunostimulation, or at least be compatible with the (re)activation of anticancer immunity by ICIs.

Because LLC cells, a common model of mouse lung cancer, are poorly sensitive to pemetrexed, and commonly employed transgene-driven lung cancer models are not particularly appropriate for tumor immunology studies, Schaer and colleagues set to investigate the effects of various pemetrexed-based (immuno)therapeutic regimens on the immunologic configuration of mouse colorectal cancer MC38 and Colon26 tumors. In this setting, single-agent pemetrexed favored tumor infiltration by immune (CD45⁺) cells, with a particular enrichment of CD8⁺lymphocytes exhibiting markers of active proliferation (Ki67), but had little numerical effects on the myeloid tumor infiltrate. Gene expression analysis confirmed that pemetrexed treatment is associated with increased levels of T-cell activation markers (e.g., Cd8b, Prf1, and Gzma), as well as with transcriptional changes that support tumor infiltration by T cells (i.e., Vegfc upregulation) and limit local immunosuppression (i.e., Nos2 downregulation). Notably, the coadministration of carboplatin (which is often used in combination with pemetrexed in the clinic) or cisplatin appeared to reduce (rather than magnify) the immunostimulatory transcriptional alterations imposed by pemetrexed (1).

Next, Schaer and colleagues evaluated the ability of pemetrexed to synergize with ICIs in the control of MC38 and Colon26 tumors. While MC38 cells displayed mild sensitivity to pemetrexed plus an ICI specific for CD274 (best known as PD-L1), Colon26 tumors were efficiently controlled by the combinatorial regimen, including cases of complete tumor eradication. Gene expression analysis of Colon26 tumors responding to pemetrexed alone, PD-L1 blockage alone, or their combination revealed that the latter had the most prominent impact on the transcriptional tumor landscape, both quantitatively and qualitatively. In particular, the combination regimen (but neither of the agents employed as standalone therapeutics) had a beneficial impact on both lymphoid (cytotoxicity) and myeloid (antigen presentation) functions, which could be confirmed by flow cytometry. Corroborating the mechanistic involvement of the immune system, the therapeutic effects of pemetrexed plus PD-L1 blockage were lost in the presence of a chemical that blocks T-cell egress from lymph nodes (1).

Finally, Schaer and colleagues analyzed the direct effects of pemetrexed on T cells. On the one hand, pemetrexed was shown to impose a mild (but largely reversible) proliferative defect on T cells. On the other hand, pemetrexed was found to improve the respiratory capacity of T cells, which is critical for their ability to mount immunologic memory. Importantly, this was accompanied by increased mitochondrial content, as well as by an improved activation status, cumulatively explaining the ability of pemetrexed-exposed T cells to mount superior cytotoxic responses against malignant cells upon priming (1). Of note, pemetrexed also induced the release of danger signals including high mobility group box 1 (HMGB1) and calreticulin (CALR) by MC38 and Colon26 cells, which the authors interpreted as a sign of immunogenic cell death (ICD; ref. 1). However, HMGB1 is released in a rather nonspecific manner by cells undergoing ICD as well as nonimmunogenic forms of cell death, and soluble CALR (at odds with its membrane-associated counterpart) is known to mediate robust immunosuppressive effects (5). Thus, in the absence of gold-standard vaccination tests assessing the ability of cancer cells responding to pemetrexed in vitro to confer long-term immunologic memory to syngeneic tumor-naïve hosts, whether pemetrexed drives bona fide ICD remains unclear.

Irrespective of this open possibility, the findings from Schaer and colleagues reveal the ability of pemetrexed to boost the metabolism and effector functions of T cells in spite of mild antiproliferative effects (Fig. 1). The negative impact of carboplatin on pemetrexed-dependent immunostimulation casts doubts on current clinical approaches as it suggests that superior therapeutic effects may be achieved with other platinum compounds. In view of the hitherto unconfirmed ability of pemetrexed to drive bona fide ICD, the ICD inducer oxaliplatin stands out as a promising candidate for the development of novel pemetrexed-based therapeutic regimens for combination with ICIs.

L. Galluzzi is an employee/paid consultant for AstraZeneca, OmniSEQ, The Luke Heller Foundation, and Boehringer Ingelheim, and reports receiving commercial research grants from Lytix and Phosplatin. No potential conflicts of interest were disclosed by the other authors.

The L. Galluzzi's laboratory is supported by a Breakthrough Level 2 grant from the U.S. Department of Defense (BC180476P1), Breast Cancer Research Program (#BC180476P1), by a startup grant from the Department of Radiation Oncology at Weill Cornell Medicine (New York, NY), by industrial collaborations with Lytix and Phosplatin, and by donations from Phosplatin, the Luke Heller TECPR2 Foundation, and Sotio a.s.