Rethinking immunotherapy in meningiomas

See article by Yeung et al. pp 1922–1935.

As the most commonly occurring tumor in the central nervous system (CNS), meningiomas account for 38% of all CNS tumors and half of all nonmalignant tumors.¹ Although about two-thirds of meningiomas are benign and classified as WHO grade I, 20%–30% are grade II and III, which usually are more aggressive in their tumor behavior. In some cases of grade III meningiomas, particularly those cannot have a gross total resection, the overall survival rate can be barely over a year.² With the current therapeutic approaches, the overall 10-year survival of grade I, II, and III tumors are 84%, 53%, and 0%, respectively.^1,3 Although grade I meningiomas are considered benign, not all tumors can have gross total resection, largely due to the location of the tumor or the involvement of the venous sinus or neurovascular tissue.³ These tumors often cause debilitating neurological deficits, such as ataxia, visual impairment, cranial nerve palsies, and exophthalmos. In addition, these typically slow-growing meningiomas can even have a linear growth rate of 2-4 mm/year.³ Although a third of all meningiomas show no growth, about one-fourth may experience exponential growth.⁴ These facts highlight the importance of effective and individualized treatments in meningiomas.

Surgery has been the primary therapeutic approach for symptomatic meningiomas. Radiation therapy is reserved for unresectable meningiomas or those likely to have disease recurrence even after the gross total resection.⁵ Unfortunately, once patients develop a progressive disease that no longer responds to surgery or radiation therapy, there are no effective systemic therapies that can be offered to these patients. In recognizing the unmet clinical need, clinical efforts have been made to develop new therapies. While clinical trials of targeted therapies based on discoveries made by several large-scale genomic profiling have been developed,^6,7 immunotherapy has been an emerging interest in the field of neuro-oncology. As summarized in Table 1, most of the ongoing immune therapy clinical trials in meningiomas are focused on targeting the PD-1/PD-L1 axis. In this issue of Neuro-Oncology, Yeung et al reported that targeting the CSF1/CSF1R axis is a potential treatment strategy for malignant meningiomas.⁸

Table 1.

Ongoing Immunotherapy Clinical Trials for Meningioma

Study Treatments	Therapeutic Target	Disease Type	Clinical Trial ID
Nivolumab	PD-1	Recurrent, grade II, III	NCT03173950
Nivolumab ± ipilimumab	PD-1 ± CTLA-4	Recurrent, grade II, III	NCT02648997
Nivolumab/SRS ± ipilimumab	PD-1 ± CTLA-4	Recurrent, grade II, III	NCT03604978
Pembrolizumab	PD-1	Recurrent, grade II, III Recurrent, grade II, III	NCT03016091 NCT03279692
Pembrolizumab w/SRS	PD-1	Recurrent, grade I, II, III	NCT04659811
Avelumab w/Proton radiotherapy	PD-L1	Recurrent, grade I, II, III	NCT03267836

Abbreviations: CTLA-4, cytotoxic T lymphocyte-associated antigen 4; PD-1, programmed death-1; PD-L1, programmed death-ligand 1; SRS, stereotactic radiosurgery.

It has been extremely challenging to develop preclinical studies in meningiomas, largely due to the lack of syngeneic animal models to recapitulate the biology and the immunology of meningiomas. Using a novel syngeneic mouse Nf2-mutant meningioma model MGS1, Yeung and colleagues demonstrated that immunotherapy targeting the CSF1/CSF1R axis, but not the PD-1/PD-L1 axis, has the potential to bring a clinical benefit to meningioma patients.⁸ The authors showed that MGS1 recapitulates the immune landscape of human meningioma with heavy infiltration of M2-like macrophages that express PD-L1. However, surprisingly, anti-PD-1 treatment provided no survival benefit to tumor-bearing mice even when it was combined with anti-4-1BB, which facilitates T-cell activation, to take a push-pull approach. In contrast, blockade of CSF1 significantly prolonged the survival of tumor-bearing mice. CSF1 is produced by both myeloid cells and tumor cells, while its receptor, CSF1R is differentially expressed only on myeloid cells. Thus, similar to previous observations in glioblastoma, the CSF1 blockade shifted the landscape of myeloid cells in tumor tissues by reducing M2-like immunosuppressive myeloid cells and increasing proinflammatory dendritic cells.⁹ However, it did not reduce the number of myeloid cells presumably be due to the production of myeloid cell growth factors such as GM-CSF (granulocyte-macrophage colony-stimulating factor) by tumor cells. While the lymphocytes can be attracted by proinflammatory myeloid cells and thus induce PD-L1 expression through IFN-γ production, it is unclear whether PD-L1 expression in meningioma cells was promoted by shifting the myeloid cells toward proinflammatory population. Furthermore, PD-1 blockade did not show any benefit to the survival of tumor-bearing mice even in the combination of CSF1 blockade. The authors also made an observation that meningioma patients had significantly higher levels of plasma CSF1, suggesting the potential of using plasma CSF1 levels as a biomarker of the disease.

Some significant discordance between the mouse model used in the preclinical study and the human disease should be noted. In human meningiomas, PD-L1 expression has been reported in multiple clinical studies, while MGS1 did not express it¹⁰. Although it is unclear whether the contribution of PD-L1 on tumor cells has a negative impact on tumor immunity, it is worth mentioning. The results of this study were mostly with subcutaneous tumors, rather than an orthotopic model. Although meningiomas usually develop outside the blood-brain barrier and the immune system should have good access to the tumor tissue, the impact of tissue-resident immune cells on the immune responses can be different between subcutaneous and orthotopic models. How much the observations made with subcutaneous tumors are relevant to orthotopic tumors remain to be elucidated.

The study by Yeung and colleagues reinforces a concept that modulation of myeloid cell populations may be the key for success to treat myeloid cell-enriched tumors, such as meningioma and glioblastoma. It is interesting to find the result of ongoing clinical trials of CSF1/CSF1R-targeted therapies. The study also reminds us that immunotherapy targeting PD-1/PD-L1 is not a magic bullet and that the expression of PD-L1 in tumor tissue is necessary but not sufficient for PD-1/PD-L1 targeted therapies to work. There are many other immune checkpoints, and there may be a hierarchy among them. The novel mouse model used in the study may also provide an opportunity to address this critical question, which may provide new insights for immune checkpoint inhibitor therapies.

Funding

This work is supported by the intramural program of the National Institutes of Health, National Cancer Institute, Center for Cancer Research, and Lasker Clinical Research Scholar Program (1 SI2 CA228571-01 to J.W.).

Conflict of interest statement. The authors declare no conflicts of interest.