CORR Insights®: Defective HLA Class I Expression and Patterns of Lymphocyte Infiltration in Chordoma Tumors

Where Are We Now?

The finding that inhibition of immune checkpoints can be used to help a patient’s immune system attack tumor cells has led to the development of a whole new family of anti-cancer drugs. Great advances in the treatment of some types of cancer have been made in the past decade, and the discovery of this concept was honored with the Nobel Prize for Physiology or Medicine in 2018 [4]. Although therapy with human antibodies such as nivolumab, atezolizumab, or ipilimumab against the involved cell surface structures of programmed cell death protein 1 (PD-1), its ligand PD-L1, or cytotoxic T-lymphocyte-associated protein 4 has become part of the standard therapy against malignant melanoma, renal cell carcinoma, and others, there is hope that this concept may also be effective against soft tissue sarcomas.

Chordoma is rare, as are other sarcomas of soft tissue and bone, and as such, they are often not the main focus of drug development. The standard therapy has been surgical removal; recently, particle radiation therapy has also become more common [5]. If the tumor is in the axial skeleton near the brain or spinal cord, however, resection or radiation is often incomplete, and tumor growth can lead to severe pain, loss of function, and death. We need new—and better—treatment alternatives.

Chordoma, osteosarcoma, and other bone malignancies express PD-L1, and are thus potential targets for PD-L1 or PD-1 antibodies [2]. Furthermore, in this month’s Clinical Orthopaedics and Related Research®, Patel et al. [6] demonstrated tumor-infiltrating lymphocytes (TILs) as well as loss of human leukocyte antigen (HLA) Class I expression. TILs are part of the effector cells once the immune checkpoint is successfully inhibited, and loss of HLA Class I expression may explain how tumor cells can hide from the immune response. Taken together, many requirements are fulfilled to make checkpoint inhibition for chordoma treatment seem sensible. However, it is not yet clear whether the available checkpoint inhibitors are effective against chordomas and under which circumstances they may be treatment options. Clinical trials with small patient numbers started in 2017 [1], and preliminary results might be reported in 2021. Based on the discoveries of Patel et al. [6], surgeons, radio-oncologists, oncologists, and pathologists should be aware of new treatment options that are currently underway or can be expected in the future. Chordomas, especially those with uncertain resectability, should be discussed in interdisciplinary tumor boards preoperatively.

Where Do We Need To Go?

The topic of checkpoint inhibition in sarcoma therapy should be addressed from both a basic research and practical therapeutic point of view. Regarding tumorigenesis and the tumor’s immune response, there are a number of important unanswered questions, including: What are the mechanisms behind the reduced expression of HLA Class I in tumor cells? To which degree do tumor cells express PD-L1, and what is the mechanism behind it? How are TILs attracted to the tumor despite HLA Class I loss and PD-L1 overexpression, and how do TILs react to the tumor? This is an incomplete list of topics that would be worth investigating using in vitro studies.

From a practical and therapeutic point of view, we need to discover whether checkpoint inhibition against chordoma and other sarcomas of soft tissue and bone is effective and safe and can be a treatment alternative in the future. Although I believe complete resection will remain the best therapeutic option in operable tumors for the foreseeable future, checkpoint inhibition may be used in inoperable, recurrent, or metastasized tumors in an adjuvant and maybe even in a neoadjuvant setting. Apart from discovering whether checkpoint inhibition has an effect on chordomas at all, we will need to determine which single drug or combination of drugs provides the most benefit. While a single checkpoint inhibitor may not be effective, a combination of substances or a combination of chemotherapy or radiotherapy might have synergistic effects, as has recently been shown in small-cell lung cancer [3].

Furthermore, we need to determine whether the response to checkpoint inhibitors depends on certain tumor characteristics. Will it be important to assess the levels of HLA Class I and/or PD-L1 expression in the tumor before making a treatment decision? If the answer is yes, we need to determine which test should be used, whether immunohistochemistry will have sufficient reproducibility and selectivity to assess the levels of HLA Class I and PD-L1 expression, and which immunohistochemical antibodies will deliver reliable results. What percentage of tumor cells need to be positive, and what cutoff is suited to predict the treatment outcome? Does the number of TILs have a predictive value, and should it be assessed before treatment? Is the type of TIL important, for example, the CD4/CD8 ratio or the presence of regulatory T-cells? Does PD-1 expression on TILs influence the outcome? These are some important questions that need to be answered for everyday practice, once we have proof of principle that chordomas respond to checkpoint inhibition.

How Do We Get There?

Most basic research has been performed on formalin-fixed, paraffin-embedded tumor samples. These samples are readily available in most hospitals, and histologic techniques such as immunohistochemistry are well-established. However, when using experimental primary antibodies and tissue with variable quality (such as over-fixed, under-fixed, decalcified, old, or improperly stored tissue), immunohistochemistry may show background staining or be false-negative and as a consequence deliver misleading findings. Therefore, it is advisable to also collect snap-frozen tissue samples (in which fresh tissue samples are frozen very rapidly, mostly using liquid nitrogen, to preserve tissue), whenever possible during chordoma surgery. Formalin-fixed paraffin-embedded and snap-frozen samples may both be used to examine the pathogenesis of the loss of HLA Class I expression and PD-L1 overexpression. It will be interesting to see whether these changes are caused on the level of DNA (mutation or methylation), RNA (transcription errors or splicing), or the protein itself (increased or delayed degradation). These examinations can be performed using next-generation sequencing, RNA expression profiling, and northern and western blotting. Recently, a study group was able to create organoids from patient-derived chordoma cells, which can be used for further in vitro studies to address the above topics [8].

To go from in vitro studies to clinical trials, the next step will be testing checkpoint inhibition in animal models. An orthotopic human-chordoma xenograft murine model has been described [7] and may be useful for testing various checkpoint inhibitors to find candidates for application in patients with chordomas.

Checkpoint inhibitors are already being used in clinical trials of other soft tissue malignancies such as osteosarcoma [9]. Because many checkpoint inhibitors are already available and have been used successfully in other tumors, it is justified to perform trials for chordoma as well [1]. Checkpoint inhibitors may be useful when surgical resection is incomplete or impossible, or when the tumor has already metastasized. Ideally, patients would be enrolled in double-blinded prospective trials and tested against standard therapy, the endpoints being progression-free survival, overall survival, and quality of life. The latter seems especially important, because checkpoint inhibitors may not eliminate the tumor but downsize it so that the brain and spinal cord are relieved and pain and neurologic symptoms are reduced. Because chordoma is a rare tumor, such studies should be undertaken in large centers or multinational research networks. A central pathology review in clinical trials on this topic seems advisable in order to obtain reliable and comparable parameters regarding the expression of PD-1, PD-L1, HLA Class I, and the number of TILs as a solid basis for exploratory research.