Commentary

Dear editors, we read with interest the article by Joerger and colleagues recently published in the Journal for ImmunoTherapy of Cancer.1 The authors reported the results of a phase 1 dose escalation study of ANV419, an antibody-cytokine fusion protein consisting of interleukin-2 (IL-2) fused to an anti-IL-2 monoclonal antibody, which limits the binding of IL-2 with the α subunit of its receptor. ANV419 has been designed to stimulate the expansion of CD8 T lymphocytes and natural killer (NK) lymphocytes while limiting that of regulatory T lymphocytes (Treg). The objective was to increase the antitumor response of cytotoxic cells and to restrict the activity of immunosuppressive and pro-tumoral Treg.2

In this study, the authors determined the maximum tolerated dose and the recommended phase 2 dose by assessing the occurrence of overall adverse events. They also investigated the pharmacodynamic effects of ANV419 on the proliferation of CD8, NK, and Treg lymphocytes 4 days after intravenous administration, measuring the expression of Ki67 by lymphocytes. At a dose of 243 µg/kg every 2 weeks (Q2W), they observed mainly fever (87.5%) and chills (62.5%) in patients. These symptoms occurred frequently during the first day of treatment and have been categorized as infusion-related reactions in 45% of patients or cytokine release syndrome in 30% of patients. The authors observed transient asymptomatic lymphopenia on day 2, which resolved between days 4 and 8. The pharmacodynamic analyses showed proliferation rates of 75.4% for CD8+T cells, 89.2% for NK cells, and 48.6% for Tregs in blood at 243 µg/kg Q2W, suggesting an efficient preferential stimulation of whole blood CD8+T and NK cells more than that of Tregs. Considering objective antitumoral responses per Response Evaluation Criteria in Solid Tumours (RECIST) v.1.1, at ANV419 doses ≥108 µg/kg, 64% of patients had at least stable disease as the best response (15 stable disease and 1 confirmed partial response).

Several pieces of information were not available in the publication and could be interesting to interpret results. In the study, patients had received between one and eight prior lines of systemic therapy, including immunotherapy in 75%, which were probably anti-programmed death-1/programmed death-ligand 1 (PD-1/PD-L1) monoclonal antibodies. It would have been important to provide more detailed information about these previous immunotherapies, at least their types and the time between the last infusion of immunotherapy and the first of ANV419. The effect of previous immunotherapies could have an impact on those of ANV419, and some patients probably kept residual anti-PD-1 monoclonal antibodies at the time of the first infusion of ANV419. Frequent eligibility criteria for phase 1 trials allowed the inclusion of patients previously treated by anti-PD-L1 within 28 days or five half-lives before the study treatment. In this study, the exclusion criterion was concurrent therapy with any other investigational drug within 1-month prior to day 1 of study drug administration. Knowing that the half-lives of pembrolizumab and nivolumab are 22 and 25 days,3 4 these antibodies persist at residual concentrations until 4 months after their last administration.

To illustrate this point, we retrospectively measured the residual concentration of nivolumab and pembrolizumab in the serum of patients (n=17) previously treated by immune checkpoint blockade therapies (ICB), included in a clinical cohort (PREMIS; NCT03984318). The mean time on the prior ICB regimen was 32.4 weeks (4.0–82.0). The mean time between the last ICB infusion and the serum collection was 20.9 weeks (2.7–91.3). The residual serum concentrations of anti-PD-1 are reported in figure 1. We observed that the mean±SD residual anti-PD-1 concentrations were 31.16±25.81 and 8.92±8.13 µg/mL in patients with previously acquired and primary resistance. The decreasing concentrations over time of nivolumab and pembrolizumab in these patients align with the known half-lives of both drugs.3 4 These results illustrate the reality of pharmacological data, which are clinically significant.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Residual dosing of previous anti-PD-1 received according to time since the last injection. ICB, immune checkpoint blockade; PD-1, programmed death-1. *Patients who achieved an objective response with the first ICB. Dashed lines represent five theoretical half-lives of the treatments dosed (blue, pembrolizumab; red, nivolumab).

It has previously been described that the proliferation of CD8+T lymphocytes induced by IL-2 treatment is accompanied by their exhaustion and a decrease in their functional capacities, associated with their expression of immune checkpoints such as PD-1, Lymphocyte Activation Gene 3 (LAG3), and T-cell immunoglobulin and mucin-domain containing (TIM3).5 Therefore, the combination of IL-2 with ICB treatment may prove essential to maintaining the antitumoral effects of amplified lymphocytes.6 To improve knowledge on proper clinical and pharmacodynamic effects of new immunotherapies, it is necessary to describe: (1) Whether patients were previously treated by ICB; (2) How many times they received it; (3) What was the time between the last ICB infusion and the first of the experimental treatment and; (4) Whether patients developed an acquired or a primary resistance to the previous ICB. As anti-PD-1 monoclonal antibody titrations are now available, it could be also important to determine whether patients kept residual treatment. Considering the current eligibility criteria of clinical trials, some patients treated in phase 1 monotherapy trials received combinations with residual anti-PD-1 treatment, at a concentration that could be associated with immunological and clinical effects.7 8