記住我
Immunosuppressive conditions within the tumor microenvironment hamper the effectiveness of immune checkpoint inhibitors like the anti-PD-L1 antibody, atezolizumab. Novel treatment combinations that make the tumor microenvironment more conducive to immunotherapy are needed. IL-8 plays a role in tumor-associated fibrosis and immunosuppression, but the anti-IL-8 antibody AMY109 reduced fibrosis and decreased immunosuppressive cells in animal studies.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICYIntroductionImmune checkpoint inhibitors have substantially improved outcomes for numerous types of advanced solid cancers.1 However, many patients still do not have adequate responses to these treatments.1 2 Immunosuppressive conditions within the tumor microenvironment (TME) hamper the activity of programmed death-1 (PD-1) and programmed death ligand 1 (PD-L1) inhibitors, reducing their effectiveness.2 Hence, there is an urgent need for novel treatment combinations involving drugs that make the TME more conducive to effective immune checkpoint inhibition.
The proinflammatory chemokine interleukin (IL)-8 is expressed by numerous solid tumors (brain, breast, cervical, colon, gastric, lung, melanoma, mesothelioma, ovarian, prostate, renal, and thyroid) to aid their growth, survival, and spread.3 4 A correlation has been found between high serum levels of IL-8 and tumor progression of several cancer types.3 5 6 High circulating IL-8 levels have been associated with reduced clinical benefit from PD-L1 blockade.5 7 8 Overall survival was shorter after immune checkpoint inhibitor treatment in patients with advanced urothelial, renal cell, and non-small cell lung cancer who had high plasma IL-8 concentrations, whereas low baseline plasma IL-8 levels were associated with improved responses to the PD-L1 inhibitor atezolizumab plus chemotherapy in patients with metastatic urothelial carcinoma.7 8 In the atezolizumab study, IL-8 was primarily expressed in intratumoral myeloid cells. This suggested that suppressing IL-8-mediated myeloid inflammation could be an essential factor in improving the clinical benefit of immune checkpoint inhibitors.
IL-8 facilitates the infiltration into the TME of neutrophils and myeloid-derived suppressor cells (MDSCs), creating an immunosuppressive state.3 7 High levels of tumor-infiltrating and intratumoral neutrophils are predictive of poor prognosis7 9 and anti-PD-1 treatment failure.10 IL-8 in tumors also promotes neovascularization3 and tumor cell epithelial-mesenchymal transition (EMT),6 11 a vital step in tumor progression and metastasis.12 Hence, suppressing IL-8-mediated myeloid inflammation could complement PD-1/PD-L1 inhibitor treatment by suppressing immuno-oncology-resistance pathways to improve clinical benefit.7
IL-8 is also involved in the formation of fibrotic interstitium in endometriotic lesions.13 Highly fibrotic cancer tissues have high densities of collagen, a key component of the extracellular matrix.14 A collagen-rich extracellular matrix inhibits the number of infiltrative CD8+T cells in tumors and downregulates their cytotoxic activity.15 The tumor tissue of patients resistant to immune checkpoint inhibitors has increased levels of fibrosis, EMT, and neovascularization.16
AMY109 is a humanized monoclonal antibody that binds to and selectively neutralizes IL-8. In a monkey model of endometriosis, AMY109 reduced fibrosis and inhibited recruitment of neutrophils to endometriotic lesions.13 In a mouse xenograft model, single-agent AMY109 did not affect tumor growth, but it did decrease polymorphonuclear MDSCs in tumor tissue and inhibited EMT. These effects of AMY109 on the TME in the mouse model suggested that this IL-8 inhibitor might enhance the activity of atezolizumab.
Reducing tumor-associated neutrophils by targeting IL-8, in combination with immunotherapy, has been suggested as a novel treatment combination to improve responses to immune checkpoint inhibitors in different malignancies.17 Hence, this phase 1 study of the safety and clinical activity of AMY109 combined with atezolizumab was conducted in patients with a range of advanced malignancies that included esophageal, colorectal, head and neck, pancreatic, breast, uterine, and ovarian cancers.
Patients and methodsStudy design and participantsThis was a multicenter, open-label, 2-part dose-escalation and -expansion phase 1 study (Japan Registry of Clinical Trials ID: jRCT2080225101) to evaluate the safety, tolerability and pharmacokinetics of AMY109 in combination with atezolizumab in patients with advanced solid tumors.
Part 1 (dose escalation) had a 3+3 design. At each dose level, if no dose-limiting toxicities (DLTs, defined in online supplemental methods) were observed, patients could be enrolled into the next dose level cohort. If a DLT was observed in one of three patients, three additional patients would be enrolled into that same cohort and all six patients would be monitored. If a DLT occurred in one of the six patients, the next dose level cohort would be enrolled. If a DLT occurred in two or more patients, that dose was judged to be intolerable and no further patients would be enrolled into that dose cohort. The maximum tolerated dose was defined as the highest dose at which <33% of at least three patients had a DLT. Each dose level cohort enrolled 3–6 patients to receive 2, 6, 15, 30, or 45 mg/kg of AMY109 and 1200 mg of atezolizumab, both administered intravenously every 3 weeks (figure 1). In part 2 (cohort expansion), 20 patients were enrolled to receive 15, 30, or 45 mg/kg of AMY109 plus 1200 mg of atezolizumab intravenously every 3 weeks. Intrapatient dose escalation to improve efficacy was permitted, provided that the previously specified conditions were met (ie, if the safety and tolerability at a dose higher than the one initially administered to the patient had been confirmed in the DLT evaluation period of part 1).
Figure 1 Study design. q3w, every 3 weeks. aMandatory tumor biopsy taken.
Eligible patients had advanced or recurrent solid tumors for which standard therapy was ineffective or inappropriate, or for which no standard therapy existed. Eligibility criteria also included written informed consent to participate, age ≥20 years, histological or cytological confirmation of tumor type, measurable lesions by Response Evaluation Criteria in Solid Tumors version 1.1 (RECIST 1.1), Eastern Cooperative Oncology Group performance status of 0 or 1, and adequate organ function. Exclusion criteria included pleural effusion, ongoing adverse drug reactions at grade ≥2 severity, concurrent or history of significant or interstitial lung disease, autoimmune disease, and grade ≥3 immune-mediated adverse drug reaction to previous cancer immunotherapy or discontinuation from cancer immunotherapy due to an immune-mediated adverse event (AE). Before study treatment initiation, the following time periods had to have elapsed: 4 weeks since surgery or cancer immunotherapy, 3 weeks since chemotherapy or radiotherapy, 1 week since molecular-targeted therapies, and 2 weeks since immunosuppressive of hematopoietic growth factor treatment. A full list of eligibility criteria is provided in the protocol, available in online supplemental materials.
EndpointsIn part 1, the primary objectives were to identify DLTs for the combination of AMY109 and atezolizumab, as well as the incidence and severity of AEs assessed according to the National Cancer Institute Common Terminology Criteria for Adverse Events version 5.0, plasma concentration of AMY109, serum concentration of atezolizumab, and pharmacokinetic parameters for AMY109.
In part 2, the primary endpoints were the incidence and severity of AEs, and the antitumor activity of AMY109 plus atezolizumab based on investigator-assessed RECIST 1.1. This comprised the confirmed overall response rate (ORR), disease control rate (DCR, defined as patients with a complete response (CR), partial response (PR), or stable disease for ≥6 weeks), and duration of response. Duration of response in responders was defined as the time from the first assessment at which CR or PR was noted until progressive disease (PD) or death.
The secondary endpoints were to evaluate the antitumor response and immunogenicity (incidence of antidrug antibodies (ADAs) to AMY109) in part 1, and pharmacokinetics and immunogenicity in part 2, for this treatment combination.
Exploratory pharmacodynamic and biomarker analyses in parts 1 and 2 included measurement of plasma IL-8 concentration and pathological evaluation of post-treatment changes in neutrophil (CD15+cells) and cytotoxic T lymphocyte (CTL; CD3+CD8+cells) density and RNA expression of multiple genes in the TME.
Procedures and assessmentsFor all cohorts and cycles, atezolizumab was administered first, followed by AMY109 (after ≥120 min on first administration and after ≥60 min at subsequent cycles). To allow for evaluation of any acute serious toxicity, a ≥24-hour interval was imposed between treating the first and second patient in each dose cohort. The first AMY109 dose in each cohort was infused over 90 (±15) min, then reduced to 60 (±10) min if the tolerability was good. Treatment was continued until study withdrawal by the patient or investigator.
AMY109 and total IL-8 concentrations were measured from plasma samples, and atezolizumab concentration was measured in serum samples collected at study visits. Plasma AMY109 concentrations were measured by ELISA with a lower limit of quantitation (LLOQ) of 50 ng/mL. Plasma total IL-8 concentrations over time were measured by ELISA with an LLOQ of 100 pg/mL. Baseline plasma IL-8 concentrations, which were difficult to quantitate using conventional ELISA methods, were measured using an ultra-sensitive ELISA (Simoa; Quanterix, Billerica, Massachusetts, USA) with an LLOQ of 0.86 pg/mL. Atezolizumab was measured by ELISA with an LLOQ of 46.9 ng/mL.
Patients were assessed by the investigator at screening and every 6 weeks thereafter for treatment response by means of physical examinations and CT/MRI scans per RECIST 1.1. Safety and laboratory parameters were assessed at the study visits throughout the trial. Sponsor-defined AEs of special interest are listed in online supplemental methods.
Serial tumor biopsies were performed to explore pharmacodynamics markers for proof of mechanism in cohorts 1–3, 1–4, 1–5, and 2–1. Pretreatment tumor biopsies were taken from patients at enrolment, and subsequent biopsies were taken from the same lesion during cycle 2 of treatment, or at the first tumor assessment after treatment initiation (figure 1).
To evaluate changes in the TME after treatment with AMY109 plus atezolizumab, the densities of tumor neutrophils (CD15+cells/mm) and CTLs (CD3+ CD8+ cells/mm) in tumor tissue samples were quantified (online supplemental methods). Additionally, RNA sequencing analyses of pretreatment and post-treatment tumor biopsies were done as described in online supplemental methods.
Statistical analysesThe sample size was planned so that 18–30 patients would be enrolled into part 1 using the 3+3 design, and 20–40 patients would be enrolled into cohort 2–1 in part 2.
Safety was assessed in the safety population, defined as patients who received ≥1 dose of study treatment. Pharmacokinetics were assessed in all patients who received ≥1 dose of study treatment and whose drug concentrations were measured at least once. Plasma concentrations of AMY109 and serum concentrations of atezolizumab were listed by cohort, with individual and mean serum drug concentrations tabulated and plotted over time. Efficacy was assessed in the full analysis set, defined as the group of patients who received ≥1 dose of study treatment and in whom ≥1 postdose efficacy assessment was done. Patients who had CR or PR were defined as responders; those who did not, or who had missing data or no assessments, were defined as non-responders. The 95% CIs for ORR and DCR were estimated using the Clopper-Pearson method. The numbers and proportions of ADA-positive patients during the treatment and follow-up periods were summarized by cohort. Statistical analyses, including pharmacokinetic analyses, were performed using SAS V.9.4.
ResultsPatient characteristicsBetween March 19, 2020 and June 8, 2021, 38 patients were enrolled at 3 study centers in Japan. In part 1, 18 patients were enrolled across 5 cohorts (3 patients in each of cohorts 1–1 to 1–4, and 6 patients in cohort 1–5) and all received study treatment. At the cut-off date (July 22, 2022), 1 patient in cohort 1–5 was still receiving treatment and 17 patients had discontinued treatment due to PD.
In part 2, 20 patients were enrolled into cohort 2–1. Of these patients, 4, 12, and 4 received 15, 30, and 45 mg/kg of AMY109, respectively. Two patients increased their AMY109 dose from 30 to 45 mg/kg: 1 patient on day 87 (this patient remained on treatment at the study cut-off) and the other on day 109. At the cut-off date, 1 patient was still receiving study treatment and 19 patients had discontinued (due to PD in 18 patients and physician decision in 1 patient).
Baseline demographic and disease characteristics are summarized in table 1. The patients ranged in age from 33 to 86 years and 24% were 65 years or older. They had esophageal, colorectal, head and neck, pancreatic, breast, uterine, ovarian, and other cancers. Overall, all 38 patients 100% had received prior systemic therapy, 25 of 38 patients (66%) had stage III or IV disease, 32 (84%) had undergone prior surgery, 15 (39%) had prior radiotherapy, and 5 (13%) had received prior anti-PD-1/PD-L1 therapy.
Table 1Baseline demographics and disease characteristics
SafetyThe median treatment duration in each dose cohort and the safety outcomes are summarized in table 2. No DLTs were observed. A maximum tolerated dose was not identified, but tolerability was confirmed up to 45 mg/kg of AMY109 combined with 1200 mg of atezolizumab in cohort 1–5. No deaths due to AEs (ie, no grade 5 AEs) occurred, and no AEs led to discontinuation of either study treatment.
Table 2Treatment exposure and safety summary
Serious AEs were reported in 10 patients (table 2). These were female genital tract fistula (one patient in cohort 1–2), abdominal pain, cytokine release syndrome, facial paralysis, and suicide attempt (one patient each in cohort 1–5), intestinal obstruction, pleurisy, pneumothorax, fever, bile duct obstruction, and viral infection (one patient each in cohort 2–1). Abdominal pain and cytokine release syndrome occurred in the same patient. All of the serious AEs were unrelated to treatment, except cytokine release syndrome and fever (reported as a symptom of infusion-related reaction), which were considered to be related to both AMY109 and atezolizumab by the investigators.
Ten AEs leading to dose interruption of both treatments in six patients were increased alanine aminotransferase and increased aspartate aminotransferase in cohort 1–1; female genital tract fistula in cohort 1–2; decreased lymphocyte count, decreased neutrophil count, decreased white cell count, and cytokine release syndrome in cohort 1–5; and intestinal obstruction, bile duct obstruction, and pleurisy in cohort 2–1 (one event each).
Grade 1–3 AEs related to either treatment occurred in 21 of 38 patients (55.3%) (table 2). No grade 4 or 5 treatment-related AEs were reported. Grade 3 treatment-related AEs occurred in 5 patients (13.2%): aspartate aminotransferase increased, alanine aminotransferase increased, white cell count decreased, and neutrophil count decreased (each in one patient), anemia in two patients, and lymphocyte count decreased in three patients.
AEs related to AMY109 with an incidence of ≥5% in all 38 patients were fever in 8 patients (21.1%); nausea, aspartate aminotransferase increased, and anemia (each in 4 patients (10.5%)); diarrhea, stomatitis, alanine aminotransferase increased, blood alkaline phosphatase increased, lymphocyte count decreased, and platelet count decreased (each in 3 patients (7.9%)); gamma-glutamyltransferase increased, and white cell count decreased (each in 2 patients (5.3%)). Infusion reactions were reported in 6 patients (15.8%; 1 patient in cohort 1–3, 3 patients in cohort 1–5, and 2 patients in cohort 2–1). One of 38 patients (2.6%; cohort 2–1) tested positive for ADA to AMY109.
PharmacokineticsThe plasma concentration of AMY109 increased dose dependently and then gradually decreased over each 21-day treatment cycle (figure 2; online supplemental table 1 and 2). A dose-proportional increase in exposure for AMY109 was indicated over the 2–45 mg/kg dose range by a power model analysis of the exposure level against the actual dose in part 1. The mean accumulation ratios calculated from the minimum and maximum concentrations at cycle 2 had a range of 1.48–1.68 and 1.02–1.24, respectively, in the 2–45 mg/kg groups in part 1 and were comparable in part 2. The pharmacokinetic parameters of AMY109 appeared similar among patients with different types of cancer in part 2 (data are not shown). No apparent change in mean predose atezolizumab serum concentrations was seen at cycle 2 across AMY109 dose groups; these ranged from 96.3 to 115 µg/mL in part 1 and from 81.7 to 112 µg/mL in part 2.
Figure 2 Plasma AMY109 concentration profile in each cohort during the first treatment cycle. Data are mean±SD.
Anti-tumor activityOverall, 2 of 38 patients had a PR, resulting in a confirmed ORR per investigator-assessed RECIST 1.1 of 5% (table 3, figure 3). The durations of these PRs were censored at 8.9 and 9.0 months. The two patients with PRs had uterocervical and pancreatic cancer, respectively, and had been treated for >500 days at the cut-off date (July 22, 2022). One patient received 45 mg/kg of AMY109 in part 1 throughout; the other received 30 mg/kg of AMY109 in part 2 until cycle 5, and 45 mg/kg thereafter. The DCR among the 38 patients was 29%: 2 patients had PRs and 9 had stable disease.
Table 3Confirmed investigator-assessed best responses per RECIST 1.1
Figure 3 Spider plot showing confirmed best overall response in each patient over time. PD, disease progression; PR, partial response; SD, stable disease.
Pharmacodynamics and biomarker studiesAccumulation of plasma IL-8 concentration after starting AMY109 administration was observed in all cohorts. In part 1, the accumulation was generally dose dependent for 2–45 mg/kg of AMY109. The mean plasma IL-8 concentrations on cycle 2 day 1 were 4530.0, 8540.0, 10460.0, 35426.7, and 40 283.3 ng/L in the 2, 6, 15, 30, and 45 mg/kg dose groups, respectively. In part 2, clear dose dependency was not observed, possibly due to high interindividual variability of baseline plasma IL-8 concentrations (online supplemental figure 1). The mean plasma IL-8 concentrations on cycle 2 day 1 were 36633.3, 36636.0, and 49 350.0 ng/L in the 15, 30-, and 45 mg/kg dose groups, respectively.
TME assessment of pretreatment and post-treatment neutrophils and CTLs did not show consistent trends in post-treatment increases or decreases and did not show AMY109 dose dependency (figure 4). Tumor biopsy samples from the two responders with a PR were not suitable for TME evaluation because they included mostly non-cancerous tissue and were, therefore, excluded from these analyses.
Figure 4 Change in immune cell density in tumor tissue after treatment. (A) Neutrophils (B) CTLs. Each colored line indicates the densities in an individual patient. CTL, cytotoxic T lymphocyte.
Fifty-seven formalin-fixed, paraffin-embedded tumor biopsy blocks were obtained before and after administration of AMY109 and atezolizumab. Seven samples that had no tumor or tissue present were excluded, along with 1 other that had low RNA yield and quality. RNA sequencing analyses of pretreatment and post-treatment biopsies that were available for 17 patients showed that changes in the expression of genes related to CD8+ tumor effector T cells, and immune and antigen presentation were strongly correlated with one another (online supplemental figure 2). By contrast, changes in the expression of these genes were negatively correlated with those related to transforming growth factor-β cancer-associated fibroblasts and angiogenesis.
DiscussionIn the first study of AMY109 combined with atezolizumab in patients with advanced malignancies, treatment with AMY109 combined with the standard 1200 mg dose of atezolizumab was well tolerated up to 45 mg/kg, the highest AMY109 dose evaluated, with no DLTs. The initial 2 mg AMY109 dose was chosen based on findings from the first-in-human AMY001JG study conducted in healthy volunteers and patients with endometriosis, as well as toxicity studies in cynomolgus monkeys. No new safety concerns or unanticipated risks were identified. Two patients, one with uterocervical cancer and the other with pancreatic cancer, both with high tumor mutational burden (TMB-H), had long-term PRs and continued receiving treatment for >500 days at the cut-off date. This may suggest that a subset of patients may respond to this treatment combination, or the finding may have been attributable to their anti-PD-L1 treatment-naïve and TMB-H status.18 19 Further investigation in an appropriately selected population would be needed to confirm this finding.
IL-8 is a key factor contributing to fibrosis, neutrophil recruitment and immune suppression in the TME.6–8 10 14 Elevated serum IL-8 is associated with increased intratumoral neutrophils and reduced clinical benefit of immune checkpoint inhibitors.20 We hypothesized that AMY109 would reduce plasma-free IL-8, leading to an antifibrotic effect as well as reduced tumor immune suppression.7 8 11 13 However, in this study, only plasma total IL-8 concentration (ie, free IL-8 plus IL-8 bound to AMY109) was measured because it is difficult to develop the highly sensitive assay system required to quantitate precisely free IL-8 without contamination of bound IL-8. A decrease in the fraction of free IL-8 could be indicated by an increase in plasma total IL-8 concentration observed after AMY109 treatment. This increase could be explained mainly by an accumulation of IL-8 bound to AMY109, potentially due to slower clearance of IL-8 bound to AMY109 compared with free IL-8, which would result in decreasing the fraction of free IL-8.21 It is uncertain whether a decrease in the fraction of free IL-8 would have led to a decrease in plasma-free IL-8 concentration; this would require further investigation.
Given that IL-8 plays a role in the infiltration of neutrophils and MDSCs in the TME that create an immunosuppressive state,3 7 8 AMY109 was expected to reduce the density of tumor tissue neutrophils that are predictive of poor prognosis9 and reduce the clinical benefit of immune checkpoint inhibitors.7 10 22 However, no consistent or dose-dependent trend in neutrophil density changes were observed after AMY109 administration. Possible reasons include the following. The variability in baseline plasma IL-8 concentrations may have impacted the ability to determine the appropriate AMY109 dose that would have successfully inhibited IL-8 in all patients. Hence, the concentration of AMY109 may have been insufficient to neutralize and inhibit tumor IL-8 signaling continuously in some patients. Furthermore, in addition to IL-8, numerous other chemokines are involved in neutrophil migration (eg, CXCL1, CXCL2, and CXCL5),22 23 suggesting that inhibiting IL-8 signaling alone may not completely inhibit neutrophil migration. This immune complexity may explain why some clinical studies combining immune checkpoint inhibitors with other therapies have not been successful to date.
It has been reported that the number of infiltrated and activated CTL increases after immune checkpoint inhibitor treatment for melanoma.24 25 The density of CTLs in the TME is essential for immune checkpoint inhibition to impair tumor growth and impacts the response to immune checkpoint inhibitor therapy.10 26 However, no consistent AMY109 dose-dependent change in CTL infiltration27 was observed in response to AMY109 combined with atezolizumab, although increased CTLs were observed in some patients after AMY109-atezolizumab treatment. The RNA sequencing results from our study are consistent with other findings26 that stromal tumor-associated macrophages, cancer-associated fibroblasts, and tumor angiogenesis are closely correlated with the negative accumulation of CTLs. These results suggest that it is likely important to suppress both fibrosis and angiogenesis to increase CTLs in the TME. The benefit of combining immune checkpoint inhibition with antiangiogenic agents has been demonstrated,28 but combining immune checkpoint inhibitors with antifibrotic agents may also improve outcomes.
A study limitation was that patients with a range of different solid tumor types were enrolled on this phase 1 trial, making it difficult to determine whether any of these tumors were more likely to respond to this treatment combination. Additionally, it was difficult to draw firm conclusions from the biomarker findings due to the small number of patients in each dose cohort, and the fact that co-clinical, TMB and PD-L1 status data were not available from the two responders with PRs because of insufficient tumor biopsy tissue.
In conclusion, AMY109 combined with atezolizumab was well tolerated in patients with advanced solid tumors, with no severe safety signals. AMY109 showed a dose-proportional increase in exposure. Although clear changes in the biomarkers were not observed, several patients, including two patients with notably durable PRs, were able to continue this combination therapy for a long period. Further investigations, including those involving selective biomarker identification, are warranted.
Ethics statementsPatient consent for publicationEthics approvalAll patients provided written informed consent to participate in the study. The study was approved by the institutional review board of each investigational site (National Cancer Center Hospital (T4772), National Cancer Center Hospital East (K0902), and The Cancer Institute Hospital of the Japanese Foundation for Cancer Research (2020-0064). It was conducted in accordance with its protocol, International Conference on Harmonization Good Clinical Practice guidelines, applicable regulations, and the Declaration of Helsinki.
AcknowledgmentsMedical writing assistance was provided by Samantha Santangelo, PhD, of Health Interaction, Inc, San Francisco, CA, USA, funded by Chugai Pharmaceutical Co, Ltd.
留言 (0)