A Phase I/II Trial of Cetuximab in Combination with Interleukin-12 Administered to Patients with Unresectable Primary or Recurrent Head and Neck Squamous Cell Carcinoma

Over 50,000 persons in the United States are diagnosed with head and neck squamous cell carcinoma (HNSCC) annually, the majority of which (>90%) overexpress the EGFR (1). The standard treatment for patients with localized but unresectable disease is cisplatin-based chemotherapy combined with radiation, an approach that is also used as adjuvant therapy in patients with high-risk pathologic findings at surgical resection. Despite this multimodality approach, the majority of patients will develop local and/or regional recurrences and 20%–30% will develop distant metastases (2). Taken together, these results suggest that there is a need for new, highly active therapeutic agents that can be combined with existing treatments.

Cetuximab is a humanized mAb that binds to EGFR on tumor cells and has activity when administered as a single agent to patients with HNSCC (3). The actions of cetuximab are multiple in that it prevents ligand binding to EGFR, promotes degradation and downregulation of EGFR on the cell surface, and provokes an immunologic antitumor effect via antibody-dependent cellular cytotoxicity (ADCC; ref. 4). The direct cytotoxic effect of cetuximab on EGFR-positive cancer cells markedly enhances the activity of standard therapies for HNSCC. The addition of cetuximab to platinum-based chemotherapy in patients with untreated recurrent or metastatic HNSCC led to a significant improvement in overall survival and increased the response rate to antibody therapy (5).

Natural killer (NK) cells are large granular lymphocytes that contain abundant cytolytic granules, express multiple adhesion molecules, and constitutively express receptors for several cytokines (6). Activated NK cells produce cytokines with antitumor actions and chemokines that recruit macrophages and T cells to sites of inflammation (7–9). Of importance, NK cells express receptors for the Fc, or constant region, of IgG (FcγRIIIa) that enable NK cells to interact with antibody-coated tumor cells and mediate both ADCC and the secretion of cytokines such as IFNγ (10–12). In addition, NK cells are unique in that they constitutively express receptors for numerous cytokines (i.e., IL12, -15, -18, and -21). Our group has shown in vitro and in murine tumor models that costimulation of NK cells via the IL12 receptor and FcγRIIIa activates the ERK, which in turn promotes the secretion of IFNγ (13). This result was confirmed in a preclinical model of HNSCC, where the stimulation of NK cells with IL12 increased the lysis of cetuximab-coated HNSCC tumor cells regardless of human papillomavirus (HPV) status of the tumor cell line, increased production of IFNγ, RANTES, MIP-1α, and IL8, induced the phosphorylation of ERK, and resulted in a reduction in tumor burden in a murine xenograft model of HNSCC (14). These results suggested that immunologically active compounds could enhance the patient immune response to therapeutic mAbs.

The aim of this phase I/II study was to determine a safe and tolerable dose of IL12 when administered in combination with cetuximab in patients with unresectable primary or recurrent squamous cell carcinoma of the oropharynx, and to test the ability of the MTD of IL12 to enhance the response rate to cetuximab in patients with advanced stage HNSCC. On the basis of preclinical studies, we hypothesized that IL12 would potentiate the activity of cetuximab in patients with HNSCC.

Patients with histologically proven recurrent and/or metastatic HNSCC that was unresectable were eligible for enrollment in this NCI-sponsored phase I/II trial (NCT01468896). Any number of prior systemic therapies for metastatic/recurrent disease were permitted in both the phase I and II portion of the study. Patients were not excluded if they had received a prior cetuximab-based chemotherapy regimen. Patients were required to be ≥18 years of age; have a life expectancy >6 months; an ECOG performance status index ≤ 2 (Karnofsky performance status index ≥60%); adequate organ and marrow function; and be capable of giving informed consent. Women of child-bearing potential and men were required to use adequate contraception prior to study entry and for the duration of study participation. Excluded from the study were patients who had chemotherapy or radiotherapy within 4 weeks prior to entering study, history of allergic reactions attributed to compounds with chemical or biological composition similar to IL12 or cetuximab, or those with an uncontrolled comorbid illness including congestive heart failure, unstable angina pectoris, cardiac arrhythmia, or psychiatric illness/social situations that would limit compliance with study requirements. All subjects gave written informed consent approved by their local institutional review board and followed the Declaration of Helsinki guidelines.

The primary objective of the phase I portion of this study was to find a safe and tolerable dose of IL12 for use in combination with cetuximab. Patients were pretreated with an H₁ blocker 30–60 minutes prior to the first dose of cetuximab. Each cycle was 14 days in length. Cetuximab was given by intravenous infusion at a dose of 500 mg/m² on day 1. Beginning with the second cycle, two subcutaneous injections of IL12 were administered on days 2 and 5, with the dose of IL12 being escalated in standard fashion in cohorts of three patients (0.2 mcg/kg then 0.3 mcg/kg) until the MTD was identified. Once the MTD was determined, a two-stage phase II design based on the phase I schema was planned to evaluate the impact of the addition of IL12 to cetuximab on response rate in patients with advanced HNSCC.

Dose-limiting toxicity (DLT) was determined during the first three cycles of therapy of the phase I portion of the trial. Patients who experienced any clearly drug-related grade 3 reaction or greater (hematologic or nonhematologic) were considered to have experienced a DLT and were removed from the study. Exceptions included: grade 3 fatigue, any grade 3 event that did not delay therapy more than 7 days, any grade 3 electrolyte abnormality that could be corrected within 48 hours via replacement therapy, or any grade 3 laboratory abnormality that did not reflect underlying organ pathology (according to the Common Terminology Criteria for Adverse Events version 4.0). The doses of IL12 and cetuximab were not escalated within an individual patient.

Peripheral blood for correlative studies was obtained pretherapy, just prior to the administration of each dose of cetuximab (day 1 every other week) and just prior to the administration of each dose of IL12 (days 2 and 5) for 24 weeks. During the phase II portion of the trial, blood for correlative studies was to be obtained pretherapy, just prior to the administration of each dose of cetuximab, and just prior to the administration of each dose of IL12 on cycles 1, 2, 6, 12, 18, and 24. Serum and peripheral blood mononuclear cells (PBMC) were procured and frozen using standard techniques as described previously (13). Further analysis of samples for correlative studies was carried out similarly to previously reported clinical trials (15).

PBMCs from patients' pretherapy, and days 1, 2, and 5 of cycle 4 were plated in 96-well V-bottom plates and treated with or without IL12 (10 ng/mL) overnight in RPMI1600 supplemented with 10% human AB serum at 37°C. Eighteen hours later, Cal27 (HPV-negative) HNSCC tumor cells were labeled with ⁵¹Cr for 1 hour, followed by incubation with either cetuximab or control IgG (50 μg/mL) for 45 minutes. Tumor cells were then added to NK cells at various effector:target (E:T) ratios. Following a 4-hour incubation, supernatants were harvested, chromium release was quantified, and percent lysis was determined as described previously (12).

Following the manufacturer's instructions, plasma from patients at baseline and cycle 8 was evaluated for levels of IFNγ, IP-10, MIP-1α, MIP-1β, RANTES, GM-CSF, IL8, and TNFα using a custom V-PLEX Assay (Meso Scale Discovery).

DNA was isolated from pretreatment patient PBMC using a QIAamp kit according to the manufacturer's instructions (Qiagen) and quantified. Genotyping of the FcγRIIIA-158V/F polymorphism was performed using a nested PCR followed by allele-specific restriction enzyme digestion as described previously (16). The amplified DNA was digested with 10 U NlaIII (New England Biolabs) at 37°C for 12 hours and separated by electrophoresis on 8% polyacrylamide gel. After staining with ethidium bromide, DNA bands were visualized under UV light. For homozygous FcγRIIIA-158F patients, only one undigested band (94 bp) was visible. Three bands (94 bp, 61 bp, and 33 bp) were seen in heterozygous individuals, whereas for homozygous FcγRIIIA-158V patients only two digested bands (61 bp and 33 bp) were obtained. All samples were analyzed in duplicate.

Frequency of T-cell subsets, macrophage subsets, and myeloid-derived suppressor cells (MDSC) in the peripheral blood was analyzed via flow cytometry. PBMCs procured from pretherapy and cycle 4 day 1 blood draws were stained with anti-CD3-V450, anti-CD8-PE, anti-CD4-FITC, anti-Foxp3-PE, anti-CD56-APC, anti-CD16-APC Cy7, anti-NKG2C, anti-CD33-APC, anti-HLA-DR-PECy7, anti-CD11b-PE, anti-CD14-V450, anti-CD15-FITC, anti-CD80-FITC, anti-CD1630-PE, anti-CD206-APC, and/or anti-CD69-PE-Texas Red (Beckman Coulter). Immune cell subsets were defined as follows: CD4 T cells CD3⁺/CD4⁺, CD8 T cells CD3⁺/CD8⁺, Treg CD3⁺/CD4⁺/Foxp3⁺, M1 macrophages CD14⁺/CD80⁺/CD163⁻/CD206⁺, M2 macrophages CD14⁺/CD80/CD163⁺/CD206⁺, NK cells CD3⁻/CD56⁺/CD16⁺, granulocytic MDSC CD33⁺/HLA-DR⁻/CD11b⁺/CD15⁺, and monocytic MDSC CD33⁺/HLA-DR⁻/CD11b⁺/CD14⁺. Data were acquired using a LSRII Flow Cytometer (BD Biosciences).

ADCC data were analyzed using a mixed effect model, incorporating repeated measures for each subject. Ratios of cytokine and chemokine production between pretreatment and cycle 8 were analyzed by a two-sample t test.

Twenty-three patients (22 male, one female) were enrolled, with an average age of 60 years (range, 41–82 years). Most patients were Caucasian (20 Caucasian, one African-American, and 1 Asian) and 18 had an ECOG performance status ≤ 1. Twelve patients had metastatic disease upon enrollment and the remainder had locally advanced disease that was unresectable. Twelve patients had HPV-positive tumors, five patients had HPV-negative tumors, and the HPV status of tumors from six patients was undetermined. Two patients were HIV-positive. The majority of patients (21/23; 91%) had received at least one prior regimen of systemic chemotherapy, including 17 patients that had previously received cisplatin-based chemotherapy. Seven patients (30%) had received prior EGFR-directed therapy, including cetuximab, and five of these seven patients achieved clinical benefit (stable disease) while enrolled on this trial that lasted an average of 268 days. Eighteen patients (78%) had prior surgery, while all 23 patients had received radiation treatment. Patient demographics are summarized in Table 1.

While a small number of patients experienced weight gain and temporary electrolyte abnormalities (e.g., hypomagnesemia, hyponatremia), no DLTs were observed during therapy cycles which included administration of IL12. Only two patients experienced significant cytopenias. As a result, after three patients had been treated, dose escalation of IL12 was initiated and advanced to 0.3 mcg/kg as planned. A few patients had their treatment regimen interrupted for health concerns, which required hospitalization including dysphagia (1), an unspecified lung infection (1), and pleuritic pain (1), yet all patients were able to resume the treatment regimen following discharge and were not considered to have had a DLT. Increases in alanine aminotransferase and aspartate aminotransferase were the only grade 4 toxicities recorded, and these occurred in patient D (phase I, dose level 2) who was removed from the trial during cycle 1 per the guidelines described above. No other DLTs were observed. All 17 patients enrolled in the phase II study were given the maximum tolerated dose of IL12 (0.3 mcg/kg). In general, the regimen was well tolerated. Common grade 2 toxicities included decreased lymphocyte count (8), fatigue (4), hypertension (2), and acneiform rash (2), and these side effects were felt to be possibly disease related in many cases. Notable grade 3 toxicities included hyponatremia (4) and weight gain (3). All grade 3 and 4 toxicities are listed in Supplementary Table S1.

Three patients enrolled in the phase I portion had stable disease (SD) that lasted for an average of 145 days (Supplementary Table S2), whereas two patients experienced progressive disease (PD). As noted above, 1 patient on dose level 2 was removed for changes in liver enzymes but was not replaced. Thirteen patients enrolled in the phase II portion had SD that lasted for an average of 206 days (Supplementary Table S2), whereas four experienced PD. Overall, there was a 69% rate of clinical benefit (SD) in the phase II portion of the study, and the average progression-free survival (PFS) in this group of patients was 195 days (Fig. 1). Patient characteristics and response evaluation are detailed in Supplementary Table S2.

Due to small patient numbers in the phase I portion of the trial, all six patients enrolled in phase I were combined for statistical analysis. The mean change in ADCC between day 45 of treatment and baseline for phase I patients was −12.74% [95% confidence interval (CI), −22.40 to −3.08; P = 0.0137], indicating a significant decrease in ADCC activity, correlating with an average PFS of 100.2 days (range, 42–269 days; Fig. 2A). The mean change in ADCC between day 45 of treatment and baseline for patients enrolled in the phase II portion of the trial was 4.21% (95% CI, 0.66–7.75; P = 0.0222), indicating a significant increase in ADCC activity, correlating with an average PFS of 174.6 days (range, 44–449 days; Fig. 2B). Patients enrolled in either the phase I or phase II part of the study who experienced PFS greater than 100 days (n = 11) had a significant increase in their ADCC activity at day 45 of treatment compared with baseline (P = 0.0386; Fig. 2C), whereas patients who experienced PFS less than 100 days (n = 11) had a significant decrease in their ADCC activity at day 45 of treatment compared with baseline (P = 0.0437; Fig. 2D).

Eight cytokines were measured in the plasma of phase I and II patients at baseline and at the beginning of cycle 8 (Supplementary Fig. S1). Cycle 8 measurements were chosen for the comparison based on the assumption that sustained cytokine production would be evident at this timepoint. Patients who experienced PFS greater than 100 days (n = 11) showed a significant increase in IFNγ production (P = 0.0059; Fig. 3A) as compared with patients who experienced PFS less than 100 days (n = 11, P = 0.1309; Fig. 3A). Likewise, patients who experienced PFS greater than 100 days (n = 11) showed a significant increase in production of IP-10 (P = 0.0195; Fig. 3B) as compared with patients who experienced PFS less than 100 days (n = 11, P = 0.4258; Fig. 3B). Patients who experienced PFS greater than 100 days (n = 11) showed a significant increase in TNFα production (P = 0.0391; Fig. 3C) as did patients who experienced PFS less than 100 days (n = 11, P = 0.0394; Fig. 3C). Production of MIP-1α, MIP-1β, RANTES, GM-CSF, and IL8 was not significantly up- or downregulated between baseline and the beginning of cycle 8 when compared between patients who experienced PFS greater than or less than 100 days (Supplementary Fig. S2).

Polymorphisms in the FcγRIIIa can affect NK-cell response to mAb therapy (15). Six patients were identified as carrying the high affinity V/V Fc gamma receptor polymorphism: 10 carried the low affinity F/F polymorphism, and the remaining seven carried the heterozygous V/F FcγRIIIa polymorphism (Supplementary Table S3). There was no significant association between polymorphism status and clinical benefit, toxicity, or ADCC activity.

Multiple immune cell subsets have been shown to express FcγRIIIa and thus have the ability to be impacted by treatment with mAbs that engage this receptor (17). Once stimulated by immobilized mAb, these cells in turn activate other immune effector populations with both positive and negative effects on tumor cell viability. To further characterize the immune cell subsets that might have impacted the increase in ADCC function exhibited by patients who experienced PFS greater than 100 days, an analysis of the PBMC compartment was performed. The following immune cell subsets were examined: NK cells, CD4⁺ T cells, CD8⁺ T cells, regulatory T (Treg) cells, M1 and M2 macrophages, and granulocytic and monocytic MDSC. Patients who experienced PFS greater than 100 days had higher levels of circulating CD4⁺ and CD8⁺ T cells and lower levels of circulating Tregs at baseline as compared with patients who experienced PFS less than 100 days (Fig. 4A). There was a slight increase in the number of circulating Tregs in patients who experienced PFS greater than 100 days between baseline and cycle 4 day 1, however the levels of all T-cell subsets in patients who experienced PFS less than 100 days remained stable between baseline and cycle 4 day 1 (Fig. 4A). Patients with PFS greater than 100 days had lower circulating levels of NK cells at baseline. Circulating NK cells increased slightly in this subset of patients over the course of treatment and remained stable for patients with PFS less than 100 days (Fig. 4A). There was not a significant presence of M1 or M2 macrophages in the peripheral blood, and these levels remained stable over the course of treatment in both subsets of patients (dns). IL12 is known to directly impact immune cell activation and expansion of CD4⁺ T cells and NK cells (18–20). There were no significant changes in T-cell activation, as measured by CD69 expression, or NK-cell expansion or activation, as measured by NKG2C and CD69 expression, in either group of patients (Fig. 4B).

Long-term secretion of cytokines and chemokines in the tumor microenvironment leads to the outgrowth of immunosuppressive cell types, including MDSC, that hinder the ability of an effective immune response (21). The percentage of MDSC present in the peripheral blood as determined by flow cytometry (CD33⁺/HLA-DR⁻) averaged 4.78% for all patients enrolled in the trial (Supplementary Table S3). Fifteen of 23 patients had a higher percentage of the monocytic subset of MDSC (CD33⁺/HLA-DR⁻/CD14⁺/CD11b⁺; Supplementary Fig. S3A) as compared with eight patients who exhibited a higher percentage of granulocytic MDSC (CD33⁺/HLA-DR⁻/CD15⁺/CD11b⁺; Supplementary Fig. S3B). The presence of a monocytic-dominant MDSC population correlated with a better prognosis, as the average PFS of monocytic-dominant patients was 182.5 days, compared with an average PFS of 95.6 days in patients who had a higher percentage of granulocytic MDSC prior to beginning treatment (Fig. 5). Of note, six of eight patients who had a high percentage of granulocytic MDSC prior to the beginning of treatment also carried the low affinity F/F FcγRIIIa polymorphism (Supplementary Table S3). Patients who experienced PFS greater than 100 days had an average of 3.6% MDSC present in the circulating peripheral blood prior to beginning treatment, compared with an average of 5.8% in patients who experienced PFS less than 100 days (Supplementary Table S3); however, these differences were not statistically significant. The 3 patients that experienced the longest PFS (range, 449–459 days) had an average of 2.9% MDSCs present prior to beginning treatment, and all 3 had a greater percentage of the monocytic subset at baseline.

Despite recent advances in surgery, radiation, and systemic therapy, the overall survival of patients with advanced HNSCC has improved only marginally over the past 3 decades. The majority of patients present with locoregionally advanced disease, with more than 50% having a recurrence within 3 years (22, 23). Although cetuximab has been FDA approved for the treatment of locally advanced and recurrent/metastatic HNSCC, response rates are only in the range of 15%–20% (3, 24). This phase I/II clinical trial was conducted to determine a safe and tolerable dose of IL12 for use in combination with cetuximab and to evaluate the effectiveness of this therapeutic strategy in patients with unresectable primary or recurrent HNSCC. The MTD of IL12 was determined to be 0.3 mcg/kg, and the combination of IL12 with cetuximab was well tolerated. The majority of grade 3 and 4 events were thought to be disease related. In this study, the combination of IL12 and cetuximab resulted in SD in 69% of patients and no partial responses. The median PFS was 6.5 months. Patients enrolled in the trial who experienced a PFS greater than 100 days (n = 11) showed increased ADCC activity at day 45 as compared with baseline, and serum cytokine analysis in these patients revealed a significant increase in circulating IFNγ, IP-10, and TNFα. Furthermore, a correlation was seen in patients who experienced a PFS greater than 100 days and the presence of fewer circulating MDSC at baseline. These MDSC were mostly of the monocytic subset.

Overall, the combination of IL12 with cetuximab was well tolerated with several patients tolerating long periods of treatment. While there were no complete or partial responses to therapy, it is important to note that all patients were heavily pretreated, with 52% having received at least two prior chemotherapy regimens. In addition, 78% had undergone prior therapeutic surgeries, and all patients enrolled had received prior radiotherapy. IL12 has been employed in the treatment of solid tumors, and our group has utilized IL12 in combination with paclitaxel and trastuzumab in HER-overexpressing patients. (25–27). These previous phase I trials have shown toxicity profiles similar to this study, with electrolyte abnormalities and cytopenias being some of the most clinically common adverse effects. Single agent cetuximab administration is associated with liver enzyme abnormalities (33%–43% of patients) and grade 3–4 cytopenias (17%–31% of patients). Thus, cetuximab and IL12 do have some overlapping adverse effects (28–30). The extent of disease and number of prior therapies in the study population may have contributed to the common grade 3 toxicities seen during therapy, as the administration of low dose IL12 as a single agent on a similar dosing schedule was generally well tolerated (26, 27, 31). While treatment with IL12 has been reported to be associated with systemic flu-like symptoms (fever, chills, fatigue, arthromyalgia, and headache) and effects on the bone marrow and liver (31–35), the absence of severe reactions to injections of low dose IL12 in combination with cetuximab in this trial was similar to our previously reported studies (25, 36).

IL12 was administered with cetuximab with the expectation that it would serve as a costimulatory signal to FcR-bearing cells that had recognized antibody-coated tumor cells. We have previously shown that costimulation of NK cells via the IL12 receptor and FcR leads to colocalization of these receptors in plasma membrane lipid rafts and a significant increase in the expression of genes encoding cytotoxicity receptors, apoptotic proteins, intracellular signaling molecules, and cytokines that could mediate enhanced cytotoxicity and interactions with other immune cells within inflammatory tissues (13, 37). While 69% of patients experienced clinical benefit (stable disease) from the administration of IL12 in combination with cetuximab, there were no clinical responses. Further, while IL12 has shown to impact the activation and expansion of Th1 and NK cells in vitro, there were no significant changes in the activation state of either of these immune cell subsets in patients between baseline and cycle 4 day 1. It is possible that IL12 administration promoted the participation of the Th1 T-cell compartment in the eradication of tumor cells in a manner that did not lead to large changes in cell numbers or activation status as measured in this study. Optimal enhancement of cytokine production by FcR-bearing cells might require the administration of multiple cytokines or immune-stimulating agents (38). There has been considerable interest in the addition of checkpoint inhibition to cetuximab therapy, as cetuximab can exert positive effects on both innate and adaptive immunity. Cetuximab has been reported to induce cross-priming of EGFR-specific circulating T lymphocytes in patients with HNSCC via stimulation of NK-cell–DC cross-talk (39). It has recently been shown, however, that cetuximab therapy increased the frequency of CD4⁺Foxp3⁺ intratumoral Tregs, which subsequently suppressed NK-cell–mediated ADCC and correlated with poor clinical outcome in patients (40, 41). However, ex vivo assays with patient samples showed that the addition of an anti-CTLA4 mAb (ipilimumab) could specifically target CTLA4⁺ Treg and restore the cytolytic potential of NK cells, thereby promoting antitumor immunity (42). These studies highlight the strategies that might be used to improve outcomes for patients with HNSCC.

Patients enrolled in the phase II portion of the trial who received the MTD of IL12 showed increased ex vivo ADCC activity of total PBMCs against cetuximab-coated tumor cells at day 45 as compared with baseline (Fig. 2). This result was also seen in patients who experienced a PFS greater than 100 days (Fig. 2). Secretion of IFNγ was observed in response to administration of IL12 (Fig. 4); however, similar to our previously reported study (25), there is no clear evidence suggesting that the higher dose of IL12 had greater antitumor activity or immune stimulatory activity as clinical effects and significant IFNγ production was observed at both dose levels. Reports suggest that a positive response to mAb therapy relies on the presence of the high-affinity VV FcγRIIIa polymorphism (43), however, while the high-affinity allele was detected in six of 23 patients, there was no correlation between this genotype and enhanced ADCC activity or prolonged PFS. Similarly, an investigation by Srivastava and colleagues of 107 patients with HNSCC treated with cetuximab who were genotyped for a polymorphic FcγRIIIa showed no association with the high-affinity allele and disease specific survival, raising the likelihood that additional immune mechanisms might play a role in the antitumor activity of cetuximab (44).

Our group and others have previously shown that enhanced levels of circulating MDSC correlate with increased disease burden and decreased survival in patients with cancer (45, 46). In the context of head and neck cancer, Li and colleagues (47) observed a higher frequency of circulating monocytic MDSC in patients with locally advanced HNSCC who responded to single-agent cetuximab treatment prior to surgery. Similarly, patients who experienced a PFS greater than 100 days in this trial had a higher frequency of monocytic MDSC compared with granulocytic MDSC prior to therapy. The association of differences in MDSC populations with PFS is of potential interest, as IL12 has been shown to modulate the immune suppressive function of MDSC and their overall frequency in mouse models of melanoma and breast cancer. The antitumor efficacy of CD8⁺ T cells engineered to secrete IL12 in the B16 murine melanoma model was found to be reliant upon the effects of IL12 on tumor-infiltrating populations of myeloid cells, including MDSC (48), resulting in the complete remodeling of the tumor microenvironment. Interestingly, the predominant tumor infiltrating MDSC population in this model was the monocytic subset and myeloid cells isolated from tumors of mice treated with IL12 had increased expression of Ifng and Il2. In addition, IL12 stimulation of MDSC and other myeloid cell populations resulted in enhanced cross-presentation of tumor antigens and stimulation of T-cell proliferation compared with MDSC not stimulated with IL12. In a breast cancer model, IL12 was shown to promote the maturation of MDSC and reduce their expression (49). These findings suggest that, in addition to stimulating NK cells, IL12 administration may also help to reverse myeloid cell–mediated peripheral immune tolerance and enhance T-cell function in patients with head and neck cancer. However, the observation that IL12 stimulation of myeloid cells increased Ifng expression suggests that IL12 could render tumors more sensitive to PD-1/PD-L1 inhibition because IFNγ is known to induce PD-L1 expression on tumor cells. Thus, blockade of an immune checkpoint such as PD-1/PD-L1 might be needed to fully harness the antitumor effect of IL12 on T cells. The combination of IL12 and cetuximab might serve as an important primer to optimize the ability of the innate immune system to stimulate adaptive antitumor immune responses in patients with HNSCC.

This study demonstrates that IL12 can be given safely in combination with cetuximab in patients with HNSCC. Clinical benefit was accompanied by an immune response that was characterized by an increase in ADCC and the secretion of IFNγ. NK-cell–directed immunotherapy represents a promising approach to the treatment of HNSCC and further studies of immune activating agents in combination with cetuximab could yield increased antitumor activity.

M.A. Caligiuri is a consultant/advisory board member for Dragonfly Therapeutics. S.V. Liu is a consultant/advisory board member for Apollomics, AstraZeneca, Bristol-Myers Squibb, Celgene, Genentech/Roche, Heron, Ignyta, Janssen, Lilly, Merck, Pfizer, Regeneron, Taiho, Takeda, G1 Therapeutics, and Boehringer Ingelheim, and reports receiving commercial research grants from AstraZeneca, Bayer, Blueprint, Bristol-Myers Squibb, Clovis, Corvus, Esanex, Genentech/Roche, Ignyta, Lilly, Lycera, Merck, Molecular Partners, OncoMed, Pfizer, Rain Therapeutics, and Threshold. No potential conflicts of interest were disclosed by the other authors.

Conception and design: E.L. McMichael, T. Olencki, M.A. Caligiuri, W.E. Carson III

Development of methodology: E.L. McMichael, B. Benner, T. Noel, T. Olencki, J.C. Byrd, W.E. Carson III

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): E.L. McMichael, L.S. Atwal, N.B. Courtney, M.E. Davis, M.C. Duggan, K. Levine, G.N. Olaverria Salavaggione, A. Ganju, S. Uppati, T.N. Teknos, P. Savvides, J.C. Byrd, S.V. Liu

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): E.L. McMichael, B. Benner, N.B. Courtney, X. Mo, A.R. Campbell, M.C. Duggan, K. Levine, T. Noel, M.A. Caligiuri, S.V. Liu, W.E. Carson III

Writing, review, and/or revision of the manuscript: E.L. McMichael, B. Benner, X. Mo, A.R. Campbell, M.C. Duggan, K. Martin, T. Olencki, S. Tridandapani, J.C. Byrd, M.A. Caligiuri, S.V. Liu, W.E. Carson III

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): K. Levine, B. Paul, T. Olencki, T.N. Teknos, W.E. Carson

Study supervision: K. Levine, T. Olencki, W.E. Carson

This work was supported by NIH Grants P01 CA095426 (to M.A. Caligiuri), P30 CA016058 (to M.A. Caligiuri), UM1 CA186712 (to W.E. Carson), and T32 CA090223 (to W.E. Carson).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.