NHS-IL2 combined with radiotherapy: preclinical rationale and phase Ib trial results in metastatic non-small cell lung cancer following first-line chemotherapy

We have previously demonstrated that NHS-IL2 is effective in a lung metastasis model and shown that the efficacy is primarily dependent on CD8+ T cells, and to a lesser extent on NK cells [10]. To test the hypothesis of whether radiotherapy affects immunity against tumor cells in metastatic disease, low-dose fractionated local radiation was combined with NHS-IL2 treatment in a syngeneic LLC tumor model. Seven days after implantation, the tumor reached a volume of 110 mm³ and 10 mice per group were treated with saline (i.e., control), NHS-IL2 alone, radiotherapy alone, cisplatin alone, NHS-IL2 plus cisplatin, NHS-IL2 plus radiation, radiation plus cisplatin, or the triple combination of cisplatin, radiation, and NHS-IL2 (study design see Figure 1A). NHS-IL2 alone did not significantly inhibit tumor growth, nor did cisplatin alone or cisplatin plus NHS-IL2. Radiotherapy alone delayed tumor growth only temporarily. Addition of cisplatin to either NHS-IL2 or radiotherapy did not achieve any significant improvement in tumor growth control compared to NHS-IL2 alone or radiotherapy alone. However, the combination of cisplatin plus radiotherapy with NHS-IL2 resulted in marked tumor reduction and delayed outgrowth that was statistically superior to all other groups (P < 0.001 by Bonferroni’s post-test following 2-way repeated-measures analysis of variance [ANOVA] on study day 14 [last day on which all study groups were complete]) (Figure 1B). At day 53 of the study, 5 of 6 mice remaining in the triple combination group had achieved complete regression. These data were reproduced in 2 experiments with similar designs (data not shown). The addition of cisplatin to the combination of radiotherapy and NHS-IL2 (ie, triple combination) contributed mainly to immune activation and tumor growth control. Following treatment, gene expression in tumors was determined by quantitative PCR, cellular changes in the periphery were monitored by fluorescence-activated cell sorting (FACS), and tumor-infiltrating lymphocytes assessed by immunohistochemistry (IHC) to comprehensively elucidate the mechanistic pathways involved in the synergistic antitumor effects of these combination therapies (data not shown; manuscript in preparation). Additional file 1: Figure S1 shows upregulated expression of genes associated with effector T cells (CD3, CD4, CD8, CD25) and cytotoxic immune cells (Fas ligand, granzyme B, perforin 1) within tumors treated with the triple combination, indicating an increased infiltration of activated T cells. This was further confirmed by FACS and IHC-based analysis of tumors, supporting the previously described immune-potentiating effects that radiation has on tumors (data not shown).

Based on the promising preclinical study results, the EMR62235-002 phase Ib trial (NCT00879866) was initiated. Between April 2009 and August 2010 a total of 15 patients with NSCLC stage IV disease who achieved at least stable disease after first-line platinum-based therapy provided informed consent. Two patients showed progressive disease during screening and were thus not enrolled into the trial, leading to a total of 13 patients treated, as part of the safety analysis set (3 patients each in the 0.15-mg/kg and 0.30-mg/kg cohort, and 7 in the 0.45-mg/kg cohort). Baseline characteristics for this population are reported in Table 1. The overall median time from metastasis diagnosis to study entry was about 4.3 months, in which the patients received 4–6 courses of platinum doublets (Table 1). The 13 patients received 5× 4 Gy radiotherapy followed by NHS-IL2 over 3 consecutive days in a 21-day treatment cycle repeated until progression of disease or unacceptable toxicity (study design see Figure 2A). Twelve patients were evaluable to assess DLTs and antitumor activity, whereas the thirteenth patient had a major protocol deviation and was replaced. This patient received incorrect dosing (0.30 mg/kg instead of 0.45 mg/kg) at cycle 1 day 1, and was excluded from the DLT analysis set. Thirteen patients received a median of 3 completed cycles NHS-IL2 (mean number of completed cycles: 3.9 [standard deviation 3.9], range 1–12). In fact, 1 patient has received over 20 treatment courses.

The treatment was generally well tolerated (Table 2). Most frequently reported AEs were fatigue, anorexia, and rash (Table 2), often presenting with erythema with mild facial and periorbital swelling, and mild urticaria on the upper part of the thorax, most prominent from day 4 to about day 7 and followed by dry skin and exfoliation. These AEs were always followed by complete recovery within the second week and tended to recur after subsequent cycles, but were very manageable. There were only 3 grade 3 AEs considered by the investigator as related to NHS-IL2 (Table 2). One patient in the 0.30-mg/kg cohort developed grade 3 normochromic normocytic anemia, while grade 3 transitory lymphopenia was observed in 1 subject at 0.45 mg/kg. One DLT was reported also at the highest dose of NHS-IL2; the patient experienced grade 3 pericarditis directly following the first course, explained by rapidly progressive disease, as confirmed by postmortem examination.

In addition to the mild transient facial edema, the only classic IL-2 immune-related AE was thyroiditis. Three out of 12 patients developed hyperthyroidism grade 1/2 during treatment, with free thyroxin levels >100 pmol/dL after the second or third cycle. High levels of thyroglobulin accompanied the thyroiditis episodes. No autoantibodies against thyroid-stimulating hormone or thyroglobulin were detected in any of these patients during or after treatment. One patient had hyperthyroidism classified as a serious AE, developed symptomatic tachycardia and had to be treated with propranolol. This patient recovered completely after the treatment was stopped because of disease progression. Two patients, in whom treatment was continued, developed subsequent hypothyroidism that required thyroid hormone therapy supplements.

Mean peak concentrations of NHS-IL2 in serum measured at the end of the 1-hour infusion on days 1–3 of cycle 1, 2, and 4 increased with dose in an almost directly proportional manner (Figure 2B). Within each cycle the peak concentrations continuously increased during the 3 days of administration until steady state was reached with the third administration. After the third administration concentrations declined steadily and reached pre-dose levels in week 2 of each cycle. No accumulation in exposure over time was observed at any NHS-IL2 dose level.

Blood analysis showed a clear immune induction as defined by the leukocyte kinetics and the production of acute-phase proteins. One week following the start of each cycle (day 1, 21, 42, and 63) an increase of leukocyte counts was noted, followed by normalization with return to basal level (Figure 3). A similar pattern was observed for lymphocytes, monocytes, and neutrophils, with exception of a short depletion of neutrophils during the first cycle. Increase of eosinophils appeared with the second cycle of treatment and showed a similar pattern for all dose levels, but with a highest response after the fourth course. Basophils presented only a modest variation during treatment.

In line with the leukocyte kinetic, immunomonitoring showed a considerable increase of proliferating (Ki67 + HLA–DR+) CD4 and CD8 T cells as well as memory (CD197–CD45RA–) CD4 and CD8 T cells during the first cycle of treatment (Figure 4A, B). From the NK population, only immature NK cells (CD16–CD56bright CD3–) increased notably during the first cycle compared to mature (CD16 + CD56 + CD3–) ones (Figure 4C). Analysis of regulatory CD4 + CD25 + Foxp3+ T cells (Tregs) was not performed due to technical issues. In the previous phase I trial in patients with advanced, refractory solid tumors in which NHS-IL2 was used as monotherapy, it was shown that circulating Tregs responded actively to NHS-IL2 with a significant but transitory expansion early after each infusion [18]. The acute-phase proteins sIL-2Ralpha, C-reactive protein, and D-dimer showed a similar cycle-dependent pattern with a strong induction of about 5-fold reaching considerably high levels (>1000 pg/mL) 1 week after infusion (Figure 4D, and data not shown). Of note, increase in sIL-2Ralpha was observed after infusion of NHS-IL2 and not post-irradiation. The concentration of sIL-2Ralpha returned to near baseline level after each cycle. Neopterin and IP-10 (CXCL10) concentrations also increased in response to treatment, although to a lower extent than sIL-2Ralpha (data not shown).

No objective responses were observed during the trial, based on RECIST v1.0 criteria for tumor response, PFS, and OS. Median PFS was 2.9 months (95% confidence interval [CI] 1.5; 3.1) and median OS was 8.6 months (95% CI 4.9; not evaluable). As of the day of reporting there are 2 long-term survivors (both in good performance status 4 years after start of first-line chemotherapy). One patient discontinued treatment with NHS-IL2 in November 2013 and there were still no signs of disease activity in August 2014. Of note, both long-term survivors developed thyroiditis during treatment.