Patient Characteristics
Between February 2017 and January 2019, we screened ten patients at Osaka University Hospital and enrolled nine in this study (Fig. 1b). Because of the enforcement of a new law on clinical research in Japan, we could complete only three of the four planned dosing levels (levels 0-2). The study was terminated on March 31, 2019. Patient characteristics at baseline are summarized in Table 1. Of the nine patients, eight had NSCLC (four adenocarcinomas, three squamous carcinomas, and one mucoepidermoid carcinoma), and one had SCLC. One patient (#08) discontinued first-line pembrolizumab therapy and then received platinum-based chemotherapy as a second-line treatment. The median time from the last treatment to enrollment in this study was 62 days. The median observation period after registration was 55 days (range: 46-181 days).
Table 1
Patient characteristics at baseline and individual clinical effects of CpG ODN (K3)
Dose level | Patient | Primary cancer | Age | Gender | PS | Smoking history | Stage | Prior therapy | Response to prior therapy | Time from the completion of prior therapy, (days) | No. of doses of CpG ODN (K3), n (n)a | Response to CpG ODN (K3) | PFS, (days) |
5 mg sc | #01 | NSCLC | 62 | F | 0 | Yes | IIIA | CRT | PR | 45 | 3 (0) | non-CR/non-PD | 398 |
| #02 | NSCLC | 59 | M | 0 | No | IV | Chemotherapy | SD | 33 | 4 (0) | PD | 50 |
| #03 | NSCLC | 70 | M | 1 | Yes | Recurrent | CRT | CR | 109 | 9 (5) | non-CR/non-PD | 891b |
10 mg sc | #04 | NSCLC | 72 | M | 1 | Yes | Recurrent | Chemotherapy | PR | 88 | 8 (4) | non-CR/non-PD | 330c |
| #05 | NSCLC | 67 | F | 0 | No | IV | Chemotherapy | PR | 71 | 4 (0) | PD | 49 |
| #06 | SCLC | 67 | M | 0 | Yes | IV | Chemotherapy | PR | 43 | 5 (1) | PD | 72 |
0.2 mg/kg iv | #07 | NSCLC | 68 | M | 1 | Yes | IV | Chemotherapy | SD | 97 | 9 (5) | SD | 294 |
| #08 | NSCLC | 72 | M | 1 | Yes | IIIB | Chemotherapy | PR | 36 | 4 (0) | SD | 410b |
| #09 | NSCLC | 77 | M | 1 | Yes | IIIB | Chemotherapy | PR | 62 | 4 (0) | SD | 402b |
NOTE: a: The letter “n” refers to the total number of administrations of CpG ODN (K3) during the study (study treatment phase plus the compassionate use phase). The maximum number is 9. The number in parentheses “(n)” refers to a total number of administrations of CpG ODN (K3) in the compassionate use phase. The maximum number is 5. b: These patients remained in progression-free survival on the cutoff date. These survival data were censored at the last observation. c: This patient received an alternative anticancer therapy before radiological confirmation of disease progression. Survival data for this patient were censored at the beginning of the other treatment. Abbreviations: sc, subcutaneous injection; iv, intravenous administration; M, male; F, female; PS, performance status; NSCLC, non-small cell lung cancer; SCLC, small cell lung cancer; CRT, chemoradiotherapy; PR, partial remission; SD, stable disease; CR, complete remission; PD, progressive disease; PFS, progression-free survival. |
Treatment Course
Three patients each were treated at each dose level of CpG ODN (K3) (Table 1). One patient (#02) was concomitantly administered bevacizumab with CpG ODN (K3). None of the patients received any prophylactic medication for adverse events of CpG ODN (K3).
All patients except patient #01 completed four CpG ODN (K3) doses during the study treatment phase (Table 1). Patient #01 discontinued the study treatment after three doses of CpG ODN (K3) because of non-symptomatic exacerbation of pneumonitis, which was probably related to prior chemoradiation therapy (CRT). Four patients continued to receive CpG ODN (K3) in the compassionate use phase, and three of them completed the planned six-month treatment (Table 1).
DLT and MTD of CpG ODN (K3)
No DLTs occurred in any of the patients at any dose level of CpG ODN (K3). We could not determine the MTD of CpG ODN (K3) at the dose settings in this study and estimated it to be greater than 10 mg sc or 0.2 mg/kg iv.
Toxicity: Treatment-related Adverse Events
Treatment-related adverse events (TrAEs) at each dose level of CpG ODN (K3) are summarized in Table 2. In total, 27 systemic adverse events were reported (Supplementary Table S1), of which 16 (59.3%) were determined to be TrAEs. All systemic TrAEs were of grade 1 or 2, were reversible without any treatment, and improved after the next administration. No symptoms or signs such as fever of >38.0°C or lymphadenopathy, which were assumed to occur as immune reactions related to CpG ODN (K3), were observed. No physical signs or symptoms suggesting the induction of autoimmunity were observed. No apparent differences in the incidence of each TrAE were noted between the dose groups.
Table 2
Treatment-related Adverse Eventsa
| Dose Level of CpG ODN (K3) |
| 5 mg sc (n = 3) | 10 mg sc (n = 3) | 0.2 mg/kg iv (n = 3) |
Adverse event | Any Grade | Grade ≥3 | Any Grade | Grade ≥3 | Any Grade | Grade ≥3 |
White blood cell decreased | 0 | 0 | 1 | 0 | 0 | 0 |
Neutrophil count decreased | 1 | 0 | 1 | 0 | 0 | 0 |
Lymphocytes count decreased | 1 | 0 | 1 | 0 | 0 | 0 |
Platelet count decreased | 0 | 0 | 1 | 0 | 1 | 0 |
Eosinophil count increased | 1 | 0 | 0 | 0 | 0 | 0 |
ALT increased | 0 | 0 | 1 | 0 | 0 | 0 |
ALP increased | 0 | 0 | 0 | 0 | 1 | 0 |
Hypocalcemia | 0 | 0 | 1 | 0 | 0 | 0 |
Hypoalbuminemia | 0 | 0 | 1 | 0 | 0 | 0 |
Urinary protein | 0 | 0 | 1 | 0 | 0 | 0 |
Hematuria | 0 | 0 | 0 | 0 | 1 | 0 |
Pneumonitis | 1 | 0 | 0 | 0 | 0 | 0 |
Maculopapular rash | 0 | 0 | 0 | 0 | 1 | 0 |
Skin reactions at the local injection site |
Redness | 2 | 0 | 3 | 0 | NE | NE |
Induration | 1 | 0 | 2 | 0 | NE | NE |
NOTE: aTreatment-related adverse events were graded according to Common Terminology Criteria for Adverse Events (CTCAE) version 4.0. Abbreviations: sc, subcutaneous injection; iv, intravenous administration; AST, aspartate aminotransferase; ALT, alanine aminotransferase; NE, not evaluated. |
All patients of the sc groups except one (#02) exhibited mild local skin reactions at the injection sites within 48 h of the administration of CpG ODN (K3) (Table 2). There were no accompanying symptoms such as itching, and no treatment was required. No apparent differences in the severity of local skin reactions were noted between the two sc groups.
Clinical Effects of CpG ODN (K3)
Six of the nine patients (66.7%) had stable disease or non-CR/non-PD for six months as the best overall response (Table 1). The median PFS was 398 days (range, 49–891; Table 1 and Supplementary Fig. S1). No apparent association was noted between the dose level of CpG ODN (K3) and clinical effects. Notably, three patients (#03, #08, and #09) remained stable for over one year after the last administration of CpG ODN (K3).
Immune Assessment 1: Effects on Peripheral Blood
We evaluated the effects of CpG ODN (K3) on peripheral blood cells such as lymphocytes, neutrophils, and monocytes. The lymphocyte counts decreased temporarily within 24 h of the administration of CpG ODN (K3) in all patients and returned to baseline by the next dosing schedule (Fig. 2a). In contrast, neutrophil counts appeared to increase temporarily within 24 h of the administration of CpG ODN (K3) (Fig. 2b), although no regular changes in monocyte counts were observed (Fig. 2c).
Immune Assessment 2: Chemokine or Cytokine Production in Response to CpG ODN (K3)
We evaluated the changes in serum cytokine and chemokine levels. Their levels increased to varying degrees within 24 h of CpG administration (Supplementary Fig. S2). We mainly focused on the time courses of IFN-α2, TNF-α, IFN-γ, and CXCL10 in response to CpG ODN (K3) (Table 3). The relative changes in these cytokines and chemokines during the acute phase within 24 h of treatment were more prominent after the third injection than after the first injection.
Table 3
Summary of cytokines and chemokines released or produced in response to CpG ODN (K3).
| | Relative change before and 24 h after the administration of CpG ODN (K3) (%)a |
Dose Level | Pt. | 1st administration | 3rd administration |
IFN-α2 | TNF-α | IFN-γ | CXCL10 | IFN-α2 | TNF-α | IFN-γ | CXCL10 |
5 mg sc | #01 | ▲24.8 | ▲37.6 | ▲26.3 | 183.9 | ▲31.4 | 24.9 | 57.2 | 471.1 |
#02 | 8.3 | ▲1.92 | ▲13.9 | ▲3.3 | 18.5 | ▲15.9 | 23.9 | 161.0 |
#03 | 100.0 | ▲12.2 | 24.7 | 160.5 | 151.7 | ▲20.9 | 95.6 | 119.5 |
10 mg sc | #04 | 62.8 | 11.2 | 21.5 | ▲8.7 | ▲38.6 | 5.0 | 8.8 | 35.54 |
#05 | ▲59.4 | ▲22.7 | ▲36.3 | ▲35.5 | 67.8 | 17.0 | 61.5 | 97.6 |
#06 | ▲26.6 | ▲10.8 | 52.9 | 119.6 | 131.0 | 11.1 | 124.7 | 184.3 |
0.2 mg/kg iv | #07 | ▲48.2 | ▲15.3 | ▲11.7 | ▲37.7 | 100.0 | ▲5.0 | 24.9 | 2.59 |
#08 | 485.1 | 32.3 | ▲6.8 | 72.5 | ▲38.4 | ▲29.1 | ▲23.1 | ▲30.9 |
#09 | ▲25.8 | ▲26.6 | ▲11.7 | ▲15.6 | 191.7 | 10.0 | 19.4 | 53.2 |
aComparison of IFN-α2, TNF-α, or IFN-γ concentrations before and 24 h after the administration of CpG ODN (K3). Each value represents the relative change between before and 24 h after the administration of CpG ODN (K3), which was defined (post level – pre level)/pre level (%). ▲ italic represents a negative value. Abbreviations: sc, subcutaneous injection; iv, intravenous administration. |
Type I IFNs and inflammatory cytokines are the markers of innate immune activation by CpG ODN (K3) [11]. In six patients, the serum levels of IFN-α2 increased mildly after the third administration of CpG ODN (K3) [mean IFN-α2 level, 2.67–3.61 pg/mL; median relative change, 67.8% (range: -38.6–191.7%)], irrespective of the dose level or administration route (Fig. 3a and Table 3). In contrast, only a slight change in TNF-α levels was observed after the third CpG ODN (K3) dose in all patients [mean TNF-α level, 48.6–47.9 pg/mL; median relative change, 4.98% (range: -29.1–24.9%)] (Fig. 3b and Table 3). Very few biologically meaningful changes were noted in the levels of other inflammatory cytokines, including IL-1β, IL-6, and IL-12 (Supplementary Fig. S2), suggesting that CpG ODN (K3) predominantly elicited type I IFN responses rather than inflammatory cytokine responses.
Next, we assessed the markers for adaptive immune activation, such as IFN-γ for Th1 response, IL-4 for Th2 response, and IL-17 for Th17 response. Eight patients had mild increases in serum IFN-γ level after the third dose [mean IFN-γ level, 9.07–12.7 pg/mL; median relative change, 24.9% (range: -23.1–124.7%)] (Fig. 3c and Table 3). Five patients (#03, #05, #06, #07, and #09) had a mild-to-moderate increase in IFN-γ level in addition to a more than 50% relative increase in IFN-α2 (Table 3 and Fig. 3c). These results suggested that the increase in IFN-γ was related to the release of IFN-α2. However, the patients in the iv group had a weaker Th1 response than those in the sc groups. The mean relative changes in IFN-α2 in the iv and sc groups were 22.2% (range: 19.4–24.9%) and 76.4% (range: 23.9–124.7%), respectively. Unlike for IFN-γ, all patients showed minimal changes in serum IL-4 and IL-17 levels (Supplementary Fig. S2).
The serum levels of several chemokines, including CXCL10, CCL2, and CXCL9, increased temporarily within 24 h of CpG ODN (K3) administration (Fig. 3d and Supplementary Fig. S2). The serum level of CXCL10, also known as IFN-γ-induced protein 10, was increased in eight patients with relatively increased serum IFN-γ levels after the third dose [mean CXCL10 level, 351–676 pg/mL; median relative change, 97.6% (range: -30.9–471.1%)] (Fig. 3d and Table 3). These responses were also considerably marked in the sc groups, but not dose-dependent.
Immune Assessment 3: Induction of Adaptive Cellular Immunity in Response to CpG ODN (K3)
We evaluated the effect of CpG ODN (K3) on adaptive cellular immunity. Overall, no notable changes were observed in lymphocyte subsets of CD4- or CD8-positive CD3+ T cells and the percentages of their immune phenotypes (Supplementary Table S2).
First, we evaluated the functional characteristics of CD4+ T cells by analyzing T-bet and Foxp3 expression (Supplementary Fig. S3a). The percentages of T-bet+ and Foxp3+ CD4+ T cells were approximately 4–5% of the CD4+ T cells in PBMCs collected from two healthy subjects as controls (Supplementary Fig. S3b). Figure 4 (a–c) presents the classifications of patients into three groups based on T-bet or Foxp3 expression at baseline. In group 1, two patients (#03 and #09) exhibited a gradual increase in T-bet-expressing CD4+ T cells during the treatment course (24.7% and 13.7%, respectively, at baseline to 45.3% and 18.0% at day 29) (Fig. 4a). In contrast, in the other two groups, T-bet or Foxp3 expression in the CD4+ T cells changed minimally or fluctuated slightly during the treatment course (Fig. 4b-c). Immunophenotypically, especially in group 1, T-bet-expressing effector memory (EM) CD4+ T cells increased (Fig. 4d), whereas Foxp3-expressing EM CD4+ T cells remained almost unchanged during the treatment course (Supplementary Fig. S4).
Next, we evaluated the functional characteristics of CD8+ T cells by analyzing T-bet expression (Supplementary Fig. S3a). The percentage of T-bet+ CD8+ T cells was approximately 30% of the CD8+ T cells in the PBMCs collected from the two healthy subjects (Supplementary Fig. S3b). At baseline, all patients exhibited higher percentage of T-bet-expressing CD8+ T cells than those in the healthy subjects (Fig. 4e). During the treatment course, the percentage gradually increased (mean [range]: 49.8% [32.5–76.2] at baseline and 59.1% [41.9–84.9] at day 29, p = 0.0273; Fig. 4e). Immunophenotypically, both T-bet-expressing EM and terminally differentiated effector memory (TEMRA) CD8+ T cells significantly increased during the treatment course (mean [range]: EM, 52.7% [18.3–80.1] at baseline and 63.7% [30.5–77.5] at day 29, p = 0.0195; TEMRA, 82.3% [61.3–91.1] at baseline and 90.0% [75.7–97.8] at day 29, p = 0.0039; Fig. 4e). The relative changes in the T-bet-expressing EM and TEMRA CD8+ T cells significantly increased on days 15 and 29 (mean: EM, 23.3% at day 15, p = 0.0195, 29.8% at day 29, p = 0.0195; TEMRA, 7.6% at day 15, p = 0.0195, 10.0% at day 29, p = 0.0039; Fig. 4f).
Case Presentation (Patient #03)
A 70-year-old man had relapsed NSCLC and received concurrent radiation therapy (Fig. 5a). He began receiving 5 mg CpG ODN (K3) via sc. He exhibited mild local skin reaction at the injection sites (Fig. 5b), but no serious systemic TrAEs. Notably, he showed disease stabilization for more than two years without any treatment (Fig. 5a).
Within 24 h of CpG ODN (K3) administration, his lymphocyte level temporally decreased (Fig. 2a) and serum IFN-α2, IFN-γ, and CXCL10 levels temporally increased (Fig. 5c). The T-bet expression in CD4+ T cells was approximately twice that of any other patients before the administration of CpG ODN (K3). During the treatment course, T-bet expression increased, whereas Foxp3 expression remained low (Fig. 5d). Notably, T-bet expression in EM CD4+ T cells increased after repeated CpG ODN (K3) administration and remained high for six months (Fig. 5e). T-bet expression also increased in CD8+ T cells (Fig. 5e). In CD8+ T cells, the percentage of TEMRA cells increased, whereas that of EM cells decreased, but both immune phenotypes showed a gradual increase in T-bet expression during the treatment course (Fig. 5e). The Th1-type cellular immune response sustained by CpG ODN (K3) possibly contributed to disease stabilization and led to better clinical results.