Harnessing the immune system by blocking the programmed cell death protein 1 (PD-1) pathway has been a major breakthrough in non-small-cell lung cancer treatment. Nonetheless, many patients fail to respond to PD-1 inhibition. Using three syngeneic models, we demonstrate that short-term starvation synergizes with PD-1 blockade to inhibit lung cancer progression and metastasis. This antitumor activity was linked to a reduction in circulating insulin-like growth factor 1 (IGF-1) and a downregulation of IGF-1 receptor (IGF-1R) signaling in tumor cells. A combined inhibition of IGF-1R and PD-1 synergistically reduced tumor growth in mice. This effect required CD8 cells, boosted the intratumoral CD8/Treg ratio and led to the development of tumor-specific immunity. In patients with non-small-cell lung cancer, high plasma levels of IGF-1 or high IGF-1R expression in tumors was associated with resistance to anti-PD-1–programmed death-ligand 1 immunotherapy. In conclusion, our data strongly support the clinical evaluation of IGF-1 modulators in combination with PD-1 blockade.
Blockade of the interaction between programmed cell death protein 1 (PD-1) and its counter ligand programmed death-ligand 1 (PD-L1) through the use of monoclonal antibodies has shown remarkable clinical activity in non-small-cell lung cancer (NSCLC)1,2. These drugs reactivate antigen-specific effector T cells, boosting the immune response to cancer. However, they are ineffective in a significant percentage of patients, and some initial responders eventually develop resistance. Mechanisms mediating resistance include the presence in the tumor microenvironment of immunosuppressive cell subsets such as myeloid-derived suppressor cells (MDSCs), immune-suppressive macrophages, regulatory T cells (Treg cells) and immature dendritic cells, which prevent the infiltration and activation of cytotoxic T cells3. Treatment modalities designed to overcome tumor-associated immunosuppressive hurdles should provide substantial therapeutic improvement in the restoration of effector T cells.
Altered metabolism has been described as a core hallmark of cancer4. Short-term starvation (STS) or fasting-mimicking diets elicit well-known metabolic adaptations, including changes in the systemic levels of hormones and growth factors such as insulin, glucagon, growth hormone, insulin-like growth factor 1 (IGF-1), glucocorticoids or adrenaline5. As a consequence, normal cells enter in a slow division mode and reprogram their metabolic and energetic requirements as a protective mechanism against stress and toxic insults. In contrast, tumor cells present difficulties in managing this metabolic switch, increasing their vulnerability to cytotoxic drugs. Consistent with this assumption, a number of studies have shown that fasting sensitizes tumors to chemotherapy, while protecting normal cells from its undesirable toxic side effects6,7. Interestingly, the synergy between fasting-like conditions and chemotherapy seems to rely on antitumor immune responses8,9. Fasting-like conditions increase tumor immunogenicity and cytotoxic CD8 tumor-infiltrating lymphocytes—two prerequisites for the antitumor response mediated by PD-1–PD-L1 blockade10.
Based on these premises, we evaluated the effects of STS in anti-PD-1 immunotherapy. Our results show that STS sensitizes tumors to anti-PD-1 treatment by reducing IGF-1 levels. Accordingly, a combination of IGF-1 receptor (IGF-1R) inhibition and PD-1 blockade results in a more permissive environment for antitumor immune responses, which impairs lung cancer progression. Finally, our data suggest that patients with lung cancer and with low levels of circulating IGF-1 or reduced expression of IGF-1R in their tumor cells are more likely to benefit from anti-PD-1/PD-L1 immunotherapy.
STS synergizes with PD-1 blockade to inhibit lung cancer progression
Recent data have shown that STS or fasting-mimicking diets enhance the immune response to cancer8,9. Therefore, we tested the effect of combining STS with the anti-PD-1 monoclonal antibody RMP1-14 on the growth of subcutaneously implanted 393P cells—a syngeneic mouse model of KRAS-driven lung adenocarcinoma11. STS was achieved by intermittent periods of complete food withdrawal while allowing access to water ad libitum. To mimic the clinical setting, treatments were performed in established tumors during a defined period (STS for days 14–16, 19–21 and 25–26, and anti-PD-1 at days 15, 20 and 26). The combination of STS with PD-1 blockade significantly reduced 393P tumor growth and improved survival compared with the effect of each treatment alone (Fig. 1a,b and Extended Data Fig. 1a). Tumor regression was observed in six out of eight mice after STS + anti-PD-1 treatment, with a complete response in four mice (Fig. 1c). The four mice with complete responses were resistant to tumor rechallenge, suggesting the generation of a long-lasting memory response (Fig. 1d). We also investigated the effect of the combination treatment in a syngeneic mouse model based on the subcutaneous implantation of Lewis lung carcinoma (LLC) cells. STS (days 6–8, 10–12 and 14–15) synergized with PD-1 (days 7, 11 and 15) to attenuate tumor growth and increase the survival of tumor-bearing mice (Fig. 1e,f and Extended Data Fig. 1b). Notably, in a highly aggressive and metastatic lung cancer model based on the intracardiac inoculation of Lacun3 cells12, the combined treatment was also able to increase the survival of mice (Fig. 1g). To further support the relevance of the PD-1–PD-L1 axis in the combined treatment, PD-L1 was inhibited with the monoclonal antibody 10F.9G2 in LLC tumor-bearing mice using the same treatment schedule as described above. The combined STS + anti-PD-L1 treatment was able to hamper tumor growth (Fig. 1h).
Next, we characterized the changes in immune cell populations resulting from the combined treatment. The experiment was performed 18 d after LLC tumor implantation, when significant differences in growth were already evident (Fig. 2a). The combination treatment led to a significant increase in tumor-infiltrating CD8 and natural killer cells, whereas the proportion of CD4 and B cells in the tumors was significantly reduced (Fig. 2b and Extended Data Fig. 2a). Moreover, the percentage of intratumoral Treg cells among total CD4 T cells was significantly decreased in the combined group, whereas the ratio of CD8/Treg was increased (Fig. 2b). Interestingly, tumor-infiltrating CD8 and CD4 T cells from the combined treatment group showed a marked reduction in the expression of the exhaustion marker PD-1 (Fig. 2c and Extended Data Fig. 2b). No remarkable differences were found in intratumoral MDSCs, macrophages or dendritic cells (Extended Data Fig. 2c). Lymphocytic subpopulations in peripheral blood were also analyzed. Differences in CD4 and B cells were consistent with those found in the tumor microenvironment, which was not the case for CD8, Treg and natural killer cells (Extended Data Fig. 3). Taken together, these data indicate that the combination of intermittent fasting with PD-1 or PD-L1 blockade reduces tumor growth and increases survival in syngeneic models of lung cancer. This effect is associated with a decrease of Treg cells and an increase of effector T cells in the tumor microenvironment.
STS-induced sensitivity to PD-1 blockade is mediated by an antitumor CD8 T cell response
Previous studies had demonstrated that STS-like conditions increase the immunogenicity of tumor cells. This was attributed to the ability of STS to induce autophagy and to downregulate the expression of heme oxygenase-1 (HO-1)8,9. To identify changes induced by STS in LLC cells, we cultured these cells under normal (2 g l−1 glucose and 10% FetalClone) or STS-like conditions (0.5 g l−1 glucose and 1% FetalClone). These culture conditions mimic those of normally fed or fasted mice, respectively7. According to previous findings8,9, STS increased the autophagy marker LC3-II/LC3-I ratio and decreased the levels of the HO-1 gene (Hmox1) in LLC cells (Fig. 3a and Extended Data Fig. 4a). Based on these results, we evaluated whether STS treatment would increase the immunogenicity of LLC cells. Following an in vivo experimental approach previously applied by Di Biase et al.8, we preconditioned LLC cells by culturing them under normal or STS-like conditions, and grafted them into the left flank (immunization side) of immunocompetent mice. After 1 week, all mice received a subcutaneous inoculation of normally cultured LLC cells on the right flank (naïve tumor side) and were treated with anti-PD-1 or vehicle as outlined in Fig. 3b. A reduction in tumor volume that did not reach statistical significance was observed in the immunization side after PD-1 blockade (Fig. 3c). This limited response may have been due to the late administration of anti-PD-1 treatment (2 weeks after tumor implantation). More importantly, a significant abscopal effect associated with anti-PD-1 treatment was observed in the naïve tumor side (Fig. 3d), pointing to the implication of the immune response in the STS-induced sensitivity to PD-1 blockade.
In relation to the identity of the effector cells, depletion of CD8 T cells completely abrogated the antitumor efficacy of the STS + anti-PD-1 treatment in LLC tumors. Depletion of CD4 T cells resulted in mild suppression of the antitumor effect, whereas depletion of natural killer cells had no effect (Fig. 3e). These results suggest that STS conditions sensitize lung tumor cells to the action of cytotoxic CD8 cells reactivated by PD-1 blockade.
Inhibition of IGF-1–IGF-1R signaling in tumor cells mediates the antitumor efficacy of the STS + anti-PD-1 combination
STS leads to reduced IGF-1 plasma levels and downstream cell signaling effects, suggesting that the IGF-1–IGF-1R axis is a major regulator of the systemic response to fasting6. Among a range of biological effects, inhibition of IGF-1 signaling increases cancer immunogenicity and boosts antitumor CD8 T cell responses13. We therefore explored the possibility that the antitumor responses generated after the combined STS + anti-PD-1 treatment were dependent on IGF-1 signaling. In tumor-bearing mice, serum IGF-1 levels were significantly reduced after STS, irrespective of anti-PD-1 treatment (Fig. 4a). Moreover, recombinant IGF-1 administration during STS cycles abrogated the antitumor activity of the STS + anti-PD-1 regimen (Fig. 4b).
The contribution of the IGF-1–IGF-1R axis to the primary resistance to anti-PD-1 treatment was evaluated in LLC cells stably transduced with lentivirus containing two different short hairpin RNAs (shRNAs) against mouse Igf1r messenger RNA (mRNA), or an empty vector (Fig. 4c). Igf1r downregulation did not affect in vitro proliferation (data not shown). However, subcutaneously implanted short hairpin IGF-1R (shIGF-1R)-transduced tumors became sensitive to anti-PD-1, whereas tumor cells transduced with the control vector remained refractory to the treatment (Fig. 4d). The IGF-1R-mediated signaling events induced by fasting were analyzed by western blotting for phosphorylated and total AKT, p70S6K and p42 or p44 (p42/44) mitogen-activated protein kinase (MAPK) in LLC cells cultured under normal or STS-like conditions for 0.5, 6, 12, 24, 48 and 72 h in the presence of either 1 µg ml−1 recombinant IGF-1 (rIGF-1) or vehicle. Under normal conditions, the presence of rIGF-1 did not alter its downstream signaling (Fig. 5a). However, STS conditions inhibited AKT, p70S6K and p42/44 MAPK activation, which was rescued with rIGF-1 addition (Fig. 5a). The effect of rIGF-1 was abrogated by genetic or pharmacological inhibition of IGF-1R (Fig. 5b). Moreover, the addition of rIGF-1 to LLC cells cultured under STS-like conditions inhibited the increase of the autophagy marker LC3-II/LC3-I ratio (Extended Data Fig. 4b) and abolished the downregulation of Hmox1 induced by the culture of LLC cells under STS-like conditions (Fig. 5c). Taken together, these results suggest that the IGF-1–IGF-1R axis on lung cancer cells contributes to the primary resistance of lung tumors to PD-1 blockade in association with the regulation of autophagy and HO-1 expression.
Levels of IGF-1 and IGF-1R predict response to PD-1 blockade in patients with NSCLC
Due to the observed association between circulating IGF-1 levels and anti-PD-1 activity, we hypothesized that the levels of IGF-1 may predict a clinical benefit to anti-PD-1 treatment. We measured the concentration of IGF-1 in plasma samples collected from 40 patients with NSCLC treated with anti-PD-1–PD-L1 monotherapy. Overall, 13 patients obtained durable clinical benefit (DCB): 11 partial responses and two disease stabilizations lasting more than 12 months. Plasma samples from patients with DCB had statistically significant lower levels of circulating IGF-1 than plasma samples from patients with no durable benefit (NDB) (Fig. 6a). Moreover, IGF-1R expression levels in tumors from patients with NSCLC who obtained DCB with anti-PD-1–PD-L1 monotherapy were significantly lower than those from patients with NDB (Fig. 6b,c). These data suggest that the activity of the IGF-1 signaling pathway is inversely associated with the efficacy of anti-PD-1–PD-L1 therapy in patients with NSCLC.
Concomitant blockade of IGF-1R and PD-1–PD-L1 is effective against established lung tumors
To better substantiate the role of IGF-1R in the efficacy of anti-PD-1 immunotherapy, we evaluated the antitumor activity of the combined administration of anti-PD-1 and PQ401—a diarylurea-based antagonist of IGF-1R14. Administration of this antagonist (day 14, then every other day until day 30) in combination with the anti-PD-1 monoclonal antibody RMP1-14 (days 15, 19 and 22) significantly reduced tumor growth of 393P established tumors compared with the effect of vehicle or each treatment alone (Fig. 7a and Extended Data Fig. 5). A significant improvement in survival was also observed in the combined group (Fig. 7b). Interestingly, complete tumor regression was observed in five out of ten mice treated with the combination (Fig. 7c). Long-term survivors were protected from tumor rechallenge, suggesting long-lasting memory response (Fig. 7d). Similar effects were found in LLC tumors treated with the anti-PD-1 antibody (Fig. 7e).
The tumor-immune infiltrate in the LLC tumors treated with anti-PD-1 in combination with PQ401 was characterized by flow cytometry (Fig. 7f and Extended Data Fig. 6a). None of the treatments modified the frequency of MDSCs (monocytic, granulocytic or total MDSCs), macrophages or dendritic cells (Extended Data Fig. 6a). However, the combined PQ401 + anti-PD-1 treatment led to a significant increase in tumor-infiltrating CD8, CD4, natural killer and B cells (Fig. 7f and Extended Data Fig. 6a). Moreover, the percentage of intratumoral Treg cells among total CD4 T cells was significantly decreased in the combined group, whereas the intratumoral CD8/Treg ratio was augmented (Fig. 7f). An analysis of splenocytes showed a significant increase in the frequency of LLC-specific interferon-γ (IFN-γ)-producing T cells in PQ401 + anti-PD-1-treated mice compared with those animals treated with the single agents or vehicle (Fig. 7g). The combined regimen also reduced the expression of the exhaustion markers PD-1 and glucocorticoid-induced tumor necrosis factor receptor (GITR) in both CD8 and CD4 T cells, as well as lymphocyte activation gene 3 (LAG-3) in CD4 T cells (Fig. 7h and Extended Data Fig. 6b). No significant differences in splenic CD8, CD4, B, natural killer and Treg cells, MDSCs, macrophages or dendritic cells were found between experimental groups (Extended Data Fig. 7). Finally, the combined blockade of PD-L1 (with the monoclonal antibody 10F.9G2) and IGF-1R (with PQ401) also attenuated tumor growth in LLC tumor-bearing mice (Fig. 7i). Collectively, these data suggest that the combined use of the IGF-1R inhibitor PQ401 and the PD-1/PD-L1-blocking antibodies reverses the LLC immune-suppressive microenvironment and promotes a specific antitumor immune response.
Resistance to PD-1–PD-L1 inhibitors poses a major challenge to the therapeutic management of patients with lung cancer. In the present study, we demonstrate that STS diminishes the levels of circulating IGF-1 to sensitize tumors to PD-1 blockade in preclinical models of NSCLC. The rationale behind this approach is that IGF-1 allows lung tumors to evade the antitumor immune response. Accordingly, inhibition or downregulation of IGF-1R synergizes with PD-1 blockade to reverse T cell exhaustion and to restore antitumor immunity. Our study opens up new avenues to stratifying and treating patients with lung cancer.
There is compelling evidence that caloric restriction has cancer-preventive effects15,16. A number of preclinical studies have shown that the combination of fasting-like conditions with chemotherapy protects the host from toxic side effects and leads to a major delay in cancer progression6,7. This effect is mediated, at least in part, by antitumor immune responses. In this regard, fasting-like conditions enhance the frequency of cytotoxic CD8 T cells and reduce the presence of Treg cells within tumors8. Similarly, treatment with the caloric restriction mimetic hydroxycitrate induces the depletion of Treg cells, thereby improving anticancer immunosurveillance and reducing tumor burden9. In accordance with these findings, an increase of activated T cells and cytolytic natural killer cells was found in the peripheral blood of fasted patients with cancer17. Using preclinical lung cancer models, we now demonstrate that STS markedly enhances the antitumor activity of an anti-PD-1 antibody, affecting lung tumor growth and metastatic spread. Interestingly, the combined treatment leads to a lower frequency of Treg cells and a higher frequency of CD8 T cells in the tumor microenvironment. Of note, in the lung cancer model used in these experiments, STS alone did not seem to affect the frequency of tumor-infiltrating Treg or CD8 cells. Depletion experiments revealed that the function of CD8 T cells is an absolute requirement for the antitumor activity of the combined regimen. Moreover, the inoculation of tumor cells grown in STS medium led to favorable effects of the anti-PD-1 antibody on distant non-fasted tumor grafts, suggesting that STS enhances anticancer immunosurveillance to facilitate the therapeutic activity of anti-PD-1 treatment. Collectively, these findings show that STS sensitizes tumors against anti-PD-1 blockade.
Our work supports the idea that IGF-1—a key factor of the mammalian systemic response to fasting5—is involved in the STS-mediated response to PD-1 blockade. Previous preclinical data had shown that IGF-1 downregulation reverses the phenotype that allows tumor cells to evade the immune response8,9. Both autophagy and the decrease of HO-1 levels in tumor cells are associated with a reduction in the number of tumor-infiltrating Treg cells and are at the core of fasting-mediated CD8-dependent tumor cytotoxicity8,9,18. Although in-depth mechanistic studies are still needed to fully understand the role of IGF-1 in starved tumor cells, we show here that rIGF-1 diminishes the autophagy marker LC3-II/LC3-I ratio and restores Hmox1 expression in lung cancer cells grown under STS-like conditions. We have also demonstrated that both genetic and pharmacological inhibition of the IGF-1–IGF-1R axis enhance the antitumor activity of anti-PD-1–PD-L1 antibodies against lung cancer. The combined treatment led to a lower frequency of Treg cells and a higher frequency of CD8 T cells in the tumor microenvironment, suggesting that the IGF-1–IGF-1R pathway sustains tumor-associated immunosuppression. Moreover, the combination treatment was associated with both decreased expression of exhaustion markers on T cells and the generation of tumor-specific IFN-γ-producing T cells, which suggests a more complete restoration of CD8 T cell effector activity.
Genetic abrogation of Igf1r expression in lung cancer cells sensitized tumors to anti-PD-1 treatment to a similar extent as STS or systemic pharmacological blockade of the receptor. This finding suggests that cancer cells are the most plausible target responsible for STS-mediated sensitization to anti-PD-1 treatment. Nevertheless, the IGF-1–IGF-1R axis regulates functional processes in many other cell types, which may as well contribute to the antitumor activities of the combined treatment. In particular, the IGF-1–IGF-1R axis represents an important stimulatory pathway in the complex process of T cell activation, proliferation and survival19. Administration of IGF-1 increases lymphopoiesis and guides the differentiation state of T cells. In fact, it has been described that the induction of Treg cells can be enhanced by the addition of IGF and suppressed by the inhibition of IGF-1R20,21. Other immune and stromal cells are critically regulated by IGF-1. In contrast with the results obtained in our study, IGF-1 promotes the development and cytotoxic activity of natural killer cells22. Depletion of IGF-1R in the myeloid lineage reduces the expression of M2 markers23. In relation to the tumor stroma, IGF stimulates endothelial cell migration, vascular endothelial growth factor synthesis and angiogenesis24. Therefore, it is clear that further investigation is required to obtain a more complete understanding of the cell-specific roles played by IGF-1 in its interaction with PD-1–PD-L1 activities. At this point, we can conclude that IGF-1R signaling in lung cancer cells is involved in the primary resistance to anti-PD-1 treatment. These preclinical data led us to evaluate the implication of the IGF-1 axis in the resistance of patients to PD-1 blockade. We found that high levels of circulating IGF-1 or high IGF-1R expression on tumor cells are associated with resistance to anti-PD-1–PD-L1 therapy in patients with advanced NSCLC. Additional studies will be required to validate these observations, as well as to evaluate the association of IGF-1 levels and IGF-1R expression with other predictive markers of response.
In conclusion, this study supports the potential of IGF-1 or its signaling pathway as a therapeutic target for lung cancer in the context of PD-1 blockade. Overexpression of IGF-1R on lung tumor cells predicts poor overall survival25, and elevated levels of circulating IGF-1 have been associated with an increased risk of lung cancer26. In clinical trials, IGF-1R antagonists have failed to increase survival in patients with advanced NSCLC when added to standard chemotherapy27,28. Our present data could constitute the basis for clinically testing these IGF-1R inhibitors in conjunction with PD-1–PD-L1 blockade therapies, although potential toxicities should be taken into consideration. Finally, following confirmation in independent cohorts, the IGF-1 pathway may be useful for predicting response to anti-PD-1 monoclonal antibodies in patients with lung cancer.
Clinical specimens were obtained at the Clínica Universidad de Navarra, Spain. The study protocol was approved by the Institutional Review Board (reference: 111/2010). All patients gave written informed consent. Plasma samples (from 40 patients) or formalin-fixed paraffin-embedded lung cancer tissues (from 13 patients) were obtained before treatment. The clinical features of the cases are shown in Supplementary Tables 1 and 2. Patients were treated from November 2013 to June 2018 with anti-PD-1 (nivolumab or pembrolizumab) or anti-PD-L1 (durvalumab) agents. The median follow-up was 9 months (range: 1–55 months). Tumor response was evaluated following Response Evaluation Criteria in Solid Tumors 1.1 guidelines29. DCB was defined as complete response, partial response or stable disease lasting at least 12 months, while NDB was defined as progressive disease or partial response lasting <12 months.
Cell lines and reagents
393P lung adenocarcinoma cells, derived from KrasLA1/+;p53R172HΔG mice, were a gift from J. M. Kurie (The University of Texas MD Anderson Cancer Center, Houston, Texas). The mouse LLC cell line was purchased from the American Type Culture Collection. The Lacun3 cell line was previously established by our group from a combined silica and N-nitrosodimethylamine-induced mouse lung adenocarcinoma12. All cells were cultured in RPMI 1640 medium (Invitrogen), 10% FetalClone (Thermo), 100 U ml−1 penicillin and 100 U ml−1 streptomycin (Invitrogen). Lentiviral transduction of shRNAs against Igf1r (TRCN0000321979 and TRCN0000023489; Sigma–Aldrich) was performed as previously described30. Transduction of the empty vector (PLKO.1-puro; Sigma–Aldrich) was used as a control.
For in vivo STS, mice underwent complete food deprivation with free access to water for two cycles of 48 h and one cycle of 24 h. In vitro STS-mimicking conditions were achieved by culturing cells in glucose-free RPMI 1640 medium (Invitrogen) supplemented with 0.5 g l−1 glucose (PanReac) and 1% FetalClone (Thermo). To mimic ad libitum conditions in vitro, cells were cultured with 2 g l−1 glucose and 10% FetalClone. This protocol has been described previously7.
Mouse lung cancer models and therapeutic schedules
All animal experiments were conducted in accordance with the protocols approved by the institutional animal care committee (references 113-13 and 049-18). 393P or LLC cells (2 × 106) were subcutaneously injected in the flanks of female Sv/129 or C57BL/6J mice, respectively. 393P engraftments were allowed to grow for 14 d, and mice were treated with STS (days 14–16, 19–21 and 25–26) and/or 100 μg anti-PD-1 (RMP1-14; Bio X Cell) at days 15, 20 and 26. For PQ401 + anti-PD-1 combination treatment, 393P tumor-bearing mice were treated with anti-PD-1 intraperitoneally at days 15, 19 and 22 and/or with 50 mg kg−1 PQ401 (MedChemExpress) or vehicle intraperitoneally at day 14 and every other day until day 30. The PQ401 treatment schedule and dose have been described previously14. The Sv/129 mice used for STS + anti-PD-1 and PQ401 + anti-PD-1 experiments were 24 and 48 weeks old, respectively.
LLC tumors were allowed to grow for 6 d before treatment. STS was performed at days 6–8, 10–12 and 14–15, and anti-PD-1 or anti-PD-L1 (10F.9G2; Bio X Cell) intraperitoneal inoculation was performed at days 7, 11 and 15. For PQ401 + anti-PD-1 combination treatment, LLC tumor-bearing mice were treated with anti-PD-1 intraperitoneally at days 7, 11 and 14 and/or with 50 mg kg−1 PQ401 or vehicle intraperitoneally at day 6 and every other day. This treatment schedule was also performed to test the PQ401 + anti-PD-L1 combination, except for anti-PD-L1 injections (days 7, 11 and 15). Depletion of CD8, CD4 or natural killer cells was performed by intraperitoneal injection of 100 μg anti-mouse CD8a (clone 2.43; Bio X Cell), CD4 (clone GK1.5; Bio X Cell) or NK1.1 (clone PK136; Bio X Cell), respectively, at days 6, 10, 14 and 18 after LLC inoculation. Recombinant IGF-1 (Prospec) was administered at a dose of 200 µg kg−1 intraperitoneally every 12 h during STS, as described previously6. C57BL/6J mice for these experiments were 9 weeks old.
The survival of Lacun3 tumor-bearing mice was analyzed after inoculation of 2 × 105 cells into the left cardiac ventricle of 9-week-old BALB/c mice, as previously described31. In this model, treatments were performed as described above, except for STS treatment (days 2–4, 6–8 and 10–11) and anti-PD-1 injections (days 3, 7 and 11).
Gene expression and western blotting
Quantification of Igf1r and Hmox1 gene expression was performed by real-time PCR as described previously32. Gapdh was used as an endogenous control. Primer sequences were as follows: Igf1r (forward: 5′-GGGGCTCCTGTTTCTCTCC-3′; reverse: 5′-GCCTTGGAGATGAGCAGGAT-3′); Hmox1 (forward: 5′-AGCCCAGTCCGGTGATGGA-3′; reverse: 5′-GCTCCTCAGGGAAGTAGAG-3′); and Gapdh (forward: 5′-ACTTTGTCAAGCTCATTTCC-3′; reverse: 5′-TGCAGCGAACTTTATTGATG-3′). Phosphorylated and total AKT, p42/44 MAPK and p70S6K (clones 108D2 and 49D7, respectively) and total LC3A/B (clone D3U4C) were analyzed by western blotting using rabbit anti-mouse antibodies from Cell Signaling Technology. Antibodies against GAPDH (clone 9484; Abcam) or β-actin (clone AC-15; Sigma–Aldrich) were used as controls.
Mouse circulating levels of IGF-1 were measured by enzyme-linked immunosorbent assay (MG100, R&D), following the manufacturer’s protocol. Human circulating levels of IGF-1 were measured using a Milliplex MAP kit (Millipore) with Luminex xMAP technology, following the manufacturer´s protocol. For IGF-1R immunohistochemical analysis, paraffin was removed and the endogenous peroxidase activity quenched as described previously32. Antigen retrieval was performed in a Lab Vision PT module for 20 min in EDTA buffer (pH 9) (Thermo Scientific). Next, sections were incubated overnight at 4 °C with an anti-human IGF-1R undiluted antibody (Ventana) followed by detection with the EnVision system (Dako). The extension and intensity of the staining was evaluated, and a H-score was calculated by assessment of both the intensity of staining (graded as: 0, non-staining; 1, weak; 2, median; or 3, strong) and the percentage of positive cells, as described previously33.
For the IFN-γ-based enzyme-linked immunospot (ELISpot), 8 × 105 splenocytes per well were cultured overnight in the presence (or not) of 8 × 104 irradiated LLC cells in ELIIP plates (Millipore) coated with purified anti-IFN-γ antibody (clone AN-18; Mabtech). After washing, wells were incubated with biotinylated anti-IFN-γ antibody (clone R4-6A2; Mabtech) and streptavidin-ALP (Mabtech). The assay was developed by adding BCIP/NBT substrate (Mabtech). Spot-forming cells were quantified using a CTL ImmunoSpot S6 micro-analyzer (Cellular Technology).
Flow cytometry analysis
Blood (50 µl) from tumor-bearing mice was collected, and tumors and spleens were harvested and mechanically disaggregated. Afterwards, erythrocytes were removed as previously described34, and the resulting single-cell suspensions were preincubated with a monoclonal antibody to mouse CD16/CD32 (Fc block; 2.4G2; BD Pharmingen), then labeled with fluorochrome-conjugated antibodies against mouse CD45 (30-F11), CD4 (RM4-5), CD8a (53-6.7), CD19 (6D5), NK1.1 (PK136), CD25 (PC61), F4/80 (BM8), CD11b (M1/70), Ly6C (HK1.4), Ly6G (1A8), CD11c (N418), LAG-3 (C9B7W), GITR (DTA-1) or PD-1 (29F.1A12) (all from BioLegend) diluted in FACS buffer (phosphate buffered saline, 0.1% NaN3 and 1% bovine serum albumin). For intracellular staining, cells were fixed using 2% formaldehyde (Polysciences), permeabilized using permeabilization buffer (eBioscience) and labeled for FOXP3 (3G3; Abcam). Cells were acquired using a BD Biosciences FACSCanto II flow cytometer. Data were analyzed using FlowJo software (Tree Star). The gating strategy for the flow cytometry analyses is shown in Supplementary Figs. 1 and 2.
Statistics and reproducibility
No statistical method was used to predetermine sample sizes. These were chosen based on estimates from previous studies, to enable significant statistical analysis. No data were excluded from the analyses. Tumor-bearing mice were randomized into experimental and control groups according to tumor burden. The investigators were not blinded to allocation during the experiments and outcome assessment, except for the analysis of clinical samples. Comparisons between two groups were performed using the Mann–Whitney U-test, except in the case of circulating human IGF-1, for which the differences in expression were analyzed by Welch’s t-test. Comparisons between treatment strategies were performed by Kruskal–Wallis test with post hoc Mann–Whitney U-test. When required, data were expressed as means ± s.e.m. Survival curves were generated using the Kaplan–Meier method, and differences were evaluated with the log-rank test. For these analyses, survival times were defined as the period from inoculation of the cells until the mice were euthanized or succumbed to tumor growth. All of the statistical tests were two sided. GraphPad Prism 5.0 software was used for the statistical analyses.
Further information on research design is available in the Nature Research Reporting Summary linked to this article.
Brahmer, J. et al. Nivolumab versus docetaxel in advanced squamous-cell non-small-cell lung cancer. N. Engl. J. Med. 373, 123–135 (2015).
Horn, L. et al. First-line atezolizumab plus chemotherapy in extensive-stage small-cell lung cancer. N. Engl. J. Med. 379, 2220–2229 (2018).
Melero, I. et al. Evolving synergistic combinations of targeted immunotherapies to combat cancer. Nat. Rev. Cancer 15, 457–472 (2015).
Ward, P. S. & Thompson, C. B. Metabolic reprogramming: a cancer hallmark even Warburg did not anticipate. Cancer Cell 21, 297–308 (2012).
Nencioni, A., Caffa, I., Cortellino, S. & Longo, V. D. Fasting and cancer: molecular mechanisms and clinical application. Nat. Rev. Cancer 18, 707–719 (2018).
Lee, C. et al. Reduced levels of IGF-I mediate differential protection of normal and cancer cells in response to fasting and improve chemotherapeutic index. Cancer Res. 70, 1564–1572 (2010).
Lee, C. et al. Fasting cycles retard growth of tumors and sensitize a range of cancer cell types to chemotherapy. Sci. Transl. Med. 4, 124ra27 (2012).
Di Biase, S. et al. Fasting-mimicking diet reduces HO-1 to promote T cell-mediated tumor cytotoxicity. Cancer Cell 30, 136–146 (2016).
Pietrocola, F. et al. Caloric restriction mimetics enhance anticancer immunosurveillance. Cancer Cell 30, 147–160 (2016).
Tang, H. et al. Facilitating T cell infiltration in tumor microenvironment overcomes resistance to PD-L1 blockade. Cancer Cell 29, 285–296 (2016).
Gibbons, D. L. et al. Contextual extracellular cues promote tumor cell EMT and metastasis by regulating miR-200 family expression. Genes Dev. 23, 2140–2151 (2009).
Bleau, A. M. et al. New syngeneic inflammatory-related lung cancer metastatic model harboring double KRAS/WWOX alterations. Int. J. Cancer 135, 2516–2527 (2014).
Trojan, J. et al. Treatment and prevention of rat glioblastoma by immunogenic C6 cells expressing antisense insulin-like growth factor I RNA. Science 259, 94–97 (1993).
Gable, K. L. Diarylureas are small-molecule inhibitors of insulin-like growth factor I receptor signaling and breast cancer cell growth. Mol. Cancer Ther. 5, 1079–1086 (2006).
Willcox, B. J. et al. Caloric restriction, the traditional Okinawan diet, and healthy aging: the diet of the world’s longest-lived people and its potential impact on morbidity and life span. Ann. NY Acad. Sci. 1114, 434–455 (2007).
Mattison, J. A. et al. Impact of caloric restriction on health and survival in rhesus monkeys from the NIA study. Nature 489, 318–321 (2012).
de Braud, F. et al. Abstract B022: metabolic and immunologic effects of the fasting mimicking diet in cancer patients. in AACR-NCI-EORTC International Conference on Molecular Targets and Cancer Therapeutics. Mol. Cancer Ther. 17(1 Suppl), Abstract nr B022 (2018).
Dey, M. et al. Heme oxygenase-1 protects regulatory T cells from hypoxia-induced cellular stress in an experimental mouse brain tumor model. J. Neuroimmunol. 266, 33–42 (2014).
Idelman, G. et al. The role of the IGF system in T-lymphocyte activation. Cancer Res. 67, 4392 (2007).
Bilbao, D., Luciani, L., Johannesson, B., Piszczek, A. & Rosenthal, N. Insulin-like growth factor-1 stimulates regulatory T cells and suppresses autoimmune disease. EMBO Mol. Med. 6, 1423–1435 (2014).
Miyagawa, I. et al. Induction of regulatory T cells and its regulation with insulin-like growth factor/insulin-like growth factor binding protein-4 by human mesenchymal stem cells. J. Immunol. 199, 1616–1625 (2017).
Ni, F. et al. IGF-1 promotes the development and cytotoxic activity of human NK cells. Nat. Commun. 4, 1479 (2013).
Spadaro, O. et al. IGF1 shapes macrophage activation in response to immunometabolic challenge. Cell Rep. 19, 225–234 (2017).
Bach, L. A. Endothelial cells and the IGF system. J. Mol. Endocrinol. 54, R1–R13 (2015).
Nakagawa, M. et al. Clinical significance of IGF1R expression in nonsmall-cell lung cancer. Clin. Lung Cancer 13, 136–142 (2012).
Yu, H. et al. Plasma levels of insulin-like growth factor-I and lung cancer risk: a case-control analysis. J. Natl Cancer Inst. 91, 151–156 (1999).
Langer, C. J. et al. Randomized, phase III trial of first-line figitumumab in combination with paclitaxel and carboplatin versus paclitaxel and carboplatin alone in patients with advanced non-small-cell lung cancer. J. Clin. Oncol. 32, 2059–2066 (2014).
Ramalingam, S. S. et al. Randomized phase II study of erlotinib in combination with placebo or R1507, a monoclonal antibody to insulin-like growth factor-1 receptor, for advanced-stage non-small-cell lung cancer. J. Clin. Oncol. 29, 4574–4580 (2011).
Eisenhauer, E. A. et al. New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1). Eur. J. Cancer 45, 228–247 (2009).
Luis-Ravelo, D. et al. A gene signature of bone metastatic colonization sensitizes for tumor-induced osteolysis and predicts survival in lung cancer. Oncogene 33, 5090–5099 (2014).
Antoń, I. et al. Receptor of activated protein C promotes metastasis and correlates with clinical outcome in lung adenocarcinoma. Am. J. Respir. Crit. Care Med. 186, 96–105 (2012).
Ajona, D. et al. Blockade of the complement C5a/C5aR1 axis impairs lung cancer bone metastasis by CXCL16-mediated effects. Am. J. Respir. Crit. Care Med. 197, 1164–1176 (2018).
Ezponda, T. et al. The oncoprotein SF2/ASF promotes non-small cell lung cancer survival by enhancing survivin expression. Clin. Cancer Res. 16, 4113–4125 (2010).
Ajona, D. et al. A combined PD-1/C5a blockade synergistically protects against lung cancer growth and metastasis. Cancer Discov. 7, 694–703 (2017).
We thank C. Zandueta and O. Rogero for technical assistance, and J. M. Kurie (The University of Texas MD Anderson Cancer Center, Houston, TX) for gifting the 393P cells. This study was supported by the Foundation for Applied Medical Research (FIMA), CIBERONC (CB16/12/00443 and CB16/12/00364), Fundación Científica de la Asociación Española Contra el Cáncer, Fundación Ramón Areces, Juan Serrano, Instituto de Salud Carlos III–Fondo de Investigación Sanitaria–Fondo Europeo de Desarrollo Regional ‘Una manera de hacer Europa’ (FEDER; PI17/00411, PI16/01352, PI16/01821 and AC14/00034), La Caixa Foundation, Caja Navarra Foundation and the Spanish Ministry of Economy and Competitiveness (SAF2015-71606R, SAF2016-78568-R and RTI 2018-094507-B-100). F.E. is funded by a predoctoral fellowship from the Asociación de Amigos de la Universidad de Navarra and from La Caixa Foundation. S.O.-E. was funded by a predoctoral fellowship from the Asociación de Amigos de la Universidad de Navarra and is now supported by an FPU fellowship. M.F.S. is supported by a Miguel Servet type I contract from Instituto de Salud Carlos III–Fondo de Investigación Sanitaria.
The authors declare no competing interests.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
a, Individual follow-up of 393P tumor volumes in mice treated with anti-PD-1 (n=10 mice), STS (n=10 mice), the combination of both (n=8 mice) or vehicle (Control; n=11 mice). b, Individual follow-up of LLC tumor volumes in mice treated with anti-PD-1 (n=8 mice), STS (n=6 mice), the combination of both (n=6 mice) or vehicle (Control; n=8 mice). Numerical source data are provided as a source data file. Source data
a, Flow cytometric analysis of the proportion of intratumoral CD4, NK (NK1.1) and B (CD19) cells at day 18 after subcutaneous inoculation of LLC cells (Control, n=8 tumors; anti-PD-1, n=8 tumors; STS, n=5 tumors; STS/anti-PD-1, n=7 tumors). b, Expression of PD-1 in tumor-infiltrating CD4 T cells of the experiment shown in section a. Data are expressed as the mean fluorescence intensity (MFI) ± s.e.m. c, Flow cytometric analysis of the proportion of intratumoral total MDSCs, monocytic MDSCs (M-MDSCs; CD11b/Ly6C+/Ly6Glow), granulocytic MDSCs (G-MDSC; CD11b/Ly6C+/Ly6G+), macrophages (F4/80) and DCs (CD11c). The number of tumors per group was: Control, n=7 tumors; anti-PD-1, n=8 tumors; STS, n=5 tumors; STS/anti-PD-1 combination, n=7 tumors. Data are expressed as the mean of the percentage of total leukocytes (CD45) ± s.e.m. In all cases, the statistical significance of the differences were evaluated using the two-sided Kruskal-Wallis test with the Mann-Whitney U-test as the post hoc test. Numerical source data are provided as a source data file. Source data
Extended Data Fig. 3 Effects of STS/anti-PD-1 combination in the proportion of immune circulating cells from LLC tumor-bearing mice.
Flow cytometric analysis of the proportion of peripheral blood CD8, CD4, Treg cells, NK (NK1.1) and B (CD19) cells in mice after inoculation of LLC cells and two cycles of 48 hours of STS (day 12 after tumor implantation) (Control, n=6 mice; anti-PD-1, n=6 mice; STS, n=5 mice; STS/anti-PD-1, n=6 mice). Data are expressed as the percentage of total lymphocytes except for Treg cells, which are expressed as the percentage of total CD4 T cells (means ± s.e.m.). The differences between experimental groups were analyzed using the two-sided Kruskal-Wallis test with the Mann-Whitney U-test as the post hoc test. Numerical source data are provided as a source data file. Source data
a, Cropping images from the western blot analysis of LC3-I and LC3-II in LLC cells cultured under normal or STS-like conditions for 48 hours. This experiment was performed twice with similar results. b, Cropping images from the western blot analysis of LC3-I and LC3-II in LLC cells cultured under normal or STS-like conditions in the presence of 1 µg ml–1 of rIGF-1 or vehicle for 48 and 72 hours. This experiment was performed once. Unprocessed images of blots are provided as source data file. Source data
Individual follow-up of 393P tumor volumes in mice treated with PQ401 (n=10 mice), anti-PD-1 (n=11 mice), or the combination of both (n=10 mice) or vehicle (Control; n=11 mice). Numerical source data are provided as a source data file. Source data
a, Flow cytometric analysis of the proportion of intratumoral total MDSCs, monocytic MDSCs (M-MDSCs; CD11b/Ly6C+/Ly6Glow), granulocytic MDSCs (G-MDSC; CD11b/Ly6C+/Ly6G+), macrophages (F4/80), DCs (CD11c), CD4 T, NK (NK1.1) and B (CD19) cells at day 19 after inoculation of LLC cells (n=8 tumors per group except for anti-PD-1 group in CD4 T, NK and B cells analyses; n=7). Data are expressed as the percentage of total leukocytes (CD45). b, Expression of the markers PD-1, GITR, and LAG-3 in tumor-infiltrating CD4 T cells. The number of tumors per group was: Control, n=8 tumors; anti-PD-1, n=7 tumors; PQ401, n=8 tumors; PQ401/anti-PD-1 combination, n=8 tumors. Data are expressed as the mean of fluorescence intensity (MFI) ± s.e.m. The differences between experimental groups were analyzed using the two-sided Kruskal-Wallis test with the Mann-Whitney U-test as the post hoc test. Numerical source data are provided as a source data file. Source data
Extended Data Fig. 7 Effects of PQ401/anti-PD-1 combination in splenic immune populations from LLC tumor-bearing mice.
Flow cytometric analysis of the proportion of splenic CD8, CD4, B cells (CD19), NK (NK1.1) and Treg cells, MDSCs (total MDSCs, M-MDSCs, and G-MDSCs), macrophages, and DCs in tumor-bearing mice at day 19 after inoculation of LLC cells (n=8 mice per group). Data are expressed as the percentage of total leukocytes (CD45), except for Treg cells, which are expressed as the percentage of total CD4 T cells (means ± s.e.m.). The differences between experimental groups were analyzed using the two-sided Kruskal-Wallis test with the Mann-Whitney U-test as the post hoc test. No statistically significant differences were found. Numerical source data are provided as a source data file. Source data
Numerical source data
Numerical source data
Numerical source data
Numerical source data
Unprocessed Western blots
Numerical source data
Numerical source data
Numerical source data
Numerical source data
Numerical source data
Numerical source data
Unprocessed Western blots
Numerical source data
Numerical source data
Numerical source data
About this article
Cite this article
Ajona, D., Ortiz-Espinosa, S., Lozano, T. et al. Short-term starvation reduces IGF-1 levels to sensitize lung tumors to PD-1 immune checkpoint blockade. Nat Cancer 1, 75–85 (2020). https://doi.org/10.1038/s43018-019-0007-9