Chemoradiation and granulocyte-colony or granulocyte macrophage-colony stimulating factors (G-CSF or GM-CSF): time to think out of the box?

Abstract

Concerns have been raised about potential toxic interactions when colony-stimulating factors (CSFs) and chemoradiation are concurrently performed. In 2006, the ASCO guidelines advised against their concomitant use. Nevertheless, with the development of modern radiotherapy techniques and supportive care, the therapeutic index of combined chemotherapy, radiotherapy, and CSFs is worth reassessing. Recent clinical trials testing chemoradiation in lung cancer let investigators free to decide the use of concomitant CSFs or not. No abnormal infield event was reported after the use of modern radiotherapy techniques and concomitant chemotherapy regimens. These elements call for further investigation to set new recommendations in favour of the association of chemoradiation and CSFs. Moreover, radiotherapy could induce anticancer systemic effects mediated by the immune system in vitro and in vivo. With combined CSFs, this effect was reinforced in preclinical and clinical trials introducing innovative radioimmunotherapy models. So far, the association of radiation with CSFs has not been combined with immunotherapy. However, it might play a major role in triggering an immune response against cancer cells, leading to abscopal effects. The present article reassesses the therapeutic index of the combination CSFs-chemoradiation through an updated review on its safety and efficacy. It also provides a special focus on radioimmunotherapy.

Introduction

With the intensification of chemotherapy doses and the increasing number of available treatment lines, it has become crucial to manage febrile neutropenia in cancer patients.^1,2 Colony-stimulating factors (CSFs, including granulocyte-colony stimulating factors (G-CSF) and granulocyte macrophage-colony stimulating factors (GM-CSF)) are growth hormones which avoid chemotherapy-induced neutropenia. They are used in frail patients undergoing chemotherapy and patients undergoing highly haematotoxic programs.^3–5 However, concerns were raised about potential toxic interactions when CSFs and chemoradiation were concurrently performed.⁶ Since 2006, the ASCO guidelines have advised against their concomitant use.⁷ Nevertheless, with the development of modern radiotherapy techniques and supportive care, the therapeutic index of combined chemoradiotherapy and CSFs is worth reassessing. Multimodal approaches involving chemoradiation became standard treatments in many cancers such as limited-stage small-cell lung cancers. Their benefit was clearly demonstrated on overall survival.^8,9 Yet, up to 51% patients experienced an acute Grade 4 haematologic toxicity. Therefore, the use of CSFs in these settings could bring significant advantages.^10,11 The safety profile of the association “CSFs plus chemo-radiotherapy” should therefore be reviewed in the light of recent data. Indeed, major evolutions occurred in radiotherapy techniques over the 30 past years. The first chemoradiation trials used 2D-radiotherapy (also known as “conventional” radiotherapy). Fields were determined with bony landmarks assessed on radiographs. Therefore, the position of moving soft tissues (lung structures, pelvic organs…) could not be assessed and large fields were used to ensure the target was not missed. Furthermore, the dose received by the organs at risk (and especially by the haematopoietic marrow) could not be precisely evaluated. At the beginning of the 21st century, the 3D-conformational radiotherapy became the gold standard: the development of the CT-scan allowed a precise definition of the anatomic structures that should be targeted (tumour) or spared (organs at risk). Progresses in ballistics have recently been made thanks to new techniques (IMRT, VMAT) allowing a higher precision and a significant decrease of radiation-induced toxicities.¹²

Recently, harnessing the power of the immune system has become a key issue for oncologists, especially with researches on the abscopal effect.^13,14 The abscopal effect is defined as the regression of non-irradiated metastases after the irradiation of another cancer location (primary or metastases). This reaction was showed to be induced by an adaptive immune process that could be stimulated by radiation or chemoradiation.¹⁵ Recent data suggested that concomitant GM-CSFs and chemoradiotherapy were potentially efficient, probably through the up-regulation of antigen-presenting cells and regulatory T-cells (Treg cells).¹⁶ Therefore GM-CSFs could be a major breakthrough in the field of radioimmunotherapy.¹⁷

The aim of this article was to reassess the therapeutic index of the combination “CSFs-chemo-radiotherapy” through an updated review on its safety, its efficacy and with a special focus on radioimmunotherapy.

Methods

Requests were performed in the Medline database (via pubmed) to identify all publications reporting the toxicity profile of CSFs with chemoradiotherapy. No restriction regarding the date of publication was used. In case of several publications for the same trial, only the most recent data was taken into account. The latest update was performed in September 2019. All reviews on the topic were also studied to ensure that major studies had not been omitted. The following keywords were used: radiotherapy, CSFs, and chemotherapy. Irrelevant papers (i.e., not reporting on efficacy/safety of the concurrent association of chemoradiation with CSFs, anecdotal case report, studies not reporting the radiation technique) were excluded by the first and the second authors. To identify studies investigating the combination of radiotherapy with CSFs, the clinicaltrials.gov database was searched with the following terms: GM-CSF, GMCSF, GCSF, G-CSF, granulocyte-colony stimulating factors, granulocyte macrophage-colony stimulating factors, radiotherapy, and radiation.

Results

G-CSF or GM-CSF: definitions and differences

Although both G-CSF and GM-CSF can stimulate the production of neutrophils, they act differently when it comes to stimulate other myeloid population. Therefore, they have different indications. The G-CSF receptor is almost exclusively expressed on neutrophils. Its activation by G-CSF leads to the proliferation of maturation of progenitors already differentiated and engaged in granulopoiesis.¹⁸ G-CSF specifically improves the survival and activity of neutrophil polynuclear cells and increases their phagocytic, antimicrobial and antibody-dependent cytotoxicity. However, GCSF has no effect on monocytes, eosinophils or basophils. Filgrastim and lenograstim are the two labelled G-CSF. They are indicated to prevent febrile neutropenia. Their most frequent side-effects are bone pain, osteopenia, dysuria, and biochemical disorders (increase of alkaline phosphatases and of dehydrogenase lactate).

GM-CSF is secreted by macrophages, T cells, mast cells, natural killer cells, endothelial cells and fibroblasts, and acts as a cytokine. Unlike G-CSF which specifically promotes neutrophil proliferation and maturation, GM-CSF additionally drives the proliferation of many myeloid cell types such as granulocytes, macrophages, eosinophils, monocytes, and dendritic cells in 3–15 days.¹⁹ It also increases the antimicrobial and antitumour functions of neutrophils, eosinophils, and macrophages (adhesion, chemotactic, and phagocytic activities). Furthermore, GM-CSF was able to regulate immunity in several preclinical studies²⁰ through the promotion of well-known signalling pathways (PI3K-Akt, ERK1/2, JAK2/STAT5, NF-kB),²¹ and the modulation of the activity of matured myeloid cells such as granulocytes, macrophage, and eosinophils. Thus, GM-CSFs are thought to act both as growth/differentiation factors as well as immune modulators.²² Sargramostim is the only labelled GM-CSF. It is indicated for myeloid reconstitution after autologous or allogeneic bone marrow transplantation and to treat neutropenia induced by chemotherapy in acute myeloid leukaemia.²³ It is also indicated in patients who have been exposed to sufficient radiation to suppress bone marrow myelogenesis.²⁴ Its main side-effects are bone pain, splenomegaly, flu-like condition, pleural, and pericardial effusion.

Finally, G-CSFs and GM-CSFs seem to have slightly different impacts on neutrophils which lead to different functional outcomes: G-CSFs promote the release of TNF receptors and of IL-1 receptor antagonist proteins whereas GM-CSFs enhance the metabolism of arachidonic acid, induce the release of B4-leucotriene and promote the production of IL-1.^22,25

The safety profile of the combination “CSFs-chemoradiation”

G-CSFs

It was initially thought that the possible pulmonary toxicity of G-CSFs—occurring when combined with chemotherapy—could be potentiated by a thoracic radiotherapy.²⁶ The association of G-CSFs with concurrent chemoradiation is still controversial. Its routine use is not recommended but current guidelines are based on old and limited data.²⁷ The safety of the association was first tested in a randomised clinical trial which used large thoracic fields of radiotherapy with or without G-CSF. The study included 13 out of the 50 intended patients and was interrupted prematurely due to a reduction of circulating progenitor cells without clinical impact.²⁸ Authors advised against the use of concomitant G-CSFs with radiation. However, peripheral leukocyte counts were similar in both groups. Moreover, the reduced number of progenitors could be explained by the action of CSFs based on the accelerated transformation of progenitors in differentiated leukocytes.²⁹ Recent retrospective and prospective Chinese trials tested the concurrent association of G-CSF with chemoradiation in lung or esophageal tumours.^30,31 Radiation-related acute and late toxicities were poorly reported, making it hard to conclude to the clear safety of the combination. A Phase II randomised clinical trial compared radiotherapy performed once a day vs twice a day, associated with concurrent chemotherapy in small-cell lung carcinoma. A total of 38 patients were included and received four cycles of chemotherapy (cisplatin: 60 mg/m2 and VP-16: 120 mg/m2). Treatment with 3D-conformationnal radiotherapy was randomised in two arms: 66 Gy in 33 daily fractions of 2 Gy performed over 45 days or 45 Gy in 30 fractions of 1.5 Gy twice a day performed over 19 days. The use of G-CSF was freely decided by the investigators. There was an increased risk of Grade 3–4 thrombocytopenia in patients using G-CSF (54% vs 11%, p was not reported). However, no Grade 3–4 acute or chronic lung toxicity was reported in patients receiving G-CSF.³² A Phase III randomised clinical trial named CONVERT was therefore designed to corroborate the results of the previous clinical trial. Similar concurrent chemoradiation programs were tested: 66 Gy in 33 daily fractions over 45 days versus 45 Gy in 30 fractions twice-a-day over 19 days, each with 4–6 courses of Cisplatin (75 mg/m2) and VP16 (100 mg/m2). A total of 547 patients were enrolled in both arms.³³ The use of G-CSF was freely decided by the investigators. Subgroup analysis confirmed that Grade 3–4 thrombocytopenia was higher in patients treated with G-CSF (29.4% vs 13%; p  <  0.001). However, the use of G-CSF with modern radiotherapy techniques during chemoradiation did not result in an increased risk of severe acute infield toxicity (oesophagitis or pneumonitis). Moreover, it had no impact on the overall survival and progression-free survival.³⁴ Authors concluded that the use of G-CSFs with modern radiotherapy techniques could now be considered safe with concurrent chemotherapy in lung cancer treatment.³⁵

These results might probably be extrapolated to abdominal and pelvic irradiation. Bone marrow is sensitive to low doses of irradiation, especially with haematopoietic stem cells.^36,37 In patients with cervical cancer, a 45 Gy pelvic and para-aortic irradiation with a 54–59.4 Gy boost on invaded lymph nodes with concurrent chemotherapy led to ≥Grade 3 leukopenia in 81% of patients.³⁸ The bone marrow volume receiving at least 40 Gy was correlated with ≥Grade 2 acute haematologic toxicity in gynecologic,³⁹ rectal,⁴⁰ and anal cancers.⁴¹ In this context, the use of CSFs could prevent discontinuing concurrent chemoradiation, which was demonstrated to be associated with poorer local and overall outcomes.^42–44 The literature is very limited on this particular topic. To our best knowledge, only one study reported the safety of the association of abdomino-pelvic radiotherapy (45 Gy on the pelvis and 22 Gy for the abdomen) with G-CSFs. 14 patients received Filgrastim in association with conventional chemoradiation for ovarian carcinoma. No abnormal infield toxicities were reported. Authors stated that the combination of chemoradiation with G-CSFs was safe in this setting.⁴⁵

GM-CSF

A Phase III randomised clinical trial tested chemoradiation in limited-stage small-cell lung cancer with or without GM-CSF.⁶ A total of 230 curative intent patients underwent a conventional 2D radiotherapy (45 Gy in 25 fractions) associated with chemotherapy. Large thoracic and supraclavicular volumes were treated, including the primary tumour and bilateral hilar, mediastinal, and supraclavicular nodes with a 1–2 cm margin. Chemotherapy consisted of six cycles of high dose VP-16 (240 mg/m² per cycle) and cisplatin (120 mg/m² per cycle), concomitantly performed with radiotherapy. Grade 3 leukopenia and Grade 4 neutropenia were significantly reduced in patients undergoing GM-CSF (40% vs 65%, p < 0.001 and 18% vs 24%, p = 0.01, respectively). The GM-CSF arm experienced more Grade 4 thrombocytopenia (54% vs 12%, p < 0.001) and more Grade 3 anaemia (33% vs 19%, p = 0.03). Median overall survival was lower in the GM-CSF arm (14 months vs 17 months) but the difference was not statistically significant (p = 0.12). Nevertheless, GM-CSF patients faced significantly more toxic deaths (9 vs 1, p < 0.01), more non-haematologic toxicities (lung infection, dyspnoea), more days in hospital (57% vs 37%, p < 0.01), a higher incidence of intravenous antibiotic usage (3.3 days vs 2.6 days, p < 0.01) and more transfusion. Toxic deaths in GM-CSF patients were related to five different pulmonary complications (pneumonitis, pulmonary oedema, embolus, aspiration pneumonia, haemoptysis, in one patient each), three infections and one cerebrovascular accident. Authors advised against the association of chemoradiation with GM-CSFs. Yet, chemotherapy doses largely exceeded the currently recommended ones.⁴⁶ This could have increased the haematologic and infectious complications. Furthermore, large fields of radiotherapy performed with a conventional 2D technique probably significantly contributed to the pulmonary toxicity.^47,48 Based on this study, the 2006 ASCO guidelines maintained a previous recommendation of not using CSFs with concurrent chemoradiotherapy, with a special warning when radiotherapy encompassed the mediastinum.⁷ However, this trial used what can now be considered as an outdated radiotherapy and chemotherapy. Therefore, banning the use of GM-CSF with concurrent chemoradiation based on the results of this trial now seems questionable. The question is still pending as no study based on modern concurrent chemoradiation has ever been performed with concurrent GM-CSF. In this context, we can wonder whether toxicity outcome applicable to G-CSF can apply to GM-CSF. Two retrospective studies reported on the safety and effectiveness of GM-CSF (sargramostim) vs G-CSF (filgrastim, perfilgrastrim) in cancer patients receiving chemotherapy.^49,50 There was no difference between sargramostim and filgrastim regarding fever related to infection or neutropenic complications requiring hospitalisation. However, the risk of hospitalisation seemed to be lower when perfilgrastim was used (vs filgrastim of sargramostim, p < 0.01)).⁵⁰ Regarding clinical toxicity, sargramostim seemed to significantly cause more non-infectious fever, fatigue, diarrhoea, dermatological disorders and oedema (all p < 0.05) than filgrastim. To conclude, G-CSF and GM-CSF seem to cause slightly different toxicities with similar efficacies. The literature does not allow us to predetermine the possible differences between G-CSF and GM-CSF when associated with radiotherapy. Based on current literature, there is no data supporting the safe use of concurrent GM-CSF with radiation. The literature reporting on severe toxicity of the association was however based on outdated chemoradiotherapy.

Take home message

As a conclusion, the combination of G-CSFs and concurrent chemoradiotherapy seems safe. No abnormal infield event was reported with modern radiotherapy techniques and chemotherapy regimens. These elements call for further investigation to set new recommendations in favour of the association of concurrent association of chemoradiation with G-CSFs. Studies assessing the toxicity profile of modern chemoradiation with GM-CSF are eagerly awaited.

Radio-immunotherapy and CSFs

Radiotherapy could enhance an antitumour specific immune response in numerous preclinical data based on animal and human cancer cells.⁵¹ The mechanisms involved were the augmentation of T regulatory cells,⁵² the stimulation of cytotoxic T lymphocytes,⁵³ the up-regulation of the class I major histocompatibility complex (MHC-I), of pro-immunogenic cytokines⁵⁴ and the release of tumour antigens.⁵⁵ This immune reaction was hypothesised to be a cornerstone to trigger the abscopal effect mediated by radiation. Further researches suggested that the mechanisms of the abscopal effect were closely linked to the effects of radiotherapy on the tumour micro-environment and the patient’s immune system. Radiation induced the death of tumour cells, resulting in the release of danger-associated molecular patterns and tumour antigens. Radiation also permeated endothelial cells. This stimulated the circulation and the activation of antigen presenting cells. An anticancer immune response based on both direct cytotoxic action of Natural Killer cells and on the specific clonogenic expansion of CD8 +T cells could then be created.⁵⁶

G-CSFs

In a preclinical murine model, authors identified a potential interest in the combination of radiation with G-CSF. First, they demonstrated that irradiated tumours attracted antitumour neutrophils. Interestingly, the neutrophils activated by radiation featured specific properties, such as the high ability to produce reactive oxygen species (ROS), which are key elements of radio-sensitisation. Authors then observed that endogenous G-CSF concentrations (i.e., produced by tumour micro-environment) increased after irradiation and hypothesised that it participated in the improved local control. Finally, they proved that the addition of exogenous G-CSF in combination with radiation had a superior antitumour effect to exclusive radiotherapy. It was provoked partly through the increase of local ROS production by activated neutrophils and partly through the activation of cytotoxic T cells.¹⁴ The antitumour immune reaction was thus up-regulated by the combination of G-CSF and radiation.

GM-CSFs

GM-CSF could play a key immune role both as a pro-inflammatory and as an immunomodulating cytokine.²⁰ However, its role in cancer is controversial. On the one hand, it was shown that prostate cancer cells secreted GM-CSF and expressed its receptors in order to proliferate.⁵⁷ On the other hand, immunotherapy based on engineered T lymphocytes producing GM-CSF (Sipuleucel-T) was the first anticancer vaccine to increase survival in prostate cancer patients.^58,59 The expression of GM-CSFs might therefore have a dual effect since it was reported both to suppress⁶⁰ and stimulate immunity against cancer in some cases.²⁰ The cytokine environment is thought to be a pivotal element that can mitigate the GM-CSF-induced reaction. It could therefore be impacted by radiotherapy. Recently, two case reports have suggested that combining GM-CSF with radiotherapy or chemoradiation could lead to abscopal responses. In the first, a patient with progressive metastatic pancreatic cancer treated with gemcitabine-paclitaxel underwent palliative radiotherapy (45 Gy/25fr) on the primary.⁶¹ GM-CSF was concurrently performed daily, from the beginning of the second week of radiation to the end. The patient experienced local and distant response one and three months after radiotherapy. In the second case, a patient with a metastatic non-small-cell lung cancer underwent local radiotherapy (47.6 Gy/28fr) associated with oncothermia, followed by GM-CSF. A major response in multiple lymph nodes that were distant from the irradiated site was evidenced after treatment.⁶² This interesting data led to clinical trials. The combination of GM-CSF and radiotherapy was first tested in clinical trials based on GVAX vaccines. Radiotherapy was used to inactivate cancer cells. Neutralised cancer cells were then engineered to produce GM-CSF and were finally injected in patients. The aim was to initiate a specific antitumour immune reaction, locally enhanced by the production of GM-CSF.⁶³ Promising results were reported in prostate⁶⁴ and pancreatic cancer patients.⁶⁵ While the aim of irradiating cancer cells was initially to neutralise them, it is important to note that irradiated cancer cells were finally hypothesised to spark immunogenic reactions due to a mechanism of radiation-induced exposure of antigens.⁵⁵ Second, a “proof of the concept” clinical trial was conducted by Golden et al to explore the abscopal effect.⁶⁶ Cancer patients with a least three measurable metastases underwent radiotherapy on two metastases with two sequential radiation courses (35 Gy/10 fractions). Concurrent chemotherapy was allowed but with smaller doses to avoid radiation-induced toxicity. Concurrent GM-CSF was performed from day 8 to day 22 of each radiation course with daily subcutaneous injection of 125 µg/m². The effect on non-irradiated metastases was assessed in order to prove the existence of an immune response triggered by the combo radiotherapy and GM-CSF. Out of the 41 enrolled patients, 11 experienced an abscopal effect and had increased overall survivals (21 months vs 8 months). Regarding radiation-induced toxicities, six patients experienced Grade 3 fatigue and one had a Grade 3 dermatitis. Regarding GM-CSF-induced complications, Grade 3 toxicities corresponded to fatigue (n = 2), syncope (n = 1), bronchospasm (n = 1), and nausea-vomiting (n = 1). One patient experienced a Grade 4 pulmonary embolism. No Grade 5 toxicity was reported. The potential links between abscopal effect and GM-CSFs are summarised in Figure 1. Currently recruiting studies investigating the role of the combination of radiotherapy with CSFs as an immunotherapy are listed in Table 1. The results are yet to be published. However, radiotherapy characteristics and primary outcome measures are heterogeneous. Only two of these studies have a control arm. All studies use GM-CSF.

Figure 1.

Biology of the interaction between radiation and CSFs, possibly triggering abscopal effect.

Table 1.

Currently recruiting clinical trials testing radiotherapy in combination with colony-stimulating factors

Trial number	Primary tumour	Primary objective	Radiotherapy	Colony-stimulating factor	Investigator	Type of study
NCT02946138	Hepatocellular carcinoma	2-year progression-free survival	Carbon-ion, 40 Gy RBE in five fractions	GM-CSF 125 µg/m2 per day, subcutaneous injection from day 1 to day 28	Shanghai Proton and Heavy Ion Center	Phase II, single arm
NCT03392545	Glioblastoma multiforme	Incidence of Treatment-related Adverse Events	Not Available	Recombinant GM-CSF, 5 days after radiotherapy	Beijing Tiantan Hospital	Phase I, single arm
NCT02663440	Glioblastoma multiforme	Progression-free survival	Hypofractionated IMRT, dose was not reported	GM-CSF plus temozolomide, doses were not reported	Zhejiang Cancer Hospital	Phase II, single arm
NCT03113851	Non-small cell lung cancer	Abscopal effect rate	Sequential irradiation of 2 distinct metastases, 35 Gy in 10 fractions	Recombinant GM-CSF 125 mg/m2 per day from day 1 to day 14, every three weeks, concurrently with radiotherapy	Beijing Cancer Hospital	Phase II, single arm
NCT02976740	Non-small cell lung cancer	Abscopal effect rate	Sequential irradiation of 2 distinct metastases, 50 Gy in 4–10 SBRT fractions	Recombinant GM-CSF 125 µg/m² per day from day 1 to day 14 concurrent with radiotherapy	The First Affiliated Hospital of Xiamen University	Phase II, single arm
NCT02940990	Non-small cell lung cancer	2-year progression-free survival	SBRT to primary or metastatic tumour	GM-CSF 125 µg/m2 for 14 days concurrent with radiotherapy	Shanghai Chest Hospital	Phase II, controlled
NCT03489616	Oligo metastatic non-small cell lung cancer	2-year progression-free survival	Sequential irradiation of 2 metastases, dose per fraction >4 Gy from day 2 to day 15 in a cycle of 21 days	Recombinant GM-CSF. 200 µg/m² per day 24 h after pemetrexed infusion	Shandong Cancer Hospital and Institute	Phase II, randomised controlled
NCT02383212	Advanced tumour (any primary)	Dose limiting toxicity	Hypofractionated but exact program was not available	GM-CSF, dose not available with Radiotherapy plus Anti-PDL1	Regeneron Pharmaceuticals	Phase I, open label

Combination of CSFs with immunotherapies

Immune response induced by concurrent radiation and CSFs may have interesting anticancer effects. Yet, cancer cells mechanisms leading to immune evasion are much more complex than the ones up-regulated by CSFs. While CSFs could contribute to the initiation of an adaptive immune response, other cancer mechanisms block the downstream immunity.⁶⁷ Immune checkpoints frequently inhibited by cancer cells were recently identified, especially with CTLA4 and PD1-PDL1. Their activation ruins the development of anticancer immune response. Therefore, overcoming such obstacles might request drug combination. The use of CTLA-4 antibodies⁶⁸ and PD-1 or PDL-1 antibodies⁶⁹ permits to stop cancer immune evasion. These treatments have led to impressive results in metastatic non-squamous cells lung cancer,⁷⁰ melanoma,^71,72 head and neck cancer,⁷³ bladder cancer⁷⁴ and mismatch-repair deficient tumours regardless of primitive cancer.⁷⁵ The proof of the concept for combining GM-CSF with checkpoint inhibitor has recently been published. In a Phase II trial, the combination of ipilimumab with GM-CSF significantly increased survival in stage III/IV melanoma patients and decreased ipilimumab toxicity.⁷⁶ So far, the association of CSFs with radiotherapy and immune checkpoints inhibitors has never been tested. No doubt future clinical trials will be devoted to assess the action of this triplet.

Conclusion

The ASCO still advises against the use of concurrent chemoradiation +CSFs. While the combination of CSFs and radiotherapy was blamed for producing toxicity in the past—especially with concurrent chemotherapy—its therapeutic index should be reassessed with modern radiotherapy techniques and chemotherapy regimens. The few modern data available about G-CSFs tends to prove otherwise. Further investigations are required in order to confirm the safety of the association, especially with GM-CSFs. Once the safety concerns are out of the way, the immune benefit of concurrent CSFs and radiotherapy could be fully explored. Indeed, pre-clinical suggested that the combination of G-CSF with radiotherapy could enhance the immune antitumour response mediated by neutrophils. Clinical data suggested that GM-CSF combined with radiotherapy could trigger abscopal effect.