A Case of Granulocyte-Colony-Stimulating Factor-Producing Non-Small Cell Lung Cancer under Steroid Treatment and with Poor Performance Status That Responded to Pembrolizumab

Granulocyte-colony-stimulating factor (G-CSF)-producing tumors induce leukocytosis and inflammation and have been reported in various cancers, such as lung cancer, liver cancer, and stomach cancer [1]. A large number of cases of G-CSF-producing lung cancer have been reported [2], and chemotherapy has been the standard treatment for G-CSF-producing lung cancer, but its efficacy is limited. This is because G-CSF itself and the associated inflammation promote the proliferation of cancer cells [3].

Recently, anti-programmed cell death-1 (PD-1) and anti-PD-ligand 1 (PD-L1) antibodies monotherapy have been used as one of the standard treatment drugs for advanced non-small cell lung cancer (NSCLC) with high PD-L1 expression. A recent study and case report showed the efficacy of the anti-PD-1 antibody, pembrolizumab, as a first-line therapy in 7 patients with G-CSF-producing NSCLC and high PD-L1 expression [4, 5]. Three of the 7 patients achieved a partial response (PR). However, the efficacy of anti-PD-1/PD-L1 antibodies for patients with G-CSF-producing NSCLC and high PD-L1 expression has been unclear.

Some patients with G-CSF-producing NSCLC may possess negative predictors for anti-PD-1/PD-L1 efficacy, including a poor performance status (PS) resulting from inflammation induced by G-CSF-producing NSCLC [6] and the use of steroid hormones to mitigate inflammation [7].

Here, we present the first case of G-CSF-producing NSCLC with high PD-L1 expression in which pembrolizumab showed high antitumor efficacy, regardless of negative factors for pembrolizumab efficacy, including poor PS and the administration of 3.3 mg of dexamethasone. The CARE Checklist has been completed by the authors for this case report, attached as online supplementary material (for all online suppl. material, see https://doi.org/10.1159/000540935).

A 77-year-old woman presented at our hospital with lower back pain in April 2020. Computed tomography (CT) suggested a diagnosis of lung cancer, and the patient was hospitalized for further examination. She was a current smoker with a Brinkman index of 150 and a PS of 2 on admission. Laboratory findings showed that white blood cell count, C-reactive protein (CRP) level, interleukin-6 (IL-6), plasma G-CSF, and cytokeratin 19 fragment were elevated to 20,720/μL, 3.06 mg/dL, 17.7 pg/mL, 9.35 pg/mL, and 21.5 ng/mL, respectively. CT showed a tumor in the upper lobe of the right lung, hilar and mediastinal lymph nodes enlargement, and multiple nodules in the liver and thoracic vertebrae (Th) 12, indicating metastasis from the primary tumor in the right lung (Fig. 1). Endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) of the mediastinal lymph nodes was performed, and immunostaining of these specimens was negative for thyroid transcription factor-1, Napsin A, p40, cytokeratin 5/6, CD56, synaptophysin, chromogranin A, and somatostatin receptor 2. Immunostaining with a rabbit polyclonal anti-G-CSF antibody (ab9691, Abcam, Inc.) was positive (Fig. 2b). Based on these observations, the patient was diagnosed with G-CSF-producing NSCLC-not otherwise specified, classified as clinical stage IVB, T2N3M1c (Union for International Cancer Control TNM Classification, 8th edition).

The treatment course is shown in Figure 3. From the 15th day after admission, she underwent radiotherapy (20 Gy in 5 fractions) for metastatic lesions in the Th12 for pain control. On the 16th day after admission, she had a slight fever, her PS worsened from 2 to 3, and her CRP increased to 6.47 mg/dL due to inflammation by a G-CSF-producing tumor. These findings were caused by G-CSF-induced hypercytokinemia. Thus, we initiated the administration of dexamethasone (3.3 mg/day) on the 21st day. Her fever abated the next day, and CRP dropped to 3.22 mg/dL after 4 days. Epidermal growth factor receptor gene mutations, anaplastic lymphoma kinase gene mutations, c-ros oncogene1 fusion gene mutations, and mesenchymal-epithelial transition factor gene Exon14 skipping mutation were negative. PD-L1, assessed by 22C3 pharmDx kit tumor proportion score (TPS), was highly expressed at 100% (Fig. 2c). Thus, pembrolizumab was started on the 26th day. Subsequently, the white blood cell count decreased, and the tumor shrank on CT, indicating a PR (Fig. 4). After three cycles of pembrolizumab therapy, the anorexia improved, and she was discharged on the 68th day after admission. On the 24th day after discharge, the patient was rushed to the emergency room with abdominal pain. Blood tests showed grade 3 elevations in aspartate aminotransferase (AST: 201 IU/L), alanine transaminase (ALT: 127 IU/L), and alkaline phosphatase (ALP: 2,258 IU/L), and grade 4 elevation in gamma-glutamyl transferase (γ-GTP: 1,505 IU/L). These gradings were performed according to the Common Terminology Criteria for Adverse Events (CTCAE) version 5.0. Abdominal ultrasound and CT showed dilation of the extrahepatic bile duct (Fig. 5a, b). Endoscopic ultrasound revealed diffuse wall thickening of the common bile duct (Fig. 5c). Endoscopic retrograde cholangiopancreatography showed sclerotic changes mainly in the intrahepatic bile ducts, with no obvious stenosis in the extrahepatic bile ducts after contrast imaging (Fig. 5d). Blood tests did not indicate viral infections that could cause liver dysfunction. Antinuclear antibody and immunoglobulin G4 were within the normal range. These findings are consistent with those of sclerosing cholangitis reported in previous studies. Therefore, we clinically diagnosed this patient with sclerosing cholangitis induced by anti-PD-1 antibodies [8]. Thereafter, the pembrolizumab treatment was discontinued. Sclerosing cholangitis improved with oral administration of ursodeoxycholic acid without increasing the dose of dexamethasone. After that, the patient was followed without cancer treatment, and the primary lesion was enlarged, indicating progressive disease. The patient and her family did not want additional treatment. Thus, we continued the best supportive care, and the patient died of lung cancer. Finally, her progression-free survival and overall survival were 6 and 7 months, respectively.

We herein present the first case of G-CSF-producing NSCLC with high PD-L1 expression showing a PR to pembrolizumab treatment, although the patient had negative factors for pembrolizumab efficacy, such as poor PS due to inflammation induced by the G-CSF-producing NSCLC [6] and administration of corticosteroids (>10 mg of prednisolone) [7]. In this case, pembrolizumab yielded a PR in G-CSF-producing NSCLC with high PD-L1 expression. A previous study and 1 case report have shown the status of PD-L1 expression in 14 cases of G-CSF-producing NSCLC (Table 1), and the proportion of the patients with PD-L1 TPS ≥50% was high at 71.4% [4, 5]. Of these, 7 cases with PD-L1 TPS ≥50% were treated with pembrolizumab, and three of the seven patients achieved a PR (Table 2). The randomized, open-label, phase 3 KEY-NOTE 024 trial, which compared pembrolizumab with chemotherapy for advanced NSCLC with PD-L1 TPS ≥50%, showed an objective response rate of 44.8% for pembrolizumab [9]. Thus, pembrolizumab may be effective for G-CSF-producing NSCLC with high PD-L1 expression, as well as for non-G-CSF-producing NSCLC. The number of patients with high PD-L1 expression and G-CSF-producing NSCLC is limited. Therefore, to validate the efficacy of pembrolizumab, a retrospective study involving a larger cohort should be conducted in the future.

In our case, the patient had negative factors for pembrolizumab efficacy, including poor PS due to inflammation induced by G-CSF-producing NSCLC [6] and the administration of 3.3 mg of dexamethasone, which reduced the activity of T lymphocytes [7]. Previous reports have shown that poor PS or steroid administration is associated with poor response to pembrolizumab, even in patients with high PD-L1 expression, a favorable factor for pembrolizumab efficacy [6, 7]. However, our patient showed a PR and progression-free survival of 6 months. In our case, the serum IL-6 level was high before pembrolizumab treatment, and IL-6 was reported to be expressed from the G-CSF-producing tumor [10]. Poor PS, in our case, was considered to be due to inflammation induced by IL-6. In addition, IL-6 was also reported to be a negative factor for anti-PD-1 antibody treatment [11]. Therefore, to suppress IL-6 expression, thereby enhancing PS and mitigating inflammation, dexamethasone was administered prior to pembrolizumab in our case. Dexamethasone is anticipated to bolster the antitumor efficacy of pembrolizumab by inhibiting IL-6 expression.

There is no clear conclusion regarding whether corticosteroids should be administered to patients with G-CSF-producing tumors. The fever of our patient improved, and her CRP level decreased after the administration of corticosteroids. In addition, a previous report showed that steroid administration improved malaise, anorexia, leukocytosis, and CRP levels in patients with G-CSF-producing duodenal cancer [12]. Clinicians should refrain from administering corticosteroids to patients scheduled for immunotherapy to prevent loss of immunotherapy efficacy. However, as mentioned above, corticosteroids may not only inhibit the efficacy of immunotherapy but also exert antitumor efficacy against anti-PD-1/PD-L1. It would be reasonable to administrate steroid hormones for patients with anorexia and poor PS due to inflammation before immunotherapy.

Here, we present the first case of G-CSF-producing NSCLC with high PD-L1 expression that showed sufficient efficacy of pembrolizumab treatment, although the patient had negative factors for pembrolizumab efficacy, such as poor PS and administration of corticosteroids. Corticosteroids generally reduce the efficacy of immunotherapy; however, there may be some cases of G-CSF-producing lung cancer in which corticosteroids exert anti-PD-1/PD-L1 efficacy.

The authors would like to thank the patient and her family. We would like to thank Editage (www.editage.com) for English language editing.

We obtained written informed consent from the patients for the publication of this case report and accompanying images. Ethical approval is not required for this study in accordance with local guidelines.

Takeshi Masuda reports personal fee from Daiichi-Sankyo Co., Ltd., Taiho Pharmaceutical Co., Ltd., Nippon Boehringer Ingelheim Co., Ltd., Kyowa Kirin Co., Ltd., Eli Lilly Japan K.K., Ono Pharmaceutical Co., Ltd., Otsuka Pharmaceutical Co., Ltd., Chugai Pharmaceutical Co., Ltd., and AstraZeneca K.K. outside the submitted work. Kakuhiro Yamaguchi reports honoraria from Ono Pharmaceutical Co., Ltd., outside the submitted work. Shinjiro Sakamoto reports honoraria from AstraZeneca K.K., Nippon Boehringer Ingelheim Co., Ltd., Chugai Pharmaceutical Co., Ltd., Kyowa Kirin Co., Ltd., and Novartis Pharma K.K. outside the submitted work. Yasushi Horimasu reports honoraria from Nippon Boehringer Ingelheim Co., Ltd., outside the submitted work. Taku Nakashima reports honoraria from AstraZeneca K.K. and Chugai Pharmaceutical Co., Ltd., outside the submitted work. Noboru Hattori reports honoraria from AstraZeneca K.K., Bristol Myers Squibb, Chugai Pharmaceutical Co., Ltd., and Ono Pharmaceutical Co., Ltd., outside the submitted work. No other disclosures are reported.

We did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors for this case report.

All authors had full access to the data in the study and take responsibility for the integrity of the data.

Hiroki Egusa and Takeshi Masuda: conceptualization, project administration, resources, and writing – original draft. Kakuhiro Yamaguchi, Shinjiro Sakamoto, Yasushi Horimasu, Taku Nakashima, Hiroshi Iwamoto, Kei Kushitani, and Hironobu Hamada: resources and writing – review and editing. Noboru Hattori: project administration, supervision, and writing – review and editing.

All data generated or analyzed during this study are included in this article. Further inquiries can be directed to the corresponding author.