Metabolic Tumor Volume as Significant Predictor for Chemotherapy Containing PD-L1 Blocker in Extensive Stage Small Cell Lung Cancer

Background/Aim: Chemo-immunotherapy, including the programmed death ligand 1 (PD-L1) antibody, is an effective treatment for patients with extensive-stage small-cell lung cancer (ES-SCLC). However, no biomarker has been established for the prediction of chemo-immunotherapy. Therefore, we investigated the potential of 18F-fluorodeoxyglucose (FDG)-positron emission tomography (PET) as a predictive marker. Patients and Methods: Forty-six patients with ES-SCLC who received 18F-FDG-PET immediately before combined platinum-based chemotherapy with PD-L1 blockade as a first-line treatment were eligible, and the maximum standard uptake value (SUVmax), metabolic tumor volume (MTV), and total lesion glycolysis (TLG) on 18F-FDG uptake were evaluated. Results: PD-L1 and tumor infiltrative lymphocytes (TILs) were immunohistochemically analyzed in 36 of the 46 patients. A high MTV was significantly associated with poor performance status and low albumin levels, and there was a significant association between low albumin and high TLG. Univariate analysis identified sex, Brinkman index, and MTV as significant predictors of progression-free survival (PFS), and sex, SUVmax, MTV, and TLG as significant factors of overall survival (OS). Multivariate analysis revealed that sex, Brinkman index, and MTV were independent prognostic factors for PFS, and sex, SUVmax, MTV, and TLG were significant predictors of OS. SUVmax was significantly higher in patients with positive PD-L1 expression than in those with negative expression but was not significantly different between positive and negative TILs. Moreover, the levels of MTV and TLG were not closely associated with the levels of PD-L1 and TILs. Conclusion: MTV or TLG metabolic tumor activity is suitable for the prediction of chemo-immunotherapy outcomes in patients with ES-SCLC.

Abstract.Background/Aim: Chemo-immunotherapy, including the programmed death ligand 1 (PD-L1) antibody, is an effective treatment for patients with extensive-stage small-cell lung cancer (ES-SCLC).However, no biomarker has been established for the prediction of chemo-immunotherapy.Therefore, we investigated the potential of 18 F-fluorodeoxyglucose (FDG)-positron emission tomography (PET) as a predictive marker.Patients and Methods: Forty-six patients with ES-SCLC who received 18 F-FDG-PET immediately before combined platinum-based chemotherapy with PD-L1 blockade as a first-line treatment were eligible, and the maximum standard uptake value (SUV max ), metabolic tumor volume (MTV), and total lesion glycolysis (TLG) on 18 F-FDG uptake were evaluated.Results: PD-L1 and tumor infiltrative lymphocytes (TILs) were immunohistochemically analyzed in 36 of the 46 patients.A high MTV was significantly associated with poor performance status and low albumin levels, and there was a significant association between low albumin and high TLG.Univariate analysis identified sex, Brinkman index, and MTV as significant predictors of progression-free survival (PFS), and sex, SUV max , MTV, and TLG as significant factors of overall survival (OS).Multivariate analysis revealed that sex, Brinkman index, and MTV were independent prognostic factors for PFS, and sex, SUV max , MTV, and TLG were significant predictors of OS.SUV max was significantly higher in patients with positive PD-L1 expression than in those with negative expression but was not significantly different between positive and negative TILs.Moreover, the levels of MTV and TLG were not closely associated with the levels of PD-L1 and TILs.Conclusion: MTV or TLG metabolic tumor activity is suitable for the prediction of chemo-immunotherapy outcomes in patients with ES-SCLC.
Systemic chemotherapy for extensive-stage small-cell lung cancer (ES-SCLC) has not been developed or improved over the last 20 years.Platinum-based regimens with etoposide are identified as standard care for this dismal neoplasm; however, most patients with ES-SCLC experience recurrence with a shorter survival time after systemic chemotherapy.Recently, immune checkpoint inhibitors (ICIs), such as programmed death-1 (PD-1) or programmed death ligand-1 (PD-L1) antibodies, have been approved for patients with ES-SCLC, and some patients could obtain a survival benefit by additional treatment with PD-L1 inhibitors (1,2).As a predictive biomarker of PD-1 blockade, the expression level of PD-L1 within tumor specimens is a well-known indicator for the treatment of patients with advanced non-small-cell lung cancer (NSCLC) (3,4).However, the expression of PD-L1 in patients with ES-SCLC is low (approximately 5%; PD-L1 ≥1%), which is not predictive of the efficacy of chemoimmunotherapy (5).Unfortunately, there are no established biomarkers for predicting the efficacy and outcome of patients with ES-SCLC who receive platinum-based chemotherapy with PD-L inhibitors (6).Further investigation is warranted to identify a significant biomarker for predicting the efficacy of therapy for ES-SCLC.
Nowadays, 2-deoxy-2-[fluorine-18]-fluoro-D-glucose ( 18 F-FDG) positron emission tomography (PET) imaging is widely used as a convenient modality to distinguish benign from malignant lesions.Recently, several researchers have reported that metabolic tumor activity assessed based on 18 F-FDG uptake can predict the outcome of ICI treatment in patients with advanced NSCLC (6)(7)(8)(9)(10).As indicators of glucose metabolism within tumor cells, maximum standardized uptake values (SUV max ), metabolic tumor volume (MTV), and total lesion glycolysis (TLG) on 18 F-FDG accumulation have been investigated to determine the potential of a significant predictor; however, MTV or TLG may be appropriate for therapeutic monitoring or outcome after immunotherapy administration based on previous clinical evidence (6)(7)(8)(9)(10)(11)(12)(13)(14).Several reports have described the relationship between SUV max by 18 F-FDG uptake and the tumor immune microenvironment including PD-L1 or tumor infiltrating lymphocytes (TILs) in patients with lung cancer (7,8,15).PD-L1 expression within tumor cells is closely correlated with SUV max by 18 F-FDG uptake in patients with surgically resected NSCLC (7,8).However, a close correlation between PD-L1 expression and 18 F-FDG uptake has not been observed in patients with SCLC, unlike in NSCLC (15).The expression level of PD-L1 in patients with SCLC is reported to be exclusive to the number of TILs (15).Thus, the clinicopathological significance of the tumor immune microenvironment may differ between SCLC and NSCLC patients.
Based on this background, we examined the prognostic significance of 18 F-FDG uptake on PET as a significant predictor for patients with ES-SCLC who received combined platinum-based regimen with PD-L1 antibody.

Patients and Methods
Patients.We evaluated 73 patients with ES-SCLC treated with platinum-based chemotherapy plus PD-L1 antibody (atezolizumab or durvalumab) at our institution between July 2019 and March 2022.The inclusion criterion was to undergo 18 F-FDG-PET before the initiation of combined platinum-based etoposide with a PD-L antibody.Of the 73 patients, 23 could not undergo 18 F-FDG-PET prior to initial chemo-immunotherapy, and one was excluded due to insufficient data.As a result, 46 patients were eligible for the study.Clinical data were extracted from the medical records.This study was approved by the institutional ethics committee of the International Medical Center of Saitama Medical University.The requirement for written informed consent was waived by the ethics committee owing to the retrospective nature of the study.
Treatment and evaluation.All the patients received combined platinum (carboplatin or cisplatin)-based regimens with anti-PD-L1 antibodies (atezolizumab or durvalumab).Impower 133 (atezolizumab 1,200 mg, area under the concentration-time curve of 5 mg/ml per min carboplatin, and etoposide 100 mg/m 2 ) and Caspian (durvalumab 1,500 mg, area under the concentration-time curve of 5 mg/ml per min carboplatin or cisplatin 80 mg/m 2 , and etoposide 100 mg/m 2 ) regimens were administered intravenously (1,2).Complete blood count, biochemical testing, physical examination, and side effects were evaluated based on the judgment of the chief physician.Any toxicity was graded based on the Common Terminology Criteria for Adverse Events version 5.0.Tumor response was examined according to the Response Evaluation Criteria in Solid Tumors version 1.1 (16).
PET imaging and data analysis.The procedure of PET imaging and data analysis were performed as previously described (11,14,17).The patients fasted for at least 6 h before the examination.They were administered 18 F-FDG at the beginning of the examination, and PET-CT was performed 60 min later.Eight bed positions were determined, based on the imaging range.The obtained PET information was reconstructed into 168×168 images with a slice width of 2 mm using an ordered subset expectation-maximization algorithm based on a point spread function after attenuation correction.The SUV was obtained semi-quantitatively based on the injected dose of 18 F-FDG, patient weight, and cross-calibration factor between PET and the dose calibrator.The SUV was defined as the radiation concentration in the volume of interest (VOI) ([MBq/g]/injected dose [MBq]/patient weight [g]).PET was analyzed using the RAVAT software on a Windows workstation (Nippon Medi-Physics Co., Ltd., Tokyo, Japan).SUV max , MTV, and TLG (=SUV mean × MTV) were calculated semi-automatically for each tumor using the SUV mean value obtained from the liver VOI plus 1.5 SD as the threshold value.Active sites other than tumors, such as the myocardium, gastrointestinal tract, kidney, and urinary tract, were manually excluded from the direction of a certified nuclear medicine physician.
Immunofluorescence signals were visualized using the OPAL™ 7-color IHC kit (Akoya Biosciences, MA, USA) TSA dyes 520, 540, 570, 620, 650, and 690, counterstained with spectral DAPI.All 60 slides (Akoya Biosciences) were imaged on the Mantra Automated Quantitative Pathology Imaging System.Color separation, tissue and cell segmentation, and cell phenotyping were performed using inForm ® Software v2.5.1 (Akoya Βioscienceσ) to extract image data.The slides were evaluated for the presence of TILs in the tumors.Statistical analysis.Statistical significance was set at p<0.05.Fisher's exact test was used to examine the association between two categorical variables.Correlations between SUV max , MTV, and TLG and 18 F-FDG uptake were analyzed using Pearson's correlation coefficient.Progression-free survival (PFS) was defined as the time from initial treatment to disease progression or death.Overall survival (OS) was defined as the time from initial treatment to death from any cause.The Kaplan-Meier method was used to estimate the survival rates.Univariate analysis was performed using the log-rank test to compare the differences in survival rates.Multivariate analysis was performed using Cox hazard ratios to calculate the hazard ratios.The optimal cutoff values of SUV max , MTV, and TLG on 18 F-FDG uptake were determined using receiver operating characteristic (ROC) curve analyses, and the sensitivity and specificity were calculated to determine the optimal cutoff value for differentiating responders from non-responders using ROC curves.Responders were defined as those with PFS of >5 months based on the median PFS in previous clinical trials (Impower133 and CASPIAN) (1,2).Survival time was a continuous variable in the ROC analysis.PD-L1 and TILs (CD4, CD8, Foxp3, and PD-1) were defined as negative expression if there was no staining (positive >0 and negative=0).All statistical analyses were performed using GraphPad Prism (v.7.0e;GraphPad Software, San Diego, CA, USA) and JMP Pro 6.0 (SAS Institute Inc., Cary, NC, USA).

Clinical features based on
Effects and survival analysis according to 18 F-FDG-PET.Among 46 patients, complete response (CR), partial response (PR), stable disease (SD), and progressive disease (PD) were observed in 0, 38, two, and five patients, respectively.One patient was not evaluated.The objective response and disease control rates were 84.4% [95% confidence interval (CI)=70.5%-93.5%]and 88.9% (95%CI=75.9%-96.3%),respectively.The median SUV max , MTV, and TLG values were not significantly different between patients with PR and SD/PD.After induction chemotherapy with platinum regimens plus PD-L1 antibody, a median of two cycles were administered as PD-L1 antibody maintenance, ranging from 0 to 24.The median SUV max , MTV, and TLG values were not significantly different between patients receiving <3 and ≥3 maintenance treatments with PD-L1 antibody monotherapy.Median PFS and OS were 178 and 344 days, respectively.Thirty-four patients experienced tumor recurrence, and 28 died due to PD. Univariate and multivariate survival analyses were performed for all patients (Table II).Univariate analysis identified sex, Brinkman index, and MTV as significant predictors of PFS, and sex, SUV max , MTV, and TLG as significant factors of OS (Table II).For the multivariate analysis, fundamental factors, such as age, sex, Brinkman index, PS, and 18 F-FDG uptake were chosen for further examination.Since MTV or TLG were calculated based on the SUV, individual 18 F-FDG uptake markers were assessed in a multivariate analysis, separate from each 18 F-FDG uptake marker.Multivariate analysis revealed that sex, Brinkman index, and MTV were independent prognostic factors for PFS, and sex, SUV max , MTV, and TLG were identified as significant predictors of OS (Table II).Kaplan-Meier survival curves for PFS and OS according to the uptake of 18 F-FDG (Figure 2).

Discussion
Tumor metabolic activity based on 18 F-FDG accumulation was identified as a significant predictor of outcomes after platinum-based chemotherapy with PD-L1 antibody in patients with ES-SCLC.In our study, MTV was closely associated with host nutrition and inflammatory changes, suggesting a strong relationship between metabolic activity assessed by 18 F-FDG uptake and tumor invasiveness.Three different markers, SUV max , MTV, and TLG, were examined to explore predictors related to the efficacy of chemoimmunotherapy, and MTV and TLG were useful as independent prognostic factors for patients with ES-SCLC.Although the relationship between 18 F-FDG uptake and the tumor immune environment was examined using immunohistochemistry, we could not find something new aside from the close correlation between SUV max and PD-L1 expression.Further patient accumulation is needed to elucidate the relationship between metabolic tumor activity and the therapeutic significance of ICIs in patients with ES-SCLC because our study is a preliminary investigation.
Recently, a meta-analysis of 19 studies described the prognostic significance of 18 F-FDG accumulation in patients with SCLC (17).The pooled analyses demonstrated that high baseline MTV was identified as a prognostic marker for OS and PFS, suggesting that pretreatment MTV yielded a prognostic factor superior to those of other PET parameters (17).Previous reports have also reported that MTV or TLG are closely associated with an objective response to first-line chemotherapy in ES-SCLC, or the sum of SUV max of malignant lesions (SUVsum) aside from MTV potentially reflects the entire tumor metabolic burden of the whole body (20,21).Based on previous investigations, the metabolic tumor burden according to 18 F-FDG parameters exhibited a potential prognostic value for chemotherapy in SCLC and NSCLC.In our study, MTV or TLG was effective in predicting outcomes after first-line chemo-immunotherapy, but not an objective response or disease control.Since chemoimmunotherapy includes both platinum-based chemotherapy and PD-L1 antibodies, it remains unclear whether 18 F-FDG parameters could accurately predict the efficacy of ICIs in patients with ES-SCLC.Currently, there are no established Figure 1.Representative imaging on 18 F-FDG PET in patients with extensive-stage small cell lung cancer (ES-SCLC).An 84-year old male with ES-SCLC presents multiple bone metastases arising from a primary site of the right lower lobe (S6), and SUV max on 18F-FDG uptake in primary site, MTV and TLG were 12,504 cm 3 /ml, and 1,721 cm 3 /ml, respectively (A).A 74-year old male with ES-SCLC shows primary tumor of the right lower lobe and pulmonary metastases in left lung, and SUV max , MTV and TLG were 8.0, 33 cm 3 /ml, and 118 cm 3 /ml, respectively (B).
parameters for predicting the therapeutic efficacy of chemoimmunotherapy in ES-SCLC; therefore, pretreatment MTV or TLG may be promising predictive biomarkers for patients with ES-SCLC receiving chemo-immunotherapy.
Only one report has focused on the relationship between 18 F-FDG uptake and the tumor immune microenvironment, such as PD-L1 and TILs (15).Generally, PD-L1 within tumor cells is weakly expressed in patients with SCLC (22); therefore, PD-L1 expression may not closely correlate with the uptake of 18 F-FDG.However, our preliminary investigation indicated that 18 F-FDG uptake significantly correlated with PD-L1 expression but not with TILs.Regarding the comparison of survival benefits, tumor metabolic activity determined as MTV or TLG seemed to be superior to PD-L1 and TILs.Thus, it remains unclear whether the tumor immune environment can be a predictive marker of ICIs for ES-SCLC.
Although an exclusive relationship between 18 F-FDG accumulation and the number of TILs in ES-SCLC has been reported (15), the uptake of 18 F-FDG is associated with the presence of glucose metabolism, tumor hypoxia, and angiogenesis (23).Currently, whether tumor glucose metabolism can modulate the number of TILs in ES-SCLC is unknown; thus, it is difficult to explain a possible mechanism to predict the outcome of PD-1 blockade by 18 F-FDG uptake

. Kaplan-Meier curves based on various maximum of standardized uptake values (SUV max ) (A, B), metabolic tumor volume (MVT) (C, D), and total lesion glycolysis (TLG) (E, F) for progression-free survival (PFS) and overall survival (OS) in all patients. No statistically significant difference in the PFS was observed between patients with high and low SUV max (A) and TLG (E), but a significant difference in the OS between high and low SUV max (B) and TLG (F). Patients with high MTV depicted a significantly worse PFS and OS than those with low MTV (C, D).
Figure 3. Continued in patients with ES-SCLC.However, the usefulness of 18 F-FDG PET has been described to be able to predict outcome after any treatment in different human neoplasms (24)(25)(26).
Further investigation is warranted to elucidate the therapeutic significance of 18 F-FDG accumulation in the immune environment of SCLC.
Study limitations.First, our sample size was limited, which might have introduced biases into our results.Although our approach followed a preliminary design, we identified the prognostic significance of MTV as a predictor of ICI therapy outcomes in patients with ES-SCLC.Second, the tumor immune microenvironment was immunohistochemically evaluated using available tumor specimens.However, adequate analysis for immunohistochemistry may be difficult due to the small sample size.If possible, surgical specimens are identified as appropriate materials, but only biopsy samples are clinically available for definite diagnosis.Finally, whether 18 F-FDG uptake within tumor cells can predict the outcome of PD-L1 antibody monotherapy in ES-SCLC remains unclear.The metabolic tumor activity identified via PET may make it difficult to distinguish the efficacy of chemotherapy from that of immunotherapy.

Conclusion
Metabolic tumor activity determined by MTV or TLG is suitable for the prediction of chemo-immunotherapy outcomes in patients with ES-SCLC.Metabolic tumor  volume can reflect the entire tumor volume related to an inflammatory and aggressive environment.

Figure 2
Figure 2. Kaplan-Meier curves based on various maximum of standardized uptake values (SUV max ) (A, B), metabolic tumor volume (MVT) (C, D), and total lesion glycolysis (TLG) (E, F) for progression-free survival (PFS) and overall survival (OS) in all patients.No statistically significant difference in the PFS was observed between patients with high and low SUV max (A) and TLG (E), but a significant difference in the OS between high and low SUV max (B) and TLG (F).Patients with high MTV depicted a significantly worse PFS and OS than those with low MTV (C, D).

Figure 3 .
Figure 3.Comparison of SUV max , MTV and TLG on 18F-FDG uptake according to numbers of CD4, CD8, Foxp3, PD-1 and PD-L1 expression (A).No statistically significant difference in the SUV max , MTV and TLG based on these markers was observed.Immunofluorescence using multiplex immunohistochemistry staining (B).CD4 (blue), CD8 (red), Foxp3 (pink), PD-1 (yellow) lymphocytes and PD-L1(green) tumor cells were observed in the stroma and intratumoral sites of tumor specimen.
max : The maximum of standardized uptake value; MTV: metabolic tumor volume; TLG: total lesion glycolysis; CR: complete response; PR: partial response; SD: stable disease; PD: progressive disease; NLR: neutrophil to lymphocyte ratio; irAE: immune related adverse event; statistically significant p-values are shown in bold.

Table II .
Univariate and multivariate analysis of progression-free survival and overall survival.
max : the maximum of standardized uptake value; MTV: metabolic tumor volume; TLG: total lesion glycolysis; statistically significant p-values are shown in bold.

Table III .
Clinical benefit of overall survival (OS) >180 days according to different biomarkers.