Prognostic value of ¹⁸F-fluorodeoxyglucose bone marrow uptake in patients with solid tumors

Supplemental Digital Content is available in the text

Abstract

Background:

Several studies have reported the prognostic value of ¹⁸F-fluorodeoxyglucose (¹⁸F-FDG) bone marrow uptake (BMU) measured by ¹⁸F-FDG positron emission tomography (¹⁸F-FDG PET) in various cancers. We performed a meta-analysis to evaluate the prognostic value of ¹⁸F-FDG BMU in patients with solid tumors.

Methods:

Systematic searches of MEDLINE and Embase databases were performed using the keywords “¹⁸F-FDG,” “bone marrow,” and “prognosis.” All published human studies of the prognostic value of ¹⁸F-FDG BMU in patients with solid tumors were searched. The primary outcome was event-free survival (EFS), and the secondary endpoint was overall survival (OS); both of these were extracted directly from each study. The effects of ¹⁸F-FDG BMU on survival were assessed by using hazard ratios (HRs).

Results:

Ten studies with 1197 patients (8 studies reporting EFS in 1096 patients and 7 studies reporting OS in 836 patients) were included. In the EFS analysis, the combined HR was 1.75 (95% confidence interval [CI]: 1.45–2.11, P < .00001) in the random effects model (I² = 51%, P = .05). The combined HR of OS was 1.40 (95% CI: 1.13–1.73, P = .002) in the random effects model (I² = 52%, P = .05).

Conclusion:

This meta-analysis has demonstrated that patients with a low level of ¹⁸F-FDG BMU have better EFS and OS than those with a high level of ¹⁸F-FDG BMU. Based on our results, we suggest that ¹⁸F-FDG BMU could be used as a biomarker for stratifying the risk of tumor progression in patients with solid tumors.

1. Introduction

Cancer progression is dependent on the tumor microenvironment as well as on the characteristics of the tumor cells, and inflammation in the tumor microenvironment can contribute to tumor cell proliferation, survival, invasion, and metastasis.^[1,2] Recently, several studies have shown that systemic inflammation is significantly associated with the prognosis of various types of cancers.^[3–5] Systemic inflammatory markers, such as neutrophil-to-lymphocyte ratio, platelet-to-lymphocyte ratio, albumin, and C-reactive protein have been shown to be significant prognostic factors for clinical outcomes in patients with solid tumors.^[3–5]

Mild bone marrow uptake (BMU) of ¹⁸F-fluorodeoxyglucose (¹⁸F-FDG) in healthy subjects is well-known; however, some patients with different diseases show relatively high ¹⁸F-FDG BMU on positron emission tomography (PET) imaging.^[6,7] Previous studies have demonstrated that in patients with malignancy, ¹⁸F-FDG BMU can reflect BM activation in response to an inflammatory condition.^[6,7] Therefore, ¹⁸F-FDG BMU can reflect the systemic inflammatory response.^[6,7] Several studies have reported the prognostic value of ¹⁸F-FDG BMU measured on PET imaging in various cancers.^[8–17] Therefore, we performed a meta-analysis to evaluate the prognostic value of ¹⁸F-FDG BMU in patients with solid tumors.

2. Materials and methods

2.1. Literature search and study selection

We performed systematic searches of MEDLINE (from inception to September 2014) and Embase (from inception to September 2014) databases for English-language publications using the following keywords: “positron emission tomography” OR “FDG” or “positron emission tomography-computed tomography” OR “positron emission tomography computed tomography” OR “PET” OR “PET-CT” OR “PET CT” OR “PET/CT” OR “fluorodeoxyglucose”; “bone marrow” OR “BM”; and “prognostic” OR “prognosis” OR “predictive” OR “survival” OR “outcome.” All searches were limited to human studies.

The inclusion criteria included the following: the study included patients with pathologically confirmed malignancies; the study was a case control or cohort study; at least 1 ¹⁸F-FDG-PET/CT scan had been performed before at least 1 relevant prognostic factor was assessed, such as overall survival (OS), disease-free survival, event-free survival (EFS), progression-free survival or disease metastasis-free survival; hazard ratios (HRs) and 95% confidence intervals (CIs) were available or could be calculated based on data from the original articles; and the publication was in English. All published studies of the prognostic value of ¹⁸F-FDG BMU in patients with solid tumors were searched. Review articles, comments, letters, case reports, conference abstracts, and editorials were excluded. Studies with insufficient data available to calculate the HRs and 95% CIs, without prognostic parameters data, or with less than 10 patients were excluded. If more than 1 published study of the same target group were found in the same institution, only 1 report with the information most relevant to this study was included. Discrepancies were resolved by consensus.^[18]

2.2. Data extraction and quality assessment

Data were independently extracted from each article by 2 reviewers and recorded on a standardized form. Univariate HRs and their 95% CIs, P-values for the log-rank test, and the necessary statistics, such as 95% CIs, the number of events, and the number included in each group assessed by using the Kaplan–Meier curves, were recorded. Then, we used the methods suggested by Parmar et al and Williamson et al to convert these data into logHRs and standard errors.^[19,20] HR values were calculated by applying a spreadsheet and using the methods suggested by Tierney et al.^[21]

Data were extracted from the publications independently by 2 reviewers, and the following information was recorded: first author, year of publication, country, type of solid tumor, parameter of ¹⁸F-FDG BMU, number of patients, and clinical endpoints.

We assessed the quality of each included article using the Newcastle–Ottawa scale (www.ohri.ca/programs/clinical_epidemiology/oxford.asp), a comprehensive and systematic reviewing tool that was designed for retrospective and prospective studies. Studies with scores of ≥6 points on the Newcastle-Ottawa scale were considered high-quality studies and included in this meta-analysis. Discrepancies were resolved by consensus (Table 1). We performed all the analyses based on previously published studies; thus, no ethical approval was required.

Table 1.

Methodological quality of the potentially included studies according to Newcastle–Ottawa scale in this meta-analysis.

2.3. Statistical analysis

The primary outcome was EFS. Data regarding disease-free survival, recurrence-free survival, and distant recurrence-free survival were obtained from the included studies, and they were redefined as EFS, which was measured from the date of initiation of therapy to the date of recurrence or the last day of follow-up.^[22] The secondary endpoint was OS, which was defined as the time from the initiation of therapy until death from any cause.

The effects of ¹⁸F-FDG BMU on survival were assessed using HRs. Survival data were extracted following a methodology suggested previously.^[19] A univariate HR estimate and the 95% CIs were extracted directly from each study if they were provided by the authors. Otherwise, P-values of the log-rank tests, 95% CIs, numbers of events, and numbers of patients at risk were extracted to estimate the HR indirectly. Survival rates calculated from Kaplan–Meier curves were read using Engauge Digitizer version 3.0 (http://digitizer.sourceforge.net) to reconstruct the HR estimate and its variance, assuming that patients were censored at a constant rate during follow-up. An HR > 1 implied worse survival for patients with ¹⁸F-FDG BMU, whereas an HR < 1 implied a survival benefit for patients with ¹⁸F-FDG BMU.

Heterogeneity among studies was assessed through unadjusted analysis and random-effects models. The following measures were calculated as described previously^[23]: I² (total heterogeneity/total variability) and v2. Because of the small number of the studies, the Cochran Q test was not used.^[23] Meta-regression analysis was performed to search for population characteristics that were unevenly distributed among the studies. Funnel plots were used to assess for publication bias.^[24]P-values <.05 were considered to be statistically significant. The data from each study were analyzed using Review Manager (RevMan, Version 5.3, Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2012) and STATA version 15.1 (STATA Corp, College Station, TX).

3. Results

3.1. Study characteristics

After the comprehensive computerized search was performed and the references lists were extensively cross-checked, our research yielded 2648 records; 587 of these records were duplicated abstracts and were excluded after reviewing the title and abstract. In addition, 935 nonrelevant studies, 87 case reports, 493 conference abstracts, 71 letters, 13 editorials, 20 notes, and 430 review articles were excluded. The remaining 13 full-text articles were assessed for eligibility, and 3 articles were excluded due to inadequate data for the calculation of survival. Finally, 10 studies, involving 1267 patients, were eligible for inclusion in the meta-analysis.^[8–17] The detailed procedure is shown in Figure 1. Quality assessment of all the 10 studies was conducted (Table 1).

Figure 1.

Flow chart of the study selection process.

The prognostic value of ¹⁸F-FDG BMU was assessed by analyzing EFS in 1096 patients and OS in 836. There were 6 types of tumors in this study: lung cancer in four studies; lymphoma in 2 studies; and cervical cancer, colorectal cancer, malignant mesothelioma, and gastric cancer each in 1 study. The parameters of ¹⁸F-FDG BMU used were BM to liver uptake ratio (BLR) in 5 studies and mean standardized uptake value (SUV_mean) in four studies.

3.2. Heterogeneity analysis

I² values were 51 (P = .05) for event and 52 (P = .05) for all-cause death. The meta-regression analysis showed that all population characteristics listed in Table 2 were not significant sources of heterogeneity among the studies. Visual inspection of the funnel plot suggested no evidence of publication bias (Supplemental Fig. 1 and Supplemental Fig 2). The study characteristics are summarized in Table 2.

Table 2.

Studies included in the current meta-analysis.

3.3. Survival analysis

Eight studies^[8–13,17] analyzed EFS in 1096 patients, and 7 studies^[9,12–17] that reported OS in 836 patients were included to calculate the combined HRs. The HR of EFS ranged between 2.07 and 26.51, and the HR of OS ranged between 0.745 and 20.69. In the EFS analysis, the combined HR was 1.75 (95% CI: 1.45–2.11, P < .00001) in the random effects model (I² = 51%, P = .05). The combined HR of death in the OS analysis was 1.40 (95% CI: 1.13–1.73, P = .002) in the random effects model (I² = 52%, P = .05). The forest plots for EFS and OS are represented in Figures 2 and 3, respectively.

Figure 2.

Forest plots of the hazard ratios for events in patients.

Figure 3.

Forest plots of the hazard ratios for deaths in patients.

4. Discussion

This is the first meta-analysis, to our knowledge, to demonstrate the value of ¹⁸F-FDG BMU in mortality prediction in patients with solid tumors. Patients with high ¹⁸F-FDG BMU had significantly shorter EFS and OS compared with those with low ¹⁸F-FDG BMU.

The BM produces blood cells, which includes leukocytes, erythrocytes, and platelets. Most of the hematopoietic cells produced in the BM are of the neutrophil cell series.^[6] The BM can show a mildly increased FDG uptake in a healthy person. Furthermore, some patients with different diseases can show relatively high BMU on ¹⁸F-FDG PET imaging.^[6,7] Previous studies have demonstrated that in patients with benign or malignant diseases, increased BMU on ¹⁸F-FDG PET imaging can reflect activation of the BM in response to an inflammatory condition.^[6,7,25] The BM can show increased ¹⁸F-FDG uptake because of the systemic inflammatory response to a malignancy.^[26] Several studies have reported that inflammation in the tumor microenvironment can contribute to tumor cell proliferation, survival, invasion, and metastasis.^[1,2] Several parameters that reflect the systemic inflammatory response to a malignancy have recently been assessed as prognostic factors in the prediction of survival in patients with solid tumors.^[27–29] Several studies have explored the relationship between FDG BMU and serum inflammatory markers, and some studies have reported a significant correlation between FDG BMU and serum inflammatory markers.^{[8,11–13,16]}

Owing to the significant correlation between FDG BMU and serum inflammatory markers, some researchers have suggested that increased BMU on ¹⁸F-FDG PET imaging might be useful for predicting survival in patients with solid tumors.^[8–17] This study has demonstrated that increased BMU on ¹⁸F-FDG PET imaging is a prognostic factor in the prediction of EFS and OS. In previous studies, the multivariate analysis also showed that increased BMU on ¹⁸F-F DG PET imaging is an independent prognostic factor in the prediction of EFS and OS.

In current study, SUV_mean of the BM and BLR were used as parameters for assessing BMU on ¹⁸F-FDG PET imaging. A few studies reported both parameters, SUV_mean of the BM and BLR, were significant prognostic factors for survival in univariate analysis; however, only BLR has been found to be a significant independent prognostic factor for EFS in multivariate analysis. The present study did not calculate the combined HRs of each parameter. In the present study, we selected the parameter with the better parameter for calculation HR. With the addition of further studies, it is expected that a meta-analysis could be conducted to determine the prognostic values for each parameter.

We acknowledge the limitations of our meta-analysis that are inherent in the data availability. First, all the studies were not prospectively conducted; therefore, selection bias may have affected study quality. Furthermore, the Berkson bias may have affected the outcome. However, although the original registries were retrospective in nature, they were conducted without any interference with the routine clinical practice. Second, many studies were conducted by the same researcher; however, the tumor types in the patients enrolled were different in different studies. Most of the target groups are Asian, and there may be bias. Finally, there is possibly a publication bias against smaller and studies with nonpositive result as suggested by the visual inspection of the funnel plot. For generation of these meta-analysis results, a larger number of prospective consecutive studies might be warranted.

5. Conclusion

In conclusion, although the parameters of ¹⁸F-FDG BMU are not standardized and the types of the tumors were different in the published studies, patients with a low level of the parameter of ¹⁸F-FDG BMU had better EFS and OS than those with a high level of ¹⁸F-FDG BMU. ¹⁸F-FDG BMU can be used as a biomarker for stratifying the risk of tumor progression in patients with solid tumors based on the results of this meta-analysis of the published data. However, further large scale cohort studies should be performed to overcome limitations of this study.

Acknowledgements

This work was supported by a grant of the Korea Health Technology R&D Project through the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health & Welfare, Republic of Korea (grant number: HI16C1501), the Korea Health Technology R&D Project through the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health & Welfare, Republic of Korea (grant number: HI15C0001), and a National Research Foundation of Korea (NRF) grant funded by the Korean government (MSIP) (number: NRF-2015M2A2A7A01045177).

Author contributions

Conceptualization: Seong-Jang Kim.

Data curation: Shin Young Jeong, Kyoungjune Pak.

Formal analysis: Shin Young Jeong.

Funding acquisition: Sang-Woo Lee, Byeong-Cheol Ahn, Jaetae Lee.

Investigation: Seong-Jang Kim, Kyoungjune Pak.

Methodology: Shin Young Jeong, Seong-Jang Kim, Kyoungjune Pak, Sang-Woo Lee.

Project administration: Jaetae Lee.

Software: Seong-Jang Kim, Kyoungjune Pak.

Supervision: Seong-Jang Kim, Sang-Woo Lee, Byeong-Cheol Ahn, Jaetae Lee.

Validation: Kyoungjune Pak, Sang-Woo Lee, Jaetae Lee.

Writing – original draft: Shin Young Jeong, Seong-Jang Kim.

Writing – review & editing: Seong-Jang Kim, Kyoungjune Pak, Sang-Woo Lee, Byeong-Cheol Ahn, Jaetae Lee.