[18F]Fluorodeoxyglucose Uptake by Positron Emission Tomography Predicts Outcome of Non–Small-Cell Lung Cancer
http://www.100md.com
《临床肿瘤学》
the Departments of Radiation Oncology, Nuclear Medicine, Thoracic and Cardiovascular Surgery, Thoracic/Head and Neck Medical Oncology, and Pathology, The University of Texas M.D. Anderson Cancer Center, Houston, TX
Division of Radiology, Kobe University Graduate School of Medicine, Hyogo, Japan
ABSTRACT
PATIENTS AND METHODS: One hundred sixty-two patients with stage I to IIIb NSCLC were analyzed. Overall survival (OS), disease-free survival (DFS), distant metastasis-free survival (DMFS), and local-regional control (LRC) were calculated by the Kaplan-Meier method and evaluated with the log-rank test. The prognostic significance was assessed by univariate and multivariate analyses.
RESULTS: There were 93 patients treated with surgery and 69 patients treated with radiotherapy. A cutoff of 5 for the SUV for the primary tumor showed the best discriminative value. The SUV for the primary tumor was a significant predictor of OS (P = .02) in both groups. Low SUVs (≤ 5.0) showed significantly better DFS rates than those with high SUVs (> 5.0; surgery group, P = .02; radiotherapy group, P = .0005). Low SUVs (≤ 5.0) indicated a significantly better DFS than those with high SUVs (> 5.0; stage I or II, P = .02; stage IIIa or IIIb, P = .004). However, using the same cutoff point of 5, the SUV for regional lymph nodes was not a significant indicator for DFS (P = .19), LRC (P = .97), or DMFS (P = .17). The multivariate analysis showed that the SUV for the primary tumor was a significant prognostic factor for OS (P = .03) and DFS (P = .001).
CONCLUSION: The SUV of the primary tumor was the strongest prognostic factor among the patients treated by curative surgery or radiotherapy.
INTRODUCTION
Recently, various methods of imaging tumor metabolism have been investigated.5,6 One of these, [18F]fluorodeoxyglucose positron emission tomography (FDG-PET), plays a central role in the staging of various types of cancer, including breast cancer, lymphoma, head and neck cancer, and NSCLC.7-13 Indeed, FDG-PET for NSCLC has been found useful for initial staging,10-12 restaging at recurrence,13 estimating radiotherapeutic or chemotherapeutic responses,13-15 and delineating radiotherapeutic targets.16 A few studies have further shown that standardized uptake values (SUVs), semi-quantitative values guided by FDG-PET, are useful for determining prognosis in patients with NSCLC treated with surgery.17,18 However, their usefulness in determining prognosis in patients treated primarily by radiotherapy remains uncertain. It is also unclear whether the SUV could predict the likelihood of either local relapse or distant metastasis in these patients. Likewise, the prognostic significance of the SUV for the regional lymph nodes and its agreement with the SUV for the primary tumor have not been clarified.
In this study, we assessed the prognostic ability of the SUVs in patients with stage I to IIIb NSCLC who were able to receive curative therapy. The ability of the SUVs to predict the outcomes of patients treated with surgery or radiotherapy was independently evaluated for each treatment group. We also studied patterns of failure and whether the SUV for the primary tumor predicted both local tumor control and the likelihood of distant metastasis.
PATIENTS AND METHODS
Patients were categorized into two treatment groups, the surgery group and the radiotherapy group. Patients in the surgery group were treated primarily by surgery with or without adjuvant therapy that consisted of or included radiotherapy, and patients in the radiotherapy group were treated with radiotherapy or chemoradiotherapy without surgery. At M.D. Anderson Cancer Center, routine follow-up evaluation was performed every 3 months for 2 years, every 6 months for 5 years, and then annually thereafter. However, in some cases, follow-up was performed at another hospital. In such cases, we contacted the patients or their physicians to obtain follow-up information. Before every follow-up visit, at least a chest x-ray or computed tomography scan of the chest was performed to check for evidence of recurrence. This study was approved by the institutional review board. Patient and tumor characteristics are listed in Table 1.
Semi-Quantitative Analysis by FDG-PET
Patients were first injected with 10 to15 mCi of FDG, and then 60-minute uptake phases and all images were scanned by one scanner (CTI-Siemens HR+ scanner; CTI-Siemens, Knoxville, TN), with vendor-provided software used for interpreting image data. The scans were all acquired in two dimensions at 5 minutes per field of view with attenuation correction at 3 minutes per field of view. The vendor-provided software used for image reconstruction performed iterative reconstruction and segmented attenuation correction. Patients fasted for at least 6 hours before the PET study. This was confirmed in all patients by a blood sugar level of less than 150 mg/dL. Measurements were conducted consistently in each patient.
For purposes of the staging work-up, the SUV was used to quantify tumor and lymph node uptake of the FDG. In our analysis, regions of interests were manually drawn on the transaxial images around the focal FDG uptake zone in the primary tumor, and the maximum SUV for each patient was used to minimize the partial-volume effects. The SUVs of the primary tumor and the regional lymph nodes were calculated using the following formula: SUV = activity concentration (μCi/mL)/(injected dose [mCi]/body weight [kg]).
Statistical Analysis
Data were analyzed using Stata 8.0 statistical software. Pearson's {chi}2 test was used to assess measures of association in a frequency table. The survival function was carried out using Kaplan-Meier estimates.19 The log-rank test was used to assess the equality of the survivor function across groups. The equality of means for continuous variables was assessed using t tests. A P value of .05 or less was considered statistically significant. Statistical tests were based on a two-sided significance level.
Survival time was measured from the date of the pathologic diagnosis to the first occurrence of the considered event (death, local-regional recurrence alone, distant metastasis alone, or any local-regional or distant recurrence). OS was defined as the time between diagnosis and death from any cause. Disease-free survival (DFS) was defined as the time between diagnosis and the first recurrence of the disease (local-regional or distant recurrence). Local-regional control (LRC) was defined as the time between diagnosis and the first local-regional failure. For the patients treated with radiotherapy, local-regional failure was defined as recurrence in the radiation field or the development of a malignant pleural effusion. Distant metastasis-free survival (DMFS) was defined as the time between diagnosis and the occurrence of the first distant metastasis.
The Cox proportional hazards model was used for the multivariate analysis to assess the effect of patients' characteristics and other prognostic factors of significance on the end points. The estimated hazard is reported. The Wald test was used to assess the role of covariates in the model.20
RESULTS
There were 69 patients (43%) in the radiotherapy group: 51 patients (74%) had stage IIIa or IIIb disease and 18 patients (26%) had medically inoperable stage I or II disease. Sixty-one patients (88%) underwent conventional radiotherapy 5 days a week, for a total dose of 60 to 66 Gy over 6 to 7 weeks. Six patients underwent hyperfractionated radiotherapy over 5.8 weeks for a total dose of 69.6 Gy delivered in 58 fractions. Two patients received higher doses of conventional radiotherapy alone (70 Gy in 35 fractions and 84 Gy in 42 fractions). In total, the median dose in the radiotherapy group was 63 Gy. In those patients who received chemotherapy, 45 patients (65%) underwent concurrent chemoradiotherapy, and two patients (3%) underwent sequential chemotherapy and radiotherapy. Of these 47 patients, 42 patients (89%) received two to seven courses of carboplatin and paclitaxel in combination. Other regimens were used in the remaining five patients (single-agent treatment with paclitaxel or cisplatin or combination treatment with cisplatin and etoposide, cisplatin and gemcitabine, or gemcitabine and vinorelbine).
FDG Uptake
The discriminative value of various cutoff SUVs for the FDG uptake by the primary tumor was analyzed in the context of OS and DFS (Fig 1). In the OS analysis, cutoffs of 4 and 5 showed significance, but 5 was the more discriminative of the two. Although in the DSF analysis dichotomization with a broad range of SUVs of between 3 and 8 gave significantly discriminative log-rank P values, the most discriminative cutoff again proved to be 5. Therefore, 5 was used as the cutoff SUV in the following analyses.
SUV for Primary Tumor and Survival
The median follow-up duration at the close-out date was 17.0 months. The 2-year OS rates in the surgery group, the radiotherapy group, and both groups combined were 76%, 71%, and 74%, respectively. However, when examined by SUV for primary tumor, the 2-year OS rate in the 43 patients with low SUVs (≤ 5.0) was significantly better than that of the 119 patients with high SUVs (> 5.0; 2-year OS, 94% v 65%; P = .02; Fig 2).
SUV for Primary Tumor and Tumor Progression
In 11 patients, it was difficult to distinguish each primary tumor accumulation of FDG from their lymph node accumulation. In these cases, the FDG accumulation was therefore regarded as the SUV for the primary tumor. As a result, the SUVs for primary tumors ranged from 0 to 27. The median SUVs for the primary tumor in the surgery group, the radiotherapy group, and both groups combined were 8.8, 8.2, and 8.4, respectively. Patients with low SUVs (≤ 5.0) showed significantly better DFS rates than those with high SUVs (> 5.0), both in the surgery group (2-year DFS, 69% v 50%; P = .02) and the radiotherapy group (88% v 29%; P = .0005; Fig 3), despite the fact that the two groups had quite different clinical stage distributions (Table 1; P < .0001).
The LRC and DMFS rates were then independently evaluated to clarify the pattern of relapse (Fig 4). In this analysis as well, patients with low SUVs showed better tumor control from the standpoint of both the LRC rate and the DMFS rate, regardless of whether they underwent surgery or radiotherapy (2-year LRC rate: surgery group, 83% v 61%, P = .05; radiotherapy group, 100% v 59% P < .0001; 2-year DMFS rate: surgery group, 75% v 66%, P = .08; radiotherapy group, 87% v 42%, P = .009).
Next, to confirm the discriminatory ability of the cutoff SUV of 5, the DFS by stage of disease (stage I or II [early] v stage IIIa or IIIb [advanced]) was assessed. Interestingly, in both stage subgroups, patients with low SUVs showed a significantly better 2-year DFS rate than those with high SUVs (stage I or II, 76% v 57%, P = .02; stage IIIa or IIIb, 90% v 31%; P = .004; Fig 5).
SUV for Regional Lymph Nodes
The prognostic ability of the SUVs for the regional lymph nodes was analyzed separately in 151 patients. The SUVs for the regional lymph nodes ranged from 0 to 3.5 in patients with N stage of 0 (median, 0), 0 to 9.1 in patients with N stage of 1 (median, 3.6), 0 to 15.7 in patients with N stage of 2 (median, 6.0), and 2.6 to 17.9 in patients with N stage of 3 (median, 5.6). In the event that several accumulations of FDG were observed in the regional lymph nodes, the maximum SUV was used for the evaluation as the SUV for the regional lymph nodes. Using the cutoff point of 5, the SUVs for the lymph nodes were not a significant prognostic factor for the 2-year DFS (58% v 30%; P = .19), the 2-year LRC (71% v 77%; P = .97), or the 2-year DMFS (71% v 37%; P = .17). Other cutoff SUVs between 3 and 10 were also evaluated, but these also did not yield significant results for DFS (0.07 < P < .52). It is noteworthy, however, that eight patients (5%) who had high (> 5.0) SUVs for their regional lymph nodes (range, 5.6 to 17.9; median, 8.4) but low (≤ 5) SUVs for their primary tumors (range, 0 to 4.8; median, 3.3) did not experience any recurrence.
SUV As a Prognostic Factor
The SUV for the primary tumor was identified as a significant prognostic factor for OS in both the univariate (P = .02) and multivariate (P = .03) analyses as well as age and N stage (Table 2). The value of SUV for the primary tumor was much greater in DFS and determined to be the strongest prognostic factor for DFS in both the univariate (P < .0001) and multivariate (P = .001) analyses (Table 3). It is important to note that treatment group was not identified as a significant prognostic factor for either OS (univariate, P = .22; multivariate, P = .94) or DFS (univariate, P = .41; multivariate, P = .92).
DISCUSSION
It is noteworthy that the treatment group (surgery or radiotherapy) did not prove to be a significant prognostic factor (Tables 2 and 3) and that the SUV for the primary tumor was a significant prognostic factor for both early-stage (stage I or II) and advanced-stage (stage IIIa or IIIb) disease (Fig 5). These data also seem to be original to our study. Although other factors, including the T stage and tumor size, N stage, and clinical stage, were also identified as significant prognostic covariables for OS and DFS in those univariate analyses, the SUV for the primary tumor showed stronger prognostic ability. Weight loss was not a significant factor in our analyses, unlike the findings of other studies.21,22 This may be because we included in our analysis only patients who could undergo curative therapy. Patients with severe weight loss therefore might not have been included in this study because they might have undergone palliative treatment or supportive care.
Vansteenkiste et al23 showed that the accuracy of FDG-PET and computed tomography in the evaluation of regional lymph nodes was 93% to 95%. However, the prognostic ability of the SUV for the regional lymph nodes remains uncertain. In our analysis, the SUV for the regional lymph nodes was not a significant prognostic factor in terms of the DFS when analyzed using cutoff SUVs of 3 to 10. It has also not been clarified whether the SUVs for primary tumors and regional lymph nodes are alike in their ability to predict prognosis. In our study, we observed that eight patients (5%) who had high SUVs for their regional lymph nodes and low SUVs for their primary tumors did not experience any local or distant relapse. Therefore, it is at least speculated that the SUVs for the regional lymph nodes do not agree with and are not stronger prognostic factors than the SUVs for the primary tumor. The SUVs for regional lymph nodes may also not be reliable because of the higher background activity within the mediastinum. Moreover, several factors, such as a history of chronic pulmonary inflammatory disease, might have affected the outcomes. Further investigation is warranted to learn more about the contribution of these factors.
On the basis of our findings that the SUV for a primary tumor predicted both local tumor control and distant metastasis, we hypothesize that tumor glucose metabolism is related to the metastatic potential of the tumor. Several investigators have also speculated that SUVs are correlated with cellular proliferation or biologic factors such as Ki-67, proliferating cell nuclear antigen, Glut-1, and hexokinase.24,25 However, the molecular mechanisms of FDG uptake in tumors are still a matter of debate. Our data further indicated that primary tumors showing high SUVs have the potential to be resistant to therapy and to metastasize. Further assessments of the correlation of SUVs with histopathology are ongoing to elucidate these molecular mechanisms.
In conclusion, the SUV for the primary tumor was the strongest prognostic factor in patients with early- and advanced-staged NSCLC, regardless of whether they underwent curative surgery or radiotherapy. Although a cutoff of 5 seems to be the most valuable in our study, analysis with larger numbers of cases or with longer follow-up is warranted for confirmation. These results thus indicated that the SUVs for primary tumors could be an important guide to decision making for patients with NSCLC.
Authors' Disclosures of Potential Conflicts of Interest
Acknowledgment
We thank Kazuro Sugimura, MD, professor in the Division of Radiology, Kobe University Graduate School of Medicine, and Peng Huang, PhD, an associate professor in the Department of Molecular Pathology, The University of Texas M.D. Anderson Cancer Center, for their generous advice and encouragement to R.S. We also thank Cora Bartholomew in the Department of Radiation Oncology, The University of Texas M.D. Anderson Cancer Center, for assisting in the preparation of this manuscript.
NOTES
R.S. is supported by a research fellowship grant from the Uehara Memorial Foundation, Kobe City, Japan.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
1. Asamura H, Nakayama H, Kondo H, et al: Lymph node involvement, recurrence, and prognosis in resected small, peripheral, non-small cell lung carcinomas: Are these carcinomas candidates for video-assisted lobectomy J Thorac Cardiovasc Surg 111:1125-1134, 1996
2. Feld R, Rubinstein LV, Weisenburger TH: Sites of recurrence in resected stage I non-small cell lung cancer: A guide for future studies. J Clin Oncol 2:1352-1358, 1984
3. Jemal A, Murray T, Samuels A, et al: Cancer statistics, 2003. CA Cancer J Clin 53:5-26, 2003
4. Jett JR, Scott WJ, Rivera MP, et al: Guidelines on treatment of stage IIIB non-small cell lung cancer. Chest 123:221S-225S, 2003 (suppl)
5. Pauwels EK, Sturm EJ, Bombardieri E, et al: Positron-emission tomography with [18F]fluorodeoxyglucose: Part I. Biochemical uptake mechanism and its implication for clinical studies. J Cancer Res Clin Oncol 126:549-559, 2000
6. Mankoff DA, Bellon JR: Positron-emission tomographic imaging of cancer: Glucose metabolism and beyond. Semin Radiat Oncol 11:16-27, 2001
7. Crippa F, Agresti R, Seregni E, et al: Prospective evaluation of fluorine-18-FDG PET in presurgical staging of the axilla in breast cancer. J Nucl Med 39:4-8, 1998
8. Kostakoglu L, Goldsmith SJ: 18F-FDG PET evaluation of the response to therapy for lymphoma and for breast, lung, and colorectal carcinoma. J Nucl Med 44:224-239, 2003
9. Allal AS, Dulguerov P, Allaoua M, et al: Standardized uptake value of 2-18F fluoro-2-deoxy-D-glucose in predicting outcome in head and neck carcinomas treated by radiotherapy with or without chemotherapy. J Clin Oncol 20:1398-1404, 2002
10. MacManus MP, Hicks RJ, Matthews JP, et al: High rate of detection of unsuspected distant metastases by PET in apparent stage III non-small-cell lung cancer: Implications for radical radiation therapy. Int J Radiat Oncol Biol Phys 50:287-293, 2001
11. Silvestri GA, Tanoue LT, Margolis ML, et al: The noninvasive staging of non-small cell lung cancer: The guidelines. Chest 123:147S-156S, 2003 (suppl)
12. MacManus MP, Hicks RJ, Ball DL, et al: F-18 fluorodeoxyglucose positron emission tomography staging in radical radiotherapy candidates with nonsmall cell lung carcinoma: Powerful correlation with survival and high impact on treatment. Cancer 92:886-895, 2001
13. Ryu JS, Choi NC, Fischman AJ, et al: FDG-PET in staging and restaging non-small cell lung cancer after neoadjuvant chemoradiotherapy: Correlation with histopathology. Lung Cancer 35:179-187, 2002
14. Hoekstra CJ, Stroobants SG, Hoekstra OS, et al: The value of [18F]fluoro-2-deoxy-D-glucose positron emission tomography in the selection of patients with stage IIIA-N2 non-small cell lung cancer for combined modality treatment. Lung Cancer 39:151-157, 2003
15. MacManus MP, Hicks RJ, Matthews JP, et al: Positron emission tomography is superior to computed tomography scanning for response-assessment after radical radiotherapy or chemoradiotherapy in patients with non-small-cell lung cancer. J Clin Oncol 21:1285-1292, 2003
16. Vanuytsel LJ, Vansteenkiste JF, Stroobants SG, et al: The impact of 18F-fluoro-2-deoxy-D-glucose positron emission tomography (FDG-PET) lymph node staging on the radiation treatment volumes in patients with non-small cell lung cancer. Radiother Oncol 55:317-324, 2000
17. Vansteenkiste JF, Stroobants SG, Dupont PJ, et al: Prognostic importance of the standardized uptake value on 18F-fluoro-2-deoxy-glucose-positron emission tomography scan in non-small-cell lung cancer: An analysis of 125 cases—Leuven Lung Cancer Group. J Clin Oncol 17:3201-3206, 1999
18. Higashi K, Ueda Y, Arisaka Y, et al: 18F-FDG uptake as a biologic prognostic factor for recurrence in patients with surgically resected non-small cell lung cancer. J Nucl Med 43:39-45, 2002
19. Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 53:457-481, 1958
20. Cox DR: Regression models and life-tables. J R Stat Soc B 34:187-220, 1972
21. Buccheri G, Ferrigno D: Importance of weight loss definition in the prognostic evaluation of non-small-cell lung cancer. Lung Cancer 34:433-440, 2001
22. Martins SJ, Pereira JR: Clinical factors and prognosis in non-small cell lung cancer. Am J Clin Oncol 22:453-457, 1999
23. Vansteenkiste JF, Stroobants SG, De Leyn PR, et al: Lymph node staging in non-small-cell lung cancer with FDG-PET scan: A prospective study on 690 lymph node stations from 68 patients. J Clin Oncol 16:2142-2149, 1998
24. Vesselle H, Schmidt RA, Pugsley JM, et al: Lung cancer proliferation correlates with [F-18] fluorodeoxyglucose uptake by positron emission tomography. Clin Cancer Res 6:3837-3844, 2000
25. Bos R, van Der Hoeven JJ, van Der Wall E, et al: Biologic correlates of 18fluorodeoxyglucose uptake in human breast cancer measured by positron emission tomography. J Clin Oncol 20:379-387, 2002(Ryohei Sasaki, Ritsuko Ko)
Division of Radiology, Kobe University Graduate School of Medicine, Hyogo, Japan
ABSTRACT
PATIENTS AND METHODS: One hundred sixty-two patients with stage I to IIIb NSCLC were analyzed. Overall survival (OS), disease-free survival (DFS), distant metastasis-free survival (DMFS), and local-regional control (LRC) were calculated by the Kaplan-Meier method and evaluated with the log-rank test. The prognostic significance was assessed by univariate and multivariate analyses.
RESULTS: There were 93 patients treated with surgery and 69 patients treated with radiotherapy. A cutoff of 5 for the SUV for the primary tumor showed the best discriminative value. The SUV for the primary tumor was a significant predictor of OS (P = .02) in both groups. Low SUVs (≤ 5.0) showed significantly better DFS rates than those with high SUVs (> 5.0; surgery group, P = .02; radiotherapy group, P = .0005). Low SUVs (≤ 5.0) indicated a significantly better DFS than those with high SUVs (> 5.0; stage I or II, P = .02; stage IIIa or IIIb, P = .004). However, using the same cutoff point of 5, the SUV for regional lymph nodes was not a significant indicator for DFS (P = .19), LRC (P = .97), or DMFS (P = .17). The multivariate analysis showed that the SUV for the primary tumor was a significant prognostic factor for OS (P = .03) and DFS (P = .001).
CONCLUSION: The SUV of the primary tumor was the strongest prognostic factor among the patients treated by curative surgery or radiotherapy.
INTRODUCTION
Recently, various methods of imaging tumor metabolism have been investigated.5,6 One of these, [18F]fluorodeoxyglucose positron emission tomography (FDG-PET), plays a central role in the staging of various types of cancer, including breast cancer, lymphoma, head and neck cancer, and NSCLC.7-13 Indeed, FDG-PET for NSCLC has been found useful for initial staging,10-12 restaging at recurrence,13 estimating radiotherapeutic or chemotherapeutic responses,13-15 and delineating radiotherapeutic targets.16 A few studies have further shown that standardized uptake values (SUVs), semi-quantitative values guided by FDG-PET, are useful for determining prognosis in patients with NSCLC treated with surgery.17,18 However, their usefulness in determining prognosis in patients treated primarily by radiotherapy remains uncertain. It is also unclear whether the SUV could predict the likelihood of either local relapse or distant metastasis in these patients. Likewise, the prognostic significance of the SUV for the regional lymph nodes and its agreement with the SUV for the primary tumor have not been clarified.
In this study, we assessed the prognostic ability of the SUVs in patients with stage I to IIIb NSCLC who were able to receive curative therapy. The ability of the SUVs to predict the outcomes of patients treated with surgery or radiotherapy was independently evaluated for each treatment group. We also studied patterns of failure and whether the SUV for the primary tumor predicted both local tumor control and the likelihood of distant metastasis.
PATIENTS AND METHODS
Patients were categorized into two treatment groups, the surgery group and the radiotherapy group. Patients in the surgery group were treated primarily by surgery with or without adjuvant therapy that consisted of or included radiotherapy, and patients in the radiotherapy group were treated with radiotherapy or chemoradiotherapy without surgery. At M.D. Anderson Cancer Center, routine follow-up evaluation was performed every 3 months for 2 years, every 6 months for 5 years, and then annually thereafter. However, in some cases, follow-up was performed at another hospital. In such cases, we contacted the patients or their physicians to obtain follow-up information. Before every follow-up visit, at least a chest x-ray or computed tomography scan of the chest was performed to check for evidence of recurrence. This study was approved by the institutional review board. Patient and tumor characteristics are listed in Table 1.
Semi-Quantitative Analysis by FDG-PET
Patients were first injected with 10 to15 mCi of FDG, and then 60-minute uptake phases and all images were scanned by one scanner (CTI-Siemens HR+ scanner; CTI-Siemens, Knoxville, TN), with vendor-provided software used for interpreting image data. The scans were all acquired in two dimensions at 5 minutes per field of view with attenuation correction at 3 minutes per field of view. The vendor-provided software used for image reconstruction performed iterative reconstruction and segmented attenuation correction. Patients fasted for at least 6 hours before the PET study. This was confirmed in all patients by a blood sugar level of less than 150 mg/dL. Measurements were conducted consistently in each patient.
For purposes of the staging work-up, the SUV was used to quantify tumor and lymph node uptake of the FDG. In our analysis, regions of interests were manually drawn on the transaxial images around the focal FDG uptake zone in the primary tumor, and the maximum SUV for each patient was used to minimize the partial-volume effects. The SUVs of the primary tumor and the regional lymph nodes were calculated using the following formula: SUV = activity concentration (μCi/mL)/(injected dose [mCi]/body weight [kg]).
Statistical Analysis
Data were analyzed using Stata 8.0 statistical software. Pearson's {chi}2 test was used to assess measures of association in a frequency table. The survival function was carried out using Kaplan-Meier estimates.19 The log-rank test was used to assess the equality of the survivor function across groups. The equality of means for continuous variables was assessed using t tests. A P value of .05 or less was considered statistically significant. Statistical tests were based on a two-sided significance level.
Survival time was measured from the date of the pathologic diagnosis to the first occurrence of the considered event (death, local-regional recurrence alone, distant metastasis alone, or any local-regional or distant recurrence). OS was defined as the time between diagnosis and death from any cause. Disease-free survival (DFS) was defined as the time between diagnosis and the first recurrence of the disease (local-regional or distant recurrence). Local-regional control (LRC) was defined as the time between diagnosis and the first local-regional failure. For the patients treated with radiotherapy, local-regional failure was defined as recurrence in the radiation field or the development of a malignant pleural effusion. Distant metastasis-free survival (DMFS) was defined as the time between diagnosis and the occurrence of the first distant metastasis.
The Cox proportional hazards model was used for the multivariate analysis to assess the effect of patients' characteristics and other prognostic factors of significance on the end points. The estimated hazard is reported. The Wald test was used to assess the role of covariates in the model.20
RESULTS
There were 69 patients (43%) in the radiotherapy group: 51 patients (74%) had stage IIIa or IIIb disease and 18 patients (26%) had medically inoperable stage I or II disease. Sixty-one patients (88%) underwent conventional radiotherapy 5 days a week, for a total dose of 60 to 66 Gy over 6 to 7 weeks. Six patients underwent hyperfractionated radiotherapy over 5.8 weeks for a total dose of 69.6 Gy delivered in 58 fractions. Two patients received higher doses of conventional radiotherapy alone (70 Gy in 35 fractions and 84 Gy in 42 fractions). In total, the median dose in the radiotherapy group was 63 Gy. In those patients who received chemotherapy, 45 patients (65%) underwent concurrent chemoradiotherapy, and two patients (3%) underwent sequential chemotherapy and radiotherapy. Of these 47 patients, 42 patients (89%) received two to seven courses of carboplatin and paclitaxel in combination. Other regimens were used in the remaining five patients (single-agent treatment with paclitaxel or cisplatin or combination treatment with cisplatin and etoposide, cisplatin and gemcitabine, or gemcitabine and vinorelbine).
FDG Uptake
The discriminative value of various cutoff SUVs for the FDG uptake by the primary tumor was analyzed in the context of OS and DFS (Fig 1). In the OS analysis, cutoffs of 4 and 5 showed significance, but 5 was the more discriminative of the two. Although in the DSF analysis dichotomization with a broad range of SUVs of between 3 and 8 gave significantly discriminative log-rank P values, the most discriminative cutoff again proved to be 5. Therefore, 5 was used as the cutoff SUV in the following analyses.
SUV for Primary Tumor and Survival
The median follow-up duration at the close-out date was 17.0 months. The 2-year OS rates in the surgery group, the radiotherapy group, and both groups combined were 76%, 71%, and 74%, respectively. However, when examined by SUV for primary tumor, the 2-year OS rate in the 43 patients with low SUVs (≤ 5.0) was significantly better than that of the 119 patients with high SUVs (> 5.0; 2-year OS, 94% v 65%; P = .02; Fig 2).
SUV for Primary Tumor and Tumor Progression
In 11 patients, it was difficult to distinguish each primary tumor accumulation of FDG from their lymph node accumulation. In these cases, the FDG accumulation was therefore regarded as the SUV for the primary tumor. As a result, the SUVs for primary tumors ranged from 0 to 27. The median SUVs for the primary tumor in the surgery group, the radiotherapy group, and both groups combined were 8.8, 8.2, and 8.4, respectively. Patients with low SUVs (≤ 5.0) showed significantly better DFS rates than those with high SUVs (> 5.0), both in the surgery group (2-year DFS, 69% v 50%; P = .02) and the radiotherapy group (88% v 29%; P = .0005; Fig 3), despite the fact that the two groups had quite different clinical stage distributions (Table 1; P < .0001).
The LRC and DMFS rates were then independently evaluated to clarify the pattern of relapse (Fig 4). In this analysis as well, patients with low SUVs showed better tumor control from the standpoint of both the LRC rate and the DMFS rate, regardless of whether they underwent surgery or radiotherapy (2-year LRC rate: surgery group, 83% v 61%, P = .05; radiotherapy group, 100% v 59% P < .0001; 2-year DMFS rate: surgery group, 75% v 66%, P = .08; radiotherapy group, 87% v 42%, P = .009).
Next, to confirm the discriminatory ability of the cutoff SUV of 5, the DFS by stage of disease (stage I or II [early] v stage IIIa or IIIb [advanced]) was assessed. Interestingly, in both stage subgroups, patients with low SUVs showed a significantly better 2-year DFS rate than those with high SUVs (stage I or II, 76% v 57%, P = .02; stage IIIa or IIIb, 90% v 31%; P = .004; Fig 5).
SUV for Regional Lymph Nodes
The prognostic ability of the SUVs for the regional lymph nodes was analyzed separately in 151 patients. The SUVs for the regional lymph nodes ranged from 0 to 3.5 in patients with N stage of 0 (median, 0), 0 to 9.1 in patients with N stage of 1 (median, 3.6), 0 to 15.7 in patients with N stage of 2 (median, 6.0), and 2.6 to 17.9 in patients with N stage of 3 (median, 5.6). In the event that several accumulations of FDG were observed in the regional lymph nodes, the maximum SUV was used for the evaluation as the SUV for the regional lymph nodes. Using the cutoff point of 5, the SUVs for the lymph nodes were not a significant prognostic factor for the 2-year DFS (58% v 30%; P = .19), the 2-year LRC (71% v 77%; P = .97), or the 2-year DMFS (71% v 37%; P = .17). Other cutoff SUVs between 3 and 10 were also evaluated, but these also did not yield significant results for DFS (0.07 < P < .52). It is noteworthy, however, that eight patients (5%) who had high (> 5.0) SUVs for their regional lymph nodes (range, 5.6 to 17.9; median, 8.4) but low (≤ 5) SUVs for their primary tumors (range, 0 to 4.8; median, 3.3) did not experience any recurrence.
SUV As a Prognostic Factor
The SUV for the primary tumor was identified as a significant prognostic factor for OS in both the univariate (P = .02) and multivariate (P = .03) analyses as well as age and N stage (Table 2). The value of SUV for the primary tumor was much greater in DFS and determined to be the strongest prognostic factor for DFS in both the univariate (P < .0001) and multivariate (P = .001) analyses (Table 3). It is important to note that treatment group was not identified as a significant prognostic factor for either OS (univariate, P = .22; multivariate, P = .94) or DFS (univariate, P = .41; multivariate, P = .92).
DISCUSSION
It is noteworthy that the treatment group (surgery or radiotherapy) did not prove to be a significant prognostic factor (Tables 2 and 3) and that the SUV for the primary tumor was a significant prognostic factor for both early-stage (stage I or II) and advanced-stage (stage IIIa or IIIb) disease (Fig 5). These data also seem to be original to our study. Although other factors, including the T stage and tumor size, N stage, and clinical stage, were also identified as significant prognostic covariables for OS and DFS in those univariate analyses, the SUV for the primary tumor showed stronger prognostic ability. Weight loss was not a significant factor in our analyses, unlike the findings of other studies.21,22 This may be because we included in our analysis only patients who could undergo curative therapy. Patients with severe weight loss therefore might not have been included in this study because they might have undergone palliative treatment or supportive care.
Vansteenkiste et al23 showed that the accuracy of FDG-PET and computed tomography in the evaluation of regional lymph nodes was 93% to 95%. However, the prognostic ability of the SUV for the regional lymph nodes remains uncertain. In our analysis, the SUV for the regional lymph nodes was not a significant prognostic factor in terms of the DFS when analyzed using cutoff SUVs of 3 to 10. It has also not been clarified whether the SUVs for primary tumors and regional lymph nodes are alike in their ability to predict prognosis. In our study, we observed that eight patients (5%) who had high SUVs for their regional lymph nodes and low SUVs for their primary tumors did not experience any local or distant relapse. Therefore, it is at least speculated that the SUVs for the regional lymph nodes do not agree with and are not stronger prognostic factors than the SUVs for the primary tumor. The SUVs for regional lymph nodes may also not be reliable because of the higher background activity within the mediastinum. Moreover, several factors, such as a history of chronic pulmonary inflammatory disease, might have affected the outcomes. Further investigation is warranted to learn more about the contribution of these factors.
On the basis of our findings that the SUV for a primary tumor predicted both local tumor control and distant metastasis, we hypothesize that tumor glucose metabolism is related to the metastatic potential of the tumor. Several investigators have also speculated that SUVs are correlated with cellular proliferation or biologic factors such as Ki-67, proliferating cell nuclear antigen, Glut-1, and hexokinase.24,25 However, the molecular mechanisms of FDG uptake in tumors are still a matter of debate. Our data further indicated that primary tumors showing high SUVs have the potential to be resistant to therapy and to metastasize. Further assessments of the correlation of SUVs with histopathology are ongoing to elucidate these molecular mechanisms.
In conclusion, the SUV for the primary tumor was the strongest prognostic factor in patients with early- and advanced-staged NSCLC, regardless of whether they underwent curative surgery or radiotherapy. Although a cutoff of 5 seems to be the most valuable in our study, analysis with larger numbers of cases or with longer follow-up is warranted for confirmation. These results thus indicated that the SUVs for primary tumors could be an important guide to decision making for patients with NSCLC.
Authors' Disclosures of Potential Conflicts of Interest
Acknowledgment
We thank Kazuro Sugimura, MD, professor in the Division of Radiology, Kobe University Graduate School of Medicine, and Peng Huang, PhD, an associate professor in the Department of Molecular Pathology, The University of Texas M.D. Anderson Cancer Center, for their generous advice and encouragement to R.S. We also thank Cora Bartholomew in the Department of Radiation Oncology, The University of Texas M.D. Anderson Cancer Center, for assisting in the preparation of this manuscript.
NOTES
R.S. is supported by a research fellowship grant from the Uehara Memorial Foundation, Kobe City, Japan.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
1. Asamura H, Nakayama H, Kondo H, et al: Lymph node involvement, recurrence, and prognosis in resected small, peripheral, non-small cell lung carcinomas: Are these carcinomas candidates for video-assisted lobectomy J Thorac Cardiovasc Surg 111:1125-1134, 1996
2. Feld R, Rubinstein LV, Weisenburger TH: Sites of recurrence in resected stage I non-small cell lung cancer: A guide for future studies. J Clin Oncol 2:1352-1358, 1984
3. Jemal A, Murray T, Samuels A, et al: Cancer statistics, 2003. CA Cancer J Clin 53:5-26, 2003
4. Jett JR, Scott WJ, Rivera MP, et al: Guidelines on treatment of stage IIIB non-small cell lung cancer. Chest 123:221S-225S, 2003 (suppl)
5. Pauwels EK, Sturm EJ, Bombardieri E, et al: Positron-emission tomography with [18F]fluorodeoxyglucose: Part I. Biochemical uptake mechanism and its implication for clinical studies. J Cancer Res Clin Oncol 126:549-559, 2000
6. Mankoff DA, Bellon JR: Positron-emission tomographic imaging of cancer: Glucose metabolism and beyond. Semin Radiat Oncol 11:16-27, 2001
7. Crippa F, Agresti R, Seregni E, et al: Prospective evaluation of fluorine-18-FDG PET in presurgical staging of the axilla in breast cancer. J Nucl Med 39:4-8, 1998
8. Kostakoglu L, Goldsmith SJ: 18F-FDG PET evaluation of the response to therapy for lymphoma and for breast, lung, and colorectal carcinoma. J Nucl Med 44:224-239, 2003
9. Allal AS, Dulguerov P, Allaoua M, et al: Standardized uptake value of 2-18F fluoro-2-deoxy-D-glucose in predicting outcome in head and neck carcinomas treated by radiotherapy with or without chemotherapy. J Clin Oncol 20:1398-1404, 2002
10. MacManus MP, Hicks RJ, Matthews JP, et al: High rate of detection of unsuspected distant metastases by PET in apparent stage III non-small-cell lung cancer: Implications for radical radiation therapy. Int J Radiat Oncol Biol Phys 50:287-293, 2001
11. Silvestri GA, Tanoue LT, Margolis ML, et al: The noninvasive staging of non-small cell lung cancer: The guidelines. Chest 123:147S-156S, 2003 (suppl)
12. MacManus MP, Hicks RJ, Ball DL, et al: F-18 fluorodeoxyglucose positron emission tomography staging in radical radiotherapy candidates with nonsmall cell lung carcinoma: Powerful correlation with survival and high impact on treatment. Cancer 92:886-895, 2001
13. Ryu JS, Choi NC, Fischman AJ, et al: FDG-PET in staging and restaging non-small cell lung cancer after neoadjuvant chemoradiotherapy: Correlation with histopathology. Lung Cancer 35:179-187, 2002
14. Hoekstra CJ, Stroobants SG, Hoekstra OS, et al: The value of [18F]fluoro-2-deoxy-D-glucose positron emission tomography in the selection of patients with stage IIIA-N2 non-small cell lung cancer for combined modality treatment. Lung Cancer 39:151-157, 2003
15. MacManus MP, Hicks RJ, Matthews JP, et al: Positron emission tomography is superior to computed tomography scanning for response-assessment after radical radiotherapy or chemoradiotherapy in patients with non-small-cell lung cancer. J Clin Oncol 21:1285-1292, 2003
16. Vanuytsel LJ, Vansteenkiste JF, Stroobants SG, et al: The impact of 18F-fluoro-2-deoxy-D-glucose positron emission tomography (FDG-PET) lymph node staging on the radiation treatment volumes in patients with non-small cell lung cancer. Radiother Oncol 55:317-324, 2000
17. Vansteenkiste JF, Stroobants SG, Dupont PJ, et al: Prognostic importance of the standardized uptake value on 18F-fluoro-2-deoxy-glucose-positron emission tomography scan in non-small-cell lung cancer: An analysis of 125 cases—Leuven Lung Cancer Group. J Clin Oncol 17:3201-3206, 1999
18. Higashi K, Ueda Y, Arisaka Y, et al: 18F-FDG uptake as a biologic prognostic factor for recurrence in patients with surgically resected non-small cell lung cancer. J Nucl Med 43:39-45, 2002
19. Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 53:457-481, 1958
20. Cox DR: Regression models and life-tables. J R Stat Soc B 34:187-220, 1972
21. Buccheri G, Ferrigno D: Importance of weight loss definition in the prognostic evaluation of non-small-cell lung cancer. Lung Cancer 34:433-440, 2001
22. Martins SJ, Pereira JR: Clinical factors and prognosis in non-small cell lung cancer. Am J Clin Oncol 22:453-457, 1999
23. Vansteenkiste JF, Stroobants SG, De Leyn PR, et al: Lymph node staging in non-small-cell lung cancer with FDG-PET scan: A prospective study on 690 lymph node stations from 68 patients. J Clin Oncol 16:2142-2149, 1998
24. Vesselle H, Schmidt RA, Pugsley JM, et al: Lung cancer proliferation correlates with [F-18] fluorodeoxyglucose uptake by positron emission tomography. Clin Cancer Res 6:3837-3844, 2000
25. Bos R, van Der Hoeven JJ, van Der Wall E, et al: Biologic correlates of 18fluorodeoxyglucose uptake in human breast cancer measured by positron emission tomography. J Clin Oncol 20:379-387, 2002(Ryohei Sasaki, Ritsuko Ko)