Article Text
Abstract
Objective The aim of this study was to investigate the prognostic value of metabolic parameters obtained at pretreatment [18F]fluoro-2-deoxy-D-glucose positron emission tomography/computed tomography ([18F]FDG PET/CT) in patients with locally advanced cervical cancer. We hypothesize that these metabolic parameters could optimize the treatment decision and thus favor the outcome of patients suffering locally advanced cervical cancer.
Methods Patients with locally advanced cervical cancer underwent pretreatment PET/CT. Standard uptake values (maximum, mean, peak), metabolic tumor volume, and total lesion glycolysis were measured in the tumor and in the hypermetabolic pelvic lymph nodes. The relationship between clinical, pathological, and PET/CT metabolic parameters with recurrence-free survival and overall survival was assessed by Cox regression analysis.
Results 115 patients with a median age of 52 years (range 23–77) presented with locally advanced cervical cancer. After a mean follow-up of 33.0 months after initiation of therapy, 26 patients (22.6%) recurred of which 17 patients had distant metastasis; 18 (15.7%) patients died. Recurrence-free survival at 2 and 5 years was 79.2% and 72.2%, respectively. The total lesion glycolysis of the tumor and the delay between diagnosis and treatment were significantly associated with recurrence-free survival in the multivariate analysis (HR 1.00, p=0.004, and HR 2.04, p=0.02, respectively). Only the total lesion glycolysis of the tumor ≥373.54 (HR 2.49, 95% CI 1.15 to 5.38; p=0.02) remained significant after log rank testing. Overall survival at 2 and 5 years was 91.7% and 68.8%, respectively. The number of PET-positive pelvic lymph nodes was the only independent prognostic factor for overall survival in the multivariate analysis (HR 1.43, 95% CI 1.13 to 1.81; p=0.003).
Conclusion Tumor total lesion glycolysis and the number of positive pelvic lymph nodes on pretreatment PET/CT appear to be independent prognostic factors for recurrence and survival in patients with locally advanced cervical cancer. This may help to select patients who may benefit from therapeutic optimization and closer surveillance.
- cervical cancer
- gynecology
- neoplasm metastasis
- neoplasm recurrence
- local
Statistics from Altmetric.com
HIGHLIGHTS
Despite the combined modality treatment, patients with locally advanced cervical cancer have a poor prognosis.
Tumor total lesion glycolysis and the number of positive pelvic lymph nodes on pretreatment [18F]fluoro-2-deoxy-D-glucose positron emission tomography/computed tomography (FDG PET/CT) may predict survival.
Metabolic parameters of pretreatment FDG PET/CT may help to individualize treatment in patients with locally advanced cervical cancer.
Introduction
Although cervical cancer is considered to be a preventable disease, it is still the fifth most common cancer and responsible for a mortality rate of 40% among European women of all ages.1–3 The latter is due to the fact that still over 50% of women with cervical cancer are diagnosed at a locally advanced stage.4 Since 1999, concurrent radiotherapy and cisplatin-based chemotherapy is the standard treatment for patients with locally advanced cervical cancer, as it has demonstrated a benefit in both disease-free survival and overall survival rates compared with radiotherapy alone.5 6 Despite advances in treatment, a substantial fraction of patients will not respond to therapy and will have a poor prognosis with an approximately 65% chance of 5 year overall survival.7 Moreover, the combined modality treatment of chemotherapy and radiotherapy is achieved at a greater cost of acute and late toxicities with quality of life impairment.8
Focusing on individualized treatment planning for patients with locally advanced cervical cancer, the challenge remains to properly select patients for addition of consolidation chemotherapy to the standard concomitant chemoradiotherapy in the case of a high-risk profile while possibly reducing the radiation dose for patients with low-risk disease.9–11
It is therefore essential to discern tumor characteristics that help predict tumor response and/or survival. Multiple studies have explored the potential prognostic value of clinical, biological, pathological, and radiological features on survival outcomes in cervical cancer patients. However, none of these features is presently integrated in the treatment decision for patients with locally advanced cervical cancer. Only International Federation of Gynecology and Obstetrics (FIGO) staging with a particular focus on lymph node involvement is used.6 12
[18F]fluoro-2-deoxy-D-glucose positron emission tomography/computed tomography ([18F]FDG PET/CT) has become increasingly more relevant in the pretherapeutic assessment of patients with locally advanced cervical cancer, and is now considered the reference imaging technique to assess metastatic disease and lymph node status.7 Although still debated, pretreatment PET/CT of patients with cervical cancer may predict disease-free survival and overall survival.13 Most studies have used intensity-based metabolic parameters, such as maximum standardized uptake value of the primary tumor, to predict survival.14 More recently, volume-based metabolic parameters such as metabolic tumor volume and total lesion glycolysis of the tumor have been correlated with clinical outcomes.15 However, consensus has not been reached on how to integrate these parameters in clinical decisions for patients with locally advanced cervical cancer.
Here, we evaluated the prognostic value of intensity- and volume-based metabolic parameters not only of the tumor but also of the “global pelvic charge” including hypermetabolic pelvic nodes in patients with locally advanced cervical cancer. Based on a homogenous series of patients we hypothesized that these metabolic parameters are independent factors of outcome in patients with locally advanced cervical cancer.
Methods
Patients
Between March 2010 and December 2016, 115 consecutive patients with locally advanced cervical cancer were retrospectively identified in two centers (University Hospital of Liège, and Hospital de la Santa Creu i Sant Pau of Barcelona). The study was approved by the Institutional Review Board and informed consent was waived. The pre-treatment work-up consisted of clinical examination, radiologic evaluation, and surgical para-aortic lymph node staging. All had histologically proven cervical cancer at FIGO 2009 clinical stage ≥1B2.16 Conventional pelvic magnetic resonance imaging (MRI) was used for the evaluation of the loco-regional extent where the largest tumor diameter was measured. PET/CT was performed to assess lymph node status and rule out distant metastases. Patients without distant metastases were included in the study and subjected to further metabolic analyses on the primary tumor and the pelvic lymph nodes.
PET/CT Acquisition
Patients fasted for at least 6 hours before the intravenous injection of fluorodeoxyglucose (120–463 mBq, according to the patient’s weight). The mean blood glucose levels at the time of PET/CT were 98.61 mg/dL (range 72–242). PET/CT studies were acquired using three-dimensional PET/CT systems; two cross-calibrated PET/CT systems were used at the University Hospital of Liège (a GEMINI TF Big Bore and a GEMINI TF TOF 16) with a GEMINI TF TOF 64 at the Hospital de la Santa Creu i Sant Pau of Barcelona (Philips Medical Systems, Cleveland, OH). The acquisition was obtained after a mean time of 67 min (range 47–148) of rest and bladder voiding. An unenhanced low-dose CT was performed, followed by a PET emission scan performed from the mid-thigh to the base of the skull.
PET/CT Analysis
Volumes of interest were manually drawn around the cervical tumor and hypermetabolic pelvic nodes by an experienced nuclear medicine physician, blinded to the patient’s outcome. The relevance of all manually delimited volume of interest was thus assessed before performing the segmentation. A segmentation using a relative threshold of 40% of the maximal activity was performed inside the volumes of interest. Metabolic tumor volume, and maximum/mean/peak standardized uptake value, were then measured inside the segmented volume of interest. The total lesion glycolysis was then calculated, corresponding to the product of mean standardized uptake value times metabolic tumor volume, thus reflecting the global glycolytic activity of the volume of interest.17 The metabolic tumor volume and the total lesion glycolysis of the pelvic lymph nodes were defined as the sum of parameters of each single nodal lesion in case of multiple lymph node metastases. The “global pelvic” metabolic tumor volume and the “global pelvic“ total lesion glycolysis were calculated as the sum of parameters of the tumor and of all nodal uptakes. The patient’s metabolic parameters are summarized in Table 1.
Study Endpoints
The study endpoints were recurrence-free survival and overall survival. Recurrence-free and overall survival were defined as the time between the initiation of treatment and disease recurrence and death, respectively. Follow-up of patients consisted of clinical examination and sequential imaging, including PET/CT and/or pelvic MRI also performed when recurrence disease was suspected or significant treatment side effects occurred. Recurrence patterns were divided into four groups: no evidence of disease, local recurrence considered as centro-pelvic, nodal and distant recurrence. Patients alive were censored at the time of the last clinical follow-up. Clinical, radiological, and survival data were retrospectively collected from the patients’ medical records.
In accordance with the journal’s guidelines, we will provide our data for the reproducibility of this study in other centers if such is requested.
Statistical Analysis
Results were expressed as medians and range for quantitative variables, while numbers and percentages were used for categorical findings. Potential clinical factors and PET/CT parameters associated with recurrence-free survival and overall survival were submitted to univariate and multivariate analyses using a Cox proportional hazard regression model. Variables with p value ≤0.1 in the univariate analysis were selected for the multivariate analysis. To ease interpretation of potential predictors of recurrence-free and overall survival, the optimal cut-off (Youden’s Index) was determined for each clinical and PET/CT parameter by receiver operating characteristics (ROC) curve analysis. Recurrence-free and overall survival were estimated using the Kaplan-Meier method, and their comparison was made with the log-rank test. For both Cox and logistic multivariate regressions, the jackknife (leaving-one-out) method was applied to the stepwise selection procedure to assess the stability of the subset of variables selected. Results were considered statistically significant at the 5% critical level (p<0.05). All statistical analyses were performed using the Statistical Analysis System (SAS) package (version 9.4) and R (version 3.6.1) for the graphics.
Results
One hundred and fifteen patients with a median age of 52 years (range 23–77) were treated for locally advanced cervical cancer. After a mean delay of 58.3 days (range 18.3–164.7) from diagnosis, treatment started with a mean duration of 63.7 days (range 15.3–301.95). One hundred and two patients (88.7%) received platinum-based concomitant chemoradiation followed by image-guided adapted brachytherapy,6 except for one patient who refused brachytherapy. Individualized extended radiation field was applied when the histological para-aortic lymph nodes were metastatic. Of the 13 (11.3%) patients treated differently according to the decision of the multidisciplinary tumor board, three had neoadjuvant chemotherapy and surgery±radiotherapy, three patients had neoadjuvant chemotherapy followed by chemoradiation and brachytherapy, two patients had surgery followed by chemoradiation and brachytherapy, and six patients had adjuvant chemotherapy after chemoradiation and brachytherapy. The neoadjuvant chemotherapy was prescribed in case of high tumor volume or peritoneal perforation by the primary neoplasm. Adjuvant chemotherapy was initiated in patients with histologically proven positive para-aortic lymph nodes. Baseline clinical data were recorded and are summarized in Table 1.
After a mean follow-up of 33.0 months (range 3.1–82.4), 26 patients (22.6%) recurred of which 17 patients developed distant metastasis. Of the other nine patients with recurrence, six presented with nodal and/or local recurrence, and three were unknown. At the time of the last follow-up, 18 (15.7%) patients were dead of whom 13 (72.2%) had died due to cervical cancer.
Recurrence-Free Survival
Recurrence-free survival was 79.2% at 2 years, 76.5% at 3 years, and 72.2% at 5 years. Non-squamous cell carcinoma, high total lesion glycolysis of the tumor, total lesion glycolysis of the tumor and pelvic lymph nodes, and multiple positive pelvic lymph nodes at PET/CT correlated with worse recurrence-free survival (p=0.04, p=0.02, p=0.02, and p=0.03, respectively) in univariate analysis. Only tumor total lesion glycolysis remained significant in the multivariate analysis (HR 1.0, 95% CI 1.001 to 1.003; p=0.004).
Of the factors showing no significant relation to worse recurrence-free survival in univariate analysis but presenting a p value ≤0.1, the delay between diagnosis and treatment remained a significant prognostic variable in the multivariate analysis (HR 2.04, 95% CI 1.01 to 3.79; p=0.02) (Table 2). The jackknife method confirmed the accuracy of the tumor total lesion glycolysis cut-off value in 96.5% of cases. Regarding the delay between diagnosis and treatment, the accuracy of the cut-off was only in 46.0% of the cases. Log rank testing showed a worse survival in patients with pretreatment tumor total lesion glycolysis ≥373.54 (HR 2.49, 95% CI 1.15 to 5.38; p=0.02) and in patients with ≥2.07 months of delay between the diagnosis and the treatment, the latter being not significant (HR 1.89, 95% CI 0.87 to 4.09; p=0.11). Figure 1 shows the Kaplan-Meier curves for recurrence-free survival in function of tumor total lesion glycolysis.
Overall Survival
Overall survival was 91.7% at 2 years, 84.1% at 3 years, and 68.8% at 5 years. Univariate analysis showed that the number of PET-positive pelvic lymph nodes correlated with decreased overall survival (p=0.003). The PET-positive pelvic lymph nodes remained the independent prognostic factor for overall survival in the multivariate analysis (HR 1.43, 95% CI 1.13 to 1.81; p=0.003) (Table 3). The selected subset proved to be highly stable, being retrieved in 100% of cases by the jackknife method. In between the groups of patients dichotomized by optimal cut-off, worse survival was inherent in patients with ≥1 pretreatment PET-positive pelvic lymph nodes (HR 2.71, 95% CI 1.01 to 7.25, p=0.047).
Discussion
The tumor total lesion glycolysis and the number of positive pelvic lymph nodes on the pretreatment PET/CT are shown to be significant prognosticators for patient survival in our study. Various clinical, morphological, and molecular biomarkers may predict outcomes of patients with locally advanced cervical cancer. However, none are conclusive enough to reach the individual risk stratification beyond the existing FIGO staging system, mainly driven by the primary tumor stage and lymph node status.7 In the current era of personalized medicine, the PET/CT is continuously gaining ground in the prognostic assessment of patients with cervical cancer eligible for chemoradiation.
The intensity of the uptake of the primary tumor is an independent prognostic factor.18 In a meta-analysis on 1150 patients with cervical cancer, Sarker et al found that high pooled maximum standardized uptake value of the tumor was associated with a significantly higher risk of relapse or death.14 However, in the majority of the studies considered in this meta-analysis, maximum tumor standardized uptake value was not a significant independent prognostic factor. Probably, this is due to the fact that tumorous maximum standardized uptake value is known to be related to other poor prognosticators such as tumor size, depth of invasion, lymphovascular space invasion, and poorly differentiated histology.19 More recently, in two retrospective studies on patients with locally advanced cervical cancer, results showed that a high tumor maximum standardized uptake value was the main prognostic factor associated with poorer overall survival, response to chemoradiation,20 21 and disease-free survival.20 In the absence of validated cut-off of tumorous maximum standardized uptake value in the literature, an arbritary cut-off value of 7 was defined in the series of Cima et al, whereas in the series of Voglimacci et al the maximum standardized uptake value as continuous variable was a significant prognostic factor.20 21 None of the latter results are concordant with our series, nor in the series of Güler et al where the multivariate analysis of 129 patients with locally advanced cervical cancer treated with chemoradiation showed that tumorous maximum standardized uptake value was not significantly correlated with overall or disease-free survival.22
Contrary to standardized uptake value, volumetric approaches may better reflect the heterogeneous nature of the tumor such as hypoxia, cellular proliferation, and blood flow.23 Despite clinical and methodological differences across the studies, Han et al,15 in a meta-analysis based on 10 studies and 541 patients with mainly locally advanced cervical cancer, demonstrated that high tumor metabolic tumor volume or total lesion glycolysis were significantly associated with worse survival. In our series, we found that the total lesion glycolysis of the tumor was a significant prognosticator for recurrence-free survival in the multivariate analysis, with a Youden’s Index cut-off estimated at 373.54. These results were stable according to the jackknife validation method. Our results and those of Han et al and Hong et al were not concordant with other series,15 24 where tumor metabolic tumor volume or total lesion glycolysis were not found to be independently predicting survival.21 22 25 This discordance with our series can probably be explained by a longer follow-up of about 40.5 months in the series of Voglimacci et al, or by a higher risk profile with a recurrence rate of 55% to 65% as in the series of Güler et al and Akkas et al.21 22 25
Aside from variables linked to the primary neoplasm, the correlation of the metabolic parameters of the lymph nodes with survival has been less studied.21 25–27 However, it is suggested that nodal metabolic parameters could be less affected by inflammation related to exogenous pathogens than the metabolic parameters of the primary tumor, and could thus be more reliable.28 A number of studies have all shown that nodal maximum standardized uptake value was a significant predictor of survival.28 29 In line with these results, Onal et al demonstrated in a series of 93 patients with locally advanced cervical cancer that maximum standardized uptake value of the pelvic lymph nodes was a significant prognosticator of survival.27 Those results were not confirmed in the series of Voglimacci et al.21 As in our series, the metabolic tumor volume and total lesion glycolysis of the pelvic lymph nodes were not found to be prognosticators for survival either.
The inhomogeneous results of the previous studies led Hong et al to perform a comprehensive analysis of the metabolic activities of both primary tumor and metastatic lymph nodes.28 They showed that FIGO stage and the sum of metabolic tumor volume of the primary tumor and metastatic lymph nodes >59.01 cm3 were independent prognostic factors of recurrence-free survival. In our series, the total lesion glycolysis of the sum of the tumor and pelvic lymph nodes was prognostic in the univariate analysis for recurrence-free survival. However, the significance was not reached in the multivariate analysis where only the tumor total lesion glycolysis was an independent metabolic prognosticator. More likely is the link between hypermetabolic lymph nodes and survival due to the number of positive lymph nodes, rather than the summation of total lesion glycolysis activity.
Others studied the value of positive pelvic lymph nodes in terms of survival. On a larger sample of patients, Bogani et al confirmed that the number of histologically positive lymph nodes significantly impacts survival in FIGO 2018 stage IIICp cervical cancer.30 In particular, patients with more than one positive lymph node had a worse survival (p<0.001). In a series of 125 patients with locally advanced cervical cancer with no para-aortic lymph node PET uptake, Martinez et al found that patients presenting with more than one pelvic lymph node with high metabolic uptake had a higher recurrence rate than the counterpart group.31 Our study showed that the number of positive pelvic lymph nodes on the pre-treatment PET/CT is a significant prognosticator for overall survival, with a Youden’s Index cut-off suggested at ≥1 for statistical significance (HR 2.71, 95%, CI 1.01 to 7.25; p=0.047). The jacknife method retained a 100% variable stability.
There are a number of limitations in our study, mainly due to its retrospective nature and degree of heterogeneity in the treatment protocols which may have affected patient outcome. Also, although the clinical profile of the 115 patients was homogeneous, only a small number of events was observed and our results may be underpowered. However, the main strength of our study is the central review of all PET/CT by an expert nuclear physician in gynecological oncology, ensuring reliability of presented values. In addition, the jackknife method validated the tumor total lesion glycolysis (96.5% variable stability) and the number of positive pelvic lymph nodes (100% variable stability) on the pretreatment PET/CT, respectively, for recurrence-free and overall survival. Overall, according to the previous literature, the current study endorses the clinical value of volumetric functional assessment.
This study supports the proposition that the number of positive pelvic lymph nodes on PET/CT and tumor total lesion glycolysis are independent prognostic factors for survival in patients with locally advanced cervical cancer. There is a need for prospective randomized trials with large numbers of patients diagnosed with locally advanced cervical cancer to validate the prognostic value of a volume-based metabolic parameter before implementation in treatment protocols.
References
Footnotes
MDC and PL are joint first authors.
Contributors PL can be considered at first co-author. Conception/design: MDC, FK. Provision of study material or patients: FG, CG, RR, JDR, FK. Collection ± assembly of data: MDC, RR, JDR. Data analysis and interpretation: MDC, PL, RH, FK. Manuscript writing: MDC, PL, RH, FK. Final approval of manuscript: MDC, FK.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Patient consent for publication Not required.
Provenance and peer review Not commissioned; externally peer reviewed.
Data availability statement Data are available upon reasonable request.