Prognostic Significance of Fluorine-18 Fluorodeoxyglucose Positron Emission Tomography in Anal Squamous Cell Carcinoma: A Systematic Review and a Meta-Analysis

Purpose Prognostic significance of fluorine-18 fluorodeoxyglucose positron emission tomography (18F-FDG-PET) in anal squamous cell carcinoma (SCC) has been evaluated in several studies; however, the results seem to be controversial and no consensus exists about its predictive capability. The current meta-analysis was carried out to comprehensively investigate the prognostic significance of 18F-FDG-PET parameters in patients with anal SCC. Methods A comprehensive literature search of PubMed/MEDLINE and Scopus databases was performed to retrieve pertinent articles published until August 5th 2018, concerning the prognostic significance of 18F-FDG-PET in patients with anal SCC. No language restriction was used. Several prognostic factors were reported for progression-free survival (PFS) and overall survival (OS) including pretreatment maximum standardized uptake value (SUVmax), metabolic tumor volume (MTV), inguinal nodal uptake, and metabolic response to therapy. Results Eleven studies (741 patients) were included. The pooled hazard ratio (HR) for the probability of PFS was 5.36 (95% confidence interval (95% CI): 3.12–9.21, p < 0.001) for metabolic response to therapy and 1.98 (95% CI: 1.26–3.12, p=0.003) for SUVmax. The pooled HR for the probability of OS was 5.87 (3.02–11.39, p < 0.0001) for metabolic response to therapy. On the other hand, the study revealed that the pooled HRs of MTV and inguinal nodal uptake for PFS were 1.56 (95% CI: 0.96–2.53, p=0.072) and 1.79 (95% CI: 1–3.21, p=0.051), respectively. Conclusions Our findings propose that some 18F-FDG-PET parameters could serve as prognostic indicators in anal SCC, but further larger studies are needed in this setting.


Introduction
Anal carcinomas represent approximately 4% of gastrointestinal cancers diagnosed annually, the majority of which are squamous cell carcinomas (SCCs) [1]. SCC of the anus is an uncommon malignancy but its incidence has increased considerably in recent years among women and men younger than 45 years. Over the past 3 decades, the incidence of anal SCC has increased by approximately 90% in men and 40% in women [2]. e rate of metastatic disease, mostly observed in the liver and the lungs, is low [3]. HIV and human papilloma virus-(HPV-) coinfected patients are at high risk of developing precancerous anal lesions (anal intraepithelial neoplastic lesions) and anal malignancies. Progression and persistence of HPV-associated lesions are known to be enhanced by human immunodeficiency virus-(HIV-) related immunosuppression, which may result in the reactivation of previously acquired HPV infection and loss of control of HPV viral replication [4], the phenomenon which explains the high risk of anal SCC in HIV-infected patients [2]. e diagnosis of anal neoplasms is usually made by physical examination and rectoscopy. In 15-20% of cases, regional lymph node metastases are present at the time of diagnosis [4][5][6]. Endorectal ultrasound (US) and/or pelvic magnetic resonance imaging (MRI) are required to evaluate tumor depth and regional spread. Haematogenous spread is rare at the time of diagnosis, but 40% of deaths during the course of the disease are due to distant metastases [7]. Anal carcinomas are categorized according to the TNM staging system [5]. e most important prognostic factors for anal cancer are known to be the tumor size and extent (T) and nodal involvement (N) [6]. Response to treatment could also be named as an important prognostic factor for anal cancer [6]. Recently, positron emission tomography with fluorine-18 fluorodeoxyglucose ( 18 F-FDG-PET) has become a valuable tool for staging, treatment, and surveillance of patients with various malignancies. Utilization of 18 F-FDG-PET for accurate staging of anal carcinomas is increasing as a result of several studies performed in this field [8]. ere are few reports, however, regarding the utility of 18 F-FDG-PET as a prognostic indicator in patients with anal cancer.
Abdominoperineal resection (APR) had been the standard method of treatment before 1980 for anal SCC, the mode of therapy which resulted in 5-year recurrence rates of 40-70% and 5-year overall survival (OS) rates of 24-62%. More recent studies, however, indicated that combined chemotherapy and radiotherapy could lead to similar OS rate to surgical treatment [7]. Currently, radiotherapy combined with 5-fluorouracil (5-FU) and mitomycin or 5-FU and cisplatin is used as the standard treatment for anal neoplasms. In the absence of significant toxicity, consolidation chemotherapy is continued either for a predefined time period or until evidence of tumor progression is noted [9]. Surgical treatment is solely considered in relapsed cases or cases with no response to chemoradiotherapy [5,10]. Adoption of such a therapeutic approach for locally advanced anal SCC has led to a 5-year OS ranging from 61% to 85%. A median OS of only 8-12 months has been reported for those patients with distant metastatic disease or recurrent locally advanced disease not amenable to APR [10].
All the data mentioned above indicates the significance of timely determination of disease recurrence and progression, so as to commence immediate therapeutic approaches, resulting in a better disease prognostication. Prognostic performance of 18 F-FDG-PET in anal SCC is unclear: several studies have been published over the years on this topic with conflicting results, not reaching a consensus. In addition, a meta-analysis of published studies was missing.
e aim of the current investigation is to systematically review and meta-analyze published data about the prognostic performance of 18 F-FDG-PET in anal SCC in order to provide evidence-based data in this setting.

Search Strategy and Study Selection.
A comprehensive computer literature search of PubMed/MEDLINE and Scopus databases was carried out to retrieve pertinent published articles concerning the prognostic significance of 18 F-FDG-PET in patients with anal SCC. We used a search algorithm based on a combination of the terms "(anus OR anal) AND (PET OR positron emission tomography)." No language restriction was used. e search was performed from inception to August 5th, 2018. e bibliographies of eligible studies were also screened to expand our search.
Studies or subsets in studies were included according to the following criteria: (1) more than 5 patients with biopsyproven anal SCC included; (2) performing at least one 18 F-FDG-PET scan before and/or after treatment; (3) containing survival data from which the hazard ratio (HR) could be extractable, providing at least one form of survival data, namely, progression-free survival (PFS) or overall survival (OS). Studies investigating the diagnostic role of 18 F-FDG-PET, in vitro studies and animal experiments, case reports, small case series, review articles, letters, editorials, conference proceedings, commentaries, and articles containing insufficient data to calculate the HRs were excluded. e studies with the most complete or recent data were included in case of data overlap in more studies.
Two researchers independently reviewed titles and abstracts of the retrieved articles, applied the above-mentioned inclusion criteria, rejected ineligible articles, and finally evaluated the full-text version of the included articles to determine their eligibility for inclusion.

Data Extraction and Quality
Assessment. Information about basic study characteristics (authors, year of publication, country of origin, study design), patients' characteristics (number of patients with anal SCC performing 18 F-FDG-PET, median age, gender, TNM staging of the anal SCC, and follow-up time), and technical aspects (injected activity of 18 F-FDG, acquisition modality, and time interval between 18 F-FDG administration and scanning) were collected. Furthermore, information about prognostic parameters were extracted including maximum standardized uptake value (SUVmax, calculated as the measure of the greatest amount of 18 F-FDG uptake in a region of interest divided by body weight), metabolic tumor volume (MTV, determined from the attenuation-corrected PET data using a software), metabolic response to therapy categorized as complete metabolic response (CMR) and partial metabolic response (PMR), inguinal nodal uptake (considering as positive a lymph node with an increased 18 F-FDG uptake, based on the criteria reported by different authors) and cut off values, as well as the survival data, including PFS and OS with HRs with 95% confidence intervals (95% CIs). Only studies providing such data were finally recruited in the meta-analysis.
Two independent reviewers assessed the methodology of the eligible studies using the Oxford Center for Evidence Based Medicine guideline to examine the quality of prognostic studies [11]. For each included paper, this tool takes into account several parameters comprising patient enrolment at a common point in the course of the disease, followup duration, method of verification of outcome, blind outcome assessment of PET findings, and adjustment for important prognostic factors, which could affect the final results [11].

Statistical Analysis.
Pooling of HRs and calculation of their 95% CI was performed using a random effects model to determine the prognostic significance of SUVmax, MTV, metabolic response to therapy, and inguinal nodal uptake.
e I 2 statistic was applied to evaluate the heterogeneity among studies, representing the percentage of total variation contributed by a between-study variation and ranging from 0% to 100% [12]. e publication bias was assessed using funnel plots and Egger's regression intercept [13]. All statistical analyses were performed by using Comprehensive Meta-analysis (version 2, Biostat Inc., USA) software. e final results were demonstrated as forest plots. Excluded (duplicates or title and abstract revealed not appropriate) (n = 394) Excluded (conference paper, review article, letter to editors, editorial, case report, inadequate data, and low sample size) (n = 24)

Study Characteristics.
e comprehensive computer literature search from PubMed/MEDLINE and Scopus databases revealed a total of 429 records, among which 394 were excluded after titles and abstracts were screened. e full-texts of the remaining 35 articles were carefully evaluated, and eventually 11 articles (741 patients) [14][15][16][17][18][19][20][21][22][23][24], found to be potentially eligible for inclusion applying the selection criteria mentioned above, were included in the current meta-analysis ( Figure 1). No additional studies were retrieved after screening the references of the selected articles. Basic study characteristics and methodological aspects of the 11 retrieved studies are summarized in Tables 1 and 2. As depicted in Table 2, the methodological quality of all included studies has been evaluated according to the Oxford Center for Evidence Based Medicine guideline to examine the quality of prognostic studies [11].
In the current review, five retrospective and six prospective studies about the prognostic significance of 18 F-FDG-PET in patients with anal SCC have been included. Table 2 demonstrates all the details regarding the PET prognostic parameters evaluated by each included study. Among the eligible articles, 5 studies (14, 15, 19, 21, and 24) evaluated the prognostic significance of metabolic response to treatment and 3 studies (16, 18, and 23) assessed the prognostic significance of inguinal nodal 18 F-FDG uptake.

Heterogeneity and Publication
Bias. Few pooled analyses of PET prognostic indices for PFS revealed mild heterogeneity (Table 3). Begg's funnel plot and Egger's test were used to examine the publication bias. e shape of the generated funnel plots seemed asymmetrical, which could signify the presence of possible clinically important publication bias. en, in order to provide statistical evidence of funnel plot asymmetry, Egger's test was carried out. However, no significant evidence of publication bias of the present metaanalysis (Egger's test p values > 0.1) was detected (Figure 4).

Discussion
18 F-FDG-PET imaging has been the focus of intensive research, revealing its ever-increasing role in the staging and management of patients with malignant diseases, and this is also the case for anal SCC [8]. Most studies on the role of this imaging modality in anal SCC have focused on its diagnostic and treatment planning significance [25][26][27]; however, only a few reports exist, analyzing and quantifying the association      between PET metabolic parameters and prognosis of anal SCC. As described earlier in this paper, some investigations have indicated the probable roles of certain PET indices, including the SUVmax on pretreatment 18 F-FDG-PET, MTV, metabolic response as determined by posttherapy 18 F-FDG-PET (categorized as complete and partial response groups), and inguinal nodal 18 F-FDG uptake, in yielding prognostic information on either OS or PFS beyond that of established prognostic markers in anal SCC [14][15][16][17][18][19][20][21][22][23][24]. e aforementioned studies on the prognostic value of 18 F-FDG-PET in patients with anal SCC have mostly revealed somehow contradictory results and could not reach a consensus. e current meta-analysis aimed to examine the prognostic significance of 18 F-FDG-PET in patients with biopsy-proven anal SCC to provide evidence-based data in this setting. Data from eleven studies (741 patients) were gathered and pooled. 18 F-FDG-PET may aid in tailoring treatment in patients with anal SCC based on data in the pretreatment and posttreatment settings, providing independently useful clinical information and improving the selection of patients who may benefit from more aggressive treatment [14,18,20,21].
Our meta-analysis demonstrated that metabolic response to therapy and preoperative SUVmax are relevant prognostic factors in patients with anal SCC; therefore, anal SCC patients with inadequate metabolic response to therapy and higher preoperative SUVmax of the anal tumor have a poorer prognosis and they could beneficiate from a more aggressive treatment (such as adequate inguinal irradiation or chemotherapy dose escalation or intensification) that cannot be routinely performed due to the expected increased toxicity [14,18,20,21].
Although the study did not indicate MTV and inguinal nodal 18 F-FDG uptake as statistically significant prognostic factors, the direction of effect was compatible with other PET metabolic indices. As the number of studies were limited, statistically nonsignificant pooled indices for MTV and inguinal nodal 18 F-FDG uptake are most likely due to low statistical power. e current study has some limitations that should be acknowledged when describing the results. Publication bias is a major concern in all meta-analyses as studies reporting significant positive findings are more likely to be published than those reporting negative results. Indeed, it is not unusual for small-sized early studies to report positive findings that subsequent larger studies fail to replicate. We cannot exclude that publication bias may have influenced the results of our analysis.
Furthermore, heterogeneity among studies may represent a potential source of bias in our meta-analysis. is heterogeneity is likely to arise through baseline differences among the patients in the included studies, diversity in methodological aspects between different studies, and different study quality. e overall quality of the studies included in our analysis was not excellent; this was partly caused as a result of lack of patients' recruitment at a common point of the disease course and the inability of the authors of eligible studies to carry out blind outcome assessment to the PET findings. ese factors make the overall findings less reliable.
Another limitation to be mentioned is the small number of patients enrolled in some of the included studies, which makes the results of our meta-analysis to be interpreted with caution. e limited number of investigations evaluating the prognostic significance of 18 F-FDG-PET parameters can be considered as a limitation. erefore, larger multicenter studies evaluating the prognostic value of several PET parameters are warranted.

Conclusions
Our meta-analysis demonstrates that the metabolic response to therapy, detected by 18 F-FDG-PET, as well as the preoperative SUVmax could serve as promising prognostic markers in patients with biopsy-proven anal SCC. ese prognostic markers could indicate which patients may beneficiate from more aggressive treatment. erefore, 18 F-FDG-PET may aid in tailoring treatment in patients with anal SCC based on data in the pretreatment and posttreatment settings, providing independently useful clinical information, but further large multicenter studies are needed to strengthen our results.

Conflicts of Interest
e authors declare no conflicts of interest.