Introduction

Cancer of the anal canal is an uncommon disease accounting for approximately 1.5 % of all gastrointestinal cancers [1, 2]. The standard treatment is a conservative chemo-radiation regimen, proposed firstly by Nigro et al. [3] and confirmed by phase III trials [4, 5]. Treatment efficacy relies on accurate staging of primary tumor and regional lymph nodes [6], that it is performed by imaging studies such as computed tomography (CT) and magnetic resonance imaging (MRI) [7–9]. At this time, the two most significant prognostic factors are tumor size and nodal status related to the TNM stage [10, 11]. (18) F-fluorodeoxyglucose positron emission tomography/CT (FDG-PET/CT) has an important role in staging and treatment planning of various tumor locations [12–14] and may provide diagnostic useful information also for the anal canal cancer [15–19]. The recently updated European Society for Medical Oncology (ESMO) clinical practice guidelines for anal cancer include PET as an “optional but often recommended” modality in the workup of patients [1].

High FDG-PET/CT uptake, measured as the maximum standardized uptake value (SUVmax), has been shown to be predictive of recurrence and survival after radiotherapy in tumors such as cervical and lung cancer [20, 21].

The present study aimed at investigating the correlation amongst the baseline SUVmax, and tumor characteristics, treatment response and tumor control in a series of anal canal cancer patients treated with chemo-radiotherapy.

Materials and methods

Patient population

Fifty-five consecutive patients, with biopsy-proven anal cancer, who underwent definitive chemo-radiotherapy after acquisition of FDG-PET/CT imaging, were prospectively enrolled in the present study after obtaining written informed consent. The local ethics committee approved the present work.

All cases were discussed in a multidisciplinary conference with gastroenterologists, surgeons, radiation oncologists and medical oncologists.

Inclusion criteria were the following: age >18 years, biopsy-proven anal canal cancer, absence of distant metastasis, absence of other concomitant tumors and no previous pelvic RT. Pre-treatment evaluation included blood chemistries, clinical examination, endoscopy with anal biopsy, contrast-enhancement CT or MRI of the abdomen and pelvis, and whole body FDG-PET/CT. SUVmax was recorded from the PET/CT. tumor size was determined by clinical examination and inguinal lymph nodes were considered suspicious if greater than 1.5 cm on clinical or CT images or if they exhibited abnormally increased FDG uptake on pre-treatment FDG-PET/CT and biopsy confirmed, as described in a previous article [19]. Final staging was according to American Joint Committee on Cancer AJCC [22].

Patients were treated with chemo-radiotherapy with curative intent. Chemotherapy was delivered with three cycles of 5-fluorouracil and mitomycin-C in 30 patients and of 5-fluorouracil and cisplatin in 25 patients. RT was given to all patients in a homogeneous way including the identification of the clinical target volume that was defined as the gross tumor volume plus the potential microscopic spread to the surrounding tissues and to the regional lymph nodes based on CT simulation and FDG-PET fused images as recommended by Myerson et al. [9]. In the present study, FDG-PET/CT imaging was therefore used for both tumor staging and target identification. Our methods for patient’s simulation and target volume delineation have been previously described [19]. All patients started chemo-radiotherapy within ten working days from PET/CT. Clinical and treatment characteristics of the patients’ series are listed in Table 1.

Table 1 Patients’ and treatment characteristics
Full size table

PET/CT imaging

FDG-PET/CT was performed within five working days from the CT simulation scan. All patients underwent FDG-PET/CT with a hybrid PET/CT scanner (Biograph 16 HI-REZ, Siemens, Hoffman Estates, IL, USA) and CT images were used both for attenuation correction of PET data and for localization of FDG uptake. The PET component is a high-resolution scanner with axial diameter of field of view of 16 mm, the number of iteration was two for 24 subsets with a 168 × 168 pixels matrix size. The CT component of the study is the Somatom Sensation sixteen-slice CT (Siemens, Hoffman Estates, IL, USA). CT images (5-mm slices), performed without administration of intravenous contrast enhancement and with a low-dose protocol for CT acquisition (120–140 kV: 29–38 mA depending on the weight of patients), were obtained from the base of skull through the proximal femur. Fasting time was at least 6 h prior to PET/CT examination. The blood glucose levels of all patients were measured before the injection of 18F-FDG. The blood glucose levels were <150 mg/dL. After injection of 8 MBq of FDG per kg of body weight, the mean time between injection and acquisition was 70 min (range 55–90 min). Emission images ranging from the proximal femur to the base of the skull were acquired for 2–5 min per bed position, depending on body mass index. The PET data were corrected for scatter, dead time, random coincidences, and attenuation using standard algorithms supplied by the scanner manufacturer. The processed images were displayed in coronal, transverse and sagittal plans. The PET/CT images were interpreted in standard clinical fashion, both separately and in fused mode. The PET/CT images were reviewed for abnormal FDG uptake at the primary tumor, lymph node regions, and distant sites. To calculate SUVmax, a master ROI was positioned around the lesion with a threshold of 2.5 excluding area of physiological uptake as bladder. SUVmax was quantitatively assessed to determine FDG-PET activity. SUV is a semi-quantitative measure of radiotracer uptake and is calculated by the following formula: SUV = tissue radioactivity concentration [nCi/mL]/[injected dose (mCi)/patient weight (g)].

Clinical endpoints and follow-up

Data were collected to determine response to treatment, loco-regional and distant failure, disease-free survival (DFS) and overall survival (OS). tumor response, defined as complete (CR), partial (decrease ≥30 %) (PR), stable (SD), and progression disease (increase >20 %) (PD), according to the RECIST criteria [23] was assessed 4–6 months after treatment completion by digital examination, endoscopy with biopsy and PET/CT. During the follow-up, patients underwent similar exams as baseline every 6 months for the first 2 years and then every year. Diagnosis of recurrent tumor and distant metastasis was based on clinical and radiologic evidence of tumor relapse, possibly confirmed by biopsy.

Statistical analysis

The Wilcoxon test was used to analyze the association of primary tumor SUVmax with T, N and clinical stage, pathological findings in terms of histological type and grading and tumor response.

Actuarial rates of disease-free and overall survival (DFS and OS) at 2 and 5 years were calculated by the Kaplan–Meier method and were correlated with treatment response, T-stage, histological type and SUVmax by using the log-rank test. A p value of less than 0.05 was considered as statistically significant. Statistical analysis was performed using the SAS package version 8.02 (SAS Institute, Inc, Cary, NC, USA).

Results

The range of SUVmax varied from 3.20 to 82.73 with a median of 14.52 (Q 1 = 8.60; Q 3 = 19.55). The distribution of SUVmax for all patients is described in Fig. 1.

Fig. 1
figure 1

The histogram shows the distribution of the primary tumor SUVmax

Full size image

The analysis of the association between SUVmax of the primary tumor and tumor stage (T1–T2 and T3–T4) showed that SUVmax was significantly associated with T-stage (p = 0.01), the median SUVmax being higher for tumors presenting with T3–T4 stage than for those with T1–T2 stage. The same analysis for tumors with N0 versus N+ and for clinical stage I–II versus stage III did not find any significant association with SUVmax (p = 0.89 and 0.46, respectively).

Histology significantly influenced SUVmax: squamous cell carcinomas had a median SUV of 14.98 while non-squamous cell tumors had a median SUVmax of 7.67 (p = 0.03). No significant association (p = 0.38) was found between SUV and histological grading (G1–G2 vs. G3).

After treatment completion, 34/55 (62 %) CR and 21/55 (38 %) PR were observed. Neither SD nor PD was found. Median SUVmax was higher for lesions with PR than for lesions with CR (15.24 vs. 12.26), but this difference was not statistically significant (p = 0.17) (Table 2).

Table 2 Analysis of SUV with clinical parameters
Full size table

After a median follow-up of 51 months (range 6–66 months), 25/55 patients (45 %) experienced tumor relapse consisting in 12 cases (48 %) in loco-regional recurrence and in 13 (52 %) in distant metastasis.

The actuarial DFS rates were 53.0 % at 2 years and 41.3 % at 5 years and the actuarial OS rates were 77.8 % at 2 years and 58.6 % at 5 years. Using the log-rank test, treatment response was significantly correlated to DFS and OS: patients with CR had a significantly higher OS and DFS rates than patients with PR (p < 0.0001 for both parameters). Also tumor stage significantly correlated with DFS: patients with a T1–T2 stage had a significantly higher DFS rate than patients with a T3–T4 stage (p = 0.02). Lower DFS and OS rates were not significantly associated with an SUVmax below or above the median value of 14.52 (p = 0.62 and 0.41).

Discussion

The present research is one of the few studies which analyze the association between tumor characteristics and the prognostic and predictive significance of FDG SUVmax in patients with anal canal cancer treated with definitive chemo-radiotherapy. As a matter of fact, data regarding the prognostic value of baseline PET SUV in anal cancer come from relatively few retrospective studies with inconsistent results [24, 25].

We found that tumor SUVmax was associated with T-stage (p = 0.01) and histological type (p = 0.03). The association of SUVmax with risk of lymph node involvement, clinical stage and histological grading at diagnosis was not statistically significant.

Kidd et al. [24] in a sample of 77 patients reported quite different results observing a significant association of higher primary tumor SUVmax with an increased risk of lymph nodal metastasis at diagnosis, but not with tumor size and histology. The authors commented that because PET was used to stage lymph nodes, it could be a bias for the relationship between anal cancer SUVmax and lymph node positivity.

In our series of patients, higher values of SUVmax correlated with higher tumor stage (T3–T4). This finding could be of interest considering that other authors found a strong association of tumor size with local control and survival [10, 11, 26, 27].

In the present study, patients with higher SUVmax were also at an increased risk of PR on their post-treatment evaluation, but this finding was not statistically significant. Moreover, treatment response was strongly correlated with DFS (77.5 % for patients with CR vs. 14 % for patients with PR at 2 years) and OS (95.7 % for patients with CR vs. 49.9 % for patients with PR at 2 years) (p < 0.0001). Primary tumor stage significantly correlated with DFS (66.5 % for T1–T2 stage vs. 41 % for T3–T4 stage at 2 years) (p = 0.02). The present study did not show a significant direct correlation between SUVmax and DFS and OS. In this regard, other literature studies [24, 25] on the use of SUVmax in anal cancer reported controversial results. Kidd et al. did not find higher baseline SUVmax significantly associated with an increased risk of persistent disease. The same authors found that DFS was worse for patients with an SUVmax >5.6; however, the data had a limited statistical power (p = 0.05) [24]. In this regard, other studies in lung, oesophageal, head and neck and cervical cancers showed that higher SUVmax correlates with worse local control and disease-free survival rate [20, 21, 28, 29]. On the other hand, the more recent study of Bazan et al., similarly to us, found no correlation between SUVmax and DFS and OS in a sample of 39 patients [25].

In clinical practice, SUVmax is considered the most reproducible parameter that can estimate metabolic activity of FDG uptake. A possible limitation is that it represents the highest voxel value within a lesion. In this direction a growing body of literature has started to analyze more in deep SUVmean and metabolic tumor volume (MTV) [25, 30–32]. Actually, SUVmean represents the average of the intensity of uptake within a designated region of interest and MTV reflects the volume of tumor tissue with increased metabolic activity on PET. In this regard, Bazan et al. [25] analyzed SUVmax and MTV that resulted in a significant prognosticator for both DFS and OS.

Our study has some limitations. We analyzed only pre-treatment PET SUVmax. The use of other metabolic parameters like MTV and SUV mean could increase the capability of predicting treatment response and long-term prognosis. Moreover, the post-treatment FDG-PET/CT could allow for a better or worse prognosis prediction as observed by other authors [25, 33]. A future planned study at our institution will include MTV in the parameters to be analyzed. Finally, although the clinical series is not small considering the tumor type, a larger number of patients could have given more statistical power to the study. For these reasons, it would be interesting to do a multi-institutional trial although a few difficulties should be overcome including differences in PET scanners, imaging protocols, and software analysis among different institutions that could affect the measurements and modify the real impact of PET parameters.

Conclusion

This is one of the few studies analyzing the use of pre-treatment PET/CT SUVmax in anal canal cancer. Our data suggest that FDG-PET/CT for initial staging and treatment planning may also be used to identify those patients at increased risk of partial response and tumor relapse. Our findings showed that pre-treatment FDG-PET/CT SUVmax cannot directly predict tumor response and survival, but it is strongly associated with tumor characteristics such as primary tumor stage, the most important and validated prognostic factor for anal cancer.

Further larger prospective studies on the predictive and prognostic value of SUVmax and maybe also MTV and SUVmean are needed to better clarify the role of FDG-PET/CT in predicting tumor response and prognosis.