Introduction

Oesophageal has an increasing prevalence worldwide and is ranked the 8^th most common cancer. As there is no population wide screening, oesophageal cancer is often diagnosed at advanced stage, and associated with significant morbidities leading to poor quality-of-life and survival outcome [1]. Treatment of esophageal cancer for locally advanced oesophageal cancer has become more standardized since the CROSS trial over the last decade with oesophatectomy with neoadjuvant therapy. Oesopahgectomy for locally advanced oesophageal cancer however carries significant peri-operative and long-term quality of life risk. Despite introudction of minimally invasive surgical approach, morbidity and mortality remain significant at 60% and 5% respectively [2,3]. This neoadjuvant multimodal therapy with Neoadjuvant Chemoradiation Therapy (NCRT) with subsequent surgical resection appeared to achieve the maximal survival outcome [4,5].

¹⁸F-Fluorodeoxyglucose Positron Emission Tomography (¹⁸F-FDG PET) combined with Computed Tomography (CT) has become an integral part of multidisciplinary assessment tool of advanced stage oesophageal cancer. PET assessment has been shown to improve accuracy to initial staging of disease in detecting distal metastatic disease metabolic acitivty over conventional trimodality approach with contrast-enhanced CT and endoscopic ultrasonography with biopsy [6-8]. Utility of FDG PET/CT extends beyond staging of disease and had been widely adapated to assess neoadjuvant treatment response. Studies have investigated the prognostic value of index FDG PET/CT in predicting patient outcomes with FDG PET derived parameters [9-11]. Some studies even investigate the role of using FDG PET/CT to guide duration of neoadjuvant therapy, namely the ad-interim FDG PET/CT. An ad-interim assessment in early stage of NCRT to evaluate response of the disease to guide duration of neoadjuvant therapy and it’s utility for ad-interim FDG PET/CT remains controversial, and this practice had slowly became out of date [12].

Decision for oesophagectomy after completion of neoadjuvant therapy must be carefully assessed to avoid unnecessary harm to patients. Surgical intervention is inappropriate for patient who had developed interval metastatic disease [13]. The reliability of interval FDG PET/CT after NCRT had been extensively investigated for this purpose. It had been reported that 8% of interval FDG PET/CT detects true distant metastases [8]. Despite this, significant portion of patients with negative FDG PET/CT who underwent oesphagectomy after a negative interval FDG PET/CT assessment developed metastatic disease shortly after surgery [14]. This reflected that the imaging had underlying significant false negative rate. This highlighted the need to design further parameters to guide decision for surgical resection after NCRT. Studies had investigated interval FDG PET/CT derived parameters to predict pathological response and prognosticate disease survival. Although some predictive parameters have been reported in several studies, the reliability of these parameters have not been adequate to change practice paradigm [15-18]. In this study, our objective is to evaluate oesophageal cancer metabolic response from interval FDG PET/CT after neoadjuvant therapy and its predictive value to pathological response.

Methods

From October 2018 to April 2024, a retrospective cohort of patients in a high-volume referral centre for patients with oesophageal cancer were recruited. Patients were identified through the state-wide electronic medical record system and prospectively followed up. All patients with biopsied confirmed oesophageal cancer underwent index FDG PET/CT for initial staging of disease. Staging laparoscopy were occasionally performed to determine presence of disseminated abdominal disease. For patients with FDG PET/CT suggesting equivocal for metastatic disease, image guided biopsy would be performed to confirm metastasis. All oesophageal cancer patients were presented at our Multidisciplinary Team (MDT) meeting consisted of radiologist, pathologist, radiation oncologist, medical oncologists and specialist surgeon in oesophagectomy. The MDT determined clinical staging of disease, and treatment. Staging of disease was determined according to American Joint Committee on Cancer 8^th Edition (AJCC 8^th) [19]. A cohort of 70 patients with AJCC 8^th Stage II or above but without distant metastatic disease and were deemed appropriate surgical candidate included in the retrospective cohort from the study period. All the patients received neoadjuvant therapy and definitive surgical resection. Patient with metastatic disease at time of diagnosis or developed distal metastases after neoadjuvant therapy were excluded from the study. Patient who was determined unfit for curative intent chemoradiation therapy and oesophagectomy by the MDT were excluded from the study.

Neoadjuvant therapy regime was decided by the MDT. Majority of patients received CROSS protocol comprising of 5 cycles of carboplatin and paclitaxel, and concurrent external bean radiotherapy with total dose 41.4 Gy over 4 weeks. All patients underwent an interval FDG PET/CT after completion of neoadjuvant therapy. Ultimately, all patients received two FDG PET/CT for initial staging and interval assessment prior to surgical resection. Index scan was performed after biopsied confirmed diagnosis of oesophageal malignancy and second (interval) imaging was performed 4 weeks after completion of neoadjuvant therapy. FDG PET/CT were acquired on Biograph mCT PET/CT scanner (Siemens Healthineers Medical). FDG PET/CT acquisitions were analysed using Syno.via universal imaging software (Siemens Healthineers Medical). All images would be reviewed by radiologist at our centre at time of the MDT. FDG-avidity variables were recorded, such as SUV_max, number of FDG-avid lymph nodes, presence of FDG-avid metastases, mediastinal blood pool SUV, and liver background SUV. All patients deemed appropriate for surgery underwent hybrid Ivor-Lewis oesophagectomy consisting of 2-stages sequentially with abdominal laparoscopy and right thoracotomy or thorascopy depending on surgeon preference.

Radiological treatment response was classified according to European Organisation for Research and Treatment of Cancer (EORTC) PET treatment response criteria [20]. SUV_max of primary oesophageal lesion was recorded from index FDG PET/CT, and post-neoadjuvant treatment SUV_max was recorded from interval FDG PET/CT. SUV_max reduction ratio was calculated using ( [Index SUV_max - Interval SUV_max]/ Index SUV_max), and assigned to appropriate EORTC metabolic response group. Pathological treatment response was classified according to the Mandard tumour regression grading system [21]. Pathological TNM staging were assigned according to AJCC 8^th Edition. Mandard score was routinely determined by pathologist at our centre. We further classified Mandard score 1-3 as good pathological response and score 4-5 as poor pathological response.

Statistical Analysis

Statistical analyses were performed using Stata Statistical Software: release 18 (Stata Corp., College Station, Texas, USA). Receiver Operating Characteristics (ROC) analyses were performed to evaluate predictive value of SUV_max changes to Mandard pathological response. Kaplan-Meier estimators and log-rank test were performed to make comparison between categories of pathological response and metabolic response. Multivariable logistic regression analyses were performed to identified covariables associated with disease recurrence and overall survival. These covariables were carried forward to be included in the adjusted Cox proportional hazard model. Unadjusted and adjusted Cox proportional hazard model was performed for metabolic and pathological response to determine risk of recurrence.

Results

A study cohort of 70 patients with summary of baseline characteristics, initial tumour pathology and index clinical staging were demonstrated. Oesophageal adenocarcinoma accounted for 90%, and squamous cell carcinoma accounted for 10% of the cohort. All patients received neoadjuvant treatment, 61.4% received CROSS, and 38.6% received FLOT. Overall, 57.1% patient received adjuvant therapy after oesophagectomy. Outcome parameters including FDG PET/CT acquired metabolic response, Mandard pathological response, final tumour histology and post-treatment disease staging. Metabolic response of disease was measured as primary tumour SUV_max reduction ratio and further classified according to the EORTC response criteria (Table 2). Majority of patients (n = 38, 54.3%) demonstrated partial metabolic response (PMR). 15 (21.4%) patients had Complete Metabolic Response (CMR). 14 patients (20.0%) had Stable Metabolic Disease (SMD). 3 patients (4.3%) had Progressive Metabolic Disease (PMD) of the primary lesion on interval imaging without developing distal metastases. Patients with stable or progressive metabolic disease were combined in subsequent analysis as the same category.

Mandard pathological response was demonstrated in Table 1. Patients with Mandard score of 1-3 was categorized as good response, and score of 4-5 was categorized as poor response. There were 50 patients (71.4%) with good Mandard response, and 20 patients (28.6%) with poor Mandard response. Association of EORTC metabolic response to Mandard response were demonstrated in Table 3. Patient with poor EORTC metabolic response may be more likely to have poor pathological response with Mandard score of 5. But this observation was statistically insignificant, and no clear distribution pattern were observed in the Madnard score range of 1-4. The predictive value of primary lesion SUV_max reduction ratio to pathological response were evaluated with ROC curve analyses (Figure 1). The value of the SUV_max reducion ratio to predict good versus poor pathological response was only modest with an AUC of 0.64 (95% CI: 0.46-0.82). We were unable to identify the value of optimal sensitivity and specificity through the ROC analysis given the low power of the finding. Unadjusted and adjusted Cox hazard survival analysis were performed to evaluate correlations between Mandard pathological responses and EORTC metabolic response to recurrence-free survival. In our adjusted analysis, we have not observed any statistically significant difference in disease free survival between the Mandard response and EORTC metabolic response group (Table 4).

Table 1: Cohort patient demographics and pathology characteristics. Treatment, treatment response, final histopathology and follow up.

Table 2: Cohort breakdown of patients with various level of metabolic response.

Table 3: Association of EORTC Radiological Response to Mandard Score, n = 70.

Table 4: Cox Proportion Hazard model for recurrence free survival between Mandard pathological response and EORTC metabolic response

Figure 1: Receiver-Operating Characteristic for the predictive value of SUV_max reduction ratio in relation to Mandard pathological response.

Discussion

In this retrospective, we observed that there was no strong association between a patient’s FDG-PET/CT metabolic response post neoadjuvant therapy to pathological response on oesophagostomy specimen. In our simple chi-squire analysis between Mandard pathological response and EROTC metabolic response, there were no observable pattern to suggest patient with good metabolic response on FDG-PET/CT were more likely to have good Mandard pathological response. The only observation was that patient with poor metabolic response appeared to be more likely have TRG 5 Mandard score, reflecting failure to any treatment response. Compounding this, one can extrapolate from prior literature to suggest that evaluating disease prognosis can be difficult through FDG-PET/CT measured metabolic response. It had been demonstrated that a poor pathological response with high Mandard score may no necessitate poor response to neoadjuvant therapy, and that patient with poor pathological response does not equate poor survival outcomes [22]. In our study, the link between FDG-PET/CT derived metabolic response and pathological response have not been established. The result demonstrated that an FDG-PET/CT may not necessarily contribute to assessing patient’s risk for developing metastatic disease post oesophagostomy. Routine use of this imaging assessment may cause a population of patient unnecessary anxiety, given FDG-PET/CT had been reported to have approximately 5% false positive rate [8]. This could potentially lead to unnecessary further biopsy and delayed of ultimate surgical treatment for this sub-group of patients. Further to this, routine FDG-PET/CT imaging of patient post neoadjuvant therapy could be costly for the health care system. Although there is cost-benefit analysis performed at present, this could be worth while investigating.

This study has several limitations. This study was conducted in a retrospective fashion, and these carries inherent risk of selection bias. Most notably in our cohort, patient with interval FDG-PET/CT demonstrating development of new metastatic disease were not captured, and this could potentially skew results. Further to this, our studies included patient receiving either CROSS (neoadjuvant chemoradiation) or FLOX (neoadjuvant radiation along). Despite adjusting for the difference in our analysis, there may still be treatment effect not accounted for. In addition, this makes comparison to other existing literature difficult as most studies use single neoadjuvant treatment protocol. Lastly, our cohort of 70 patients were relatively small compared to other existing studies.

Conclusion

Our studies demonstrated that interval FDG-PET derived assessment of metabolic response to neoadjuvant therapy for oesophageal cancer did not predict patient’s pathological response outcome. Further to this, it appeared that both FDG-PET metabolic response and Mandard pathological response were poor predictor of patient’s survival outcomes.

References

1. Pennathur A, Gibson MK, Jobe BA, Luketich JD (2013) Oesophageal carcinoma. Lancet 381: 400-412.
2. Hii MW, Smithers BM, Gotley DC, Thomas JM, Thomson I, et al. (2013) Impact of postoperative morbidity on long-term survival after oesophagectomy. Br J Surg 100: 95-104.
3. Bras Harriott C, Angeramo CA, Casas MA, Schlottmann F (2022) Open versus hybrid versus totally minimally invasive Ivor Lewis esophagectomy: Systematic review and meta-analysis. J Thorac Cardiovasc Surg 164: e233-e254.
4. Eyck BM, van Lanschot JJB, Hulshof M, van der Wilk BJ, Shapiro J, et al. (2021) Ten-Year Outcome of Neoadjuvant Chemoradiotherapy Plus Surgery for Esophageal Cancer: The Randomized Controlled CROSS Trial. J Clin Oncol 39: 1995-2004.
5. van Hagen P, Hulshof MC, van Lanschot JJ, Steyerberg EW, van Berge Henegouwen MI, et al. (2012) Preoperative chemoradiotherapy for esophageal or junctional cancer. N Engl J Med 366: 2074-2084.
6. Noble F, Bailey D, Tung K, Byrne JP (2009) Impact of integrated PET/CT in the staging of oesophageal cancer: a UK population-based cohort study. Clin Radiol 64: 699-705.
7. Gillies RS, Middleton MR, Maynard ND, Bradley KM, Gleeson FV (2011) Additional benefit of ¹⁸F-fluorodeoxyglucose integrated positron emission tomography/computed tomography in the staging of oesophageal cancer. Eur Radiol 21: 274-280.
8. Kroese TE, Goense L, van Hillegersberg R, de Keizer B, Mook S, et al. (2018) Detection of distant interval metastases after neoadjuvant therapy for esophageal cancer with 18F-FDG PET(/CT): a systematic review and meta-analysis. Dis Esophagus 31.
9. Mantziari S, Pomoni A, Prior JO, Winiker M, Allemann P, et al. (2020) (18)F- FDG PET/CT-derived parameters predict clinical stage and prognosis of esophageal cancer. BMC Med Imaging 20: 7.
10. Malik V, Johnston C, O'Toole D, Lucey J, O'Farrell N, et al. (2017) Metabolic tumor volume provides complementary prognostic information to EUS staging in esophageal and junctional cancer. Dis Esophagus 30: 1-8.
11. Han S, Kim YJ, Woo S, Suh CH, Lee JJ (2018) Prognostic Value of Volumetric Parameters of Pretreatment 18F-FDG PET/CT in Esophageal Cancer: A Systematic Review and Meta-analysis. Clin Nucl Med 43: 887-894.
12. Cremonesi M, Garibaldi C, Timmerman R, Ferrari M, Ronchi S, et al. (2017) Interim (18)F-FDG-PET/CT during chemo-radiotherapy in the management of oesophageal cancer patients. A systematic review. Radiother Oncol 125: 200-212.
13. Lagergren J, Smyth E, Cunningham D, Lagergren P (2017) Oesophageal cancer. Lancet 390: 2383-2396.
14. Monjazeb AM, Riedlinger G, Aklilu M, Geisinger KR, Mishra G, et al. (2010) Outcomes of patients with esophageal cancer staged with [¹⁸F]fluorodeoxyglucose positron emission tomography (FDG-PET): can postchemoradiotherapy FDG-PET predict the utility of resection? J Clin Oncol 28: 4714-4721.
15. Abraham AG, Riauka T, Hudson M, Ghosh S, Zebak S, et al. (2023) (18)F-Fluorodeoxyglucose Positron Emission Tomography Parameters can Predict Long-Term Outcome Following Trimodality Treatment for Oesophageal Cancer. Clin Oncol (R Coll Radiol) 35: 177-187.
16. Hamai Y, Emi M, Ibuki Y, Murakami Y, Nishibuchi I, et al. (2020) Predictions of Pathological Features and Recurrence Based on FDG-PET Findings of Esophageal Squamous Cell Carcinoma after Trimodal Therapy. Ann Surg Oncol 27: 4422-4430.
17. Gillies RS, Middleton MR, Blesing C, Patel K, Warner N, et al. (2012) Metabolic response at repeat PET/CT predicts pathological response to neoadjuvant chemotherapy in oesophageal cancer. Eur Radiol 22: 2035-2043.
18. Kroese TE, Ruurda JP, Bakker AS, Jairam J, Mook S, et al. (2022) Detecting Interval Distant Metastases With 18F-FDG PET/CT After Neoadjuvant Chemoradiotherapy for Locally Advanced Esophageal Cancer. Clin Nucl Med 47: 496-502.
19. AJCC (2017) The American Joint Committee on Cancer) Cancer Staging Manual. 8 ed: Springer Cham 2017.
20. Young H, Baum R, Cremerius U, Herholz K, Hoekstra O, et al. (1999) Measurement of clinical and subclinical tumour response using [18F]-fluorodeoxyglucose and positron emission tomography: review and 1999 EORTC recommendations. European Organization for Research and Treatment of Cancer (EORTC) PET Study Group. Eur J Cancer 35: 1773-1782.
21. Mandard AM, Dalibard F, Mandard JC, Marnay J, Henry-Amar M, et al. (1994) Pathologic assessment of tumor regression after preoperative chemoradiotherapy of esophageal carcinoma. Clinicopathologic correlations. Cancer 73: 2680-2686.
22. Knight WRC, Baker CR, Griffin N, Wulaningsih W, Kelly M, et al. (2021) Does a high Mandard score really define a poor response to chemotherapy in oesophageal adenocarcinoma? Br J Cancer 124: 1653-1660.