Treatment and clinical outcome of clinical T4 esophageal cancer: A systematic review

Abstract Background Survival of patients with cT4 esophageal cancer is dismal. Although the optimal treatment strategy remains to be established, two treatment options are available for cT4 esophageal cancers: definitive chemoradiation (dCRT) and induction treatment followed by conversion surgery (CS). However, little is known concerning the differences in clinical outcome between patients with T4 esophageal tumors treated with dCRT and those eventually treated with CS. Methods A systematic search of the scientific literature on PubMed/MEDLINE was carried out using the keywords “T4 esophageal cancer,” “invading (involving) adjacent organ,” “definitive chemoradiation,” “induction therapy,” “salvage surgery,” and “conversion surgery,” obtaining 28 reports published up to July 2018. Results/Conclusion We found that CS was superior to dCRT with respect to local disease control and short‐term survival; however, CS was associated with relatively higher perioperative mortality and morbidity. Alternatively, although dCRT might often cause fistula formation, a clinical complete response to dCRT is likely to lead to a better prognosis. Recent advances in chemotherapeutic agents have led to triple induction chemotherapy, with docetaxel, cisplatin, and 5‐fluorouracil (DCF), which has shown promise as an initial induction treatment for cT4 esophageal cancer. Indeed, this regimen could control both local and systemic disease, which enables curative resection without preoperative CRT. Moreover, some appropriate changes in perioperative management and intensive systemic chemotherapy might enhance patient outcome. Randomized controlled trials with a large sample size are needed to establish the standard treatment for cT4 esophageal cancer.


| INTRODUC TI ON
Esophageal cancers tend to invade adjacent organs, including the trachea, bronchus, lung, and aorta, as a result of the lack of serosa in the esophagus and the fact that this conduit is located in a very narrow mediastinal space. 1,2 Tumors that invade adjacent organs are classified as T4, according to the TNM staging system of the International Union against Cancer (UICC). Despite recent advances in multidisciplinary treatments, the prognosis of patients with T4 esophageal cancer remains unsatisfactory. 3 Although esophageal cancer is associated with a high incidence of morbidity and mortality, treating with surgery alone, where neighboring organs are resected together with an esophagectomy, has not improved survival. [4][5][6][7] Similarly, definitive chemoradiation (dCRT), a maximum-dose irradiation together with chemotherapy used as a curative treatment which many investigators consider the most suitable treatment for T4 esophageal cancer, has not dramatically contributed to improving patient survival. 8 As a result of a paucity of evidence, a treatment strategy for T4 esophageal cancer has not been established to date. According to the Guidelines for Diagnosis and Treatment of Carcinoma of the Esophagus, 2017, the current standard chemotherapeutic regimen for treating esophageal cancer is 5-fluorouracil (5-FU) combined with cisplatin 9 (CF) because of their synergistic radiosensitizing effects. 10 Previous studies have reported that concurrent CRT with a CF regimen was effective for treating advanced esophageal cancers, including T4 tumors. 1,11 Thus, two modalities are currently in use for the treatment of cT4 esophageal tumors: 12,13 dCRT [14][15][16][17][18][19] and induction chemotherapy or CRT, followed by conversion surgery (CS). 12,13,[20][21][22][23][24][25][26][27][28][29] Theoretically, when surgery completely removes the tumor, survival should be prolonged, regardless of the T stage. Thus, effective induction treatments must be established for T4 tumors to achieve curative resections, even for initially unresectable tumors. 30 However, to our knowledge, there is little or no information on the differences in clinical outcome between patients with T4 esophageal tumors treated with dCRT and those eventually treated with CS.
Recently, new triplet chemotherapy regimens have been reported to yield high response rates for esophageal squamous cell carcinoma (ESCC). 31,32 In particular, docetaxel plus CF (DCF) was shown to be more effective for treating ESCC than the standard treatment of CF or CF plus adriamycin (ACF). 31 Some studies showed promising results when induction chemotherapy with the DCF regimen was applied before carrying out CS for cT4 ESCC. 20,24,30 In the present review, we focus on these treatments and the outcomes in patients with T4 esophageal cancer, and we discuss future perspectives regarding these modalities.

| MATERIAL S AND ME THODS
We conducted a systematic search of the scientific literature on PubMed/MEDLINE to obtain all relevant articles involving T4 esophageal cancers published up to July 2018. In the searches, we excluded all non-English articles. To avoid duplications of data, articles from the same unit or hospital were included only once, when data were being updated in a later publication. The search terms were "T4 esophageal cancer," "invading (involving) adjacent organ," "definitive chemoradiation," "induction therapy," "salvage surgery," and "conversion surgery." All available major publications (primarily from high-volume surgical centers) were considered. Articles were selected when the abstract indicated that data were collected on patients with T4 esophageal cancer included in randomized controlled trials (RCT), other cohorts, or comparative studies. We also reviewed the reference lists of these articles to find additional candidate studies. For the present study, data were taken from the published reports; authors were not contacted to obtain additional information.
Therefore, articles that lacked necessary data, including survival information according to each treatment group, were excluded from this systematic review. Reports with fewer than 10 cases were also excluded from this study.

| Studies included in the present review
A fiow chart of the article selection process is shown in Figure S1.
A total of 28 articles regarding dCRT or/and induction treatment, followed by CS for cT4 esophageal cancer were finally selected (Table 1).

| Chemoradiation regimen
As summarized in Tables 1 and 2

| Toxicity and mortality as a result of induction CRT or chemotherapy
Yano et al 29 reported that the most common major toxicities (grade 3-4) caused by CRT (40 Gy/CF) were leukocytopenia (49%), followed by gastrointestinal toxicities (47%). In that study, one patient (2%) died of a treatment-related cause (pancytopenia). Ikeda et al 28
Miyata et al 24

| Triplet chemotherapy as an initial induction treatment
The standard regimen for induction treatment in locally advanced T4 esophageal cancer is concurrent CRT with CF. The CF regimen has not changed in decades, but it is possible that a stronger regi-  Calculated with intention-to-treat analysis.
aimed to test downstaging the tumor and, subsequently, converting to surgery as a multidisciplinary strategy for treating cT4 ESCC.
In that trial, the first-line chemotherapy regimen consisted of three

| SUMMARY AND PER S PEC TIVE S
A possible algorism of treatment for cT4 esophageal cancer is summarized in Figure 1. In the case of dCRT, patient prognosis depends on whether or not cCR can be achieved. However, it is often difficult to determine a treatment strategy after achieving cCR with F I G U R E 1 Possible algorism of a treatment strategy for cT4 esophageal cancer. Different treatment strategies, including (A) definitive chemoradiation (CRT), potentially followed by salvage surgery, in the absence of a complete response (CR); or (B) induction treatments potentially followed by conversion surgery. BSC, best supportive care; DCF, docetaxel, cisplatin, and 5-fluorouracil dCRT. It is also clinically difficult to make a diagnosis of CR based on endoscopic biopsies, which sometimes give false-negative results, or imaging tools, due to CRT-induced inflammation, fibrosis, or edema. In contrast, it remains controversial whether surgery should play a role in a treatment modality carried out after achieving CR with dCRT. Two randomized trials 45,46 have compared preoperative dCRT, followed by surgery, versus dCRT alone to assess the role of surgery in T3 and/or T4 diseases. They found that adding surgery to dCRT provided no survival benefit. Furthermore, significantly higher operative mortality rates and major morbidities, including anastomotic leaks and pulmonary complications, were reported in both trials. These findings were presumably due to the adverse effects of CRT, including radiation-induced fibrosis, which affected thoracic tissue and patient performance status. Meanwhile, as patients that achieved cCR after dCRT sometimes developed disease recurrence, careful follow up is necessary even after achieving cCR.
In addition, for recurrence or persistent disease (non-CR) after cCR, salvage surgery (optional) or palliation including chemotherapy may be indicated, depending on the clinical situation and the patient's general condition ( Figure 1A). However, when curative resection is considered possible after induction CRT or DCF, CS might be scheduled. When the tumor remains unresectable (persistent T4), chemotherapy or CRT might subsequently be given, depending on the type of initial induction treatment. In cases with persistent T4 tumors after an initial induction with DCF, a second-line induction CRT might be indicated to pursue any chance of carrying out CS as an optional treatment strategy; this latter option is practiced in our institute ( Figure 1B).
Older patients are often excluded, or at least underrepresented, in clinical trials. Thus, it is reasonable to question whether the results are generally transferable to the older population. Although it is true that some older patients are not suitable for intensive multimodality treatment, age alone should not be taken as the decisive factor in making treatment decisions in T4 esophageal cancer. In fact, according to a recent analysis by Pultrum et al., 47 older age did not significantly influence the overall outcome or the complication rate in patients treated with extended esophagectomies, However, the presence of comorbidity had a significant impact on survival.
Thus, it might be more appropriate to base treatment decisions on comorbidity and/or performance status, rather than chronological age alone. 48 Although we proposed a possible treatment algorithm for cT4 esophageal cancer (Figure 1), the tolerance for each treatment should first be evaluated, considering comorbidity, performance status, and general condition, in addition to the patient's age.
Radiation alone or palliation might be indicated for older patients at high risk; alternatively, a potentially curative treatment strategy might be considered for carefully selected older patients without severe comorbidity.
This review has shown that CS appeared to be superior to dCRT for treating T4 esophageal cancer with respect to local control and short-term prognosis despite the relatively high association with perioperative morbidities. However, although the fistula formation rate was relatively high in dCRT, a CR to CRT might lead to a better prognosis. When more powerful chemotherapy, such as a DCF regimen, is tolerable concurrent with definitive radiation, this is the most promising option for treating T4 esophageal cancer. Also, as an initial induction therapy, triplet chemotherapy, including a DCF regimen, can yield both significant local control and systemic control, which enables the application of CS for T4 esophageal cancer, without preoperative radiation. DCF chemotherapy can also be used for chemoselection, followed by CS or dCRT, as a multidisciplinary treatment strategy. In addition, a number of clinical trials are currently testing immune-checkpoint inhibitors with/without chemotherapy or radiation. These treatments might become viable treatment options for T4 esophageal cancer in the near future. Randomized controlled trials that include a large population are needed to define a standard treatment for T4 esophageal cancer.

D I SCLOS U R E
Authors declare no conflicts of interest for this article.