Treatment Options for Patients With Persistent Disease After Initial Endotherapy for Esophageal Neoplasia

Abstract

There are numerous treatment options for esophageal dysplasia and neoplasia. Initial endoscopic therapy is the standard of care for high-risk lesions of the esophagus. However, residual and recurrent disease may be seen after initial endoscopic treatment, and selection of the optimal salvage treatment may be critical in achieving cure. Patients may be high risk for esophagectomy after assessment of risk factors such as age, frailty, and nutritional status. A wide array of salvage endoscopic eradication therapies may also be available, but consideration should be given to the initial endoscopic treatment modality, patient related risk factors, and expertise at the regional center. Chemoradiation may also play a key role in treating high risk patients and should be planned in conjunction with surgery or endoscopic therapy. Multidisciplinary team care improves survival in high-risk patients with high-risk lesions. Consideration should be given to a multidisciplinary approach after the diagnosis of dysplastic and malignant esophageal disease.

Key Learning Points

1. Decisions regarding operability should be made by a multidisciplinary team involving an esophageal surgeon and consider factors such as age, frailty, and comorbidities.

2. Endoscopic eradication therapy should be considered as both first line and salvage treatment for remnant and recurrent high-risk lesions.

3. Multidisciplinary Teams enhance survival and improve outcomes in patients with dysplastic and neoplastic esophageal disease.

Introduction

Therapeutic options for the management of dysplastic and neoplastic lesions of the esophagus have increased in both number and complexity over the last 2 decades. Endoscopic eradication therapy (EET), definitive chemoradiation and esophagectomy are established multimodal options that are recommended in a stepwise and sequential manner to treat early esophageal cancer. ¹ The spectrum of patients that can be treated have also widened, with a shift to endoscopic therapies for mucosal disease.^2,3

Esophageal adenocarcinoma (EAC) has established precursor conditions originating from GERD, which can transform normal mucosa to a lining containing intestinal metaplasia that is at risk of developing dysplasia. ⁴ Esophageal squamous cell carcinoma (SCC) also arises from squamous dysplasia. ⁵ As early dysplastic lesions have a low risk of lymph node metastasis, EET has emerged as an effective and minimally invasive set of techniques that consist of endoscopic resection (endoscopic mucosal resection (EMR) and endoscopic submucosal dissection (ESD)), radiofrequency ablation (RFA), cryoablation, argon plasma coagulation (APC), and photodynamic therapy (PDT). ⁶ If nodular mucosa is identified, endoscopic resection should be the initial technique chosen as it can simultaneously provide staging as well as curative therapy.^4,7

Even after successful EET, recurrence of both dysplastic and neoplastic esophageal mucosa can occur. ⁸ Sami et al analyzed a cohort of patients that underwent successful endoscopic ablation therapy in 5 centers. The annual rate of recurrence after treatment to CRIM (Complete remission of intestinal metaplasia) was 9.6% in the 594 patients studied. Of the 390 that initially had high grade dysplasia (HGD) or cancer, the recurrent HGD/cancer rate was 2.3%. ⁹ Thus, an organized and thorough post endoscopic surveillance plan should be discussed and implemented at the outset of EET. ⁸

High-Risk Lesions

During initial endotherapy, there may be lesion specific histologic factors that should be considered as high-risk. A finding of Barrett’s esophagus (BE) with HGD or T1a adenocarcinoma, and squamous dysplasia, typically leads to further EET. High-risk histological findings with an increased risk of metastases include:

• T1b adenocarcinoma (sm2 or sm3)

• Deep margin positivity

• Lymphovascular invasion

• Poor differentiation

There are also endoscopic resection specific factors that may prevent successful EET. Some of these factors include: 1-inability to lift a scarred area; 2-recurrent HGD/cancer despite multiple EET attempts; 3-proximal strictures that prevent passage of an endoscope or ablation accessories; and/or 4-inability to maintain a high level of acid suppression in the endoscopic treatment zone. ³ Additionally, endoscopic therapy may not be as effective in cases where:

• poor healing is seen after endoscopic resection

• technical problems are encountered due to previous EET

• there is an incorrect choice or sequence of EET

• secondary cancers are discovered ¹⁰

High-Risk Patients

Determination of high-risk surgical status for patients with esophageal malignancy is a complex and multi-factorial process. All aspects of the patient’s physiology, their past medical and surgical history, as well as their own goals of care must be considered together to make this decision. Declaring a patient inoperable may mean withholding the most effective treatment to cure a given cancer. Because of the impact of that decision, it should be made in a multidisciplinary fashion and must always include an experienced esophageal surgeon.

Age is both a significant risk factor for the development of esophageal cancer, and a major consideration when determining patient operability. Age is a predictor of non-cancer-specific mortality and is often associated with other comorbidities. While advanced age is an independent and strong factor associated with increased risk of short term morbidity and long term mortality after esophagectomy,^11-15 esophagectomy can be performed safely in select elderly patients.^16-18 Life expectancy and competing risks should be considered when selecting patients for surgical treatment. However, all risks factor, especially age, should be considered collectively in the overall context of the patient’s medical condition.

Performance status and frailty are closely tied to age but are separate considerations. Performance status, using the ECOG or Karnofsky scales, predicts mortality after esophagectomy, ¹¹ although there may be significant confounding from more specific risk factors. Frailty, an “age related cumulative decline in multiple physiological systems” leading to a “multidimensional state of increased vulnerability,” ¹⁹ is a stand-alone concept related to performance status. Many tools exist to measure frailty, ²⁰ which predict post-operative morbidity and mortality better than age alone after esophagectomy.^21,22

Malnutrition and sarcopenia are increasingly recognized as key factors in determining patient prognosis and response to therapy. ²³ This population is especially prone due to the combined effects of tumor-related dysphagia, deleterious impact of neoadjuvant treatment, and cancer-related cachexia. The importance of sarcopenia may approach that of tumor stage in determining overall and disease-free survival after esophagectomy. ²⁴

Each patient’s unique profile of comorbidities must also be evaluated, but few are absolute contraindications to esophagectomy. Cardiorespiratory disease, ²⁵ liver cirrhosis,^26,27 renal failure,^28,29 and diabetes ³⁰ are examples of comorbidities that are associated with increased perioperative risk in esophagectomy but can be managed in select patients to permit safe surgery. In addition, patient anatomy, past history of surgery and radiation may have a major impact on the complexity of esophagectomy surgery and reconstruction. A patient with multiple comorbidities who requires a complex esophageal replacement in the form of a colon or jejunal interposition, or who has had remote high-dose radiation to the mediastinum, may present a prohibitively high risk for surgery.

Because of the complex nature of selecting patients for esophagectomy, several risk-calculators have been developed to aid clinicians in evaluating patient fitness for surgery. These tools can help to predict both in-hospital ³¹ and 90 day mortality after esophagectomy ³² but should be viewed as a supplement and not a replacement for the clinical judgment of a multidisciplinary team. Ultimately, the decision of whether to offer surgery must be made by the treating surgeon and consider the above factors in a frank discussion with the patient about their own wishes and expectations.

Initial Endoscopic Resection Therapy

After confirmation of a high-risk lesion on endoscopy (squamous dysplasia, Barrett’s HGD or any nodular lesion within Barrett’s), endoscopic resection is often first-choice therapy. ³³ As well, patients with advanced lesions such as suspected T1b EAC may be referred for endoscopic therapy if deemed a poor candidate for surgery. In a multicenter cohort study that followed 141 patients with T1b EAC, 52% were treated endoscopically. This group was typically older, had more comorbidities, and had lower overall survival than the esophagectomy group. However, these patients were likely not candidates for esophagectomy and the 3 year survival rate if no high-risk histologic features were found was 93.8%. ³⁴ Thus, initial endoscopic therapy may be offered to high-risk patients due to clinical factors such age, cancer stage, comorbidity, or a patient’s own informed choice.^3,35

After embarking on EET in patients with suspected high-risk lesions, regular treatment intervals are strongly recommended.^6,33 Multiple studies have demonstrated the need for regular cycles of EET and ongoing surveillance. In a radiofrequency ablation versus sham controlled RCT for BE, 77.4% of 127 ablation patients achieved CRIM. ³⁶ Further follow up of 106 subjects from this initial cohort for a mean of 3.05 years revealed overall complete resolution of dysplasia (CRD) in 95%, CRIM in 93% and among patients with initial HGD, CRD in 93%. While this was an intention to treat study design, it highlights the intensity of follow up required after successful endoscopic therapy or biopsies starting at 3 months intervals, and then at least annually thereafter in the HGD group. ³⁷ A similar observation was found in the UK EMR and RFA registry where 355 patients with BE and early EAC were followed for a median follow up of 19 months. In this study, 96% of patients had HGD or intramucosal carcinoma (IMC) on entry, and 49% had an initial EMR. Follow up endoscopy was scheduled every 3 months for the first year, every 6 months for the second year and then annually for the duration of the study unless dysplasia recurred. ³⁸ The current ACG surveillance recommendations after achieving CEIM in patients starting with HGD or IMC are biopsies at 3, 6, and 12 months after EET, and then annually therafter. ⁴ Thus, prior to initiation of EET, it is imperative that the patient understands and consents to multiple procedures that may span months if not years.

The selection of initial EET in high-risk esophageal lesions may influence the overall treatment course and subsequent therapy. ³⁹ While guidelines agree on endoscopic resection for any visible lesions within a BE field, the recommendation of EMR or ESD as the initial treatment has not been as clear.^4,33

EMR may consist of focal resection, radical stepwise resection, multiband (ligator) mucosectomy, and cap assisted EMR. In a systematic review and meta-analysis, complete EMR of BE achieved 85% complete eradication with the primary adverse event being stricture formation (pooled estimate 37.4%). ⁴⁰ Comparative analysis of patients undergoing different types of cap assisted EMR revealed no difference between cap and snare compared to a band ligator device. ⁴¹

ESD as initial therapy for high-risk lesions of the esophagus has been implemented in Asian and European Societies.^42,43 ESD is recommended in SCC if lesion size >10 mm and there is no obvious submucosal involvement. In BE with visible lesions, ESD should be considered if: 1-size >15 mm; 2-poor submucosal lifting and; 3-risk of submucosal involvement. ⁴² The ability to provide en bloc resections may lead to more accurate staging, especially at lateral margins. However, ESD in the esophagus also has several limitations including: 1-steep learning curve; 2-increased complication rate; 3-longer case duration time; 4-paucity of formal training programs in the West. ⁴⁴

In comparing the 2 techniques, ESD appears to have a higher rate of CRD in Barrett’s with dysplasia ⁴¹ and complete resection rate in early esophageal cancer ⁴⁴ than EMR. The Japan Gastroenterological Endoscopy Society (JGES) recommends ESD over EMR for superficial esophageal cancers. ⁷ However, a small RCT comparing EMR and ESD for early Barrett’s neoplasia revealed no difference between the techniques for complete remission of neoplasia. ⁴⁵ Overall, the choice of EMR or ESD as an initial treatment technique relies on the availability and expertise of a regional/local ablation program.

The complexity of the patient’s presentation may also affect the result of initial endoscopic therapy. A recently developed model that predicts if a patient will have a complex treatment course was analyzed and externally validated. BE patients were considered at risk for a complex treatment course if overall length was >9 cm, initial HGD/adenocarcinoma was found, and/or poor squamous regeneration was observed. Proper identification of complex patients could lead to earlier identification of high-risk lesions, affect the EET chosen by the endoscopist and temper patient expectations after initial therapy. ⁴⁶

Endoscopic Therapy After Initial Treatment in High-Risk Patients

After initial endoscopic therapy, treatment of residual disease may depend on histologic and patient factors. Residual or remnant disease can be found on initial interval endoscopy and should be distinguished from late recurrent disease (after CRD or CRIM). A wide array of endoscopic techniques may be used individually or multimodally to treat remnant dysplasia or metaplasia.

Endoscopic Resection (ER) should be the initial consideration in cases where a mucosal defect or nodularity are seen after initial therapy.^3,47 In a study that followed 1000 consecutive patients with mucosal adenocarcinoma, 96.3% achieved remission of neoplasia with a median of 1 endoscopic resection and within 1 month of initiating therapy. Of the 3.7% that did not achieve remission, 32.4% underwent esophagectomy, while the remainder either continued endoscopic therapy, stopped all treatments due to comorbidities, or died of unrelated causes. ¹⁰ A systematic review that examined ESD for squamous cell carcinomas of the esophagus revealed that for T1a (m1/m2) cancers, complete resection was achieved in 86% to 94%, and the overall 5 years survival was 84% to 96.6%. ⁴⁸ Thus, initial endoscopic therapy for superficial EAC or SCC resulted in high rates of cure and remission.

In a series that followed 32 patients who had ESD after initial endoscopic resection of BE related neoplasia, 75% achieved R0 resection compared to 80% in a non-salvage naïve group. There were statistically significant technical challenges seen in the ESD salvage compared to the non-salvage group, with 92% (vs 25%) encountering submucosal fibrosis and 33% (vs 0%) with intra-procedural bleeding. However, the resection time was not significantly different (112 vs 90 minutes respectively) and the median time from initial therapy to the salvage ESD was 162 days. ⁴⁹ This small series demonstrates that salvage ESD of high-risk lesions after initial endoscopic therapy is safe and potentially curative.

Radiofrequency Ablation (RFA) can be safely used in patients undergoing initial endoscopic resection. Subramaniam et al retrospectively analyzed a cohort of 91 patients that had RFA-alone, or RFA after initial EMR or ESD. The post resection histology among ESD patients revealed 81.4% with T1a or T1b cancer, and 30.2% of EMR patients had T1a cancer. The CRIM rate among the ESD and EMR followed by RFA groups was 85.2% and 81.4%, respectively. Adverse events in the ER group included 1 bleed managed endoscopically, and 5 strictures that were treated with dilation. Overall, RFA after endoscopic therapy was shown to be safe and effective. ⁵⁰

In another series that followed 537 patients that had ER, RFA was used in 100% of patients who had EMR (n = 456) and 64.2% of patients that had ESD (n = 81). The overall rate of CRIM and adverse events did not differ between the groups. RFA was the preferred modality after ER to achieve CRIM. ⁴¹

Cryotherapy as an endoscopic therapy for esophageal disease is available as a spray or balloon delivery device. This thermal modality relies on cycles of freezing and then thawing tissue to induce cell damage and death. Cryotherapy has been studied as an adjunct therapy in esophageal cancer and as a primary ablative technique. ⁵¹ In an open label registry from 11 centers, 88 patients with T1a to T2 EAC received spray cryotherapy after initial treatment with another modality. The initial treatment was EMR in 34.1% and 19.3% had either APC, RFA or PDT. While a subgroup analysis was not reported, the overall complete response rate in T1 cancers was 66.2%. No specific adverse events were attributed to prior endoscopic therapy. ⁵² Another group of patients that may benefit from cryotherapy are those with persistent BE post RFA. The utility of salvage cryotherapy was analyzed in a small cohort of 18 patients from 2 centers. Following RFA, 11 patients had persistent dysplastic BE, while 7 patients had nondysplastic BE. Cryospray ablation achieved CRIM in 50% patients in this study. ⁵³ In another retrospective cohort, 46 patients had either spray cryotherapy or extended sequential RFA after initial EMR and circumferential RFA. Patients enrolled had BE with LGD, HGD, or EAC. The overall CRIM and CRD rates were higher in the sequential RFA group compared to cryotherapy (87% vs 57%, P = .04). As well, the BE length reduction was superior in the RFA group (median C0M0 vs C0M1). However, salvage cryotherapy was shown to be a viable option for treating persistent BE. ⁵⁴ Cryotherapy can also be applied as part of multimodal endoscopic management plan for dysplastic BE. In a review of 57 patients that underwent spray cryotherapy as part of a treatment regimen that included EMR and RFA, CRD was achieved in 98.1% and CRIM in 75%. In this group, 33.3% had persistent Barrett’s despite 3 or more treatments, and 68.4% had prior EMR. ⁵⁵

Photodynamic therapy was a pioneering non-ER ablative endoscopic tool used to treat BE and early esophageal cancer. Initial case series demonstrated the ability to destroy dysplastic and neoplastic tissue due to the concentrated uptake of a photosensitizer which is activated by a laser light.^56,57 PDT was shown in an international multicenter trial to provide superior treatment of BE with HGD compared to a proton pump inhibitor. ⁵⁸ As well, it was an effective therapy for palliation of obstructing or bleeding esophageal cancers. ⁵⁹ PDT was shown in systematic reviews to be safe, efficacious and cost effective in treating early esophageal cancer. ⁶⁰ However, due to the high cost of porfimer sodium, side effects from photosensitization, and the increased uptake of technologies such as RFA, PDT is not as readily available or used in many endoscopic ablation centers. Recently, there has been recognition that PDT may play a role in multimodal and salvage therapy. There is a paucity of recent data on the use of PDT after initial ER. In a series that compared PDT to ESD, patients with BE HGD, SCC and EAC were followed after ESD or PDT. While there was no difference in disease-free survival, the PDT group had significantly lower procedure time and complications (decreased bleeding, perforation, and stricture formation). ⁶¹ Salvage PDT was used in patients with local recurrence after chemoradiation for esophageal cancer. In a study that followed 77 patients who had EMR or PDT after chemoradiation, 65.8% of the PDT group (22 of 38) had a complete response with a 5 year survival of 41.6%. ⁶² More recently, the introduction of second-generation photosensitizers such as talaporfin sodium along with portable diode lasers may allow for treatment of high-risk lesions of the esophagus with less toxicity and in conjunction with other modes of endoscopic therapy. ⁶³

Chemoradiation Therapy in High-Risk Patients

Chemoradiation therapy after initial endoscopic therapy has been reported in several series. Hamada et al retrospectively reviewed 66 patients with superficial esophageal squamous cell cancer initially treated by ER (cap EMR or ESD). High risk histologic features after initial ER included T1b tumor depth in 65%, deep margin positivity in 8% and lymphovascular involvement in 55%. Patients were then treated with cisplatin and 5-fluorouracil and concurrent radiotherapy with a total of 40.0 to 41.4 Gy. While metastases occurred in 9% (all had lymphovascular involvement), the overall survival at 5 years was 75%. ⁶⁴ Nagami et al ⁶⁵ followed 84 patients who had ESD for HGD or T1a and T1b squamous cell cancer. Chemoradiation was administered after endoscopic therapy in 18 patients. The overall 5 year survival of this cohort was 99%, demonstrating that ESD of high-risk lesions followed by CRT in specific patients can be a curative multimodal strategy.

Patients with locally advanced resectable disease may be at risk of local as well as distant recurrence and hence, trimodality therapy involving the combination of surgery and chemoradiation has been recommended.^66,67 Currently, those eligible for trimodality therapy are T1N1 or T2-3N0-1 M0 patients. However, those with an WHO performance score of >2 or weight loss of ≥10% are not eligible. Patients also must have adequate renal, pulmonary, hepatic, and hematologic function to proceed with trimodality treatment. A subset of patients with locally advanced non-metastatic disease may have surgically unresectable disease due to medical causes or tumor location (cervical region). Definitive chemoradiation is the standard locoregional therapy for this group of patients with non-metastatic unresectable disease. The RTOG 85-01 study reported a 5 year overall survival rate of 25% and locoregional control rate up to 40% to 60% following definite chemoradiation. ⁶⁶

Multidisciplinary Team (MDT) Management of High-Risk patients

After the diagnosis of an invasive cancers, most regional cancer programs use MDTs to enhance patient care. Some of the key reasons they are used to enhance clinical decision making include: 1-increasingly complex and numerous treatment options; 2-involvement of all key professional groups; 3-to facilitate timely access of appropriate treatments; and 4-ensure continuity of care. ⁶⁷ The essential elements of a multidisciplinary team including the composition, clinic structure, and conference structure are well described in a descriptive study by Boniface et al. The focus on proper staging, direct communication of different specialties, and a single visit approach reduces wait times and expedites initiation of treatment. ⁶⁸

In patients with gastro-esophageal cancers, MDTs have been shown to improve staging accuracy and treatment selection. Davies et al followed 118 patients with histologically confirmed gastro-esophageal cancer. Patients with local and metastatic disease were assessed by the MDT. In the esophageal cancer subgroup, there was prevention of 9% under-treatment and 25% over-treatment due to the MDT process. ⁶⁷ In a UK study, 107 patients with esophageal cancer and undergoing R0 resections were either followed by a MDT or had surgery alone. Patients in the MDT arm had significantly lower operative mortality (5.7% vs 26%) and improved 5 year survival (52% vs 10%) compared to the surgery alone group. Multivariate analysis revealed that MDT management, lymph node metastases, and ASA grade were independently associated with duration of survival. ⁶⁹ Even in patients that required radiation therapy, discussion of the treatment plan through an MDT improved overall survival significantly (27 vs 17 months). ⁷⁰ In a recent nationwide analysis from Taiwan, 5906 patients with esophageal cancer were either treated through an MDT or did not have an MDT care plan. Patients who had a MDT had a significantly lower risk of death (HR = 0.73; 95% CI 0.67-0.79) with the greatest benefit in Stage 3 cancers (HR = 0.72; 95% CI 0.67-0.79 ). ⁷¹

Many MDTs are structured around regional cancer centers or hospitals, and often for invasive cancer. In the treatment of early esophageal dysplasia and neoplasia, our center has developed MDTs for early-stage disease (dysplasia, early esophageal cancer) as well as having a traditional MDT Tumor Group for invasive disease.

We have developed a Foregut team that consists of thoracic surgeons, interventional/oncologic gastroenterologists, motility gastroenterologists, minimally invasive foregut general surgeons, radiologists, and expert GI pathologists to discuss high-risk esophageal disease. Most patients with Barrett’s and early esophageal cancer are managed based on a provincial Clinical Practice Guideline. ⁷² However, at any stage of initial workup or endoscopic staging, patients can be discussed at this Foregut MDT. Further diagnostic testing, urgent consultations and alteration of the treatment plan are implemented after a consensus based, patient centered care approach. Patients may shift from medical, endoscopic, to surgical management based on team recommendations. This open model where opinions from all specialties are considered has reduced variance and provided patients with a unified MDT decision.

Patients found to have invasive esophageal cancer are brought forward to the GI Tumor Group Rounds. In addition to the Foregut team, medical and radiation oncologists as well as allied health support (dieticians, social services, psychologists, pain and palliative teams) are also present. A team consensus is obtained, and the plan is presented to the patient by one of the team members, eliminating unnecessary visits and expediting treatment. Treatment algorithms follow a provincial Clinical Guideline for invasive esophageal cancer. ⁷³

Conclusion

The management of high-risk esophageal lesions has increased in both complexity and the number of options available. After initial endoscopic therapy, residual disease may be identified as a high-risk lesion. The choice of initial therapy may influence subsequent therapy. Patients may also be identified as high-risk for esophagectomy. Critical factors such as age, performance status, frailty, malnutrition, and comorbidities may guide in the selection of appropriate patients for salvage esophagectomy. For those patients that are not fit for surgery, numerous endoscopic therapies can be considered. Further endoscopic resection, radiofrequency ablation, cryotherapy, and photodynamic therapy can be used in a stepwise or salvage approach to treat high risk lesions. Chemoradiation therapy after initial endoscopic therapy has been shown to be safe and provide survival benefit. Finally, the use of a multidisciplinary team should be considered at all stages of esophageal dysplasia or neoplasia treatment. While most cases may follow a prescribed algorithm, stage shifting or a change in the patient’s overall condition may prompt a change in the treatment plan. In summary, high-risk patients with residual high-risk esophageal lesions have many potential treatment options that are best facilitated by a multidisciplinary approach.