Article Text

Download PDFPDF

Long-term results and risk factor analysis for recurrence after curative endoscopic therapy in 349 patients with high-grade intraepithelial neoplasia and mucosal adenocarcinoma in Barrett’s oesophagus
  1. O Pech1,
  2. A Behrens1,
  3. A May1,
  4. L Nachbar1,
  5. L Gossner1,
  6. T Rabenstein1,
  7. H Manner1,
  8. E Guenter1,
  9. J Huijsmans1,
  10. M Vieth2,
  11. M Stolte2,
  12. C Ell1
  1. 1
    Department of Internal Medicine II, HSK Wiesbaden, Wiesbaden, Germany
  2. 2
    Institute of Pathology, Bayreuth Hospital, Bayreuth, Germany
  1. Professor C Ell, Department of Internal Medicine II, HSK Wiesbaden, Ludwig-Erhard-Strasse 100, 65199 Wiesbaden, Germany; ell.hsk-wiesbaden{at}


Objective: Endoscopic therapy is increasingly being used in the treatment of high-grade intraepithelial neoplasia (HGIN) and mucosal adenocarcinoma (BC) in patients with Barrett’s oesophagus. This report provides 5 year follow-up data from a large prospective study investigating the efficacy and safety of endoscopic treatment in these patients and analysing risk factors for recurrence.

Design: Prospective case series.

Setting: Academic tertiary care centre.

Patients: Between October 1996 and September 2002, 61 patients with HGIN and 288 with BC were included (173 with short-segment and 176 with long-segment Barrett’s oesophagus) from a total of 486 patients presenting with Barrett’s neoplasia. Patients with submucosal or more advanced cancer were excluded.

Interventions: Endoscopic therapy.

Main outcome measures: Rate of complete remission and recurrence rate, tumour-associated death.

Results: Endoscopic resection was performed in 279 patients, photodynamic therapy in 55, and both procedures in 13; two patients received argon plasma coagulation. The mean follow-up period was 63.6 (SD 23.1) months. Complete response (CR) was achieved in 337 patients (96.6%); surgery was necessary in 13 (3.7%) after endoscopic therapy failed. Metachronous lesions developed during the follow-up in 74 patients (21.5%); 56 died of concomitant disease, but none died of BC. The calculated 5 year survival rate was 84%. The risk factors most frequently associated with recurrence were piecemeal resection, long-segment Barrett’s oesophagus, no ablative therapy of Barrett’s oesophagus after CR, time until CR achieved >10 months and multifocal neoplasia.

Conclusions: This study showed that endoscopic therapy was highly effective and safe, with an excellent long-term survival rate. The risk factors identified may help stratify patients who are at risk for recurrence and those requiring more intensified follow-up.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

In recent years, increasing attention has focused on adenocarcinoma arising in Barrett’s oesophagus, mainly because of the dramatic increase in the incidence of the disease. Recent publications have reported that the incidence of cancer in Barrett’s oesophagus has risen approximately 6-fold in the United States, from 4 to 23 cases per million. Oesophageal adenocarcinoma is now even more frequent than oesophageal squamous cell cancer in Western countries.1

The development of carcinoma is causally linked to gastro-oesophageal reflux disease, probably passing through premalignant stages known as Barrett’s metaplasia, low-grade intraepithelial neoplasia, and finally high-grade intraepithelial neoplasia (HGIN).24 Due to physicians’ increasing awareness of this causal link, carcinoma in Barrett’s oesophagus is being identified more often at an early stage, particularly in patients who are receiving surveillance.5 6

The appropriate management of patients in whom early carcinoma or high-grade intraepithelial neoplasia has been detected continues to be a subject of controversy. Until now, radical oesophageal resection has been regarded as the treatment of choice, but because of the high morbidity and mortality rates associated with the procedure, less radical treatment strategies, including photodynamic therapy and endoscopic resection, have been advocated by many groups. Endoscopic treatment is considered to be curative when Barrett’s neoplasia is limited to the mucosa, since there is almost no risk of lymph-node metastasis in these patients.79 However, only limited data have been available until now regarding the results of endoscopic treatment for early neoplasia in Barrett’s oesophagus, particularly in relation to the long-term follow-up. For definitive conclusions to be drawn regarding the safety and efficacy of a new oncological treatment method, such as endoscopy, prospective studies including large groups of patients and a long-term follow-up period of at least 5 years are needed, and it is only when these are available that it will become possible to consider this new modality as an alternative method to the current standard treatment.

The present study was conducted in order to investigate the safety and efficacy of endoscopic treatment for early neoplasia in Barrett’s oesophagus and to provide long-term follow-up data on the outcomes of the treatment based on a large patient cohort.


Between October 1996 and September 2002, 486 patients presented at Wiesbaden Hospital with suspected intraepithelial neoplasia or early adenocarcinoma arising in Barrett’s oesophagus. All the patients underwent intensive staging, and endoscopic ultrasound, abdominal ultrasound and computed tomography of the chest were carried out in all patients with confirmed adenocarcinoma. Video endoscopy and chromoendoscopy with methylene blue staining and/or acetic acid staining were also carried out using Fujinon EG-450HR instruments (Fujinon Europe, Willich, Germany) or Olympus 130 and 140 instruments during the early phase of the study (Olympus, Hamburg, Germany). In addition to conventional endosonography with a radial scanner to assess lymph-node status, miniprobe endosonography with 20 MHz probes was also carried out in all patients with raised lesions, to assess the depth of infiltration.9 Biopsies were taken from all of the identified lesions, as well as four-quadrant biopsies every 1–2 cm over the entire Barrett’s segment. Assessment of the biopsies taken during the diagnostic procedures was usually carried out by at least two different experienced pathologists. The histological criteria, classification and assessment of the grade of differentiation corresponded to the World Health Organization classification.10

Patients were excluded if the staging examinations did not confirm the suspected diagnosis of Barrett’s carcinoma or HGIN, or if they showed a more advanced tumour stage (>T1), lymph-node involvement, or metastasis (n = 91). Submucosal cancer was diagnosed in 30 patients during the staging procedures by endoscopic ultrasound and histological confirmation by diagnostic endoscopic resection (ER). These patients received non-curative treatment, as they were unfit for surgery or declined it. In addition, histological evaluation of the resected specimen after endoscopic resection in 16 patients revealed submucosal cancer, and these patients were referred for surgery. All patients with submucosal adenocarcinoma were excluded from the study, due to the increased risk of lymph-node metastasis. A total of 349 patients with HGIN (n = 61) and mucosal Barrett’s adenocarcinoma (n = 288) were ultimately included (fig 1). Details of the characteristics of the patients included are presented in table 1.

Figure 1 Flow diagram for the patients in this study. BC, Barrett’s carcinoma; CR, complete response; ER, endoscopic resection; ET, endoscopic treatment; HGIN, high-grade intraepithelial neoplasia.
Table 1 Characteristics of the patients and tumours

During one or more information discussions, all of the patients received written and oral information indicating that oesophageal resection is the current “gold standard” and that endoscopic therapy has to be carried out under research conditions.

Endoscopic therapy

Treatment of neoplastic lesions

Patients with localised neoplasia were offered endoscopic resection. Patients with lesions that were not clearly localised, those with superficial subtle multifocal neoplasia, and patients who had no neoplasia found in oesophageal biopsies taken at our institution but who had verified oesophageal neoplasia in oesophageal biopsy specimens taken at the referring institutions (n = 10) were treated using photodynamic therapy with 5-aminolevulinic acid (5-ALA).

Endoscopic resection with the “suck-and-cut” technique, using a ligation device or cap, was used.11 The ligation devices used were the Speedband (Boston Scientific, La Garenne-Colombes, France) and the reusable EuroLigator (WMT, Wiesbaden, Germany). Procession of the resected specimens and histopathological assessments were performed as described previously by highly experienced pathologists (MS and MV).12

For photodynamic therapy (PDT), 5-ALA was used as a photosensitiser. The light delivered had a wavelength of 635 nm, an energy dosage of 150 J/cm2 and a power density of 100 mW/cm2. PDT was performed as described previously.13 Patients were discharged 2 days after treatment.

Complete local remission was defined as an R0 resection plus one normal endoscopic follow-up examination. In R1 or Rx situations (Rx meaning evaluation of the margin is not possible due to coagulation artifacts) on the lateral margin of the resected specimen, two consecutive endoscopic check-up examinations without evidence of residual tumour were required in order to conclude that there was a complete response. In R1 or Rx situations at the base of the resected specimen, patients were defined as having treatment failure, and operable patients were scheduled for radical oesophagectomy.

Ablation of remaining non-neoplastic Barrett’s epithelium

If a complete response of neoplastic changes (HGIN and mucosal cancer) was achieved, ablation of the remaining non-neoplastic part of the Barrett’s epithelium was not generally performed in all patients; during the last 3 years, however, ablative treatment was performed in most of the patients who agreed to it. We have changed our concept and performed elective ablation of the remaining Barrett’s epithelium that did not contain any neoplastic changes as proven by at least two negative follow-up endoscopies. The reason for this was the high rate of metachronous neoplasias detected during follow-up, as shown in a previous analysis.14

Elective ablation was carried out using argon plasma coagulation (APC) in patients with short segments of remaining Barrett’s oesophagus, and using PDT in those with longer (>3 cm) and circumferential segments Barrett’s oesophagus in combination with APC of small remnants of Barrett’s mucosa after PDT.

The primary end points of this study were recurrence and tumour-associated death. The initial results, short-term follow-up, and medium-term follow-up for about half of the patients in the study group were previously described in our preliminary reports on endoscopic resection and PDT for early Barrett’s cancer.1215

Treatment with proton-pump inhibitors

All patients were placed on 40 mg omeprazole or pantoprazole before endoscopic treatment. After treatment, all of the patients received omeprazole or pantoprazole intravenously to begin with for 2 days and then 2×40 mg orally for at least 10 days, after which the dose of proton-pump inhibitor (PPI) was reduced to 40 mg/day. Long-term PPI treatment was carried out in accordance with the 24 h pH-metry results, performed 6–8 weeks after the first treatment.

Follow-up programme

All of the patients were included in a strict follow-up programme monitored in collaboration with the referring external gastroenterologists or hospitals. Follow-up examinations were planned 1, 2, 3, 6, 9 and 12 months after treatment and then at 6 month intervals up to the end of a 5 year period after treatment. Annual check-ups were planned thereafter. The check-ups included endoscopy with high-resolution endoscopes and biopsies of any suspicious lesions, as well as four-quadrant biopsies and/or chromoendoscopy of residual Barrett’s mucosa. In addition, every second check-up included endosonography, computed tomography and abdominal ultrasonography.

Metachronous lesions were defined as high-grade intraepithelial neoplasia or early cancer detected after two negative follow-up high-resolution (HR) endoscopies with chromoendoscopy and four-quadrant biopsies performed at our department during the follow-up after the patient had achieved a complete response. The term “metachronous lesion” included local neoplastic recurrences or possible remnants at the margin of the prior resection zone, synchronous lesions not detected during the initial staging, and new lesions that developed during the follow-up in previously non-neoplastic residual Barrett’s tissue.


Statistical analyses of the patients’ data and clinical parameters are given as means with the standard deviation, or as medians and 25% and 75% percentile. Kaplan–Meier estimates of the survival curves were calculated for time to recurrence. Patients who did not complete the maximum follow-up period, or who were lost to follow-up, were censored at the last follow-up date (figs 2 and 3). To analyse risk factors for recurrence, multivariate analysis, proportional hazards tests, and diagnosis based on weighted residuals were used.

Figure 2 Kaplan–Meier plot for estimating the tumour freedom in 349 patients with high-grade intraepithelial neoplasia and early adenocarcinoma in Barrett’s oesophagus treated with endoscopic therapy.
Figure 3 Kaplan–Meier plot for estimating the overall survival compared to the average German population with matched gender and age (dashed line) (p = not significant).


Endoscopic therapy was initially performed with curative intent in 365 patients, but histopathological assessment of the resected specimens revealed submucosal carcinoma in 16 after diagnostic endoscopic resection. These 16 patients were referred for surgery. The data relating to the remaining 349 patients were analysed (96%). Data for the 349 patients included are listed in table 2.

Table 2 Acute and long-term results

Endoscopic resection was performed in 279 patients, with a total of 734 resections (2.1 per patient; median 1, range 1–16); the procedure had to be performed using a piecemeal technique (R1 at the lateral margin) in 100 patients (35.8%) due to neoplastic lesions with a diameter of more than 2 cm. In 184 patients, ER was performed during the initial endoscopies and in the remaining 95 patients ER was performed also during repeat endoscopies. PDT with 5-ALA was used in 55 patients who had widespread areas of superficial neoplasia, or when proven neoplasia could not be redetected (n = 10). The two procedures were combined in 13 patients (PDT was used to ablate remaining areas of HGIN), and two patients with very small areas of neoplasia underwent primary treatment with APC in the first phase of the study period.

The overall complication rate was 17.2% (60 of 349). Major complications occurred in two patients (0.6%), involving spurting bleeding with a fall in the haemoglobin level of more than 2 g/dl. Minor complications consisted of bleeding not affecting the haemoglobin level (n = 40), stenosis (n = 15), long-lasting odynophagia (n = 1), and sunburn after PDT (n = 2). All of the complications were managed conservatively. Eighty-seven per cent of patients who developed a stenosis received more than two ERs.

A complete response was achieved in 337 patients (96.6%). Surgery was performed in 13 patients (3.7%) after failure of endoscopic therapy. Reasons for failure were technical problems performing ER due to scars after prior ER (n = 2), unexpected submucosal cancer after several ER (n = 2) and difficulties in achieving complete remissions after several treatment attempts mainly due to poor healing despite high-dose PPI (n = 9). During the median follow-up period of 63 months (25% percentile: 49.5; 75% percentile: 80.0) up to September 2006, metachronous lesions were observed in 74 patients (21.5%). Metachronous lesions were detected after a median of 15 months (25% percentile: 12.0; 75% percentile: 24.0). As a result of the close follow-up schedule, which was arranged in collaboration with the referring physicians, only 13 patients dropped out of the study after a median follow-up of 62 months.

The rate of apparent long-term eradication achieved can be calculated as 94.5% (330 of 349 patients) (fig 1). The remaining patients did not develop any suspicious lymph nodes on follow-up endoscopic ultrasonography or computed tomography. Details of the results of treatment are summarised in table 2.

Fifty-eight patients have so far died of concomitant disease, but none of the patients died due to Barrett’s neoplasia. The overall 5 year survival rate in all patients was calculated as 84%. The recurrence-free rate at 5 years was 77% (fig 2). There were no significant differences in survival and recurrence between patients with different tumour stages (HGIN, T1m1, T1m2, T1m3, T1m4), macroscopic types, and tumour grading (G1, G2, G3). The overall survival did not differ significantly from the survival of the average German population with the same age and gender distribution (fig 3).

Eighty-five per cent of the patients (63 of 74) who developed metachronous neoplasia received a second treatment and achieved a second complete remission. Three patients were referred to surgery because a complete response was not achieved with endoscopic therapy. Two patients were still receiving treatment at the time of analysis, and two patients died of other causes before a complete remission could be achieved again.

Patients with mucosal cancer treated by endoscopic resection

When excluding those patients treated by PDT and those with HGIN 231 patients (mean age 64.1 years) with mucosal BC treated by only ER are remaining. One hundred and sixty-two patients had a well-differentiated carcinoma, 62 patients a moderately differentiated carcinoma and seven patients a poorly differentiated carcinoma. Complete response could be achieved in 225 patients (97.4%) after a median of 3 months (25% percentile: 1; 75% percentile: 7). After a median follow-up of 61 months (25% percentile: 49.8; 75% percentile: 77.0), 201 patients (87%) are still alive and 30 (13%) died from other causes. Metachronous neoplasia was found in 49 patients (21.2%) but after endoscopic re-treatment, 221 patients (95.7%) achieved long-term complete response.

Ablative therapy of remaining non-neoplastic Barrett’s mucosa

After complete eradication of neoplasia, 200 patients underwent ablative therapy of the remaining non-neoplastic Barrett’s epithelium to prevent development of metachronous neoplasia. One hundred and thirty-six patients received APC and 64 patients underwent PDT as ablative treatment. In 60 of 64 patients treated by PDT, APC was added to ablate small remnants of Barrett’s mucosa. At time of analysis, 117 patients (86%) achieved complete removal of Barrett’s epithelium after a mean of 3.1 sessions (range 1–12) and 19 patients (14%) were still under treatment. In the ablation group, in 33 patients (16.5%) a metachronous neoplasia was detected. In those patients receiving no ablation (n = 137), metachronous neoplasia (n = 41; 29.9%) was significantly more often found during follow-up endoscopies (p = 0.0014). Especially late occurrence of metachronous neoplasia (⩾24 months after complete remission from neoplasia) was present significantly more often in the group without ablation than in those patients who received ablation of non-neoplastic remaining Barrett’s mucosa (13/41 (32%) vs 6/33 (18%) patients; p = 0.0053). Metachronous neoplasia in the ablation group was only found in those patients with residual Barrett’s mucosa. Metachronous lesions were not found in any of the patients who achieved a complete ablation of the Barrett’s epithelium.

Risk factor analysis

The results of the analysis of risk factors associated with recurrence after endoscopic therapy are shown in table 3.

Table 3 Multivariate analysis of factors potentially associated with recurrence after endoscopic resection of early Barrett’s neoplasia

The multivariate analysis showed that long-segment Barrett’s oesophagus, multifocal neoplasia, piecemeal resection, no ablation of Barrett’s epithelium after complete response, and a period of more than 10 months until achieving a complete response were independent risk factors associated with recurrence. The analysis did not reach significance for other variables (local tumour stage, macroscopic type, lesion size ⩽2 cm or >2 cm, treatment modality (PDT or ER)).


Radical oesophageal resection has until now been regarded as the treatment of choice in patients with HGIN and mucosal adenocarcinoma arising in Barrett’s metaplasia, due to the lack of long-term results with endoscopic therapy. Recent publications by other authors and our group only reported short- or mid-term data with a maximum mean follow-up of 3 years in selected and significantly smaller patient cohorts. The present study is the first to report long-term data in a large and unselected group of patients with Barrett’s neoplasia including a large group of patients with mucosal Barrett’s cancer treated by ER and/or PDT. Although this was not a randomised trial comparing surgery with endoscopic therapy, the excellent results make it possible to conclude that endoscopic therapy for HGIN and mucosal Barrett’s carcinoma is a safe and effective form of treatment, with long-term results that are not inferior to those with surgical options. Radical oesophageal resection is known to be associated with a mortality rate of 3–5% in experienced centres, which may increase to 20.3% in less-experienced centres.16 17 In addition, surgery is associated with a morbidity rate of 40–50%.18 19 In contrast, the mortality rate with endoscopic therapy in the present series is zero, and the rate of relevant complications usually does not exceed 1–3%.20

Randomised trials comparing endoscopic resection with surgery would be desirable, but there are two major reasons why this type of study would be almost impossible to conduct. First, it would be hard to find patients willing to agree to randomisation, particularly in view of the excellent long-term results with endoscopic treatment. Second, to obtain significant results with regard to the key parameter of tumour-related survival, more than 200 patients would have to be included in each group.

Recurrences or metachronous neoplasia have been shown to be the major problem with endoscopic therapy in early Barrett’s neoplasia, although successful repeat endoscopic treatment is possible in almost all patients.20 Recent studies have shown that this risk ranges between 3 and 19% and can increase to more than 30% after a follow-up period of 3 years.13 21 22 In the present study, the rate of metachronous lesions found after more than 5 years of follow-up was 21.5%, but only three of the patients in whom a metachronous lesion was found (4%) had to be referred for surgery, and none of the patients who received endoscopic treatment died of Barrett’s neoplasia. The apparent long-term eradication rate was 94.5%. The reasons for the high rate of recurrence appear to be to a small percentage of undetected neoplasias in the residual Barrett’s segment after treatment. Recurrences in our study were detected after a median of 15 months after the last treatment. After this time, patients usually had at least three control endoscopies at our institute without detection of residual neoplastic Barrett’s mucosa, suggesting that the majority of those malignant lesions found during follow-up are de novo neoplasias rather than residuals of the initially treated lesions. Those neoplasias usually develop due to genetic abnormalities within the remaining Barrett’s mucosa, not influenced by the PDT or focal ER. Those findings again underline the importance of a close follow-up with HR endoscopy and chromoendoscopy by experienced endoscopists, especially in those patients with one or more of the identified risk factors for recurrence and especially during the first 2 years after successful treatment. Those patients with long-segment Barrett’s oesophagus who had piece-meal resection of the neoplasia and/or multifocal neoplasia were the patients with the highest risk for developing metachronous neoplasia. A close follow-up schedule seems to be important to detect newly developed neoplasia at an early stage in order to remove those lesions endoscopically, again. In high-risk patients we would suggest follow-up endoscopies every 3 months in this subgroup during the first 2 years and after that twice annually, as performed in our study. In contrast, follow-up endoscopies in patients with short-segment Barrett’s esophagus with unifocal lesions treated with only one ER could be performed in larger intervals (eg, every 6 months within the first year and after that annually).

Figure 4 Kaplan–Meier plot for recurrence after complete local remission in patients who underwent ablative therapy for the remaining non-neoplastic Barrett’s oesophagus (solid line) in comparison with those without ablative therapy (dashed line) (p = 0.0003; odds ratio, 0.26–0.66).

An interesting finding of our analysis was that ablation of the remaining non-neoplastic Barrett’s epithelium after complete eradication of HGIN and mucosal cancer seems to be capable of significantly reducing the risk for the development of metachronous neoplasia. In those patients who received ablative therapy, only 16.5% developed neoplasia during follow-up compared to 28.3% in the group of patients without ablation. Why was the rate of metachronous neoplasia not zero in the ablation group? One explanation would be that ablative therapy was performed within the follow-up intervals (every 6–12 months) and complete ablation was not usually the aim during one treatment session. All neoplasias in the ablation group were observed within the ablation period and none of those patients achieving complete remission of Barrett’s mucosa have developed neoplasia during follow-up. These retrospective data support the theory that complete removal of the Barrett’s mucosa at risk after successful treatment of neoplasia should be the aim of endoscopic therapy. Well-established ablation techniques are APC and PDT. However, with both methods complete ablation could only be achieved in 60–90% and are associated with a relevant complication rate with stricture formation or perforation.2325 A further problem of those ablation techniques are small remnants of Barrett’s epithelium under the newly developed squamous epithelium (so-called “buried glands”). After APC and PDT those “buried glands” are found in about 20% of patients.24 25

A recently developed and promising new method to remove the Barrett’s mucosa at risk seems to be radio-frequency ablation. First series from the USA and Amsterdam are demonstrating the safety and efficacy of this system and further studies are ongoing to test its ability to eliminate neoplasia and to reduce the rate of neoplastic recurrence.2628

Several attempts have therefore been made to reduce the rate of recurrent malignancy after successful treatment. Circumferential endoscopic resection to eradicate the entire Barrett’s mucosa at risk was reported previously.29 Endoscopic therapy was carried out in 12 patients with HGIN or mucosal carcinoma. The complete Barrett’s segment was resected in one to five sessions, with a median of five endoscopic resections per session (range, 1–19). Complications occurred in six cases (four cases of bleeding and two strictures), all of which were managed endoscopically. The median follow-up in this small series was 9 months, and no recurrences were observed. Other groups have also followed the approach of using circumferential endoscopic resection to eradicate the entire area of Barrett’s metaplasia, and the short-term results have been promising.30 31 A complete eradication of neoplasia was achieved in 86–100% of patients, but malignancy recurred in up to 11%, however, in 0% in the series reported by the Amsterdam group.30 One problem with this method is the significant rate of strictures, which could occur in up to 70% of cases,32 and the fact that the goal of resecting the entire area of metaplastic epithelium is usually only achieved in about 75–89% of cases.30 31 As in all patients treated endoscopically, further surveillance endoscopies are also mandatory in these patients. Whether this approach is capable of significantly reducing the risk of recurrence needs to be investigated in further studies with longer follow-up periods and more patients. However, the high rate of relevant strictures occurring following use of the piecemeal resection technique is an argument against this strategy. En bloc resection using endoscopic submucosal dissection (ESD) would be another option, but no data are currently available concerning this technique in Barrett’s neoplasia.

The present study was able to demonstrate that metachronous neoplasia is strongly associated with the presence of long-segment Barrett’s oesophagus, multifocal neoplasia, the need for more than one endoscopic resection to resect a neoplastic lesion (piecemeal resection), and a longer period needed to achieve complete remission. The risk of recurrence was significantly reduced in patients who underwent ablative therapy of the remaining non-neoplastic Barrett’s epithelium with APC or PDT. Although this study was not designed to investigate the effect of Barrett’s ablation after successful eradication of neoplasia, the multivariate analysis reached significance and supports the approach of eradicating the remaining area of Barrett’s metaplasia at risk after successful endoscopic therapy once neoplasia has been found. A prospective randomised trial to investigate this issue is ongoing at our institute.

In the present study, the majority of patients were treated with endoscopic resection. PDT was only used in 16% of cases, mainly in patients with larger areas of superficial neoplasia. The advantages of endoscopic resection are obvious: it allows histological assessment of the resected specimen in order to assess the depth of infiltration of the tumour and freedom from neoplasia at the lateral and (more importantly) basal margins, imitating the surgical situation. These significant advantages of ER are the main reason why it should be preferred to ablative treatment methods, even PDT, whenever possible, particularly in view of the low accuracy of endoscopic ultrasound for local tumour staging.9 3336 Nevertheless, PDT has proved to be an effective form of treatment for neoplastic Barrett’s epithelium in several studies. In a study by our own group, 66 patients were treated with PDT using 5-ALA as a photosensitiser.14 A complete response was achieved in 98.5% of the patients, and recurrences developed in 17% during a median follow-up period of 37 months. No major side effects of PDT were observed. A recently published, prospective, randomised, multicentre study compared medical treatment with omeprazole and PDT with porfimer sodium (Photofrin) in patients with HGIN.37 Complete response was achieved in 77% of the patients in the PDT group and 39% of those in the omeprazole group. Adverse effects of PDT were observed in 94% of the patients, and strictures occurred in 36%. A complication rate as high as this does not appear to be acceptable, particularly in view of the limited effectiveness of PDT in comparison with PPI treatment (with a 38% difference). However, the Mayo group recently reported their results of a retrospective study comparing photofrin–PDT with those patients who were referred to surgery.38 The mean follow-up in both groups was 5 years. The overall mortality and long-term survival were comparable in both groups suggesting that PDT is a safe and effective treatment modality. Prospective randomised trials are needed in order to allow definitive conclusions to be reached regarding whether endoscopic resection is preferable to PDT.

Another important finding of the present study was that there is a significantly higher risk of metachronous neoplasia in patients who undergo piecemeal resection. This observation was already known from endoscopic resection of early gastric cancer, as a result of which ESD was introduced in the treatment of early gastric cancer in order to allow en bloc resection even in lesions larger than 20 mm in diameter. However, there is as yet a lack of prospective comparative studies on this issue, particularly with regard to the recurrence rate.3942 There is at present almost no experience with the use of ESD in patients with Barrett’s carcinoma.43

In conclusion, this is the first study reporting the long-term follow-up of more than 5 years after endoscopic therapy in patients with HGIN and mucosal adenocarcinoma in Barrett’s oesophagus in a large group of patients treated by ER and/or PDT. The study demonstrates that endoscopic therapy is highly effective and safe and that the long-term results are excellent. It was also possible to analyse risk factors associated with recurrence, providing important data for risk stratification and potential improvement of endoscopic therapy and surveillance in these patients.



  • Competing interests: None.

  • Ethics approval: This study was approved by the Ethics Commission of the Medical Council of the State of Hesse, Germany, in 1996 and was carried out in accordance with the criteria of good clinical practice.

Linked Articles

  • Digest
    Robin Spiller Magnus Simren