Keywords Endoscopy Upper GI Tract - Diagnosis and imaging (inc chromoendoscopy, NBI, iSCAN,
FICE, CLE) - Endoscopic resection (ESD, EMRc, ...) - Endoscopy Lower GI Tract - Endoscopic
resection (polypectomy, ESD, EMRc, ...)
Introduction
Anal squamous cell carcinoma (ASCC) is a rare disease. However, incidence is increasing
in Europe, Australia and the United States, with an annual incidence of 0.5 to 2.0
cases per 100,000 inhabitants [1 ].
Local excision (LE) is considered the treatment of choice when ASCC is diagnosed at
an early stage [2 ]
[3 ]. The main objective is to achieve an “en bloc” resection to facilitate adequate
histological analysis, decrease recurrence risk, and preserve sphincter functionality.
In this sense, it has been observed that piecemeal resection is associated with high
rates of local recurrence, making its use inadvisable [4 ]
[5 ].
Furthermore, chromoendoscopy and magnifying endoscopy enable clear visualization of
superficial microvascular structures to better assess depth of invasion and risk of
lymph node metastasis, similar to abnormal intrapapillary capillary loop (IPCL) patterns
for superficial esophageal SCC [6 ]
[7 ]. This advanced endoscopic assessment may contribute to better selection of candidates
for endoscopic resection, but its diagnostic yield in this location is unknown.
Endoscopic submucosal dissection (ESD) permits “en bloc” resection regardless of tumor
size but evidence about its use in ASCC is limited to case reports [6 ]
[7 ]
[8 ]
[9 ]
[10 ]
[11 ]
[12 ]
[13 ]
[14 ]. Therefore, we aimed to evaluate safety and long-term efficacy of ESD for ASCC.
Patients and methods
Study design
This retrospective study was promoted by the Mucosal Resection and Third Space Endoscopy
working group of the Spanish Society of Gastrointestinal Endoscopy. International
centers with prospective ESD databases were invited to participate. We included patients
who underwent ESD for ASCC at 11 tertiary referral centers in Europe (n = 8) and Japan
(n = 3) between November 2015 and August 2024.
The study was approved by the ethics committees for clinical research of the participating
centers (institutional review board code: HRYC-DSE-19). Written informed consent was
obtained from all patients before inclusion in the prospectively maintained ESD registries.
Additional study-specific informed consent was obtained when deemed necessary by local
regulations. The study is reported following the STROBE guidelines [15 ].
ESD procedure
ESD technique and materials were at the discretion of each endoscopist ([Fig. 1 ], [Fig. 2 ], [Video 1 ]). Decisions about ESD for LE and/or need for complementary treatment (surgery, radiotherapy
[RT] and/or chemotherapy) were discussed at the local multidisciplinary oncologic
committees of each institution. Pretreatment assessment studies and follow-up were
carried out following oncological clinical guidelines at each center.
Fig. 1 Squamous cell carcinoma of the anal canal. a White light (WL) image showing 20 × 12 mm flat elevated lesion over the dentate line.
b Blue light imaging (BLI) with a well-defined lesion border. c Magnified BLI images showing dilation, weaving, and elongation of IPCL-like micro
vessels. d Lugol stain-negative lesion with clearly defined lesion margins. e Incision in the anal canal with submucosal exposure. f Endoscopic submucosal dissection scar g Specimen resected “en bloc”. h, i Three months of follow-up (WL & BLI).
Fig. 2 Histological evaluation. a Hematoxylin and eosin staining (x5). b Immunohistochemical staining for Ki67 throughout mucosal thickness (x5). c High p16 staining showed a diffuse-positive pattern. throughout mucosal thickness
(x2).
Endoscopic submucosal dissection of squamous cell carcinoma of the anal canal using
conventional ESD technique.Video 1
Definitions and endpoints
Primary endpoints were the curative, “en bloc” and R0 resection rates. Curative resection
was defined as free margin resection (R0) and absence of high-risk histologic criteria
for lymph node metastasis (lymphovascular invasion, submucosal invasion, or poorly
differentiated tumor). Secondary endpoints were the adverse event (AE) rate, use of
chromoendoscopy and magnification for pretreatment assessment, and analysis of overall
and progression-free survival. Delayed bleeding was defined by at least one of the
following criteria: 1) rectal bleeding or hematochezia associated with clinical signs
of bleeding after 24 hours of the procedure; 2) a drop-in hemoglobin ≥ 2 g/dL related
to ESD (> 24 hours after the procedure); and 3) need for therapeutic intervention
such as endoscopic hemostasis, surgery, or blood transfusion due to bleeding (24 hours
after the procedure). AE severity was graded using the Adverse events in GastRointEstinal
Endoscopy (AGREE) classification [16 ].
Statistical analysis
Data analysis was performed using IBM SPSS Statistics, Version 20.0 (Chicago, Illinois,
United States). The Wilson method was used to calculate 95% confidence intervals (CIs).
Categorical variables were expressed as absolute frequencies and percentages. Continuous
variables were expressed as medians with ranges. Missing values are described in [Table 1 ] and [Table 2 ].
Table 1 Study population and patient characteristics.
Characteristic
Value
ASA, American Society of Anesthesiologists; HPV, human papillomavirus; MSM, men who
have sex with men; SCC, squamous cell carcinoma; SCC, squamous cell carcinoma; SD,
standard deviation; WSW, women who have sex with women.
Number of patients
23
Age (years), median (range)
65 (42–85)
Female, n (%)
17 (73.9)
Active smoker, n (%)
4 (17.4)
History of previous SCC in other location, n (%)
2 (8.7)
1 (4.3)
HIV, n (%)
2 (8.7)
HPV, n (%)
11 (47.8)
8 (34.8)
4 (17.4)
MSM/WSW, n (%)
12 (52.1)
3 (13)
8 (34.8)
ASA functional status, n (%)
11 (47.8)
7 (30.4)
5 (21.7)
Charlson Comorbidity Index, median (range)
2 (0–7)
2
Patient fit for radical surgery, n (%)
20 (87)
3 (13)
Anticoagulant therapy, n (%)
0 (0)
Antiplatelets, n (%)
19 (82.6)
3 (13.1)
1 (4.3)
Table 2 Lesion characteristics and endoscopic submucosal dissection procedure details.
Lesion characteristics
Value
ESD, endoscopic submucosal dissection; JES IPCL, Japan Esophageal Society Intrapapillary
Capillary Loop Classification.
Preprocedure biopsy findings, n (%)
8 (34.8)
4 (17.4)
8 (34.8)
3 (13)
JES IPCL Classification, n (%)
2 (8.7)
5 (21.7)
4 (17.4)
12 (52.1)
Morphology (Paris classification), n (%)
4 (17.4)
12 (52.1)
4 (17.4)
1 (4.3)
2 (8.7)
ESD procedure
Endoscopist previous ESD experience, n (%)
1 (4.3)
3 (13)
19 (82.6)
Sedation, n (%)
12 (52.2)
7 (30.4)
4 (17.4)
ESD strategy, n (%)
16 (69.6)
2 (8.7)
5 (21.7)
Traction, n (%)
22 (95.7)
1 (4.3)
Fibrosis, n (%)
8 (34.8)
10 (43.5)
1 (4.3)
4 (17.4)
Injection of local anesthetic, n (%)
17 (73.9)
6 (26.1)
Type of knife used, n (%)
19 (82.6)
1 (4.3)
1 (4.3)
2 (8.7)
Results
The study included a total of 23 patients (17 women [73.9%]; median age 65 years (range:
42–85) with ASCC. Eight of 19 (42.1%) were HPV-16-related lesions and only two (8.7%)
presented HIV coinfection. Two patients presented SCC HPV-related tumors in other
locations:
one had a history of laryngeal (cT2N0M0) and vulvar (pT1N0M0) SCC treated with RT
and LE,
respectively, and the other had a synchronous vulvar lesion (pT1N0M0) treated by LE.
Additional baseline characteristics are shown in [Table 1 ].
Pretreatment assessment
Pretreatment workup to rule out locoregional and/or metastatic disease was heterogeneous:
four of 23 patients (17.4%) underwent both thoraco-abdomino-pelvic computed tomography
(CT) scan and pelvic magnetic resonance imaging (MRI), eight (34.78%) had a CT scan
and two (8.7%) a MRI. No positron emission tomography (PET)-CT scans were performed
and only one patient from the first group above received an endoscopic ultrasound
examination.
High-quality endoscopic assessment (high definition and virtual chromoendoscopy) was
performed in all of them. Lugol stain was used in seven patients (30.4%) to clearly
delineate tumor margins. Optical magnification was available in 15 (65.2%) and the
Japan Esophageal Society (JES) intrapapillary capillary loop (IPCL) Classification
for SCC was used in 11 (47.8%) with a diagnostic accuracy to predict final histology
of seven of 11 (63.6%, 95% CI 35%-84%) (3/11 underestimated; 1/11 overestimated).
Previous biopsies were taken in 15 patients (65.2%), with final histology concordance
in 10 of 15 (66.7%, 95% CI 42%-85%) (5 of 15 underestimated the final histology).
Regarding anatomical ASCC location, 14 lesions (60.9%) were located in the anal canal
with rectal extension, seven (30.4%) in the anal canal extending to both the rectum
and anal margin, one (4.3%) was limited to the anal canal, and one (4.3%) was located
in the anal canal with extension to the anal margin.
ESD procedure
Twelve procedures (52.2%) were performed under general anesthesia. Initial circumferential
cutting was the most common strategy (69.6%) and the needle-type knives were the preferred
ones (82.6%) ([Table 2 ]). Six patients (26.1%) had topical anesthetic injection during ESD to prevent postoperative
anal pain. One ESD case was performed after recurrence of a previous surgical LE.
Resection outcomes
En bloc and R0 resection rates were 95.6% (22 of 23, 95% CI 79%-99%) and 78.3% (18
of 23, 95% CI 58%-90%), respectively. Curative resection rate was slightly lower (73.9%,
17 of 23, 95% CI 54%-88%). Reasons for non-curative ESD were: three carcinomas with
submucosal invasion, one positive vertical margin with lymphovascular invasion, and
two positive horizontal margin (1 with free margin < 1 mm and 1 affected with dysplasia).
Additional treatment was performed in three of six non-curative ESDs. Median follow-up
for these patients receiving adjuvant treatment was 17.9 months (range, 10.9–46.2).
Median interval between ESD and adjuvant treatment was 2.63 months (range, 2–4). Details
of non-curative resections and further management are provided in [Table 3 ].
Table 3 Summary of cases treated with endoscopic submucosal dissection.
Age
Gender
HPV/genotype
Paris
JES
Previous biopsy
Size (mm)
Final histology
Depth (micras)
LV invasion
Differentiation
Lateral/vertical margins
Additional treatment
Follow-up (months)
CRT, chemoradiotherapy; HGD, high-grade dysplasia; HPV, human papillomavirus; JES
IPCL, Japan Esophageal Society Intrapapillary Capillary Loop Classification; LGD,
low-grade dysplasia; LV, lymphovascular invasion; RT, radiotherapy.
*Endoscopic submucosal dissection for managing recurrence following surgical local
excision.
85
M
No
IIa+IIc
-
No
15
Invasive carcinoma
4150
Yes
G3
+/-
No
10.9
81
F
Unknown
IIa
B1
LGD
55
Invasive carcinoma
400
No
G1
+/+
CRT
46.2
74
F
No
Is
-
HGD
40
HGD
-
No
G1
+/-
No
36.1
73
F
Yes/-
IIa
A
No
65
HGD
-
No
-
-/-
No
11.2
72
F
Unknown
IIa
B1
Carcinoma
20
Intramucosal
-
No
G1
-/+
No
8
70
F
No
Is
-
No
50
Intramucosal
-
No
G1
+/-
No
70
70
F
Yes/16
IIb
-
HGD
25
HGD
-
No
-
-/-
No
53
69
F
Yes/16
Is
-
HGD
9
Invasive carcinoma
3000
No
G1
-/-
No
29
67
M
No
IIa
-
HGD
30
HGD
-
No
-
-/-
No
13
67
F
No
IIa
-
No
20
Invasive carcinoma
<200
No
G1
-/-
RT
11
65
F
No
IIa
-
No
15
HGD
-
No
G1
-/-
No
63
60
F
No
Is+IIb
B2
Carcinoma
30
Intramucosal
-
Yes
G1
-/-
No
7.2
59
M
Yes/16
IIb
-
LGD
20
HGD
-
No
-
-/-
No
4.6
59
F
Yes/16
IIa
B2
LGD
25
HGD
-
No
G1
-/-
No
7.4
56
F
Yes/-
IIb
-
HGD
20
HGD
-
No
-
-/-
No
2.8
56
M
Unknown
Is
B2
Carcinoma
12
Intramucosal
-
Yes
G1
+/+
CRT
24.7
52
F
Unknown
IIa+IIc
-
No
25
Invasive carcinoma
3000
Yes
G1
-/-
No
0
46
F
No
IIa
B1
LGD
20
LGD
-
No
G1
-/-
No
9.4
42
F
Yes/16
IIa
A
No
15
LGD
-
No
G1
-/-
No
1
62
M
Yes/-
IIa
B1
No
60
HGD
-
No
G1
-/-
No
2.5
63
M
No
IIa
B2
HGD
40
HGD
-
No
G1
+/-
No
1
72
F
No
IIa
-
HGD
20
Intramucosal
-
No
G1
-/-
No
1.3
43*
F
No
IIb
B1
HGD
20
HGD
-
No
G1
-/-
No
1.1
Adverse events
There were no clinically relevant bleeds (those not controlled with hemostatic forceps
or that led to suspension of the procedure, surgery, or radiological intervention).
Delayed bleeding occurred in four patients (17.4%, 95% CI 8%–28%): three had previous
intraprocedural bleeding, (two were managed conservatively, and one required endoscopic
hemostasis) while one experienced a de novo bleed 9 days after ESD in a patient without
antithrombotic agents, requiring readmission and endoscopic treatment. Postprocedural
anal pain occurred in nine patients (39.1%, 95% CI 22%–59%): six (26.1%) were managed
with non-opioids, two (8.7%) with minor, and one (4.3%) with major opioids. Only one
of the three patients who underwent postoperative RT experienced anal pain as a complication
which was also treated with oral analgesics. One patient reported fecal incontinence
(4.3%, 95% CI 1%–21%) during the perioperative period, which gradually resolved. One
other patient developed anal stenosis 2 months after ESD, which was successfully resolved
after a single session of digital plus bougie dilation and subsequent topical steroid
treatment. None of the patients receiving adjuvant therapy developed stenosis ([Table 4 ]).
Table 4 Outcomes and adverse events.
Outcomes
En bloc resection, n (%)
22 (95.6)
R0 resection achieved, n (%)
18 (78.3)
Specimen major axis (mm), median (range)
40 (15–75)
Specimen minor axis (mm), median (range)
24 (13–60)
Curative resection, n (%)
17 (73.9)
Procedure time (min), median (range)
62.5 (26–210)
Final histology, n (%)
2 (8.7)
11 (47.8)
5 (21.7)
5 (21.7)
Adverse events
Anal pain, n (%)
9 (39.1)
2 (9.5)
3 (14.3)
Delayed bleeding, n (%)
4 (17.4)
1 (4.3)
1 (4.3)
2 (8.7)
Fecal incontinence, n (%)
1 (4.3)
1 (4.3)
Stenosis, n (%)
1 (4.3)
1 (4.3)
Follow-up
Seventeen of 23 patients underwent follow-up endoscopy (73.9%); one of them was also
followed with a CT scan and pelvic MRI. Two of 23 (8.7%) underwent CT scan and MRI
exclusively, and one (4.3%) MRI and PET scan. All patients who underwent complementary
treatment after ESD were followed with radiological imaging tests. Median follow-up
was 10.1 months (range, 0–69.6). There were neither deaths nor tumor progression at
the end of follow-up.
Discussion
Results from this multicenter study demonstrate that ESD is an effective and safe
technique for treatment of superficial ASCC. Although endoscopic techniques have evolved
significantly in recent years, their roles in current management of superficial ASCC—both
for diagnosis and treatment—are not yet well-established, despite improvements in
safety and technology [2 ]
[3 ].
LE is an effective treatment for T1N0M0 and some selected T2N0M0 ASCC cases, with
recurrence rates similar to chemoradiotherapy (CRT) but with a lower rate of late
AEs (hematologic or gastrointestinal toxicity, dermatitis, proctitis, anal or vaginal
stenosis, incontinence, fecal urgency, or sexual dysfunction). Moreover, by choosing
LE as the first treatment option, RT and/or CRT is saved for hypothetical locoregional
recurrences and/or treatment of other pelvic malignancies [4 ]
[17 ]. However, because risk of locoregional recurrence after LE is still not well defined,
need for adjuvant treatment should be evaluated on a case-by-case basis, and close
follow-up should always be performed. Radical surgery in T1–2N0M0 is currently reserved
as salvage treatment after progression to RT/CRT or LE [3 ].
Despite good accessibility, sinuous morphology and nerve sensitivity, together with
high vascularization of the hemorrhoidal plexus can make ESD challenging in this location.
Nonetheless, en bloc, R0, and curative resection rates in our series were high at
95.7%, 78.3% and 73.9% respectively, supporting use of ESD as a first-line full-biopsy
resection treatment. No local recurrences were detected during surveillance after
ESD, which outperforms recurrence rates reported after surgical excision (9%–63%)
[4 ]
[18 ], and other less invasive topical (trichloroacetic acid, imiquimod, 5-fluorouracil)
and ablative techniques [19 ]. Furthermore, treating ASCC with ESD theoretically eliminates need for reintervention
and subsequent increased associated risk of stricture and/or fecal incontinence.
In consonance with what was previously reported, the majority of cases in our series
were women (80%), and were related to HPV-16, for which gynecological evaluation to
rule out concomitant lesions is highly recommended. Special attention should be paid
during routine colonoscopies in patients with other well-known associated risk factors
for ASCC, such as smoking, HIV, receptive anal sex, cervical or perineal cancer, or
immunosuppression [3 ]. Involvement of other areas such as the oropharyngeal region should be individually
assessed.
Pretreatment staging in our series was heterogeneous, mainly based on MRI and CT to
exclude locoregional and distant disease. Endoscopic ultrasound (EUS)-guided fine
needle aspiration was not performed in any patient because biopsy was not deemed necessary
to clarify the nature of suspicious lymphadenopathy, either prior to ESD or during
follow-up. Although use of EUS to assess tumor invasion depth remains controversial,
it may be considered in carefully selected cases. All patients underwent high-quality
(high-definition and chromoendoscopy) endoscopic assessment. Use of Lugol stain in
this series was infrequent (30.4%). However, when used either alone or in combination
with acetic acid, it may improve lesion delineation, aid in detection of small lesions
that are not visible to the naked eye, and enhance accuracy of targeted biopsies.
Although vital stains have demonstrated good sensitivity (80%–90%), false negatives
may still occur, especially in small lesions where iodine absorption is insufficient.
Virtual chromoendoscopy techniques offer slightly higher sensitivity (90%–95%) and
are particularly useful for identifying small, superficial, or anatomically challenging
lesions. For these reasons, both methods are regarded as complementary tools that
can further improve early detection of SCC throughout the gastrointestinal tract.
Accuracy of JES IPCL Classification for predicting final histology was moderate (63.6%),
closely matching the 66.7% accuracy obtained from prior biopsies. Although this classification
has not been formally validated for this anatomical location, by analogy with findings
in the esophagus, presence of aberrant microvessels—characterized by dilation, tortuosity,
caliber changes, and irregular shapes—may indicate increased risk of deep tumor invasion.
Interestingly, two cases were initially classified as JES type A, and no confirmatory
biopsies were performed. Although current ASCC clinical guidelines recommend performing
a diagnostic biopsy, eight of 23 patients (34.8%) did not undergo one prior to ESD
due to concerns that biopsy-related fibrosis in flat lesions might hinder successful
resection. In such instances, the ESD specimen was regarded as a high-quality biopsy
in itself. Given the lower rate of early SCC detection in Western countries, it is
crucial to assess endoscopist experience and proficiency with the JES classification
before relying on it for diagnostic purposes. These findings highlight the added value
of optical endoscopic diagnosis performed by trained endoscopists, which provides
complementary information in pretreatment staging and should always be considered.
AEs were medically managed with no protocol deviation. Most of the lesions had rectal
extension, which implied performing ESD through the hemorrhoidal plexus and increased
risk of bleeding [20 ]. However, delayed bleeding occurred in only four patients after ESD; two were managed
conservatively, whereas the other two required endoscopic intervention. Anal pain
occurred in 39.1% of patients, all of whom were successfully managed with perioperative
analgesia and the pain gradually disappeared. Post-ESD anal pain is a relevant clinical
concern influenced by several factors, including individual sensitivity, dissection
technique employed, and thermal damage to the muscularis propria. In this series,
only six patients (26.1%) received topical anesthetic injections for pain prevention;
among them, three (50%) still experienced postoperative pain, which was managed with
analgesics. Of the 17 patients (73.9%) who did not receive any form of preventive
treatment, six (35.3%) reported postoperative pain, all of whom also responded well
to analgesic therapy. These findings suggest that routine preventive analgesia may
not be necessary as a standard approach and should instead be tailored on a case-by-case
basis.
Regarding worrying risk of stricture or fecal incontinence, each occurred only temporarily
in one patient, including those who required complementary treatment. Although use
of steroids may help prevent rectal stenosis, their efficacy in this anatomical location
remains unclear. Therefore, their use may be considered selectively in high-risk cases.
Fecal incontinence is a potential complication that must be anticipated. Thorough
evaluation of patient medical history is essential, and in cases of uncertainty, anorectal
manometry should be considered to assess sphincter integrity and tone. Risk factors
such as history of instrumental vaginal delivery, previous anal surgery, advanced
age, neurological disorders, or diabetes mellitus may impair sphincter function and
increase risk of incontinence. In addition, procedure factors related to ESD—such
as involvement of the internal anal sphincter, resection of more than 50% of the anorectal
circumference, or extension to the dentate line—can further contribute to development
of fecal incontinence. Another possible contributing factor is reduced rectal compliance
due to fibrosis, which may lead to urgency-related incontinence. To minimize risk
of these complications, a multimodal preventive strategy is recommended. This may
include local steroid injections, limited or targeted dissections, a high level of
technical expertise in ESD, and use of non-aggressive electrosurgical settings. These
measures are essential to reduce tissue injury, reduce risk of fibrosis, and preserve
anorectal function.
The main strengths of our study are that it represents the largest ESD cohort of superficial
ASCC involving 11 referral centers and that the data are derived from prospectively
collected databases. However, we acknowledge that our study has limitations. First
is the limited sample size and its retrospective design. Second, the institutions
that participated have extensive experience in ESD, and therefore, results should
be interpreted with caution. Finally, the follow-up period was short and staging and
surveillance methods after ESD were heterogeneous, and therefore, long-term conclusions
cannot be drawn.
Conclusions
In conclusion, ESD seems an effective and safe resection technique for treatment of
superficial ASCC and should be considered as a first-line therapeutic option. In expert
hands, “en bloc” resection rates are high, reducing recurrences and risks associated
with reintervention. Furthermore, in cases of non-curative ESD, adjuvant treatments
can be performed maintaining sphincter functionality and quality of life of these
patients. Larger prospective comparative studies with longer follow-up are needed
to assess the role of ESD and to compare it with other modalities of treatment in
this clinical setting.
Bibliographical Record
