Non‐surgical interventions for late rectal problems (proctopathy) of radiotherapy in people who have received radiotherapy to the pelvis

Fleur T van de Wetering; Leen Verleye; H Jervoise N Andreyev; Jane Maher; Joan Vlayen; Bradley R Pieters; Geertjan van Tienhoven; Rob JPM Scholten

doi:10.1002/14651858.CD003455.pub2

. 2016 Apr 25;2016(4):CD003455. doi: 10.1002/14651858.CD003455.pub2

Non‐surgical interventions for late rectal problems (proctopathy) of radiotherapy in people who have received radiotherapy to the pelvis

Fleur T van de Wetering ^1,^✉, Leen Verleye ², H Jervoise N Andreyev ³, Jane Maher ⁴, Joan Vlayen ², Bradley R Pieters ⁵, Geertjan van Tienhoven ⁶, Rob JPM Scholten ¹

Editor: Cochrane Gynaecological, Neuro‐oncology and Orphan Cancer Group

PMCID: PMC7173735 PMID: 27111831

Abstract

Background

This is an update of a Cochrane review first published in 2002, and previously updated in 2007. Late radiation rectal problems (proctopathy) include bleeding, pain, faecal urgency, and incontinence and may develop after pelvic radiotherapy treatment for cancer.

Objectives

To assess the effectiveness and safety of non‐surgical interventions for managing late radiation proctopathy.

Search methods

We searched the Cochrane Central Register of Controlled Trials (CENTRAL) (Issue 11, 2015); MEDLINE (Ovid); EMBASE (Ovid); CANCERCD; Science Citation Index; and CINAHL from inception to November 2015.

Selection criteria

We included randomised controlled trials (RCTs) comparing non‐surgical interventions for the management of late radiation proctopathy in people with cancer who have undergone pelvic radiotherapy for cancer. Primary outcomes considered were: episodes of bowel activity, bleeding, pain, tenesmus, urgency, and sphincter dysfunction.

Data collection and analysis

Study selection, 'Risk of bias' assessment, and data extraction were performed in duplicate, and any disagreements were resolved by involving a third review author.

Main results

We identified 1221 unique references and 16 studies including 993 participants that met our inclusion criteria. One study found through the last update was moved to the 'Studies awaiting classification' section. We did not pool outcomes for a meta‐analysis due to variation in study characteristics and endpoints across included studies.

Since radiation proctopathy is a condition with various symptoms or combinations of symptoms, the studies were heterogeneous in their intended effect. Some studies investigated treatments targeted at bleeding only (group 1), some investigated treatments targeted at a combination of anorectal symptoms, but not a single treatment (group 2). The third group focused on the treatment of the collection of symptoms referred to as pelvic radiation disease. In order to enable some comparison of this heterogeneous collection of studies, we describe the effects in these three groups separately.

Nine studies assessed treatments for rectal bleeding and were unclear or at high risk of bias. The only treatments that made a significant difference on primary outcomes were argon plasma coagulation (APC) followed by oral sucralfate versus APC with placebo (endoscopic score 6 to 9 in favour of APC with placebo, risk ratio (RR) 2.26, 95% confidence interval (CI) 1.12 to 4.55; 1 study, 122 participants, low‐ to moderate‐quality evidence); formalin dab treatment (4%) versus sucralfate steroid retention enema (symptom score after treatment graded by the Radiation Proctopathy System Assessments Scale (RPSAS) and sigmoidoscopic score in favour of formalin (P = 0.001, effect not quantified, 1 study, 102 participants, very low‐ to low‐quality evidence), and colonic irrigation plus ciprofloxacin and metronidazole versus formalin application (4%) (bleeding (P = 0.007, effect not quantified), urgency (P = 0.0004, effect not quantified), and diarrhoea (P = 0.007, effect not quantified) in favour of colonic irrigation (1 study, 50 participants, low‐quality evidence).

Three studies, of unclear and high risk of bias, assessed treatments targeted at something very localised but not a single pathology. We identified no significant differences on our primary outcomes. We graded all studies as very low‐quality evidence due to unclear risk of bias and very serious imprecision.

Four studies, of unclear and high risk of bias, assessed treatments targeted at more than one symptom yet confined to the anorectal region. Studies that demonstrated an effect on symptoms included: gastroenterologist‐led algorithm‐based treatment versus usual care (detailed self help booklet) (significant difference in favour of gastroenterologist‐led algorithm‐based treatment on change in Inflammatory Bowel Disease Questionnaire–Bowel (IBDQ‐B) score at six months, mean difference (MD) 5.47, 95% CI 1.14 to 9.81) and nurse‐led algorithm‐based treatment versus usual care (significant difference in favour of the nurse‐led algorithm‐based treatment on change in IBDQ‐B score at six months, MD 4.12, 95% CI 0.04 to 8.19) (1 study, 218 participants, low‐quality evidence); hyperbaric oxygen therapy (at 2.0 atmospheres absolute) versus placebo (improvement of Subjective, Objective, Management, Analytic ‐ Late Effects of Normal Tissue (SOMA‐LENT) score in favour of hyperbaric oxygen therapy (HBOT), P = 0.0019) (1 study, 150 participants, moderate‐quality evidence, retinol palmitate versus placebo (improvement in RPSAS in favour of retinol palmitate, P = 0.01) (1 study, 19 participants, low‐quality evidence) and integrated Chinese traditional plus Western medicine versus Western medicine (grade 0 to 1 radio‐proctopathy after treatment in favour of integrated Chinese traditional medicine, RR 2.55, 95% CI 1.30 to 5.02) (1 study, 58 participants, low‐quality evidence).

The level of evidence for the majority of outcomes was downgraded using GRADE to low or very low, mainly due to imprecision and study limitations.

Authors' conclusions

Although some interventions for late radiation proctopathy look promising (including rectal sucralfate, metronidazole added to an anti‐inflammatory regimen, and hyperbaric oxygen therapy), single small studies provide limited evidence. Furthermore, outcomes important to people with cancer, including quality of life (QoL) and long‐term effects, were not well recorded. The episodic and variable nature of late radiation proctopathy requires large multi‐centre placebo‐controlled trials (RCTs) to establish whether treatments are effective. Future studies should address the possibility of associated injury to other gastro‐intestinal, urinary, or sexual organs, known as pelvic radiation disease. The interventions, as well as the outcome parameters, should be broader and include those important to people with cancer, such as QoL evaluations.

Keywords: Humans, Anti‐Inflammatory Agents, Anti‐Inflammatory Agents/therapeutic use, Electrocoagulation, Electrocoagulation/methods, Fatty Acids, Fatty Acids/therapeutic use, Formaldehyde, Formaldehyde/therapeutic use, Hyperbaric Oxygenation, Pelvic Neoplasms, Pelvic Neoplasms/radiotherapy, Proctitis, Proctitis/therapy, Radiation Injuries, Radiation Injuries/therapy, Randomized Controlled Trials as Topic, Rectum, Rectum/radiation effects, Sucralfate, Sucralfate/therapeutic use

Plain language summary

Non‐surgical interventions for late rectal consequences of radiotherapy in people who have received radical radiotherapy to the pelvis

Background

Radiotherapy is often used to treat cancer in the pelvic area. Several organs in the pelvis, such as the anus, rectum, bladder, prostate, gynaecological organs (womb, ovaries, cervix, and vagina), small bowel, and pelvic bones may be exposed to the effects of radiotherapy, which can lead to pelvic radiation disease. Symptoms from pelvic radiation disease may occur around the time of treatment (early effects) or over a period of time, often many years after treatment (late effects) due to long‐term changes secondary to scarring (fibrosis), narrowing (stenosis), and bleeding due to new blood vessel formation (telangiectasia). Damage to the rectum (radiation proctopathy) is the most often investigated late radiation effect to the pelvis, which affects a small but but still important group of people who undergo pelvic radiotherapy. The common symptoms are rectal urgency, rectal incontinence, pain, mucus discharge, and rectal bleeding.

The aim of the review

The aim of this review was to assess the effect of non‐surgical treatments on late rectal damage.

Main Findings

We found 16 (quasi) randomised controlled trials (RCTs) including 993 participants that assessed non‐surgical treatments for radiation proctopathy. Although some treatments look promising (including rectal sucralfate, adding metronidazole to an anti‐inflammatory regimen, and hyperbaric oxygen therapy), the quality of evidence was low to very low. Furthermore, outcomes important to people with cancer, including quality of life (QoL), and long‐term effects were often not addressed in these studies.

Conclusions

Although some interventions for late radiation rectal damage are promising, the evidence was of low quality and we can draw no firm conclusions. We could not combine data from the studies to compare different treatments, since the trial designs and outcome measures differed. The episodic and variable nature of late radiation rectal damage requires larger RCTs to establish whether treatments are effective. Future studies should address the possibility of associated injury to other pelvic structures, collectively known as pelvic radiation disease. Ideally outcome measures should be standardised across studies and include QoL evaluations and other outcomes important to people with cancer .

Quality of the evidence

The quality of the evidence for the majority of outcomes was low or very low, mainly due to the small size of most studies and study limitations.

Background

This review is an update of the previously published Cochrane review 'Non‐surgical interventions for late radiation proctopathy in people who have received radical radiotherapy to the pelvis' (Denton 2002), which was last updated in 2007.

Description of the condition

Radiotherapy is often used to treat cancer in the pelvic area. It can be used as single therapy or in combination with chemotherapy, as primary treatment or before or after surgery. Examples of cancer in the pelvis that can be treated with radiotherapy include prostate, bladder, cervix, endometrial, vaginal, rectal, and anal cancer. One drawback of radiotherapy is the development of late radiation toxicity. Several organs in the pelvis, including the anus, rectum, bladder, prostate, gynaecological organs, pelvic bones, and sometimes small bowel may be exposed to the effects of radiotherapy, which can lead to a variety of symptoms. The term 'pelvic radiation disease' has been proposed, defined as “transient or long term problems, ranging from mild to severe, arising in non‐cancerous tissues resulting from radiotherapy treatment to a tumour of pelvic origin” (Andreyev 2010; Denham 2002). If symptoms are confined to the rectum or the anorectal complex, we tend to speak of radiation‐associated proctopathy, or radiation‐induced proctopathy, formerly referred to as late radiation proctopathy. The current review, which is an update of a Cochrane review first published in 2002 and subsequently assessed as up to date in 2008, about the treatment of late radiation proctopathy is mainly focused on late radiation‐associated proctopathy, further referred to as radiation proctopathy, although more recent studies may examine treatment for pelvic radiation disease (Andreyev 2013). Radiation proctopathy is the most often investigated late‐radiation effect to the pelvis. The common symptoms are rectal urgency, rectal incontinence, pain, strictures, mucus discharge, and rectal bleeding.

There have been extensive investigations into the association between rectal dose and the occurrence of late radiation proctopathy when treating the prostate gland (Boersma 1998; Peeters 2006; Pollack 2002). Since several dose‐escalation studies found an improved biochemical control for prostate cancer at higher radiation doses, the standard care for treatment is a prescribed dose of at least 75 Gy (Peeters 2006b; Pollack 2002; Zietman 2005). However, in these dose‐escalation studies, an increase of radiation proctopathy was found with higher doses to the prostate (and therefore also to the rectum). It is reasonable to assume that the risk of radiation proctopathy also applies to other cancer sites in the same region that are being radiated. Toxic effects of radiotherapy are often graded according to the Radiotherapy Therapy Oncology Group/European Organisation for Research and Treatment of Cancer (RTOG/EORTC) Late Radiation Morbidity Scoring Schema on a scale of 0 to 5 (www.rtog.org, Appendix 1). The Common Terminology Criteria for Adverse Events (CTCAE) and other adjusted local scorings schemas are also often used. Radiation proctopathy greater than grade 2 has been reported in 20% to 35% of men treated for prostate cancer (Al‐Mamgani 2008; Pollack 2002; Zietman 2005). However, these are physician‐reported outcomes. Patient‐reported outcomes (outcomes derived directly from people about how they function or feel in relation to a health condition and its therapy, without interpretation of the patient’s responses by a clinician) could be very different, and thus these toxicity outcomes could be underestimated (Higgins 2011).

The pathophysiology and symptomatology of radiation proctopathy are rather complex because different anorectal subregions can be involved (Heemsbergen 2005; Smeenk 2012b). For example, rectal incontinence and urgency are caused by reduced rectal capacity and tissue compliance resulting in impaired anal sphincter function (Kushwaha 2003; Smeenk 2012b; Yeoh 2009). Incontinence and urgency symptomatology originates from different muscle groups of the anorectal sphincter complex (Smeenk 2012). However, in this patient group incontinence can also be induced by loose stool consistency and speed of transit through the small bowel (Putta 2005). Rectal bleeding is usually the result of vulnerable mucosa combined with loss of submucosal capillaries, leading to new, but abnormal, blood vessels formation (teleangiectasia) in the rectal wall.

With modern radiotherapy techniques such as intensity modulated radiotherapy (IMRT) and volumetric modulated arc therapy (VMAT), higher conformal dose distribution to the prostate gland is reached with consequently lower doses to the anorectal complex. Yet, even with IMRT radiation, proctopathy incidence is 5% to 65% (Bekelman 2011; De Meerleer 2007; Zelefsky 2008).

Description of the intervention

The medical treatment of radiation proctopathy is not clearly defined and in the absence of recommendations, management is often unsatisfactory. This is due in part to difficulties in recognising and establishing the diagnosis and also because a proportion of the biological changes are not reversible. At present there is no 'best' treatment for this clinical scenario, and the outcomes of both medical and surgical management can be disappointing. The literature suggests a number of treatment options, including low‐residue or elemental diets, pain control, and replacement transfusion. Other therapies are reported to be of variable benefit in controlling symptoms, with the option of surgery if medical management fails or is inappropriate (Babb 1996).

Current non‐surgical treatment options include:

Aminosalicylic acid derivatives: Anti‐inflammatory agents in this group, such as sulfasalazine and mesalazine, have been reported to have a role in the management of this condition (Baum 1989). Another agent with anti‐inflammatory properties is WF10 (Veerasarn 2006).
- Sulfasalazine is a prodrug that is composed of 5‐aminosalicylic acid (5‐ASA) and sulfapyridine. In the colon, sulfasalazine is divided by the bacterial enzyme azoreductase into the two components. 5‐ASA (mesalazine), the active component, is poorly absorbed from the colon and is largely secreted in the stool. Sulfapyridine is largely absorbed in the colon and is associated with many side effects. Sulfasalazine is available as an oral and as a topical (rectal) compound. Intake is distributed in two to three times over day. At the beginning of therapy, a dose of 2 g to 4 g a day is given, which can be lowered to 1 g to 2 g a day for maintainance. Sulfazalasine, 5‐ASA (mesalazine), is the active component against inflammatory bowel disease. The precise mechanism of action is not known, but is mainly attributed to anti‐inflammatory and immunosuppressive properties. The considered mechanisms of action are: inhibition of cytokine synthesis (Bantel 2000; Rousseaux 2005), inhibition of prostaglandin and leukotriene synthesis (Hawkey 1985), free radical scavenging (Ahnfelt‐Ronne 1990), immunosuppressive activity by inhibition of T‐cell and B‐cell activation (MacDermott 1989a; Stevens 1995), and impairement of white cell adhesion and function (Neal 1987).
- Mesalazine is the active component of sulfasalazine. It is rapidly absorbed in the jejunum. Several compounds are on the market with delayed release of mesalazine to increase the availability for the colon.
Short chain fatty acid (SCFA) preparations: SCFA enema is a solution of sodium acetate, sodium propionate, sodium n‐butyrate with additional sodium chloride. As butyrate is the most effective SCFA for colonic regeneration, an only‐butyrate enema as a 80 mmol solution is also effective (Hille 2008; Vernia 2000). This solution is placed endorectally once or twice a day. Vernia et al. found that an application course of three weeks or more is needed for optimal result (Vernia 2000). An improvement can be observed after four to six weeks (Cook 1998). SCFAs are organic fatty acids that are produced by colonic bacterial metabolism. SFCAs are produced predominantly in the colon by anaerobic bacterial fermentation of non‐absorbed carbohydrates. The majority is absorbed in the colon. When SCFA is absorbed by colonocytes, sodium and water absorption is stimulated. Besides this, SCFA dilates resistance arteries, increase mucosal blood flow and oxygen uptake, reduces mucosal permeability, and enhances production and release of mucus (Kvietys 1981; Mortensen 1990). SCFAs are necessary for optimal growth of the colonic mucosa and stimulate cell proliferation.
Sucralfate preparations: Sucralfate is an aluminium salt of sucrose octasulfate. It can be used twice a day in a 1 g to 2 g enema as a 10% suspension in water (Kochhar 1991; McElvanna 2014). Relief of symptoms can be expected after one to two weeks. Other possible suspensions are in propylcellulose or glycerine. The dose varies from 1 g to 10 g (Carling 1986; Kochhar 1988). Sucralfate works by a cytoprotective effect because of an adherent complex binding to tissue proteins of the mucosa and protecting the colonic mucosa. There is evidence for promoting angiogenesis and reduction of microvascular injury (O'Brien 1997). Another possible mechanism is because of lowering of prostaglandin‐E2 levels (Zahavi 1989).

Coagulation therapy:
- Bipolar electrocoagulation: Coagulation is achieved with a probe by heating the contact tissue. Direct contact of the probe with the treated tissue is necessary. At the tip of the probe positive and negative electrodes are located that pass electricity through the tissue. Coagulation depth is controlled by probe size, duration of heating, and choice of energy. The applied current is locally between the electrodes of the probe. The coagulation is aimed at bleeding telangiectatic areas. At 70ºC tissue coagulation will occur and bleeding is stopped. Once the tissue is desiccated, tissue resistance increases and prevents deep coagulation. Bipolar electrocoagulation is effective for superficial lesions. Penetration depth can be regulated by the probe size, applied energy level, and pressure of the probe on tissue.
- Thermal coagulation therapy: The heater probe is a thermocouple to heat up the exposed tissue by cauterisation. At the tip of the probe there is a heat‐generating device that converts electric energy to heat energy (Protell 1978).
- Argon plasma coagulation uses ionised argon gas for a thermal reaction. A probe is introduced to the treatment area, and argon gas, which itself is non‐flammable, is sprayed on the surface to be treated and ionised by 6000 volts via electric wires in the probe. A high‐frequency current develops between the electrode and the underlying tissue, resulting in coagulation and desiccation. Desiccated tissue loses electric conductivity, prohibiting deep devitalization of the treated tissue. Coagulation is achieved by heating up the treated tissue. Heating is not influenced by tissue resistance, making deep coagulation possible. The mechanism of coagulation and desiccation of the treated tissue is by direct heat transfer.
Corticosteroids: Several steroid enemas are available for radiation proctopathy treatment. Hydrocortisone, prednisolone, and betamethasone were used in this review. The enema is introduced into the rectum for several minutes once a day. The treatment can be continued for 2 to 4 weeks. Hydrocortisone is available as, for example, 100 mg in a 60 ml aqueous solution. Betamethasone, which has a more powerful action than hydrocortisone, is available as, for example, 5 mg in a 100 ml aqueous solution. The working mechanism of topical corticosteroid application (steroid enema) is an anti‐inflammatory effect with the inhibition of prostaglandin synthesis.

Formalin applications: Formalin is applied topically in a 4% to 10% solution on the affected rectal mucosa. The application can be either by irrigation of the rectum or direct application with a soaked gauze. Formalin acts by hydrolysing proteins leading to chemical cauterisation. Another mechanism is by coagulation of the tissue (Haas 2007; Leiper 2007; Parikh 2003).
Pentoxyfilline: Pentoxyfilline is a xanthine derivative. It is almost completely resorbed by oral admission. It is also used for intermittent claudication and administered orally three times a day 400 mg. Pentoxifylline increases the deformability of erythrocytes, inhibits the aggregation of thrombocytes, and reduces fibrinogen level in plasma. In this way microvascular blood flow is enhanced.
Antibiotic treatment: Metronidazole is an antibiotic that is especially useful in the treatment of anaerobic infections. Metronidazole is effective against obligate anaerobes and against facultative anaerobes such as Helicobacter pylori and Gardnerella vaginalis. It is orally well absorbed and distributed evenly into body tissues. The dosage is dependent on the indication for treatment. Usual oral administration is 1 g to 2 g distributed in one to four intakes. The working mechanism of metronidazole is by its anti‐inflammatory action to improve mucosal healing in combination with other therapeutic measures.
Hyperbaric oxygen: Hyperbaric oxygen therapy is performed in specially designed chambers. One or more people are treated in the chamber with 100% oxygen. Pressure is increased to 25 x 10⁴ to 30 x 10⁴ Pa. Treatment duration is approximately 60 to 90 minutes, 5 to 7 days a week. Number of treatments varies with the degree of severity and response effect, usually 30 to 40 treatments. Plasma oxygen level increases by inhalation of 100% high‐pressure oxygen. As a result, tissue oxygenation will improve. Hyperbaric oxygen facilitates fibroblast proliferation, angiogenesis, and wound healing (Marx 1990; Roth 1994; Wattel 1998).
Retinol palmitate: Retinol palmitate, or vitamin A, is completely absorbed via the bowel. The recommended daily allowance for an adult male is 3000 IU (900 micrograms) and for an adult female 2300 IU (700 micrograms). Therapeutic dosage for retinol in case of deficiency is 25,000 to 50,000 IU per day for a limited period. The prophylactic dosage is 2500 to 5000 IU per day. Retinol palmitate increases fibroblast secretion of mucopolysaccharide and collagen and increases fibronectin synthesis. These mechanisms assist in wound healing (Hein 1984).
Chinese traditional medicine in combination with Western medicine: consists of Shen Ling Bai Zhu powders (herbal ingredients) and Western medicine (smectite powder 6 g, dexamethasone 5 mg, levofloxacin hydrochloride 0.2 g, anisodamine 10 mg, and physiological saline 100 ml).

How the intervention might work

See Description of the intervention

Why it is important to do this review

As described above, late radiation proctopathy is common, with incidence rates of 5% to 65% (Bekelman 2011; De Meerleer 2007; Zelefsky 2008), and potentially a major burden to people with cancer. The medical treatment of radiation proctopathy is not clearly defined, and in the absence of recommendations, management of the condition is often unsatisfactory. While the literature suggests a number of treatment options, there is no 'best' treatment choice. We are therefore assessing the effectiveness of various non‐surgical treatment options in managing late radiation proctopathy in this review.

Objectives

To assess the effectiveness and safety of non‐surgical interventions for managing late radiation proctopathy.

Methods

Criteria for considering studies for this review

Types of studies

All randomised controlled trials (RCTs) or quasi‐RCTs, irrespective of language and publication status, that compared any non‐surgical intervention for late radiation proctopathy to no intervention, placebo, or any other intervention were eligible for inclusion. In previous versions of this review, non‐randomised studies were also eligible. For this update, as RCT data was available, we decided to focus on RCTs and quasi‐RCTs only, as non‐randomised studies suffer from high risk of bias.

Types of participants

People diagnosed with a pelvic malignancy, who had undergone pelvic radiotherapy as part of their treatment schedule (primary radiotherapy, pre‐ or postoperative radiotherapy, with or without chemotherapy, or as a palliative treatment) and subsequently developed late radiation proctopathy, defined as radiation proctopathy of any grade, continuing from completion of radiotherapy for more than three months, or occurring more than three months after completion of radiotherapy.

Types of interventions

Experimental: any non‐surgical intervention for late radiation proctopathy, such as:

Aminosalicylic acid derivatives
Short chain fatty acid (SCFA) preparations
Sucralfate preparations
Coagulation therapy
Corticosteroids
Formalin applications
Pentoxyfilline
Hyperbaric oxygen
Antibiotic treatment

Control: no intervention, placebo, or any other non‐surgical intervention.

Types of outcome measures

Primary outcomes

We determined primary outcome measures by the presenting symptoms as recorded retrospectively or prospectively with diaries and scoring systems, for example:

Episodes of bowel activity
Bleeding
Pain
Tenesmus
Urgency
Sphincter dysfunction

Secondary outcomes

Mortality
Morbidity
Quality of life (QoL): 12‐Item Short Form Health Survey (Ware 1996), Subjective, Objective, Management, Analytic ‐ Late Effects of Normal Tissue (SOMA‐LENT) score (Pavy 1995), Visual Analogue Scale, (Scott 1976), mean Inflammatory Bowel Disease Questionnaire–Bowel subset score (IBDQ‐B) (Cheung 2000)

Search methods for identification of studies

Electronic searches

The literature searches from inception to 2007 have been updated and were re‐run in November 2015.

The Cochrane Central Register of Controlled Trials (CENTRAL) (Issue 11, 2015) (Appendix 2)
MEDLINE (Ovid) (April 2007 to Nov week 1 2015) (Appendix 3)
EMBASE (Ovid) (April 2007 to 2015 week 45) (Appendix 4)

We did not apply a search filter due to the range of interventions searched for. This basic strategy was expanded for text and MeSH terms before being applied to the described databases.

Searching other resources

In addition, we searched the prospective trial register ClinicalTrials.gov with the key words 'proctitis' AND 'radiation' and 'proctopathy' AND 'radiation', and checked reference lists.

Data collection and analysis

We downloaded all titles and abstracts retrieved by the electronic searches to a reference management database (Reference Manager) and removed duplicate references. Two pairs of review authors (FW and RS; LV and JV) independently examined the remaining references. We excluded those studies that clearly did not meet the inclusion criteria, and we obtained copies of the full text of potentially relevant references. Two pairs of review authors (FW and RS; LV and JV) independently assessed the eligibility of the retrieved reports/publications. We resolved any disagreements through discussion or, if required, by consulting a third review author (RS). We identified and excluded duplicates. We recorded the selection process in sufficient detail to complete a PRISMA flow diagram and Characteristics of excluded studies table.

Data extraction and management

We (FW, LV, JV) independently abstracted the data using a pre‐designed data abstraction form (Appendix 5). A second review author (LV, JV, or RS) checked data abstraction for accuracy. One review author (FW) transferred data into the RevMan 5 file (Review Manager 2014). We double‐checked that data was entered correctly by comparing the data presented in the systematic review with the study reports.

Assessment of risk of bias in included studies

Two review authors (FW, RS) independently assessed the risk of bias of all included studies according to the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). We grouped outcomes as ‘subjective’ and ‘objective’ for the purposes of assessing blinding and incomplete outcome data. We resolved disagreements through consensus. For each relevant comparison we summarised the evidence per outcome in an additional table. We allocated the level of evidence using the the Grading of Recommendations Assessment, Development and Evaluation (GRADE) system (GRADEpro 2014).

Measures of treatment effect

We described dichotomous data using the risk ratio (RR) with 95% confidence interval (CI). We expressed continuous data, that is symptom scores, as mean differences (MDs).

Dealing with missing data

If outcome data were missing, we planned to contact trial authors.

Assessment of heterogeneity

We planned to formally test statistical heterogeneity using the natural approximate Chi² test, which provides evidence of variation in effect estimates beyond that of chance. Since the Chi² test has low power to assess heterogeneity where a small number of participants or trials are included, we planned to set the P value conservatively at 0.1. We planned to test heterogeneity using the I² statistic, which calculates the percentage of variability due to heterogeneity rather than chance.

We planned to interpret the I² statistics in relation to the size of the included studies. We used the following interpretation as a rough guide:

0% to 40%: might not be important;
30% to 60%: may represent moderate heterogeneity;
50% to 90%: may represent substantial heterogeneity;
75% to 100%: considerable heterogeneity.

We planned to examine potential sources of clinical heterogeneity through the use of analyses as specified above. We planned to examine potential sources of methodological heterogeneity through the use of sensitivity analyses (Sensitivity analysis).

Assessment of reporting biases

If more than 10 included studies were available, we planned to use funnel plots to assess the potential for small‐study effects such as selective publication.

Data synthesis

We had planned to pool outcome data from studies that were sufficiently similar in participant characteristics and methodology followed (length of follow‐up, diagnostic criteria) and to use the random‐effects model for meta‐analysis, as we expected diversity in cancer types, interventions, and definitions of radiation proctopathy across included studies. However, due to the heterogeneous nature of the included studies, this was not feasible for this update but may be implemented in the future.

Subgroup analysis and investigation of heterogeneity

We planned to conduct subgroup analyses to investigate possible differences between groups, considering factors such as age, stage, type of intervention, and length of follow‐up in interpreting any heterogeneity.

Sensitivity analysis

We planned sensitivity analyses by excluding studies with high risk of bias. However as no meta‐analyses were performed, no sensitivity analyses were undertaken.

Results

Description of studies

Results of the search

In the previous version of this review (Denton 2002), which was assessed as up to date in 2008, seven trials were included (Cavcic 2000; Ehrenpreis 2005; Jensen 1997; Kochhar 1991; Pinto 1999; Rougier 1992; Talley 1997). Through the update, we found a total of 1443 trials as a result of the literature search. After we removed duplicates, 1221 were left for evaluation. We (FW, LV, JV) identified a total of 48 potential titles and abstracts for full‐text evaluation. After review of full text and discussion (FW, LV, JV, RS), we excluded a further 31 trials (Characteristics of excluded studies). We (FW, LV, JV, RS) eventually decided to include an additional nine trials (Figure 1), increasing the total number of included trials to 16 (including 993 participants). One eligible study found through the last update was moved to the Studies awaiting classification section (Guo 2015). Details of the participants, interventions, and outcomes in these trials are in the Characteristics of included studies table.

Included studies

Study design

Five trials were placebo controlled (Chruscielewska 2012; Clarke 2008; Ehrenpreis 2005; Pinto 1999; Talley 1997), three trials assessed the effect of the addition of one intervention to another intervention, (Cavcic 2000; Kochhar 1991; Venkitaraman 2008) and eight trials compared two or more active interventions (Andreyev 2013; Jensen 1997; Lenz 2010; Nelamangala 2012; Rougier 1992; Sahakitrungruang 2012; Tian 2008; Yeoh 2013). All but two trials addressed single comparisons (Pinto 1999; Talley 1997).

Types of interventions

Four trials compared different types of anti‐inflammatory agents as a treatment for late radiation proctopathy (Cavcic 2000; Kochhar 1991; Rougier 1992; Venkitaraman 2008). In two trials, short‐chain fatty acids (SCFA) enemas were compared to placebo (Pinto 1999; Talley 1997). Three trials assessed the effects of sucralfate compared to either anti‐inflammatory agents (Kochhar 1991), formalin therapy (Nelamangala 2012), or placebo (Chruscielewska 2012). Another three trials assessed the effects of formalin therapy compared to colonic irrigation (Sahakitrungruang 2012), sucralfate steriod (Nelamangala 2012), or argon plasma coagulation (Yeoh 2013). Two trials assessed the effect of thermal coagulation therapy to either heater probe, in Jensen 1997, or bipolar coagulation, in Lenz 2010. One trial assessed the effect of hyperbaric oxygen therapy versus placebo (Clarke 2008), and three trials assessed other interventions: integrated Chinese traditional medicine (Tian 2008), retinol palmitate (Ehrenpreis 2005), and a gastroenterologist‐led algorithm‐based treatment with a nurse‐led algorithm‐based treatment or with usual care (Andreyev 2013).

Participant characteristics

Three trials included participants with prostate cancer (Cavcic 2000; Venkitaraman 2008; Yeoh 2013), one trial included participants with cervical cancer (Nelamangala 2012), seven trials included a mixed population of participants with prostate, cervical, endometrial, uterine, vaginal, or rectal cancer (Andreyev 2013; Chruscielewska 2012; Jensen 1997; Kochhar 1991; Lenz 2010; Sahakitrungruang 2012; Talley 1997), and five trials did not specify cancer types of their participants (Clarke 2008; Ehrenpreis 2005; Pinto 1999; Rougier 1992; Tian 2008).

Fourteen trials were published in English, one was published in French, and one in Chinese.

In addition, we searched the prospective trial register ClinicalTrials.gov with the key words 'proctitis' AND 'radiation' and 'proctopathy' AND 'radiation' and found four potentially relevant registered trials that are ongoing (Characteristics of ongoing studies).

Excluded studies

A total of 47 titles and abstracts seemed to fulfil our inclusion criteria and required further discussion. Through discussion we excluded 31 of these studies (Characteristics of excluded studies). We excluded 11 studies because of their design (no randomised studies), 10 studies because the population did not fulfil our inclusion criteria (no late rectal consequences of radiotherapy), 9 studies as they were prevention studies, and 1 study because the intervention did not fulfil our inclusion criteria.

Risk of bias in included studies

The results of the 'Risk of bias' assessment of the included studies are presented in the Characteristics of included studies tables and summarised in Figure 2 and Figure 3.

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Allocation

Appropriate sequence generation to ensure randomisation seemed likely in eight studies (Andreyev 2013; Chruscielewska 2012; Clarke 2008; Ehrenpreis 2005; Nelamangala 2012; Sahakitrungruang 2012; Venkitaraman 2008; Yeoh 2013). Cavcic 2000 and Tian 2008 had a high‐risk score as they were quasi‐randomised trials (treatment was allocated according to the date on which their treatment began, in Cavcic 2000, and according to treatment order, in Tian 2008). The remaining six studies did not adequately describe the method of randomisation, and thus were judged to be of unclear risk of selection bias (Jensen 1997; Kochhar 1991; Lenz 2010; Pinto 1999; Rougier 1992; Talley 1997).

Allocation concealment was adequate in six studies (Andreyev 2013; Chruscielewska 2012; Clarke 2008; Ehrenpreis 2005; Lenz 2010; Yeoh 2013; ). We judged the method of allocation concealment to be inappropriate, resulting in a high risk of bias, in three studies, as two used quasi‐randomisation (Cavcic 2000; Tian 2008), and one used an open list of random numbers (Nelamangala 2012). The method of allocation concealment was not clearly described in the remaining seven studies.

Blinding

The participants and personnel were not blinded in six studies (Andreyev 2013; Cavcic 2000; Nelamangala 2012; Sahakitrungruang 2012; Venkitaraman 2008; Yeoh 2013), which we therefore judged as having a high risk of performance bias. Participants and personnel were adequately blinded in eight studies (Chruscielewska 2012; Clarke 2008; Ehrenpreis 2005; Jensen 1997; Kochhar 1991; Lenz 2010; Pinto 1999; Talley 1997). For the remaining two studies, insufficient information was provided to judge the potential risk of performance bias and we therefore considered them to be at an unclear risk of bias.

Outcome assessors for subjective outcomes were not blinded in six studies (Andreyev 2013; Cavcic 2000; Nelamangala 2012; Sahakitrungruang 2012; Venkitaraman 2008; Yeoh 2013), and blinded in eight studies (Chruscielewska 2012; Clarke 2008; Ehrenpreis 2005; Jensen 1997; Kochhar 1991; Lenz 2010; Pinto 1999; Talley 1997). In the remaining two studies, the risk of bias on this item was unclear due to insufficient information.

As for the objective outcomes, the outcome assessors were blinded in seven studies (Cavcic 2000; Clarke 2008; Ehrenpreis 2005; Lenz 2010; Nelamangala 2012; Pinto 1999; Sahakitrungruang 2012;). In six studies the risk of bias on this item was again unclear due to insufficient information, and three studies did not assess objective outcomes (Andreyev 2013; Chruscielewska 2012; Yeoh 2013).

Incomplete outcome data

For subjective outcomes, we judged eight studies as at low risk of attrition bias (Chruscielewska 2012; Ehrenpreis 2005; Jensen 1997; Lenz 2010; Nelamangala 2012; Pinto 1999; Sahakitrungruang 2012; Tian 2008). We scored one study as at high risk of attrition bias, as a substantial number of participants (25% after three months, 40% after one year, 52% after two years) dropped out of the study for unknown reasons (Cavcic 2000). For the remaining seven studies, insufficient information was provided to draw a safe conclusion.

As for the objective outcomes, we judged six studies as at low risk of attrition bias (Ehrenpreis 2005; Jensen 1997; Lenz 2010; Nelamangala 2012; Sahakitrungruang 2012; Tian 2008). We scored one study as at high risk for attrition bias because the number of dropouts was not evenly distributed between the two groups (at three months: 10% versus 40%; after one year: 20% versus 60%; after two years: 37% versus 67%), and reasons for lost to follow‐up were not reported (Cavcic 2000). In three studies (Andreyev 2013; Chruscielewska 2012; Yeoh 2013), objective outcomes were not assessed. The remaining six studies provided insufficient information to draw a safe conclusion and we therefore judged them as at unclear risk of bias.

Selective reporting

We judged only two studies to be at low risk of selective reporting bias (Andreyev 2013; Lenz 2010). We judged two studies to be at high risk of reporting bias (Rougier 1992; Sahakitrungruang 2012). In the remaining studies no protocol was available, and we therefore judged the risk of bias on this item as unclear.

Other potential sources of bias

We judged all but two of the studies to be at low risk of other bias. In one study there were substantial differences in baseline characteristics (more aggressive grade of disease in the betamethasone group), resulting in a high risk of bias (Rougier 1992). We judged a second study to be at unclear risk of bias, as they failed to achieve the expected rate of enrolment, which may have influenced the results (Venkitaraman 2008).

Effects of interventions

Since radiation proctopathy is a condition with various symptoms or combinations of symptoms, the studies are heterogeneous in their intended effect. Some studies investigated treatments targeted at bleeding only (group 1), some investigated treatments targeted at a combination of anorectal symptoms, but not a single treatment (group 2). The third group focused on the treatment of the collection of symptoms referred to as pelvic radiation disease. In order to enable some comparison of this heterogeneous collection of studies, we described the effects in these three groups separately.

1. Treatments for rectal bleeding

(see Table 1)

1. Treatment for bleeding only.

Study ID	Participant characteristics	Intervention(s)	Results	Level of evidence (GRADE)
Chruscielewska 2012	Adults with chronic radiation proctopathy or proctosigmoiditis were included if all of the following criteria were met: radiotherapy for a pelvic tumour completed at least 3 months before enrolment, the presence of rectal bleeding, radiation‐induced telangiectasia in the rectum and/or sigmoid colon on endoscopy, informed written consent by the person to participate in the study	Endoscopic APC followed by oral sucralfate (6 g twice daily) for 4 weeks (n = 60) versus APC with placebo administration for 4 weeks (n = 62)	Changes in chronic radiation proctopathy severity score (based on scales previously proposed by Chutkan et al. and Kochhar et al., median (interquartile range)) Overall severity score Week 8: 2 (1.3) versus 2 (1.3) Week 16: 2 (1.3) versus 2 (1.3) Week 52: 1 (1.2) versus (1.2) P > 0.05 for all between‐group comparisons	Moderate due to imprecision (precision not quantified)
			Diarrhoea score Week 8: 1 (1.1) versus 1 (1.1) Week 16: 1 (1.1) versus 1 (1.1) Week 52: 1 (1.1) versus 1 (1.1) P > 0.05 for all between‐group comparisons	Moderate due to imprecision (precision not quantified)
			Bleeding score Week 8: 1 (0.1) versus 0 (0.1) Week 16: 0 (0.1) versus 0 (0.1) Week 52: 0 (0.1) versus 0 (0.0) P > 0.05 for all between‐group comparisons	Moderate due to imprecision (precision not quantified)
			Tenesmus ⁄ rectal pain score Week 8: 0 (0.1) versus 0 (0.1) Week 16: 0 (0.0) versus 0 (0.1) Week 52: 0 (0.0) versus 0 (0.0) P > 0.05 for all between‐group comparisons	Moderate due to imprecision (precision not quantified)
			Endoscopy scores (graded according to the Gilinsky scale as endoscopic severity score) Score 6 to 9: Week 8: 20 to 59 versus 9 to 60 (RR 2.26, 95% CI 1.12 to 4.55) Week 16: 10 to 57 versus 8 to 60 (RR 1.32, 95% CI 0.56 to 3.10)	‐
			Complications Number (%) of people with complications: 38/60 (63%) versus 36/62 (58%) (RR 1.09, 95% CI 0.82 to 1.45)	Moderate due to imprecision (CI includes both benefits and harms)
			APC‐related complications Asymptomatic rectal ulcer: 30 versus 30 Symptomatic rectal ulcers: 7 versus 5 Rectovaginal fistula: 2 versus 0 Adynamic ileus: 0 versus 1	Low due to very serious limitations for imprecision (CI includes both benefits and harms)
			Other complications Severe constipation: 4 versus 0 Urticaria: 1 versus 0	‐
			Complication severity (APC‐related) Severe + fatal: 2/60 versus 0/62 (RR 5.17, 95% CI 0.25 to 105.4) P > 0.05 for all between‐group comparisons regarding complications	Low due to very serious imprecision (CI includes both benefits and harms)
Jensen 1997	People being considered for surgery, having failed 1 year of medical therapy, pelvic radiotherapy for a cancer at least 2 years earlier, rectal bleeds at least 3 times per week, anaemia, and a life expectancy of at least 2 years	Bipolar electrocoagulation (n = 12) versus Heater probe (n = 9)	Severe bleeding episodes after 1 year (measured by participant interview) 3/12 (33%) versus 1/9 (11%) (RR 2.25, 95% CI 0.28 to 18.22)	Low due to very serious imprecision (CI includes both benefits and harms)
			Mean number of severe bleeds after 1 year, mean (SD) (measured by participant interview) 0.3 (0.3) versus 0.4 (0.9) (MD ‐0.10, 95% CI ‐0.71 to 0.51)	Low due to very serious imprecision (CI includes both benefits and harms)
			Mean haematocrits after 1 year, mean (SD) 38.2 (4.8) versus 37.6 (3.0) (MD 0.60, 95% ‐2.75 to 3.95)	Low due to very serious imprecision (CI includes both benefits and harms)
			Mean units of red blood cells transfused after 1 year, mean (SD) 0.0 (0.0) versus 0.2 (0.6) (MD not estimable)	Low due to very serious imprecision (OIS not reached)
			Complications No major complications occurred
			QoL QoL informally assessed with participant responses, which improved with treatment
Kochhar 1991	Symptomatic radiation‐induced proctosigmoiditis	Oral sulfasalazine (500 mg tds) + rectal prednisolone (20 mg bd) for 4 weeks (n = 8) versus Oral placebo + rectal sucralfate suspension (2 g bd) for 4 weeks (n = 19)	Clinical improvement at 4 weeks (an in‐house scoring system was used measuring diarrhoea, bleeding per rectum, bleeding requiring blood transfusion, and tenesmus) 8 to 15 versus 16 to 17 (RR 0.57, 95% CI 0.35 to 0.92)	Low due to unclear RoB and imprecision (OIS not reached)
			Endoscopic improvement at 4 weeks (injury was graded according to the criteria of Gilinsky et al.) 7 to 15 versus 12 to 17 (RR 0.66, 95% CI 0.35 to 1.23)	Very low due to unclear RoB and very serious imprecision (CI includes both benefits and harms)
			Side effects at 4 weeks (participants were questioned and examined for any side effects of the medication) 2 participants in the sulfasalazine group did not tolerate the drugs due to myalgia, nausea, and headaches	Very low due to unclear RoB and very serious imprecision (very small sample size)
Lenz 2010	Recurrent rectal bleeding, started 6 months after radiotherapy with at least 1 bleeding episode in the week before and endoscopically confirmed radiation telangiectasias	APC (n = 15) versus Bipolar electrocoagulation (n = 15)	Eradication of all telangiectasias at 1 year 12/15 versus 14/15 (RR 0.86, 95% CI 0.64 to 1.14)	Low due to indirectness and imprecision (CI includes both benefits and harms)
			Complications Minor: 5/15 versus 10/15 (RR 0.50, 95% CI 0.22 to 1.11) Major: 1/15 versus 5/15 (RR 0.20, 95% CI 0.03 to 1.51)	Low due to very serious imprecision (CI includes both benefits)
			Relapse 1/12 versus 2/14 (RR 0.58, 95% CI 0.06 to 5.66)	Low due to very serious imprecision (CI includes both benefits and harms)
Nelamangala 2012	Rectal bleeding as a result of chronic haemorrhagic radiation proctopathy, following radiotherapy for carcinoma of the cervix	Formalin dab treatment (4%) as an outpatient procedure (formalin applied directly to each lesion via a rigid sigmoidoscope or a proctoscope under local anaesthesia. Under direct vision, a small piece of gauze soaked in 4% formalin was applied to the haemorrhagic areas for 2 min until the mucosa turned pale) (n = 51) versus Sucralfate steroid retention enema (100 mg of prednisolone and 1 g of sucralfate in 100 ml of normal saline), twice daily for 7 to 10 days (n = 51)	Symptom score after treatment (graded by the RPSAS), median (range) 9 (6 to 24) versus 13 (8 to 27) (P < 0.001)	Low due to high risk of bias and imprecision (OIS not reached)
			Sigmoidoscopic score (median (range) after treatment) 1 (0 to 3) versus 2 (0 to 3) (P < 0.001)	Very low due to high risk of bias, imprecision and indirectness (OIS not reached)
			Adverse events “Mild pain occurred in 33.3% patients in Group 1 during the application of formalin but this subsided within 1 day. There were no complications in Group 2.”	Low due to high risk of bias and imprecision (OIS not reached)
Pinto 1999	Clinical and histological diagnosis of chronic radioproctitis: Grade III Pourquier classification with rectal bleeding at least once a week, persistent symptoms for a minimum of 12 months (except 1 participant)	SCFA enema (as per Harig preparation 60 ml bd 5 weeks) (n = 10) versus Placebo: identically appearing enemas containing saline isotonic solution, bd 5 weeks (n = 9)	Mean endoscopic score after 5 weeks, mean (SD) 2.6 (1.8) versus 1.6 (4.2) (MD 1.00, 95% CI ‐2.33 to 4.33)	Low due to indirectness and imprecision (CI includes both benefits and harms)
			Number of days of rectal bleeding at the end of treatment (after 5 weeks), mean (SD) (participant diary) 1.4 (2.2) versus 3.4 (2.6) (MD ‐2.00, 95% CI ‐4.40 to 0.40)	Low due to very serious imprecision (CI includes both benefits and harms)
			Haemoglobin levels after 5 weeks, mean (SD) 13.1 ± 0.9 g/dl versus 10.7 ± 2.1 g/dl (MD 2.40, 95% CI 0.74 to 4.06)	Low due to imprecision (very small sample size)
Rougier 1992	Chronic radiation proctopathy confirmed and graded on sigmoidoscopy	Hydrocortisone acetate mousse 90 mg bd for 4 weeks (n = 16) versus Betamethasone lavage 5 mg od for 4 weeks (n = 16)	Improvement of endoscopic appearance 12/16 versus 5/14 (RR 2.10, 95% CI 0.98 to 4.48)	Very low due to high RoB, indirectness, and imprecision (CI includes both benefits and harms)
			Reduction of degree of bleeding (an in‐house scoring system was used) 6/16 versus 3/14 (RR 1.75, 95% CI 0.53 to 5.73)	Very low due to high RoB and very serious imprecision (CI includes both benefits and harms)
			Poor tolerance of enema 2/16 versus 10/14 (RR 0.17, 95% CI 0.05 to 0.67)	Low due to high RoB and imprecision (OIS not reached)
Sahakitrungruang 2012	Symptomatic haemorrhagic radiation proctopathy for more than 6 months without complications of rectal stricture, deep ulceration, fistula formation, or sepsis	Colonic irrigation (with 1000 ml of tap water via a 20F Foley catheter) plus ciprofloxacin (500 mg twice daily) and metronidazole (500 mg 3 times daily) by mouth for 1 week (n = 25) versus Formalin application in an office setting (4% formalin‐soaked gauze applied for 3 minutes under direct vision by using proctoscopy followed by immediate cleansing with water irrigation) (n = 25)	Comparison between the 2 treatment groups after 8 wks of treatment (measured by participant survey) Bleeding (days/week) median (min,max) ‐5 (‐7,0) versus ‐2 (‐7,4) (P = 0.007)	Low due to high RoB and imprecision (OIS not reached)
			Frequency (times/day) ‐2 (‐8,2) versus ‐2 (‐4,2) (P = 0.09)	Low due to high RoB and imprecision (OIS not reached)
			Urgency (days/week) ‐2 (‐7,3) versus 0 (‐2,7) (P = 0.0004)	Low due to high RoB and imprecision (OIS not reached)
			Diarrhoea (days/week) ‐2 (‐6,0) versus 0 (‐7,2) (P = 0.007)	Low due to high RoB and imprecision (OIS not reached)
			Tenesmus (days/week) ‐2 (‐7,0) versus 0 (‐4,4) (P = 0.07)	Low due to high RoB and imprecision (OIS not reached)
			Haematocrit (mg/dL) 0 (‐9,10) versus 0 (‐7,18) (P = 0.86)	Low due to high RoB and imprecision (OIS not reached)
			Vienna rectoscopy score after 8 wks of treatment No significant differences between groups (P = 0.78)	Low due to high RoB and imprecision (OIS not reached)
			Adverse events after 8 wks of treatment "There were no serious adverse drug reactions." "Four patients in the irrigation group and 8 patients in the formalin group required blood transfusion during the study, but this finding did not reach the statistic difference."	Low due to high RoB and imprecision (OIS not reached)
			Patient satisfaction 20/24 versus 10/23 (RR 1.92, 95% CI 1.16 to 3.16)	Low due to high RoB and imprecision (OIS not reached)
Venkitaraman 2008	Symptomatic late morbidity with at least 1 episode of rectal bleeding more than 6 months since pelvic radiotherapy	Standard therapies for late radiation‐induced bleeding plus oral pentoxifylline (400 mg) 3 times daily for 6 months (n = 20) versus Standard treatment for late radiation‐induced bleeding (n = 20)	Cessation of rectal bleeding (measured using a daily participant symptom diary) 16/20 versus 12/20 (RR 1.33, 95% CI 0.88 to 2.03)	Very low due to high RoB and very serious imprecision (CI includes both benefits and harms)
			Median time to cessation of bleeding (measured using a daily participant symptom diary) 22 days (range 1 to 119 days) versus 95 days (range 13 to 172) (P = 0.12) “However, at least one episode of recurrent bleeding occurred in 14 of the 16 patients in the study group and in all the 12 patients who had cessation of bleeding in the control group.”	Low due to high RoB and imprecision (OIS not reached)
			The median duration of freedom from bleeding (measured using a daily participant symptom diary) 12 days (range 8 to 290) versus 11 days (range 7 to 133)	Very low due to high RoB and very serious imprecision (precision not quantified; very small sample size)
			Adverse events “Pentoxifylline was well tolerated with minor side‐effects and the compliance rate was satisfactory. Seven patients required a dose reduction or temporary discontinuation of pentoxifylline, due to dyspepsia in five patients and rashes and chest pain in one patient each, whereas one patient required permanent discontinuation of pentoxifylline due to dyspepsia. Contrary to expectations, one patient had a transient, but significant, increase in serum fibrinogen levels.”	Very low due to high RoB and very serious imprecision (OIS not reached)

Date	Event	Description
12 November 2015	New citation required but conclusions have not changed	Review updated and new studies added, but overall conclusions unchanged.
12 November 2015	New search has been performed	Nine new studies (RCTs) included. We have assessed the risk of bias of all included studies with the Cochrane 'Risk of bias' tool and divided the review into treatments targeted at bleeding, treatments targeted at very localised signs or symptoms, and treatments targeted at the more global collection of symptoms. In addition, we have changed the title from 'Non‐surgical interventions for late radiation proctitis in patients who have received radical radiotherapy to the pelvis' to 'Non‐surgical interventions for late rectal problems (proctopathy) of radiotherapy in people who have received radiotherapy to the pelvis'.

Date	Event	Description
12 November 2007	New search has been performed	New studies found and included or excluded: 25 April 2007. The literature search as described in the search strategy sections was updated on 25 April 2007. The time frame used was from the April 2007 back to 1999. Fifteen publications were identified that had not been previously included, apparently dealing with the various therapeutic options for management of late radiation proctitis. Initial examination confirmed that 10 were case reports or case series, two were reviews, one was a phase II study, and one was a systematic review containing an as yet unpublished RCT of hyberbaric oxygen in the management of late radiotherapy proctitis. The other reports were excluded, leaving the phase II prospective series and one possible randomised comparative trial.    Hence seven controlled studies are currently reported: three compared different rectal steroid preparations and anti‐inflammatory agents versus rectal sucralfate, and were not placebo controlled; two evaluated short chain fatty acid enemas and were placebo controlled but of different designs; one compared the effect of bipolar electrocoagulation versus the heater probe; one unpublished study examined the effect of hyberbaric oxygen versus placebo; another reported a double‐blind placebo RCT using vitamin A; and lastly a phase II series was included on people who had received WF10 therapy.    As in the original review, all the sections describe different interventions and outcome parameters so that they cannot be compared and a summary statistic cannot be derived.
29 October 2003	New citation required and conclusions have changed	Substantive amendment.

Organ tissue	0	Grade 1	Grade 2	Grade 3	Grade 4	5
Skin	None	Slight atrophy Pigmentation change Some hair loss	Patch atrophy Moderate telangiectasia Total hair loss	Marked atrophy Gross telangiectasia	Ulceration	D E A T H D I R E C T L Y R E L A T E D T O R A D I A T I O N L A T E E F F E C T S
Subcutaneous tissue	None	Slight induration (fibrosia) and loss of subcutaneous fat	Moderate fibrosis but asymptomatic Slight field contracture < 10% linear reduction	Severe induration and loss of subcutaneous tissue Field contracture > 10% linear measurement	Necrosis
Mucous membrane	None	Slight atrophy and dryness	Moderate atrophy and telangiectasia Little mucous	Marked atrophy with complete dryness Severe telangiectasia	Ulceration
Salivary glands	None	Slight dryness of mouth Good response on stimulation	Moderate dryness of mouth Poor response on stimulation	Complete dryness of mouth No response on stimulation	Fibrosis
Spinal cord	None	Mild Lhermitte's syndrome	Severe Lhermitte's syndrome	Objective neurological findings at or below cord level treated	Mono, para, quadraplegia
Brain	None	Mild headache Slight lethargy	Moderate headache Great lethargy	Severe headaches Severe central nervous system dysfunction (partial loss of power or dyskinesia)	Seizures or paralysis Coma
Eye	None	Asymptomatic cataract Minor corneal ulceration or keratitis	Symptomatic cataract Moderate corneal ulceration Minor retinopathy or glaucoma	Severe keratitis Severe retinopathy or detachment Severe glaucoma	Panopthalmitis/Blindness
Larynx	None	Hoarseness Slight arytenoid oedema	Moderate arytenoid oedema Chondritis	Severe oedema Severe chondritis	Necrosis
Lung	None	Asymptomatic or mild symptoms (dry cough) Slight radiographic appearances	Moderate symptomatic fibrosis or pneumonitis (severe cough) Low‐grade fever Patchy radiographic appearances	Severe symptomatic fibrosis or pneumonitis Dense radiographic changes	Severe respiratory insufficiency/Continuous O₂/Assisted ventilation
Heart	None	Asymptomatic or mild symptoms Transient T wave inversion and ST changes Sinus tachycardia > 110 (at rest)	Moderate angina on effort Mild pericarditis Normal heart size Persistent abnormal T wave and ST changes Low ORS	Severe angina Pericardial effusion Constrictive pericarditis Moderate heart failure Cardiac enlargement electrocardiogram abnormalities	Tamponade/Severe heart failure/Severe constrictive pericarditis
Esophagus	None	Mild fibrosis Slight difficulty in swallowing solids No pain on swallowing	Unable to take solid food normally Swallowing semi‐solid food Dilatation may be indicated	Severe fibrosis Able to swallow only liquids May have pain on swallowing Dilatation required	Necrosis/Perforation Fistula
Small/large intestine	None	Mild diarrhoea Mild cramping Bowel movement 5 times daily Slight rectal discharge or bleeding	Moderate diarrhoea and colic Bowel movement > 5 times daily Excessive rectal mucus or intermittent bleeding	Obstruction or bleeding requiring surgery	Necrosis/Perforation Fistula
Liver	None	Mild lassitude Nausea, dyspepsia Slightly abnormal liver function	Moderate symptoms Some abnormal liver function tests Serum albumin normal	Disabling hepatitic insufficiency Liver function tests grossly abnormal Low albumin Oedema or ascites	Necrosis/Hepatic coma or encephalopathy
Kidney	None	Transient albuminuria No hypertension Mild impairment of renal function Urea 25 to 35 mg% Creatinine 1.5 to 2.0 mg% Creatinine clearance > 75%	Persistent moderate albuminuria (2+) Mild hypertension No related anaemia Moderate impairment of renal function Urea > 36 to 60 mg% Creatinine clearance (50% to 74%)	Severe albuminuria Severe hypertension Persistent anaemia (< 10 g%) Severe renal failure Urea > 60 mg% Creatinine > 4.0 mg% Creatinine clearance < 50%	Malignant hypertension Uraemic coma/Urea > 100%
Bladder	None	Slight epithelial atrophy Minor telangiectasia (microscopic haematuria)	Moderate frequency Generalised telangiectasia Intermittent macroscopic haematuria	Severe frequency and dysuria Severe generalised telangiectasia (often with petechiae) Frequent haematuria Reduction in bladder capacity (< 150 cc)	Necrosis/Contracted bladder (capacity < 100 cc) Severe haemorrhagic cystitis
Bone	None	Asymptomatic No growth retardation Reduced bone density	Moderate pain or tenderness Growth retardation Irregular bone sclerosis	Severe pain or tenderness Complete arrest of bone growth Dense bone sclerosis	Necrosis/Spontaneous fracture
Joint	None	Mild joint stiffness Slight limitation of movement	Moderate stiffness Intermittent or moderate joint pain Moderate limitation of movement	Severe joint stiffness Pain with severe limitation of movement	Necrosis/Complete fixation

Date form completed(dd/mm/yyyy)
Name/ID of person extracting data

	Description Include comparative information for each group (i.e. intervention and controls) if available
Population description (from which study participants are drawn)
Setting (including location and social context)
Inclusion criteria
Exclusion criteria
Notes

	Descriptions as stated in report/paper
Aim of study
Design(e.g. parallel, crossover, cluster)
Period of study (year of study)
Notes

Domain	Risk of bias			Support for judgement
Domain	Low risk	High risk	Unclear	Support for judgement
Random sequence generation (selection bias)
Allocation concealment (selection bias)
Blinding of participants and personnel Subjective outcomes (performance bias)
Blinding of participants and personnel Objective outcomes (if required)
Blinding of outcome assessment Subjective outcomes (detection bias)				Outcome group: All/
Blinding of outcome assessment Objective				Outcome group:
Incomplete outcome data Subjective outcomes – short term (attrition bias)
Incomplete outcome data Subjective outcomes –long term (attrition bias)
Incomplete outcome data Objective outcomes – short term
Incomplete outcome data Objective outcomes – long term
Selective outcome reporting? (reporting bias)
Other bias (baseline imbalances, early study ending …)

	Description as stated in report/paper
Total no. randomised (or total pop. at start of study for NRCTs, or total per cluster)
Total no. analysed Loss to FU beschrijven
Age
Sex
Race/Ethnicity
Smoking/non smoking
Severity of illness (type of malignancy and grade of radiation proctitis)
Weight
Co‐morbidities
Other relevant sociodemographics

	Description as stated in report/paper Intervention	Control
Group name
No. randomised to group (specify whether no. people or clusters)
Description (which category of agent used, method of administration and dose)
Duration of treatment period
Providers (e.g. no., by physiotherapist , unclear, profession, training, ethnicity etc. if relevant)
Co‐interventions
Notes:

	Description as stated in report/paper
Outcomes
Time points measured

	Description as stated in report/paper
Results	Intervention result	SD (or other variance)	Control result	SD (or other variance)

	Overall results		SE (or other variance)

No. participants	Intervention		Control

Outcome or subgroup title	No. of studies	Statistical method	Effect size
1 Eradication of all telangiectasias	1	Risk Ratio (M‐H, Random, 95% CI)	Totals not selected
2 Complications ‐ Minor	1	Risk Ratio (M‐H, Random, 95% CI)	Totals not selected
3 Complications ‐ Major	1	Risk Ratio (M‐H, Random, 95% CI)	Totals not selected
4 Relapse	1	Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

Outcome or subgroup title	No. of studies	Statistical method	Effect size
1 Bowel subset score (IBDQ‐B) at 6 months (higher score indicates better)	1	Mean Difference (IV, Random, 95% CI)	Totals not selected
1.1 Gastroenterologist group vs booklet group	1	Mean Difference (IV, Random, 95% CI)	0.0 [0.0, 0.0]
1.2 Nurse group vs booklet group	1	Mean Difference (IV, Random, 95% CI)	0.0 [0.0, 0.0]
1.3 Gastroenterologist group vs nurse group	1	Mean Difference (IV, Random, 95% CI)	0.0 [0.0, 0.0]
2 Change in rectal SOMA‐LENT at 6 months	1	Mean Difference (IV, Random, 95% CI)	Totals not selected
2.1 Gastroenterologist group vs booklet group	1	Mean Difference (IV, Random, 95% CI)	0.0 [0.0, 0.0]
2.2 Nurse group vs booklet group	1	Mean Difference (IV, Random, 95% CI)	0.0 [0.0, 0.0]
2.3 Gastroenterologist group vs nurse group	1	Mean Difference (IV, Random, 95% CI)	0.0 [0.0, 0.0]
3 Change in small intestine SOMA‐LENT at 6 months	1	Mean Difference (IV, Random, 95% CI)	Totals not selected
3.1 Gastroenterologist group vs booklet group	1	Mean Difference (IV, Random, 95% CI)	0.0 [0.0, 0.0]
3.2 Nurse group vs booklet group	1	Mean Difference (IV, Random, 95% CI)	0.0 [0.0, 0.0]
3.3 Gastroenterologist group vs nurse group	1	Mean Difference (IV, Random, 95% CI)	0.0 [0.0, 0.0]
4 SF12: change in physical component score at 6 months	1	Mean Difference (IV, Random, 95% CI)	Totals not selected
4.1 Gastroenterologist group vs booklet group	1	Mean Difference (IV, Random, 95% CI)	0.0 [0.0, 0.0]
4.2 Nurse group vs booklet group	1	Mean Difference (IV, Random, 95% CI)	0.0 [0.0, 0.0]
4.3 Gastroenterologist group vs nurse group	1	Mean Difference (IV, Random, 95% CI)	0.0 [0.0, 0.0]
5 SF12: change in mental component score at 6 months	1	Mean Difference (IV, Random, 95% CI)	Totals not selected
5.1 Gastroenterologist group vs booklet group	1	Mean Difference (IV, Random, 95% CI)	0.0 [0.0, 0.0]
5.2 Nurse group vs booklet group	1	Mean Difference (IV, Random, 95% CI)	0.0 [0.0, 0.0]
5.3 Gastroenterologist group vs nurse group	1	Mean Difference (IV, Random, 95% CI)	0.0 [0.0, 0.0]

	Description as stated in report/paper
Results	Intervention		Comparison
	No. events	No. participants	No. events	No. participants

Outcome or subgroup title	No. of studies	Statistical method	Effect size
1 Mean reduction of endoscopic score after treatment	1	Mean Difference (IV, Random, 95% CI)	Totals not selected
2 Mean number of days of rectal bleeding after treatment	1	Mean Difference (IV, Random, 95% CI)	Totals not selected
3 Mean haemoglobin levels at the end of treatment period (5 weeks)	1	Mean Difference (IV, Random, 95% CI)	Totals not selected

Outcome or subgroup title	No. of studies	Statistical method	Effect size
1 Endoscopy scores Gilinsky scale grade 6‐9 at 8 weeks	1	Risk Ratio (M‐H, Random, 95% CI)	Totals not selected
2 Endoscopy scores Gilinsky scale grade 6‐9 at 16 weeks	1	Risk Ratio (M‐H, Random, 95% CI)	Totals not selected
3 Number of participants with complications	1	Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

Outcome or subgroup title	No. of studies	Statistical method	Effect size
1 Severe bleeding episodes	1	Risk Ratio (M‐H, Random, 95% CI)	Totals not selected
2 Mean number of severe bleeds after first year	1	Mean Difference (IV, Random, 95% CI)	Totals not selected
3 Mean haematocrit after first year	1	Mean Difference (IV, Random, 95% CI)	Totals not selected
4 Mean units of blood transfused after first year	1	Mean Difference (IV, Random, 95% CI)	Totals not selected

Outcome or subgroup title	No. of studies	Statistical method	Effect size
1 Improvement of endoscopic appearance	1	Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
2 Reduction of rectal bleeding	1	Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
3 Poor tolerance of enema	1	Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Clinical improvement	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
2 Endoscopic improvement	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

Outcome or subgroup title	No. of studies	Statistical method	Effect size
1 Rectal bleeding score < 2 after 1 year	1	Risk Ratio (M‐H, Random, 95% CI)	Totals not selected
2 Diarrhoea score < 2 after 1 year	1	Risk Ratio (M‐H, Random, 95% CI)	Totals not selected
3 No rectal erythema after 1 year	1	Risk Ratio (M‐H, Random, 95% CI)	Totals not selected
4 No rectal ulceration after 1 year	1	Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Grade of radioproctitis after treatment Grade I	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected
2 Cured ‐ better after 37 days	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

Methods	Single‐centre, prospective, 3‐arm, non‐blinded, randomised controlled trial Duration of study: 2007 to 2012 Follow‐up: 12 months
Participants	218 people (aged over 18 years) who had troublesome, persisting gastrointestinal symptoms that started during or after radiotherapy given with curative intent for histologically proven prostatic, bladder, vulval, vaginal, cervical, endometrial, anal, or rectal malignant neoplasia or para‐aortic irradiation for metastatic disease from any of those primary sites or the testis. Radiotherapy should have been completed at least 6 months before enrolment. Sex (M/F): 168/50 Similar groups with respect to demographic and clinical characteristics. Dropouts: 25 (11%) participants were withdrawn from the study before completion.
Interventions	Group I: gastroenterologist group (management according to the algorithm by a consultant gastroenterologist) (n = 70) Group II: nurse group (management according to the algorithm by a specially trained research nurse) (n = 80) Group III: booklet group (detailed advice booklet on self management of bowel symptoms) (n = 68) N.B.: participants in the booklet group whose symptoms continued 6 months after recruitment were offered consultation with the gastroenterologist and, if appropriate, investigation and treatment. Participants in the nurse‐led care group were crossed over to the gastroenterologist‐led care group if they had gastrointestinal issues that were beyond the scope of the algorithm.
Outcomes	Change in IBDQ‐B subset score, QoL, anxiety and depression scores, and pelvic symptom scores
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"Using a computer‐generated randomization sequence and random permuted blocks, we allocated patients to one of the three groups."
Allocation concealment (selection bias)	Low risk	Central randomisation was performed (the randomisation office of the Institute of Cancer Research, which had no further involvement in the trial, generated the randomisation sequence)
Blinding (performance bias and detection bias)	High risk	Non‐blinded trial (interventions are impossible to blind), high risk of performance and detection bias
Blinding of outcome assessment subjective	High risk	Non‐blinded trial (interventions are impossible to blind), high risk of subjective outcome assessment (participant‐reported outcomes)
Blinding of outcome assessment objective	Unclear risk	Study item not assessed in the trial
Incomplete outcome data: subjective outcomes	Unclear risk	The study was analysed on an ITT basis, but 25 (11%) participants were withdrawn from the study before completion (if a participant had a recurrence of cancer requiring treatment or was admitted to hospital for gastrointestinal symptoms, they were withdrawn from the study). However, it is unclear what happened with the data of these withdrawn participants (such as LOCF, etc.). The number of withdrawn participants is not similar between the groups, so this could have led to bias
Incomplete outcome data: objective outcomes	Unclear risk	Study item not assessed in the trial
Selective reporting (reporting bias)	Low risk	This trial is registered with ClinicalTrials.gov, number NCT00737230. All assumed outcomes except cost‐effectiveness are reported in the paper. The authors state that a full cost‐effectiveness analysis was embedded within the trial, but that the results will be reported separately
Other bias	Low risk	No indications of other bias

Methods	Quasi‐randomised study. Duration of study: 1990 to 2000 Follow‐up: 12 months
Participants	60 people with chronic radiation proctopathy and cytologically proven prostatic carcinoma staged to the TNM classification as T2N0M0 stage Sex (M/F): not reported Comparable groups with respect to age and previous treatment time Dropouts (after 3 months) IG: 3 CG: 12. Reasons for dropouts not explained
Interventions	Intervention: metronidazole (3 x 400 mg orally per day), mesalazine (3 x 1 g orally per day), and betamethasone enema (once a day) Comparator: same doses of mesalazine and betamethasone enema, but without metronidazole
Outcomes	The efficacy of metronidazole was assessed using rectal bleeding, diarrhoea, and endoscopy
Notes	The incidence of diarrhoea, rectal bleeding, ulcers and oedema was significantly reduced in the metronidazole group up to 12 months after treatment. No QoL
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk	Participants were allocated to 2 groups according to the date on which their treatment began (quasi random)
Allocation concealment (selection bias)	High risk	Allocation concealment could be foreseen, as participants were allocated according to the date on which their treatment began
Blinding (performance bias and detection bias)	High risk	Participants in intervention group received their medication orally, it seemed the intervention had not been blinded
Blinding of outcome assessment subjective	High risk	Participants in intervention group received their medication orally, it seemed the intervention had not been blinded
Blinding of outcome assessment objective	Low risk	Objective participant response was documented by rectal bleeding score and diarrhoea score. The participants were scored the same way before and after treatment. The same physician interviewed participants once a week during the treatment period, however unclear whether outcome assessment was blinded
Incomplete outcome data: subjective outcomes	High risk	Assessments before and after 4 weeks treatment: I: N = 30; C: N = 30. After 3 months dropouts: I: N = 3; C: = 12 (25%). After 1 year dropouts: I: N = 6; C: N = 18 (40%). After 2 years dropouts: I: N = 11; C: N = 20 (52%) Reasons for lost to follow‐up not reported. Not all participants were followed up for the same period of time: "The longest follow‐up period after 4 weeks treatment was 3 years, whereas the shortest one was 2 years."
Incomplete outcome data: objective outcomes	High risk	Assessments before and after 4 weeks treatment: I: N = 30; C: N = 30. After 3 months dropouts: I: N = 3; C: = 12 (25%). After 1 year dropouts: I: N = 6; C: N = 18 (40%). After 2 years dropouts: I: N = 11; C: N = 20 (52%) Reasons for lost to follow‐up not reported. Not all participants were followed up for the same period of time: "The longest follow‐up period after 4 weeks treatment was 3 years, whereas the shortest one was 2 years."
Selective reporting (reporting bias)	Unclear risk	No study protocol available, however all outcomes mentioned in the methods sections have been addressed in the results section
Other bias	Low risk	No indications of other bias

PERMALINK

Non‐surgical interventions for late rectal problems (proctopathy) of radiotherapy in people who have received radiotherapy to the pelvis

Fleur T van de Wetering

Leen Verleye

H Jervoise N Andreyev

Jane Maher

Joan Vlayen

Bradley R Pieters

Geertjan van Tienhoven

Rob JPM Scholten

Abstract

Background

Objectives

Search methods

Selection criteria

Data collection and analysis

Main results

Authors' conclusions

Plain language summary

Background

Description of the condition

Description of the intervention

How the intervention might work

Why it is important to do this review

Objectives

Methods

Criteria for considering studies for this review

Types of studies

Types of participants

Types of interventions

Types of outcome measures

Primary outcomes

Secondary outcomes

Search methods for identification of studies

Electronic searches

Searching other resources

Data collection and analysis

Data extraction and management

Assessment of risk of bias in included studies

Measures of treatment effect

Dealing with missing data

Assessment of heterogeneity

Assessment of reporting biases

Data synthesis

Subgroup analysis and investigation of heterogeneity

Sensitivity analysis

Results

Description of studies

Results of the search

1.

Included studies

Study design

Types of interventions

Participant characteristics

Excluded studies

Risk of bias in included studies

2.

3.

Allocation

Blinding

Incomplete outcome data

Selective reporting

Other potential sources of bias

Effects of interventions

1. Treatments for rectal bleeding

1. Treatment for bleeding only.

Short‐chain fatty acid (SCFA) versus placebo

1.1. Analysis.

1.2. Analysis.

Sucralfate versus placebo following argon plasma coagulation

2.1. Analysis.

2.2. Analysis.

2.3. Analysis.

Endoscopic bipolar electrocoagulation versus heater probe

3.1. Analysis.

3.2. Analysis.

3.4. Analysis.

Bipolar eletrocoagulation versus argon plasma coagulation

4.1. Analysis.

4.2. Analysis.

Methods	RCT. Duration of study: participants were recruited between June 2003 and March 2006. Follow‐up: 52 weeks
Participants	122 adults with chronic radiation proctopathy or proctosigmoiditis were included if all of the following criteria were met: radiotherapy for a pelvic tumour completed at least 3 months before enrolment, the presence of rectal bleeding, radiation‐induced telangiectasia in the rectum or sigmoid, or both. The baseline characteristics of the 2 groups were comparable
Interventions	Intervention: endoscopic APC followed by oral sucralfate (6 g twice daily) for 4 weeks (n = 60) Comparator: APC with placebo
Outcomes	Changes in chronic radiation proctopathy severity score, changes in endoscopy scores, and complications
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer random number generator was used
Allocation concealment (selection bias)	Low risk	The study investigators were not involved in the preparation of the computer‐generated allocation code or in the generation of consecutively numbered containers with the study medication (drug or identical‐appearing placebo tablets)
Blinding (performance bias and detection bias)	Low risk	Study participants and investigators remained blinded to group assignment until the conclusion of the study, which was the visit at week 52
Blinding of outcome assessment subjective	Low risk	Study participants and investigators remained blinded to group assignment until the conclusion of the study, which was the visit at week 52
Blinding of outcome assessment objective	Unclear risk	Study item not assessed in the trial
Incomplete outcome data: subjective outcomes	Low risk	ITT analysis was used throughout the study. Only 3 dropouts in intervention group
Incomplete outcome data: objective outcomes	Unclear risk	Study item not assessed in the trial
Selective reporting (reporting bias)	Unclear risk	No registered study protocol available, however, all outcomes mentioned in the methods sections have been addressed in the results section
Other bias	Low risk	No indications of other bias

Methods	Multi‐centre RCT. Duration of study: unclear Follow‐up: 5 years
Participants	120 people with rectal late radiation tissue injury. The diagnosis must have been present for more than 3 months and responded insufficiently to other therapies Sex (F/M): 106/14 Comparable groups with respect to demographic and clinical characteristics Dropouts: at 1 year, 5 participants (4%) had died and 9 (8%) had been lost to follow‐up
Interventions	Intervention: 100% oxygen at 2.0 ATA for 90 min, once daily, 5 times weekly Comparator: 21% oxygen (normal air) at 1.1 ATA for 90 min, once daily, 5 times weekly
Outcomes	The primary outcome measures were the SOMA‐LENT score and standardised clinical assessment. The secondary outcome was the change in QoL
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Biostatisticians at the University of South Carolina generated the randomisation sequence, which was uploaded into, and concealed within, the study database software. Participants were randomly assigned (1:1) to receive HBO or normobaric air, using a "blocking" process. The block size was 4 and was equally stratified with 2 of each treatment options (A or B).
Allocation concealment (selection bias)	Low risk	Not reported on how the allocation concealment was done, but referring physicians (who also were the outcome assessors) as well as the participants are described as being unaware of the allocation
Blinding (performance bias and detection bias)	Low risk	Both assessors and participants had been blinded
Blinding of outcome assessment subjective	Low risk	Participants had been blinded (sham treatment for control group)
Blinding of outcome assessment objective	Low risk	Referring physicians (who were also the outcome assessors) are described as being unaware of the allocation
Incomplete outcome data: subjective outcomes	Unclear risk	At 1 year (minimum follow‐up time), 5 participants (4%) had died and 9 (8%) had been lost to follow‐up. No clear reasons for these dropouts provided (reasons only well described for dropouts before randomisation)
Incomplete outcome data: objective outcomes	Unclear risk	At 1 year (minimum follow‐up time), 5 participants (4%) had died and 9 (8%) had been lost to follow‐up. No clear reasons for these dropouts provided (reasons only well described for dropouts before randomisation)
Selective reporting (reporting bias)	Unclear risk	No study protocol available, however all outcomes reported in the methods section have been addressed in the results section
Other bias	Low risk	No indications of other bias

Methods	Double‐blind RCT. Duration of study: not reported Follow‐up: at 90 days
Participants	19 people with radiation proctopathy (at least 2 symptoms with a severity score of 3 on at least a weekly basis), however, one participant from each group did not take a single dose of study medication and were therefore excluded from analysis (n = 17). Sex (M/F): 15/2 Comparable groups with respect to baseline RPSAS Dropouts: 1 (reason not explained), but included in final analysis
Interventions	Intervention: Retinol palmitate 10,000 IU by mouth for 90 days Comparator: Identical placebo capsules
Outcomes	Reduction in the RPSAS scores after 30 days for 90 days
Notes	5 placebo non‐responders who were crossed over to Tx responded
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Random number system
Allocation concealment (selection bias)	Low risk	Neither the investigators nor participants were aware of who was receiving intervention treatment or placebo
Blinding (performance bias and detection bias)	Low risk	Double‐blinded study (participants and investigators were blinded)
Blinding of outcome assessment subjective	Low risk	Double‐blinded study (participants and investigators were blinded).
Blinding of outcome assessment objective	Low risk	Double‐blinded study
Incomplete outcome data: subjective outcomes	Low risk	ITT analyses was performed, number of drop‐outs reported, however reason for drop‐out not reported
Incomplete outcome data: objective outcomes	Low risk	ITT analyses was performed, number of drop‐outs reported, however reason for drop‐out not reported
Selective reporting (reporting bias)	Unclear risk	No study protocol available, however all outcomes mentioned in the methods section had been addressed in the results section
Other bias	Low risk	No indications of other bias

Methods	RCT. Duration of study: not reported Follow‐up, days (mean): 770 versus 704. However, major outcomes were compared at 12 months after randomisation
Participants	21 people with CRP, RT 2 years ago, failed medical treatment, and per rectum bleeds at least 3 times a week Sex (M/F): 18/3 Similar groups with respect to demographic and clinical characteristics Dropouts: none
Interventions	Intervention: Heater probe (9) Comparator: Bipolar electrocoagulation probe (12)  Treatment with the same probe till the bleeding stopped, mean 4 treatments
Outcomes	Sigmoidoscopies 4 to 6 weekly till bleeds stopped and follow‐up requirements  Fall in severe bleeds and the need for follow‐up in both groups. Significant decrease in both groups
Notes	No side effects reported. QoL informally assessed with participant responses, which improved with treatment
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Randomisation method not described
Allocation concealment (selection bias)	Unclear risk	Insufficient information to permit judgement
Blinding (performance bias and detection bias)	Low risk	Unclear whether participants had been blinded, but as both arms received a similar active intervention it seems unlikely that this would have introduced performance bias
Blinding of outcome assessment subjective	Low risk	Unclear whether participants had been blinded, but as both arms received a similar active intervention it seems unlikely that this would have introduced bias for subjective outcomes
Blinding of outcome assessment objective	Unclear risk	Managing physician was blinded to treatment, however, it was not clearly described whether the research nurse who evaluated and followed all participants was blinded.
Incomplete outcome data: subjective outcomes	Low risk	No loss to follow‐up
Incomplete outcome data: objective outcomes	Low risk	No loss to follow‐up
Selective reporting (reporting bias)	Unclear risk	Study was published in 1997, so no study protocol available. However, all outcomes prespecified in the methods section were reported in the results section
Other bias	Low risk	No indications of other bias

Methods	Double‐blind RCT. Duration of study: 4 weeks Follow‐up: at 4 weeks
Participants	37 people with CRP, 36 cervix and 1 prostate  Dx confirmed on symptoms (graded) and mucosal appearance (graded). Cumulative score
Interventions	Intervention (group 1): oral sulfasalzine 1 g tds + prednisolone enemas 20 mg bd for 4 weeks  18 people entered, 1 dropped out (unclear why), and 2 did not tolerate the drug  Comparator (group 2): oral placebo + rectal sucralfate 2 g bd for 4 weeks  19 entered, with 2 dropouts (unclear why)
Outcomes	Groups' responses:  1. Clin improvement 8/15, P < 0.01; endoscopic changes 7/15, P < 0.01  2. Clin improvement 16/17, P < 0.001 (clin response better P < 0.05); endoscopic changes 12/17, P < 0.001
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Randomised, but methods of sequence generation not described
Allocation concealment (selection bias)	Unclear risk	Insufficient information to permit judgement
Blinding (performance bias and detection bias)	Low risk	Both the participants and the physician were blinded to the type of therapy
Blinding of outcome assessment subjective	Low risk	Participants were blinded, and filled out their own diary
Blinding of outcome assessment objective	Unclear risk	Not described
Incomplete outcome data: subjective outcomes	Unclear risk	Group 1: 1 drop out, reasons not discussed. 2 others dropped out because they did not tolerate the drug. Group 2: 2 drop outs, reasons not discussed
Incomplete outcome data: objective outcomes	Unclear risk	Group 1: 1 drop out, reasons not discussed. 2 others dropped out because they did not tolerate the drug. Group 2: 2 drop outs, reasons not discussed
Selective reporting (reporting bias)	Unclear risk	Study was published in 1991, so no study protocol available. However, all outcomes prespecified in the methods section were reported in the results section
Other bias	Low risk	No indications of other bias

Methods	RCT. Duration: 2005 to 2008 Follow‐up: 1 year
Participants	30 people with recurrent rectal bleeding, started 6 months after radiotherapy Mean age 67.4 (SD 11.8); 10% grade 1; 43.3% grade 2; 26.7% grade 3; 20% grade 4. Comparable groups Dropouts: APC group: 2 (1 died, 1 refused further therapy after successful reduction of her rectal bleeding (score from 4 to 2 points))
Interventions	Intervention: BEC Comparator: APC
Outcomes	Efficacy, complications, relapse
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Randomisation was done by sequential opening of numbered, opaque envelopes
Allocation concealment (selection bias)	Low risk	Randomisation was done by sequential opening of numbered, opaque envelopes
Blinding (performance bias and detection bias)	Low risk	Unclear whether participants had been blinded, but as both arms received a similar active intervention, it seems unlikely that this would have introduced performance bias
Blinding of outcome assessment subjective	Low risk	Unclear whether participants had been blinded, but as both arms received a similar active intervention, it seems unlikely that this would have introduced performance bias
Blinding of outcome assessment objective	Low risk	No information on blinding. However, it is unlikely that objective outcomes were influenced
Incomplete outcome data: subjective outcomes	Low risk	2 participants (APC group) did not complete the treatment, despite clinical improvement. The clinical score of 1 of these participants improved from 4 to 3 points, but he died due to pneumonia. The other refused further therapy after successful reduction of her rectal bleeding (score from 4 to 2 points)
Incomplete outcome data: objective outcomes	Low risk	2 participants (APC group) did not complete the treatment, despite clinical improvement. The clinical score of 1 of these participants improved from 4 to 3 points, but he died due to pneumonia. The other refused further therapy after successful reduction of her rectal bleeding (score from 4 to 2 points)
Selective reporting (reporting bias)	Low risk	Protocol was registered on ClinicalTrials.gov (NCT00725244), and all prespecified outcomes have been reported in the article
Other bias	Low risk	There was no significant difference between the groups regarding pelvic cancer, age, gender, aspirin use, previous medical therapy, or disease severity

Methods	RCT. Duration of study: August 2005 to May 2007 Follow‐up: 6.3 (range 2 to 18) months
Participants	102 participants with rectal bleeding as a result of chronic haemorrhagic radiation proctopathy, following radiotherapy for carcinoma of the cervix Mean age: 50.8 ± 5.0 years in Group 1 (formaline dab); 51.8 ± 5.1 years in Group 2 (sacralfate steroid enema) Symptoms: Bleeding per rectum: 51 (100) versus 51 (100) Rectal pain: 37 (72.6) versus 35 (68.6) Loose stools: 5 (9.8) versus 4 (7.8) Tenesmus: 10 (19.6) versus 8 (15.7) Urgency: 8 (15.7) versus 11 (21.6) Faecal incontinence: 0 versus 0 Comparable groups Dropouts: none
Interventions	Intervention: formalin dab Comparator: sucralfate steroid retention enema
Outcomes	Symptom score and sigmoidoscopic grade
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Permuted block randomisation, fixed
Allocation concealment (selection bias)	High risk	An open list of random numbers was used, participants or investigators could possibly foresee assignments and thus introduce selection bias (information retrieved from registered protocol)
Blinding (performance bias and detection bias)	High risk	No information on blinding described, but blinding most likely impossible considering the characteristics of the interventions
Blinding of outcome assessment subjective	High risk	No information on blinding described, but blinding most likely impossible considering the characteristics of the interventions
Blinding of outcome assessment objective	Low risk	Sigmoidoscopy was repeated by a consultant who was blinded to the treatment received, and the sigmoidoscopic grading was recorded
Incomplete outcome data: subjective outcomes	Low risk	None were lost to follow‐up
Incomplete outcome data: objective outcomes	Low risk	None were lost to follow‐up
Selective reporting (reporting bias)	Unclear risk	Trial was registered in the Clinical Trials Registry of India (CTRI/2011/12/002184). Adverse events were not pre‐specified in the protocol. In the protocol it was also stated that all participants were female, which was not reported in the article (gender was not mentioned at all)
Other bias	Low risk	No indications of other bias

Methods	Double‐blind RCT. Duration of study: 1992 to 1994 Follow‐up: at 5 weeks
Participants	19 people with Grade III CRP Pourquier classification, persistent symptoms for a minimum of 12 months Mean age 59 years (range 36 to 75), Sex (F/M): 18:1 Both groups had equivalent baseline characteristics  1 (SCFA arm) and 2 (placebo) arm dropped out due to heavy per rectum bleeds
Interventions	Intervention: SCFA enema (as per Harig preparation 60 ml bd, 5 wks) Comparator: placebo
Outcomes	No. of days of per rectum bleeding, colonoscopic scores, and DNA and protein content  Reduction of days/week of bleeding in SCFA P = 0.001. Increased Hb in SCFA vs placebo P = 0.02. Colonoscopic scores reduced in both arms, but more in SCFA P = 0.02
Notes	Side effects were recorded and there were none. No QoL assessment either
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Participants were randomly allocated, but randomisation method not described
Allocation concealment (selection bias)	Unclear risk	Insufficient information to permit judgement
Blinding (performance bias and detection bias)	Low risk	Study was double blinded "allocated to receive either SCFA enemas or an indentically appearing saline isotonic solution..."
Blinding of outcome assessment subjective	Low risk	Participants were blinded to the treatment
Blinding of outcome assessment objective	Low risk	Outcome assessors were blinded to the treatment "...by two well‐trained physicians who were not aware of the study phase or the nature of the treatment administered." "All biopsy specimens were interpreted in a random and blinded manner by two pathologists."
Incomplete outcome data: subjective outcomes	Unclear risk	3 drop outs: 1 in the intervention group and 2 in the placebo group. Reasons for dropouts not discussed
Incomplete outcome data: objective outcomes	Unclear risk	3 drop outs; 1 in the intervention group and 2 in the placebo group. Reasons for dropouts not discussed
Selective reporting (reporting bias)	Unclear risk	Study was published in 1999, so no study protocol available. However, all outcomes prespecified in the methods section were reported in the results section
Other bias	Low risk	No indications of other bias

Methods	RCT. Duration of study: October 2010 to January 2012 Follow‐up: at 2 months
Participants	50 participants with symptomatic haemorrhagic radiation proctopathy for more than 6 months without complications of rectal stricture, deep ulceration, fistula formation, and sepsis. People who were allergic to ciprofloxacin and metronidazole were excluded. Mean age (range): 64 (31 to 85) versus 64 (27 to 80) Comparable groups Dropouts: 3 participants were lost to follow‐up: 1 in the irrigation group and 2 in the formalin group
Interventions	Intervention: Colonic irrigation Comparator: Formalin application
Outcomes	Episodes of rectal bleeding (days/week), frequency (times/week), urgency (days/week), diarrhoea (days/week), and tenesmus (days/ week) before and after 8 weeks of treatment
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"Patients were consecutively assigned to each treatment group according to a computer‐generated randomization list.”
Allocation concealment (selection bias)	Unclear risk	Insufficient information to permit judgement
Blinding (performance bias and detection bias)	High risk	No information on blinding, no placebo
Blinding of outcome assessment subjective	High risk	No information on blinding, no placebo, and different active interventions. As most outcomes were participant reported, high risk of bias.
Blinding of outcome assessment objective	Low risk	Repeated sigmoidoscopy (VRS) and haematocrit values and number of people requiring packed red blood cell transfusion are ‘objective’, however these outcomes are not of interest to this review
Incomplete outcome data: subjective outcomes	Low risk	3 participants were lost to follow‐up: 1 in the irrigation group and 2 in the formalin group
Incomplete outcome data: objective outcomes	Low risk	3 participants were lost to follow‐up: 1 in the irrigation group and 2 in the formalin group
Selective reporting (reporting bias)	High risk	No study protocol available. Outcomes every 2 months after treatment not reported
Other bias	Low risk	There are no indications of other bias

Methods	Double‐blind, randomised, placebo‐controlled, cross‐over pilot. Duration of study: not reported Follow‐up: at 5 weeks
Participants	15 people: 12 prostate, 1 cervix, and 2 rectal carcinomas diagnosed as chronic radiation proctopathy (over 2 months). Mean age 67.7 years, Sex (F or M) 2 to 13 Group comparability: unclear (no separated group characteristics presented) Dropouts: 3 participants were lost without explanation, so 12 completed
Interventions	Intervention: Butyric acid enema (60 ml containing 40 mmol butyric acid) twice a day Comparator: Normal saline placebo enema Interventions for 2 weeks with a 1‐week wash‐out period and then alternate enema
Outcomes	Scores of symptoms, endoscopy, and biopsy. No significant change in the SCFA or placebo arm for any of the 3 scores
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Participants were randomly allocated, but randomisation method not described
Allocation concealment (selection bias)	Unclear risk	Insufficient information to permit judgement
Blinding (performance bias and detection bias)	Low risk	Study was double blind study
Blinding of outcome assessment subjective	Low risk	Both participants and outcome assessors were unaware of the treatment given
Blinding of outcome assessment objective	Unclear risk	Both participants and outcome assessors were unaware of the treatment given
Incomplete outcome data: subjective outcomes	Unclear risk	3 dropouts (not specified why and in which group)
Incomplete outcome data: objective outcomes	Unclear risk	3 dropouts (not specified why and in which group)
Selective reporting (reporting bias)	Unclear risk	Study was published in 1997, so no study protocol available. However, all outcomes prespecified in the methods section were reported in the results section
Other bias	Low risk	No indications of other bias