MASCC recommendations on the management of constipation in patients with advanced cancer

The Oxford Concise Medical Dictionary defines constipation as “a condition in which bowel evacuations occur infrequently, or in which the faeces are hard and small, or where passage of faeces causes difficulty or pain” [5]. Indeed, the term “constipation” means different things to different people [6], which has implications for the management of constipation. For example, a Swedish survey of the general population reported that 23.8% female respondents and 24.3% male respondents considered “straining in connection with bowel movement” was not indicative of constipation [7]. Similarly, a Korean survey of constipated individuals reported that 48.3% respondents considered “using fingers to help empty your bowel” was not indicative of constipation [8].

The Rome Foundation categorises constipation as being either functional constipation or secondary constipation [9]. Functional constipation is a heterogeneous phenomenon and can be subdivided into normal-transit constipation, slow-transit constipation, and defaecatory or rectal evacuation disorders [9]. The risk factors for functional constipation include a positive family history, low levels of dietary fibre, and low levels of physical activity [9]. Secondary constipation may have a variety of different causes, including gastrointestinal diseases (e.g. diverticulosis, irritable bowel syndrome), neurological/psychiatric diseases (e.g. dementia, depression), other systemic diseases (e.g. diabetes, hypothyroidism), and/or medications (e.g. antacids, diuretics) [6].

Opioid-induced constipation is a subtype of secondary constipation, with a distinct pathophysiology (Box 1), which is primarily a peripheral effect and primarily a mu-opioid receptor effect [10]. Opioid-induced constipation appears to be more common in patients with cancer pain (than non-malignant pain) [9], may be influenced by genetic factors [11], and may be influenced by the type of opioid utilised [12]. However, opioid-induced constipation does not appear to be particularly influenced by the dose of opioid utilised [13]. Opioid-induced constipation tends to be a chronic side effect, although some patients appear to develop tolerance over time [13]. Unsurprisingly, the pathophysiology has important implications in relation to the management [14].

Constipation is a common problem, with most people experiencing the symptom at some point in their life. The reported mean prevalence of functional constipation in the general population is 14% (range 1.9–40.1%), with a mean prevalence of 15% using self-assessment and a mean prevalence of 6.8% using the Rome Foundation diagnostic criteria (i.e. Rome III diagnostic criteria) [15].

The reported prevalence of constipation in patients with advanced cancer is 32–87%, which reflects variable methods of assessment as well as different study populations (e.g. inpatients generally have a higher prevalence than outpatients) [16]. Similarly, the reported prevalence of opioid-induced constipation in patients with cancer pain is 5–97%, which again reflects variable methods of assessment as well as different study populations [17].

The clinical features vary from patient to patient and include symptoms relating to constipation and/or symptoms resulting from the complications of constipation (i.e. local, systemic).

An American survey of constipated (functional constipation) individuals reported the frequency of different symptoms relating to constipation: straining (79%), “gas” (74%), hard stool (71%), abdominal discomfort (62%), infrequent defaecation (57%), bloating (57%), sensation of incomplete evacuation (54%), abdominal pain (48%), rectal pain (41%), and sudden urge for defaecation (35%) [18]. Moreover, 52% of individuals reported that constipation significantly affected their quality of life (i.e. “somewhat”, “a lot”, or “a great deal”) [18].

The local complications of constipation include faecal impaction, faecal leakage/“overflow diarrhoea”, gastrointestinal obstruction, gastrointestinal perforation, rectal prolapse, haemorrhoids, anal tears, rectal bleeding, urinary tract infection, and urinary retention [16]. Patients may also experience upper gastrointestinal problems such as halitosis, anorexia, early satiety, nausea and vomiting, and gastro-oesophageal reflux (“heartburn”). The systemic complications of constipation include general malaise, confusion/delirium, and headache [16]. Uncommonly, constipation can indirectly cause the death of an individual (e.g. faecal impaction causing gastrointestinal perforation, straining causing pulmonary embolism) [19, 20].

Constipation is associated with various psychological problems, including anxiety, depression, and even “catastrophic thinking” (e.g. “I thought you would die, blow up inside”) [21]. Moreover, it can lead to negative social outcomes such as avoidance of family/friends and avoidance of public places (and so social isolation) [22].

Constipation is also associated with a significant health economic burden: constipation leads to increased “direct costs” (e.g. costs of intervention, costs of healthcare professional) and increased “indirect costs” (e.g. costs of travel, costs of decreased productivity), which impacts on the patient, the health service, and the wider society [23].

Of concern, patients with opioid-induced constipation may reduce the dose of opioid analgesic or even stop the opioid analgesic, in order to overcome the associated distress and discomfort (which usually results in worse cancer pain) [24].

The objectives of assessment are to determine (a) the presence of constipation; (b) the aetiology of constipation (i.e. functional, secondary); and (c) other factors that may influence the choice of intervention. Inadequate assessment may result in initiation of inappropriate interventions (or even contraindicated interventions).

The assessment of constipation involves primarily taking a detailed history and performing an appropriate examination (i.e. clinical assessment). The history should include the questions outlined in Box 2 [6, 9] and the use of the Bristol Stool Chart (to assess stool consistency) [30]. The examination should include an abdominal examination and ideally a digital rectal examination [6, 31].

The Rome Foundation diagnostic criteria are generally employed to diagnose functional gastrointestinal disorders. The Rome IV criteria for functional constipation are shown in Table 2 [9]. Of note, the validation studies of these criteria reported a relatively low sensitivity of 32.2%, a relatively high specificity of 93.6%, and “moderate” reliability [32]. The new Rome IV criteria for opioid-induced constipation are also shown in Table 2 [9]. Currently, there are no published validation studies of these criteria.

Observational studies suggest that plain abdominal X-rays may be helpful in assessing patients with constipation [33], including patients with advanced cancer [34]. However, more specialised gastrointestinal investigations should be limited to patients with “resistant” constipation in patients with advanced cancer (and should be undertaken by gastroenterologists with an interest in constipation) [9].

The management of constipation should be individualised and depends on: (a) the aetiology of constipation (e.g. functional, opioid-induced); (b) the clinical features of constipation (e.g. faecal impaction, stool in rectum); and (c) patient-related factors (e.g. personal preferences, co-morbidities).

Observational studies of patients with advanced cancer have shown an association between constipation and lack of privacy (for defaecation) [35]. Hence, whenever possible, patients should be supported to defaecate in a private toilet, rather than at the bedside (or within the bed). Indeed, bedside commodes should only be used if the patient has mobility problems, and bed pans should only be used if the patient is on strict bed rest. Defaecation may be facilitated by adopting the “correct” toilet position (i.e. semi-squatting position, knees above hips, leaning slightly forward) [36], and this can be facilitated by the use of a foot stool [9]. It is essential that a foot stool is used whenever a raised toilet seat is employed.

Observational studies of patients with advanced cancer have shown an association between constipation and inadequate nutrition (low-fibre diet), inadequate hydration, and decreased physical activity [35]. Moreover, there is evidence that increasing fibre intake, increasing fluid intake (in dehydrated patients), and increasing physical activity may improve functional constipation (but not opioid-induced constipation) [9]. However, these strategies are generally unsuitable for patients with advanced cancer [16], and indeed there is no evidence that these strategies are effective in this group of patients.

In some cases of secondary constipation, it will be possible to treat the underlying condition, or discontinue the constipating medication, and so ameliorate the constipation. It should be noted that reducing the dose of opioid invariably does not improve opioid-induced constipation, although “switching” the opioid sometimes improves opioid-induced constipation [37].

Currently, there is relatively little data on the use of conventional laxatives in patients with advanced cancer. Thus, the Cochrane systematic review of laxatives for the management of constipation in people receiving palliative care (almost exclusively patients with advanced cancer) concluded that “there was no evidence on whether individual laxatives were more effective than others or caused fewer adverse effects” [38].

However, on the basis of data in the general population [39, 40], expert opinion (gastroenterology) [6, 41], expert opinion (palliative care) [42], and extensive clinical experience in patients with advanced cancer, the group recommend the use of polyethylene glycol formulations as the first-line treatment for constipation in patients with advanced cancer. (It should be noted, that none of the studies in the aforementioned Cochrane systematic review involved polyethylene glycol formulations [38].)

Nevertheless, as stated above, the management of constipation should be individualised, since certain practical issues can limit the use of polyethylene glycol formulations (e.g. volume of fluid, taste/consistency), as well as certain adverse effects (e.g. diarrhoea, abdominal distension) [2]. The same considerations apply to all other conventional laxatives (and indeed all interventions for constipation). Table 3 shows the British National Formulary categories of laxatives with examples for each category [43].

Management of secondary constipation primarily involves treating the underlying condition. If this is not possible, or if the patient remains constipated, then management of secondary constipation should follow that of functional constipation (with the exception of opioid-induced constipation—see below).

If patients do not respond to optimal dosing of first-line conventional laxatives, then the options for ongoing management involve the use of conventional laxatives from a different class of drug (Table 3) or a more “specialist” medication (e.g. linaclotide, lubiprostone, and prucalopride) [44]. The latter should generally be prescribed/monitored by clinicians with experience in utilising such specialist medication.

The peripherally acting mu-opioid receptor antagonists (PAMORAs) reverse the effects of opioids on the gastrointestinal tract and so in theory should be the optimum treatment for opioid-induced constipation. A number of different PAMORAs have been developed to treat opioid-induced constipation, including subcutaneous methylnaltrexone [45], oral methylnaltrexone [46], oral naloxegol [47], and oral naldemedine [48]. Studies of PAMORAs have been undertaken in patients with advanced disease and also in patients with cancer pain (as opposed to non-malignant pain). Of note, there is no data to suggest that the efficacy/tolerability of PAMORAs is population-dependent.

The Cochrane systematic review of mu-opioid antagonists for opioid-induced bowel dysfunction in people with cancer and people receiving palliative care concluded that “there is moderate-quality evidence that methylnaltrexone improves bowel function in people receiving palliative care in the short term and over two weeks, and low-quality evidence that it does not increase adverse events” and that “there is moderate-quality evidence to suggest that, taken orally, naldemedine improves bowel function over two weeks in people with cancer and OIBD (opioid-induced bowel dysfunction) but increases the risk of adverse events” [49].

Another systematic review of treatments for opioid-induced constipation concluded “mu-opioid receptor antagonists to be safe and effective for the treatment of OIC” [50]. This systematic review included a larger range of studies than the Cochrane systematic review (23 versus eight randomised controlled trials) [49] and calculated numbers needed to treat (NNT) of 3.4 (95% CI 3–6) for methylnaltrexone, 7 (95% CI 4–26) for naloxegol, and 5 (95% CI 4–8) for naldemedine and a number needed to harm of 20 for all medications [50]. The response to PAMORAs was affected by the dose of opioid (increased efficacy in patients on higher doses) and the previous response to laxatives (increased efficacy in patients “refractory to laxatives”) [50]. It should be noted that the aforementioned NNTs were calculated using different primary endpoints (and so are not directly comparable) [50].

In 2015, the National Institute for Health and Care Excellence (NICE) reviewed the clinical and health economic data on oral naloxegol and concluded that “naloxegol is recommended, within its marketing authorisation, as an option for treating opioid induced constipation in adults whose constipation has not adequately responded to laxatives” [29]. The technology appraisal guidance states that “an inadequate response is defined as opioid-induced constipation symptoms of at least moderate severity in at least 1 of the 4 stool symptom domains (that is, incomplete bowel movement, hard stools, straining or false alarms) while taking at least 1 laxative class for at least 4 days during the previous 2 weeks”.

Conventional opioid antagonists (e.g. naloxone) have been used to manage OIC [51] and have even been incorporated into analgesic formulations to prevent/manage OIC [52]. However, conventional opioid antagonists, particularly in high doses, may reverse analgesia and precipitate withdrawal (and so their role is somewhat limited) [51].

Systematic review data suggests that many patients with opioid-induced constipation do not achieve an adequate response with PAMORAs [49, 50], which may reflect the fact that the constipation is not opioid-induced or more likely that the constipation is multifactorial in aetiology [53]. Thus, some patients will benefit from the addition of a conventional laxative to the PAMORA, whilst others will require the substitution of a conventional laxative for the PAMORA. Of note, lubiprostone and prucalopride have efficacy in opioid-induced constipation (as well as functional constipation) [54, 55].

As constipation is a common adverse effect of opioid analgesics, it is recommended that patients starting opioid analgesics should be co-prescribed conventional laxatives [56, 57]. Surprisingly, there is only limited evidence to support this recommendation [58].

Currently, there is no data to support the co-prescribing of PAMORAs, although in theory this would be a more effective strategy to prevent opioid-induced constipation (than the co-prescribing of conventional laxatives) [14].

The evidence base for rectal interventions (i.e. suppositories, enemas) is somewhat limited, and there are no randomised controlled trials in patients with advanced cancer [38]. Rectal interventions are generally reserved for patients that have not responded to other interventions [6, 42], with suppositories used in patients with stool in the rectum and enemas used for patients with stool in the descending colon [4]. A variety of different formulations are available (Table 3) [4, 43], and there are a number of local and systemic contraindications (e.g. intestinal obstruction, thrombocytopenia) [4].

Trans-anal irrigation (TRI) may have a role in some patients, but the evidence to support its use in clinical practice is limited [59].

A wide range of other interventions have been used to manage constipation in the general population and in patients with advanced cancer, i.e. traditional remedies, complementary therapies, and other pharmacological interventions. Interventions that have been reported as being effective in patients with advanced cancer include acupuncture (and associated techniques) [60], aroma massage [61], petroleum jelly [62], amidotrizoate/diatrizoate [63], and neostigmine [64]. However, the evidence base (and indeed experience) with these interventions is more limited than with previously discussed interventions.

The objectives of re-assessment are to: (a) determine changes in the clinical condition; (b) determine the effectiveness of any intervention; and (c) assess the tolerability of any intervention. Inadequate re-assessment may result in continuation of ineffective interventions (and persistence of constipation).

An expert consensus panel has recommended that the Bowel Function Index be used to assess the effectiveness of interventions for opioid-induced constipation, with a score of > 30 triggering a change in intervention (as this represents an “inadequate response”) [65, 66]. However, other outcome measures may be more appropriate in some individuals [65].