Symptomatic Treatment of Cough Among Adult Patients With Lung Cancer

Evidence and Recommendations/Suggestions

Clinical research question: In adults with lung cancer experiencing cough beyond initial cancer treatments, what are the most effective pharmacologic and nonpharmacologic interventions?

Summary of the Evidence and Interpretation

For patients with lung cancer experiencing cough, its control often depends on the treatment of the cancer, treatment of associated comorbidities, and antitussive therapy. The cancer and noncancer-related causes of cough may include a direct effect of the tumor mass (eg, infiltration or obstruction), pleural or pericardial effusion, atelectasis, infections, gastroesophageal reflux disease, pulmonary emboli, exacerbation of coexisting COPD or congestive heart failure, esophagorespiratory fistulas, lymphangitic carcinomatosis, superior vena cava syndrome, or treatment-induced cough due to radiotherapy or (more rarely) chemotherapy.^{12, 19} Such causes of cough, for example, may be treated with oncological treatment of cancer, pleural drainage (if pleural effusion exists), antibiotics (when an infection is present), or steroid therapy in cases of COPD or asthma. Physicians should also differentiate between productive and nonproductive cough, where the aim of treatment is different (eg, using mucolytics in productive cough vs suppressing nonproductive cough), as suppression of cough is not always the aim of treating cough among patients with lung cancer. Nevertheless, many patients with lung cancer report typically a dry tickling cough with mechanical and environmental triggers.⁹

Therefore, on the basis of the variety of causes linked with cough among patients with lung cancer, it is essential to start the management of cough with a comprehensive assessment, first targeting the treatable causes of cough (Table 2).²⁰ A comprehensive list of suggestions, not necessarily evidence based, is available through the British Thoracic Society²¹ and the ACCP guidelines on chronic cough due to lung tumors.¹¹

1. In adult patients with cough associated with lung cancer that persists despite cancer treatment, we suggest, as a first step, that a comprehensive assessment according to a published, evidence-based management guideline be undertaken to identify any co-existing causes linked with cough and initiate treatment accordingly (Ungraded, Consensus Based Statement).

Summary of the Evidence and Interpretation

Vertigan and Gibson²² reviewed the work done in chronic refractory cough and the role of speech pathology and cough suppression interventions that consist of education, strategies to control cough, vocal hygiene training, and psychoeducation. Results presented in a review of a single trial of patients with respiratory diseases²³ showed that 88% of patients improved their cough in the speech pathology group vs 14% in the control group. In a systematic review of cough management trials in respiratory diseases other than lung cancer, cough suppression interventions also showed promising results.²⁴ The same is further supported by another systematic review of five trials in refractory chronic cough²⁵ in which a package of cough suppression exercises given over three to four sessions was linked with significant improvements in cough frequency, cough severity, and cough-related quality of life. Cough suppression exercises refer to a number of approaches, including education, identifying cough triggers, cough suppression techniques (ie, pursed lip breathing, swallowing, sipping water), improvements in laryngeal and vocal hygiene and hydration, breathing exercises, and counseling.²⁶

In a more recent feasibility trial in 101 patients with lung cancer experiencing a respiratory distress symptom cluster (breathlessness, cough, fatigue), participants found benefit from an intervention package educating them on how to manage this symptom cluster with specific nonpharmacologic approaches, including cough suppression techniques and diaphragmatic breathing.²⁷ A large trial to test this feasibility trial further is now ongoing in the United Kingdom. Speech pathology training is minimally used in current health care settings primarily because of limited availability of speech therapy services. However, the use of broader cough suppression exercises may be an area to enhance in future provision of care and identify further roles for speech therapists and respiratory physiotherapists in clinical practice, if results from future trials in lung cancer are consistently positive.

2. In adult patients with lung cancer experiencing cough despite anticancer treatment, we suggest cough suppression exercises as alternative or additional to pharmacological therapy where such services are available (Grade 2C).

Summary of Evidence and Interpretation

Endobronchial brachytherapy in a variety of doses seemed to improve cough in selected participants, suggesting that possibly the lowest effective dose should be used to minimize side effects.¹⁵ The studies reviewed^{28, 29, 30, 31, 32, 33, 34} were all uncontrolled comparative trials (mostly prospective ones), although it would have been unethical to carry out randomized trials with such a treatment approach. Photodynamic therapy was examined in one study,³⁵ and, although improvements in cough were observed, its role in relation to other therapies for cough was unclear. When brachytherapy is indicated but not locally available, patients should be transferred to a facility where it is available. Endobronchial brachytherapy should be considered particularly for patients with small or endobronchial tumors or limited disease or in whom the tumor has extended into the large airways; otherwise, external-beam radiotherapy (or a combination of the two) may be more effective.

In using endobronchial brachytherapy, we suggest the lowest dose and fractionated schedule (eg, a single fraction of 10 Gy, two fractions of 7-8 Gy, or three fractions of 5 Gy), as this approach is linked with a good response and a lower number of side effects. It is worth noting that external-beam radiation of one or two fractions can also improve thoracic symptoms, as shown in a systematic review,³⁶ and this approach should be considered if facilities for endobronchial brachytherapy are not available. Nevertheless, as endobronchial brachytherapy can also be associated with significant risk of hemoptysis and other complications, depending on the type of lesion and the area of application, a pharmacologic therapy trial may be more appropriate to start with.

3. In adult patients with cough due to localized endobronchial disease for whom surgery, chemotherapy, or external beam radiation are not indicated, we suggest the use of endobronchial brachytherapy where such specialist facilities are available and in suitable patients (Grade 2C).

Summary of Evidence and Interpretation

It may be appropriate to start pharmacologic treatment at the same time with the previous two recommendations, especially if the cough is severe. This takes into consideration the fact that such specialist services may not be widely available, and, if they are, there are practical considerations in their delivery and their effect may not be immediate. The majority of trials reviewed in the Cochrane systematic review¹⁵ referred to a variety of pharmacologic approaches. All of them had a high risk of bias. There was one double-blind randomized trial suggesting significant improvements in cough management from the use of butamirate citrate linctus (Sinecod syrup)³⁷ only in the subgroup of patients with lung cancer (n = 14). In the general respiratory disease field, there were also two additional trials on glycerol-based cough syrups,^{38, 39} also showing cough decreases, and this may also be appropriate treatment for patients with lung cancer. There is a variety of these cough syrups on the market, sold over the counter, including Sinecod; Benylin Tickly Coughs; Benylin Dry Coughs; Actifed Multi-Action Dry Coughs; Meltus Dry Coughs; Robitussin for dry coughs; Day & Night Nurse (GSK); and others (some of them are not available in the United States and some other countries). Many of them include dextromethorphan in variable concentrations, and the Day & Night Nurse also includes pholcodine. Because of their low cost, some evidence of effect, and low side effect profile, this approach could be tried initially, although for patients with profound cough, demulcents may be less effective, particularly as some over-the-counter preparations contain active drugs at subthreshold therapeutic levels.

When patients do not respond to this approach, opioids should be considered next. Opioids are the drugs that have most evidence in the management of cough among patients with lung cancer, albeit of low methodologic quality, and they are perhaps not as effective as physicians would ideally like. In the Cochrane systematic review,¹⁵ there was a double-blind randomized trial with positive results from using codeine 30 mg twice a day with phenyltoloxamine 10 mg⁴⁰; one with dihydrocodeine,⁴¹ and one comparing morphine with codeine.⁴² However, there have been a number of case reports providing some evidence of the beneficial effect of morphine, methadone, pholcodine, and hydromorphone, summarized in a scholarly review,⁴³ and a phase II trial using hydrocodone.^{44, 45} For patients with lung cancer experiencing cough for whom treatment with an opioid derivative is indicated, we suggest pholcodine or hydrocodone (where available) or dihydrocodeine or morphine. Codeine is less preferred (despite being the most researched drug in this field) because of its greater side effect profile compared with those of other opioids, as commented on by many experts (personal communication with palliative medicine experts). Morphine should be used if the cough is not suppressed by other opioid derivatives or other means, including other centrally acting antitussives such as dextromethorphan. The patient’s previous exposure to opioids will dictate the initial starting dose.

Many patients with advanced lung cancer may already be receiving opiates for other symptoms (eg, pain or breathlessness). If patients are already receiving morphine, sometimes increasing the dose by 20% may be helpful, although this is based on experience rather than any evidence. For patients with lung cancer who are experiencing nonspecific cough and who are in the palliative stage of their illness, we suggest a bedtime dose of codeine/pholcodine or morphine, as this approach may help suppress cough and induce an undisturbed sleep, although again there is no evidence for this (but physicians invariably use this approach).

Among peripherally acting antitussives, levodropropizine is probably equally effective to dihydrocodeine or moguisteine and with possible earlier cough reductions than with dextromethorphan. This was suggested by a systematic review of four trials (two randomized trials testing levodropropizine against dihydrocodeine and moguisteine and two nonrandomized placebo-controlled studies, all with important limitations and high risk of bias).⁴⁶ The major methodologic limitations of the studies reviewed make these results less convincing, and further research is necessary before any concrete conclusions are derived. Because some of these drugs are not available in many countries, the choice of treatment may be dictated primarily by availability rather than pharmacologic parameters.

Local anesthetics, such as nebulized lidocaine, have been suggested to be helpful in case studies^{47, 48} and are commonly used in palliative care for intractable cough that has not responded to any other approaches. This is also supported by findings from a systematic review of cough management approaches.²⁴ Hence, we suggest that such local anesthetics be tried when other pharmacologic approaches have failed to manage cough among patients with lung cancer. As local anesthetics can increase the risk of aspiration, which can be prevalent in frail patients with cancer, aspiration risk should be assessed prior to the use of this type of treatment for cough.

Doses used in the medications mentioned earlier vary from country to country. Table 3, shown originally in a previous cough guideline,¹³ provides indicative doses.

4. In adult patients with lung cancer who require a pharmacological approach for the treatment of cough, we suggest an initial trial with demulcents such as butamirate linctus (syrup) or simple linctus (syrup) or glycerol-based linctus (syrup) where available (Grade 2C).

5. In adult patients with lung cancer experiencing cough that does not respond to demulcents, we suggest pharmacological management using an opiate-derivative titrated to an acceptable side-effect profile (Grade 2C).

6. In adult patients with lung cancer experiencing opioid-resistant cough, we suggest a peripherally-acting antitussive (where available), such as levodropropizine, moguisteine, levocloperastine or sodium cromoglycate (Grade 2C).

7. In adult patients with lung cancer experiencing opioid-resistant cough that does not respond to peripheral antitussives, we suggest a trial with local anesthetics, including nebulized lidocaine/bupivacaine or benzonatate (Ungraded, Consensus Based Statement).

As the evidence described earlier is of generally low quality and the level of confidence is fairly low, it is likely that, despite best efforts as provided according to the previous recommendations, some patients will not respond to the suggested treatment. Physicians need to be aware that these cough management strategies, although based on the available evidence, are not necessarily optimal or effective enough, and discretion in their use should be exercised. The duration of treatment is an issue to consider, too: Although the evidence for this is minimal, if a short course of treatment does not lead to improvements, the treatment should be discontinued and another approach should be tried. Hence, ongoing research on the unmet need for better antitussive approaches in the lung cancer population is urgently needed. In managing cough among patients with lung cancer, who also often have advanced cancer, controlling cough and providing cough-free periods is highly important for patients from a quality-of-life perspective. In these cases, other experimental approaches may be used and trialed. We have examples in the literature of case studies in which physicians have used gamma aminobutyric acid agonists (such as baclofen),²¹ diazepam,¹⁹ paroxetine (in concomitant pruritus and cough),⁴⁹ amitriptyline,⁵⁰ gabapentin,^{50, 51} carbamazepine,⁵⁰ and thalidomide,⁵² although many of these refer to chronic cough or cough in respiratory diseases other than cancer and could be tried in the lung cancer setting in an N-of-1 trial.

On occasion, complications of coughing⁵³ may be debilitating and not responsive to any medications. In such circumstances, physicians may consider trying other medications even when treatment decisions cannot be based on existing evidence. In such cases, single case experiments, also referred to as “N-of-1 randomized controlled trials (RCTs),”⁵⁴ should be considered because N-of-1 RCTs are the most rigorous design for establishing efficacy of treatment in individual patients.⁵⁴ The key elements in performing such trials are (1) obtaining informed consent, (2) randomizing two treatments (eg, active therapy or placebo or alternative therapy) determined by random allocation preferably with crossover, (3) double blinding, and (4) measuring a cough outcome important to the patient.⁵⁴ Because ethical issues of N-of-1 RCTs are different than those of standard RCTs involving other patients, institutional review board approval and informed written consent may not be necessary. However, consultation with the institutional review board and reading more about how to set up an N-of-1 RCT program are advised prior to starting such a program.⁵⁴ Also, the costs of preparing the placebo should be considered because it may be a barrier to developing the program. N-of-1 RCTs with placebo control groups may not be practical in a private or clinical setting, in which case an unblended N-of-1 trial, although less scientifically rigorous and robust, may be an alternative approach.

8. In adult patients with intractable cough due to lung cancer in whom surgery, chemotherapy, external beam radiation, brachytherapy and the previously mentioned nonpharmacological and pharmacological approaches are ineffective or not indicated, we suggest that clinicians consider performing N-of-1 randomized controlled trials to determine if any of the following drugs might be of benefit in controlling cough because none have been definitively shown to be effective nor devoid of side effects: diazepam, gabapentin, carbamazepine, baclofen, amitriptyline, thalidomide (Ungraded, Consensus Based Statement).