Difficulty with maintaining nutritional intake is common in patients with cancer. European guidance suggests that wherever possible nutritional support should first be provided by the oral or enteral route. Where this is not possible, for example, in malignant small bowel obstruction, parenteral nutrition (PN) may be considered. In palliative care, it is generally accepted that the significant risks and burdens of PN outweigh the potential benefits in patients with an expected survival <2 months. Determining prognosis is crucial when helping patients to make decisions regarding appropriate care pathways; however, this remains challenging. An overview of clinical issues and prognostic indicators related to selecting patients appropriately for palliative PN is given to cover the relevant advanced competencies of the 2010 Gastroenterology Curriculum. The organisation of Home Parenteral Nutrition (HPN) services in England is described including the associated risks and burdens of HPN in the palliative patient.
- clinical decision making
- gastric cancer
- parenteral nutrition
- quality of life
Statistics from Altmetric.com
Difficulty with maintaining nutritional intake is common in patients with cancer. The European Society for Clinical Nutrition and Metabolism (ESPEN) suggest that wherever possible nutritional support should first be provided by the oral or enteral route.1 Where this is not possible, for example, in malignant small bowel obstruction (MSBO), parenteral nutrition (PN) may be considered. ESPEN guidelines1 recommend offering nutritional intervention after considering with the patient the expected prognosis of the disease, if their quality of life is acceptable and will be stabilised or improved and if the patient wants the intervention. Generally, patients should be free of any form of organ dysfunction that may complicate PN.2 The purpose of PN should be clear; whether to support the patient through palliative treatment or to palliate until death. The British Intestinal Failure Alliance (BIFA)3 recognises three palliative groups of patients who may be referred for Home Parenteral Nutrition (HPN). Group 1 are those with expected short survival of <12 weeks but who are young and with good performance indicators. Group 2 are those who require PN to support them during palliative treatments such as chemotherapy or palliative surgery. Group 3 are those with slow growing tumours who are expected to survive for a number of years. However, it is generally accepted that the significant risks and burdens of PN outweigh the potential benefits in patients who have an expected survival of <2 months.
Although it may be difficult, determining prognosis is crucial when helping patients to make decisions regarding the most appropriate care pathway for them. Where patients report feeling reassured having nutritional needs met via PN, it has also been shown that home PN can significantly restrict family life and social contact.2 This article gives an overview of clinical issues related to selecting patients appropriately for palliative PN to cover the relevant advanced competencies of the 2010 gastroenterology curriculum (table 1).4
Home Parenteral Nutrition services
The provision of HPN in England is subject to the NHS Commercial Medicines Unit National Framework Agreement for Home Parenteral Nutrition.5 This framework sets out the organisational requirements for centres providing HPN services; and aims to ensure equity of access to specialised care, minimum clinical standards and designation of centres to ensure appropriate resourcing. Only approved HPN homecare companies identified within the framework may be commissioned by HPN centres to provide homecare to patients.
HPN homecare companies provide services which generally encompass:
Compounding/provision of bespoke PN and other intravenous fluids;
Supply of all equipment (infusion pumps, administration sets, dressings, pharmacy refrigerator for storage of PN, etc);
Delivery service for all equipment;
Nursing service (available 7 days/week either for training of patients/carers or provision of full nursing service for administration of PN).
Lead time to discharge
Patients accepted for HPN will generally require an inpatient stay at the HPN centre to establish suitable central venous access and an individual PN regimen that meets their requirements. In some cases this may take 2–3 weeks. Once the nutrition support team at the HPN centre has identified that the patient is stable and suitable for HPN and completed all necessary homecare company registration documents, the company aim to compound PN and install all equipment within 5 working days.5 However, if homecare company nursing services are required, the availability of nurses within the patient’s home area will impact on discharge plans. In areas outside of England company nursing services may not be available and so training for District Nurses may need to take place.3 Discharge may also be delayed if the patient’s clinical condition changes or if they require organisation of other support services at home.
Impact of HPN
The impact and burdens on the patient and their family must be taken into account when considering HPN as an option. Important factors must be discussed with the patient in order for them to give informed consent to HPN. Table 2 outlines some of the key impacting factors.
Risks of parenteral nutrition
A study of 82 patients receiving palliative chemotherapy and PN for MSBO showed that PN-related complications (catheter-related sepsis or hyperbilirubinaemia) occurred in 32.9% of patients.6 A recent poster presentation of a systematic review, showed complications in 13%±6% of patients.7 Palliative patients may wish to have an agreed plan in place regarding how they would like complications managed in the terminal phase of their disease, for example, re-admission to hospital versus palliation at home or hospice.
Central venous access devices
Central venous access is needed in patients who are considered for HPN. Prolonged use and HPN usually require a long-term device such as a tunnelled catheter or a totally implantable device (often called a port).8 Peripherally Inserted Central Catheters are acceptable for short (<6 weeks) to medium (6 weeks to 6 months) term PN.8 A single, dedicated lumen for PN is required.8 Factors to consider when deciding on the type of central venous access devices (CVAD) include frequency of infusions, duration of therapy, prognosis and previous history related to obtaining central venous access.9 The following need to be taken into account when selecting the most appropriate type and site for CVAD insertion9:
Patient-specific factors, for example, pre-existing CVAD, anatomical abnormalities, coagulopathy;
Insertion-related risks, for example, bleeding, pneumothorax, thrombosis;
Ongoing risk of infection including catheter-related bloodstream infections.
Catheter-related bloodstream infections
Catheter-related bloodstream infections (CRBSI) are a major complication of HPN, with NHS England targets of <1 per 1000 catheter days. Diagnosis of CRBSI is achieved by (a) culture of the catheter tip (when CVAD is removed) or b) paired blood cultures from a peripheral vein and from the CVAD with continuous monitoring of the differential time to positivity (if CVAD has not been removed).9 Successful management of CRBSI with line salvage may require 7–14 days in hospital, impacting on successful home palliation, with septic complications having significant morbidity and potential mortality.
Management of CRBSI
ESPEN guidelines9 suggest that in patients with catheter-related bacteraemia the CVAD can be salvaged by administering antibiotic lock therapy with systemic antibiotics. Treatment should continue for 2 weeks. However, removal of a CVAD is advised in complicated catheter-related bacteraemia such as8:
Tunnel infection or port abscess;
Clinical signs of septic shock;
Positive blood cultures for fungi or highly virulent bacteraemia;
Complicated infection, for example, endocarditis, septic thrombosis or other metastatic infection.
A systematic review showed that intestinal failure-related liver disease (IFALD), previously often referred to as PN-associated liver disease, is common in approximately 15%–40% of adults receiving PN for a prolonged period (>2 weeks).10 IFALD is often defined by hyperbilirubinaemia.10 The review by Zu and Li10 suggest possible PN-related causative factors including: excess calories given, phytosterols found in the lipid component of PN, lack of enteral feeding leading to cholestasis and sepsis, which may cause or exacerbate underlying cholestasis.
A study by Diamond et al 11 in infants receiving PN showed that septic episodes led to a 3.2-fold increase in the risk of developing IFALD. It must be noted that altered hepatic function is multifactorial and in some cases will not be attributable to PN, thus the change in nomenclature to IFALD, for example, sepsis from an unrelated source, the patient’s underlying disease, and relative shortness of remaining small bowel. Current hepatic function must be considered when determining if palliative PN is of benefit to the patient, given the risk of exacerbating any pre-existing hepatic dysfunction.
The impact of PN on hepatic function can be minimised to some extent by manipulation of the PN prescription, for example, reduction in the amount of lipid given. However, it must be noted that sacrificing lipid calories leads to a requirement to increase the amount of glucose given to meet energy requirements. This can lead to glucose intolerance and fatty liver.10 In some cases, it may be necessary to significantly underfeed patients to reduce the risk of IFALD. In these circumstances, it is necessary to consider if PN is still of benefit in the palliative patient, although many may not survive long enough to develop significant IFALD complications. The use of fish oil containing lipids in PN has shown to reduce the impact of lipids on hepatic function.10 Maintaining even small amounts of enteral intake is a further strategy to reduce the risk of cholestasis as is the prevention of CRBSI.10
Quality of life
Stabilising or improving quality of life (QOL) and/or performance status for patients receiving PN in palliative care is one of the guiding factors of ESPEN guidelines.1 QOL can be determined by using performance status scores and QOL questionnaires. A systematic review6 involving 18 studies suggested an improvement in QOL, however, over half the papers reported patients given PN for cachexia rather than MSBO. The review found the mean duration of HPN was 3.8±1.4 months7. Studies by Vashi et al 12 and Cotogni et al 13 have also suggested an improved QOL for palliative patients receiving HPN.
A cohort of 52 patients with advanced cancer receiving oncological treatments and HPN were monitored monthly using a QOL questionnaire and Karnofsky Performance Scale (KPS).12 Significant improvements in QOL were measured after 1 month but the greatest benefit was for those patients who received 3 months of HPN.12 Patients who initially had a poor KPS still benefited from HPN.12 In another study of HPN in palliative patients, over half of the 111 patients were receiving further oncological treatments.13 The study reports patients felt relief knowing their nutritional needs were being met with HPN as it caused distress to them and their families when they were not able to take sufficient orally.13 It is notable that in both studies all patients or their carers had been trained to administer the PN rather than relying on external nursing input. Both studies concluded HPN was beneficial and improved QOL in patients with advanced cancer. However, a recent Cochrane review of patients with MSBO suggested that available evidence was low quality and thus, they could not determine if palliative HPN improved QOL or longevity in this group.14
The modified Glasgow Prognostic Score (mGPS) is a validated prognostic score for use with patients diagnosed with several types of cancer and has been shown to be accurate in patients with, for example, gastric cancer.15 The mGPS is an inflammation-based score and considers concurrently raised C reactive protein and hypoalbuminaemia as risk factors for poor prognosis. The scores are intended to give an indication of likelihood of survival at 5 years. In gastric cancer, this is score 0=74.6%, score 1=61.4% and score 2=34.6% overall survival.15 It is unclear how useful this score is in determining prognosis for patients who are recognised to be palliative and so possibly are not expected to survive for 5 years.
BIFA suggest that performance indicators such as KPS (preferably score >70) or the Eastern Cooperative Oncology Group (ECOG) scale (<2) may be of help in some cases.3 The KPS score 70 equates to a patient who is able to care for themselves but who is unable to carry on normal activity or to do active work. The ECOG score description equates to a patient who is ambulant and capable of all self-care but unable to carry out any work activities around 50% of the day.3 These performance scores would seem to offer a useful assessment of the patient in determining likely survival and QOL on HPN.3
Decision making for patient selection for palliative HPN remains challenging. Essentially, teams need to employ a variety of assessment tools to arrive at a likely prognosis for individual patients. However, BIFA recognise that the sensitivity and specificity of prognostic and performance scores may be poor in individual patients.3 Further research into realistic prognostic indicators is warranted particularly in patients with a younger age at diagnosis.
Given the risks and burdens associated with HPN, the focus should be on maintaining or improving QOL for patients. While some patients will perceive a benefit to their QOL, in other patients HPN simply cannot achieve this. Thus, the limitations of HPN must be recognised. The impact on home life and the lead time required to initiate HPN, particularly in patients with an expected short survival time, can be prohibitive. The patient must be involved in a full and frank discussion regarding all issues related to HPN. This is essential both in aiding patients' decision making and in establishing realistic expectations of likely outcomes.
At the initiation of HPN, it is important that a plan is made for the patient’s end-of-life care with regard to when and if they would want to stop having HPN. Management of PN-related complications must also be discussed. The occurrence of CRBSI will require urgent admission to hospital for a period of treatment, which will inevitably reduce the amount of time patients are at home. PN-related complications are likely to be in addition to any disease-related complications that require hospital treatment. Early discussion of these issues will prevent difficult decisions having to be made when the patient’s condition is declining and ensure the patient’s wishes are respected.
Gastroenterology teams are likely to encounter these challenges in their practice. Having an understanding of what HPN involves, and the potential burdens, will enable teams to weigh risk and benefits carefully before discussing palliative care options in patients with MSBO.
Multiple Choice Questions
1. A patient aged 23 years is referred to the nutrition support team for consideration of HPN. They have been diagnosed with gastric adenocarcinoma. What additional information do you require at this point?
Full history of the diagnosis and treatments to date
If there is a plan for any palliative treatment
If enteral nutrition has been attempted
Patient’s current nutritional status
What the patient understands about his or her disease and likely prognosis
Answer: all of the above
More information will be required before determining if HPN is suitable for this patient. However, from the outset it is important to understand the disease process and treatments to help determine a likely prognosis. As per ESPEN guidelines, enteral nutrition should always be used in preference to PN so if this has not been trialled it is important to do so. It is easier, quicker, safer and less intrusive to discharge a patient home on enteral feed than HPN. Current nutritional status is required to determine any feeding plan. Given the burden of HPN, it is essential that patients have realistic expectations regarding the course of their disease and prognosis. It is important that patients understand HPN will not alter the course of their disease.
2. What services would you expect to be provided to a patient by a HPN homecare company?
Dressing of a central venous access device site
Help with personal hygiene
Administration of enteral feed
HPN companies will only deal with aspects directly related to the patient’s parenteral nutrition or the central venous access device that it is administered through. Any other care needs must be provided by other services.
3. A patient is receiving HPN while they undergo palliative chemotherapy for a gastric carcinoma. The patient calls the nutrition support team from home complaining of ‘shivers’ and feeling unwell. They report that several members of their family have bad colds at the moment. Which of the following is the incorrect/unsafe plan?
Ask the patient to check their temperature and report back
Watch and wait—they have probably caught a cold from a family member
Admit the patient as an emergency as she is at high risk of developing central line-related sepsis
Take peripheral and central blood cultures
Commence intravenous antibiotics immediately
Answer: A, C, D, E
There is a possibility that the patient just has a cold. However, patients receiving PN are at high risk of developing central line-related infections so sepsis must always be suspected.
Fever, rigours particularly when PN is infusing and patient reports of feeling generally unwell are indicators of possible sepsis in PN patients. Treatment for sepsis must be started immediately until blood culture results are available but after the blood cultures have been obtained. A decision regarding CVAD salvage versus removal should be made based on whether the patient has septic shock, response to antimicrobial therapy and the organisms grown at blood culture. Unnecessary line removal increases morbidity, number of invasive procedures a patient has to undergo (and complications thereof) and may result in future loss of central access.
4. What would an mGPS of 0 suggest to you in a patient who has been referred for palliative HPN?
The patient has about a 75% chance of surviving for 5 years
HPN would definitely benefit this patient
The patient has responded well to their treatment
The patient has responded poorly to their treatment
The mGPS indicates in gastric cancer 5-year overall survival is likely to be 0=74.6%, 1=61.4% and 2=34.6%. The score on its own does not give an accurate indication if the patient will benefit from HPN. However, it may aid decision making in determining a likely prognosis when used in conjunction with other performance indicators and a thorough assessment of the patient’s history and current clinical status.
5. A patient with non-small cell lung cancer has not responded to second line chemotherapy. He is losing weight rapidly. The lung cancer MDT has referred for HPN as the patient is only in their mid-30s. Which statements are true?
HPN is the best choice of treatment
Enteral feed must be tried first
Prognosis from the cancer is not important because the patient is young
The patient does not need to understand about HPN as it will be connected and disconnected by specialist nursing staff
The performance status does not need to be considered as the patient is young
HPN offers no advantages to patients with cancer cachexia, but may benefit those with intestinal failure. As per ESPEN guidelines, enteral nutrition should always be used in preference to PN so if this has not been trialled it is important to do so. Enteral support is likely to be as successful as PN in cancer cachexia but has the advantages of being simpler to achieve, more physiological, result in less serious complications, more rapid to facilitate discharge home and cheaper. A younger age does not by itself confer an improved prognosis and longevity must still be considered, in combination with performance status.
The patient should be fully informed and understand the benefits and drawbacks of HPN.
Contributors JF made substantial contributions to the article in the conception and design of the content, drafting the article and revising it critically for important intellectual content. DW, MD and SCC made substantial contributions in the conception and design of the article content. All authors have reviewed and approved the final version of the article to be submitted.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests JF: participated in a BD Medical Expert Advisory Board and received an honorarium. DW: received educational sponsorship from Calea/Fresenius-Kabi. MD: none declared. SCC: received honoraria and participated in advisory boards for Eli Lilly, honoraria from Baxter and educational sponsorship from Shire and Calea/Fresenius-Kabi.
Provenance and peer review Not commissioned; externally peer reviewed.
Patient consent for publication Not required.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.