Abstract
Objectives. To examine the outcome in patients with SSc requiring parenteral nutrition (PN), and to compare their clinical characteristics with those of other SSc patients and of patients requiring PN/home parenteral nutrition (HPN) for other conditions.
Methods. Retrospective review of SSc and Intestinal Failure Unit databases at a tertiary referral centre for SSc/national unit for intestinal failure over a 13-yr period.
Results. Eight patients with SSc requiring PN during the study period were identified (2 males, 6 females: median age at commencement of PN 51 yrs, range 42–56 yrs). All patients commencing PN had bacterial overgrowth and malabsorption not responding to antibiotic therapy. The median duration of PN therapy in the eight patients was 40 months (range 0.8–192 months). Between them the eight patients had a total of 13 851 catheter-use days and only two line infections (0.14/1000 catheter days), a lower rate of line infection than in other HPN-treated patients at Hope Hospital (0.52/1000 catheter days). Three patients died during the 13-yr period, none of causes related to their PN. Six were unable to manage their HPN regime themselves, mainly because of problems with hand function.
Conclusions. Although patient numbers were small, our findings suggest that HPN can be safely and successfully used long-term in patients with SSc and should be considered for patients unable to maintain their nutritional status because of severe gastrointestinal involvement. Impaired hand function should not preclude SSc patients from receiving HPN: family members or community nurses may be trained in the care of the HPN line.
Background
Approximately 90% of patients with the multisystem connective tissue disease SSc develop gastrointestinal involvement of their disease, although many are asymptomatic [1, 2]. Gastrointestinal problems occur in both the lcSSc and dcSSc disease subtypes [3]. Any part of the gastrointestinal tract may become involved and an autopsy study showed that approximately half of the patients have histological involvement of their small bowel [4]. A small minority of patients develop intestinal failure. The usual scenario in this minority is that small intestinal involvement, with impaired peristalsis, results in bowel dilatation and stasis that in turn lead to bacterial overgrowth, malabsorption, and sometimes pseudo-obstruction. Gastrointestinal involvement in SSc can therefore be life-threatening: of 264 patients with dcSSc from 29 centres, with an average duration of follow-up of 5.2 yrs, 13 (5%) died of gastrointestinal disease, representing 15% of deaths definitely related to SSc [5].
When a patient is unable to maintain his/her nutritional status, as in SSc-related intestinal failure, parenteral nutrition (PN) may be required to prevent, or correct, malnutrition. Other forms of artificial feeding, including oral supplementation, naso-jejunal feeding and feeding via percutaneous enterogastrostomy (PEG)/percutaneous jejunostomy tubes may be tried before introducing PN. Complications of PN administration include central venous catheter infections, occlusions and fractures of the central line, hepatic disease and osteoporosis. If the indication for PN suggests that long-term treatment will be required, as in SSc, home parenteral nutrition (HPN) may be considered. PN is a potentially life-saving treatment but training the patient as carer for the central line for HPN may pose particular problems in SSc because of concomitant problems including impaired hand function and cardiorespiratory disease.
The aim of this retrospective study was to examine the outcome in patients with SSc requiring PN, and to compare their clinical characteristics with those of other SSc patients and of patients requiring PN/HPN for other conditions.
Patients and methods
The study relates to the cohort of patients with SSc attending Hope Hospital, Salford, that began receiving, or were currently receiving, PN to maintain their nutritional status between 1993 and 2006. Hope Hospital is a tertiary referral centre for SSc and one of the only two existing intestinal failure units in the UK. Once identified, the patients receiving PN were compared with other SSc patients entered on to the SSc database at Hope Hospital during the same period of time, looking at factors such as gender, age at diagnosis and autoantibody status. In addition, the PN regimes, management issues and complication rates, including central line infections, were compared with those in patients requiring HPN for other conditions.
Patients with SSc requiring PN during the study period were identified from one or both of the two Hope Hospital databases: the SSc database (containing details of those attending with SSc) and the Intestinal Failure Unit database (containing details of those attending the Intestinal Failure Unit). For each patient identified:
His/her medical records were retrospectively reviewed for details of gastrointestinal problems leading up to their treatment with PN.
His/her clinical details (including gender, SSc subtype, mean age at physician's diagnosis, autoantibody status) were compared with those of the other patients with details entered on to the SSc database at Hope Hospital during the study period. Gender, subtype and antibody status were compared between the two SSc groups (PN vs non-PN) using a 2 × 2 chi-square test.
His/her PN regime was reviewed. The PN regimes of the patients with SSc were compared with those of other patients receiving PN at the Intestinal Failure Unit.
Results
Eight patients with SSc (2 male, 6 female) requiring PN during the study period were identified.
Gastrointestinal problems and their treatment prior to commencing PN
The median age at commencement of PN in the eight patients was 51 yrs, range 42–56 yrs. Table 1 gives details of gender, age at onset of Raynaud's phenomenon/first non-Raynaud's manifestation of SSc/recalled onset of skin thickening/diagnosis of SSc, SSc subtype, autoantibody status, length of time between onset of Raynaud's phenomenon/skin thickening and initiation of PN, age at commencement of PN, duration of PN treatment and the PN situation at the end of the study period (June 2006). The median time between the onset of Raynaud's phenomenon and commencement of PN was 54 months (range 4–251 months). Patient 8 had very rapidly progressive dcSSc and was on PN for less than 1 month prior to her death in hospital.
Clinical characteristic of the eight patients with SSc-related intestinal failure
| Patient | Gender | Age at recalled onset of Raynaud's phenomenon (yrs) | Age at onset of 1st non-Raynaud's manifestation of SSc (yrs) | Age at recalled onset of skin thickening (yrs) | Age at SSc diagnosis (yrs) | SSc subtype | Auto-antibody status | Time between onset of Raynaud's phenomenon and PN (months) | Time between onset of skin thickening and commencement of PN (months) | Age at commencement of PN (yrs) | Duration of PN (months) | PN outcome |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | F | 44 | 48 | 48 | 48 | Limited | Negative | 76 | 28 | 50 | 192 | Current |
| 2 | M | 41 | 42 | 42 | 42 | Diffuse | ANA 1/10000 | 44 | 32 | 44 | 48 | Current |
| 3 | M | 32 | 37 | 37 | 37 | Limited → Diffuse | ANA 1/100 | 251 | 189 | 52 | 18 | Current |
| 4 | F | 38 | 38 | 38 | 38 | Diffuse | ANA 1/10000 Anti-topoisomerase I positive | 59 | 57 | 42 | 9 | Died |
| 5 | F | 53 | 54 | SSc sine scleroderma | 55 | Limited | Anti-centromere positive | 39 | n/a | 56 | 103 | Current |
| 6 | F | 40 | 39 | 40 | 40 | Limited | ANA 1/1000 Anti-RNP positive | 49 | 41 | 44 | 32 | Stopped October 2004 |
| 7 | F | 37 | 37 | 37 | 39 | Limited | Anti-centromere positive | 232 | 230 | 56 | 59 | Died |
| 8 | F | 55 | 54 | 55 | 56 | Diffuse | ANA 1/20 | 4 | 4 | 56 | 0.75 | Died |
| Patient | Gender | Age at recalled onset of Raynaud's phenomenon (yrs) | Age at onset of 1st non-Raynaud's manifestation of SSc (yrs) | Age at recalled onset of skin thickening (yrs) | Age at SSc diagnosis (yrs) | SSc subtype | Auto-antibody status | Time between onset of Raynaud's phenomenon and PN (months) | Time between onset of skin thickening and commencement of PN (months) | Age at commencement of PN (yrs) | Duration of PN (months) | PN outcome |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | F | 44 | 48 | 48 | 48 | Limited | Negative | 76 | 28 | 50 | 192 | Current |
| 2 | M | 41 | 42 | 42 | 42 | Diffuse | ANA 1/10000 | 44 | 32 | 44 | 48 | Current |
| 3 | M | 32 | 37 | 37 | 37 | Limited → Diffuse | ANA 1/100 | 251 | 189 | 52 | 18 | Current |
| 4 | F | 38 | 38 | 38 | 38 | Diffuse | ANA 1/10000 Anti-topoisomerase I positive | 59 | 57 | 42 | 9 | Died |
| 5 | F | 53 | 54 | SSc sine scleroderma | 55 | Limited | Anti-centromere positive | 39 | n/a | 56 | 103 | Current |
| 6 | F | 40 | 39 | 40 | 40 | Limited | ANA 1/1000 Anti-RNP positive | 49 | 41 | 44 | 32 | Stopped October 2004 |
| 7 | F | 37 | 37 | 37 | 39 | Limited | Anti-centromere positive | 232 | 230 | 56 | 59 | Died |
| 8 | F | 55 | 54 | 55 | 56 | Diffuse | ANA 1/20 | 4 | 4 | 56 | 0.75 | Died |
ANA, antinuclear antibody; RNP, ribonucleoprotein.
Of the eight patients, five complained of abdominal pain, four of abdominal distension and four experienced constipation. Six patients experienced symptoms suggestive of intestinal pseudo-obstruction including distension, bloating and a change in bowel habit with diarrhoea and intermittent constipation. All eight patients had positive glucose hydrogen breath tests suggesting the presence of bacterial overgrowth [6] and positive faecal fat tests, suggesting co-existing malabsorption. All eight patients experienced weight loss and a change in bowel habit. Four patients described diarrhoea as the main symptom and four constipation. All eight patients were initially managed with long-term cyclical antibiotics prior to PN treatment. Three of the eight patients were initially managed with PEG tubes but failed to gain weight. The median body mass index of the eight patients was 18.1, range 15.1–20.6, reflecting their poor nutritional state.
Of the eight patients, all but one had sclerodactyly, five had digital pitting, two had had digital amputations, three had pulmonary fibrosis and two had clinically significant cardiac problems including previous pericardial effusions. None had clinically significant renal involvement. None of the patients were on immunosuppressive therapy.
Comparison with SSc patients not requiring PN
Table 2 compares gender, SSc subtype, age at time of physician's diagnosis and autoantibody status [anti-centromere, anti-topoisomerase I (Scl-70)] between the eight patients on PN and the other 286 patients on the SSc database. There were no statistically significant differences between groups in gender, SSc subtype or autoantibody status.
Comparision between eight patients with SSc on PN and 286 patients with SSc not on PN
| PN group (n = 8) | Non-PN group (n = 286) | |
|---|---|---|
| Male/female (%) | 2/6 (25/75) | 58/228 (20/80) |
| LcSSc/dcSSc (%) | 5/3 (63/37) | 214/72 (75/25) |
| Mean age at diagnosis of SSc (yrs) | 44 | 48 |
| Anti-centromere positive (%) | 2 (25) | 97 (34) |
| Anti-topoisomerase positive (%) | 1 (13) | 25 (9) |
| PN group (n = 8) | Non-PN group (n = 286) | |
|---|---|---|
| Male/female (%) | 2/6 (25/75) | 58/228 (20/80) |
| LcSSc/dcSSc (%) | 5/3 (63/37) | 214/72 (75/25) |
| Mean age at diagnosis of SSc (yrs) | 44 | 48 |
| Anti-centromere positive (%) | 2 (25) | 97 (34) |
| Anti-topoisomerase positive (%) | 1 (13) | 25 (9) |
PN
Seven of the eight patients received HPN.
Duration of PN and treatment outcome
The median duration of PN treatment in the eight patients was 40 months (range 0.8–192 months). Three patients died although none as a direct result of their PN. Patients 4 and 8 died as a result of cardiorespiratory failure secondary to SSc. Patient 7 died from respiratory failure following aspiration. Patient 4 had her PN stopped before her death because of problems with fluid overload, but her nutritional status had improved. Patient 6 was able to stop HPN treatment as she was able to maintain her nutritional status, her weight was steady and the bacterial overgrowth began to respond to cyclical antibiotics. The remaining four patients were still on HPN at the end of the study period.
HPN complications and comparison with those in patients receiving HPN for indications other than SSc-related intestinal failure
The eight patients had a total of 13 851 catheter-use days from the commencement of therapy through to the end of the research period. There were two line infections during this time. This gave a HPN line infection rate of 0.14 infections per 1000 catheter-use days. The HPN database showed that between 1 April 2005 and 31 March 2006 there were 165 patients on HPN giving a total of 51 484 catheter-use days. There were a total of 27 line infections during this time (and none in the SSc patients). This gives a rate of 0.52 line infections per 1000 catheter-use days.
In the eight SSc patients, there were a total of eight central line blockages and four thrombus formations. These four patients were all anti-coagulated with tinzaparin. Two patients developed calcification of their HPN lines. In one patient the calcification of the line made it impossible and hazardous to remove the line and so it was buried under the skin and a new line inserted on the opposite side. As stated above, one patient experienced severe fluid overload while on HPN and this, combined with her improved nutrition, meant that HPN was stopped.
HPN management issues
All seven patients on HPN infused their HPN overnight. The bags were infused over a longer period of time than other HPN patients (in whom the standard time is 12 h including the weaning on and off periods), so as to reduce the risk of fluid overload. The volume of HPN was on average 1500 ml, which is less than the average in most HPN patients (1500–5000 ml/day). All patients were given approximately 4 weeks of training on HPN line management before discharge. This training was the same, and of similar duration, to that of other HPN patients. Five of the patients were unable to be trained. In four cases, family members were trained in the management techniques. In one case, district nurses were responsible for the entire HPN management. The reasons the patients could not manage their own HPN were the presence of digital ulcerations, fixed flexion deformities due to sclerodactyly and digital amputations. One of the patients who could manage had good manual dexterity and no ulcerations or amputations. The other patient had SSc sine (without) scleroderma and was free of any skin complications.
Discussion
The key finding from this study is that HPN can be successfully used on a long-term basis for patients with SSc and should therefore be considered in patients unable to maintain their nutritional status because of severe gastrointestinal involvement of their disease. In these patients, PN is life-saving. Peripheral feeding may be possible initially but if the symptoms deteriorate or the nutritional status of the patient fails to improve, long-term HPN may be necessary. While resection of affected bowel might sometimes be considered in patients with SSc-related intestinal failure [7], this is seldom performed as the disease process tends to affect the whole bowel, and there is also the concern of adhesions complicating diagnosis of subsequent episodes of pseudo-obstruction. Therefore, treatment of intestinal failure usually relies on nutritional support.
There are very few previous reports in the literature describing experience with HPN in SSc-related intestinal failure. Our findings confirm those of Ng et al. [8] who also reported successful HPN therapy in SSc. In Ng et al.'s cohort of 15 patients, studied over a total of 15 700 catheter days (compared with 13 851 catheter days in our eight patients), two experienced septicaemia, two had superior vena caval obstruction and seven died (none from gastrointestinal disease or from HPN-related complications). The mean duration of HPN treatment in Ng et al.'s cohort was 2.9 yrs (range 2 months to 7.5 yrs) [8]. In our study, one of the patients had been treated with HPN for 16 yrs. A paper in the Japanese literature reported a poorer outcome in patients with SSc on PN—two of four patients on HPN died of septicaemia [9].
The main indications for PN in our eight patients were intestinal pseudo-obstruction and malabsorption. All eight patients had bacterial overgrowth with malabsorption and weight loss and were initially managed with cyclical antibiotics, but symptoms failed to improve. Feeding via a PEG tube was tried in three patients but was not an option for the longer term.
Significant small bowel involvement, such as recurrent pseudo-obstruction and malabsorption, is associated with a poor prognosis [10]. During the 13-yr study period three of our patients died. This was as a result of cardiac or respiratory failure secondary to their SSc and none of the patients died as a direct result of their HPN use. Four patients were still using HPN at the end of the study period, and the fifth surviving patient no longer required HPN, highlighting how some patients require only temporary parenteral support. These observations suggest that the use of HPN improves survival for patients with intestinal failure secondary to SSc.
There was no increase in line infection rate compared with patients on HPN for other indications despite the propensity of patients with SSc to digital ulceration and infection. Although the line infection rate in our eight patients over the 13-yr study period was compared with the line infection rate in ‘other’ patients on HPN over only a 1 yr period at the end of the study period, there is no suggestion that the line infection rate has increased in the Intestinal Failure Unit during the 13-yr period. Therefore, even though patient numbers are small, we can be confident that patients with SSc are not at an increased risk of line infection when the care of the line is optimized. There may be a greater risk of line calcification affecting the HPN lines in patients with SSc, with two of the eight patients developing this rare complication. One of the patients developing line calcification had clinically apparent subcutaneous calcinosis elsewhere.
The HPN regimes in the patients with SSc included lower volumes of fluid infused over longer periods of time. This is because the amount of fluid losses in patients with SSc is relatively small compared with in those with other indications for HPN, many of whom have high output stomas, for example, in Crohn's disease. Also, large amounts of fluid can result in fluid overload in patients with SSc because of the concomitant pulmonary and cardiac involvement. All patients infused their HPN overnight. This meant that the HPN regime had less impact on their daily lives and ultimately on their quality of life than had it been infused during the day.
Limitations of our study were first that it was retrospective (with the result that it is possible that patients might not have been included if data entry on to either the SSc or Intestinal Failure Unit databases was incomplete) and second that the small number of SSc patients requiring HPN makes it difficult to make meaningful comparisons with those patients not requiring HPN. The small differences between ‘PN’ and ‘non-PN’ patients (higher proportion of patients with dcSSc, younger age at diagnosis) are unlikely to be clinically relevant. Nonetheless, a strength of the study is that is was based on patients attending a tertiary referral centre for both SSc and intestinal failure. A key point is that patients with either lcSSc or dcSSc may develop gastrointestinal involvement that can be severe.
In conclusion, HPN appeared to be well tolerated in our small cohort of patients and should be considered in patients with SSc who are unable to maintain their nutritional status.
Disclosure statement: The authors have declared no conflicts of interest.
Comments