-
PDF
- Split View
-
Views
-
Cite
Cite
M. Cuesta, A. Ortolá, A. Garrahy, A.L. Calle Pascual, I. Runkle, C.J. Thompson, Predictors of failure to respond to fluid restriction in SIAD in clinical practice; time to re-evaluate clinical guidelines?, QJM: An International Journal of Medicine, Volume 110, Issue 8, August 2017, Pages 489–492, https://doi.org/10.1093/qjmed/hcx036
Close - Share Icon Share
Abstract
Fluid restriction is recommended as first line therapy for Syndrome of Inappropriate Antidiuresis (SIAD), despite of lack of good evidence base to support its use, and poor efficacy in clinical practice and in the literature.
We set out to determine how many patients with well-defined SIAD had pre-treatment criteria which would predict failure to fluid restriction.
This was a consecutive, prospective evaluation of 183 patients with a diagnosis of SIAD in two different hospitals. Full ascertainment of the diagnostic criteria for SIAD was obtained in all patients.
About 47% of patients had a urine volume <1500 ml in 24 h, 41% had initial urine osmolality > 500 mOsm/kg, 26% a Furst-equation ratio > 1. About 59% had one criterion predicting failure to respond to fluid restriction, 37% two criteria, and 3% three criteria.
Our data suggest that up to 60% of patients with SIAD had criteria which recent clinical guidelines suggest would predict nonresponse to fluid restriction. This may explain why the recommended first line therapy for SIAD has been shown to be ineffective.
Background
Hyponatraemia is a frequent finding in clinical practice affecting between 15% and 30% of hospital in-patients.1 Hyponatraemia has been consistently associated with an increase in mortality in almost every condition with which it is associated, including patients admitted to internal medicine services,2 patients in the intensive care unit,3 patients with pneumonia,4 ambulatory patients,5 and oncological patients.6 In addition, there is evidence of increased morbidity associated with hyponatraemia of all degrees of severity, including mild hyponatraemia.7
The commonest cause of hyponatraemia in clinical practice is the Syndrome of Inappropriate Antidiuresis (SIAD). The causes of hyponatraemia due to SIAD are very heterogeneous, with the most common groups being secondary to central nervous disorders, malignancy, respiratory disorders, and drugs.8 The diagnosis of SIAD in hyponatraemia is critical, as the therapies to target hyponatraemia due to SIAD are different, usually opposed, to hyponatraemia secondary to hypo or hypervolaemia. Furthermore, inappropriate treatment of hyponatraemia has been associated with worse clinical outcome, when compared to patients who received correct treatment.9
Clinical guidelines from USA and Europe are consistent in their recommendation that fluid restriction should be first line therapy for hyponatraemia due to SIAD.10,11 However, the evidence base for the use of fluid restriction is very limited,10 with no randomized controlled trials in SIAD to support the recommendations for fluid restriction. In addition, fluid restriction is difficult to implement, particularly if patients require intravenous antibiotics or nutritional supplementation, and many patients are unable to tolerate the degree of restriction of fluid intake that is necessary to elevate plasma sodium concentration. Recent data, from a large multinational database of SIAD patients showed that in routine clinical practice, fluid restriction was no more effective than no treatment at all, in the management of SIAD.12 The US guidelines have recommended a number of parameters which can be used to predict nonresponse to fluid restriction.10 These include elevated urine osmolality > 500 mOsm/kg and urine output < 1500 ml/24 h, both of which reflect marked antidiuresis and the Fürst equation ratio (an estimation of the electrolyte-free water clearance through the formulae: urine sodium + potassium/plasma sodium > 1). The Fürst equation ratio is recommended to ascertain the degree of fluid restriction needed to correct hyponatraemia in SIAD. If the Fürst equation is 0.5–1; the patient should be commenced of fluid restriction up to 500 ml/day, if its <0.5 fluid restriction <1000 ml/day should be recommended. Moreover, if the Fürst equation result is > 1, there is no excretion of electrolyte-free water and fluid restriction is unlikely to be beneficial.13
<0.5 → restrict to <1000 ml all fluids
0.5–1 → restrict to <500 ml all fluids
>1 → fluid restriction unlikely to be effective enough to correct hyponatraemia. Consider other therapies.
Rationale of the Fürst equation ratio: Dilute urine can be divided in electrolyte-free water and an isotonic component. The proportion of electrolyte free water and isotonic urine can be measured by the ratio of effective solutes (principally sodium and potassium, with associated anions) between the plasma and the urine.14 Restriction of water intake to less than the amount of electrolyte-free water excreted will increase plasma tonicity and serum sodium. The ability of the kidneys to excrete electrolyte-free water will impact in the patient’s response to fluid restriction. Adapted from Grant et al.15
In view of the limited efficacy of fluid restriction to correct hyponatraemia in clinical practice,12 we designed the present study to ascertain how many patients with SIAD display pre-treatment criteria which predict failure to respond to fluid restriction in clinical practice.
Aim
We set out to determine how many patients with well-defined, consecutively recruited SIAD had pre-treatment criteria which would predict failure of fluid restriction
Methods
This is a cross-sectional, two centre-study. About 183 patients were consecutively and prospectively recruited, from acute admissions to Hospital Clínico San Carlos, Madrid (n = 51) and Beaumont Hospital, Dublin (n = 132) with a diagnosis of SIAD. The selection of patients with SIAD was performed in parallel in Madrid and Dublin over a 3-month period. The cohort from Madrid was recruited by A.O and M.C recruited the patients in Dublin. All patients with euvolaemic hyponatraemia were potentially included, if they fulfilled the criteria for SIAD (Table 1).16
Standard criteria for the diagnosis of SIAD16
| 1. Plasma hyposomolality (<275 mOsm/kg) |
| 2. Inappropriate urine concentration (urine osmolality >100 mOsm/kg) |
| 3. Urine sodium >30 mmol/l |
| 4. Clinical euvolemia |
| 5. Exclusion of hypothyroidism, secondary adrenal failure and recent use of diuretics. |
| 1. Plasma hyposomolality (<275 mOsm/kg) |
| 2. Inappropriate urine concentration (urine osmolality >100 mOsm/kg) |
| 3. Urine sodium >30 mmol/l |
| 4. Clinical euvolemia |
| 5. Exclusion of hypothyroidism, secondary adrenal failure and recent use of diuretics. |
Standard criteria for the diagnosis of SIAD16
| 1. Plasma hyposomolality (<275 mOsm/kg) |
| 2. Inappropriate urine concentration (urine osmolality >100 mOsm/kg) |
| 3. Urine sodium >30 mmol/l |
| 4. Clinical euvolemia |
| 5. Exclusion of hypothyroidism, secondary adrenal failure and recent use of diuretics. |
| 1. Plasma hyposomolality (<275 mOsm/kg) |
| 2. Inappropriate urine concentration (urine osmolality >100 mOsm/kg) |
| 3. Urine sodium >30 mmol/l |
| 4. Clinical euvolemia |
| 5. Exclusion of hypothyroidism, secondary adrenal failure and recent use of diuretics. |
Only patients with full ascertainment of the diagnostic criteria for SIAD were included; 100% of patients had recorded results for urine osmolality (>100 mOsm/kg) and urine sodium concentration (> 30 mmol/l), normal free thyroxine and TSH and 09:00 h cortisol > 300 nmol/l (10.8 mcg/dl). The plasma cortisol cut off was selected on the basis of our own published data, which showed no case of adrenal insufficiency in SIAD patients with a random cortisol > 300 nmol/l.8 Basic clinical data to predict failure to respond to fluid restriction was obtained but clinical decisions were left to admitting physicians. The study was approved as on noninterventional, observational audit by the ethics section of Beaumont Hospital Medical Research Committee.
Results
Median age was 75 years (IQR 64–82), 109 (59%) were female. The most common causes of SIAD were central nervous disorders (n = 38, 20.9%), pulmonary diseases (n = 36, 19.8%), malignancy (n = 36, 19.8%) and drugs (n = 16, 8.8%).
The median and IQR concentration at diagnosis were, respectively: plasma sodium 128 mmol/l (IQR 125–130 mmol/l), urine osmolality 451 mOsm/kg (347–590 mOsm/kg), urine sodium 62 mmol/l (43–88 mmol/l), urine potassium 31 mmol/l (21–44 mmol/l).
Choice of intervention was left to the discretion of the medical team managing the patient. About 82 patients (45%) received no therapy for SIAD. About 43 patients received fluid restriction (23%), 15 patients (8%) salt tablet supplements, 23 (13%) isotonic saline infusion, 3 patients (2%) oral Tolvaptan, and 17 patients (9%) several therapies.
Median plasma sodium at baseline was 129 mmol/l (IQR 127–130 mmol/l) in the group who received no therapy vs. 126 mmol/l (IQR 124–129 mmol/l) in the group with fluid restriction (P < 0.0001), suggesting that the decision not to treat may have been made on the basis of less severe hyponatraemia. About 13/43 (30%) patients in the group who received fluid restriction achieved a plasma sodium ≥ 135 mmol/l compared to 11/23 (48%) in the group receiving intravenous isotonic saline (P = 0.016) and 35/83 (43%) in the group without therapy (P = 0.24).
About 65/183 (35%) were receiving intravenous fluids at volumes ≥ 1500 ml in 24 h. In 115/183 (62%), the total administration of fluids (including isotonic/dextrose, oral, intravenous medication and nutritional support, oral/enteral/parenteral) was ≥ 1500 ml, and in 89/183 (48%) was ≥ 2000 ml; for all patients median total administration of fluids was 2000 ml (IQR 1200–2718 ml). A description of the fluid intake at the time of SIAD-diagnosis is provided in Table 2.
Description of the fluid intake in the cohort of 183 patients with SIAD at the time of diagnosis
| . | N (%) . |
|---|---|
| Intravenous fluids ≥1500 ml in 24 h | 65 (35) |
| Total fluid administration (intravenous fluid, oral fluid, nutritional support,) ≥1500 ml in 24 h | 115 (62) |
| Total fluid administration ≥2000 ml in 24 h | 89 (48) |
| Number of patients receiving oral or parenteral nutritional support | 20 (11) |
| . | N (%) . |
|---|---|
| Intravenous fluids ≥1500 ml in 24 h | 65 (35) |
| Total fluid administration (intravenous fluid, oral fluid, nutritional support,) ≥1500 ml in 24 h | 115 (62) |
| Total fluid administration ≥2000 ml in 24 h | 89 (48) |
| Number of patients receiving oral or parenteral nutritional support | 20 (11) |
Description of the fluid intake in the cohort of 183 patients with SIAD at the time of diagnosis
| . | N (%) . |
|---|---|
| Intravenous fluids ≥1500 ml in 24 h | 65 (35) |
| Total fluid administration (intravenous fluid, oral fluid, nutritional support,) ≥1500 ml in 24 h | 115 (62) |
| Total fluid administration ≥2000 ml in 24 h | 89 (48) |
| Number of patients receiving oral or parenteral nutritional support | 20 (11) |
| . | N (%) . |
|---|---|
| Intravenous fluids ≥1500 ml in 24 h | 65 (35) |
| Total fluid administration (intravenous fluid, oral fluid, nutritional support,) ≥1500 ml in 24 h | 115 (62) |
| Total fluid administration ≥2000 ml in 24 h | 89 (48) |
| Number of patients receiving oral or parenteral nutritional support | 20 (11) |
About 109/183 (59%) of patients had one predictor of failure to respond to fluid restriction, while 69/183 (37%) had two predictors. The results of the predictors of failure of fluid restriction are shown in the Table 3.
Number and percentage of patients with SIAD who had pre-treatment criteria which would predict failure to fluid restriction
| Predictor of failure of fluid restriction . | N . | % . |
|---|---|---|
| Urine volume <1500 ml/24 h | 49/103 | 47 |
| Urine osmolality >500 mOsm/kg | 75/183 | 41 |
| Fürst equation ratio>1 | 48/183 | 26 |
| Presence of one predictor | 109/183 | 59 |
| Presence of two predictors | 69/183 | 37 |
| Presence of three precitors | 6/183 | 3 |
| Predictor of failure of fluid restriction . | N . | % . |
|---|---|---|
| Urine volume <1500 ml/24 h | 49/103 | 47 |
| Urine osmolality >500 mOsm/kg | 75/183 | 41 |
| Fürst equation ratio>1 | 48/183 | 26 |
| Presence of one predictor | 109/183 | 59 |
| Presence of two predictors | 69/183 | 37 |
| Presence of three precitors | 6/183 | 3 |
Number and percentage of patients with SIAD who had pre-treatment criteria which would predict failure to fluid restriction
| Predictor of failure of fluid restriction . | N . | % . |
|---|---|---|
| Urine volume <1500 ml/24 h | 49/103 | 47 |
| Urine osmolality >500 mOsm/kg | 75/183 | 41 |
| Fürst equation ratio>1 | 48/183 | 26 |
| Presence of one predictor | 109/183 | 59 |
| Presence of two predictors | 69/183 | 37 |
| Presence of three precitors | 6/183 | 3 |
| Predictor of failure of fluid restriction . | N . | % . |
|---|---|---|
| Urine volume <1500 ml/24 h | 49/103 | 47 |
| Urine osmolality >500 mOsm/kg | 75/183 | 41 |
| Fürst equation ratio>1 | 48/183 | 26 |
| Presence of one predictor | 109/183 | 59 |
| Presence of two predictors | 69/183 | 37 |
| Presence of three precitors | 6/183 | 3 |
Discussion
Our data demonstrate that a significant proportion of a well-defined, prospective and consecutive cohort of SIAD patients had clinical parameters that the US guidelines for hyponatraemia would suggest could predict failure of response to fluid restriction. About 59% of patients had one criterion and 37% of patients had two predictors. This would suggest that almost two-thirds of SIAD patients can be predicted at the time of diagnosis to fail to respond to fluid restriction. This might explain why the observations on clinical efficacy of fluid restriction in the large SIAD cohort in the SIAD registry were so disappointing;12 if two-thirds of patients could be predicted to fail to respond to fluid restriction, it would hardly be surprising that the response to fluid restriction was no better than to no treatment at all. The basal characteristics and biochemical results at time of diagnosis are similar to recent reports of patients with SIAD.12 Our cohort differs from published cohorts in that there was higher fulfilment of the diagnostic criteria recommended for SIAD, and in particular the exclusion of underlying glucocorticoid deficiency. In addition, as our cases were recruited consecutively and without selection; our data contains no bias, and therefore the internal validity of this observational study is likely to be high, and applicable to a broad SIAD patient population.
As the ethical approval was for an observational study only, we had no influence over treatment choices; the therapies selected were so varied, that we could not test whether the subsequent response to fluid restriction validated the advice in the US guidelines. This might be seen as a limitation of our data, but response to therapy was not a primary endpoint of the study which was to establish the frequency of negative predictive data; to document the response to therapy would require a randomized controlled prospective study. The current article provides the relevant data with which to apply a power calculation for the intervention study. The results of our series and those from a larger cohort of patients documented in the International Registry of hyponatraemia12 highlight that a significant subgroup of patients, especially those with hyponatraemia > 125 mmol/l, receive no therapy to target hyponatraemia due to SIAD in routine clinical practice. As mortality has been clearly demonstrated to be elevated in hyponatraemic patients, this therapeutic nihilism reflects the lack of intervention data to convince clinicians of the value of treatment. As the treatment/no treatment groups in our cohort were nonrandomized, and therefore contained clear clinical differences, such as the severity of hypnatraemia, assessment of the response to treatment would not have been appropriate, and may have produced misleading conclusions based on unrandomized options. In addition, a significant proportion of patients were receiving inappropriate high volumes of intravenous fluids, which may have; triggered or sustained hyponatraemia, assessment of interventions would have been erroneous. Avoiding fluid overload in SIAD is of paramount importance, since it can aggravate dilutional hyponatraemia. The electrolyte concentration of the fluid given must exceed the electrolyte concentration of the urine, nor simply of the plasma.17 For example, in one report, 22 patients with postoperative SIAD received either isotonic saline or lactated Ringeŕs solution, and 21 of them experienced a fall in serum sodium concentration by mean of 4 mmol/l.18 Finally as the urine collections for electrolytes were taken in real world clinical practice, the calculated Fürst equation ratio may have been affected by prior volume expansion or fluid restriction, which should be acknowledged in the analysis of the data.
Interventional studies of patients with SIAD of different aetiologies, to test the real validity of the negative predictors of fluid restriction, both in ambulatory and inpatients and in acute and chronic hyponatraemia will be of high clinical value. Fluid restriction remains the primary intervention recommended by current guidelines,10 and although it can be useful in a subgroup of patients, it can be difficult to implement because of the co-prescription of intravenous antibiotics or nutritional supplements. As a result, noninterventional studies suggest that fluid restriction is not superior to no treatment in treating SIAD, although this is based on low quality evidence.
There remains a need for a randomized controlled trial of fluid restriction in SIAD, and our data suggest that measurement of the predictors of response should be incorporated into this trial. Until then, the role of fluid restriction as first line therapy for SIAD should be treated with caution, if over 50% of patients can be predicted to fail to respond to therapy.
Authorships: All authors had access to the data and a role in writing the article.
Conflict of interest: Prof. Isabelle Runkle has worked in an advisory capacity for Otsuka and given talks sponsored by Otsuka. Dr Martin Cuesta and Professor Thompson have received honoraria for lectures from Otsuka.
