Symptoms, Electrolyte Disturbances and Serum Albumin Levels in Palliative Oncology Patients Admitted Through Emergency: Characteristics and Survival Outcomes

Long-term outcomes

Sixty-six patients died within the hospitalisation period (30 had abnormal Na, 16 had abnormal K, 15 with abnormal Mg, 32 had abnormal Ca, and 47 had abnormal albumin). On long-term follow-up one year after the pre-defined study period (in January 2023), 138 patients died, eight (5.1%) survived, and follow-up was lost for seven patients. The proportion of patients with haematological malignancies who were surviving was more (27%) than the patients with solid tumours (3.7%).

Hyponatraemia was the most common electrolyte imbalance (43%), hypoalbuminaemia was observed in about two-thirds of the patients (66%), and one-third of patients had deranged renal function (33%). On analysis of each subgroup in terms of survival, we observed a weak positive correlation between Na levels and outcome (r = 0.199, P = 0.016) and a strong positive correlation between albumin levels and survival outcomes (r = 0.329, P = 0.000). Patients with normal albumin had a higher chance of survival (odds ratio: 33.1225, 95% confidence interval: 3.415–321.20, P = 0.003). Patients who had long-term survival had normal sodium levels and had no history of electrolyte imbalance. The results are shown in Table 2 (long-term outcome), Table 3 (median days of survival), and Table 4 (bivariate analysis: Levels of electrolytes with outcome: Spearman’s rho), and the survival curve plotted against sodium and albumin levels is shown in Figures 1 and 2, respectively.

Close

Figure 1:

Survival curve depending on sodium levels (attached).

Export to PPT

Close

Figure 2:

Survival curve depending on albumin levels (attached).

Export to PPT

Hyponatraemia

Hyponatraemia is the most common tumour-related electrolyte imbalance.^[1] Nair et al. documented the prevalence of hyponatraemia in hospitalised patients after reviewing 2666 inpatient charts and found out that it was significant in palliative care patients.^[2] Similar studies document the prevalence of hyponatraemia in patients with cancer receiving palliative care.^[10] Castillo et al. reviewed the diagnosis and management of hyponatraemia in cancer patients. They identified hyponatraemia as a negative prognostic factor, usually caused by the syndrome of inappropriate antidiuretic hormone secretion.^[11] They observe variable reported rates of the prevalence of hyponatraemia, ranging from 15% in some studies to 25–44% in others. In our study, we observed the prevalence of hyponatraemia to be 43.3%, and it was the most common electrolyte abnormality observed. It was the most common electrolyte abnormality in the past medical history (11% of patients). There was a weak correlation between sodium levels and mortality.

Hypernatraemia

Hypernatraemia is relatively uncommon in cancer patients, as a study by Salahudeen et al. involving 3446 patients reported a frequency of 2.6%, as opposed to 46% for hyponatraemia.^[3] The study went on to point out that most cases were hospital-acquired, with an impact in terms of higher morbidity, longer stay, and higher financial costs. In this context, it may be an anticipated and preventable dyselectrolytemia. In our study, we have observed comparable numbers, with a prevalence of 2.5%. Seo et al. conducted a retrospective cohort study, from which they concluded that hypernatremia was predictive of shorter patient survival and emphasised the need for greater clinical attention to its prognostic utility.^[4] We could not draw any conclusion from our observations on hypernatremia in the present study.

Hypokalaemia

Although reviews mention hypokalaemia as a common electrolyte problem in cancer with multiple possible mechanisms that may cause it, there is a dearth of literature on the prevalence of hypokalaemia in the cancer population.^[1] Treatment in these conditions addresses the underlying cause, addresses the symptoms, and is aimed at preventing life-threatening complications that may potentially arise secondary to hypokalaemia.^[12,13] In the update by Kardalas et al., they cited the prevalence in hospitalised patients to be 20%, with about 5% being clinically significant.^[14] We observed the prevalence of hypokalaemia to be lower in our study (12.1%). In patients with oncology, symptoms of hypokalaemia, such as fatigue and ileus may mimic cancer fatigue or malignant bowel obstruction. More data are needed on the prevalence of this symptom in the cancer palliative care population to plan and guide treatment strategies.

Hyperkalaemia

Hougen et al. reported a retrospective cohort study of 96,337 patients, concluding that hyperkalaemia was an independent risk factor for all-cause mortality, cardiovascular events, hospitalisations, and ICU admissions.^[5] As in hypokalaemia, treatment remains directed towards preventing the accompanying life-threatening complications.^[15] Our study observed a prevalence of 7% in palliative oncology admissions. In our clinical practice, we observe that new-onset hyperkalaemia is commonly associated with worsening renal function, and patients with advanced cancers may have obstructive uropathies, urosepsis, or sepsis as a proximal cause. There remains a dearth of literature on the prevalence of hyperkalaemia in hospitalised cancer patients, and more studies may help anticipate and correct this silent symptom. Magnesium: There is a dearth of literature related to the prevalence of hypo- and hypermagnesaemia in hospitalised cancer patients.

Hypomagnesaemia

Magnesium is an intracellular cation, and its deficiency frequently occurs (7–12%) in hospitalised patients, and it is correlated to an increased risk of death.^[1] Our observations of prevalence are slightly higher (15.9%). Although the symptoms of hypomagnesaemia may be non-specific, we have observed it to occur in our practice in patients with poor oral intake, alongside hypokalaemia. In such cases, clinical correction of magnesium precedes that of potassium. In a study by López-Saca et al., two cases of opioid refractory pain attacks were described, which were corrected with magnesium replacement.^[16]

Hypermagnesaemia

Hypermagnesaemia is a rare electrolyte abnormality.^[1] Takahashi and Uchino. reported a cross-sectional study of 533 patients admitted to a palliative care centre, out of whom hypermagnesaemia was observed in 123 patients (23.08%).^[17] In this study, predictors of hypermagnesaemia included renal dysfunction, short prognosis prediction, and oral magnesium oxide laxative use, with a plausible incidence of hypermagnesaemia being 20%. However, our observations on hypermagnesaemia are much lower 1.9% than in this study.

Hypocalcaemia

Hypocalcaemia can arise from multiple causes.^[18] In the cancer population, the use of bone-directed treatments such as bisphosphonates and Denosumab are specific examples of causative factors for hypocalcaemia, with a reported incidence of 9.6% in patients treated with Denosumab.^[19] In a review of the literature, Kreutle et al. reported that the true incidence of hypocalcaemia was difficult to measure with bisphosphonate therapy.^[20] Broadbent et al. have described a case report of severe symptomatic hypocalcaemia following bisphosphonate therapy in a patient with advanced cancer, pointing to the significance of assessment of this ion status in palliative care.^[21] Hypocalcaemia can be caused by myriad mechanisms, and there is a dearth of literature about its prevalence in the palliative care cancer population. Our study reported a prevalence of 34.4% of hypocalcaemia.

Hypercalcaemia

Asonitis et al. reviewed the literature on hypercalcaemia of malignancy and cited how cancers were the more prevalent cause of hypercalcaemia in patients presenting to the emergency department.^[22] This electrolyte disturbance is the most common metabolic life-threatening emergency encountered in advanced cancers and is associated with a poor prognosis. It has been estimated to occur in 20% of all patients with cancer, and despite it being a common electrolyte disorder with emergency implications, there is a need to report the incidence of hypercalcaemia in patients with cancer. In this study, we noted the prevalence to be much lower in our hospitalised palliative care population presenting to the emergency (3.8%). Our findings are in line with studies reported in the literature.^[23]

Albumin

In a systematic review of 59 studies, the authors concluded that pre-treatment serum albumin levels provided prognostic significance in oncology.^[6] Hypoalbuminaemia may be classified using common terminology criteria for adverse events criteria:

1. Grade 1: Lower limit of normal (LLN) - 3 g/dL

2. Grade 2: LLN < 3–2 g/dL

3. Grade 3: 2 g/dL

4. Grade 4: Life-threatening consequences; urgent intervention indicated

5. Grade 5: Death.^[24]

The relationship between albumin, protein levels, malnutrition, and survival is documented in the literature. In a study by Ñamendys-Silva, 164 out of 200 patients with advanced malignancy (82%) had serum albumin levels below 35 g/L, and 55.5% of patients had levels of albumin ≤20 g/L. The latter group had the highest mortality rate (73%).^[25] Serum albumin is known for its association with prognosis in oncology and cancer cachexia.^[6,7,25-28] In the nested case-control study by Liu et al., serum albumin had an independent negative linear association with 1-year mortality in cachectic patients in oncology.^[7] When serum albumin level was divided into quartiles, the authors observed that quartiles 3 and 4 (albumin levels >36.9 g/L) had higher survival than quartile 1 (albumin levels <33.1 g/L). Their observation was free of external influences such as age/cancer type/hospital/sex.

In our study, most patients had low albumin levels (66.2%), and hypoalbuminaemia was associated with a long-term mortality of 99%. Our study demonstrates the strong association of this cheap and easily available test with mortality and reinforces its utility as a practical prognostic marker in palliative oncology.

Our study reports the prevalence of electrolyte imbalance, hypoalbuminaemia, and associations with mortality in hospitalised patients in palliative oncology admitted through the emergency. There are limitations of our study due to its retrospective observational nature. Multiple factors may influence morbidity and mortality in patients with advanced disease. We included all patients in palliative oncology as one group for the purpose of the study; this may not reflect the study population in other centres. We cannot comment on prognosis based on the primary cancer type. The symptom clusters at the time of presentation to the emergency may differ possibly between palliative healthcare delivery systems and may not be generalised to the palliative oncology population. However, the associations noted between albumin levels and the high mortality in these patients seem to be consistent across patients with advanced cancer. With larger data and the increasing use of artificial intelligence, we may presuppose that more robust prognostic and predictive indicators may be developed using these variables.