It is well known that indiscriminate chemotherapy administration close to the end of life may be strongly associated with a greater number of invasive procedures preceding death, including ICU admissions, insertion of probes, deep venous access, and use of vasoactive drugs, among other interventions considered as “invasive measures” [8–11]. Some authors believe that chemotherapy in the last weeks of life is a low-quality factor in the assessment of patient care [12]. Nevertheless, urgent palliative chemotherapy is a treatment option for advanced disease in hospitalized patients of cancer, provided that there is some expectation of a positive response with an increase in survival or improvement of symptoms [3–5,13].
One biggest challenge in oncology practice is knowing which patients can benefit from oncological treatment when hospitalized in a severe clinical condition. The ECOG-PS score can help make this decision by assessing the functional status and allowing a certain prediction of treatment toxicity in patients with cancer. Moreover, other criteria previously reported in the literature as predictors of worse prognosis and unfavorable outcomes can be considered for patients with severe cancer. These criteria include age, tumor sensitivity to chemotherapy, previous comorbidities, hypercalcemia, and hyperbilirubinemia [7,12,14].
Our results showed that patients hospitalized with advanced solid neoplasia and poor clinical performance had a mean post-chemotherapy survival time of 74 days, which is longer than that reported in the literature 7,14–17]. Nevertheless, the number remains low, highlighting the poor prognosis of these patients, regardless of whether they receive systemic cancer treatment or complementary invasive measures during hospitalization. Another important point is that most previous studies used overall survival rather than post-chemotherapy survival as the main outcome. We consider that the option for the latter form was more appropriate in the temporal evaluation of these individuals, as it was possible to more reliably estimate the influence of an “urgent” chemotherapy regimen immediately after its institution. Given that all included patients would invariably die due to their advanced disease, we chose death or the last outpatient visit as the temporal endpoint. Surprisingly, this time interval exhibited great variability among the patients, reflecting the intrinsic differences between them.
We also evaluated post-chemotherapy survival or death within 30 days in relation to the clinical characteristics of the patients and observed a statistically significant association between their ECOG performance status and a reduction in the prescribed chemotherapy. ECOG, a validated clinical performance scale, was related to lower survival, as expected. This association is consistent with the literature, with previous studies reporting a clear relationship between poor clinical performance and poor prognosis [2,7,12]. Reduction in the prescribed chemotherapy dose was associated with a lower survival rate. This finding can be explained as a confounding factor because physicians administering treatment during hospitalization tend to be more cautious when prescribing chemotherapy to frail patients or those with poor overall clinical performance. However, this association reinforces the idea that even clinically borderline patients may need full-dose treatment when hospitalized with poor clinical performance, which can bring benefits in the short and medium terms. However, our study was not designed to answer this question, and prospective studies are required to confirm this hypothesis.
We also found an association between the outcome of death within 30 days after chemotherapy and the following laboratory variables: lower hemoglobin values, elevated direct bilirubin values, increased leukocyte and neutrophil counts, altered renal function with lower creatinine clearance and higher urea values, lower albumin, and higher C-reactive protein levels. These findings are consistent with the previous data on worse clinical outcomes associated with poor nutritional status (e.g., albumin levels), anemia, renal and liver dysfunction, and increased inflammatory marker levels [2,7,12,18–21]. Another biomarker that has been recently studied as a prognostic factor in solid cancers is the lymphocyte-to-monocyte ratio, with a higher ratio related to longer overall survival [22–26]; however, we did not analyze this in our work, given that there is no ideal cutoff established to safely use it objectively.
We also found an association between post-chemotherapy early death and systemic arterial hypertension. Although no study reports a direct association between hypertension and worse survival, we believe that this reflects the worse survival reported in patients with comorbidities in general [7,12].
However, in the multivariate analysis, variables with a statistically significant association in the binary evaluation lost their significance, with a positive association found for only albumin. This may have occurred due to the confounding factors and interactions between variables and may indicate a low power of association, perhaps due to the sample size, reinforcing the need for a new prospective and larger studies to further test these relationships.
As for the number of previous lines of chemotherapy, to our surprise, more than three quarters of the patients had not received systemic treatment previously (we call them “chemotherapy-sensitive”). We believe that these patients can be considered as being neglected by the system, because although they are part of the public healthcare system, their access to tertiary hospitals and medical subspecialties is often difficult and time consuming. These data motivated us to conduct an exploratory analysis of the subgroups of patients included in the study regarding whether they had previously undergone treatment; the aim was to assess whether there was a difference between the results of the tests performed. We observed that only “chemotherapy-sensitive” patients had lower survival rates with treatment dose reduction (i.e., there was a relationship between a lower dose and the worse outcome in only the first chemotherapy treatment), and with higher ECOG scores. These data suggest that treatment-naïve patients have worse outcomes and clinical performance; however, once the choice is made to prescribe chemotherapy to these hospitalized patients, it may not be worth reducing the dose. In the “polytreated” patient subgroup, greater survival was observed in patients with higher serum albumin levels, which may indicate that nutritional status plays a more important role in the final outcome for this subgroup.
In our study, approximately one-third of the patients were evaluated by the palliative care team during hospitalization. We questioned whether this low percentage was because we were being very “aggressive” in the treatment, viewing the patient primarily as a candidate for oncological treatments. This result even served as a point of reflection for improving our service, since we already have data in the literature relating to early access to the palliative care team to improve the quality of life and even ensure longer survival of patients with cancer [2,7,8,27].
A limitation of this study is its retrospective observational nature, which made it difficult to establish causal relationships, and the lack of a control group that did not receive chemotherapy during hospitalization. Nevertheless, this profile of patients with poor clinical performance is not routinely addressed in clinical studies; thus, given the low-quality and scarce data in medical literature, we believe our results are relevant.