Ovarian cancer is the eight cancer for incidence and at the fifth place for mortality in Europe in 2022 (1). The prognosis, although improved in recent years, remains poor, especially in advanced stages, and most of the patients diagnosed with Fédération Internationale de Gynécologie et d’Obstétrique (FIGO) stage III or IV tumors will develop recurrence within 18 months (2, 3). At the fifth recurrence, survival is around five months (4). The term ‘end of life’ is frequently used but poorly defined (5). The European Society for Medical Oncology (ESMO) Guideline refers to ‘end of life’, for people with advanced disease, as ‘point of rapid physical decline, typically the last few weeks or months before an inevitable death as a natural result of a disease’ (6). ESMO Guidelines explicitly contraindicate chemotherapy and immunotherapy in the last weeks of life (6). Chemotherapy in the last month of life is associated with adverse outcomes including poor quality of care, emergency department attendance, cardiopulmonary resuscitation, mechanical ventilation and dying in an intensive care unit. These items are defined as indicators of aggressive end of life care by several authors in the literature (7, 8).

Despite these indications, there are large retrospective registry-based American studies showing aggressive care in the final months of life in patients with ovarian cancer, including chemotherapy administration, emergency room access, hospital and intensive care unit admissions (9–14). Other retrospective and registry-based trials conducted in different countries around the world have shown a similar scenario of aggressive treatment at the end of life (15–17). In contrast, there is little data on the European reality on this topic. A large Dutch registry-based study has been published in recent years: of the 1775 patients included, only 12% received chemotherapy in the last month of life (18). The only data on the Italian scenario come from the EOLO study, a retrospective single-center Italian study, that included 110 patients dead of advanced/recurrent ovarian cancer and 38% of these had chemotherapy during the last month of life (19). The cut-off of chemotherapy administration at the end of life varies between studies (9–19). 30 days is the one most used regarding ovarian carcinoma (6, 9, 12, 16, 18–21).

Several parameters representing systemic inflammatory responses have been reported to be prognostic parameters in patients with ovarian cancer. Among these the neutrophil/lymphocyte ratio has been described as negative prognostic factor in ovarian cancer, as well as in other tumor types (22). Furthermore, in a study of patients with recurrent ovarian cancer, higher neutrophil/lymphocyte ratio was identified as prognostic factors for mortality within 100 days of failure of last line of chemotherapy (23).

The modified Glasgow Prognostic Score, which is classified as 0, 1, or 2 calculated by combining C-reactive protein and albumin values, has also been identified as an independent predictor of overall survival in patients with recurrent ovarian cancer (24). Also, C-reactive protein value has been described in literature as negative prognostic parameter in advanced cancer and in ovarian cancer (16, 25) The aim of this study is to describe the Italian reality about chemotherapy in the end of life in a cohort of ovarian cancer patients, and to identify factors associated with the probability of receiving chemotherapy in this period.

A total of 603 patients from 10 MITO centers participating (response rate 6%, 10/160 contacted centers) were analyzed ( Table 1 ).

The TO CARE/MITO 42 is the first multicenter study that analyzed the Italian practice about chemotherapy in the end of life in ovarian cancer patients. Our case series had 603 patients, from ten Italian centers, 5 centers in Northern Italy and 5 centers in Central and Southern Italy, all belonging to the MITO group, a cooperative research group in the field of gynecologic oncology. The fact that we chose only centers belonging to the MITO group allowed us to select a homogeneous case series, both from the surgical and oncological treatment point of view and reasonably representative of the Italian landscape. About one patient out of four (25.7%, 155/603) received chemotherapy in the last thirty days of life, in contrast to ESMO Guidelines 2021 (6), highlighting how this aggressive treatment at the end of life is still to be considered a very complex and under-studied topic.

The neutrophil/lymphocyte ratio is reported to be a prognostic factor in ovarian cancer, as well as in other tumor types (23). In a meta-analysis including nearly 3000 patients from 10 trials, patients with high pre-treatment neutrophil/lymphocyte ratio were found to have worse overall survival and progression free survival than patients with low pre-treatment neutrophil/lymphocyte ratio (22). In our case series, this subgroup of patients with neutrophil/lymphocyte ratio ≥5 at last-line initiation, using the cut-off used in the study by Roncolato et al. (20), in addition to having worse prognosis, in agreement with the literature, also showed to be more likely to receive, improperly, chemotherapy at the end of life (p<0.001). Thus, this is a selection of worse prognosis patients, potentially at risk of death in the short term, as also shown in the study by Kiuchi et al. (21)

Elevated C-reactive protein is also considered a negative prognostic factor for ovarian cancer patients (25, 26). We observed how patients with elevated C-reactive protein (cut-off of 5 mg/dL) at the start of the last line of chemotherapy seem to be more likely to receive chemotherapy at the end of life (p=0.05). This result is borderline statistical significance; however, it should be considered that this data was only available for 139 patients (23.1% of the total case series), reducing the statistical power of the analysis. In the study by Utsumi et al. (16) a lower performance status, high white blood cell count and high C-reactive protein were significantly correlated with shorter survival after the last treatment performed (p=0.004, 0.006 and 0.027, respectively). These observations spur us to perform complete blood tests on patients before starting a new line of treatment, including, of course, complete blood count but also laboratory parameters indicative of systemic inflammation, such as, indeed, C-reactive protein.

Death in hospital is among factors considered from the National Quality Forum as an indicator of aggressive medical care (27). In our case series, patients who died in the hospital (99 patients), compared with patients who died having activated palliative care at home or admitted to hospice (232 patients), were more likely to be treated with chemotherapy at the end of life (p<0.001). This result also agrees well with the literature, as shown by studies by Kajiyama and colleagues conducted in Japan, albeit on a smaller case series (28, 29). Recognizing when a patient can benefit from early activation of palliative care even during active treatment, is crucial and can make the transition between active treatment and best supportive care less difficult and sharp, avoiding unnecessary and harmful treatments for the patient, but also costly for public health (10).

The association between the geographical distribution of the center where the patient was treated (North vs. Center-South of Italy) and chemotherapy administration in the last 30 days is interesting: only 19.5% of patients enrolled in centers in the North received chemotherapy at the end of life, compared with 41.7% of patients from the Center-South (p<0.001). There are objective and documented differences in palliative care services between Northern and Southern Italy, as well as different management of resources in the territory (30). It should also be noted that, in our case series, the majority of patients enrolled in centers in the north were from university hospitals or research centers compared to those enrolled in the south (n=391 from university centers in the north versus n=108 from the south). This imbalance in source centers could further explain the difference found in the management of these patients, as clinicians working at university hospitals or research centers may have more experience in treating patients at the end of life.

These data urge us to have a more in-depth discussion among colleagues and, in the context of a multicenter group such as MITO, to look for ways to improve and try to increasingly smooth out such differences between centers. Patient enrollment within clinical protocols is a significant finding, but in contrast with already published data in the literature (31). In fact, being enrolled in at least one clinical protocol during disease history would lead to a lower likelihood of chemotherapy at the end of life (p=0.027). Certainly, the experience of “investigator” clinicians in MITO centers may have given reason for how they are able, especially for patients enrolled in research trials, to recognize when it is necessary to stop with active treatments. It is important to emphasize that these conclusions should be applied with caution to first-line patients. In fact, an outcome benefit has also been described for ovarian cancer patients who, despite their low performance status (Performance Status Eastern Cooperative Oncology Group ≥3), initiate first-line chemotherapy (32).

The retrospective nature of the study is a weakness and, in addition, the analyzed case series suffers from missing data. On the other hand, there is no Italian work of this magnitude in the literature (more than 600 patients). Our case series also included 63 patients with an interval from last chemotherapy to death of more than one year. It is more plausible that these patients did not die from ovarian cancer, although the retrospective nature of the case history does not allow us to be certain, as ‘cause of death’ data was not collected. However, we consider that since this population represents only 10% of all potential patient, this data did not impair the main analysis.