User login
Lung cancer patients could soon have their risk of dying over the following 3 months accurately predicted by analyzing their urine samples, allowing them to better prepare for their end of life, say U.K. researchers.
Dr. Seamus Coyle, consultant in palliative medicine, the Clatterbridge Cancer Centre, Liverpool, and colleagues studied urine samples from more than 100 lung cancer patients, deriving a model based on their metabolite profile.
This allowed patients to be divided into high- and low-risk groups for dying over the following 3 months, with an accuracy of 88%.
The model “predicts dying … for every single day for the last 3 months of life,” Dr. Coyle said.
“That’s an outstanding prediction,” Dr. Coyle added, “based on the fact that people actively die over 2 to 3 days on average,” while “some die over a day.”
He continued: “It’s the only test that predicts dying within the last 2 weeks of life, and that’s what I’m passionate about: The earlier recognition of dying.”
The research was presented at the 2021 American Society of Clinical Oncology Annual Meeting on June 4.
‘Promising and important pilot study’
Dr. Nathan Pennell, an ASCO expert, told this news organization that “predicting the actual ‘time’ someone has left is more of an art than a science.”
“For people who may be closer to death, this would potentially allow more focus on supportive care and allow families and patients to plan more accurately for supporting their loved one through the dying process.”
He continued that “while this is a promising and important pilot study, there is more work to be done before this could be used in practice.”
For example, the treatment status of the patients was not clear.
“Were these patients all in hospice, or were some undergoing treatment which, if effective, could ‘rescue’ them from their poor prognostic state?”
Dr. Pennell continued: “Would measuring kidney function be just as good? Is this something that could be intervened upon?
“For example, if someone has a high-risk score for dying, could medical intervention to treat an infection or some other modifiable action change that ‘fate’?”
Death ‘difficult to predict’
Dr. Coyle began by saying that, while for him recognizing that a patient is dying is the start of good end of life care, “recognizing dying accurately, when someone is in the last days of life, is difficult.”
He noted that the 2019 National Audit of Care at the End of Life found that people were recognized to be dying at median of 34 hours before death, with 20% recognized in the last 8 hours.
Moreover, 50% of people who are dying “are unconscious and unable to be involved in any conversation that [is] pertinent to them.”
In an attempt to better predict the onset of dying, the researchers conducted a prospective, longitudinal study in which 424 urine samples were collected from 162 lung cancer patients from six centers.
Of those, 63 patients gave a sample within the last 28 days of life, and 29 within the last week of life.
Urine samples were analyzed using a liquid chromatography quadrupole time-of-flight mass spectrometer for 112 patients, who had a median age of 71 years and a range of 47-89 years, and 40.2% were female. The most common diagnosis was non–small cell lung cancer, in 55.4%, while 19.6% had small cell lung cancer.
Performing Cox Lasso regression analysis on the “hundreds of metabolites” identified in the urine samples, the team developed an End of Life Metabolome (ELM) that predicted an individual’s risk of dying over the following 3 months.
Kaplan-Meier analysis allowed the patients to be divided into five risk groups based on their ELM (P < .001 for trend), which showed that all patients in the lowest-risk group were still alive after more than 2 months following the urine sample.
In contrast, more than 50% of patients in the highest-risk group died within 1 week of their urine sample being taken, and 100% had died within 3 weeks.
Calculating the area under the receiver operating characteristic curve revealed that the ELM was able to predict the risk of dying for every day for the last 3 months of life with an accuracy of 88%.
ELM is being validated in a new cohort of lung cancer patients and it is being assessed in multiple cancers.
The study was funded by the Wellcome Trust UK and North West Cancer Research UK.
No relevant financial relationships were declared.
A version of this article first appeared on Medscape.com.
Lung cancer patients could soon have their risk of dying over the following 3 months accurately predicted by analyzing their urine samples, allowing them to better prepare for their end of life, say U.K. researchers.
Dr. Seamus Coyle, consultant in palliative medicine, the Clatterbridge Cancer Centre, Liverpool, and colleagues studied urine samples from more than 100 lung cancer patients, deriving a model based on their metabolite profile.
This allowed patients to be divided into high- and low-risk groups for dying over the following 3 months, with an accuracy of 88%.
The model “predicts dying … for every single day for the last 3 months of life,” Dr. Coyle said.
“That’s an outstanding prediction,” Dr. Coyle added, “based on the fact that people actively die over 2 to 3 days on average,” while “some die over a day.”
He continued: “It’s the only test that predicts dying within the last 2 weeks of life, and that’s what I’m passionate about: The earlier recognition of dying.”
The research was presented at the 2021 American Society of Clinical Oncology Annual Meeting on June 4.
‘Promising and important pilot study’
Dr. Nathan Pennell, an ASCO expert, told this news organization that “predicting the actual ‘time’ someone has left is more of an art than a science.”
“For people who may be closer to death, this would potentially allow more focus on supportive care and allow families and patients to plan more accurately for supporting their loved one through the dying process.”
He continued that “while this is a promising and important pilot study, there is more work to be done before this could be used in practice.”
For example, the treatment status of the patients was not clear.
“Were these patients all in hospice, or were some undergoing treatment which, if effective, could ‘rescue’ them from their poor prognostic state?”
Dr. Pennell continued: “Would measuring kidney function be just as good? Is this something that could be intervened upon?
“For example, if someone has a high-risk score for dying, could medical intervention to treat an infection or some other modifiable action change that ‘fate’?”
Death ‘difficult to predict’
Dr. Coyle began by saying that, while for him recognizing that a patient is dying is the start of good end of life care, “recognizing dying accurately, when someone is in the last days of life, is difficult.”
He noted that the 2019 National Audit of Care at the End of Life found that people were recognized to be dying at median of 34 hours before death, with 20% recognized in the last 8 hours.
Moreover, 50% of people who are dying “are unconscious and unable to be involved in any conversation that [is] pertinent to them.”
In an attempt to better predict the onset of dying, the researchers conducted a prospective, longitudinal study in which 424 urine samples were collected from 162 lung cancer patients from six centers.
Of those, 63 patients gave a sample within the last 28 days of life, and 29 within the last week of life.
Urine samples were analyzed using a liquid chromatography quadrupole time-of-flight mass spectrometer for 112 patients, who had a median age of 71 years and a range of 47-89 years, and 40.2% were female. The most common diagnosis was non–small cell lung cancer, in 55.4%, while 19.6% had small cell lung cancer.
Performing Cox Lasso regression analysis on the “hundreds of metabolites” identified in the urine samples, the team developed an End of Life Metabolome (ELM) that predicted an individual’s risk of dying over the following 3 months.
Kaplan-Meier analysis allowed the patients to be divided into five risk groups based on their ELM (P < .001 for trend), which showed that all patients in the lowest-risk group were still alive after more than 2 months following the urine sample.
In contrast, more than 50% of patients in the highest-risk group died within 1 week of their urine sample being taken, and 100% had died within 3 weeks.
Calculating the area under the receiver operating characteristic curve revealed that the ELM was able to predict the risk of dying for every day for the last 3 months of life with an accuracy of 88%.
ELM is being validated in a new cohort of lung cancer patients and it is being assessed in multiple cancers.
The study was funded by the Wellcome Trust UK and North West Cancer Research UK.
No relevant financial relationships were declared.
A version of this article first appeared on Medscape.com.
Lung cancer patients could soon have their risk of dying over the following 3 months accurately predicted by analyzing their urine samples, allowing them to better prepare for their end of life, say U.K. researchers.
Dr. Seamus Coyle, consultant in palliative medicine, the Clatterbridge Cancer Centre, Liverpool, and colleagues studied urine samples from more than 100 lung cancer patients, deriving a model based on their metabolite profile.
This allowed patients to be divided into high- and low-risk groups for dying over the following 3 months, with an accuracy of 88%.
The model “predicts dying … for every single day for the last 3 months of life,” Dr. Coyle said.
“That’s an outstanding prediction,” Dr. Coyle added, “based on the fact that people actively die over 2 to 3 days on average,” while “some die over a day.”
He continued: “It’s the only test that predicts dying within the last 2 weeks of life, and that’s what I’m passionate about: The earlier recognition of dying.”
The research was presented at the 2021 American Society of Clinical Oncology Annual Meeting on June 4.
‘Promising and important pilot study’
Dr. Nathan Pennell, an ASCO expert, told this news organization that “predicting the actual ‘time’ someone has left is more of an art than a science.”
“For people who may be closer to death, this would potentially allow more focus on supportive care and allow families and patients to plan more accurately for supporting their loved one through the dying process.”
He continued that “while this is a promising and important pilot study, there is more work to be done before this could be used in practice.”
For example, the treatment status of the patients was not clear.
“Were these patients all in hospice, or were some undergoing treatment which, if effective, could ‘rescue’ them from their poor prognostic state?”
Dr. Pennell continued: “Would measuring kidney function be just as good? Is this something that could be intervened upon?
“For example, if someone has a high-risk score for dying, could medical intervention to treat an infection or some other modifiable action change that ‘fate’?”
Death ‘difficult to predict’
Dr. Coyle began by saying that, while for him recognizing that a patient is dying is the start of good end of life care, “recognizing dying accurately, when someone is in the last days of life, is difficult.”
He noted that the 2019 National Audit of Care at the End of Life found that people were recognized to be dying at median of 34 hours before death, with 20% recognized in the last 8 hours.
Moreover, 50% of people who are dying “are unconscious and unable to be involved in any conversation that [is] pertinent to them.”
In an attempt to better predict the onset of dying, the researchers conducted a prospective, longitudinal study in which 424 urine samples were collected from 162 lung cancer patients from six centers.
Of those, 63 patients gave a sample within the last 28 days of life, and 29 within the last week of life.
Urine samples were analyzed using a liquid chromatography quadrupole time-of-flight mass spectrometer for 112 patients, who had a median age of 71 years and a range of 47-89 years, and 40.2% were female. The most common diagnosis was non–small cell lung cancer, in 55.4%, while 19.6% had small cell lung cancer.
Performing Cox Lasso regression analysis on the “hundreds of metabolites” identified in the urine samples, the team developed an End of Life Metabolome (ELM) that predicted an individual’s risk of dying over the following 3 months.
Kaplan-Meier analysis allowed the patients to be divided into five risk groups based on their ELM (P < .001 for trend), which showed that all patients in the lowest-risk group were still alive after more than 2 months following the urine sample.
In contrast, more than 50% of patients in the highest-risk group died within 1 week of their urine sample being taken, and 100% had died within 3 weeks.
Calculating the area under the receiver operating characteristic curve revealed that the ELM was able to predict the risk of dying for every day for the last 3 months of life with an accuracy of 88%.
ELM is being validated in a new cohort of lung cancer patients and it is being assessed in multiple cancers.
The study was funded by the Wellcome Trust UK and North West Cancer Research UK.
No relevant financial relationships were declared.
A version of this article first appeared on Medscape.com.