Method for identification of cancer patients with durable benefit from immunotehrapy in overall poor prognosis subgroups

A method of guiding treatment of lung cancer patient, a renal cell carcinoma patient, or a melanoma patient comprising the steps of:

• (a) conducting a mass spectrometer test on a blood-based sample of the cancer patient to obtain a mass spectrum;
• obtaining integrated intensity values of selected features in the mass spectrum at one or more m/z ranges as shown in Table 25;
• using the integrated intensity values in a first stage classification algorithm using a training set comprising class-labeled spectra produced from blood-based samples from the same type of cancer patients to identify the patient as being in a class of patients determined to be a poor prognosis subgroup, and
• (b) identifying the patient as being in the class of patients determined to be a poor prognosis subgroup, obtaining integrated intensity values of selected features in the mass spectrum at one or more m/z ranges as shown in the table and using the integrated intensity values in a second stage classification algorithm using a training set comprising class-labeled spectra produced from blood-based samples from the same type of cancer patients treated with an immunotherapy drug to obtain a class label of Late or the equivalent which guides treatment of the patient to an immunotherapy drug.

A method indicating the relative likelihood of success of an immunotherapy treatment for a lung cancer patient, a renal cell carcinoma patient, or a melanoma patient comprising the steps of:

• (a) conducting a mass spectrometer test on a blood-based sample of the cancer patient to obtain a mass spectrum;
• obtaining integrated intensity values of selected features in the mass spectrum at one or more m/z ranges as shown in Table 25 using the integrated intensity values in a first stage classification algorithm using a training set comprising class-labeled spectra produced from blood-based samples from the same type of cancer patients to identify the patient as being in a class of patients determined to be a poor prognosis subgroup,
• (b) identifying the patient as being in the class of patients determined to be a poor prognosis subgroup, using the integrated intensity values as shown in the table in a second stage classification algorithm using a training set comprising class-labeled spectra produced from blood-based samples from the same type of cancer patients treated with an immunotherapy drug to a class label of Late or the equivalent which identifies the patient as likely to have durable benefit from an immunotherapy drug;
• and
• (c) determining that the patient having a class label of Late or the equivalent is likely to respond to an immunotherapy drug.