Key Summary Points
Why carry out this study? Cytokine release syndrome (CRS) is a significant adverse event associated with immuno-oncology (IO) therapies that lack standardized patient-reported outcome measures (PROMs) to capture patient experiences. Understanding patient-reported symptoms and impacts of CRS is crucial for improving patient care and monitoring in clinical trials. This study aimed to develop a content-valid PROM, the IO-induced CRS patient diary, to capture the incidence, severity, and impact of CRS symptoms from the patient’s perspective. What was learned from the study? The study identified various CRS-related signs, symptoms, and impacts through literature review, clinical expert validation, and patient interviews, leading to the development and refinement of the IO-induced CRS patient diary. The IO-induced CRS patient diary is a comprehensive tool that effectively captures patient-reported symptoms and impacts of CRS, providing valuable insights for clinical trials and improving patient management. Future research should focus on further validating this PROM in broader and larger patient samples and exploring its use in clinical practice.
| Why carry out this study? |
| Cytokine release syndrome (CRS) is a significant adverse event associated with immuno-oncology (IO) therapies that lack standardized patient-reported outcome measures (PROMs) to capture patient experiences. |
| Understanding patient-reported symptoms and impacts of CRS is crucial for improving patient care and monitoring in clinical trials. |
| This study aimed to develop a content-valid PROM, the IO-induced CRS patient diary, to capture the incidence, severity, and impact of CRS symptoms from the patient’s perspective. |
| What was learned from the study? |
| The study identified various CRS-related signs, symptoms, and impacts through literature review, clinical expert validation, and patient interviews, leading to the development and refinement of the IO-induced CRS patient diary. |
| The IO-induced CRS patient diary is a comprehensive tool that effectively captures patient-reported symptoms and impacts of CRS, providing valuable insights for clinical trials and improving patient management. |
| Future research should focus on further validating this PROM in broader and larger patient samples and exploring its use in clinical practice. |
Introduction
Immuno-oncology (IO) therapy is a type of cancer treatment that leverages or enhances the immune system’s ability to recognize, target, and destroy tumor cells [ 1 , 2 ]. Clinical development of IO therapies has advanced exponentially in the past decade, with regulatory bodies worldwide approving the use of chimeric antigen receptor T (CAR-T) cells, monoclonal and bispecific antibodies, cytokines, and immune checkpoint inhibitors [ 3 , 4 ]. Compared with traditional anticancer therapies, such as chemotherapy, molecularly targeted therapies, and radiotherapy, IO therapy is considered to be more specific with a wider breadth of response [ 5 ]. Particularly in the case of metastatic cancers, IO therapy is associated with improved clinical outcomes and better quality of life than these abovementioned therapies [ 5 ].
However, because of immune hyperactivation with IO therapy, patients may experience various adverse events. Such adverse events can be either systemic or localized, affecting specific anatomical regions, including the stomach and intestines, skin, liver, endocrine system, and lungs [ 6 ]. IO therapy has been associated with cytokine release syndrome (CRS), a constellation of signs and symptoms that can present as early as a few hours after treatment administration [ 7 ].
CRS involves the overwhelming release of cytokines, which causes the immune system to go into a hyperinflammatory, unregulated state [ 8 ]. Its clinical presentation is heterogeneous: it can manifest as mild, with fever, nausea, fatigue, headache, rash, arthralgia, myalgia, and chills; or severe, with vascular leakage, disseminated intravascular coagulation, shock, and respiratory compromise, which may lead to multisystem organ failure or death [ 9 , 10 ].
At present, there is no standardized tool to collect information on patients’ experience of CRS signs, symptoms, and impacts. Existing clinician-reported grading systems do not fully capture the patient perspective, and there is a general lack of understanding of CRS, especially in non-CAR-T therapies [ 11 ]. According to clinical criteria, CRS events are mild to moderate; however, the patient-perceived severity of CRS signs or symptoms and impacts on their daily lives is not well understood. This gap highlights the need for tools that can reliably and consistently capture patient-reported experiences to support a more holistic and patient-centered approach to care [ 12 , 13 ].
In the present qualitative research study, we aimed to address this unmet need by exploring patients’ experiences of CRS, including the most important symptoms and impacts on how patients feel and function. In alignment with the US Food and Drug Administration (FDA)’s Patient-Focused Drug Development (PFDD) guidance [ 14 ], we sought to develop a novel, content-valid patient-reported outcome measure (PROM) to track the onset and resolution of IO therapy-induced CRS events (and their impacts on patients’ lives) in clinical trials.
Methods
This research study involved several steps: a targeted literature review (TLR) of published research and safety data, input from clinical experts, and concept elicitation (CE) interviews with patients informed the development of a conceptual model (CM) of IO therapy-induced (CRS), outlining the signs, symptoms, and impacts on people’s lives. The CM was used to draft a PROM for evaluating the onset and resolution of IO therapy-induced CRS events (and their impacts) in clinical trials; the draft PROM was refined and content-validated through cognitive debriefing (CD) interviews with clinicians and patients, and the PROM was then finalized for exploratory use in clinical trials (Fig. S1 ).
Ethical Approval
This study was conducted in accordance with the ethical principles that have their origin in the Declaration of Helsinki 1964 and its later amendments and is consistent with Good Clinical Practice and applicable regulatory requirements. Centralized approval was obtained from the Western IRB-Copernicus Group (WCG) IRB, including for all recruitment materials. IQVIA is responsible for submitting the application and all study documents for study approval to the IRB. The study was initiated only after IRB approval was granted, including the use of all recruitment materials.
Informed consent was obtained prior to performing any study-related procedures, including screening, and patients provided consent via the online consent form. Patients were informed that, if the results of the study were published for scientific purposes, the patient’s identity would remain confidential. Clinicians who participated in the CD interviews provided verbal consent at the start of the interview after the objectives and logistics of the interview were shared with them. They provided verbal consent to both participate and to also be audio-recorded.
A standard ethics approved informed consent process was followed for all patients involved in this study. This required their review and signature of an informed consent form document that clearly outlined the study background, objectives, and procedures (including screening, risks and benefits, compensation, confidentiality, information use and data sharing, and voluntary participation and withdrawal). Completed and signed forms were reviewed by the study team for completeness prior to any interviews being conducted. For participating clinicians, informed consent was obtained verbally at the start of each interview. Prior to providing consent, clinicians were informed of the study background and objectives, interview logistics, and information regarding their confidentiality and how their data will be used.
Developing a Conceptual Model of IO Therapy-Induced CRS
Targeted Literature Review
A TLR of published literature on patients receiving IO therapies from 2017 to 2022 was conducted to identify patient-reported signs and symptoms potentially associated with IO therapy-induced CRS events. The TLR was enriched with data from anonymized safety reports (Sanofi clinical trial adverse event data). Aggregated weighted prevalence, when available, was calculated for each sign and symptom, along with average onset and duration values. Details of the prevalence of each sign and symptom from the TLR are provided in the Supplementary Materials (Table S1 ).
Concept Elicitation Research with Clinical Experts
Semi-structured clinical expert interviews were conducted to further understand signs and symptoms of IO therapy-induced CRS events. Specifically, four clinical oncology or hematology specialists from the USA were asked to define IO therapy-induced CRS in 1:1 CE interviews. They were also asked to provide dimensional descriptors (onset, duration, frequency, and resolution) of common signs and symptoms they encounter frequently in the clinic, identify patient-reportable ones, and describe the impacts of CRS on patients’ lives. Following the clinical expert interviews, an advisory board meeting was held with five academic oncologists (four from the USA and one from Australia). They provided further input on the list of patient-reportable symptoms of an IO therapy-induced CRS event that were identified in the TLR and clinical expert interviews. Advisory board members also further expanded on the list of possible impacts experienced by patients who experience these events.
Concept Elicitation Interviews with Patients
Recruitment
In total, 14 patients were recruited via two third-party vendors and through partnership with two patient advocacy groups to add the lived-experience perspective to the list of signs, symptoms, and impacts of CRS, as well as to understand language used by patients. Each patient’s initial eligibility for the study was assessed using the patient-completed screening form and subsequently confirmed by the clinician in the confirmation of diagnosis form, on the basis of the specified inclusion and exclusion criteria. Inclusion criteria were: age > 18 years; diagnosis of any stage of advanced, metastatic, relapsed, or refractory cancer or hematologic malignancy (any indication); a CRS event ≤ 12 months before screening; and currently on or previously treated with non-CAR-T (including blinatumomab, pembrolizumab, atezolizumab, avelumab, dostarlizumab, durvalumab, ipilimumab, and nivolumab) or CAR-T therapy in the past 12 months (the first 11 interviews included only patients on non-CAR-T therapy, with the remaining 3 including patients receiving CAR-T therapy). Meanwhile, exclusion criteria were: currently participating in any investigational study for oncological treatment or participated in such a study in the past 12 months, or being unable to participate in a 60-min teleconference interview.
Interviews
Semi-structured CE interviews were conducted with patients on a virtual conferencing platform, following a semi-structured interview guide. These were audio recorded. Patients were first asked to describe and define what they understood to be a CRS event. They were thereafter asked open-ended questions that were designed to elicit spontaneous mentions of signs, symptoms, and impacts they experienced at the time of the CRS event. They were asked to rate the severity and bothersomeness of each symptom they reported and how the CRS event was managed, both by themselves and any healthcare professionals.
Analysis
The interviews were transcribed, with all personal information redacted. Transcripts were coded in MAXQDA Plus® qualitative data analysis software (2022 version) by trained, experienced qualitative analysts. The coding process identified and categorized patients’ descriptions of concepts regarding their CRS experience using both deductive (on the basis of a predefined coding framework developed from prior steps in this study) and inductive coding approaches.
Initial transcripts were dual coded until intercoder agreement was achieved. Overall, about 36% of transcripts from CE interviews were dual coded by assigned coders to establish and maintain intercoder alignment throughout the coding process. Discrepancies were discussed with the study team, and decisions implemented consistently across coders for all future transcripts. The number of patients who reported particular symptoms and impacts was tabulated using a quantitative descriptive approach.
As mentioned in the second FDA PFDD guidance [ 15 ], saturation is defined as the point when no new relevant or important information emerges, and collecting additional data will not likely add to the understanding of how patients perceive the concept of interest. In the present study, saturation of concepts was assessed by comparing the reporting of concept codes derived from one group of interviews with the next group of interviews, identifying if the same or new concepts were appearing in subsequent interviews.
CRS Conceptual Model Development
Data from the TLR, clinical expert interviews, oncologist advisory board, and CE interviews with patients were qualitatively collected to inform development of a CM of IO therapy-induced CRS, outlining signs, symptoms, and impacts on patient lives.
PROM Development
A list of prioritized patient-reportable symptoms and impacts was generated for inclusion in the PROM on the basis of the CM. Specifically, symptoms were considered on the basis of weighted prevalence in the TLR and prevalence in anonymized safety reports; whether clinical experts flagged the symptom as common; whether the advisory board recommended its inclusion in the PROM; the average time of onset and resolution identified from anonymized safety reports; and the frequency, severity ratings, and most bothersome symptom reported by patients. For each of these signs and symptoms, prevalence data were extracted from the source reporting the highest prevalence and from two safety reporting datasets; while no minimum prevalence threshold was set, signs and symptoms were organized and examined from highest to lowest weight prevalence (Table S1 ).
When a list of prioritized symptoms and impacts had been identified, they were mapped to potentially applicable existing PROM item libraries from the European Organisation for Research and Treatment of Cancer, Functional Assessment of Chronic Illness Therapy, and National Cancer Institute groups. Item wording and response scales of mapped PROM items were examined against the language patients used in the CE interviews, and the appropriateness of the recall period was evaluated on the basis of onset and resolution data and on collected patient descriptions. Ultimately, items were drafted de novo for the PROM using FDA guidance on clinical outcome assessment development as guidance [ 14 ] as no existing item libraries had items and features applicable to all prioritized concepts.
Refinement of the PROM
Cognitive Debriefing Interviews with Clinical Oncologists
CD interviews were held with three clinical oncologists who participated in the advisory board to obtain feedback on the draft PROM. Before each interview, the clinician was informed of the development history, purpose, and context of use for the PROM. They were then asked to confirm the comprehensiveness, grouping, clinical relevance, and clinical meaningfulness of each symptom item and response scale. Finally, they were asked to comment on the impact items and their specific phrasing, and for their perspectives on the proposed schedule of assessment and potential patient burden. The PROM was refined prior to patient interviews.
Cognitive Debriefing Interviews with Patients
Recruitment
A total of 15 patients who received non-CAR-T therapy were recruited from among those who participated in the CE interviews, those who were approached for CE but did not qualify, and additional patients from a patient advocacy group.
Interviews
Interviews followed a semi-structured interview guide. Patients were asked to complete each section of the draft PROM using think-aloud techniques. They were thereafter asked for their interpretation of and feedback on instructions, items, and response options in the draft PROM. The interviews were transcribed, with all personal information redacted.
Analysis
Patient CD interviews were conducted in three waves of five patients each. Between each wave, the feedback obtained was analyzed thematically to identify trends across patients. Modifications were made to the draft PROM, interview guide, and approach between each wave. Feedback from the entire sample was reviewed after the third wave to determine whether any final modifications or updates to the PROM were needed.
Finalization of the PROM
The draft PROM was finalized as the IO-induced CRS patient diary. Formatting was updated across the PROM to add emphasis to different components (e.g., recall period, concepts, and severity).
Results
Developing a Conceptual Model of IO Therapy-Induced CRS
Targeted Literature Review
Overall, 19 signs and symptoms from CRS-like events were identified from the TLR, and 26 were identified from anonymized safety data. In total, 10 of the 19 signs and symptoms identified in the TLR were also identified from anonymized safety data, meaning that 35 unique CRS-related signs and symptoms were identified across the sources.
Concept Elicitation Interviews with Clinical Experts
The clinical experts confirmed the relevance of all signs and symptoms from the TLR, except dysfunctional blood regulation, which they felt was nonspecific and poorly understood.
The experts described eight new signs or symptoms that were not identified in the TLR but were considered to be related directly to CRS: malaise, arthralgia, diarrhea, myalgia, edema, redness at site, change in mental status, and brain fog. The advisory board generated a shortlist of the most clinically important CRS signs and symptoms and sorted them by organ class. They recommended removal of four signs (desaturation, capillary leakage, dysfunctional blood regulation, and hypertension) and five symptoms (flu-like symptoms, lethargy, lack of vigor, drowsiness, and loss of following instructions) from the list derived from the TLR and clinical expert interviews. This left 17 signs and 29 symptoms. They then proposed the addition of one further symptom (chest pain). They differentiated between patient-reportable and nonpatient-reportable signs and symptoms. In addition, they identified and described eight broad impacts likely relevant across patients: residual or lingering fatigue, increased desire to rest, physical limitation owing to feeling unwell, impact on desire to do instrumental activities of daily living, avoidance of additional healthcare interactions, anxiety regarding future treatment, emotional burden on family, and symptom-related distress.
Concept Elicitation Interviews with Patients
A total of 14 US-based patients who experienced IO therapy-induced CRS were interviewed for CE: 11 patients experienced CRS while on non-CAR-T therapy and three patients experienced CRS while on CAR-T therapy. Two of the non–CAR-T patients had also received CAR-T therapy (Table 1 ). Table 1 Demographic characteristics of interviewed patients Characteristic Patients in CE interviews ( n = 14) Patients in CD interviews ( n = 15) Gender Male 6 5 Female 5 7 Prefer not to say 2 2 Non-binary 1 1 Race White 6 8 Black/African American 4 4 Preferred not to say 3 3 Native Hawaiian 1 – Ethnicity Non-Hispanic 10 10 Hispanic 2 3 Prefer not to say 2 2 Age Mean (years) 56.8 55.3 Standard deviation (years) 8.0 10.6 Range (years) 44.0–70.0 35.0–71.0 Marital status Married 10 11 Divorced 2 4 Widowed 2 – Employment status Retired 5 4 Disabled 3 2 Homemaker 2 2 Unemployed 2 2 Prefer not to say 2 2 Employed full time – 2 Employed part time – 1 Highest level of education attainment a High school diploma 3 4 College degree 3 2 Some college 3 2 Associate degree 2 - General educational development (GED) b – 1 Graduate degree 1 – Bachelor’s degree – 4 Master’s degree 1 1 Postgraduate – 1 Doctorate 1 – Cancer type Multiple myeloma 4 c Lung 3 Melanoma 2 Kidney 2 Colorectal 1 Lymphoblastic leukemia 1 c Cancer stage 1 1 2 1 3 3 3/4 2 d 4 7 Site of metastases c Lymph nodes 7 Bones 4 Lungs 1 IO therapy Pembrolizumab e 6 Nivolumab e 4 Teclistamab e 1 Idecabtagene vicleucel f 1 Brexucabtagene autoleucel f 1 CAR-T cell therapy (unspecified) + teclistamab f 1 g CD cognitive debriefing, CE concept elicitation, GED general educational development, IO immuno-oncology a “Highest education attained” data were obtained from the screener when data from interviews were not available b GED is a high school equivalency certification c One patient was recorded as having multiple diagnoses at the time of screening, including multiple myeloma and lymphoblastic leukemia d Two patients were recorded as having stage 3 cancer at the time of screening and stage 4 cancer at the time of the interview e Non-CAR-T cell therapy f CAR-T cell therapy g One patient received both an unspecified CAR-T cell therapy and teclistimab and reported experiencing CRS attributed to the CAR-T therapy. Although the CRS event occurred more than 12 months prior to the study, the patient was included on the basis of their receipt of IO therapy (teclistamab) within the past 12 months and their strong recollection of the CRS
Table 1
Demographic characteristics of interviewed patients
CD cognitive debriefing, CE concept elicitation, GED general educational development, IO immuno-oncology a “Highest education attained” data were obtained from the screener when data from interviews were not available b GED is a high school equivalency certification c One patient was recorded as having multiple diagnoses at the time of screening, including multiple myeloma and lymphoblastic leukemia d Two patients were recorded as having stage 3 cancer at the time of screening and stage 4 cancer at the time of the interview e Non-CAR-T cell therapy f CAR-T cell therapy g One patient received both an unspecified CAR-T cell therapy and teclistimab and reported experiencing CRS attributed to the CAR-T therapy. Although the CRS event occurred more than 12 months prior to the study, the patient was included on the basis of their receipt of IO therapy (teclistamab) within the past 12 months and their strong recollection of the CRS
| Characteristic | Patients in CE interviews ( n = 14) | Patients in CD interviews ( n = 15) |
| Gender | ||
| Male | 6 | 5 |
| Female | 5 | 7 |
| Prefer not to say | 2 | 2 |
| Non-binary | 1 | 1 |
| Race | ||
| White | 6 | 8 |
| Black/African American | 4 | 4 |
| Preferred not to say | 3 | 3 |
| Native Hawaiian | 1 | – |
| Ethnicity | ||
| Non-Hispanic | 10 | 10 |
| Hispanic | 2 | 3 |
| Prefer not to say | 2 | 2 |
| Age | ||
| Mean (years) | 56.8 | 55.3 |
| Standard deviation (years) | 8.0 | 10.6 |
| Range (years) | 44.0–70.0 | 35.0–71.0 |
| Marital status | ||
| Married | 10 | 11 |
| Divorced | 2 | 4 |
| Widowed | 2 | – |
| Employment status | ||
| Retired | 5 | 4 |
| Disabled | 3 | 2 |
| Homemaker | 2 | 2 |
| Unemployed | 2 | 2 |
| Prefer not to say | 2 | 2 |
| Employed full time | – | 2 |
| Employed part time | – | 1 |
| Highest level of education attainment a | ||
| High school diploma | 3 | 4 |
| College degree | 3 | 2 |
| Some college | 3 | 2 |
| Associate degree | 2 | - |
| General educational development (GED) b | – | 1 |
| Graduate degree | 1 | – |
| Bachelor’s degree | – | 4 |
| Master’s degree | 1 | 1 |
| Postgraduate | – | 1 |
| Doctorate | 1 | – |
| Cancer type | ||
| Multiple myeloma | 4 c | |
| Lung | 3 | |
| Melanoma | 2 | |
| Kidney | 2 | |
| Colorectal | 1 | |
| Lymphoblastic leukemia | 1 c | |
| Cancer stage | ||
| 1 | 1 | |
| 2 | 1 | |
| 3 | 3 | |
| 3/4 | 2 d | |
| 4 | 7 | |
| Site of metastases c | ||
| Lymph nodes | 7 | |
| Bones | 4 | |
| Lungs | 1 | |
| IO therapy | ||
| Pembrolizumab e | 6 | |
| Nivolumab e | 4 | |
| Teclistamab e | 1 | |
| Idecabtagene vicleucel f | 1 | |
| Brexucabtagene autoleucel f | 1 | |
| CAR-T cell therapy (unspecified) + teclistamab f | 1 g |
Overall, 22 symptoms were reported by the 14 patients—10 by both CAR-T therapy patients and non-CAR-T therapy patients, 7 by just CAR-T therapy patients, and 5 by just non-CAR-T therapy patients (Fig. 1 ). Patients on non-CAR-T therapy reported a median of 3 symptoms (range, 1–14), whereas those on CAR-T therapy reported a median of 10 symptoms (range, 3–12). Eight symptoms were identified for the first time in patient CE interviews: itching, sweating, weakness, memory loss, difficulty speaking, difficulty writing, constipation, and tiredness. Fig. 1 Signs and symptoms of patients reporting IO-induced CRS. CRS Cytokine release syndrome
Participants also reported the most bothersome symptom on the basis of their CRS experience. A total range of eight different symptoms were reported. Out of ten participants who reported fatigue as a symptom, three reported it as the most bothersome. Of participants who reported body aches or pain as a symptom, two ( n = 2/8) reported it as the most bothersome symptom. Two out of seven participants reported fever as the most bothersome symptom. Out of six participants reporting tiredness, three reported it as the most bothersome symptom. Lightheadedness was reported as the most bothersome symptom by two out of six participants. Diarrhea was reported as the most bothersome symptom by one out of five participants. Nausea was reported as the most bothersome symptom by two out of three participants. Lastly, one out of one reported memory loss as the most bothersome symptom.
Patients also rated signs/symptoms on a 4-point Likert severity scale (mild, moderate, severe, and very severe) during interviews. A wide range of symptom severities were reported across patients receiving non-CAR-T cell therapies, with the highest severity ratings observed for fatigue, body aches or pain, tiredness, diarrhea, nausea, and itching. Among patients receiving CAR-T cell therapy, all signs/symptoms were rated as moderate or severe, with the exception of one mild rating each for fever and headache. No patients receiving CAR-T cell therapy rated any signs/symptoms as very severe.
Patients also reported various impacts of CRS that centered around activities of daily living, physical impacts, social and role functioning impacts, emotional impacts, and work/financial impacts. Table 2 provides the number of patients reporting each of these key symptoms and impacts, by therapy type, along with patient quotes. Table 2 Key symptoms and impacts of CRS reported by patients in the CE interviews ADL , activities of daily living, CAR-T chimeric antigen receptor T-cell, CE concept elicitation, CRS cytokine release syndrome, IO immuno-oncology a The patient reported the concept themselves without prompting b The patient reported the concept after being prompted Dark gray = reported only by patients on CAR-T therapy. Pale gray = reported only by patients on non-CAR-T therapy
Table 2
Key symptoms and impacts of CRS reported by patients in the CE interviews
ADL , activities of daily living, CAR-T chimeric antigen receptor T-cell, CE concept elicitation, CRS cytokine release syndrome, IO immuno-oncology a The patient reported the concept themselves without prompting b The patient reported the concept after being prompted Dark gray = reported only by patients on CAR-T therapy. Pale gray = reported only by patients on non-CAR-T therapy
CRS Conceptual Model Development
A CM was developed on the basis of the above research to provide a comprehensive understanding of CRS events that could be described and reported by patients in a PROM. In acknowledgement of the fact that CRS can present heterogeneously, the CM included all signs, symptoms, and impacts identified by patients and/or clinical experts in the CE research, as both are important to track to understand IO-induced CRS and patient-reportable signs or symptoms. Signs or symptoms identified only in the TLR but not endorsed through subsequent research were not included. The final CM included 44 signs and symptoms across five organ systems and 31 impacts across six domains (Fig. 2 ). This was the basis from which the new PROM was drafted. Fig. 2 Final conceptual model to identify IO-induced CRS and its impacts on patients’ lives. CRS Cytokine release syndrome, IO Immuno-oncology
Patient-Reported Outcome Measure Development
Through an iterative process, 12 patient-reportable symptoms and four impacts were ultimately selected from the final CM for inclusion in the PROM (Fig. 3 ). The PROM was designed to measure the incidence and severity of each symptom, and the impact of symptoms generally on regular daily activities, worry, and physical ability. Manageability of symptoms and overall perceived health were also prioritized for inclusion in the PROM. Fig. 3 Patient-reportable symptoms and impacts from the conceptual model selected to create the patient PROM. PROM patient-reported outcome measures, CD Cognitive debreifing
Six PROM development criteria were followed: (1) include measurement of a common set of symptoms representing the most commonly experienced CRS-related symptoms as reported by patients, clinical experts, and the investigator advisory board; (2) measure severity of these symptoms, rather than frequency; (3) capture information about symptoms prior to first treatment, on the day of each treatment, and on subsequent days after treatments; (4) include measurement of a limited set of impacts representing those that are most relevant to patients; (5) capture information about impacts prior to first treatment and weekly after each treatment; and (6) use terminology consistent with (a) language used by patients in CE interviews, and (b) language proposed for measurement of signs and symptoms by physicians in clinical expert interviews and the investigator advisory board.
The structure of the draft PROM is presented in Fig. 4 . Three versions of the PROM were developed to be administered at different time points (baseline, day of treatment, and subsequent days) (Fig. 5 ). The recall period depended on which version of the PROM was administered. An electronic (e)PROM format was used for all three versions. Fig. 4 Administration schedule for the IO-induced CRS patient diary. CRS Cytokine release syndrome, IO Immuno-oncology Fig. 5 Participant-reported onset of symptoms associated with CRS ( N = 14). CRS Cytokine release syndrome
In the first question, patients are presented with a single item showing all 12 symptoms. Patients are asked to indicate which (if any) they have experienced over the recall period. There is also a free-text box for patients to add “other” symptoms that they experienced. This symptom list is a new item drafted for this PROM. For each symptom that they have experienced, patients are then presented with an item about the severity of that symptom on a four-point severity response scale of “mild,” “moderate,” “severe,” and “very severe.” All items used this simple structure to measure the severity of each concept, utilizing the language patients used in the CE interviews. Additional items were developed de novo utilizing the language patients used in the CE interviews to capture the overall impact of symptoms on patients. A numeric rating scale (NRS), Likert scale, or dichotomous scale were used on the basis of the item wording. Not all impacts are measured in all three versions (see Fig. 5 ).
This PROM is not intended for use as a diagnostic tool but rather to supplement existing grading scales by tracking the onset and resolution of CRS and impacts on patients’ lives.
Refinement and Finalization of the Patient-Reported Outcome Measure
Cognitive Debriefing Interviews with Clinical Oncologists
Clinical oncologists proposed adding headache to the draft PROM, and to separate chills and shaking. They also proposed removing items “stay at home or rest” (measuring symptoms experienced since receiving the study medication that caused the patient to stay at home or rest more than normal), “regular daily activities” (measuring symptoms that had an impact on regular daily activities), and “worry because of symptoms impact” (measuring how much experienced symptoms caused worry), as well as clarifying recall periods to reduce confusion. The response options for impact on “regular daily activities” and “worry due to symptoms” were changed from a NRS to a 4-point Likert scale. Lastly, formatting was amended to remove capitalization and to boldface only symptom severity and recall period.
Cognitive Debriefing Interviews with Patients
CD interviews were conducted with 15 patients in three waves with five different patients each: wave 1 included patients P1–P5, wave 2 included patients P6–P10, and wave 3 included patients P11–P15. Nine of these patients had also participated in the CE interviews (five in wave 1, three in wave 2, and one in wave 3). Patients’ demographic characteristics are provided in Table 1 . The 15 total patients who received non-CAR-T therapy were recruited from among those who participated in the CE interviews, those who were approached for CE but did not qualify, and additional patients from a patient advocacy group.
Three versions of the IO-induced CRS patient diary were developed, to be administered at different time points in the treatment journey. This ensures that the patient experience throughout their symptom journey is holistically captured. Overall, the patients found the PROM easy to understand and use, and they approved of the schedule of assessment. “Yeah, what you’ve developed is nice. There’s no problems with it.”—P1 “I think it’s pretty simple. It’s pretty easy and quick. It doesn’t have anything dramatic—wording or anything—that seems overwhelming. It seems quite simplified… I think most people should be able to answer it without a concern.”—P1 “… it [schedule of assessment] totally makes sense, because there’s the prior, there’s the completed once on day of each infusion, completed daily for a set number of days... it makes sense.”—P5
PROM Baseline Version
The overall instructions for this version were interpreted as intended by all patients across the three waves ( n = 15/15). No modifications were recommended or made to the instructions. All patients ( n = 15/15) interpreted the symptom list item as intended, and 14 patients ( n = 14/15) reported that the symptom list was clear; one patient reported being unclear about whether “symptoms” in the question were referring to symptoms from the IO therapy or symptoms from their cancer. In total, 13 patients ( n = 13/15) interpreted the recall as intended, whereas 1 ( n = 1/15) expressed some difficulty thinking back over the past week, and 1 patient ( n = 1/15) considered the recall period to be more than 7 days. Since three patients reported entering symptoms unrelated to IO therapy in the free-text box, the “other” response option and the corresponding screen with the text box were removed from all versions of the PROM to avoid complications with data analysis to track CRS onset and offset.
All patients providing feedback ( n = 14/14), interpreted the symptom severity instructions as intended, and considered the instructions clear. All patients ( n = 15/15) interpreted the severity scale for each symptom as intended and were able to report their experience. Of 12 patients who were probed on their preference between a frequency scale and a severity scale for gastrointestinal symptoms (diarrhea, vomiting, and nausea), 5 preferred a frequency scale, 3 preferred the severity scale, and 4 suggested retaining the severity scale but adding an item for gastrointestinal symptoms with a frequency scale. Since this would prolong PROM completion, the severity scale was retained for all symptoms. To improve the clarity and consistency of the introductory instruction, “symptom(s)” was revised to “symptoms.” This change was implemented across all three versions.
The overall perceived health impact item was debriefed for this version of the PROM. In wave 1, two patients ( n = 2/5) interpreted this item as intended (linking overall health to the symptoms they reported). Changes were made to the wording of this item for clarity. In waves 2 and 3, nine patients ( n = 9/10) interpreted the question in the item as intended, and one patient primarily considered their functionality in relation to their symptoms. Seven patients ( n = 7/9) interpreted the response options as intended; two patients ( n = 2/9) had some difficulty with the “neither poor nor good” response options. Eight patients ( n = 8/9) considered the item to be clear, and one patient ( n = 1/9) considered it to be unclear because they did not perceive the response options as relating to health. No further changes to this item were deemed necessary after wave 3.
PROM Day-of-Treatment Version
The instructions for the day-of-treatment version of the PROM were the same as for the baseline version, except for the recall period, which was “since receiving the study medication today” (instead of “in the past week”). Overall,. 13 patients ( n = 13/14) interpreted the overall instructions for the day-of-treatment version and the associated recall period as intended. No changes to the instructions were considered necessary.
Patients were not probed on symptom items, since this section in the day-of-treatment version was similar to the baseline version. Impact items were tested in the baseline version (overall health) and the subsequent-days version (other impacts).
PROM Subsequent-Days Version
The instructions for the subsequent-days version were the same as for the baseline except for the recall period, which was “the past 24 h.” All patients providing feedback ( n = 14/14), interpreted the overall instructions for the subsequent-days version as intended, and considered the instructions to be clear. After wave 1, “in the past 24 h” was rephrased to “over the past 24 h.” All patients providing feedback ( n = 13/13) interpreted the recall as intended.
All impacts except “overall perceived health” were debriefed for this version. The “stay at home or rest” impact item was revised twice to improve clarity and reporting during CD. After wave 1, the response scale was changed from “yes/no” to a Likert scale, which was preferred by eight patients (three preferred a dichotomous scale), and to “…over the past 24 h” from “…in the past 24 h” on the basis of discussions within the study team. After wave 3, the item was revised only to ask about “rest” rather than “stay at home and rest.” Since this change was made after wave 3, the final item has not undergone CD.
For the “regular daily activities” impact item, all patients interpreted the question ( n = 15/15) and recall ( n = 11/11) as intended, and considered it to be clear ( n = 15/15). In total, 13 patients ( n = 13/15) interpreted the response options as intended, 1 patient ( n = 1/15) reported difficulty with quantifying the response options, and 1 patient ( n = 1/15) reported difficulty with differentiating between “quite a bit” and “very much.”
For the “manageability of symptoms” impact item, 13 patients ( n = 13/15) interpreted the question in the item as intended, 1 patient ( n = 1/15) interpreted the question slightly differently as the ability to withstand symptoms, and 1 patient ( n = 1/15) interpreted it to mean one’s physical ability while experiencing symptoms. Overall, 12 patients ( n = 12/14) interpreted the response options as intended, 1 patient ( n = 1/14) reported difficulty with defining “neither difficult nor easy” and recommended removing the response option, and 1 patient ( n = 1/14) reported difficulty differentiating between “easy” and “very easy” or “difficult” or “very difficult.” In total, 14 patients ( n = 14/15) interpreted the item as clear, although 1 patient suggested replacing “manage” with “handle symptoms” on the basis of their interpretation.
Fourteen patients ( n = 14/15) interpreted the question in the “worry” impact item as intended, and one patient selected a response option on the basis of their experience with their cancer. Twelve patients ( n = 12/14) interpreted the response options as intended, and two patients ( n = 2/14) interpreted “quite a bit” and “very much” as similar response options. Fourteen patients ( n = 14/15) considered the item to be clear, and one patient ( n = 1/15) suggested that the question could be shortened.
Apart from a revision after wave 1 to improve the consistency of phrasing across the PROM, no other changes were considered necessary to the “regular daily activities,” “manageability of symptoms,” and “worry” impact items.
Overall
Overall, patients described the PROM as easy to understand and complete. When asked whether they would be able to complete the PROM daily after receipt of an IO treatment, 11 patients ( n = 11/14) indicated that they would, while 3 patients ( n = 3/14) indicated that they would not be able to complete a daily diary immediately after IO treatment owing to the way they may feel.
Six patients ( n = 6/14) reported that they would be able to complete the PROM even while experiencing severe symptoms, five patients ( n = 5/14) reported that they would not be able to complete it, and three patients ( n = 3/14) reported that they would be able to complete it but would require assistance. Patients most frequently reported vomiting, diarrhea, headaches, and pounding or racing heartbeat as symptoms (two mentions each) which, if they became severe, would hamper PROM completion.
Discussion
In the present study, a novel PROM—the IO-induced CRS patient diary—was developed to track the onset, resolution, and impact of IO therapy-induced CRS events in clinical trials. This study ties into FDA initiatives to better understand the tolerability of IO therapies [ 16 ].
The development of this PROM followed the FDA’s best-practice recommendations for qualitative research methods [ 15 ], as the methodology included an interview plan tailored to the intended population, study characteristics, and design; study materials such as a study protocol, interview guide, and training materials; and a detailed qualitative analysis plan. In line with the relevant FDA guidance, the content validity of this PROM was established through CD interviews with patients [ 15 ] and its clinical relevance was ensured through CD interviews with clinical oncologists.
One of the key strengths of this study was that the patient population interviewed showed considerable diversity in various aspects. Patients had a mix of cancer diagnoses (including solid and hematologic cancers) of different stages, the sites of metastases were varied, and they were receiving various types of cancer treatments. CRS presentation and symptom severity were also remarkably different among the patients in the present study, and patients receiving CAR-T and non-CAR-T therapy were included. Thus, our findings generate a comprehensive picture of what it is like to experience a CRS event.
The IO-induced CRS patient diary therefore demonstrates potential for use alongside existing clinician-reported tools, offering a more detailed and time-sensitive account of CRS experiences. By capturing symptom onset, progression, and resolution in real time, the diary enhances the granularity of adverse event reporting. Together, the clinician-reported tools and the diary converge to comprehensively and accurately identify and monitor CRS, contributing to a more patient-centered understanding of IO therapy tolerability.
This study has certain limitations. A few concepts included in the CM may correspond to CRS-like adverse events which cannot be distinguished from clinically confirmed CRS events. Although the patient population was diverse in terms of race and ethnicity, the research was US-based, so any further cultural factors may not have been captured. Lastly, minor modifications were made to the PROM after the CD interviews, and these were not debriefed.
Conclusions
The IO-induced CRS patient diary is a first-of-its-kind, well-developed, content-valid PROM for tracking the onset, resolution, and impact of IO-induced CRS on patients’ lives. It provides a comprehensive picture of the patient experience of IO therapy-induced CRS. Although this PROM was developed for use in clinical trials, it may also help to understand CRS events in clinical practice by complementing the clinical description with the patient perspective. Future studies should test the validity of this PROM in a broader and larger sample, and confirm what constitutes onset and resolution in the PROM. Its psychometric performance in clinical trials should also be tested. Finally, additional CD interviews are required to confirm the interpretability and relevance of the PROM, especially among patients on CAR-T therapy.
Supplementary Information
Below is the link to the electronic supplementary material. Supplementary file1 (PDF 211 KB)
Medical Writing/Editorial Assistance
Support for medical writing, editing, and graphic design was provided by Rucha Kurtkoti (MSc), Harkiran Gill (PharmD), Chrysi Petraki (PhD), Nathaniel Grubbs (PhD), and Alexandra Guanique (Designer) from IQVIA, funded by Sanofi.
Author Contributions
Edward (Ted) Wells, Laurence Lucats, Giovanni Abbadessa, James Turnbull, Joyce R. Talavera, Sophie R. Van Tomme, Paul Cordero, Benoit Arnould, Matthew Reaney, and Catherine Coulouvrat all contributed to the conception, design, execution, and/or interpretation of the study. All authors reviewed and approved the final version of the manuscript and meet the International Committee of Medical Journal Editors (ICMJE) criteria for authorship.
Funding
Sponsorship for this study and Rapid Service Fee were funded by Sanofi.
Data Availability
All data generated or analyzed during this study are included in this published article/as supplementary information files.
Declarations
Conflict of Interest
Edward Wells, James Turnbull, and Matthew Reaney are employees of IQVIA, which received funding from Sanofi to conduct this study. Laurence Lucats, Joyce R. Talavera, Sophie R. Van Tomme, Benoit Arnould, Catherine Coulouvrat, and Paul Cordero are employees of Sanofi and may hold stocks or stock options in the company. Giovanni Abbadessa was an employee of Sanofi when this study was conducted and may hold stocks or stock options in the company.
Ethical Approval
This study was conducted in accordance with the ethical principles that have their origin in the Declaration of Helsinki 1964 and its later amendments and is consistent with Good Clinical Practice and applicable regulatory requirements. Centralized approval was obtained from the Western IRB-Copernicus Group (WCG) IRB, including for all recruitment materials. IQVIA is responsible for submitting the application and all study documents for study approval to the IRB. The study was initiated only after IRB approval was granted, including the use of all recruitment materials.
Informed Consent
Informed consent was obtained prior to performing any study-related procedures, including screening, and patients provided consent via the online consent form. Patients were informed that, if the results of the study were published for scientific purposes, the patient’s identity would remain confidential. Clinicians who participated in the CD interviews provided verbal consent at the start of the interview after the objectives and logistics of the interview were shared with them. They provided verbal consent to both participate and to also be audio-recorded.
A standard ethics approved informed consent process was followed for all patients involved in this study. This required their review and signature of an informed consent form document that clearly outlined the study background, objectives, and procedures (including screening, risks and benefits, compensation, confidentiality, information use and data sharing, and voluntary participation and withdrawal). Completed and signed forms were reviewed by the study team for completeness prior to any interviews being conducted. For participating clinicians, informed consent was obtained verbally at the start of each interview. Prior to providing consent, clinicians were informed of the study background and objectives, interview logistics, and information regarding their confidentiality and how their data will be used.
Thanking Patients
The authors would like to thank SparkCures and Rare Patient Voice for their patient advocacy and support during interviews and the patients for their contributions to the study.
References
1. Papaioannou NE Beniata OV Vitsos P Tsitsilonis O Samara P Harnessing the immune system to improve cancer therapy Ann Transl Med 2016 4 14 261 10.21037/atm.2016.04.01 27563648 PMC4971375 Papaioannou NE, Beniata OV, Vitsos P, Tsitsilonis O, Samara P. Harnessing the immune system to improve cancer therapy. Ann Transl Med. 2016;4(14):261. 27563648 10.21037/atm.2016.04.01 PMC4971375
2. Kamta J Chaar M Ande A Altomare DA Ait-Oudhia S Advancing cancer therapy with present and emerging immuno-oncology approaches Front Oncol 2017 7 64 10.3389/fonc.2017.00064 28459041 PMC5394116 Kamta J, Chaar M, Ande A, Altomare DA, Ait-Oudhia S. Advancing cancer therapy with present and emerging immuno-oncology approaches. Front Oncol. 2017;7:64. 28459041 10.3389/fonc.2017.00064 PMC5394116
3. Institute CR. Regulatory approval timeline of active immunotherapies. 2025 [cited; Available from: https://www.cancerresearch.org/regulatory-approval-timeline-of-active-immunotherapies
4. Dobosz P Dzieciatkowski T The intriguing history of cancer immunotherapy Front Immunol 2019 10 2965 10.3389/fimmu.2019.02965 31921205 PMC6928196 Dobosz P, Dzieciatkowski T. The intriguing history of cancer immunotherapy. Front Immunol. 2019;10:2965. 31921205 10.3389/fimmu.2019.02965 PMC6928196
5. Kaufman HL Atkins MB Subedi P The promise of immuno-oncology: implications for defining the value of cancer treatment J Immunother Cancer 2019 7 129 10.1186/s40425-019-0594-0 31101066 PMC6525438 Kaufman HL, Atkins MB, Subedi P, et al. The promise of immuno-oncology: implications for defining the value of cancer treatment. J Immunother Cancer. 2019;7:129. 31101066 10.1186/s40425-019-0594-0 PMC6525438
6. Schneider BJ Naidoo J Santomasso BD Lacchetti C Adkins S Anadkat M Management of immune-related adverse events in patients treated with immune checkpoint inhibitor therapy: ASCO guideline update J Clin Oncol 2021 39 36 4073 4126 10.1200/JCO.21.01440 34724392 Schneider BJ, Naidoo J, Santomasso BD, Lacchetti C, Adkins S, Anadkat M, et al. Management of immune-related adverse events in patients treated with immune checkpoint inhibitor therapy: ASCO guideline update. J Clin Oncol. 2021;39(36):4073–126. 34724392 10.1200/JCO.21.01440
7. Lee DW Gardner R Porter DL Louis CU Ahmed N Jensen M Current concepts in the diagnosis and management of cytokine release syndrome Blood 2014 124 2 188 195 10.1182/blood-2014-05-552729 24876563 PMC4093680 Lee DW, Gardner R, Porter DL, Louis CU, Ahmed N, Jensen M, et al. Current concepts in the diagnosis and management of cytokine release syndrome. Blood. 2014;124(2):188–95. 24876563 10.1182/blood-2014-05-552729 PMC4093680
8. Cobb DA Lee DW Cytokine release syndrome biology and management Cancer J. 2021 27 2 119 25 10.1097/PPO.0000000000000515 33750071 Cobb DA, Lee DW. Cytokine release syndrome biology and management. Cancer J. 2021;27(2):119–25. 33750071 10.1097/PPO.0000000000000515
9. Cosenza M Sacchi S Pozzi S Cytokine release syndrome associated with t-cell-based therapies for hematological malignancies: pathophysiology, clinical presentation, and treatment Int J Mol Sci. 2021 22 14 7652 10.3390/ijms22147652 34299273 PMC8305850 Cosenza M, Sacchi S, Pozzi S. Cytokine release syndrome associated with t-cell-based therapies for hematological malignancies: pathophysiology, clinical presentation, and treatment. Int J Mol Sci. 2021;22(14):7652. 34299273 10.3390/ijms22147652 PMC8305850
10. Shimabukuro-Vornhagen A Godel P Subklewe M Stemmler HJ Schlosser HA Schlaak M Cytokine release syndrome J Immunother Cancer 2018 6 1 56 10.1186/s40425-018-0343-9 29907163 PMC6003181 Shimabukuro-Vornhagen A, Godel P, Subklewe M, Stemmler HJ, Schlosser HA, Schlaak M, et al. Cytokine release syndrome. J Immunother Cancer. 2018;6(1):56. 29907163 10.1186/s40425-018-0343-9 PMC6003181
11. Tay SH Toh MMX Thian YL Vellayappan BA Fairhurst AM Chan YH Cytokine release syndrome in cancer patients receiving immune checkpoint inhibitors: a case series of 25 patients and review of the literature Front Immunol 2022 13 807050 10.3389/fimmu.2022.807050 35154124 PMC8831742 Tay SH, Toh MMX, Thian YL, Vellayappan BA, Fairhurst AM, Chan YH, et al. Cytokine release syndrome in cancer patients receiving immune checkpoint inhibitors: a case series of 25 patients and review of the literature. Front Immunol. 2022;13:807050. 35154124 10.3389/fimmu.2022.807050 PMC8831742
12. Haslam A Herrera-Perez D Gill J Prasad V Patient experience captured by quality-of-life measurement in oncology clinical trials JAMA Netw Open 2020 3 3 e200363 10.1001/jamanetworkopen.2020.0363 32129865 PMC7057133 Haslam A, Herrera-Perez D, Gill J, Prasad V. Patient experience captured by quality-of-life measurement in oncology clinical trials. JAMA Netw Open. 2020;3(3):e200363. 32129865 10.1001/jamanetworkopen.2020.0363 PMC7057133
13. Watts T Roche D Csontos J Patients’ experiences of cancer immunotherapy with immune checkpoint inhibitors: a systematic review and thematic synthesis J Clin Nurs 2024 33 8 2885 2904 10.1111/jocn.17154 38589992 Watts T, Roche D, Csontos J. Patients’ experiences of cancer immunotherapy with immune checkpoint inhibitors: a systematic review and thematic synthesis. J Clin Nurs. 2024;33(8):2885–904. 38589992 10.1111/jocn.17154
14. Administration USFaD. Patient-focused drug development: selecting, developing, or modifying fit-for-purpose clinical outcome assessments guidance for industry, Food and Drug Administration staff, and other stakeholders. US FDA; 2022.
15. Administration USFaD. Patient-focused drug development: methods to identify what is important to patients. 2022.
16. Research FoC. Broadening the definition of tolerability in cancer clinical trials to better measure the patient experience. 2018 [cited]; Available from: https://friendsofcancerresearch.org/wp-content/uploads/Comparative-Tolerability-Whitepaper_FINAL.pdf#:~:text=Tolerability%20is%20a%20complex%20concept,are%20most%20burdensome%20to%20patients .