Patient-reported cognitive function among hematopoietic stem cell transplant and cellular therapy patients: A scoping review

Introduction

Cognitive dysfunction, such as memory impairment, difficulty concentrating, and impaired ability to multitask or make decisions, has been recognized as a common complication for patients undergoing treatments for cancer.¹ Cognitive dysfunction is a particular concern after cellular therapies (CT) such as hematopoietic cell transplantation (HCT) and chimeric antigen receptor (CAR) T-cell therapy. CAR T-cell therapy carries a risk of cytokine release syndrome (CRS) and neurologic toxicity, also referred to as CAR-related encephalopathy syndrome (CRES) or immune effector cell-associated neurotoxicity syndrome (ICANS).^2,3 In the clinical setting, a primary approach to assessing neurocognitive toxicity are clinician-reported assessments, and in CAR-T these are often in the form of consensus grading approaches (ASTCT Consensus grading of CRS)⁴ with associated management guidelines. An expert review of neurocognitive dysfunction among HCT patients reported several risk factors associated with neurocognitive impairment including conditioning regimen (i.e. fludarabine), graft-versus-host-disease and associated immunosuppressive therapies, infections, primary disease, and other risk factors (i.e. female gender, higher body mass index, absence of social partner).⁵

In research settings, two other types of assessments are often used: performance-based testing and patient-reported outcome (PRO) measures. Performance-based measures of cognitive function provide a test- or task-based assessment of processing speed/attention, memory, verbal fluency, fine motor speed, executive functioning, and educational achievement.⁵ A 2018 expert review of over 140 studies that used performance-based cognitive dysfunction measures among adult and pediatric HCT survivors showed an incidence of neurocognitive dysfunction in 60% of patients 2–6 years post-HCT.⁵ A similar study has not been conducted in CAR-T recipients. PROs are patients’ reports of their own symptoms and functioning without interpretation by anyone else.⁶ For cognitive function, questions typically ask patients to rate their difficulty remembering things, difficulty concentrating, or difficulty thinking or speaking clearly. A recent review of PROs among patients receiving CAR-T therapy found only 2 published studies of adult patients and suggested that 40% of adults report some cognitive difficulties after CAR-T.^7–9 Each of these studies used a different measurement approach.

Performance-based measures are time-consuming for patients and research staff to complete, making them a more resource intensive option. Relative to performance-based measures, PROs are quicker and less costly. Yet, it is unlikely that the two are interchangeable. A systematic review of 24 studies that directly compared performance-based neuropsychological test scores and patient-reported measures of cognitive dysfunction found that in studies of people who had been treated with chemotherapy for cancer, 60% did not find an association between the two types of measures.¹⁰ Studies that did demonstrate a significant correlation (range r= .33-.66) were mostly in breast cancer patients.¹⁰ Moreover, multiple studies have demonstrated that patient-reported cognitive concerns are more strongly with depression and anxiety,^11,12 perhaps reflecting negative affect more generally.¹³

There are many PRO measures of cognitive dysfunction available, but much of the work on measurement of cognitive function has been conducted outside the transplant setting. Within the transplant setting, PROs are beginning to be collected, within two cellular therapy registries (Center for International Blood and Marrow Transplant Research [CIBMTR] and European Society for Blood and marrow Transplantation [EBMT]) utilizing the Patient-Reported Outcomes Measurement Information System (PROMIS). PROMIS is the NIH’s initiative to standardize measurement of common PROs in clinical research across chronic conditions¹⁴ including oncology.¹⁵ Both registries are collecting cognitive function using PROMIS. With increasing utilization of PROs in the cellular therapy setting, it is critical to identify which PRO measures are being assessed and when. Though a recent review assessed five measures of cognitive impairment among cancer populations and recommended PROMIS measures¹⁶, there has yet to be a comprehensive review of self-reported cognitive function measures among HCT patients. Therefore, our aims in this systematic scoping review were to understand 1) which PRO measures of cognitive function have been used in CT settings, 2) the effects of CT therapies on patient-reported cognitive function, and 3) pre and post treatment factors associated with cognitive function.

Methods

Our scoping review was guided by the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) Extension for Scoping Reviews.¹⁷

Search strategy:

The library scientist on the team (E.S) designed an extensive search of the literature, which was conducted on February 4^th, 2020. Medline (Ovid), PsycINFO, Web of Science, Scopus, and Cochrane Database of Systematic Reviews, and Cochrane Central Register of Controlled Trials were queried using a combination of natural language terms and controlled vocabulary for immunotherapy, stem cell transplant, patient reported outcome, and cognitive function. Articles were also identified by review of relevant review bibliographies and trial protocols for publications. Results were limited to English-only publications. The complete search strategy that was replicated in other databases is available in Supplementary Table 1.

Study Selection

Database search results were screened using the Rayyan platform.¹⁸ 2801 records were returned from the database search and 33 were identified through other sources. After removing duplicates,1952 individual studies remained, and were each screened by title and abstract by two members of the research team (R.C. and J.B.). Discrepancies among reviewers were resolved through regular meeting and discussion. After the initial screening, 278 studies remained, which were divided between the two reviewers and underwent full-text review. Throughout the full-text review process, any study that that did not clearly meet inclusion criteria was discussed by both reviewers. Furthermore, 25% of full texts were double screened intermittently throughout the review process to ensure consistency in reviewers’ decisions (R.C. and J.B.). Any disagreement between the two reviewers was resolved through regular discussions and consultation of additional members of the research team as needed. After the full text screening process was complete, 56 studies remained and are included in this review. The study selection process is available in the PRISMA flow diagram in Figure 1.

Figure 1. PRISMA Flow Diagram.

Inclusion Criteria

Eligibility criteria were specific to the (1) patient population: adult, received/receiving hematopoietic cell transplantation or received/receiving chimeric antigen receptor t-cell therapy, (2) the type of survey measure used: self-reported survey, patient-reported outcomes, (3) construct specific: self-reported cognitive function and abilities, including cognitive dysfunction/cognitive complaints, and (4) cognitive function results were reported on their own, not part of a summary score. For the initial screening step, we included original studies as well as cases that were published in abstract format or clinical trial protocol format. In subsequent steps, the full texts of potentially eligible studies were retrieved and assessed for further eligibility. Studies were removed if only published in abstract format.

Outcomes:

In this scoping review, we identified studies with full published results that include results uniquely reporting on cognitive function. Therefore, studies that used measures that include survey items about cognitive function but did not report the cognitive function results separately were excluded. Studies most often removed for this reason used the EORTC-QLQC30 (n=4) where an overall quality of life score was reported but not a cognitive-function specific score.

Results were reported according to the three goals of this study. First, we catalog the different cognitive function PRO measures used and the frequency with which they were used in the HCT and CT settings. We also report which studies used performance based cognitive function measures in conjunction with patient-reported measures. Second, we report whether outcomes related to receiving HCT/CT were statistically significantly related to changes in patient-reported cognitive function. An association was determined to be significant if the results were statistically significant at p < .05. Results were divided into two categories: (1) The differences in cognitive function scores between two groups, with one group receiving a CT compared to another group(s) (e.g.., health controls, another treatment, the control arm of a randomized trial). (2) The change in cognitive function over time for those receiving a CT. Third, we assess which factors, both pre-HCT/CT and post-HCT/CT, were associated with patient-reported cognitive dysfunction. These are defined as any symptom, function, or clinical outcome that was statistically significantly associated with patient-reported cognitive function at p < .05. Given a previous evaluation of cognitive function measures among cancer patients¹⁶, we did not evaluate the PRO measures used.

Data extraction:

Two research team members (R.C. and J.B) used a shared data extraction form to collect data for each eligible study. Extracted variables included: study characteristics specific to how cognitive function was assessed (which patient-reported cognitive function measure, performance-based cognitive function measure, timepoint(s) assessed), cognitive function results. Three members (J.B, V.K, and Z.U) used a separate data extraction form to collect study population characteristics (age, gender, race, employment, income, treatment and disease characteristics, country) for eligible studies. Data are summarized using frequencies and percentages and results are presented in both narrative and tabulated form.

Results

Study characteristics

Of the 278 full text records reviewed, 56 met the criteria of the study. Reasons for the 213 excluded articles were primarily wrong article type (e.g., conference abstracts, trial protocols; n = 72), wrong population (e.g. not HCT or CT, pediatric; n = 65), did not use PROs (n = 26), or did not use or uniquely report on cognitive function (n = 59), see Figure 1 for flow diagram. Table 1 shows the majority looked at both autoHCT and alloHCT (n=37; 66%), and only one study⁹ looked at CAR-T recipients. The studies included were conducted in 15 countries, with the most in the United States (n=27, 48%). Study designs included observational (n = 48; 86%) and randomized controlled trials (n = 8; 14%). The majority of studies were longitudinal (31; 55%) and follow-up post-HCT ranged up to 5 years¹⁹ and as many as 12 timepoints collected.²⁰ For cross-sectional studies, post-HCT follow up ranged up to 20 years.^21,22 Samples sizes of the studies ranged from 16²³ to 670²⁴ (though this study only had 30 HCT recipients). Fifteen of the studies assessed performance-based measures of cognitive function in addition to PROs. The largest fully HCT sample was 415.²⁵ Studies included a mix of proportions of women and men, with most being within 10 percentage points of 50/50. Average/median ages ranged from 35^26,27 to 64.²⁸

Table 1.

Study Characteristics for included articles using self-reported cognitive function measures among adult hematopoietic cell transplant and cellular therapy patients

Study	Country	Study Design	Time of survey administration	Sample Size (n)	Sex	Mean/Median Age	Diseases Included in Study	Treatment (HCT = both)	Number of time points	Self-reported cognitive function measure
Acaster, S. 2013	United Kingdom	Observational	N/A	370	M=52% F=48%	64.6 (mean)	MM	AlloHCT	1	EORTC-QLQ-C30
Ahmedzai, S. 2019	United Kingdom	Observational	Before registration and after completion of reinduction: (T1) 100 days (T2) 6 months (T3) 12 months (T4) Annually.	288	M= 70% F=30%	61 (median)	MM	HCT	Less than or equal to 5	EORTC-QLQ-C30
Alaloul, F. 2015	Jordan	Observational	≥ 3 months after HSCT	126 (HCT=63)	M=68%F=32%	35.4 (mean)	Leukemia, Lymphoma, MM	HCT	1	EORTC-QLQ-C30
Andresen, S. 2011	Germany	Observational	Median follow-up for the HDCT + ASCT group = 8.5 years, R-CHOP group = 4.5 years	124 (HCT=63)	M=54%F=46%	60.4 (median)	Follicular lymphoma	AutoHCT	1	EORTC-QLQ-C30
Andrykowski, M. A.; Altmaier, E. M. 1990	United States	Observational	≥ 1 year post BMT	30	M=50% F=50%	34.3 (mean)	Leukemias	AlloHCT	3	SIP; POMS
Andrykowski, M. A.; Henslee, P. J. 1989	United States	Observational	T1: mean of 28 months post-transplant, T2: mean of 52 months post-BMT	16 (at final assessment)	M=44%F=56%)	35.6 (mean)	Any cancer treatable with BMT	AlloHCT	3	SIP
Baker, F 2003	United States	Observational	Post-transplant; mean months 35.6	99	M=52%F=49%		Several hematologic diseases	HCT	1	CPILS
Bartley, E. 2014	United States	Observational	1 week prior to transplant, at least one post-transplant: 3 months and/or 6 months post-transplant	70	M=56% F=44%	57.3 (mean)	Several hematologic diseases	HCT	2–3	PROMIS
Basinski, J. 2010	United States	Observational	Prior to HCT, 6 months post, 12 months post	52	M=56% F=44%	22–35 (17), 36–45 (16), 46–55 (12), 56–62 (7)	Several hematologic diseases	HCT	3	NBRS
Bishop, M. 2007	40 North American transplantation centers	Observational	Mean of 6.7 years post-transplant, SD 3.1, Range 1.9 to 19.4.	487 (HCT=177)	M=50% F=50%	50 (mean)	Any cancers treatable with BMT	HCT	1	Sickness Impact Profile (SIP)
Bonifazi, F. 2019	Italy, Spain, Germany, Israel	Randomized control trial	T0: Baseline admission to hospital, T1: 3–6 months post-HCT, T2: 12–24 months post-HCT	155	NR	NR	AML	AlloHCT	3	EORTC-QLQ-C30
Booth-Jones, M. 2005	United States	Observational	At least 6 months post BMT	65	M=22%F=78%	47 (mean)	Several hematologic diseases	HCT	1	MAQ
Braamse, A.M. 2014	The Netherlands	Observational	Up to 5.5 years post HCT	248	M=41% F=59%	56.6 (mean)	Several hematologic diseases	HCT	3	The problem list
Bush, N. 2000	United States or Canada	Observational	Annually, at 1,2,3 and 4 years post HCT	415	M=48% F=52%	41.8 (mean) +	Several hematologic diseases	HCT	4	EORTC-QLQ-C30
Caocci, G. 2006	Italy	Observational	at least 300 days post-HCT; median 43 months post HCT	19	M=58%F=42%	23 (median)	Thalassemia	AlloHCT	1	EORTC-QLQ-C30
Chang, G. 2009	United States	Observational	Baseline, 12 months	106 (HCT=45)	M=47% F=53%	40.6 (mean)	CML or Primary MDS	HCT	2	POMS
Clavert, A. 2017	France	Observational	At least 2 years post-HCT	110	M=58% F=42%	55 (median)	Myeloid and Lymphoid	AlloHCT	1	EORTC-QLQ-C30
Cleeland, S. 2000	United States	Observational	While inpatient for HCT	670 (HCT = 30)	M=50% F=50%	47 (mean)	Several hematologic diseases	HCT	1	MDASI
Correa, D. 2019	United States	Observational	Baseline, Post-induction chemo; Y1, Y2, Y3, Y4, Y5	29 (HCT=15)	M=40% F=60%	53 (median)	Cancers treatable with HCT	autoHCT	6	FACT-BR
Crooks, M. 2013	United States	Observational	Baseline pre transplant (allo-HCT); T2 = at discharge; T3 = 3 months post-tx; T4 = 6 month	37	M=62%F=38%	53.7 (mean)	ALL, AML, NHL, Other	AlloHCT	4	Patient Problem List
Danaher, E. 2006	United States	Observational	5 days before HCT and 4–8 days after HCT	20	M=45% F=55%	48.65 (mean)	Several hematologic diseases	HCT	2	EORTC-QLQ-C30
Dean, H. 2012	United States	Observational	Median time post tx = 10.5 (range 2.2–20 years)	51 (HCT=37, Controls=14)	M=59% F=41%	52 (median)	Several hematologic diseases	AutoHCT	1	EORTC-QLQ-C30
Deeg, H. 1998	United States	Observational	2, 5-, 10-, 15-, and 20-years post-transplant	212	M=55% F=45%	32 (median)	Severe aplastic anemia	HCT	5	Unspecified questionnaire
El-Banna, M. 2004	Jordan	Observational	Baseline; Day −2; Day 2; Day 7; Day 14	27	M=56% F=44%	49 (mean)	Several hematologic diseases	AutoHCT	5	PFS Piper Fatigue Scale
Epstein, J. 2002	Canada	Observational	Day 90–100	50	M=50% F=50%	39.5 (mean)	Several hematologic diseases	AlloHCT	1	EORTC-QLQ-C30
Fann, J. 2002	United States	Observational	Prospectively from pre-transplant to Day 30; 3 assessments a week up to 30 days	90	M=60% F=40%	Among those with no episode of delirium: 39.1 (mean) Among those with episode of delirium: 43.4 (mean)	Several hematologic diseases	HCT	12	Profile of Mood States
Ghazikhanian, S. E. 2017	United States	Observational	pre-transplant and post-transplant (3–12 months post)	138	M=60% F=40%	60.4 (mean)	Several hematologic diseases	HCT	2	PROMIS Applied Cognitive – General Concerns Scales
Hacker, E. D. 2011	United States	Randomized control trial; strength training intervention	Prior to hospitalization for HCT, during hospitalization D4-D8, Six weeks post discharge	19	M=74% F=26%	46 (mean)	Cancers treatable with HCT	HCT	3	EORTC-QLQ-C30
Harder, H. Cornelissen, J. 2002	The Netherlands	Observational	avg= 2.7 (4) months from treatment	40	M=60% F=40%	40.8 (mean)	Several hematologic diseases	HCT	1	EORTC-QLQ-C30
Harder, H. Van Gool, A. 2005	The Netherlands	Observational		183 (HCT=101, reference group=82)	M=61% F=39%	42 (mean)	Several hematologic diseases	HCT	1	EORTC-QLQ-C30; Cognitive Failure Questionnaire (CFS)
Harder, H. Van Gool, A. 2007	The Netherlands	Observational	22– 82 months after HCT, Baseline before undergoing HSCT (time1 [T1]), and at intervals of 8 months (time 2 [T2]) and 20 months (time 3 [T3]) post-HCT	183 (HCT=101, reference group=82)	M=61% F=39%	42 (mean)	Several hematologic diseases	HCT	3	EORTC-QLQ-C30, Cognitive Failure Questionnaire (CFS)
Hayden, P. J. 2004	Ireland	Observational	Median follow-up of 98 months (range 34–217 months).	51 (were alive in 2003 to receive questionnaire)	M=61% F=39%	35 (median)	Leukemias	AlloHCT	1	EORTC-QLQ-C30
Hendriks, M. 2002	The Netherlands	Observational	About 2.5 and 4.5 years post-HCT	52 (patients)	M=42% F=58%	41 (mean)	Malignant lymphoma, Breast CA, Acute leukemia	HCT	2	EORTC-QLQ-C30
Hjermstad, M. Evensen 1999	Norway	Observational	T1: pre-treatment T2: 1 year after transplant	177 (HCT=92(41=SCT, 51=ASCT))	AlloHCT group: M=56% F=44%; AutoHCT group: M=71% F=29%	SCT group=36 (median), HCT group 41 (median)	Several hematologic diseases	HCT	2	EORTC-QLQ-C30
Hjermstad, M., Holte, H. 1999	Norway	Observational	1 year after transplant	177 (HCT=92(41=SCT, 51=ASCT))	female AlloHCT group: M=56% F=44%; AutoHCT group: M=71% F=29%	SCT group=36 (median), HCT group 41 (median)	Several hematologic diseases	HCT	1	EORTC-QLQ-C30
Hong, F. 2013	United States	Observational	T1: Pretreatment T2: During treatment (at first post-hospital discharge clinic visit)	627 (HCT=191)	M=58% F=42%	49 (mean)	Several hematologic diseases	HCT	2	EORTC-QLQ-C30
Hung, Y. 2014	Australia	Randomized control study; nutrition/exercise intervention	T1: Hospital discharge, T2: 100 days post-HCT	37	Usual care group: M=53% F=47% Extended care group: M=56% F=44%	Usual care 60, extended care 58 (mean vs. median not reported)	Lymphoma and Myeloma	AutoHCT	2	EORTC-QLQ-C30
Jacobs, S. R. 2007	United States	Observational	Either 6 month or 12-month post-HCT	101	M=56% F=44%	53 (mean)	Several hematologic diseases	HCT	1	EORTC-QLQ-C30, FACT-COG
Jarden, M. 2009	Denmark	Randomized control study; exercise, relaxation intervention	Prior to hospitalization, weeks 1, 2, 3, 4, 5, 6, three months; six months	42	M=62% F=38%	39 (mean)	Several hematologic diseases	AlloHCT	9	EORTC-QLQ-C30; Stem Cell Transplantation Symptom Assessment Scale (SCT-SAS)
Jones, D. 2013	United States	Observational	1 month and 3 months	53	M=62% F=38%	57.8 (mean)	MM	AutoHCT	2	M.D. Anderson Symptom Inventory multiple myeloma module (MDASI-MM)
Kav, S. 2009	Turkey	Observational	post-HCT; mean time since transplant = 16 months (4–43)	67	M=46% F=54%	37.5 (mean)	Several hematologic diseases	AutoHCT	1	EORTC-QLQ-C30
Larsen, J. 2007	Sweden	Observational	T0: Admission to Tx unit; T1: at discharge from unit; T2: 3 mo., T3: 6 mo., T4: 12 mo.	41	M=34% F=66%	44 (median)	Several hematologic diseases	HCT	5	Symptom Frequency, Intensity, and Distress Questionnaire SFID-SCT
Mayo, S. 2016	Canada	Observational	T0: within 14 days pre-HCT conditioning; T1: 100 days post, T2: 6 months post	58	M=53% F=47%	47.9 (mean)	Several hematologic diseases	alloHCT	3	EORTC-QLQ-C30
Molassiotis, A. 2011	United Kingdom	Observational	>1 year from diagnosis (mean 5 years)	225 (MM patients=132, Partners=93). 68% of MM patients had received HCT.	M=61% F=39%*	62 (mean)*	MM	HCT	1	EORTC-QLQ-C30
Naegele, M. 2018	Germany	Observational	T0: hospital admission, T1: leucocyte nadir T2: discharge, T3: 30 days post discharge	29	M=65% F=35%	61 (median)	MM	AutoHCT	4	PROVIVO
Persoon, S. 2017	The Netherlands	Randomized control study, exercise intervention	Pre/Post exercise intervention	109 (54 exercise intervention, 55 usual care)	Exercise group: M=59% F=41%; Usual care group: M=67% F=33%	Exercise: 53.5 (median), Usual care: 56 (median)	MM	AutoHCT	2	EORTC-QLQ-C30
Ruark, J. 2019	United States	Observational	At least one year post CAR-T therapy, +/− two months yearly anniversary of t-cell infusion	40	M=63% F=38%	57 (median) (54 at time of CAR-T)	CLL, ALL, NHL	CAR-T	1	Investigator developed measure
Sanders, J. E. 2010	United States	Observational	at least 5 years post-HCT	214 (Non-malignant=53, Lymphoid malignancy=69, Myeloid malignancy=68, CML=24)	M=55% F=45%	Range of subgroups: 25– 32 (mean)	Several hematologic diseases	HCT	1	Neurobehavioral rating scale (NBRS), Modified Memory Questionnaire (MMQ)
Sarkar, S. 2014	Germany	Observational	T0: Before HCT, T1: 100 days post HCT, T2: 12 months post HCT	239	M=62% F=38%	50.4 (mean)	Several hematologic diseases	AlloHCT	3	EORTC-QLQ-C30
Schoulte, Joleen C. 2011	United States	Randomized trial of alternate approaches to prevent GVHD	T0: Baseline pre-transplant, T1 6 months post-transplant	105	M=53% F=47%	35.9 (mean)	Leukemia and Lymphoma	HCT	2	Bush Bone Marrow Transplant Module
Schulz-Kindermann, F. 2007	Germany	Observational Study	T0: Baseline pre-HCT, T1: 100 days post HSCT	19 (at T1, participated)	M=63% M=37%	46.5 (mean)	Several hematologic diseases	AlloHCT	2	EORTC-QLQ-C30
Stewart, A. K. 2013	Canada and United States	Randomized Control Trial of thalidomide/prednisone therapy	Baseline; every 2 months for 6 months, every 3 months from months 6–48, and every 6 months thereafter.	332	M=66% =34%	57.8 (median)	MM	AutoHCT	At least 2	EORTC-QLQ-C30
Watson, M. 2004	United Kingdom	Randomized control trial of BMT vs CCT	1 year post end of treatment	481 (HCT=171)	M=43% F=57%	AlloHCT: 32 AutoHCT: 37 (median at entry)	AML	HCT	1	EORTC-QLQ-C30
Wu, L. M.; Austin, J.; Hamilton 2012	United States	Observational	Between 9 months to 3 years post HCT	245	M=42% F=58%	54.20 (“average”)	Cancers treatable with HCT	HCT	1	FACT-cog
Wu, L. M.; Austin, J.; Valdim 2014	United States	Observational	Between 9 months to 3 years post HCT	42	M=55% F=45%	54.4 (mean)	Leukemia, multiple myeloma, lymphoma	HCT	2	Frontal Systems Behavior Scale (FrSBe)
Wu, L. M.; Kuprian, N. 2019	United States	Observational	Between 9 months to 3 years post HCT	69	M=33% F=67%	57.8 (mean)	Cancers treatable with HCT	HCT	2	FACT-Cog

When HCT group is identified in sample- sex/age are only for that subgroup

^**Denotes that the sex/age columns may include non-HCT sample participants

When the HCT group size is specified in the sample size, average age and gender percentages are for the HCT group only unless noted

^*+Denotes that age at time of transplant was reported in study

PRO measure characteristics

Of the 56 included studies, 21 different cognitive function PRO measures were used (Table 2). The most frequently used measure was the EORTC-QLQ-C30 (n = 32, 57%). Many measures, including the EORTC-QLQ-C30, used 2 or fewer items to measure cognitive function (n = 10; 48%), 5 measures include between 3 and 10 items, and 6 measures had over 15 items. The measure with the largest number of items was the FACT-Cog with 42 items. Fourteen measures asked about cognitive dysfunction or cognitive difficulties, 1 framed questions around cognitive abilities, and 5 used a combination of both (with 1 unknown because item wording was not available to be assessed²²).

Table 2.

Self-reported cognitive function measures used in hematopoietic cell transplant and cellular therapy studies, n=21

Measure	Number of Studies	Percentage of Studies	Number of Items	Number Positive Worded Items	Number Negative Worded Items	Example Item(s)
European Organization for Research and Treatment of Cancer, Quality of Life Questionnaire Core 30 (EORTC-QLQ-C30)	31	55%	2	0	2	(1) Have you had difficulty in concentrating on things, like reading a newspaper or watching television? (2) Have you had difficulty remembering things?
Sickness Impact Profile (SIP)	3	5%	2	0	2	(1) I have difficulty reasoning and solving problems, for example, making plans, making decisions, learning new things. (2) I sometimes behave as if I were confused or disoriented in place or time, for example, where I am, who is around, directions, what day it is.
Cognitive Failure Questionnaire (CFS)	3	5%	25	0	25	(1) Do you find you confuse right and left when giving directions? (2) Do you have trouble making up your mind?
Profile of Mood States (POMS) confusion-bewilderment scale	3	5%	7	1	6	Confused; Unable to Concentrate; Bewildered; Forgetful; Efficient
Patient Reported Outcomes Measurement Information System - Cognitive Ability Scale (PROMIS - CAS)	2	4%	8	8	0	(1) In the past 7 days - My thinking has been slow; (2) In the past 7 days - I have had trouble concentrating
MD Anderson Symptom Inventory (MDASI)	2	4%	1	0	1	“Your problem with remembering things at its worst?”
Functional Assessment of Cancer Therapy - Cognitive Scale (FACT-COG)	2	4%	42	9	33	(1) I have trouble forming thoughts; (2) I have had trouble concentrating; (3) I have had trouble speaking fluently
Cancer Problems in Living Scale (CPILS)	1	2%	2	0	2	(1) Diminished ability to concentrate, (2) Having difficulty in making long-term plans
Neurobehavioral Rating Scale (NBRS)	1	2%	27	0	27	Asked to Place an Z in the appropriate box to represent level of severity (Not Present to Extremely severe): Inattention/Reduce alertness - fails to sustain attention, easily distracted, fails to notice aspects of environment, difficulty directing attention, decreased alertness.
Multiple Abilities Questionnaire (MAQ)	1	2%	38	18	20	(1) I am alert to things going on around me; (2) I can do simple calculations in my head quickly.
The Problem List	1	2%	1	0	1	Has the patient experienced problems with concentration: Yes or No.
FACT-BR	1	2%	23	10	13	(1) I am able to concentrate; (2) I am able to remember new things
Patient Problem List	1	2%	1	0	1	Has the patient experienced problems with memory/concentration: Yes or No.
Unspecified Questionnaire	1	2%	1	unknown	unknown
Piper Fatigue Scale (PFS) cognitive/mood subscale	1	2%	3	3	3	Items use a 0–10 response option with 0 = “Able to concentrate, think clearly, remember” and 10 = “Not able to concentrate...etc.”
Stem Cell Transplantation Symptom Assessment Scale (SCT-SAS)	1	2%	1	0	1	Diminished concentration and memory problems.
Symptom Frequency, Intensity, and Distress Questionnaire (SFID-SCT)	1	2%	2	0	2	(1) Difficulty in concentrating; (2) Difficulty in remembering
Patient Reported Outcomes of Long-Term Survivors after allogenic HCT (PROVIVO)	1	2%	2	0	2	Problems with memory; Problems with concentration
Investigator Developed Measure (Ruark)	1	2%	4	0	4	Difficulty concentrating, finding words, memory, or solving problems y/n
Bush Bone Marrow Transplant Module	1	2%	3	0	3	Difficulty maintaining a train of thought
Frontal Systems Behavior Scale (FrSBe) - Executive Dysfunction	1	2%	17	0	17	Not available; Measure needs to be purchased

PROs and Performance-based Assessments

Eleven studies (20%) collected a combination of performance based and self-reported measures to assess cognitive function. Seven validated PROs were used across the 11 studies in conjunction with performance based, with most using the EORTC-QLQ-C30 (Supplemental Table 2). Less than half of these studies directly compared self-reported and performance-based assessments (5/11; 78.6%) (Table 3). One study using the EORTC-QLQ-C30 was moderately to highly correlated (r=.58) with a composite score of cognitive impairment comprised of performance-based measures²⁹ while another did not find any association pre-HCT or 100 days post-HCT (study does not report r).³⁰ Using the FACT-Cog, Jacobs and colleagues found none of the subscales were significantly related to performance measures of cognitive function except the scale Other People Noticed Deficits, which was still only weakly to moderately related (r= −0.25). The Multiple Abilities Questionnaire demonstrated similar findings, with few significant correlations to performance measures and no correlation above r=0.3.³¹ Harder et al. (2002) found a composite score of performance measures of cognitive impairment was negatively associated with self-reported cognitive function. Jones identified learning/memory and motor function performance measures were positively associated with self-reports of cognitive function at one-month post-HCT.

Table 3.

Effects of hematopoietic cell transplant and cellular therapy on cognitive function

Self-reported cognitive function measure	Study	Mean and SD of self-reported cognitive function measure for study group (HCT recipients)	Self-reported cognitive function scores compared to alternate population (different treatment or general population)	Change in self-reported cognitive function score over time in treatment group
Bush Bone Marrow Transplant Module	Schoulte, J 2011	6 months post HSCT cog difficult subscale: 1.48, SD 0.75	No comparisons	N/A (Baseline data on cog difficulty subscale not provided)
CPILS	Baker, F 2003	7–71 months post-HCT: 38.9% endorse item on cognitive function as somewhat or severe problem; Mean 0.43	Significantly more patients endorsed the item on cognitive function with self-reported Karnofsky scores >=2 compared to those with scores <2.	N/A
CFS	Harder, H. Van Gool, A 2007	Baseline: 26.2 (12.9); 8 months: 27.4 (14.3); 20 months: 28.4 (15.3)	No significant differences between HCT and disease-specific references at baseline, 8 months, or 20 months post-HCT.	No significant change between pre-HCT, 8 months, and 20 months post-HCT.
CFS	Harder, H. Van Gool, A 2005	HCT group (n=101): 26.2 (12.6);	No significant difference between HCT group and reference group (patients with hematological malignancies treated with systemic chemotherapy and/or involved-field radiotherapy).	N/A
CFS	Mayo, S. 2016	T0 71.19 (22.07); T1 73.46 (20.09); 73.93 (23.51)	No comparisons	Analysis of statistical significance of change over time not reported.
EORTC-QLQ-C30	Acaster, S. 2013	Exact numbers NR (graph depicts means) 70–80	No difference in cognitive function among patients at different phases of myeloma (first line treatment, 2nd line treatment, later phase, first treatment free interval)	N/A
EORTC-QLQ-C30	Ahmedzai, S. 2019	Baseline: 74.8 (CI 68.1; 81.6); 100 days: 80 (CI 74.8; 85.2) 6 months: 83.1 (CI 78.4; 87.8) 1 year: 83.5 (CI 78.5; 88.5) 2 year: 83.0 (CI 76.2; 90.3)	No difference between secondary HCT arm and nontransplantation consolidation control arm	HCT arm: Lowest at baseline then increase at 100 days and again 6 months and stabilizes year 1 and 2. Analysis of significance not provided.
EORTC-QLQ-C30	Alaloul, F. 2015	M 79.9; SD 23.6	No difference compared to healthy controls 4–60 months post-HCT	N/A
EORTC-QLQ-C30	Andresen, S. 2011	M 77.0; SD 27.8	Compared to conventional chemo group significantly higher cognitive function (median follow up 8.5 (HCT) and 4.5 years (R-CHOP)); Compared to general German population, HCT group had significantly lower cognitive function post-HCT	N/A
EORTC-QLQ-C30	Bonifazi, F. 2019	Presented in graph form only	No significant difference between treatment group and control.	No significant change in treatment group, presented in graph form only.
EORTC-QLQ-C30	Bush, N. 2000	1 Year post-HCT: 23.4 (1.18); 2 Year post-HCT: 22.3 (1.41); 3 Year post-HCT: 24.4 (1.61); 4 Year post-HCT: 21.7 (2.83)	No comparisons	Analysis of statistical significance of change over time not reported.
EORTC-QLQ-C30	Caocci, G. 2006	92.6 (4.4) 16–137 months post-HCT)	No significant difference between patients with (n=4) and without cGVHD (n=15) 16–137 months post-HCT.	N/A
EORTC-QLQ-C30	Clavert, A. 2017	83.3 range [33.3–100]	Patients with cGVHD (n=46) report significantly lower cognitive function compared to patients who did not have cGVHD (n=15) a median 4.5 years post-HCT.	N/A
EORTC-QLQ-C30	Danaher, E. 2006	Pre-HCT: 88.9 (15.0); Up to 5 days post-HCT: 73.3 (25.8)	No comparisons	Statistically significant decrease in cognitive function between pre-HCT and up to 5 days post-HCT.
EORTC-QLQ-C30	Dean, H. 2012	> 3 years post-HCT: ~80 (5)	HCT survivors not significantly different to healthy controls.	N/A
EORTC-QLQ-C30	Epstein, J. 2002	Day 90–100 post-HCT: 77.2	No comparisons	N/A
EORTC-QLQ-C30	Hacker, E. D. 2011	Pre-HCT: Strength Training: 95.8 (7.7) Usual Activity: 81.5 (25.6); Hospitalization (D4–8): Strength Training: 83.3 (12.6) Usual Activity: 70.4 (21.7); 6 weeks post-HCT: Strength Training: 95.8 (7.7) Usual Activity: 85.2 (21.2)	No significant difference between strength training and usual activity.	Statistically significant decrease in cognitive function between pre-HCT and through hospitalization (day 4–8 post-HCT). Statistically significant increase in cognitive function from hospitalization to 6 weeks post-HCT
EORTC-QLQ-C30	Harder, H. Cornelissen, J 2002	Post-HCT 22–80 months: 74.2 (22.3)	No comparisons	N/A
EORTC-QLQ-C30	Harder, H. Van Gool, A. 2005	HCT group (n=101): 76.5 (10.0)	Significant differences between HCT group and reference group (patients with hematological malignancies treated with systemic chemotherapy and/or involved-field radiotherapy), with reference group reporting higher cognitive function.	N/A
EORTC-QLQ-C30	Harder, H. Van Gool, A. 2007	Baseline: 76.5 (10.0); 8 months: 82.6 (20.5); 20 months: 82.1 (19.1)	No significant differences between HCT and disease-specific references at baseline, 8 months, or 20 months post-HCT.	No significant change between pre-HCT, 8 months, and 20 months post-HCT.
EORTC-QLQ-C30	Hayden, P. J. 2004	78.3 (25.8)	HCT recipients had significantly lower cognitive function compared to general population.	N/A
EORTC-QLQ-C30	Hendriks, M. 2002	Patient initial cognitive function score was 71.21, follow up score (approximately 18–24 months later) 70.20 (SD not provided).	The follow up timepoint of the cognitive score for the Dutch HCT group (70.20) was compared to the general healthy Norwegian population (88.25). An analysis of significance was not provided.	Both timepoints are post SCT. Initial mean cognitive score was 71.21, follow up 70.20, was not significant at <0.01.
EORTC-QLQ-C30	Hjermstad, M. Evensen 1999	Allo-HCT baseline: 90; 1 year: 89; Auto-HCT baseline: 77; 1 year: 80	Auto-HCT recipients had significantly lower cognitive function at baseline compared to allo-HCT recipients. Auto-HCT recipients had significantly lower baseline cognitive function compared to chemotherapy-only recipients. At one year, auto-HCT patients had significantly lower cognitive function compared to chemotherapy-only recipients.	No significant change from baseline to 1 year for auto, or allo-HCT recipients.
EORTC-QLQ-C30	Hjermstad, M., Holte, H. 1999	1-year post-HCT: Leukemia treated with high-dose chemo + allo-HCT: ~89; Lymphoma treated with high-dose chemo + autoHCT: ~80	Lymphoma patients treated with high-dose chemo + autoHCT has significantly lower cognitive function compared to Norwegian general population. Leukemia patients treated with high-dose chemo + alloHCT has no significant difference with Norwegian general population. Each at 1-year post-HCT	N/A
EORTC-QLQ-C30	Hong, F. 2013	means/SD NR; Change from pre-HCT to 14–107 days post-HCT reported: −9.51	No comparisons	Change of −9.51 from pre-HCT to 100 post-HCT timepoint.
EORTC-QLQ-C30	Hung, Y. 2014	means/SD NR; Change from pre-HCT to 100 days post-HCT reported: Usual Care: 14.6 (15.9); Extended Care: 24.0 (29.5)	No significant difference between usual care and extended care 100 days post-HCT.	Change of 14.6 reported for usual care group, and 24.0 reported for extended care group.
EORTC-QLQ-C30	Jacobs, S. R. 2007	Dichotomized items to complaint/non-complaint. Most common complaints were endorsed by 34%−47% of patients.	NR	N/A
EORTC-QLQ-C30	Jarden, M. 2009	Converted to severity scores (not on same scale as EORTC); Percent prevalence of symptoms reported: Diminished concentration - Intervention: 64%; Control: 85%; Memory problems - Intervention: 43%; Control: 72%	Intervention group experienced significantly lower cognitive symptoms	Significant decrease in cognitive function from baseline through first 4 weeks post-HCT for both intervention and control arm. Cognitive function returns to baseline 3 and 6 months post-HCT.
EORTC-QLQ-C30	Kav, S. 2009	81.5, SD 22.5	No comparisons	N/A
EORTC-QLQ-C30	Mayo, S. 2016	T0 (14 days before transplant conditioning) 26.88 (SD 15.31), T1 (100 days after transplant) 27.67 (SD 15.49), T2 (6 months post-transplant) 27.69 (SD 13.97).	No comparisons	Cognitive function lowest at baseline, 14 days pre-HCT, small increase at 100 days and remains stable thereafter to 6 months. Statistical analysis of significance of change over time not provided.
EORTC-QLQ-C30	Molassiotis, A. 2011	Raw cognitive function score 3.6, SD 1.4. Transformed score 73.3 *	Cognitive function score in study sample was lower than a Danish Myeloma sample (77), Austrian Hematology sample (76.8), and an Austrian healthy subjects sample (87.7). Analysis of significance of difference not provided.	N/A
EORTC-QLQ-C30	Persoon, S. 2017	Exercise group T0: 87.3(15.6) T1: 83.7(18.0) Usual care group: T0: 80.5(21.8) T1: 82.3(20.4)	No significant beneficial effects of the exercise program on cognitive function when compared to usual care were found.	Analysis of statistical significance of change over time not reported.
EORTC-QLQ-C30	Sarkar, S. 2014	Estimated means shown in graph form only.	Over the three assessment points, patients with high FCR had a significantly lower quality of life in cognitive function compared with patients with low FCR (p= 0.003)	Only shown in graph form; no statistical tests.
EORTC-QLQ-C30	Schulz-Kindermann, F. 2007	Before HSCT (T0): 85.1, 18.3 (SD) 100 days post-HSCT (T1): 86.0, 21 (SD)	No comparisons	No significant difference in cognitive function between T0 and T1.
EORTC-QLQ-C30	Stewart, A. K. 2013	Mean and SD for cognitive function subscale NR.	Patients assigned to thalidomide-prednisone had inferior HRQoL scores, including cognitive function, when compared to control group.	NR
EORTC-QLQ-C30	Watson, M. 2004	53% of patients reported problems in the cognitive functioning subscale.	No significant cognitive differences between treatments (CCT, AlloHCT, AutoHCT).	N/A
FrSBe	Wu, L. M.; Austin, J.; Valdim 2014	NR	No comparisons	NR
FACT-BR	Correa, D. 2019	Baseline: ~120; post-treatment: ~135; Year 1: ~150; Year 2: ~160; Year 3: ~160; Year 4: ~155; Year 5: ~158	No statistically significant difference in cognitive function over time (baseline to 5 years) between autoHCT + high-dose chemo compared to reduced-dose whole brain radiotherapy.	Cognitive function significantly improved from baseline to 1-year post-treatment.
FACT-COG	Jacobs, S. R. 2007	Mean/SD NR; Percent endorsing complaints by item ranged from 34%−47%; Most endorsed (47%) My memory is as good as it has always been (reverse coded).	No comparisons	N/A
FACT-COG	Wu, L. M.; Austin, J.; Hamilton 2012	NR	No comparisons	N/A
FACT-COG	Wu, L. M.; Kuprian, N. 2019	NR	No comparisons	NR
Investigator developed measure	Ruark, J. 2019	15 patients (37.5%) reported at least one cognitive difficulty in the post CAR-T survey.	No comparisons	N/A
MDASI-MM	Jones, D. 2013	NR	No comparisons	NR
MDASI	Cleeland, S. 2000	Overall Sample = Remembering: 2.82 (2.93); Attention: 2.49 (2.81); BMT Sample = Remembering: 1.77 (2.11)	No comparisons	N/A
MMQ	Sanders, J. 2010	0.55	No significant difference compared to healthy controls.	N/A
MMQ	Basinski, J. 2010	6 months post-HCT: Pts with no delirium episode: 0.45 (0.43); Pts with delirium episode 0.81 (0.52); 12 months post-HCT: Pts with no delirium: 0.42 (0-.36); Pts with delirium episode 0.77 (0.49)	Patients who experienced a delirium episode at 6 months post-HCT reportedly significantly worse cognitive function compared to those who did not. Patients who experienced a delirium episode at 12 months post-HCT reportedly significantly worse cognitive function compared to those who did not.	Analysis of statistical significance of change over time not reported.
MAQ	Booth-Jones, M. 2005	Average number of complaints 14.8 (10.4); Range 1–42	MAQ self-reported cognitive complaints were not significantly associated with performance-based measures, though there was a trend for lower performance related to more complaints 5 to 10 months post-HCT	N/A
NBRS	Basinski, J. 2010	6 months post-HCT: Pts with no delirium episode: 1.03 (0.99); Pts with delirium episode 0.88 (1.47); 12 months post-HCT: Pts with no delirium: 0.82 (0.79); Pts with delirium episode 1.65 (1.38)	Patients who had experienced a previous delirium episode reported significantly worse neurocognitive functioning on the NBRS (P= .004)	Analysis of statistical significance of change over time not reported.
NBRS	Sanders 2010	1.64	HCT patients had significantly lower cognitive function compared to healthy age and gender matched controls.	N/A
PPL	Crooks, M 2013	“Memory concentration” on problem list: Baseline: 6/37 (16%), At discharge: 8/36 (22%), 3 months post-HCT: 1/28 (3.6%), 6 months post-HCT: 2/16 (12.5%)	No comparisons	Analysis of statistical significance of change over time not reported.
PROMIS-AC GCS	Ghazikhanian, S. E. 2017	NR	No comparisons	Cognitive function did not significantly change over time, from pre-HCT to 3–12 months post-HCT.
PROMIS - CAS	Bartley, E. 2014	3 months post-HCT: 18.2 (8.5); 6 months post-HCT: 17.8 (8.0)	No comparisons	Analysis of statistical significance of change over time not reported.
PROVIVO	Naegele, M. 2018	At T3 (30 days post discharge) the cognitive measures (difficulty with concentration and memory) were among the top reported symptoms, with 59% and 47% reporting them.	No comparisons	Paper presents a bubble graph in which cognitive symptom intensity and distress appear to become greater over the timepoints (from hospital admission, leukocyte nadir, to 30 days post discharge). Numbers data and analysis of significance not presented.
PFS	El-Banna, M. 2004	Pre chemo: ~4; Post-chemo, but before HCT: ~4; 2 days post-HCT:~5; 7 days post-HCT: ~6.5; 14 days post-HCT: 4	No comparisons	Differences in cognitive function scores over time were not statistically significant.
POMS	Andrykowski, M. A.; Altmaier, E. M. 1990	NR	No comparisons	Analysis of statistical significance of change over time not reported.
POMS	Andrykowski, M. A.; Henslee, P. J. 1989	Mean 28.3 (8–52) months post-HCT: 7.8 (6.2); Mean 37.3 (12–57) months post-HCT: 9.2 (5.9) Mean 51.7 (26–71) months post-HCT: 10.2 (5.0)	No comparisons	Analysis of statistical significance of change over time not reported.
POMS	Bush, N. 2000	NR	Compared with normative sample, HCT patients at year 1–4 showed less confusion.	Analysis of statistical significance of change over time not reported.
POMS	Chang, G. 2009	Baseline: 1.11 (1.53); 12 months 0.68 (1.14); 18 months 0.84 (1.28)	No comparisons	Analysis of statistical significance of change over time not reported.
POMS	Fann, J. 2002	From Day −7 through Day 30: Those with no delirium episode (n=41): 0.5 (0.3); Those with delirium episode (n=49): 0.7 (0.4) range [0.1–1.8]	Those who experienced a delirium episode, on average, reported significantly worse cognitive function compared to those who did not experience a delirium episode.	Analysis of statistical significance of change over time not reported.
SIP	Bishop, M. 2007	Mean of 6.7 years post-BMT (1.9–19.4): Partners: 1.42; Survivors: 2.42; Controls: 0.86	HCT survivors and partners reported significant decline in cognitive function compared to control population. HCT survivors reported significantly more decline in cognitive function compared to partners.	N/A
SIP	Andrykowski, M. A.; Altmaier, E. M. 1990	NR	Those who received higher doses of total body irradiation (TBI) were significantly more likely to endorse cognitive dysfunction compared to those who received less TBI among patients 1–8 years post-HCT	Analysis of statistical significance of change over time not reported.
SIP	Andrykowski, M. A.; Henslee, P. J. 1989	Mean 37.3 (12–57) months post-HCTT: 31.9 (29.4); Mean 51.7 (26–71) months post-HCT: 43.5 (37.2)	No comparisons	Significant difference from assessment 2 (ave. 37.3 months post-HCT) to assessment 3 (ave. 51.7 months post-HCT), indicating significant decline in cognitive function over time.
SCT-SAS	Jarden, M. 2009	NR	No comparisons	Analysis of statistical significance of change over time not reported.
SFID-SCT	Larsen, J. 2007	At discharge, 4 patients reported “difficult remembering” and 4 reported associated distress. At 1 year post HCT it was 3 and 3. At discharge 9 reported “difficulty concentrating” and 7 reported it as distressing. At 1 year post it was 4 and 4. (n at discharge was 20, at 1 year post was 9)	No comparisons	Analysis of statistical significance of change over time not reported.
The Problem List	Braamse, A.M. 2014	AlloHCT patients: ~68% report cognitive-emotional problem; AutoHCT patients: ~59% report cognitive-emotional problem	No comparisons	Analysis of statistical significance of change over time not reported.
Unspecified questionnaire	Deeg, H. 1998	2–20 years post-HCT: Mean 7.6–8.1 range [1–10]	No comparisons	Analysis of statistical significance of change over time not reported.

^*Mean and SD include not HCT patients

Abbreviations: Cancer Problem in Living Scale (CPILS), Cognitive Failure Questionnaire (CFS), European Organization for Research and Treatment of Cancer, Quality of Life Questionnaire Core 30 (EORTC-QLQ-C30), Frontal Systems Behavior Scale (FrSBe) - disinhibition and apathy subscales, Functional Assessment of Cancer Therapy - Brain (FACT-BR), Functional Assessment of Cancer Therapy - Cognitive Scale (FACT-COG), Investigator developed measure, M.D. Anderson Symptom Inventory - Multiple Myeloma (MDASI-MM), MD Anderson Symptom Inventory (MDASI), Modified Memory Questionnaire (MMQ), Neurobehavioral Rating Scale (NBRS), Patient Problem List (PPL), Patient Reported Outcomes Measurement Information System- Applied Cognitive – General Concerns Scales (PROMIS-AC GCS), Patient Reported Outcomes Measurement Information System - Cognitive Ability Scale (PROMIS - CAS), Patient Reported Outcomes of Long Term Survivors after allogenic HCT (PROVIVO), Piper Fatigue Scale (PFS), Profile of Mood States (POMS), Sickness Impact Profile (SIP), Stem Cell Transplantation Symptom Assessment Scale (SCT-SAS), Symptom Frequency, Intensity, and Distress Questionnaire (SFID-SCT).

Comparison Results

Some studies identified significant differences between groups in self-reported cognitive function, while others did not (Table 3). Among those using the EORTC, half did not find a significant difference among those studies that reported comparisons (n=11; 50%). Comparisons were often made to the general population, with results indicating patients who received an HCT had significantly lower cognitive function scores. Another study comparing HCT patients to healthy age and gender matched controls found HCT patients reported significantly lower cognitive function using the Neurobehavioral Rating Scale.³² Similarly, a study using partners as controls found HCT patients reporting larger declines in cognitive function using the Sickness Impact Profiles.³³ In contrast, when other studies compared to healthy controls, they found no differences using the EORTC-QLQ-C30 or the Modified Memory Questionnaire.^21,26,32

Longitudinal Results

Most longitudinal studies (11/18, 61%; 12 longitudinal but did not report cognitive changes over time) did not see significant changes in cognitive function over time (Table 3). Of those that did find significant change, general patterns emerged around the timing of statistically significant increasing and decreasing function that align with clinical expectations. Though few studies collected data within the first month post-HCT, those that did noticed significant decline in cognitive function during hospitalization³⁴ and within the first month³⁵ compared to pre-HCT. Those collecting data at timepoints 100 days post-HCT and later found significant increases in cognitive function, with plateaus happening around 1 year post-HCT.^19,36,37 Using the EORTC-QLQ-C30, Jarden et al. found returns to baseline cognitive function occurring at 3 and 6 months post-HCT. There was not much change identified one year or more out from transplant (7 of 9 studies assessing 1 yr+ timepoints did not see significant change between pre-HCT and 1 year), though one study using the Sickness Impact Profile found cognitive function declined from 12 month to later timepoints (an average of 37 months post-HCT to 52 months post-HCT).²³

Associated Factors PRE-HCT

We reviewed whether any clinical, sociodemographic, or treatment-related factors were significantly associated with self-reported cognitive function (Table 4). We divided the factors into those that are pre-HCT factors (e.g. sociodemographic) and those that are post-HCT (e.g. clinical outcomes). Overall, there were more post-HCT factors associated with cognitive function compared to pre-HCT. For those pre-HCT factors, the most frequent were baseline depression,^9,38 age, sex, ^32,39,40and education.^40,41 Other pre-HCT factors included pre-HCT treatment factors such as increased dose of total body irradiation pre-BMT,⁴² receiving cranial irradiation pre-HCT,³² as well as higher fear of cancer recurrence⁴³ and higher income were all associated with better cognitive function.⁴⁰

Table 4.

Primary endpoint of study, pre- and post-treatment factors associated with self-reported cognitive function

Study	Primary Endpoint of Study	Symptoms, functioning, quality of life, and clinical outcomes associated with self-reported cognitive function
		BEFORE HCT/CT	AFTER HCT/CT
Acaster, S. 2013	HRQoL during a treatment free interval compared to other treatment phases.
Ahmedzai, S. 2019	Time to progression between autoHCT and nontransplant consolidation
Alaloul, F. 2015	Difference between HCT survivors and healthy comparisons in QOL and social support post-HCT at least 3 months post-HCT		Social support (+) Overall symptoms (−), (average scores of fatigue, nausea and vomiting, pain, dyspnea, insomnia, appetite loss, constipation, diarrhea)
Andresen, S. 2011	Differences in QoL between HCT and conventional chemo, and general German population post-HCT (median 8.5 years)
Andrykowski, M. A.; Altmaier, E. M. 1990	Correlations between total body irradiation (TBI) dosage and long-term cognitive impairment	Total body irradiation dosage (−) Education (−) (with POMS- confusion subscale)	Psychological adjustment to illness scale - distress (−)
Andrykowski, M. A.; Henslee, P. J. 1989	Longitudinal assessment of psychosocial status and functional quality of life and affective status among HCT patients between 8 and 71 months post-HCT
Baker, F 2003	Develop and assess construct validity of objective problems-in-living scale with HCT patients 7–71 months after transplant		Self-reported Karnofsky score 7–71 months post-HCT (−)
Bartley, E. 2014	Assess moderating effect of pre-HCT holding back on impact of health symptoms on social well-being 3 and 6 months post-HCT
Basinski, J. 2010	Associations between having a delirium episode in acute phase of HCT and cognitive function, distress, and HRQoL at 6 and 12 months post-HCT
Bishop, M. 2007	Compare QoL, marital adjustment, and personal growth of HCT survivors, their spouses, and matched controls.
Bonifazi, F. 2019	Assessment of long-term outcomes, including QOL on HCT patients treat with ATLG to prevent GVHD.
Booth-Jones, M. 2005	Assess prevalence of self-reported cognitive complaints, identify associations with performance based cognitive function, and determine the relationship with QoL in HCT patients 5–10 months post-HCT	Age (+)	Depressive symptomology (CES-D) (−); Fatigue - average daily rating & level of interference (FSI) (−); Physical Health (SF-36) (+); Mental Health (SF-36) (+)
Braamse, A.M. 2014	Assess prevalence, distribution, and associated risk factors with distress and problems up to 5 years post-HCT.
Bush, N. 2000	Evaluate the multidimensional course of QoL at years 1, 2, 3, 4 post-HCT
Caocci, G. 2006	Assessing pre-transplant communication and post-transplant quality of life 16–137 months post-HCT
Chang, G. 2009	Quantify neuropsychological and mental status at the start of HCT, 12 and 18 months after.
Clavert, A. 2017	Assess QoL in long term adult survivors (2+ years post-HCT) of alloHCT in malignant and nonmalignant disorders.		cGVHD (−)
Cleeland, S. 2000	Development and validation of M.D. Anderson Symptom Inventory
Correa, D. 2019	Assess differences in cognitive function, QoL, white matter abnormalities and cortical atrophy among patients who received reduced-dose whole brain radiotherapy and high-dose chemo + autoHCT		Cognitive Function at year 3, 4, 5: White matter abnormalities grades 2/3 (−); Cortical atrophy (−)
Crooks, M. 2013	Monitor patients self-reported psychosocial distress and patient problems overtime baseline to 6 months post-HCT
Danaher, E. 2006	Examine patterns of fatigue, physical activity, health status, and QoL 5 days before and 5 days after HCT
Dean, H. 2012	Investigate long term (>3 years) immunological status of patients treated with autoHCT for malignant lymphoma.
Deeg, H. 1998	Long term outcomes (up to 20 years) in patients with aplastic anemia who survived at least 2 years post-transplant.
El-Banna, M. 2004	Associations between fatigue and depression among autoHCT recipients over time (prior to chemo initiation through 14 days after transplant).	Depression (−)	Depression (−)
Epstein, J. 2002	Assess taste and smell, and QoL among HCT patients 90–100 days post-HCT>
Fann, J. 2002	Determine prevalence, incidence, and severity of delirium for HCT patients in acute phase (−7 days through Day 30) and identify any associated risk factors.
Ghazikhanian, S. E. 2017	Examine change in sleep and cognitive problems from pre to post-HCT (3 to 12 months post) and assess relationship between sleep and cognitive problems.		Sleep problems (−); Depressive symptoms (−); Greater fatigue (−)
Hacker, E. D. 2011	Pilot test the effects of strength training compared to usual activity on physical activity, muscle strength, fatigue, perceive health, and QoL post-HCT (six weeks post-HCT).
Harder, H. Cornelissen, J. 2002	Assess cognitive functioning and quality of life among long term adult survivors of HCT (22–82 months post-HCT).		Fatigue (−); Composite score of performance measures of cognitive impairment (−)
Harder, H. Van Gool, A. 2005	Study pre-HCT cognitive functioning in a large sample (n= 193) of HCT patients and relations to potentially confounding factors 2.7 months post-HCT.
Harder, H. Van Gool, A. 2007	Compare cognitive function longitudinally (up to 20 months after baseline) between HCT recipients and disease-specific reference group ages 16 to 65.		Anxiety (−); Depression (−); Feelings of intrusion (−) 20 months post-HCT
Hayden, P. J. 2004	Measure long term (time not specified) QOL for CML patients who receive HCT and compare results to general population
Hendriks, M. 2002	Evaluate QOL in HCT patients as well at their partners and physicians’ perceptions of the patient’s QOL about 2.5- and 4.5-years post-transplant
Hjermstad, M. Evensen 1999	Evaluate HRQoL in adult HCT patients and baseline (pre-HCT) and 1-year post-HCT compared to chemotherapy patients.
Hjermstad, M., Holte, H. 1999	Compare HRQoL at 1-year post-HCT with reference general Norwegian population.
Hong, F. 2013	Report and interpret HRQoL change among various types of cancer to determine minimally clinically meaningful differences.
Hung, Y. 2014	Evaluate feasibility and effectiveness of home-based, telephone nutrition and exercise intervention on nutritional status, body composition, QoL 100 days post_HCT compared to usual care.
Jacobs, S. R. 2007	Describe frequency of cognitive complaints, assess whether greater cognitive complaints were associated with poorer HRQoL, and assess relationship between self-report cognitive complaint and cognitive performance.	Female (−); Education (+);	Physical well-being (+); FACT-Cog and EORTC Cognitive Function scores (+); Anxiety (−)
Jarden, M. 2009	Assess the effects of a multimodal intervention (exercise, progressive relaxation, psychoeducation) on longitudinal patterns (4–6 weeks and 3–6 months post-HCT) of symptom severity.
Jones, D. 2013	Assess acute effect of HCT on neuropsychological functioning of multiple myeloma patients		Performance based cog function (learning/memory and motor function) at one month post HCT (+)
Kav, S. 2009	Assess quality of life and other difficulties faced by HCT patients 100 days post-transplant
Larsen, J. 2007	Describe functional status, general health, and symptom distress in HCT patients from admission to 1-year post-transplant, and identify variables associated with patient perceived general health.		Poor general health, self-reported (−) (with difficulty concentrating)
Mayo, S. 2016	Understanding the relationship between neurocognitive functioning and medication management ability in HCT patients.
Molassiotis, A. 2011	Identify nature of needs and quality of life in myeloma patients and their partners.		Actively receiving maintenance therapy (−)
Naegele, M. 2018	Explore the experience of multiple myeloma patient’s symptom frequency, intensity and distress when treated with mephelan followed by HCT
Persoon, S. 2017	Evaluate the effectiveness of an exercise program on physical fitness and fatigue in HCT patients
Ruark, J. 2019	Assess neuropsychiatric and other patient-reported outcomes of 40 patients with relapse/refractory chronic lymphocytic leukemia, non-Hodgkin lymphoma, and acute lymphoblastic leukemia 1 to 5 years after treatment with CD19-targeted CAR T cells	Depression (−)	Acute neurotoxicity (trend only, p 0.08) (−) Global mental health score (−) Global physical health score (−) Pain interference (−) Sleep disturbance (−) Fatigue (−) Depression (−) Anxiety (−) Physical function (−) Social function (−) White (−) *n=40 >80% of sample was white, no non-white people reported cog. difficulties
Sanders, J. E. 2010	Determine the extent of physical limitations, psychosocial issues, and cognitive symptoms among survivors (> 5 yrs. post-HCT) of childhood HCT by comparing to age and gender matched controls.	More therapy = second or subsequent remission therapy or relapse or cranial irradiation (−); Receiving cranial irradiation pre-HCT (−); Age at transplant >13 (+)
Sarkar, S. 2014	To examine the course and the prevalence of a high fear of cancer recurrence (FCR) in patients undergoing HCT, before HCT, 100 days, and 12 months post HCT.	Higher fear of cancer recurrence (−)	Higher fear of cancer recurrence (FCR) (+)
Schoulte, Joleen C. 2011	To investigate the influence of coping style on interference caused by a variety of common post-treatment symptoms post HCT.		Avoidant coping styles (−)
Schulz-Kindermann, F. 2007	Assess cognitive performance among hematological malignancy patients before and 3 months after HCT.
Stewart, A. K. 2013	Comparing thalidomide-prednisone as maintenance therapy and melphalan 200 mg/m2 in HCT patients, primary endpoint of study was overall survival, secondary end points included health-related quality of life, with a median follow up of 4.1 years.		Thalidomide-prednisone therapy (−)
Watson, M. 2004	Evaluate the impact of CCT vs BMT on QOL, 1 year post HCT	Age >35 (−)
Wu, L. M.; Austin, J.; Hamilton 2012	Investigate if adverse effects of subjective cognitive impairment occur because cognitive difficulties reduce survivors’ confidence that they can manage HSCT-related symptoms	Age (+) Men (+) Income (+)
Wu, L. M.; Austin, J.; Valdim 2014	To examine patient’s neurobehavioral function pre and post HCT (9 months to 3 years) and its impact on HCQOL
Wu, L. M.; Kuprian, N. 2019	Explore HCT survivors’ perceptions of cognitive impairment and its impact on daily life functioning at 9 months to 3 years post HCT.

Note: Blank cell indicates the study either did not assess associations with patient reported cognitive function or no statistically significant associations (p < .05) were identified. (+) = factor is positively associated with perceived cognitive function; (−) = factor is negatively associated with perceived cognitive function

Associated Factors POST-HCT

The most common factors associated with higher self-reported cognitive function following transplant were less depression^9,31,44,45 and other post-HCT symptoms and functioning, including decreased overall symptoms,²⁶ better general health,⁴⁶ decreased psychological distress,⁴² lower self-reported Karnofsky scores,⁴⁷ lower fatigue,^9,29,44 higher physical^9,31,41 and mental health,^9,31 fewer sleep problems,^9,44 less anxiety,^9,41,45 fewer feelings of intrusion,⁴⁵ pain interference,⁹ physical function,⁹ and social function.⁹ Additional factors included a negative association with avoidant coping style,²⁷ as well as a positive association with higher fear of cancer reoccurrence⁴³ and increased social support.²⁶

Treatment factors that were both negatively associated with cognitive function included actively receiving maintenance therapy⁴⁸ and thalidomide-prednisone therapy.⁴⁹ Very few studies statistically assess the relationship between self-reported cognitive function and clinical outcomes such as relapse, chronic and acute graft-versus host disease (GVHD), and/or death. Only one study reported a negative association with chronic GVHD, though most studies did not assess the relationship between cognitive function and chronic GVHD.⁵⁰

Discussion

There were 21 different measures of self-reported cognitive function used in 56 studies among adult CT patients. Most of these studies used the EORTC-QLQ-C30, which measures health-related quality of life more broadly, but includes a 2-item cognitive function subscale. Importantly, none of the studies included the supplementary module specific to high-dose chemotherapy and transplantation (EORTC QLQ-HDC29), which was validated among the transplant population with items specific to the transplant experience.⁵¹ Among the studies using other measures, several had little previous validation. The use of so many different measures made it difficult to compare across studies. There are resources available, such as PROsetta stone, that allow for direct comparison between different PRO measures, but little has been done with cognitive function measures. In order to make direct comparisons across CT studies, there is a need for crosswalks with cognitive function measures. Since the EORTC-QLQ-C-30 cognitive function subscale is the dominant measure in this therapeutic context, it may be prioritized for a crosswalk to other measures, though it should be noted that with only 2 items, it was not originally intended to provide a comprehensive assessment of cognitive function.

A review of studies collecting patient-reported cognitive function longitudinally identified consistent patterns in when patients experienced declines in cognitive function and when function returned to baseline levels. Results coincide with what is clinically expected and previously published using performance-based measures; cognitive function declines soon after treatment, namely within the first month post-HCT with improvement by 6 and 12 months post-HCT.⁵² Self-reported cognitive function was found to return to pre-HCT levels by a year post-HCT and in some studies as early as 3-to-6-months post-HCT. Importantly, those studies not collecting data at earlier timepoints may have failed to capture the changes in patient’s cognitive function. It is also important to consider how changes in cognitive function may affect self-reported measurement of cognitive impairment, for example, Howland et al found that in a sample of older adults, self-reported cognitive function did not align with proxy-reported function. Another study by Gruters et al found that subjective cognitive decline was associated with both performance-based cognitive decline and depressive symptoms- highlighting an additional important clinical consideration.^53,54

Few studies (20%) collected both performance-based and self-reported cognitive function, with even fewer reporting on correlations between these measures. Though there are few, the studies capture a range of diseases, both autologous and allogeneic transplant patients, and a wide span of timepoints from every three days post-HCT up to day 30²⁰ to yearly, five years after HCT.¹⁹ Additionally, several self-reported measures (five different measures) were used with four studies using the EORTC-QLQ-C30. Findings from these studies support other literature that performance-based and self-report cognitive function are not well correlated among cancer patients, leading to the idea that self-reported cognitive function may be capturing a unique perspective of patients experience that is otherwise not captured in performance-based tests. Further, evidence confirms previous conclusions that self-reported cognitive function was more consistently associated with depression as oppose to performance-based measures.¹⁰ Whether collected at baseline or post-HCT, our study found depression was the most consistent factor significantly associated with cognitive function. This suggests self-reported cognitive function may be an indicator of psychological distress more so than cognitive impairment.

The assessment of self-reported cognitive function among adult CAR-T patients is rare in the published literature. The one published study is a smaller sample (n=40) of patients at least one-year post-therapy, utilizing an investigator-developed measure to assess cognitive function.⁵⁵ Though CAR-T is a relatively new therapy that is rapidly increasing in use, there remains much to be desired for assessing cognitive function, especially given the unique neurotoxicity associated with CAR-T. Indeed, two studies published recommendations for the routine collection of PROs among CAR-T patients, chiefly among those PROs is cognitive function.^7,56 It will be essential for future studies to assess patients perceptions of cognitive (dys)function in a way that promotes translatability, consistency, and comparability by collecting measures at earlier timepoints (within first three months post-therapy) and using validated measures.

This scoping review comes with important limitations. We limited our search to only studies written in English and to those published in peer reviewed journals. With CAR-T being so new, there are likely protocols currently utilizing cognitive function measures that were not captured in this review. This review was limited to adult patients, which leaves out cognitive function results for pediatric patients who are also impacted by neurocognitive impacts after CT. As previous findings have shown evidence of cognitive disfunction prior to HCT in oncologic populations (suggesting that some amount of dysfunction may be attributed to the disease itself)⁵, one must consider that as a potential confounding factor, especially given the heterogenicity of diagnosis and treatments among study populations included in this review. Finally, we limited our inclusion criteria to studies that reported results for cognitive function subscales, which leaves out a number of studies that only report overall scores of quality of life, where items asking about cognitive function are part of a larger set of quality of life items that comprise a single score. Though these scores may provide helpful insight into the symptom and functioning trajectories of CT patients, it would be too difficult to meaningful compare to other cognitive function-specific measures. Despite these limitations, this review offers the first comprehensive assessment of self-reported cognitive function measures used among CT patients. It is also among the first to review evidence for comparing performance-based and self-reported cognitive function measures in this patient population.