The Effect of Chemotherapy on Balance, Gait, and Falls Among Cancer Survivors: A Scoping Review

Abstract

Background: 

Evidence has mounted that cancer survivors demonstrate impaired balance, gait, and increased risk of falls following chemotherapy treatment. This relationship must be clearly defined in order to screen for functional impairments, intervene appropriately, effectively advise stakeholders, and plan for future research.

Purpose: 

To conduct a scoping review to systematically map the current literature on the effect of chemotherapy on balance, gait, and falls among cancer survivors.

Methods: 

Data Sources: Four databases were searched for studies. Study Selection: In total, 3769 abstracts were identified through literature searches. After screening and full-text review, 30 articles were included in this scoping review. Data Extraction: Data related to study design, temporality, sample characteristics, cancer/chemotherapy type(s), outcome measures, and relevant findings were extracted from full texts. Data Synthesis: Descriptive quantitative summaries were calculated and a narrative analysis was performed.

Discussion and Limitations: 

The current body of literature is largely in agreement that chemotherapy negatively impacts static balance, dynamic balance, and gait and results in increased risk of falls throughout the survivorship continuum. The relationship between chemotherapy-induced peripheral neuropathy (CIPN) and these outcomes is evident. Males, non–breast cancer diagnoses, and causal pathways other than CIPN are underrepresented in the literature.

Conclusion: 

This scoping review summarized the current body of literature related to the effect of chemotherapy on balance, gait, and falls. Clinical implications are proposed and suggestions made for future research to include more prospective studies with emphases on chemotherapy type and dosage, fewer sex-biased diagnoses, and to explore alternative causal pathways.

As advances in the management of cancer have led to improved prognoses and survival rates, the population of cancer survivors in the United States has surged to more than 16.9 million.¹ Contrasted with these advances are the challenges often faced by individuals throughout cancer survivorship. More than 50% of adult-onset cancer survivors report physical performance limitations secondary to side effects and long-term sequelae of cancer treatments.² These functional limitations are more prevalent closer to diagnosis and treatment, but even long-term survivors remain more functionally limited than healthy peers.³ The gravity of physical limitations among cancer survivors is underscored by the correlation between low levels of physical activity, poorer outcomes, increased risk of recurrence, and increased mortality for certain cancers.⁴

Common physical limitations that have been linked to decreased overall physical function and quality of life (QOL) among cancer survivors include imbalance,⁵ gait dysfunction,⁶ and increased risk of falls.^7–9 The risk of falls increases after a cancer diagnosis when compared with prediagnosis fall rates,¹⁰ and fall rates among cancer survivors have been shown to be twice those of cancer-free peers¹¹ and community-dwelling older adults.¹² Imbalance, gait dysfunction, and falls may have particularly severe consequences in this population due to comorbidities increasing the risk of fall-related injury, secondary decline in physical activity, and subsequent functional decline.^13–17

The maintenance of balance—also known as postural control—requires the integration of visual, vestibular, and somatosensory inputs.¹⁸ Evidence has shown that chemotherapy often causes multisystem morbidities, including neurotoxic effects that impact these sensory systems.^8,19–21 Some classes of chemotherapy, known to have neurotoxic effects, have been linked to peripheral neuropathy, visual impairment, and inner ear dysfunction including hearing and vestibular loss.^22–26 Evidence has mounted that patients who are treated with chemotherapy demonstrate impaired balance and are at a greater risk of falling.^3,27,28 These secondary effects can critically impact QOL^8,29 and functional mobility.⁵ Given the profound effects that imbalance, gait dysfunction, and falls have on functional independence, QOL, and long-term health and survivorship, the treatment or prevention of these symptoms is an essential component of cancer rehabilitation for all survivors.

Efforts have been made to describe the incidence and clinical characteristics of imbalance, gait dysfunction, and falls among cancer survivors following chemotherapy exposure.^11,22 While some trends have become clear, the heterogeneity of methodology, clinical populations, and outcome measures in the published research has resulted in an ambiguously defined relationship between chemotherapy and balance, gait, and falls. Without a clear description of the effect of chemotherapy on these variables in respect to time and patient characteristics, we cannot assume generalizability across clinical populations or the survivorship continuum. Furthermore, our ability to screen for functional impairments, intervene appropriately, effectively advise stakeholders, and plan for future research is limited. For this reason, a scoping review was conducted to systematically map the research done in this area and to identify any existing gaps in knowledge.

To guide this review, the following research question was formulated: What is known from the literature about the effect of chemotherapy on balance, gait, and falls among cancer survivors? Secondary objectives included in this inquiry were as follows: (a) to examine what clinical populations are represented in the literature; (b) to explore what outcomes have been used to quantify the effect; (c) to map what is known in respect to time; (d) to consider what causal pathways have been identified; and finally, (e) to identify the current gaps in the literature.

METHODS

The method of this scoping review was principally informed by the framework proposed by Arksey and O'Malley.³⁰ Based on these guidelines, after developing our research question, we sought to search for and identify relevant studies; select studies to be included in this review; chart the data; and collate, summarize, and report the results. In addition, stakeholder feedback was solicited from content experts (1 physical therapist and 1 oncologist—both of whom are cancer rehabilitation researchers) to gain insight into how the findings of this scoping review could be applied within their clinical practices and within future research priorities. Stakeholder feedback was collected via e-mail and incorporated into this articleʼs “Discussion” section. We describe our methods and results in accordance with the guidelines and checklist of the PRISMA Extension for Scoping Reviews (PRISMA-ScR).³¹ No protocol for this scoping review has been previously published.

Data Sources and Searches

We searched the following 4 electronic databases for relevant research articles: PubMed (including MEDLINE), EMBASE, Web of Science, and CINHAL. Filters were applied where possible to retrieve articles in the English language, with human participants, and time of publication since 2006. We selected this time frame as it followed the publication of the Institute of Medicine report, “From Cancer Patient to Cancer Survivor: Lost in Transition,” which raised awareness about the functional consequences of cancer and its treatments.³² This publication demarcated a new era of cancer survivorship research, and we were most interested in the literature published since that time.

Our key words and initial search strategy were developed in partnership with an experienced librarian and were further refined through team discussion and in light of preliminary results. Our search strategy, which was modified depending on the database searched, aimed to capture any article that addressed chemotherapy side effects in the realm of balance, gait, and falls. In addition to searching electronic databases, we hand-searched relevant journals, reference lists of included articles, as well as those of excluded review articles,^6,22,25 until saturation was achieved. Our search strategies, which were most recently implemented in January 2020, are reported in Table 1. Following January 2020, active searching beyond reference lists and relevant journals was not performed.

TABLE 1 - Search Strategies

Database	PubMed (MEDLINE)	EMBASE	Web of Science	CINHAL
Search terms	“cancer survivors”[MeSH Terms] OR “neoplasms”[Mesh Terms] OR “survivors”[MeSH Terms]) “neoplasms/drug therapy”[Mesh Terms] OR “antineoplastic combined chemotherapy protocols/adverse effects”[Mesh Terms] OR “antineoplastic agents”[Mesh Terms] OR “neurotoxins/adverse effects”[Mesh Terms] OR “neurotoxins/the rapeutic use”[Mesh Terms] “postural balance”[Mesh Terms] OR “posture”[Mesh Terms] OR “peripheral nervous system diseases/chemically induced”[Mesh Terms] OR “gait”[Mesh Terms] OR “sensation disorders/chemically induced”[Mesh Terms] OR “accidental falls”[MeSH Terms] OR vestibular[All Fields] OR (“Balance”[Journal] OR “balance”[All Fields]) 1 and 2 and 3	‘postural control’/exp OR ‘postural control’ OR ‘body equilibrium’/exp OR ‘body equilibrium’) AND (‘chemotherapy’/exp OR ‘chemotherapy’ OR ‘cancer therapy’/exp OR ‘cancer therapy’)	((TS=(postural control OR postural balance OR body equilibrium OR sway OR gait) AND TS=(chemotherapy OR cancer therapy))	(chemotherapy) AND (postural control or postural balance or balance or postural stability)
Results	2978 (Search #4)	140	500	151

Study Selection

We imported citations yielded from our search strategy into the open-source reference management software, Zotero,³³ and then into Covidence,³⁴ a systematic review production tool. Title and abstract screening to determine whether citations were within the scope of our research question was undertaken independently by both authors. Screening results were compared, and citations that were jointly agreed upon were advanced to full-text review. Discrepancies were resolved through discussion and/or review of the citation's full text with eventual consensus.

Prior to full-text screening, we developed a list of inclusion and exclusion criteria. To be eligible for inclusion, articles needed to meet the following criteria: (a) exposure of all participants to chemotherapy or stratification of results to demonstrate effect of chemotherapy; and (b) objective measurement of balance-, gait-, or falls-related outcomes during or after chemotherapy exposure. Articles were excluded for the following reasons: (a) study examined outcomes following exposure to a medical intervention other than chemotherapy; (b) study assessed outcomes of exercise/rehabilitation intervention; (c) participants with medical histories/comorbidities likely to confound outcomes of interest (ie, tumors of central nervous system, cranial irradiation, lower-limb amputation); (d) ambiguous reporting of chemotherapy exposure and/or outcomes; (e) subjective report of function/balance as primary outcome; (f) pediatric population (<18 years old); (g) review article; and (h) conference abstract with full text not published. Only articles employing performance-based measures were reviewed to study the effect of chemotherapy on balance and gait, though studies identifying falls via self-report were included. The authors then independently reviewed each full text and voted to “include” or “exclude.” Discrepancies were resolved through discussion and consensus.

In keeping with the iterative process of scoping review methodology, our study selection process evolved as we considered what studies were most relevant to our research question. For example, several screened articles used the 6-Minute Walk Test (6MWT) as a primary outcome during and after exposure to chemotherapy. Although the 6MWT could be considered an objective measure of gait, given our focus on balance, gait, and falls—as opposed to ambulation endurance—we refined our criteria regarding gait outcomes to only include gait speed and kinematic measures of gait. In addition, we adapted our exclusion criteria regarding the pediatric population to allow for inclusion of studies in which participants were treated for pediatric cancers as long as they met the following criteria: (1) participants were older than 18 years at time of the study; and (2) adult outcome measures were used.

Finally, we found instances where multiple articles pertained to a single study. In these cases, if the results of the article provided novel information, we opted to include the study in our review. In cases where an article provided only pilot data for a future article that was included in our study, we opted to exclude it. See the Figure for a PRISMA-ScR flow diagram.

Fig.: PRISMA flow diagram. This figure is available in color online (www.rehabonc.com).

Data Extraction and Quality Assessment

Before performing data extraction, we convened to specify characteristics relevant to the main research question and secondary objectives. A data extraction form was developed using Microsoft Excel to capture the corresponding data from each included article. The authors piloted this form by independently extracting data from 6 articles (20%) that met inclusion criteria and comparing results.

The finalized data extraction form included the following characteristics: first author, year of publication, study design, time of study in regard to chemotherapy exposure/diagnosis, sample size, control group, participant age and sex, cancer type(s) included in the sample, chemotherapeutic agents, outcome measures, and relevant findings. Following validation and finalization of the data extraction form, the lead author independently extracted data from the remaining articles. These data were fully reviewed with the other author prior to data synthesis. In accordance with the guidelines and checklist of the PRISMA-ScR,³¹ quality assessment of individual articles, including risk of bias, was not performed.

Data Synthesis and Analysis

The literature was examined for clinical characteristics, outcome measures, temporal themes, and causal pathways in the relationship between chemotherapy and balance, gait, and falls. We felt that identifying these traits across the literature would provide greatest clinical utility and help identify gaps in the literature.

First, when available, descriptive quantitative summaries such as sample sizes and breakdown of participants' clinical characteristics were aggregated to examine overall representation across the included studies. Articles were then organized 3 ways: by outcome measure (static balance vs dynamic balance vs gait vs falls); temporally in relation to chemotherapy exposure or diagnosis (acute [during, ≤3 months following chemotherapy] vs subacute [3-12 months following chemotherapy] vs chronic [1-5 years following chemotherapy] vs long term [>5 years following chemotherapy]); and by causal pathway (chemotherapy-induced peripheral neuropathy [CIPN] vs other).

As many articles reported multiple outcomes, it was possible for an article to be represented in more than 1 outcome measure category. It was also possible for an outcome measure to be represented in more than 1 category. For example, since all studies that reported results for the Short Physical Performance Battery (SPPB), which contains elements of static balance assessment, gait speed, and functional capacity, reported results for individual subscales, the SPPB was included in both static balance and gait outcome sections. In addition, although the Fullerton Advanced Balance (FAB) scale contains items assessing static balance, the majority of individual items measure dynamic balance and therefore we considered it a measure of dynamic balance.

For studies with temporal overlap, a decision was made on the basis of the emphasis of the articles' results and discussion sections. If the only temporal statistic provided by an article was time since diagnosis, it was grouped with articles reporting outcomes from the same time period since chemotherapy exposure. Within each organization scheme, studies were also assessed for their methodology and were defined as case report (longitudinal study, N = 1), prospective cohort (longitudinal study with control), prospective observation (longitudinal study without control), cross-sectional (single assessment postexposure), or retrospective cohort (retrospective analysis with control). Narrative analysis was used to summarize findings related to participant characteristics, outcome measures, temporality, and causal pathways.

RESULTS

A total of 3769 abstracts were identified through initial database searches, with 2 additional articles discovered via hand-searching reference lists³⁵ and relevant journals.³⁶ After duplicates were removed, 3597 studies were left for screening. Of these, 3523 studies were deemed irrelevant based on their titles and/or abstracts, leaving 74 articles for full-text review. After full-text review, 44 articles were excluded, ultimately leaving 30 articles to be included for data extraction and analysis in this scoping review. ^{3,5,27,28,35–60} Included articles with extracted data were organized temporally in Table 2.

TABLE 2 - Description of Included Studies

Author (Year)	Study Type	Timing	Participant Characteristics	Outcome Measures	Relevant Findings
During, ≤3 mo s/p chemotherapy (n = 10)
Müller et al (2020)36	Prospective cohort	Median 21 d (3-55) since final chemotherapy	N = 35 Control: 35 age-matched healthy controls Median age: 51 y (38-73) Sex: M = 2, F = 33 Cancer type: Breast, rectal, pancreatic, oral Chemotherapy type: Taxane-based, platinum-based, vinca alkaloids, taxane-platinum combo	Postural control: Quiet standing COP on the force plate (5 conditions). Falls: FES-I; Number of falls during chemotherapy (patients) or within last 6 mo (controls)	Following chemotherapy, patients showed significantly increased temporal and spatial measures of COP compared with baseline scores as well as compared with healthy matched controls. Withdrawal of visual input during balance testing resulted in greater temporal and spatial COP displacements following chemotherapy than with baseline and healthy matched controls.
Wampler et al (2007)57	Cross-sectional	Within 30 d of final chemotherapy	N = 20 Control: 20 age-matched healthy controls Mean age: 50.35 ± 9.34 y Sex: F = 20 Cancer type: Breast Chemotherapy type: Taxane-based	Postural control: SOT; COP velocities (Kistler force plate test) under 4 conditions; FAB scale; TUG test. Visual acuity.	Women in the breast cancer group had significantly poorer low-contrast visual acuity. Women who were treated with taxanes showed poorer static and dynamic postural control. The largest group mean differences on the SOT were observed for conditions where vision is either occluded or altered.
Tofthagen et al (2012)56	Prospective observation	During chemotherapy	N = 109 Control: None Mean age: 58.4 ± 11.8 y Sex: M = 40, F = 69 Cancer type: Breast, colon, lung, ovarian, prostate, pancreatic, esophageal, other (78.9% stage III or IV across diagnoses) Chemotherapy type: Taxane-based, platinum-based	Falls: Self-reported incidence since initiating chemotherapy. Self-report: CIPNAT	Participants who reported falls received higher cumulative doses of chemotherapy, had a significantly higher number of neuropathic symptoms, more severe muscle weakness and loss of balance, and higher interference with walking. Participants who received a taxane were more likely to have fallen than those who had received a platinum-based chemotherapy drug.
Monfort et al (2017)52	Prospective observation	During chemotherapy through 1-3 mo after final infusion	N = 33 Control: None Mean age: 47.8 ± 11.2 y Sex: M = 1, F = 32 Cancer type: Breast Chemotherapy type: Taxane-based	Postural control: Quiet standing balance COP (on the force plate); 10-m walk test (average step length and gait speed). Self-report: EORTC QLQ-CIPN20	Decreased balance was observed after the first chemotherapy cycle and progressed with cumulative exposure. Patients also demonstrated slower walking speeds as they progressed through treatment. These functional deficits were mirrored with increased patient-reported symptom severity for all EORTC QLQ-CIPN20 subscales.
Kolb et al (2016)48	Prospective observation	During chemotherapy	N = 116 Control: None Mean age: 55.5 ± 11.9 y Sex: M = 7, F = 109 Cancer type: Breast, ovarian, lung Chemotherapy type: Taxane-based, platinum-based, taxane-platinum combo	Falls: Daily self-report of falls and near falls via the automated telephone system during full course of chemotherapy. Self-report: symptoms of CIPN	After controlling for age, sex, and length of follow-up, participants with self-reported CIPN symptoms were 2.5 times more likely to have a fall event (near falls and falls together) than those without CIPN symptoms.
Klepin et al (2016)45	Prospective observation	During inpatient examination, within 8 wk s/p hospital discharge after induction chemotherapy	N = 49 Control: None Mean age: 70 ± 6.2 y Sex: M = 28, F = 21 Cancer type: Acute myeloid leukemia Chemotherapy type: Anthracycline + cytarabine ± etoposide, other	Postural control and gait speed: SPPB	For SPPB component scores, significant declines were detected in gait speed and balance. Participants with depressive symptoms at baseline and follow-up had a significant decline in SPPB score, but those without depressive symptoms at either time point did not. Participants with cognitive impairment at baseline and at follow-up had a greater decline in SPPB score than those without cognitive impairment at either time point.
Hile et al (2010)27	Case report	Baseline (initiation of chemotherapy) through 2.5 y	N = 1 Control: None Age: 81 y Sex: F = 1 Cancer type: Breast Chemotherapy type: Paclitaxel	Postural control and gait speed: SPPB	After 3 cycles, paclitaxel therapy was stopped by the oncologist because of neurotoxicity. Declines as large as 50% were seen in performance-based measures at 12 wk and persisted at 2.5 y, and the patient reported recurrent falls, cane use, and mobility-related disability.
Monfort et al (2019)51	Cross-sectional	Approximately 6 wk post-chemotherapy	N = 14 Control: 6 cancer survivors, no chemotherapy exposure Mean age: Control: 59.3 ± 9.6 y −CIPN: 55.9 ± 9.0 y +CIPN: 50.0 ± 15.0 y Sex: M = 3, F = 17 Cancer type: Breast, colorectal Chemotherapy type: Taxane-based, oxaliplatin	Postural control: Single-/dual-task treadmill walking tests; kinematics measured using a motion capture system. Self-report: EORTC QLQ-CIPN20. Cognitive test: Groton Maze Learning Test	Participants with combined peripheral sensory and cognitive impairments demonstrated worsened gait impairments during dual-task walking. The no/mild CIPN group demonstrated decreased orbital stability during the dual-task condition at initial contact compared with the control group, with the moderate/severe CIPN group showing a similar but smaller effect that did not reach statistical significance. The no/mild CIPN group also demonstrated decreased orbital stability compared with the moderate/severe CIPN group during dual-task gait.
Monfort et al (2019)28	Cross-sectional	Approximately 6 wk post-chemotherapy	N = 14 Control: 6 cancer survivors, no chemotherapy exposure Mean age: Control: 59.3 ± 9.6 y −CIPN: 55.9 ± 9.0 y +CIPN: 50.0 ± 15.0 y Sex: M = 3, F = 17 Cancer type: Breast, colorectal Chemotherapy type: Taxane-based, oxaliplatin	Postural control: Sensory organization protocol (COP on the force plate). Self-report: EORTC QLQ-CIPN 20	The +CIPN group demonstrated significant deficits in summary COP measures compared with the control and −CIPN groups. The majority of significant differences in balance parameters were between the +CIPN and control groups during vision-deprived condition. The asymptomatic −CIPN group also demonstrated increased COP dispersion compared with the control group during select conditions.
Bahcaci et al (2019)37	Prospective observation	During chemotherapy	N = 32 Control: None Mean age: 47.5 y (31-63) Sex: F = 32 Cancer type: Breast Chemotherapy type: Not indicated	Postural control: Double-/single-leg/tandem stance tests with eyes opened/closed; Romberg test; STS test; TUG test (an accelerometer was used in all tests for postural sway). Fear of falling: FES-I	Patients demonstrated worse postural balance performance through chemotherapy cycles in most of the tests. Significant postural changes were observed in single-leg and tandem stance tests with both closed and opened eyes, as well as in the TUG and STS tests. Fall efficacy scale surprisingly demonstrated a significantly decreased fear of falling between the chemotherapy cycles 1-3, 1-4, and 2-4.
3-12 mo s/p chemotherapy (n = 5)
Kneis et al (2016)46	Cross-sectional	Mean 5.2 ± 4.4 mo s/p chemotherapy	N = 20 Control: 16 age-matched healthy controls Mean age: 48.8 ± 4.5 y Sex: F = 20 Cancer type: Breast Chemotherapy type: Taxane-based	Postural control: COP displacement during bi-/monopedal stance on the force plate. Self-report: FACT&GOG-Ntx	The breast cancer group's COP displacement was greater than controls' and correlated significantly with the level of antagonistic co-contraction indices (an electromyographic measure of articular stiffening during balance reactions) and self-reported CIPN symptoms.
Ward et al (2014)60	Retrospective cohort	≤12 mo s/p first dose of chemotherapy	N = 65 311 Control: None Mean age: 65-69 y: 27% 70-74 y: 31% 75-79 y: 26% 80+ y: 16% Sex: M = 29 390 (45%), F = 35 921 (55%) Cancer type: Breast, colon, lung, prostate Chemotherapy type: Singlet vs doublet vs non-neurotoxic	Falls: Fall-related injury, ie, code for hip fracture (excluding codes for pathologic or spontaneous fractures), head injury, joint dislocation, or other traumatic fracture codes (ICD-9 codes: 800-839, 850-854).	The rate of fall-related injuries (after censoring for death) was significantly higher among patients who received a neurotoxic doublet (2 chemotherapies) than those who received a single neurotoxic agent or those who received a non-neurotoxic agent.
McCrary et al (2019)5	Cross-sectional	Mean 12 mo s/p completion of chemotherapy	N = 190 Control: None Mean age: 57 ± 13 y Sex: M = 65, F = 125 Cancer type: Breast, colorectal, lymphoma, myeloma, ovarian, other Chemotherapy type: Taxane-based, platinum-based, vinca alkaloids, taxane-platinum combo, bortezomib, thalidomide	Postural control: Modified CTSIB with sway meter. Self-report: Total Neuropathy Score clinical version, EORTC QLQ-CIPN20 questionnaire	In total, 68% of patients reported CIPN symptoms at assessment. Symptomatic patients displayed increased balance deficits, with degree of balance impairments consistent with a healthy elderly population (aged ≥65 y) reporting multiple falls over the subsequent year. Increasing CIPN severity correlated with increasing balance deficits. Patient-reported numbness/tingling, weakness, and balance deficit; age; and vibration perception—were strongly linked to balance deficits.
Niederer et al (2014)54	Cross-sectional	13 participants with ongoing chemotherapy; 8 participants completed ≤12 mo prior	N = 21 Control: 2 control groups (n = 24 each): one age-matched and one senior group Mean age: 51 ± 7 y Sex: M = 1, F = 20 Cancer type: Breast, other Chemotherapy type: Taxane-based, platinum-based	Falls: FES-I. Postural control: COP in upright stance with eyes covered; gait speed (comfortable fluid walking). Strength: Quadriceps MVIF	Fear of falling and postural sway were higher and gait speed and MVIF were lower in cancer patients than in age-matched controls. Patients and senior controls differed significantly in gait speed and MVIF (higher values in the patient group) but not in fear of falling and postural sway.
Jing et al (2016)44	Prospective observation	Initiation of chemotherapy through 12 mo s/p chemotherapy	N = 106 Control: None Mean age: 50.2 ± 8.6 y Sex: F = 106 Cancer type: Ovarian Chemotherapy type: Carboplatin, paclitaxel	Falls: Self-reported incidence recorded in diary. Self-report: Severity of CIPN	At 1-y follow up, the rate of peripheral neuropathy still existed with a rate of 88.5%, 57.3%, and 38.7% at 3, 6, and 12 mo, respectively, s/p drug withdrawal. Life accidents occurred in 31 patients, including falls (9.4%).
1-5 y s/p chemotherapy (n = 8)
Zahiri et al (2019)59	Cross-sectional	Mean 2.75-5.5 y s/p diagnosis	N = 82 Control: 57 age-matched healthy controls Mean age: 71.1 ± 9.7 y Sex: M = 36, F = 46 Cancer type: Breast, lung, multiple myeloma, colorectal Chemotherapy type: Taxane-based, platinum-based, vinca alkaloids	Falls: History of falls in the past year; FES-I. Gait: measured via LEGSys: spatial parameters (stride velocity and stride length) and temporal parameters (stride time and double support). Postural control: Measured via BalanSens: Quiet standing eyes open/closed	All gait parameters of interest were significantly degraded in the group with cancer when compared with the control group. Both groups had deterioration in balance with the eyes closed condition compared with the eyes open condition, but the magnitude of increase was on average more than 2-fold higher in the group with cancer than in the control group. Having +CIPN increased the OR of falls by 1.55 and 2.07 compared with −CIPN and controls, respectively.
Winters-Stone et al (2011)58	Cross-sectional with prospective observation	≤2 y s/p chemotherapy	N = 59 Control: None Mean age: 58.5 ± 9.7 y Sex: F = 59 Cancer type: Breast Chemotherapy type: Not indicated	Postural control: SOT. Gait: gait speed (4-m walk). Falls: self-reported for the past year (retrospective) and monthly for 6 mo (prospective). Vision: Visual assessment battery	Subjects who took longer to read letters on contrast sensitivity chart were more likely to have fallen. Past participants with a history of falls had lower scores on condition 5 of the SOT. Vestibular score on the SOT mediated the relationship between treatment and falls among breast cancer survivors who received chemotherapy-only but not adjuvant endocrine therapy.
Petrovchich et al (2019)55	Cross-sectional	Mean durations of CIPN: 3.56-4.09 y across groups	N = 416; divided into 3 BMI groups: Normal, overweight, and obese Control: None Mean age: Normal weight: 60.56 ± 11.42 y Overweight: 61 ± 10.26 y Obese: 61.15 ± 9.41 y Sex: M = 57, F = 359 Cancer type: Breast, colon, lung, ovarian, other Chemotherapy type: Taxane- based, platinum-based, taxane-platinum combo	Postural control: TUG test, FAB scale. Self-report: CIPNAT	Compared with the normal weight group, both overweight and obese survivors had higher scores for the self-reported severity and frequency of balance problems. Compared with the normal weight group, survivors in the other 2 BMI groups had significantly worse balance scores.
Miaskowski et al (2018)50	Cross-sectional	Mean durations of CIPN: 3.19-4.98 (SD = 3.24-5.27) y across study groups	N = 371 Control: None Mean age: CIPN: 59.68 ± 10.39 y CIPN + HL: 64.13 ± 9.16 y CIPN + HL + TIN: 62.5 ± 9.78 y Sex: M = 50, F = 321 Cancer type: Breast, colon, ovarian, lung, other Chemotherapy type: Taxane- based, platinum-based, taxane-platinum combo	Postural control: TUG test, FAB scale	Compared with the survivors with only CIPN, a higher percentage of survivors in the other 2 groups reported trouble with balance. In addition, compared with survivors with CIPN-only, survivors with CIPN/HL/TIN reported high severity scores for balance problems and demonstrated worse TUG scores.
Miaskowski et al (2017)49	Cross-sectional	Mean duration of CIPN: 3.94 ± 4.24 y	N = 623 Control: None Mean age: +CIPN: 60.9 ± 10.52 y −CIPN: 58.38 ± 12.27 y Sex: M = 96, F = 527 Cancer type: Breast, colon, ovarian, lung, other Chemotherapy type: Taxane-based, platinum-based, taxane-platinum combo	Postural control: TUG test, FAB scale	Survivors with CIPN had a significantly longer TUG test and a significantly lower FAB score.
Kober et al (2018)47	Cross-sectional	Mean duration of CIPN: 3.8 ± 4 y	N = 211 Control: None Mean age: +CIPN: 59.59 ± 9.82 y −CIPN: 53.36 ± 9.55 y Sex: M = 1, F = 210 Cancer type: Breast, ovarian, other Chemotherapy type: Paclitaxel	Postural control: TUG test, FAB scale. Self-report: CIPNAT	Survivors with peripheral neuropathy were more likely to report trouble with balance as well as higher severity and frequency scores associated with balance problems. In addition, these survivors demonstrated worse TUG and FAB scores.
Gewandter et al (2013)42	Cross-sectional	Mean 15.5 ± 21.4 mo s/p completion of chemotherapy	N = 471 Control: None Mean age: 60 y (28-86) Sex: M = 137, F = 334 Cancer type: Breast, other Chemotherapy type: Taxane, other	Falls: History of recent (3 mo) falls. Self-report: EORTC QLQ-CIPN20	Those reporting a fall had higher (worse) sensory and motor neuropathy scores on EORTC QLQ-CIPN20. After adjustment for demographic and cancer characteristics, more severe motor neuropathy was the only variable significantly associated with reported falls.
Extermann et al (2017)41	Cross-sectional	Mean 15.3 ± 3.2 mo s/p completion of chemotherapy	N = 27 Control: 29 cancer survivors, no chemotherapy exposure Mean age: 70.5 ± 3.6 y Sex: F = 56 Cancer type: Breast (ER-positive) Chemotherapy type: Not indicated	Postural control and gait: SPPB; 400-m walk test	There were no significant differences in the physical function tests or physical activity. Survivors in both groups accumulated daily steps slightly above the national average for their gender and age class. They also spent 50 min/d in moderate-vigorous physical activity, which is significantly greater than the normative population.
>5 y s/p chemotherapy (n = 6)
Bao et al (2016)38	Cross-sectional	Mean 5.6 ± 3 y s/p chemotherapy	N = 296 Control: None Mean age: 62 ± 9.0 y Sex: F = 296 Cancer type: Breast (early-stage ER-positive) Chemotherapy type: Taxane-based	Falls: Incidence of falls in past 12 mo	Severity of CIPN was associated with higher rates of falls, with 23.8%, 31.9%, and 41.5% in the ‘‘no CIPN,’’ ‘‘mild,’’ and ‘‘moderate-to-severe’’ groups, respectively, experiencing falls. Compared with women with no CIPN, moderate to severe CIPN was associated with a higher incidence of falls, with adjusted OR of 2.27.
Hsieh et al (2019)43	Cross-sectional	Mean years s/p diagnosis: 5.8 ± 7.4 (CIPN group), 6.8 ± 2.9 (no CIPN group)	N = 17 Control: 12 healthy controls Mean age: 53.5 ± 11.8 y Sex: M= 2, F= 15 Cancer type: Breast cancer, lymphoma, prostate, leukemia, salivary gland Chemotherapy type: Not indicated	Gait: 6-m walk across (gait speed, step length, and step width). Falls: Physiological Profile Assessment	No group differences in gait speed. Step length variability was significantly less in cancer survivors with neuropathy than in healthy controls. Cancer survivors with neuropathy had significantly greater step width variability than healthy controls.
Fino et al (2019)40	Cross-sectional	Mean 61.0 ± 51.5 mo s/p chemotherapy	N = 434 Control: 49 healthy controls Mean age: 62.5 ± 6.4 y Sex: F = 434 Cancer type: Breast, colon, ovarian, lymphoma, uterine, lung, other Chemotherapy type: Not indicated	Postural control: Quiet standing with inertial sensor. Falls: Self-reported falls in previous 6 mo. Self-report: CIPN symptoms	Cancer survivors had worse sway than healthy control subjects in components related to sway magnitude and mediolateral frequency of sway, but no difference in the component related to resultant/AP sway jerk and frequency. Cancer survivors who reported neuropathy were more likely to have higher resultant/AP sway frequencies and jerk than asymptomatic survivors, whereas survivors who reported a fall were more likely to have lower frequencies of mediolateral sway than participants without a history of falls. Falls were more strongly associated with mediolateral sway in survivors with more severe neuropathy; whereas falls were more strongly associated with resultant/AP sway frequency in survivors with less severe neuropathy.
Winters-Stone et al (2017)3	Cross-sectional	Average 6 y s/p chemotherapy	N = 512 Control: None Mean age: 62.6 ± 6 y Sex: F = 512 Cancer type: Breast, colon, ovarian, lymphoma, uterine, lung, other Chemotherapy type: Not indicated	Postural control: Timed chair stand, SPPB. Gait: speed; step number, rate, and length; base of support. Self-report: falls in the past year	The +CIPN group took significantly more, but shorter and slower, steps, had shorter strides, and spent more time in stance and double-support phase than asymptomatic women. Increasing CIPN symptom severity was linearly associated with slower chair stand times, slower walking speeds, and worse SPPB scores. Having CIPN symptoms significantly increased the odds of falling in the past year compared with being symptom free, even after adjusting for all covariates. Symptomatic women were 1.8 times more likely to report a fall than were asymptomatic women.
Einarsson et al (2016)39	Cross-sectional	Mean 15 y s/p chemotherapy	N = 16 Control: 25 healthy controls Mean age: 26.1 ± 6.9 y Sex: F = 56 Cancer type: Malignant solid (non-CNS) tumors Chemotherapy type: Not indicated	Postural control: Quiet stance; perturbed stance (both with eyes open/eyes closed) on the force plate	During balance perturbations, those treated with chemotherapy had significantly poorer stability than controls. Subgroup analyses showed that survivors who were treated before the age of 12 y had poorer stability than controls and survivors who were treated after the age of 12 y.
Ness et al (2013)53	Cross-sectional	Median years since diagnosis: 25.2 (10.7-48.2)	N = 475 Control: None Median age: 31.6 y (18.7-63.8) Sex: M = 237, F = 238 Cancer type: Carcinoma, Ewing sarcoma, neuroblastoma, osteosarcoma, retinoblastoma, rhabdomyosarcoma, Wilms tumor Chemotherapy type: Vincristine, vinblastine, cisplatin, carboplatin	Postural control: SOT; TUG test	Among participants without a lower-limb amputation, 25% did not meet the normative threshold of <6 s to complete the TUG test. 12% of the participants demonstrated a composite score of <70% on the SOT. Participants who demonstrated sensory impairment demonstrated an elevated risk for decreased mobility on the TUG test, even after controlling for ankle dorsiflexion strength, age, sex, weight, and height.
Unspecified (n = 1)
Marshall et al (2017)35	Cross-sectional	S/p chemotherapy (time not specified)	N = 8 Control: 8 age-matched healthy controls Mean age: 61.38 ± 7.42 y Sex: F = 8 Cancer type: Breast, colorectal Chemotherapy type: Taxane-based, oxaliplatin-based	Gait: Velocity, step length/time, base of support, swing time, single-/double-support time measured by GAITRite. Falls: TUG test	Gait velocity was significantly slower and step length was significantly shorter in the subjects with CIPN. In addition, CIPN participants had a significantly higher TUG score than the controls.

Abbreviations: AP, anteroposterior; BMI, body mass index; CIPN, chemotherapy-induced peripheral neuropathy; CIPNAT, Chemotherapy-Induced Peripheral Neuropathy Assessment Tool; CNS, central nervous system; COP, center of pressure; EORTC QLQ-CIPN20, European Organization for Research and Treatment of Cancer Quality of Life Questionnaire–CIPN 20-item scale; CTSIB, Clinical Test for Sensory Interaction in Balance; FAB, Fullerton Advanced Balance; FACT&GOG-Ntx, Functional Assessment of Cancer Therapy/Gynaecology Oncology Group–Neurotoxicity; FES-I, Falls Efficacy Scale–International; HL hearing loss; ICD-9, International Classification of Diseases, Ninth Revision; MVIF, maximum voluntary isometric force; OR, odds ratio; SOT, Sensory Organization Test; s/p, status-post/after; SPPB, Short Physical Performance Battery; STS, Sit-To-Stand; TIN, tinnitus; TUG, Timed Up and Go.

Clinical Characteristics and Participants

Aggregated quantitative data can be seen in Table 3. We opted to exclude one study from this aggregation,⁶⁰ as the Ward et al⁶⁰ database sample of 65 311 participants was an outlier in almost all categories. Across the 29 aggregated studies, sample sizes ranged from 1 to 623 participants, with a median of 59. Females accounted for 83.92% of all participants. Samples ranged in average age from 26.1 to 81 years old, with a weighted average of 57.3 years. Among the Ward et al⁶⁰ 65 311 participants, 57% were between 70 and 79 years old, all were older than 65 years, and 55% were female.⁶⁰

TABLE 3 - Participant Characteristics

Demographics (n = 29 studies)
Sample size, median (range)	59 (1-623)
Sex, %
Male	16.08
Female	83.92
Age, weighted mean (range), y	57.3 (26.1-81)
Cancer type (n = 26 studies)	Frequency in Aggregated Sample, %
Breast	55.4
Extracranial childhood tumors	9.9
Ovarian	6.8
Colorectal	5.3
Lymphoma, leukemia, myeloma	4.2
Lung	2.1
Other (including uterine, pancreatic, esophageal, salivary gland, prostate)	16.1
Chemotherapeutic agents (n = 21 studies)
Taxane-based	47.5
Platinum-based + taxane-based combo	19.1
Platinum-based	15.4
Vinca alkaloid	8.5
Other	9.5

All 30 articles provided information regarding study participants' oncologic diagnoses. Twenty-six articles (87%) included participants with a history of breast cancer, whereas 8 (26.7%) samples contained only participants with breast cancer. Twenty-six articles provided descriptive statistics of participants' oncologic diagnoses; of these, samples were aggregated to determine overall frequency of each cancer diagnosis (Table 3). Although extracranial childhood tumors represent the second largest frequency of cancer type (9.9%), this is attributable to one study,⁵³ which had the fourth largest sample size (n = 475). Diagnoses among the Ward et al⁶⁰ 65 311 participants included lung cancer (54%), colon cancer (27%), breast cancer (18%), and prostate cancer (1%).

Twenty-two (73.3%) articles provided information regarding which chemotherapeutic agents participants were exposed to, and 21 (70%) provided detailed descriptive statistics. Of those 21, samples were aggregated to determine the overall frequency of chemotherapeutic agents (Table 3). Taxane-based chemotherapies were the most frequent exposure (47.5%), which likely correlates with this being a common line of treatment of breast cancer. While 3 articles reported the cumulative effects of chemotherapy exposure on balance,^37,52 gait,⁵² or falls,⁵⁶ only one article reported a difference in outcomes by chemotherapy type.⁵⁶ Ward et al⁶⁰ examined the difference in fall-related injuries based on exposure to 1 versus 2 versus none neurotoxic chemotherapies.

Outcome Measures

Outcome measures used are listed in Table 4.

TABLE 4 - Outcome Measures

Outcome Measures Used Among Included Studies
Static Balance (n = 17)	Dynamic Balance (n = 9)	Gait (n = 10)	Falls (n = 15)
Prospective (n = 5)	Prospective (n = 1)	Prospective (n = 3)	Prospective (n = 5)
Quiet standing on the force plate	Sit-to-Stand (with accelerometer)	Gait speed SPPB (gait speed subscale) 10-m walk test	Fear of falling (FES-I)
Double/single/tandem stance with eyes open/eyes closed	TUG test (with an accelerometer)	Kinematics Step length	Incidence of falls
Romberg test			Retrospective cohort (n = 1)
SPPB			Fall-related injuries
Cross-sectional (n = 12)	Cross-sectional (n = 8)	Cross-sectional (n = 7)	Cross-sectional (n = 9)
Sensory Organization Test	Fullerton Advanced Balance scale	Gait speed SPPB subscale 6-m walk across 4-m walk (fast and usual pace) 400-m walk test Comfortable fluid walking	Fear of falling (FES-I)
Quiet/perturbed stance on the force plate	TUG test	Kinematics Step length/width Stride velocity/length Stride time Single-/double-support time Step number/rate/length Base of support	Incidence of falls (prior 3-12 mo)
SPPB	Single/dual-task treadmill walking		TUG test
			Activities-Specific Balance Confidence Scale
			Physiological Profile Assessment

Abbreviations: FES-I, Falls Efficacy Scale–International; SPPB, Short Physical Performance Battery; STS, Sit-To-Stand; TIN, tinnitus; TUG, Timed Up and Go.

Static Balance

Seventeen (56.7%) studies included at least one measure of static balance. Of these, 5 were prospective studies, including 1 case report of a breast cancer survivor exposed to paclitaxel,²⁷ 3 prospective studies without control groups following patients with breast cancer and acute myeloid leukemia (AML) during and after taxane or anthracycline chemotherapy, respectively,^37,45,52 and 1 prospective cohort study examining a mixed sample during the acute period following taxane, platinum, or vinca alkaloid chemotherapy.³⁶ Four of the 5 prospective studies performed assessments during chemotherapy exposure, with follow-up times of all 5 ranging from a few weeks to 2.5 years postexposure. The remaining 12 studies reporting static balance measures were cross-sectional, with assessment points ranging from within 30 days to 25 years following completion of chemotherapy.^* Chemotherapeutic agents within these 12 cross-sectional studies included taxanes, platinums, and vinca alkaloids.

All 5 of the prospective studies reported declines in static balance measures during and after chemotherapy, with some evidence of withdrawal of visual input resulting in greater imbalance.³⁶ Eleven of the 12 cross-sectional studies reported decreased static balance measures, whereas the one study reporting nonsignificant differences cited significantly above-average levels of physical activity among participants that may have accounted for nonsignificant differences.⁴¹ Among the 11 studies with significant findings, there was evidence of worse static balance impairments among participants with CIPN^3,5,28,40 and under altered/absent visual testing conditions.^28,57–59 Ness et al⁵³ reported impaired performance on the Sensory Organization Test (SOT) in a sample of childhood cancer survivors with median 25.2 years since diagnosis, indicating possible long-term effects of chemotherapy on static balance.

Dynamic Balance

Nine (30%) studies included at least one measure of dynamic balance. Of these, one was a prospective study that also reported static balance outcomes among breast cancer survivors during unspecified chemotherapy.³⁷ The other 8 studies—6 of which did not report static balance measures—were cross-sectional, with assessment points ranging from within 30 days to 25 years following completion of chemotherapy.^{35,47,49–51,53,55,57} Chemotherapeutic agents reported in the 8 cross-sectional studies included only taxanes and platinums.

The prospective study reported significant postural changes using an accelerometer during Sit-to-Stand and Timed Up and Go (TUG) tests among patients with breast cancer undergoing chemotherapy.³⁷ Each of the 8 cross-sectional studies reported decreased dynamic balance measures, including decreased FAB scale scores,^{47,49,50,55,57} increased TUG test times,^{35,47,49,50,53,55,57} and instability while performing dual-task treadmill walking.⁵¹ A similar trend was found as that with static balance testing: participants with neuropathy generally demonstrated worse dynamic balance.^35,47,49,53 A novel finding in this group of studies was that compared with cancer survivors with normal weight, survivors who were overweight or obese demonstrated significantly worse dynamic balance.⁵⁵ Ness et al⁵³ reported impaired performance on the TUG test, indicating possible long-term effects of chemotherapy on dynamic balance.

Gait

Ten (33.3%) studies reported at least one measure of gait. Three of these studies were prospective with assessments during and after chemotherapy exposure,^27,45,52 whereas 7 were cross-sectional with assessment points ranging from within 1 to 6 years following completion of chemotherapy.^{3,35,41,43,54,58,59} Cancer diagnoses varied among these studies while chemotherapeutic agents included taxanes, platinums, vinca alkaloids, and anthracycline.

All 10 studies included a measurement of gait speed, though outcome measures varied (Table 4). In addition, a variety of technologies were employed to assess gait kinematics: GAITRite, Zeno pressure-sensitive walkway, and LEGSys.

The 3 prospective studies reported significant declines in gait speed as participants progressed through chemotherapy treatment. Results were more heterogeneous among the cross-sectional studies. Four of the 7 cross-sectional studies reported significant differences in gait speed,^3,35,54,59 whereas one study that reported no difference in gait speed between cancer survivors and healthy controls did report significant changes in step length and width variability.⁴³ Of the 2 remaining studies with nonsignificant gait speed findings, one reported no difference in fast or usual walking speed between participants with a history of falls and participants without a history of falls⁵⁸ and the other attributed the nonsignificant findings to above-average physical activity levels among participants.⁴¹ All 5 studies that measured gait kinematics reported significant differences, especially noted for survivors with CIPN, with changes seen acutely during chemotherapy and as long as 6 years beyond treatment.^{3,35,43,52,59}

Falls

Fifteen (50%) studies reported at least one outcome measure related to falls. Five of these studies were prospective, assessing fear of falling (FOF) and incidence of falls during and in the acute period following chemotherapy treatment.^{36,37,44,48,56} One study was a retrospective cohort study examining fall-related injuries during the first year following treatment with neurotoxic chemotherapy.⁶⁰ The remaining 9 studies were cross-sectional, with assessment points ranging from within 1 to 6 years following completion of chemotherapy.^† There was a wide variety of cancer diagnoses among the 15 studies; however, chemotherapeutic agents included only taxanes, platinums, or vinca alkaloids.

Among the 5 prospective studies, 2 reported that participants with higher cumulative doses of chemotherapy and symptoms of CIPN were more likely to have a fall (or near fall).^48,56 Interestingly, 2 prospective studies reported contradicting results regarding FOF, with one study reporting significantly increased FOF following chemotherapy³⁶ and one reporting significantly decreased FOF during chemotherapy.³⁷ The fifth prospective study found evidence of ongoing CIPN at 1-year follow-up, with 9.4% of participants reporting falls.⁴⁴ The retrospective study reported that the rate of fall-related injuries (hip fracture, head injury, joint dislocation, or other traumatic fracture code) within 1 year of a patient receiving his or her first dose of chemotherapy was significantly higher among patients who received 2 neurotoxic agents than among those who received one or none.⁶⁰

Among the 9 cross-sectional studies, outcome measures and results were heterogeneous. Two studies found that participants with cancer and/or CIPN were more fearful of falling following treatment with chemotherapy.^54,59 Six of the studies suggested that greater symptoms of CIPN increased the odds of having fallen in the previous 3 to 12 months,^{3,38,40,42,59} whereas one study went further to indicate that the vestibular score on the SOT mediated the relationship between treatment and falls.⁵⁸ One study related TUG test times of participants with CIPN to fall risk and found that participants with CIPN took significantly longer than healthy controls and on average scored above the cutoff for increased risk of falls.³⁵ Finally, one study assessed scores on the Activities-Specific Balance Confidence Scale and the Physiological Profile Assessment and found no significant differences between participants with and without CIPN and healthy controls.

Outcomes Through the Continuum

Twenty-six (86.7%) articles provided descriptions of either when the study assessment(s) took place in respect to chemotherapy exposure or how long participants had experienced symptoms related to chemotherapy (ie, CIPN). Three articles provided only time since diagnosis.^43,53,59 The author of the one article that did not provide temporal description was contacted without response.³⁵ On the basis of temporal descriptions, the 29 articles were organized into one of 4 time periods: acute, subacute, chronic, or long-term (Table 5).

TABLE 5 - Temporality of Studies

Time Point (n = 29)	Articles, n (%)
Acute (during, ≤3 mo following chemotherapy)	10 (34.5)
Subacute (3-12 mo following chemotherapy)	5 (17.2)
Chronic (1-5 y following chemotherapy)	8 (27.6)
Long term (>5 y following chemotherapy)	6 (20.7)

Among the 10 “acute” phase studies, there were 5 prospective studies, 4 cross-sectional studies, and 1 case report. All 10 articles reported concordant results during this time period, with 8 articles reporting decreased balance, 3 reporting decreased gait speed, and 3 reporting increased falls. Three articles reported evidence of withdrawal/alteration of visual input increasing imbalance, whereas 2 articles reported the effect of impaired cognition on balance. Three articles reported cumulative effects of chemotherapy on respective outcomes.

Among the 5 “subacute” phase studies, there were 3 cross-sectional studies, 1 retrospective cohort study, and 1 prospective study. Regarding this time period, 3 articles reported decreased balance, with one reporting correlation with CIPN severity. The retrospective cohort study reported a higher rate of fall-related injuries in participants who received 2 neurotoxic chemotherapies compared with one or none, whereas the remaining study described a natural history of CIPN and a 9.4% incidence of falls in the year following chemotherapy.

All 8 “chronic” phase studies were cross-sectional. Outcome measures and results were more heterogeneous among studies during this time period, with 6 articles reporting decreases in balance, 1 reporting gait dysfunction, 2 reporting a positive correlation of falls and neuropathy, and 1 reporting correlation of falls and the vestibular score of the SOT. Two articles reported on the effect of vision on balance or falls, whereas one study found that higher body mass index correlated with worse dynamic balance. One study found no significant findings but cited significantly above-average physical activity levels among participants.

All 6 “long-term” phase studies were cross-sectional. In respect to long-term survivors, 4 studies reported balance impairments, 2 reported increased falls associated with CIPN, and 2 reported gait dysfunction associated with CIPN.

Causal Pathways

Twenty (66.6%) articles reported results related to CIPN. Of those 20 articles, 14 related CIPN to balance, 4 related CIPN to gait dysfunction, and 7 related CIPN to falls. Five studies reported a correlation of CIPN severity with these variables, with 3 of those 5 reporting a correlation between increased symptom severity and decreased balance,^3,5,50 1 reporting a correlation between increased symptom severity and decreased gait speed,³ and 3 reporting a correlation between increased symptom severity and increased falls.^3,38,42

While many of the reviewed articles assessed balance with varying visual conditions, only 2 articles included specific visual outcome measures.^57,58 Wampler et al⁵⁷ reported that breast cancer survivors demonstrated significantly poorer low-contrast acuity following treatment with taxane-based chemotherapy, whereas Winters-Stone et al⁵⁸ reported that subjects who took longer to read letters on a contrast sensitivity chart were more likely to have fallen, though the article does not specify to what chemotherapeutic agent participants were exposed. While these 2 studies did not include dedicated objective measurements of vestibular function, they were the only articles to suggest possible contribution of vestibular impairment through highlighting performance deficits on the SOT consistent with vestibular dysfunction.

Two articles highlighted the potential influence of affective and/or cognitive domains on balance and gait.^45,51 Monfort et al⁵¹ reported that participants with combined peripheral sensory and cognitive impairments following treatment with oxaliplatin or taxane-based chemotherapy demonstrated worsened stability during dual-task walking. Klepin et al⁴⁵ reported that, although treatment with anthracycline did not worsen cognitive function or depression scores, participants with persistent depressive symptoms had a significant decline in SPPB score, whereas participants without symptoms did not, and participants with persistent cognitive impairment demonstrated greater decline in SPPB score than those without cognitive impairment. Only 4 articles reported or controlled for the potential influence of strength on balance and falls.^5,53,54,56 No articles reported on the relationship of fatigue and post-chemotherapy imbalance, gait dysfunction, or falls.

DISCUSSION

To our knowledge, this is the first scoping review of the effect of chemotherapy on balance, gait, and falls among cancer survivors. Despite heterogeneous methodologies, the current body of literature is largely in agreement that chemotherapy negatively impacts balance and gait and results in increased risk of falls. While this result is not entirely surprising, this review allowed us to draw specific conclusions based on established knowledge.

First, evaluation of balance, gait, and falls in the existing literature included a wide variety of objective measures. Many studies included the FAB scale, TUG test, and gait speed, which are all validated measures in this population and have been recommended for clinical use.⁶¹ Other studies reported outcomes that have yet to be validated in adult cancer survivors and therefore should be interpreted with caution. The overwhelming majority of studies reported negative effects of chemotherapy on static balance, dynamic balance, gait, and falls among cancer survivors. This finding emphasizes the importance of clinicians selecting appropriate outcome measures to screen each of these domains among individuals who have been exposed to chemotherapy. Clinicians are encouraged to consider the recent systematic review of balance outcome measures for adult cancer survivors.⁶¹

When examining the literature for temporal themes, we found evidence of the negative effects of chemotherapy on balance, gait, and falls beginning during chemotherapy, remaining present during the subacute and chronic phases, and in many cases lasting beyond 5 years of survivorship. Clinicians should be aware of this finding as they encounter patients throughout the survivorship continuum and attention should be given to individuals' chemotherapy exposures. Clinicians should also take note of the clear evidence linking CIPN to deficits in these functional outcomes throughout the continuum and should screen for sensory impairments and intervene appropriately according to best available evidence in their profession.^62,63

Knowledge Gaps and Implications for Further Research and Clinical Practice

The sex and oncologic diagnosis disparity in this body of research must be acknowledged first. More than 83% of all participants in this body of literature were female, which aligns with the disproportionate percentage of participants with breast cancer. While these percentages may be rationalized, given the prevalence and increased survival rate of breast cancer leading to a growing population of female participants with a history of breast cancer, we must recognize that males are underrepresented in this literature. In addition, this review highlights underrepresented oncologic diagnoses. For example, in our aggregated sample, colorectal cancer—despite often being treated with oxaliplatin (a known neurotoxic chemotherapy)—only accounted for 5.3% of participants. Excluding breast cancer, only 4 studies recruited dedicated samples of oncologic diagnoses (ovarian, AML, and childhood cancers). Increased attention to fewer sex-biased cancer diagnoses in future research may achieve 2 necessary end points: improving representation of currently underrepresented diagnoses as well as inherently increasing the number of males in the literature. This would enhance generalizability of results to all clinical populations.

Only one article stratified outcomes by chemotherapeutic agent, which presents another gap in the research: comparison of outcomes by chemotherapy. Disparities in the clinical characteristics and natural histories of neuropathic symptoms following exposure to different chemotherapies have been demonstrated.⁶⁴ If differences in balance, gait, and falls outcomes can be described according to chemotherapy, our ability to monitor and intervene appropriately would be enhanced. This suggests opportunity for future research including secondary analyses of existing studies where possible.

Considering study design, this review illustrates that the vast majority of studies in this body of literature are cross-sectional. A cross-sectional study is unable to demonstrate natural history and is easily challenged with the claim that a causal relationship cannot be drawn. This body of literature would benefit from development of controlled prospective studies to demonstrate causality and further elucidate the natural history of balance, gait, and falls during and after chemotherapy exposure. Despite robust evidence demonstrating the cumulative neurotoxic effects of chemotherapy,⁶⁵ only 3 studies in this review reported cumulative effects on function. Prospective studies with attention to chemotherapy dosing would assist in bolstering evidence in this area. Furthermore, organizing these studies temporally reveals a gap in studies during the subacute (3-12 months following chemotherapy) and long-term phases (>5 years). Prospective studies with longer-term follow-up would assist in filling these gaps.

Finally, this review illuminates the intense focus of the literature on the condition and causal pathway of CIPN. While the effect of impaired somatosensation on balance, gait, and falls cannot be understated, the paucity of research exploring other potential causal pathways must be acknowledged. Given some studies' implication of chemotherapy-induced visual and vestibular impairments, and with a full understanding of the necessary integration of sensory systems for postural control, these sensory systems deserve further investigation. Acknowledging recent evidence linking cancer-related fatigue, CIPN, cognition, and impaired function,^66,67 future research should also consider other potential causal pathways such as chemotherapy-induced fatigue and cognitive impairment.

Limitations

The results of this study should be interpreted with an understanding of its limitations. First, the filters that were applied to our searches limited our results to articles published in English since the year 2006. These filters may have resulted in language bias and/or incomplete search results due to missing relevant literature published in 2005 or earlier. Second, as this is a scoping review, we performed no evaluation of the quality of the studies included. For this reason, the reader should be cautious in interpreting the results we have presented and those of the individual articles.

CONCLUSIONS

This scoping review summarized the current body of literature related to the effect of chemotherapy on balance, gait, and falls among cancer survivors. The overwhelming majority of studies in this review report negative effects of chemotherapy on static balance, dynamic balance, gait, and falls, spanning the survivorship continuum. This review also identifies knowledge gaps including underrepresented participants and causal pathways other than CIPN. Finally, this review suggests opportunities to strengthen the current body of literature with controlled prospective studies, improved sex ratios, and consideration for additional causal pathways.