Support Care Cancer (2013) 21:847–856
DOI 10.1007/s00520-012-1591-8
ORIGINAL ARTICLE
The pediatric-modified total neuropathy score:
a reliable and valid measure of chemotherapy-induced
peripheral neuropathy in children with non-CNS cancers
Laura S. Gilchrist & Lynn Tanner
Received: 30 April 2012 / Accepted: 30 August 2012 / Published online: 20 September 2012
# Springer-Verlag 2012
Abstract
Background Neurotoxicity is a common side-effect of can-
cer treatment, but no scales have been validated for the
pediatric population. The objective of this study was to test
the reliability and validity of the pediatric modified-Total
Neuropathy Scale (ped-mTNS) to measure chemotherapy-
induced peripheral neuropathy in school-aged children.
Methods Forty-one subjects aged 5–18 years undergoing
chemotherapy with vincristine or cisplatin and 41 age- and
gender-matched controls completed study measures. Sub-
jects were tested with the ped-mTNS at a specified time
during treatment. Standardized measures of balance and
hand function were completed concurrently. Internal consis-
tency of the ped-mTNS was evaluated using Chronbach’s
alpha. Validity was tested by comparing case and control
ped-mTNS scores as well as testing the hypothesis that ped-
mTNS scores would be associated with scores on tests of
balance and manual dexterity. Inter-rater and test–retest
reliability were each assessed in a subset of 10 subjects.
Results Twenty-three subjects with acute lymphoblastic leu-
kemia, six with lymphoma, and 12 with solid tumors com-
pleted measures along with 41 age- and gender-matched
controls. Internal consistency was acceptable with a Chron-
bach’s alpha of 0.76. Children undergoing treatment for
cancer had significantly worse scores on the ped-mTNS
compared to controls (subjects, 8.7±4.2; controls, 1.4±0.9;
L. S. Gilchrist (*) : L. Tanner
Hematology and Oncology Program,
Children’s Hospitals and Clinics of Minnesota,
2545 Chicago Ave S, Suite 412,
Minneapolis, MN 55404, USA
e-mail: lsgilchrist@stkate.edu
L. S. Gilchrist
Physical Therapy Program, St. Catherine University,
601 25th Ave S,
Minneapolis, MN 55454, USA
p<0.001). As hypothesized, scores on the ped-mTNS were
associated with measures of balance and manual dexterity.
Inter-rater and test–retest reliability was acceptable (intraclass
correlation coefficients >0.9 each).
Conclusions The ped-mTNS is a reliable and valid measure
of chemotherapy-induced peripheral neuropathy in school-
aged children that is associated with relevant functional
limitations.
Keywords Cancer . Neuropathy . Chemotherapy . Pediatrics
Introduction
Many pediatric cancer diagnoses now have 5-year survival
rates of over 85 % due to the use of multimodal treatment
that often includes chemotherapy [1]. Unfortunately, a num-
ber of chemotherapeutic medications have potential long-
term effects, including neurotoxicity. In the peripheral nerv-
ous system (PNS), this toxicity often manifests as
chemotherapy-induced peripheral neuropathy (CIPN); while
in the central nervous system, toxicities range from cogni-
tive and intellectual impairments to encephalopathy and
coma. Symptoms of neurotoxicity in the PNS are often
under-recognized and undiagnosed in adult patients [2].
Even less is known about the prevalence and impact of
CIPN in pediatrics. One major reason for sparse data has
been the lack of specific and sensitive measurement tools for
CIPN in the pediatric population.
The Total Neuropathy Scale (TNS) was developed as a
clinical assessment of peripheral neuropathy in adults and
has demonstrated reliability and validity in measuring CIPN
in adult cancer patients [3–5]. A modified version of the
TNS, the modified-Total Neuropathy Scale (mTNS) [6], was
developed to determine the severity of CIPN in adult women
with breast cancer. It was adapted from the original TNS to
848 Support Care Cancer (2013) 21:847–856

make it more feasible for use in the rehabilitation or oncology
clinic settings. This study was designed to determine if a
similar measure could be used to detect CIPN and its severity
in children during treatment for cancer. There is no evidence
that indicates that CIPN is a different entity in children than in
adults. Similar to adults undergoing cancer-related chemo-
therapy, children treated with chemotherapy often report
symptoms of neuropathy, and demonstrate balance and gait
disorders [7, 8]. Thus, the clinical components of the mTNS
should be applicable for and meaningful to a pediatric pop-
ulation if appropriately modified.
Our initial work modifying the mTNS to create the
pediatric modified-Total Neuropathy Scale (ped-mTNS)
has been published previously [9]. The ped-mTNS is com-
posed of three sets of questions on sensory symptoms and
pain, motor function, and autonomic function, and a five-
part neurologic exam that includes light touch testing, vibra-
tion sensation, pin sensation, distal strength assessment by
manual muscle test, and deep tendon reflexes so that it
assesses multiple components of the PNS (Table 1). Each
category is rated from 0 to 4, with 0 indicating no symptom
or deficit on clinical exam to 4 indicating worse symptoms
or proximal spread of neurologic symptoms (Appendix A).
The scale has 32 possible points and a higher score indicates
worse neuropathy.
The purpose of this study was to investigate the psycho-
metric properties and validity of the ped-mTNS as a meas-
ure of CIPN in school-aged children and adolescents. We
examined internal consistency as well as test–retest and
inter-rater reliability of the ped-mTNS. Construct validity
was assessed by testing the hypothesis that the ped-mTNS
would distinguish between a group of children who have
received known neurotoxic chemotherapy and a group of
age- and gender-matched controls. Additionally, we tested
the hypothesis that patients with worse neuropathy (higher
scores on the ped-mTNS) would have worse function on
specific measures of physical function for which distal sensa-
tion and motor control are needed (lower scores on measures
of balance and manual dexterity).
Methods
Subject recruitment
Patients with a new diagnosis of acute lymphoblastic leuke-
mia (ALL), lymphoma, or other solid tumor except for brain
tumors, who met the inclusion/exclusion criteria (Table 2)
attending visits at a children’s oncology clinic, were
approached for participation in this study during outpatient
appointments. All recruitment material, consent forms, and
testing procedures were reviewed by the institutional review
board at Children’s Hospitals and Clinics of Minnesota.
In total, 83 patients met invitation criteria, 4 declined
participation, and 13 were not approached for recruitment
due to scheduling issues. Of the 66 subjects recruited, 60
completed measures. Of the 60 children completing measures,
we obtained age- and gender-matched controls for 41 subjects,
and thus this group is included in this study. This sample size
was adequate for hypothesis testing based on pilot study
results. Controls were recruited from siblings attending clinic
visits (n035) as well as children of clinic staff (n06).

Table 1 Components of the

pediatric-modified Total Neuro- Scale item Testing method Primary nerve fibers and
pathy Scale receptor types examined

Sensory symptoms Scripted interview Sensory fibers—multiple

types
Motor symptoms Scripted interview Motor fibers—somatic
voluntary control
Autonomic symptoms Scripted interview Autonomic fibers—sensory
and motor
Light touch sensation Light touch threshold using Sensory—Aβ fibers,
Semmes–Weinstein monofilaments Meissner’s corpuscles
Pin sensation exam Sharp/dull sensation using MediPin Sensory—C and Aδ fibers,
free nerve endings
Vibration exam Vibration perception threshold Sensory—Aβ fibers,
using biothesiometer Pacinian corpuscles
Strength Manual muscle tests of finger Motor—alpha motor neurons
abduction, wrist extension, great
toe extension, ankle dorsiflexion
Deep tendon reflexes Deep tendon reflexes of the Sensory—muscle spindle
achilles and patellar tendons receptors, type Ia and II fibers
Motor—alpha and gamma
Support Care Cancer (2013) 21:847–856
Table 2 Inclusion and exclusion criteria
Patient inclusion • Child between the ages of 5 and 18
criteria • New diagnosis of leukemia, lymphoma
(Hodgkins or non-Hodgkins), or solid tumor
(Wilm’s, rhabdomyosarcoma, neuroblastoma)
• On a chemotherapy protocol utilizing
vincristine or cisplatin
• English-speaking subject and guardian
• Deemed appropriate to participate in the study
by their physician, nurse, or physical therapist
Patient exclusion • Diagnosis of relapsed or second cancer
criteria • CNS tumors
• Other developmental disorders (such as: Down’s
syndrome, other chromosomal disorders)
• Other neuromuscular disorders (such as:
traumatic brain injury, cerebral palsy)
• Non-English-speaking subject or guardian
• Upper or lower extremity amputations
Control • Child age 5–18 without a cancer history
inclusion • English-speaking subject and guardian
criteria
• Consent of the parent and assent of the child
Control • Previous diagnosis of cancer (except skin
exclusion cancers with no chemotherapy treatment)
criteria • Previous treatment with chemotherapy
for any condition
• Non-English-speaking subject or guardian
• Developmental disorders (Down’s syndrome,
chromosomal disorders, etc.)
• Neuromuscular disorders (traumatic
brain injury, cerebral palsy, etc.)
• Upper or lower extremity amputation
or limb deficiency
Demographic and treatment data
Parents of cases and controls filled out a demographic
questionnaire regarding age, sex, racial background, and
past medical history. Parents of controls also indicated
current medications. Information regarding cancer diag-
nosis and treatment was abstracted from the medical
record by a trained clinical research associate. Informa-
tion abstracted included: diagnosis and staging, surgical
interventions, cumulative dose of each chemotherapeutic
agent used, radiation treatments number and cumulative dose,
duration of cancer treatments, and comorbid conditions.
Timing of measures
Measures were completed for children undergoing can-
cer treatment a minimum of 2 months after the initiation
of neurotoxic chemotherapy (n 041). Specifically, in
children with ALL, patients were measured within
849
2 weeks of the end of delayed intensification. In chil-
dren with lymphoma and solid tumors, patients were
measured 2–3 months after the initiation of induction
therapy.
Validity
Both cases and controls underwent the same testing proce-
dures. On the date of testing, a physical therapist adminis-
tered the ped-mTNS followed by the balance and manual
dexterity subscales of the Bruininks–Osteretsky Test of Motor
Proficiency version 2 (BOT-2; NCS Pearson; Minneapolis,
MN, USA) [10].
Ped-mTNS
The sensory, motor function, and autonomic symptom
questions were read aloud to all participants (Appendix
A). If the child answered yes to any symptom question, they
were asked to rate the symptom according to the scales in
Appendix A.
Light touch sensation was assessed using Semmes–
Weinstein monofilaments (Rolyan–Ability One, German-
town, WI, USA). For the fingers, the 2.83 log force
filament was presented on the pad of the index finger
with the child’s eyes closed. The subject was asked to
respond each time they felt the stimulus, and the stim-
ulus was repeated three times. If the participant was
able to correctly identify the three stimuli sequentially,
smaller monofilaments were used until the subject no
longer was able to perceive all stimuli. If the subject
was unable to identify three stimuli of the first mono-
filaments, monofilaments of increasing size were used
until the subject was able to correctly identify all three
stimuli. The lowest level of stimulus that the child was
able to identify correctly was recorded as the light touch
sensory threshold. This procedure was completed for
index fingers of both hands, and then great toe pad of
both feet (starting with the 3.61 log force monofilament)
with repetition in more proximal segments (wrist/ankle,
elbow/knee) if thresholds were above published values
[11]. Scores higher than published normal values indi-
cated a sensory deficit.
Pin sensibility was tested using a Medipin™ (Medipin
Ltd, Hertfordshire, UK). Testing was demonstrated on the
participant’s forearm and then initiated on the distal aspect
of the extremities (palmar pads of the fingers and plantar
pads of the toes) with the patient’s eyes closed. Four stimuli
were given on each extremity randomly alternating between
the “pointy” and “not pointy” ends and the child was
instructed to differentiate between the two stimuli. The test
was scored as abnormal if the child incorrectly identified
even one stimulus, and the next proximal segment was then
850
tested until the participant obtained 100 % accuracy. Any
area not receiving a score of 100 % was noted to have a
deficit.
Vibratory thresholds were assessed using a handheld
Biothesiometer (Biomedical Instruments, Newbery OH,
USA). The stimulator ending vibrates at 100 Hz with
amplitude proportional to the square of the applied
voltage. The voltage was started at 0 and slowly
increased until the child indicated s/he felt the stimulus
(“vibration or buzz”). An introductory trial was con-
ducted at the wrist before continuing with trials at the
distal plantar surface of the great toe and distal palmar
pad of the index finger. If the reading was >8 V for the
fingers and/or >10 V for the toes, further trials were
conducted at the medial malleolus/radial wrist and
medial knee/medial elbow until a normal reading was
obtained. The normal values were adapted from a com-
bination of data from Hilz et al. [12] and Coutinho dos
Santos et al. [13] based on data from control subjects in
pediatric diabetic neuropathy studies and used to determine
deficits.
Muscle strength of the great toe extensors, ankle
dorsiflexors, finger abductors, and wrist extensors was
assessed using manual muscle tests in all children
according to Medical Research Council guidelines and
graded accordingly [14]. A deficit was noted if the
examined muscle did not receive a grade 5 (normal)
muscle test.
Deep tendon reflexes (DTR) were assessed with the
participant seated with lower extremities free swinging.
Achilles and patellar tendon reflexes were elicited with
the use of Jendrassic maneuver if no reflex could be
elicited. DTRs were graded according to American
Academy of Neurology definitions with a deficit indi-
cated if reflexes were not grade 2+ [15]. Upper extrem-
ity reflexes were not tested, as no distal reflex is
standard for the upper extremity, and intravenous lines were
in place in some of the children.
Functional measures (BOT-2)
The BOT-2 is designed to assess motor proficiency in
children with mild to moderate functional deficits. Inter-
rater reliability, test–retest reliability, a standard error of
measurement, and intercorrelations are established for
each subtest and each age category in a group of
1,520 subjects [16]. Points are tallied for each subtest
and converted to age- and gender-based scores [10],
with a standardized score mean of 15 and standard
deviation of 5. Balance and manual dexterity subtests
each have inter-rater reliability above 0.9 [16]. The
Support Care Cancer (2013) 21:847–856
BOT-2 was administered according to published stand-
ardized protocols [10].
Reliability
During the one study visit, 10 children were retested on the
ped-mTNS, by the same investigator to assess test–retest
reliability. An additional 10 children were re-assessed on the
ped-mTNS by another trained physical therapist researcher
to investigate inter-rater reliability. Test–retest or inter-rater
reliability was conducted only on 10 subjects each due to
scheduling constraints during visits to the oncology
clinic. This number was chosen to yield adequate power
for analysis with minimal disruption to the clinic. For
both of these groups, at least 1 h between measures was
given to lessen the impact of subject recall. In addition,
it was required that no new chemotherapy be administered
in this interval to decrease the chance of new symptoms
developing.
Analysis
The internal consistency was assessed using Chronbach’s
alpha and item–total score correlations. A Chronbach’s
alpha of >0.7 was set as the threshold for internal consis-
tency and the item–total correlation threshold for accept-
ability of individual items was set at ≥0.3 according to
established criteria [17]. Ceiling and floor effects were
examined by assessing the number of cases receiving the
lowest and highest scores. Test–retest and inter-rater reli-
abilities were assessed using intraclass correlation coefficients
(ICC, model 3), with a correlation of 0.9 considered to be
acceptable for clinical measures [17].
Construct validity of the ped-mTNS was evaluated by
comparing the case and control groups using an independent
sample t test. Spearman correlations were used to evaluate
the hypothesis that neuropathy scores on the ped-mTNS
would be associated with functional measures of balance
and manual dexterity. Differences between groups on indi-
vidual items within the ped-mTNS were evaluated using
multivariate analysis of variance. All statistics were assessed
using SPSS version 15.
Results
Twenty-three subjects with ALL, six with lymphoma, and
12 with solid tumors completed the measures along with 41
age- and gender-matched controls. No significant differen-
ces were found between groups in height or weight (Table 3).
Subjects ranged from 5 to 18 years in age and had been in
Support Care Cancer (2013) 21:847–856 851

Table 3 Population characteristics Table 4 Internal Consistency of the ped-mTNS

Cases (n041) Controls (n041) Ped-mTNS item Corrected item–total Chronbach’s alpha
correlation if item deleted
Age (years) 9.56±3.1 9.61±3.1
Sex (% male) 36.6 36.6 Sensory symptoms 0.350 0.754
Height (cm) 139.3±19.5 143.4±19.2 Motor function 0.507 0.728
Weight (kg) 36.6±13.8 36.6±12.0 Autonomic symptoms 0.311 0.758
Cancer diagnosis Light touch 0.602 0.704
ALL 23 (56 %) NA Pin sensibility 0.325 0.756
Lymphoma 6 (15 %) Vibration sensation 0.393 0.746
Other solid tumor 12 (29 %) Strength 0.525 0.721
Wilms’ tumor 6 Deep tendon reflexes 0.694 0.682
Ewings sarcoma 3
Other (ovarian, liver, muscle) 3
Treatment factors: chemotherapy exposurea
Validity
Vincristine 40 (96 %) NA
Cisplatin 2 (5 %)
Children undergoing treatment for cancer had signifi-
Radiation treatment:
cantly worse scores on the ped-mTNS compared to
Cranial radiation 2 (5 %)
controls (subjects, 8.7±4.2; range, 2–18; controls, 1.4±
Other radiation (Chest wall, 5 (12 %)
abdomen, or pelvis) 0.9; range, 0–4; p<0.001; Fig. 1). For the total scale
score, 0 % of the subjects but 9.8 % of the controls
a
One participant received both vincristine and cisplatin received the lowest score of 0. None of the subjects or
ALL acute lymphoblastic leukemia controls received the highest score of 32 points possible.
Significant differences between subjects and controls were
found for all individual scale items except autonomic symp-
treatment for an average of 5.3±2.4 months, with a range of
toms and pin sensibility (Table 5). For the subjects receiving
2–10 months of treatment at the time of examination. All but
one subject were exposed to vincristine (mean cumulative
dose, 16.6 mg/m2; range, 3.0–28.5 mg/m2), one patient
received cisplatin (cumulative dose, 360 mg/m2), and one 20 Controls
subject received both viscristine and cisplatin (cumulative Cancer
doses of vincristine (3.0 mg/m2) and cisplatin (360 mg/m2)).
A small percentage of children undergoing treatment had
received radiation to the cranium, chest wall, abdomen, or 15
pelvis (Table 3).

Reliability
Frequency

10
The ped-mTNS demonstrated acceptable internal consis-
tency with no items scoring less than 0.3 on the cor-
rected item–total correlation and an overall Chronbach’s
alpha of 0.76 (Table 4) Removal of any item did not
improve the Chronbach’s alpha score. Test–retest was 5

conducted on children ranging in age 6–18 years (n0

10) with a mean age of 10.2 ± 3.7 years. Test–retest
reliability of the ped-mTNS score was acceptable with
an ICC00.99 (95 % CI 0.96–0.99). Inter-rater reliability 0
was conducted on children ranging in age from 5 to 0.00 5.00 10.00 15.00 20.00

18 years (n010) with a mean age of 9.8±4.4 years. Total pedmTNS Score

Inter-rater reliability of the total scale score was accept- Fig. 1 Histogram of ped-mTNS scores for subjects being treated for
able with an ICC00.98 (95 % CI 0.95–0.99). cancer and controls
852 Support Care Cancer (2013) 21:847–856

Table 5 Comparison of ped-mTNS item scores (range for each item 0–4) balance and manual dexterity. Thus, the ped-mTNS
Ped-mTNS item Cases (n041) Controls (n041) meets many of the criteria needed for an acceptable
clinical outcome measure for CIPN. Because it does
% with Mean % with Mean not require excessive equipment, specialized practi-
deficit deficit tioners, or excessive time (less than 10 min in our pilot
study), it is a useful instrument for obtaining subjective
Subjective symptoms
symptoms and objective measures that indicate periph-
Sensory symptoms 27 0.39* 2 0.02
eral neurotoxicity across multiple nerve fiber types.
Motor function 49 0.78** 10 0.10
Although the ped-mTNS has some differences from
Autonomic 37 0.73 39 0.46
symptoms other versions of the TNS, this study demonstrates that
Clinical examination children undergoing potentially neurotoxic chemotherapy
Light touch 44 1.12** 5 0.05 have comparable levels of neuropathy to adults under-
Pin sensation 46 0.51 32 0.32 going treatment for cancer. For example, Wampler et al.
Vibration sensation 37 0.73* 2 0.10 [6] reported that women treated with taxanes for breast
Distal strength 98 1.88** 39 0.39 cancer had a mean score of 7 ± 3 the mTNS (score
Deep tendon reflexes 100 2.60** 0 0.00 range, 0–24) and a mean score of 7 ± 3 on the full
TNS scale (score range, 0–32). These values are com-
*p<0.05 parable to the mean score of 8.7±4.2 that we found in
**p<0.001 the children undergoing cancer treatment with vincris-
tine or cisplatin. Our pediatric patients had much lower
scores than those reported by Cornblath et al. [3] in
vincristine, there was no significant correlation between patients with severe diabetic neuropathy (25±7) on the
cumulative dose and neuropathy score. There was also no full version of the TNS. Similar to the findings of
significant difference in ped-mTNS scores between children Wampler et al. [6], even though the neuropathy in these
receiving (n07) and not receiving (n035) radiation treatment children was relatively mild, it was associated with
(radiation, 7.4±4.3; no radiation, 8.9±4.1; p00.39). functional deficits in balance and manual dexterity, sug-
In terms of motor performance, children treated for gesting clinical importance.
cancer scored significantly worse on both scales of Few other measures of CIPN for use in children are
balance (cases (7.8 ± 3.2) vs. controls (14.0± 4.1); p < available. Most studies of the neurotoxic impact of chemo-
0.001) and manual dexterity (cases (10.7±4.1) vs. controls therapy in children have used the National Cancer Institute
(15.5±4.2); p<0.001). Measures of neuropathy demonstrated Common Terminology Criteria for Adverse Events
significant negative associations with balance (rs 0−0.626, p< (CTCAE) [18] as a measure of CIPN. The CTCAE are
0.001) and manual dexterity (r s 0−0.461, p < 0.001) as generally applied as a chart review of the oncologist’s clin-
hypothesized. No significant association was found between ical findings during routine clinical care. It is an extensive
cumulative dose of vincristine and balance or manual dexter- list of potential complications across all major body systems
ity score. and has severity ratings of 1–5 for each potential event. In
the CTCAE, sensory and motor peripheral neuropathies are
rated separately and pain from neuropathy could be rated
Discussion under many different categories. In a key study completed
by Postma et al. [19], pitfalls of the CTCAE were high-
The ped-mTNS is the first measure designed to assess lighted. When two independent neurologists used a version
CIPN in school-aged children. It demonstrates both of the CTCAE to rate the same patients, inter-observer
acceptable reliability and internal consistency. In agreement was only 46 %. While the criteria have been
school-aged children, ped-mTNS scores are signifi- refined over the past 10 years, the CTCAE remains a
cantly higher for children undergoing neurotoxic che- method to broadly look for the impact of cancer treatment,
motherapy as compared to controls. The scale does not but is limited in its ability to specifically measure CIPN and
suffer from ceiling or floor effects in the cancer pop- accurately measure change over time. Thus, specific and
ulation and also displays an ample range of scores that sensitive rating scales for CIPN are needed that have been
may allow it to be more sensitive to change than validated in the pediatric cancer population.
currently used scales. As hypothesized, ped-mTNS While we have demonstrated good reliability for the
scores correlate well with functional measures of ped-mTNS, it must be noted that both testers were
Support Care Cancer (2013) 21:847–856
physical therapists trained in neurological examination.
Whether or not the ped-mTNS is reliable between dis-
ciplines, such as nurses and physical therapists has yet
to be determined. This study does demonstrate that
these examinations can be done by non-neurologist
healthcare professionals trained in neurologic assessment
techniques. Similarly, other versions of the TNS have
been used successfully by nurses in adult populations
[20] making the case that multiple professionals can
conduct these types of assessments.
The validity of the ped-mTNS was tested in this
study by comparing values between known groups and
also by comparing its associations to relevant functional
limitations. We were unable to perform formal neuro-
physiological testing or concurrent testing with other
CIPN rating scales, measures that would have added
to the validity testing of this measure. Previous studies
of different versions of the TNS have been compared to
other measures of peripheral neuropathy in adults such
as the neuropathy symptom score, the neurologic
impairment score, and the Michigan diabetic neuropathy
score and found good correlations among the TNS
variants and these other measures of peripheral neuro-
pathy [3, 6]. Survivors of childhood ALL have previ-
ously been reported to display neuropathy using a
reduced version of the TNS, the Neuropathy Impairment
Scale, and nerve conduction studies [21] lending credi-
bility to the argument that children being treated for
non-CNS cancers with vincristine or cisplatin will dis-
play CIPN. Concurrent testing with CIPN rating scales
validated in the pediatric population could not be com-
pleted for this study as no other scales were available at
the time of study design.
Each of the individual items in the ped-mTNS dem-
onstrated fair to good levels of item–total score correla-
tion indicating that all items contribute to the total
score. Removal of any item would not improve the
internal consistency of the scale and the Chronbach’s
alpha of 0.76 demonstrated acceptable internal consis-
tency [22]. In examining the individual items in the
scale, two items (autonomic symptoms and pin sensitiv-
ity) did not differ significantly between cancer patients
and controls. There was consideration to delete these
items, but was decided against for theoretical consider-
ations. In looking at the items in the scale (Table 1), no
other item except for autonomic symptoms tests the
autonomic portion of the nervous system, and thus this
item is critical to survey all portions of the PNS. Like-
wise, pin sensation is the only scale item that tests for
small fiber (C and Aδ) sensory neuropathy that may be
important in the development of neuropathic pain, and
853
thus we chose to retain the item. Because it is important
that this measure be able to detect neuropathy in multi-
ple fiber types, some of which may not demonstrate
significant damage with routine vincristine therapy, we
are electing to retain all items at this time. This study
also included only those patients undergoing initial
treatment for non-CNS tumors. Patients undergoing
treatment with higher levels of neurotoxic potential,
such as patients with relapsed disease or requiring bone
marrow transplantation may indeed have positive find-
ing with these items if tested in the future.
The current study gives initial evidence on reliability
and validity of the ped-mTNS as a measure of CIPN in
children with non-CNS tumors. Because it was not
feasible to test children prior to initiation of chemo-
therapy, we could not use cases as their own controls.
Thus, we matched age and gender in order to limit
confounding factors. Height and weight was also similar
between groups, also limiting the effect of patient size
on the data. Some of the subjects did receive radiation
treatment prior to their study measures that could have
lead to radiation-induced damage of the PNS or CNS.
In this small subgroup (n07), scores on the ped-mTNS
were not significantly worse and on average had lower
ped-mTNS scores than the non-irradiated group indicat-
ing that radiation damage was not a major confounding
factor in this study. We did not conduct a longitudinal
assessment as a part of this study, and thus no informa-
tion on sensitivity of the measure to change over time
can be implied.
Measurements such as the ped-mTNS will give
physicians, nurses, and rehabilitation personnel a tool
to measure the impact of chemotherapy on the PNS.
As more studies are being conducted to reduce symp-
tom burden and improve function and quality of life in
children treated for cancer, these types of specific meas-
ures will be especially important endpoints for clinical
trials. Further work will need to be completed to inves-
tigate the ped-mTNS’s sensitivity to change, relation to
multimodal chemotherapy exposure, and the clinical sig-
nificance of scores.
Acknowledgments The following individuals were instrumental in
the success of this research: MC Hooke RN, PhD; A. Logelin, CRA; K.
Ites, PT; and K. Ness PT, PhD. In addition, we would like to thank the
patients, families, physicians, and nurses of the hematology/oncology
clinic at Children’s Hospitals and Clinics of Minnesota.
Funding source Pine Tree Foundation, St. Paul MN
Disclosure statement The authors declare the absence of any con-
flicts of interest in this publication.
854 Support Care Cancer (2013) 21:847–856

Appendix A

Pediatric-modified total neuropathy scale

Sensory Symptoms: ______ (record worst score for the three sensations)
“ Do you have any parts of your body that are tingly, numb (can hardly feel), or hurt?”
______ Tingly ______ Numb ______ Hurt (record number for each)
If yes, “Where you have those feelings?”
0 None
1 Symptoms limited to fingers or toes
2 Symptoms extend to ankles or wrists
3 Symptoms extend to knee or elbow
4 Symptoms above knee or elbow

Functional Symptoms: ______ (record worst score of the three questions)

“Do you have trouble buttoning shirts or zipping zippers?” ______
“Do you have trouble walking such as tripping frequently?” ________
“Do you have trouble going up or down stairs?” _______
If yes to any, “Is it…..(read choices)” and record after each question
0 Not Difficult
1 A little difficult
2 Somewhat difficulty
3 I need help
4 I can’t do that at all

Autonomic Symptoms: _____ (record worst score of the three questions)

“Do you feel dizzy or light-headed when you get up out of bed?”______
“Do your hands or feet feel hotter or colder than normal?” ______
0 Never
1 A little bit
2 Sometimes
3 Very much
4 Almost always

Clinical Testing :
Light Touch Sensation: ______

0 Normal Semmes Semmes

1 Reduced in fingers/toes Toes R Finger R
2 Reduced up to L L
wrist/ankle Med Mal R Wrist R
3 Reduced up to L L
elbow/knee Knee R Elb R
L L
4 Reduce to above
elbow/knee

Pin Sensibility: _____

0 Normal
1 Reduced in fingers/toes
2 Reduced up to wrist/ankle
3 Reduced up to elbow/knee
4 Reduce to above elbow/knee
Support Care Cancer (2013) 21:847–856 855

Vibration Sensibility: _______ (worst score)

0 Normal
Bioesth Bioes
1 Reduced in
Toes R / Finger R /
fingers/toes
L / L /
2 Reduced up to
Med Mal R / Wrist R /
wrist/ankl L / L /
3 Reduced up to Knee R / Elb R /
elbow/knee L / L /
4 Reduce to above
elbow/knee

Strength: _____ Worst Score (MRC Score R / L)

MRC level: Great Toe ___/___ ankle DF___/___ finger abd___/___ wrist ext___/___
0 Normal
1 Mild weakness (MRC 4)
2 Moderate weakness (MRC 3)
3 Severe weakness (MRC 2)
4 Paralysis (MRC 1-0)

DTR: _____ (Achilles, Patellar)

0 Normal
1 Ankle reflex reduced (Achilles +1)
2 Ankle reflex absent (Achilles 0, Patellar +2)
3 Ankle reflex absent, others reduced (Achilles 0, Patellar +1)
4 All reflexes absent (all 0)

Total Score: / 32

References 7. Wright MJ, Galea V, Barr RD (2005) Proficiency of balance in

1. Ries LAG, Smith MA, Gurney JG, et al. (eds) (1999) Cancer
incidence and survival among children and adolescents: United
States SEER Program 1975–1995, National Cancer Institute,
SEER Program. NIH Pub. No. 99–4649. Bethesda, MD
2. Markman M (2006) Chemotherapy-induced peripheral neuropa-
thy: underreported and underappreciated. Curr Pain Headache Rep
10:275–278
3. Cornblath DR, Chaudhry V, Carter MS et al (1999) Total Neuro-
pathy Scale: validation and reliability study. Neurology 53:1660–
1664
4. Cavaletti G, Bogliun G, Marzorati L et al (2003) Grading of
chemotherapy-induced peripheral neurotoxicity using the Total
Neuropathy Scale. Neurology 61:1297–1300
5. Cavaletti G, Jann S, Pace A et al (2006) Multi-center assessment of
the Total Neuropathy Scale for chemotherapy-induced peripheral
neurotoxicity. J Periph Nerv Syst 11:135–141
6. Wampler MA, Miaskowski C, Hamel K et al (2006) The modified
total neuropathy score: a clinically feasible and valid measure of
taxane-induced peripheral neuropathy in women with breast can-
cer. J Support Oncol 4:9–16
children and youth who have had acute lymphoblastic leukemia.
Phys Ther 85:782–790
8. Hartman A, van den Bos C, Stijnen T, Pieters R (2009) Decrease in
motor performance in children with cancer is independent of the
cumulative dose of vincristine. Cancer 106:1395–1401
9. Gilchrist LS, Tanner L, Hooke MC (2009) Measuring chemotherapy-
induced peripheral neuropathy in children: development of the Ped-
mTNS and pilot study results. Rehabil Oncol 27:7–15
10. Bruininks RH, Bruininks BD (2005) BOT-2: Bruininks–Oseretsky
test of motor proficiency, 2nd edn. Pearson Assessments, Minneapolis
11. Bell JA (1984) Semmes–Weinstein monofilament testing for deter-
mining cutaneous light touch/deep pressure sensation. Star 44:8–11
12. Hilz MJ, Axelrod FB, Hermann K, Haertl U et al (1998) Norma-
tive values of vibratory perception in 530 children, juveniles, and
adults aged 3–79 years. J Neurolog Sci 159:219–225
13. Coutinho dos Santos LH, Bruck I, Antoniuk SA, Sandrini R (2002)
Evaluation of sensorimotor polyneuropathy in children and ado-
lescents with type I diabetes: association with microalbuminuria
and retinopathy. Pediatric Diabetes 3:101–108
14. Medical Research Council of the United Kingdom (1978) Aids to
examination of the peripheral nervous system: memorandum no
45. Pedragon House, Palo Alto
856
15. Dick JPR (2003) The deep tendon and the abdominal reflexes. J
Neurol Neurosurg Psychaitry 74:150–153
16. Dietz JC, Kartin D, Kopp K (2007) Review of the Bruininks–
Oseretsky test of motor proficiency, second edition (BOT-2). Phys
Occup Ther Pediatr 27:87–102
17. Streiner DL, Norman GR (2008) Health measurement scales: a
practical guide to their development and use, 4th edn. Oxford
University Press, New York
18. DCTD, NCI, NIH, DHHS (2010) Cancer Therapy evaluation pro-
gram, common terminology criteria for adverse events, Version
4.03. http://evs.nci.nih.gov/ftp1/CTCAE/CTCAE_4.03_2010-06-
14_QuickReference_5x7.pdf. Published June 14, 2010. Accessed
23 January 2011
Support Care Cancer (2013) 21:847–856
19. Postma TJ, Heimans JJ, Muller MJ et al (1998) Pitfalls in grading
severity of chemotherapy-induced peripheral neuropathy. Ann
Oncol 9:739–744
20. Lavoie Smith E, Cohen JA, Pett MA, Beck SL (2010) The reli-
ability and validity of a modified total neuropathy score-reduced
and neuropathic pain severity items when used to measure
chemotherapy-induced peripheral neuropathy in patients receiving
taxanes and platinums. Cancer Nurs 33:173–183
21. Ramchandren S, Leonard M, Mody RJ et al (2009) Peripheral
neuropathy in survivors of childhood acute lymphoblastic leuke-
mia. J Periph Nerv Syst 14:184–189
22. DeVellis RF (2003) Scale development: theory and application,
2nd edn. SAGE, Thousand Oaks