IMPACT OF LYMPHEDEMA AND ARM SYMPTOMS ON QUALITY …

1

Lymphology 43 (2010) 1-13

IMPACT OF LYMPHEDEMA AND ARM SYMPTOMS ON QUALITY OF LIFE IN BREAST CANCER SURVIVORS

J.M. Hormes, C. Bryan, L.A. Lytle, C.R. Gross, R.L. Ahmed, A.B. Troxel, K.H. Schmitz

Department of Psychology (JMH) and School of Nursing (CB), University of Pennsylvania; Divisionof Epidemiology and Community Health (LAL,RLA), College of Pharmacy and School of NursingCRG), Department of Dermatology (RLA), University of Minnesota, Minneapolis, Minnesota; andCenter for Clinical Epidemiology and Biostatistics (KHS), Department of Biostatistics andEpidemiology, University of Pennsylvania, Philadelphia, Pennsylvania, USA

ABSTRACT

Lymphedema is one of many armproblems reported by breast cancer survivors.Understanding the impact of lymphedema onquality of life requires consideration that armsymptoms may occur with or withoutlymphedema. It was hypothesized that specificarm symptoms and pain, related or unrelatedto lymphedema, would be more associatedwith quality of life outcomes than armswelling. The relation of arm swelling and ofarm symptoms and associated severity with arange of quality of life outcomes followingbreast cancer treatment was assessed in adiverse sample of 295 women, 141 of whomhad a clinical diagnosis of lymphedema. Armswelling (as defined by interlimb volume orcircumference differences) and lymphedemaseverity (defined by Common ToxicityCriteria) were less correlated with quality oflife than total number of arm symptoms andspecific individual symptoms. Pain in theaffected arm correlated with poor quality oflife outcomes, regardless of arm swelling.When evaluating the impact of lymphedemaon quality of life, arm swelling may not be asimportant as the total number and specifictypes of arm symptoms present, as these may

be more informative about quality of lifeoutcomes in survivors of breast cancer withand without lymphedema.

Keywords: lymphedema, breast cancer,quality of life, pain, arm symptoms

Lymphedema is a chronic lymphaticcondition that consists of interstitialaccumulation of protein-rich fluid andsubsequent inflammation, adipose tissuehypertrophy, and fibrosis, resulting inswelling, disfigurement, heightened risk ofinfection, and decreased mobility andfunction (1,2). Lymphedema of the ipsilateralupper extremity is a common side-effect inbreast cancer patients undergoing radiationand surgical resection of axillary lymphnodes. It arises in approximately 15-30% ofindividuals treated for breast cancer, withlower incidence following sentinel nodebiopsy (3,4). About 60% of patientsundergoing radical mastectomy developlymphedema, compared to approximately30% of women treated with modified radicalmastectomies or breast-conserving surgery(2). Onset varies but a majority of patientswho develop arm edema do so within fouryears following treatment (5); an estimated

Permission granted for single print for individual use. Reproduction not permitted without permission of Journal LYMPHOLOGY

2

38% of survivors struggle with arm problemsat five years post- treatment (6).

Studies of long-term, recurrence-freebreast cancer survivors suggest no significantdifferences from healthy controls in self-reported quality of life (QOL) (7). Armproblems and related complaints, however,have emerged as an exception and are one of the symptoms related to poorer QOLoutcomes in survivors (7). Breast cancerpatients and survivors who developlymphedema following axillary lymph nodedissection report decreased QOL, as indicatedby low scores on the Medical Outcome Study– Short Form (SF-36) Emotional and RoleFunctioning and Bodily Pain subscales (8), as well as the subscales of the FunctionalAssessment of Cancer Therapy-Breastquality-of-life instrument (9). A subgroup ofpatients with lymphedema who also reportconsiderable levels of pain tend to experiencesignificant functional interference, psychiatricsymptoms, intrusive thoughts, body imagedisturbance, and decreased sexual drive (10).

There is mixed evidence regardingwhether the impact of lymphedema onquality of life is due to arm swelling versusarm symptoms. In a study of the effects ofcomplete decongestive therapy on limbvolume reduction and QOL in breast cancersurvivors with lymphedema, QOL improvedsignificantly following treatment (11).Interestingly, however, improvement wasunrelated to actual volume decrease in theaffected arm (11). Even though the incidenceof lymphedema differs depending on the typeof surgery performed QOL outcomes at oneyear post-surgery are not worse in womenwho had more invasive operations (12).Therefore, the question remains: is the effectof lymphedema on quality of life due to armswelling? Or is it due to arm symptoms? Theanswer to this question will assist clinicians in deciding what is more important to treat to improve quality of life among breast cancer survivors.

This report, which uses baseline datafrom the recently completed Physical Activity

and Lymphedema (PAL) Trial, assesses theway in which different definitions of armswelling, and number and type of armsymptoms (13) differentially relate to a rangeof QOL outcomes in a diverse sample ofalmost 300 female survivors of breast cancer.The PAL trial cohort is uniquely able toaddress this issue as it includes breast cancersurvivors with and without clinicallydiagnosed lymphedema who underwentassessments of limb volume, arm symptoms,pain, and QOL. It is hypothesized that armsymptoms will be more associated with QOLoutcomes than arm swelling among breastcancer survivors with and withoutlymphedema. It is furthermore expected thatpain will be significantly correlated withpoorer outcomes in QOL domains insurvivors with and without lymphedema.

METHODS

All methods were approved by theinstitutional review board at the University of Pennsylvania, and all participants providedwritten consent for their participation in the study.

Sample Characteristics

Participants were 295 female breastcancer survivors (mean age=55.98 years,SD=8.83, range: 36-80 years) taking part in a randomized controlled strength trainingintervention trial. Of the 295 participants, 141 entered the study with a prior diagnosisof lymphedema. Women with lymphedemawere randomized into two equal sized groups,with a primary aim to assess the safety ofstrength training by comparing rates oflymphedema exacerbations (“flare-ups”) inthe treatment and control groups. The other154 participants entered the study with noevidence of lymphedema and were alsorandomized into two equal sized groups toassess the safety of strength training with aprimary outcome of lymphedema onset.Women were primarily recruited through


3

letters sent in collaboration with thePennsylvania and New Jersey state cancerregistries. Study eligibility criteria for theentire sample included female gender, historyof unilateral breast cancer, and a body massindex ≤ 50 kg/m2. Eligibility criteria specificto those with lymphedema were that theywere one to 15 years post-diagnosis, had atleast one lymph node removed, and eitherhad current evidence of lymphedema (10%interlimb difference or obscuration ofanatomic architecture upon clinical exam bya qualified lymphedema therapist or pittingedema), or clinical documentation of a priordiagnosis. Lymphedema had to be “stable,”defined as having, over the past threemonths, no therapist-delivered treatment, no more than one arm infection requiringantibiotics, no changes in the ability tocomplete activities of daily living due tolymphedema symptom changes, and noverified changes greater than 5% in armswelling. There were no upper limits placedon the magnitude of interlimb differenceallowable among women with lymphedema.Eligibility criteria specific to those withoutlymphedema were that they must have had at least two lymph nodes removed as part oftreatment and a cancer diagnosis between one and 5 years prior to study entry. Basedon the exclusion criteria specified, 1411 of the 3200 women who contacted study staffabout participation were excluded. Of these,13 women were excluded solely because theypresented with unstable lymphedema.Women who were interested in the study who were currently experiencing a flare-up or cellulitic infection were asked to gettreatment and call back after the arm hadstabilized. The majority of the 13 womenexcluded due to unstable lymphedema calledless than 3 months before the end of therecruitment period, so this option was notpossible for them. The study startedrecruitment in October 2005 and completedrecruitment in February 2007. All dataincluded for this report were collected atbaseline.

Measures

At baseline all participants completed a self-report battery of QOL measures,including the eight subscales of the MedicalOutcome Study – Short Form (SF-36),Coopersmith Self-Esteem Inventory (SEI) –Adult Short Form, the Fatigue SymptomInventory – Disruption Index, LifeOrientation Test – optimism and pessimismsubscales, the Medical Outcome Study (MOS) Social Support Scale, the TemporalSatisfaction with Life Scale, the PittsburghSleep Quality Index (PSQI), and the threeBody Image and Relationships Scale (BIRS)subscales (Table 1). All measures – with theexception of the recently developed BIRS –are widely used instruments; all haveevidence for good reliability and validity.

Lymphedema presence and severity was assessed through objective measurementand clinician rating. Objective measures were made by study research staff at pre-randomization baseline, and used tocategorize participants into lower, middle,and upper tertiles according to measures ofinterlimb difference in water volume (%) andof largest arm circumference difference (%).Both have been found to be reliable measuresof lymphedema in clinical practice (14).Clinician ratings were assigned by a studyclinician pre-randomization based on aphysical examination. Participants withlymphedema were categorized according to the National Cancer Institute’s CommonToxicity Criteria (CTC), ranging from 0:“normal or sub-clinical” to 4: “severelymphedema limiting function withulceration” (15).

Number and perceived severity of armsymptoms, regardless of lymphedema, wereassessed using a 14-item self-report measuredeveloped by Norman and colleagues (13),which describes the presence, frequency,severity and distress caused by individualarm symptoms commonly associated withlymphedema (symptoms listed in Tables 3aand 3b). Finally, participants were asked to


4

indicate whether they experienced any painon the side of the body that was treated, andif so, to rate the severity of their pain on ascale of 0: “no pain” to 10: “worst pain

imaginable.” All measurements wereadministered to women with and withoutlymphedema.

TABLE 1Description of Quality of Life Assessment Instruments Administered to Respondents


5

Statistical Analyses

The relationship between QOL outcomes,lymphedema severity, and arm-relatedsymptoms was assessed using Pearson’scorrelation coefficients. Time since diagnosisand race correlated significantly with anumber of QOL measures and werecontrolled for in all analyses. Correlationanalyses were conducted separately forgroups of participants with and without aprior diagnosis of lymphedema, and for allparticipants combined. Groups werecompared using chi-square, t-test analyses,and one-way analyses of variance (ANOVA).All significance tests were two-sided andconducted using a Type I error of 0.05. It wasassumed here that lymphedema and arm-related symptoms impact QOL, and thusQOL was considered the outcome variable inall analyses.

RESULTS

Demographics

Respondents self-identified as Caucasian(n=194, 65.8%), African-American (n=95,32.2%), Asian (n=6, 2.0%), AmericanIndian/Alaskan Native (n=5, 1.7%), Hispanic(n=2, 0.7%), Native Hawaiian (n=1, 0.3%),and “other” (n=1, 0.3%). Approximately athird of all respondents had completed somecollege (34.2%, n=101), 23.1% (n=68) had afour-year college degree, and 26.8% (n=79)had a post-graduate degree. A majority ofparticipants was married or living with apartner (61.3%, n=179); fewer respondentsreported being divorced or separated (16.8%,n=49), widowed (11.0%, n=32), or nevermarried (11.0%, n=32). Fewer than half ofthe respondents (42.4%, n=125) were primarycaregivers for children living with them.

Description of Arm Swelling and ArmSymptoms in the PAL Cohort

The average interlimb volume difference

among women with lymphedema was 16.12%(SD=15.19), and among women withoutlymphedema was -0.07% (SD=5.00; t(293)=-12.07, p≤0.001). Circumferential measure-ments were also taken. The differencesbetween the affected and unaffected limbs atthe point of the largest visible difference hadan average of 14.14 cm (SD=10.91) amongthe women with lymphedema and 3.63 cm(SD=3.03; t(285)=-10.96, p≤0.001) amongthose without lymphedema. Using CTCcategories of lymphedema severity a majorityof all participants were categorized as “0”(i.e., “normal” or controlled lymphedema)(n=184, 62.4%), followed by 21.4% (n=63) ofparticipants in category 2 (i.e., “moderatelymphedema, requiring compression”), 8.8%(n=26) in category 3 (i.e., “severe lymphe-dema limiting function”), and 7.5% (n=22) incategory 1 (i.e., “mild lymphedema”). Of thewomen with a prior diagnosis of lymphedema44.7% (n=63) fell into the CTC category “2,”21.3% (n=30) were categorized as “0,” 18.4%(n=26) as “3,” and 15.6% (n=22) as “1.”Thus, the sample recruited represented arange of lymphedema severity.

Regardless of prior lymphedema statusall participants were categorized intoclinically defined tertiles based on interlimbvolume differences as lowest (-11.0%-9.9%,n=203, 68.8%), middle (10.0%-19.9%, n=49,16.6%), and highest tertiles (≤ 20.0%-, n=43,14.6%). The values for the largest interlimbpercent difference in arm or hand circum-ference ranged from 0.16% to 44.35%(M=8.72%, SD=9.48), yielding clinicallydefined tertiles of lowest (0.16%-9.9%, n=210,71.2%), middle (10.0%-19.9%, n=41, 13.9%),and highest tertiles (≤20.0%, n=36, 12.2%) in the entire sample of participants.

Participants on average reported threeout of a possible 14 arm symptoms (M=3.31,SD=3.17, range: 0-12). Participants who hadpreviously been diagnosed with lymphedemaon average reported a significantly highernumber of symptoms (M=5.48, SD=2.83) than those who entered the study without adiagnosis (M=1.32, SD=1.92; t(282)=-14.35,


6

p≤0.001). Of those without a prior lymphe-dema diagnosis, 50.7% reported at least one symptom, 30.4% reported at least twosymptoms, and 20.3% reported at least three symptoms; 7.4% reported five or more symptoms.

About half of the full PAL cohort (n=148,50.5%) reported currently experiencing painon the side of their body treated for breastcancer, including 50.7% of the womenwithout lymphedema. Respondents whoreported experiencing any pain rated theseverity of their pain as relatively mild(M=2.30, SD=2.09, range: 0-8 on a scale of“0=no pain” to “10=worst pain imaginable”).Respondents with any pain symptomsreported significantly more total armsymptoms (M=4.15, SD=3.24) than those whodid not currently experience any pain[M=2.47, SD=2.86; t(280)=-4.62, p≤0.001].

Criteria Overlap

Arm volume as measured by percentinterlimb difference in water volume andpercent difference in circumference correlatedsignificantly with CTC categorization(r=0.88, p≤0.001 for water volume; r=0.80,p≤0.001 for circumference), total number ofsymptoms (r=0.50, p≤0.001; r=0.52,pœ0.001), and self-reported pain severity(r=0.24, p=0.003; r=0.32, p≤0.001). CommonToxicity Criteria categorization correlatedsignificantly with number of arm symptoms(r=0.57, p≤0.001) and pain severity (r=0.24,p=0.01), and number of symptoms correlatedsignificantly with pain severity (r=0.49,p≤0.001).

Lymphedema and Quality of Life:Lymphedema Defined by Interlimb WaterVolume Difference

One-way ANOVAs that included allparticipants (with and without lymphedema)yielded no significant differences between thethree tertile groups of largest, medium, andsmallest percent interlimb water volume

difference in terms of any of the QOLoutcomes. In a separate analysis thatincluded only those participants with a priordiagnosis of lymphedema, the continuousmeasure of water volume was also notsignificantly correlated with any QOLoutcomes.

Lymphedema and Quality of Life:Lymphedema Defined by LargestCircumference Difference

In an analysis that included allparticipants (with and without lymphedema),the three tertile groups of largest, medium,and smallest percent interlimb circumferencedifference did not differ in any QOLoutcomes, with the exception of scores on theSF-36 Physical Role Functioning Scale[F(2,281)=3.17, p=0.04], where those in thethird tertile scored significantly higher(M=89.71, SD=24.71) than those in thesecond tertile (M=69.51, SD=41.65, p=0.03),with the first tertile mean scores falling inbetween (M=79.39, SD=34.52). In a separateanalysis that included only those participantswith a prior diagnosis of lymphedema, thecontinuous measure of circumferencedifference was not correlated significantlywith any of the QOL outcomes, with theexception of scores on the PSQI (r=-0.24,p=0.04).

Lymphedema and Quality of Life:Lymphedema Defined by Common ToxicityCriteria

In the full cohort (those with and withoutlymphedema) there were no differencesacross CTC categories in any QOL outcomes,with the exception of scores on the Cooper-smith Self-Esteem Inventory [F(3,257)=2.60,p=0.05], where those in CTC category “0”scored significantly lower than those incategory “1” (p=0.04). In those with anexisting diagnosis of lymphedema significantdifferences emerged on the SF-36 PhysicalRole Functioning subscale [F(3,134)=2.90,


7

p=0.04], with those with grade 3 lymphedemascoring significantly higher than respondentsin grade 0 (p=0.02).

Arm Symptoms and Quality of Life: Number and Type

In an analysis that included allparticipants (with and without lymphedema)and controlling for time since diagnosis andrace, total number of arm symptomscorrelated significantly with the PhysicalFunction, Physical Role Functioning,Emotional Role Functioning, Social, BodilyPain, Vitality, and General HealthPerceptions subscales of the SF-36, the

Fatigue Scale Disruption Index, scores on the Quality of Life visual analog scale, andthe Strength and Health, Social Barriers, andAppearance and Sexuality subscales of theBIRS in the entire sample of respondents(Table 2). In a separate analysis that includedonly those respondents with a prior diagnosisof lymphedema and controlling for time sincediagnosis and race, total number of armsymptoms was correlated significantly withthe Physical Function, Physical RoleFunctioning, Emotional Role Functioning,Social, Bodily Pain, and General HealthPerceptions subscales of the SF-36, and theSocial Barriers and the Appearance andSexuality subscales of the BIRS (Table 2).

TABLE 2Number of Arm Symptoms and Quality of Life Outcomes


TA

BL

E 3

aC

orre

lati

ons

of I

ndiv

idua

l S

ympt

oms

wit

h Q

uali

ty o

f L

ife

Out

com

es (

shad

ed b

oxes

ind

icat

e si

gnif

ican

ce o

f p<

0.05

)

8


9

In yet another separate analysis that onlyincluded respondents without a prior diag-nosis of lymphedema and again controllingfor time since diagnosis and race, there were significant correlations between totalnumber of symptoms and the PhysicalFunction, Physical Role Functioning, Pain,Mental Health, Vitality, and General HealthPerceptions subscales of the SF-36, theFatigue Disruption Index, scores on theCoopersmith SEI, Life Orientation TestOptimism and Pessimism subscales,Satisfaction with Life, and the Strength and Health subscale of the BIRS (Table 2).

In terms of specific arm symptomsreported by the full sample of respondents,experiencing “pain in hand or arm on oneside” was the most strongly associated withQOL outcomes, correlating significantly with 16 of the 20 measures administered.Having “rings too tight on one side” andexperiencing “difficulty writing” eachcorrelated significantly with 12 of the 20 QOL measures. Having one’s “arm feelingtired, thick, or heavy on one side” correlatedsignificantly with 10, and reporting “skinfeeling different, firmer, or leathery on oneside” correlated significantly with eightmeasures of QOL. Not being able to see yourknuckles, noticing indentations in skin whenleaning against something, and not being ableto see veins in the hand on one side were theleast associated with QOL outcomes,correlating only with one or two measuresadministered (Tables 3a,3b).

Lymphedema and Quality of Life: The Role of Pain

In an analysis that included allparticipants, regardless of lymphedemastatus, respondents with any pain differedsignificantly from individuals with no pain on all but one (MOS Social Support) measureof QOL outcomes. Pain severity correlatedsignificantly with all SF-36 subscales, theFatigue Symptom Inventory DisruptionIndex, the Coopersmith SEI, the Quality of

Life Visual Analogue Scale, the PessimismIndex of the Life Orientation Test, the PSQI,and the Strength and Health and SocialBarriers subscales of the BIRS (Table 4). Itshould be noted that there is some contentoverlap with items assessing pain included onthe SF-36 as well as the total symptom scoreand that the scale for the BIRS is the reverseof the other scales (higher score is worse).

DISCUSSION

This report evaluated the relationship of arm swelling and arm symptoms with abattery of QOL measures to determinewhether swelling or symptoms would be most associated with poor QOL adjustmentfollowing breast cancer treatment. It shouldbe noted that participants with and withoutexisting lymphedema diagnosis were assessed,thus results regarding the influence of armsymptoms on QOL reflect the full spectrum of experiences in breast cancer survivors.Number of arm symptoms correlated signifi-cantly with QOL outcomes in respondentswith and without a prior diagnosis oflymphedema. Arm swelling, on the otherhand, was only significantly correlated withan overall sleep score and a physical functionsubscale of the SF-36. A significantproportion of participants who had not beendiagnosed with lymphedema previouslyreported experiencing symptoms, includingpain. These findings are consistent with thehypothesis that arm symptoms and pain aremore closely correlated with a negativeimpact on quality of life than arm swelling.The results emphasize the importance ofrecognizing that lymphedema is not justabout the swelling, as well as the fact thatlymphedema is only one arm-related issue ona spectrum of diverse upper body problemsand pain in breast cancer survivors.

Several symptoms emerged as beingparticularly strongly correlated with poorerQOL outcomes. Pain in hand or arm on theside treated for breast cancer was the moststrongly associated individual symptom in


10

TA

BL

E 3

bC

orre

lati

ons

of I

ndiv

idua

l S

ympt

oms

wit

h Q

uali

ty o

f L

ife

Out

com

es (

shad

ed b

oxes

ind

icat

e si

gnif

ican

ce o

f p<

0.05

)


11

TABLE 4Pain Symptoms and Severity and Quality of Life Measures


12

the full cohort, regardless of lymphedemadiagnosis. The importance of pain wasreinforced by the fact that pain assessed inother ways (yes/no and severity scale) wasalso associated with poorer QOL. Otherindividual symptoms correlated with worseQOL included measures of functioning, suchas difficulty writing and inability to wearrings, as well as heaviness in one arm or handand changes in skin texture and appearance.These findings should be noted by cliniciansas indicating the symptoms that would bemost important to resolve to improve qualityof life among breast cancer survivors,regardless of lymphedema diagnosis.

This study has many strengths that speak for the generalizability of findings. Thelarge sample surveyed was highly diverse interms of age, race, and ethnicity, making thisone of the only studies highlighting QOLoutcomes in breast cancer survivors of diversebackgrounds. Participants entered the studyat a variety of time points from their initialdiagnosis, represented a broad cross sectionof lymphedema severity (including 26 womenwith grade 3 lymphedema and 63 womenwith grade 2 lymphedema), and reported arange of arm symptoms, swelling, and pain,as well as widely varying levels of QOL.Measures used were manifold and of highquality, yielding a valid, reliable, and nuancedpicture of the ways in which lymphedemaand QOL are related. A limitation to be notedis that due to its cross-sectional design thepresent study cannot speak to the direction ofassociation between the variables of interest.It is however unlikely that poorer QOL has asignificant impact on upper body symptomsand pain, as opposed to vice versa.

In summary, the findings herein supportthe conclusion that arm symptoms and painare more closely correlated to multipledomains of quality of life among breastcancer survivors than arm swelling. This wasobserved in a diverse cohort of breast cancersurvivors with and without a diagnosis oflymphedema, including 63 women with inter-limb differences of 10-20% and 26 women

with interlimb differences over 20%.Lymphedema is clearly not just aboutswelling. Clinicians who work with breastcancer survivors are urged to focus on armsymptoms and pain as much if not more thanon arm swelling. Assessment and treatmentof arm symptoms is warranted, even insurvivors who do not meet full diagnosticcriteria for lymphedema, as these may pointto possible complications in psychosocialadjustment and QOL following treatment.

ACKNOWLEDGMENT

This investigation was supported by theNational Cancer Institute (R01-CA106851 toSchmitz) and by the Public Health ServicesResearch Grant RR024134 from the NationalInstitutes of Health to the University ofPennsylvania.

REFERENCES

1. Warren, AG, H Brorson, LJ Borud, et al:Lymphedema: A comprehensive review. Ann.Plast. Surg. 59 (2007), 464-472.

2. Ganz, PA: The quality of life after breastcancer: Solving the problem of lymphedema.N. Engl. J. Med. 340 (1999), 383-385.

3. Vignes, S, M Arrault, A Dupuy: Factorsassociated with increased breast cancer-related lymphedema volume. Acta Oncologica46 (2007), 1138-1142.

4. Seifart, U, US Albert, ME Heim, et al:Lymphedema in patients with breast cancer –A consensus regarding diagnostics andtherapy in patients with postoperativelymphedema after primary breast cancer.Rehabilitation 46 (2007), 340-348.

5. Werner, RS, B McCormick, J Petrek, et al:Arm edema in conservatively managed breastcancer: Obesity is a major predictive factor.Radiology 180 (1991), 177-184.

6. Engel, J, J Kerr, A Schlesinger-Raab, et al:Axilla surgery severely affects quality of life:Results of a 5-year prospective study in breastcancer patients. Breast Cancer Res. Treat. 79(2003), 47-57.

7. Dorval, M, E Maunsell, L Deschenes, et al:Long-term quality of life after breast cancer:Comparison of 8-year survivors withpopulation controls. J. Clin. Oncol. 16 (1998),487-494.


13

8. Velanovich, V, W Szymanski: Quality of lifeof breast cancer patients with lymphedema.Am. J. Surg. 177 (1999), 84-187.

9. Beaulac, SM, LA McNair, TE Scott, et al:Lymphedema and quality of life in survivorsof early-stage breast cancer. Arch. Surg. 137(2002), 1253-1257.

10. Newman, ML, M Brennan, S Passik:Lymphedema complicated by pain andpsychological distress: A case with complextreatment needs. J. Pain Symptom Manage.12 (1996), 376-379.

11. Weiss, JM, BJ Spray: The effects of completedecongestive therapy on the quality of life ofpatients with peripheral lymphedema.Lymphology 35 (2002), 46-58.

12. Shimozuma, K, PA Ganz, L Petersen, et al:Quality of life in the first year after breastcancer surgery: Rehabilitation needs andpatterns of recovery. Breast Cancer Res.Treat. 56 (1999), 45-57.

13. Norman, SA, LT Miller, HB Erikson, et al:Development and validation of a telephonequestionnaire to characterize lymphedema inwomen treated for breast cancer. Phy. Ther.81 (2001), 1192-1205.

14. Chen, YW, HJ Tsai, HC Hung, et al:Reliability study of measurements forlymphedema in breast cancer patients. Am. J. Phy. Med. Rehab. 87 (2008), 33-38.

15. Common Toxicity Criteria, Version 2.0, 1998.16. Coopersmith, S: Self Esteem Inventories. Palo

Alto, CA, Consulting Psychologists Press, 1981.17. Hann, DM, PB Jacobsen, LM Azzarello, et al:

Measurement of fatigue in cancer patients:Development and validation of the FatigueSymptom Inventory. Qual. Life Res. 7 (1998),301-310.

18. Scheier, MF, CS Carver, MW Bridges:Distinguishing optimism from neuroticism(and trait anxiety, self-mastery, and self-esteem): A re-evaluation of the Life Orienta-tion Test. J. Pers. Soc. Psychol. 67 (1994),1063-1078.

19. Sherbourne, CD, AL Stewart: The MOS social support survey. Soc. Sci. Med. 32(1991), 705-714.

20. Ware, J, M Kosinski, J Dewey: How to scoreversion 2 of the SF-36(R) Health Survey.Lincoln, RI: QualityMetric Inc., 2000.

21. Tomich, PL, VS Helgeson: Five years later: A cross-sectional comparison of breast cancersurvivors with healthy women. Psychooncology11 (2002), 154-169.

22. Ganz PA, Desmond KA, Belin TR, MeyerowitzBE, Rowland JH. Predictors of sexual healthin women after a breast cancer diagnosis. J.Clin. Oncol. 17 (1999), 2371-2380.

23. Rowland JH, Desmond KA, Meyerowitz BE,Belin TR, Wyatt GE, Ganz PA. Role of breast reconstructive surgery in physical andemotional outcomes among breast cancersurvivors. J. Natl. Cancer Inst. 92(2000),1422-1429.

24. Buysse, DJ, CF Reynolds, TH Monk, et al:The Pittsburgh Sleep Quality Index: A newinstrument for psychiatric practice andresearch. Psychiatry Res. 28 (1989), 193-213.

25. Carpenter, JS, MA Andrykowski:Psychometric evaluation of the PittsburghSleep Quality Index. J. Psychosom. Res. 45(1998), 5-13.

26. Pavot W, Diener E, Suh E. The temporalsatisfaction with life scale. J. Pers. Assess. 70 (1998), 340-354.

27. Hormes, JM, LA Lytle, CR Gross, et al: The Body Image and Relationships Scale:Development and validation of a measure ofbody image in female breast cancer survivors.J. Clin. Oncol. 26 (2008), 1269-1274.

Kathryn H. Schmitz, Ph.D., MPH903 Blockley Hall423 Guardian DriveCenter for Clinical Epidemiology

and BiostatisticsDepartment of Biostatistics andEpidemiologyUniversity of PennsylvaniaPhiladelphia, PA 19104-6021Phone: 215-898-6604Fax: 215-573-2265E-mail: [email protected]


Documents

IMPACT OF LYMPHEDEMA AND ARM SYMPTOMS ON QUALITY …