Abstracts
Conclusions: We report the first validation of clinicogenetic risk models
in a population-based cohort. Furthermore, we describe the role of PD-L2
and several additional TME markers which predict survival with conven-
tional therapies in FL. To our knowledge this is the first description of
PD-L2 as a prognosticator in any tumour subtype. Importantly, the
methods used to define TME biomarkers were applicable with archival tis-
sue, which is essential for translational application. These TME bio-
markers are independent and additive to new state-of- the-art prognostic
models. Incorporation of immune-effectors and immune-checkpoints into
a new clinco- immuno-genetic prognosticator has the potential to improve
risk stratification of FL patients and should be further validated in addi-
tional cohorts.
References
1 Casulo C, Byrtek M, Dawson KL, Zhou X, Farber CM, Flowers CR,
et al. Early Relapse of Follicular Lymphoma After Rituximab Plus Cyclo-
phosphamide, Doxorubicin, Vincristine, and Prednisone Defines Patients
at High Risk for Death: An Analysis From the National LymphoCare
Study. J Clin Oncol. 2015;33(23):2516-22.
2 Jurinovic V, Kridel R, Staiger AM, Szczepanowski M, Horn H, Dreyling
MH, et al. Clinicogenetic risk models predict early progression of follicular
lymphoma after first-line immunochemotherapy. Blood. 2016;128
(8):1112-20.
3 Dave SS, Wright G, Tan B, Rosenwald A, Gascoyne RD, Chan WC,
et al. Prediction of survival in follicular lymphoma based on molecular fea-
tures of tumor-infiltrating immune cells. N Engl J Med. 2004;351
(21):2159-69.
4 Keane C, Gould C, Jones K, Hamm D, Talaulikar D, Ellis J, et al. The
T-cell receptor repertoire influences the tumor microenvironment and is
associated with survival in aggressive B-cell lymphoma. Clinical Cancer
Research. 2016.
SOUTH AUSTRALIA
EPICARDIAL, ABDOMINAL, AND OVERALL ADIPOSITY:
IMPLICATIONS FOR ATRIAL FIBRILLATION
Wong CX and Sanders P
1
Centre for Heart Rhythm Disorders, South Australian Health and
Medical Research Institute, Royal Adelaide Hospital and University
of Adelaide, South Australia, Australia
Background: There is limited previous data on the role of obesity in atrial
fibrillation (AF). We sought to better understand this through a research
program exploring epicardial, abdominal and overall adiposity.
Methods: A series of systematic reviews and meta-analyses of adiposity
and AF was undertaken. Cardiac magnetic resonance imaging (CMRI)
was employed to characterise the relationship between overall and epicar-
dial adiposity with the presence, severity and recurrence of AF. The preva-
lence of AF in Indigenous and non-Indigenous Australians was studied,
and the potential role of epicardial adiposity assessed using computed
tomography (CT).
Results: In 626,603 individuals, every 5-unit BMI increase was associated
with 29% greater incident AF. Similar increases conferred a 10% and 13%
risk of post-operative and post-ablation AF. CMRI in 130 individuals
showed that epicardial fat associated with the presence, severity, and recur-
rence of AF. Further analyses from 352,275 individuals compared AF risk
for one-standard deviation (1-SD) increases in epicardial fat, waist circum-
ference (WC), waist-to-hip ratio (WHR) and BMI. A 1-SD increase in
epicardial fat was associated with 2.1 to 5.4-fold higher odds of AF out-
comes; in contrast, associations of abdominal and overall adiposity with
AF were considerably less extreme. In 204,668 hospitalised individuals
under 60, AF was almost twice as prevalent in Indigenous compared to
non-Indigenous patients (9.26 vs. 4.61%). A CT study showed signifi-
cantly more voluminous epicardial fat in Indigenous patients.
Conclusions: Associations of adiposity and AF appear to be strong, have
a dose-response with AF severity, are greater for epicardial as opposed to
abdominal and overall adiposity, and are consistent for numerous AF out-
comes and across clinical settings. AF is more prevalent in Indigenous
Australians and greater visceral adiposity may be one contributing mecha-
nism. Given the increasing burden of AF, adiposity may be a profitable
target for preventing and managing AF.
6 Copyright of individual abstracts remains with the authors.
TASMANIA
SMART WATCHES FOR HEART RATE ASSESSMENT IN
ARRHYTHMIAS
Koshy A1,2, Sajeev J2, Nerlekar N3, Brown A3, Rajakariar K2,
Wong M2, Cooke J2 and Teh A2
1
Royal Hobart Hospital, Department of Cardiology, Hobart,
Tasmania, Australia
2
Monash University, Eastern Health Clinical School, Department of
Cardiology, Box Hill, Victoria, Australia
3
Monash Cardiovascular Research Centre, Department of Medicine,
Monash University and Monash Heart, Monash Health, Clayton,
Victoria, Australia
Background: Wrist-worn smart watches (SW) that estimate heart rate
(HR) via photoplethysmography are increasingly popular. HR estimation
using SW has been evaluated in healthy controls in sinus rhythm with vari-
able accuracy depending on the device used, activity undertaken and HR
attained.1-3 Although not marketed for medical use, patient-initiated pre-
sentations for medical assessment due to device-detected HR abnormali-
ties are increasing, with a paucity of data on their validity in cardiac
arrhythmias. We aimed to assess the accuracy of two leading SWs for HR
estimation in a range of heart rhythms.
Methods: This prospective cohort study included 112 hospitalized
patients (cardiac arrhythmias (n=60), sinus rhythm (n=52) recruited from
a tertiary-care hospital. Patients were evaluated at rest using continuous
electrocardiogram (ECG) monitoring as reference standard, with concom-
itant SW-HR monitoring for 30- minutes, with HR recorded every
15-seconds. Participants wore an Apple Watch (AW) (Apple Inc., Cuper-
tino, USA) and Fitbit Blaze (FB) (Fitbit Inc., San Francisco, USA) on
either wrist, secured firmly above the ulnar styloid. Spearman rank correla-
tion (rs) and Bland-Altman analysis with 95% limits of agreement (LoA)
were used to assess agreement between SW-HR and ECG-HR. Bias was
calculated as the mean difference between ECG-HR and SW-HR.
Results: Across all devices 40,720 HR values were collected. Sinus
rhythm demonstrated strong correlation with ECG- HR (FB rs 0.90, AW
rs 0.99, both p<0.01), with no differences observed between sinus brady-
cardia and tachycardia. AW demonstrated a stronger correlation with
arrhythmia than FB (AW rs =0.87, FB rs =0.68, both p<0.01) with lower
LoA. Subgroup arrhythmia analysis demonstrated the highest correlation
for atrial flutter in both devices (AW rs =0.99, FB rs =0.98, both p<0.01)
with a mean bias <1 beat (Figure 1). However, in atrial fibrillation (AF),
there was significant HR underestimation with both devices with only a
moderate correlation for AW (rs =0.63, p<0.01, mean bias 8 beats) and
very weak for FB (rs =0.09, p<0.01, mean bias 28 beats) (Figure 1). Very
weak, non-significant correlations were seen with both devices for com-
plete heart block (FB rs =-0.05, AW rs =0.17, p 0.35).
Conclusion: SW estimation of HR demonstrated most accuracy in
patients with sinus rhythm and atrial flutter, but was reduced for other
arrhythmias. The discrepancy between devices was most marked in AF,
with AW outperforming FB. Further improvements in these devices are
needed before they can be reliably used for chronotropic assessment and
arrhythmia detection.
Trial Registration: Australian New Zealand Clinical Trials Registry
(ANZCTR 1261001374459)
Smart Watch Heart Rate
References
1 Stahl SE, An HS, Dinkel DM, Noble JM, Lee JM. How accurate are the
wrist-based heart rate monitors during walking and running activities? Are
they accurate enough? BMJ Open Sport Exerc Med. 2016;2(1):e000106.
2 Wallen MP, Gomersall SR, Keating SE, Wisloff U, Coombes
JS. Accuracy of Heart Rate Watches: Implications for Weight Manage-
ment. PLoS One. 2016;11(5):e0154420.
3 Wang R, Blackburn G, Desai M, et al. Accuracy of Wrist-Worn Heart
Rate Monitors. JAMA Cardiol. 2017;2(1):104-106.
Editorial material and organization © 2018 Royal Australasian College of Physicians.
Internal Medicine Journal (2018) 48 (Suppl. 5): 5–8

Smart Watch Heart Rate
Figure 1: Spearman correlations (rs) between SW devices and ECG for both atrial fibrillation and atrial flutter. Measures are HR recorded
every 15 seconds.
VICTORIA
CORRELATION BETWEEN A 10-COLOUR FLOW CYTOMETRIC
MINIMAL RESIDUAL DISEASE (MRD) ANALYSIS AND MOLECULAR
MRD IN ADULT ACUTE LYMPHOBLASTIC LEUKAEMIA (ALL)
Singh J1, Gorniak M1, Grigoriadis G1,2, Westerman D3, McBean M3,
Sutton R4, Morgan S1, and Fleming S1,2
1 therapies such as blinatumomab. This technique offers rapid and afford-
Laboratory Haematology, Alfred Pathology, Melbourne, Victoria,
Australia
2 assay cannot be developed.
Department of Clinical Haematology, Monash Health, Melbourne,
Victoria, Australia
3 WESTERN AUSTRALIA
Department of Pathology, Peter MacCallum Cancer Centre at the
Victorian Comprehensive Cancer Centre, Melbourne, Victoria,
Australia
4
Children’s Cancer Institute, Sydney, NSW, Australia
Background: MRD monitoring in ALL is a strong predictive factor and a
stratification tool for treatment intensification. The currently accepted
standard of molecular monitoring with either immunoglobin heavy or
kappa chain (IG) or T-cell receptor (TCR) quantitative PCR (qPCR) in
Philadelphia negative (Ph-) ALL offers high sensitivity, but accessibility is
limited by expertise, cost and turnaround time. Flow cytometric assays are
increasingly utilised and improved sensitivity is seen with multi-parameter
flow cytometry at 8 or more colours.
Methods: We developed a 10-colour single tube flow cytometry assay
(Figure 1). Samples were subject to bulk ammonium chloride lysis to max-
imise cell yields with a target of 1 x 106 events. Once normal maturation
patterns were established, patient samples were analysed in parallel to
standard molecular monitoring with either IG/TCR qPCR in Ph- disease
or BCR-ABL qRT-PCR in Ph+ disease. Statistical analysis was performed
in Graphpad Prism v7.0.
Results: Flow cytometry was performed on 47 samples from 16 patients.
13 samples were at diagnosis or morphologic relapse. An informative
immunophenotype was identifiable in all patients; however a molecular
assay could not able to be developed in one patient. 38 samples were
tested for MRD by flow cytometry (Figure 2). In 2 samples, flow cyto-
metric MRD was detected despite blinatumomab (anti-CD19) therapy.
27 samples were tested concurrently for MRD by both molecular and flow
cytometric methods (Figure 3A and 3B). There was a strong correlation
between molecular and flow cytometric MRD (R2=0.909, p<0.001;
Figure 3B). Correlation was strong with both IG/TCR- based (n=16;
R2=0.955, p<0.001) and BCR-ABL-based (n= 11; R2=0.957, p<0.001)
Editorial material and organization © 2018 Royal Australasian College of Physicians.
Copyright of individual abstracts remains with the authors.
Internal Medicine Journal (2018) 48 (Suppl. 5): 5–8
Abstracts
assays. The cost was significantly lower than IG/TCR qPCR (eg. Cost for
four time-points per patient approximately $1200 vs $3700).
Conclusion: Our 10-colour flow cytometric MRD assay attained sensitiv-
ity of ≤0.01% in 87% of samples, and correlated strongly with molecular
assays. By including CD22 in our panel, we demonstrated that we were
also able to detect MRD in patients who had been treated with anti-CD19
able testing in B-ALL patients, including cases where a suitable molecular
DPP-4 INHIBITOR MONOTHERAPY AND INSULIN RESISTANCE IN
POST TRANSPLANT DIABETES MELLITUS
Thiruvengadam S, Bennett K and Chakera A
Sir Charles Gairdner Hospital, Perth, Western Australia, Australia
Aim: To compare the insulin resistance in patients treated with DPP-4
therapy to a historical cohort of patients receiving other treatment for
NODAT.
Background: The use of corticosteroids and calcineurin-inhibitors to pre-
vent rejection after renal transplant predisposes patients to insulin resis-
tance. Within the first year, almost half of patients are affected by new
onset diabetes after transplant (NODAT). Although DPP-4 inhibitors
have shown efficacy in protecting beta cells from tacrolimus-induced toxic-
ity, there is limited data on their clinical application in this population.
Methods: Seventeen out of the 36 patients who received a renal transplant
between October 2015 and August 2016 had NODAT. Nine of them were
treated with a DPP-4 inhibitor as part of a new clinical pathway and had
Homeostatic Model Assessment – Insulin Resistance (HOMA-IR) scores
available. A second historical cohort of 43 patients transplanted between
2008 and 2015 had HOMA-IR scores available. Nine patients in this
cohort developed NODAT and received metformin, insulin and/or glicla-
zide. The HOMA-IR scores between these two cohorts of patients were
compared.
Results: The HOMA-IR scores of patients without NODAT did not dif-
fer significantly between the two cohorts suggesting that their baseline
characteristics were similar. Patients with NODAT treated with DPP-4
inhibitor therapy had significantly better (p<0.02) HOMA-IR scores
(mean 2.35) than those in the historical cohort treated with other agents
(mean 4.07). Furthermore, there was no difference in HOMA score
7