82 / 244
participants with median level of urinary NAG and below (4.89
(3.70
–
6.21) U/gCr). In participants with carotid plaques, the
levels of urinary NAG were significantly higher than those
without plaques (7.53 (5.24
–
12.0) vs 6.35 (4.40
–
8.35) U/gCr). In
the multiple regression analysis, age (STD
β
= 0.22), hyperten-
sion (STD
β
= 0.13), and above median level (7.21 U/gCr) of
urinary NAG (STD
β
= 0.13) predicted higher values of
maximum carotid IMT. Odds ratio for presence of carotid
plaques after adjustment for age, hypertension, albuminuria,
serum cholesterol, and estimated glomerular filtration rate
was 1.86 (95% CI, 1.02
–
3.38) for increase in urinary NAG. In
conclusion, urinary NAG was independently associated with
carotid atherosclerosis in patients with T2D.
OL10-8
Validation of a novel biomarker panel, DNlite, for
management of renal complication in type 1 diabetes
Yann-Jinn LEE
1
, Wei-Ya LIN
2
, Hsiang-Chi WANG
2
,
Chi-Yu HUANG
1
*, Wei-Hsin TING
1
, Lee-Ming CHUANG
3
,
Tzu-Ling TSENG
2
*.
1
Department of Pediatrics, Mackay Memorial
Hospital,
2
Bio Preventive Medicine Corp.,
3
Department of Internal
Medicine, National Taiwan University Hospital, Taiwan
Background:
In the previous studies, we have developed a
novel biomarker panel, DNlite, for detecting kidney disease
in patients with type 2 diabetic mellitus (T2DM). We also
establish a novel scoring system, diabetic nephropathy score
(DN_Score). DN_Score is a composite score built from fitting
several urinary biomarkers, including alpha2-HS-glycoprotein
precursor (AHSG), alpha-1-antitrypsin (A1AT) and acid-
1-glycoprotein (AGP) in DNlite, to a statistical model that
correlates highly with the stage of kidney disease in T2M.
However, the development of diabetic nephropathy (DN)
of T2DM is not as straightforward as it is in type 1 diabetic
mellitus (T1DM) where there is a clear progression from
normal renal function to hyperfiltration after about 5 years.
In order to strengthen the application of DNlite, we conduct a
large scale validation of DNlite in patients with T1DM.
Material and methods:
447 patients with T1DM were enrolled
and tested with DNlite. There were 206 male and 241 female
participants. The mean age was 21.15 ± 9.5 years. Related
clinical parameters were well recorded. To access the severity
and risk of kidney disease, patients were further categorized by
GFR and albuminuria (KDIGO 2012 Clinical Practice Guideline).
Results:
The difference between patients with normo- and
clinical albuminuria was very significant (p < 0.0001), and
diagnostic accuracy using AUROC is up to 0.92. The DN_Score
and stage in KDIGO is highly correlated. The difference
of DN_Score between the low risk (1 if CKD) and the high risk
(1
–
4+) is very significant (p < 0.0001). We also evaluated the
correlation of DN_Score with metabolic variables. DN_Score
was highly correlated with BMI, blood pressure, fasting plasma
glucose, HbA1c and plasma triglyceride level.
Conclusion:
DN_Score is correlated significantly with the
traditional indicators of DN in all stages of the disease in
Type 1 DM. The application of DNlite in T1DM for detecting of
DN has been demonstrated in this study. Furthermore, the
application of DNlite for managing the DN prognosis in T1DM
is under investigation.
Oral Presentations / Diabetes Research and Clinical Practice 120S1 (2016) S40
–
S64
S64