لخّصلي

We provide an up-to-date meta-analysis of trials showing

the benefits of SGLT2i for the reduction of cardiorenal mor-
bidity and mortality in individuals living with HF or CKD

and we reappraise the comparison of SGLT2i in people with

T2D as well. In people with HF, the results show benefits

across the spectrum of EF with reductions in the composite

endpoint of CV Death or hospitalization for HF in the range

of 21 to 25% and for reduction of hospitalization for HF

in the range of 31 to 26%. However, despite the additional

DELIVER trial data, reduction in all cause or CV mortality

and reduction in the renal composite endpoint remain evi-
dent only in the HFrEF population (Fig. 2 and Supplemental

Figures S9 – S16).
The EMPA-Kidney trial is unique in adding informa-
tion to participants with CKD defined by an eGFR of at

least 20 mL/min/1.73m2 but less than 45 mL/min/1.73m2

of body-surface area, or who had an eGFR of at least 45

mL/min/1.73m2 but less than 90 mL/min/1.73m2 with a

urinary albumin-to-creatinine ratio (with albumin mea-
sured in milligrams and creatinine measured in grams) of

at least 200. However, the report does not provide infor-
mation regarding the primary or individual endpoints for

these two disparate groups of participants. Subset analyses

stratified by eGFR or by albuminuria separately are, strictly

speaking, not heterogeneous but the report suggests that

benefits are seen largely in participants with higher levels

of proteinuria. We believe, therefore, that the existing CCS

guidelines recommending use of an SGLT2i in adults with

CKD (UACR>20  mg/mmol, eGFR≥25 mL/min/1.73m2)

to reduce the composite of a significant decline in eGFR,

progression to end-stage kidney disease or death due to kid-
ney disease, all-cause and CV mortality, non-fatal MI, and

hospitalization for HF remain largely unchanged although

a slightly lower eGFR of 20 mL/min/1.73m2 would be con-
sidered reasonable. In this group, the endpoint of non-fatal

stroke shows a point estimate reduction of 22% but this

remains non-significant in the new analysis even with the

new trial (Fig. 2 and Supplemental Figure S14). The effect on

the combined outcome of CV mortality or HF hospitaliza-
tion also remained unchanged (25% reduction) despite the

data from two additional trials (DELIVER and EMPULSE,

Fig. 2 and Supplemental Figure S11).
In individuals living with T2D with ASCVD or with mul-
tiple risk factors for ASCVD, the recommendations regard-
ing the use of either SGLT2i or GLP-1 RA for reduction of

all-cause or CV death or MACE are substantiated. Similarly,

the inclusion of new trial data does not alter our prior con-
clusion that there is a strong signal for the reduction of non-
fatal stroke associated with the GLP-1 but not the SGLT2i

class. Additionally, the significant but modest reduction in

non-fatal MI of 10% with SGLT2i remains not significantly

different from the 6%-point estimate reduction associated

with GLP-1 RA, which itself is not significant. Accordingly,

the new data still support the reticence of the CCS cardiore-
nal guideline to make any recommendations regarding the

reduction of non-fatal MI using SGLT2i. Finally, it is impor-
tant to note that the prior CCS cardiorenal guideline com-
mittee did not feel that recommendations regarding either

HF or CKD protection using GLP1-RA were warranted in

the absence of published, dedicated trials in these popula-
tions. In the population of people with T2D, the signal for

the reduction in the renal composite endpoint of 22% asso-
ciated with GLP-1 RA use was inferior to the 35% reduction

associated with the use of SGLT2i. In the current analy-
sis, however, the risk reduction associated with SGLT2i