Hydrochlorothiazide and metformin cross-over study for attenuating aquaretic side-effects in ADPKD patients treated with tolvaptan

Autosomal dominant polycystic kidney disease (ADPKD) is characterized by the
formation of numerous cysts in both kidneys and progressive renal function
decline leading to renal replacement therapy (RRT) in 70% of the patients at a
median age of 58. Despite a relatively low prevalence of 3-4:10.000, ADPKD is
the most common hereditary renal disease and accounts for 10% of all dialysis
patients. Recently, the first drug to slow down renal function decline,
vasopressin 2 receptor antagonist tolvaptan was approved to be prescribed to
ADPKD patients. Tolvaptan slows renal function decline by 26%, thereby
postponing RRT by one year every four years it is being used. Tolvaptan slows
down cyst growth by 49%.

As a vasopressin (antidiuretic hormone) receptor antagonist, aquaresis
associated side effects such as polyuria, thirst, nycturia and polydipsia occur
in the majority of patients. Aquaresis-related side effects are the most common
reason for discontinuing tolvaptan, and in our personal experience patients
regard the prospect of polyuria, which can amount to over 8 litres per day, as
the most important threshold for starting treatment. Currently, there are no
adjuvant treatment options to reduce these side effects. If the side effects
prove to be intolerable to a patient the only possibilities are to either cease
tolvaptan treatment, or to reduce the dose. In the TEMPO trial the average dose
of tolvaptan was 95mg, the majority of patients completed the study on the
maximum dose of 120mg per day. Therefore, there is little evidence for the
efficacy of reduced doses of 90mg or even 60mg per day.

Tolvaptan causes aquaresis by competitive antagonism of the vasopressin V2
receptor, thereby inhibiting migration of aquaporin 2 (AQP2) to the apical cell
membrane in the principal cells of the distal collecting duct. As a result the
distal collecting duct remains impermeable to water, which consequently cannot
be reabsorbed and is thus urinated. The mechanism by which tolvaptan causes
aquaresis is similar to the pathophysiology of nephrogenic diabetes insipidus
(NDI). In NDI The collecting duct is (relatively) impermeable to water due to
insensitivity to vasopressin, caused either by defects in the V2 receptor or
defects downstream, inhibiting migration of AQP2 to the apical cell membrane of
principal cells. Because of these similarities it is likely that the same
treatments that are effective to reduce polyuria in NDI may also be effective
to reduce polyuria in patients who use tolvaptan.

Hydrochlorothiazide has been an established treatment for polyuria in NDI for
over 50 years and is known to lower urine output by up to 30-50%. We
hypothesize that additional treatment with hydrochlorothiazide will lead to
reduction of polyuria in ADPKD-patients using tolvaptan by similar margins as
treatment with hydrochlorothiazide in NDI.

Recently, metformin has been shown to be a vasopressin-independent activator of
water transport in the inner kidney medulla through stimulation of
AMP-activated protein kinase (AMPK) in rats. This made metformin a viable
candidate as treatment for polyuria in NDI. in a study by Efe et al. rats were
treated with tolvaptan as a model for NDI, increasing 24-hour urine volume from
10mL to 22mL. Addition of metformin normalized urine volume within five days.
We hypothesize that metformin will attenuate polyuria and increase urine
osmolality in humans through nonvasopressor activation of AMPK.

To demonstrate whether hydrochlorothiazide or metformin can diminish aquaresis
in patients with ADPKD who are treated with tolvaptan as measured by 24-hour
urine volume.

A placebo-controlled, double-blind, double dummy, cross-over trial in subjects
with ADPKD that are using tolvaptan at the maximum dose of 120mg per day.
Treatment consists of metformin, hydrochlorothiazide and placebo in random
order. The study will consist of three treatment periods, each treatment period
contains two treatment weeks and one wash-out week, save for the last treatment
period, which does not have a wash-out week. Total study duration will be 8
weeks.

Subjects will be treated with hydrochlorothiazide, metformin and placebo for
two weeks each, followed by one wash-out week, in random order.
Hydrochlorothiazide will be initiated at 12,5mg QD, after one week the dose
will be increased to 25mg QD if well tolerated. Metformin will be initiated at
500mg BID, after one week the dose will be increased to 1000mg BID, if well
tolerated.

Burden and risk associated with participation are:
-Patients who are using tolvaptan as part of regular clinical care usually have
to come to the out-patient clinic every month during the first 18 months of
treatment. Participating in this study entails a minimum of 1 and a maximum of
4 extra study visits within 8 weeks of treatment plus the variable screening
period (5 study visits versus 1- 4 in regular care). A maximum of 5 additional
times venous blood has to be drawn (8 times versus 3 or 4 in regular care).
-4 times GFR measurement using the iohexol plasma clearance technique
-8 times 24-hour urine collection
-Patients will be exposed to metformin and hydrochlorothiazide (not combined),
while already using tolvaptan

Potential benefit of participation:
The potential benefit of this study is to reduce aquaretic side-effects of
tolvaptan. Potentially leading to less polyuria, nycturia, thirst and
polydipsia. Due to reduced polyuria, patients may be able to tolerate higher,
more effective doses of tolvaptan, which could ultimately lead to postponement
of or preventment of renal replacement therapy.