Dr. Watkins' Report

Question 2

Have any liver concerns been raised previously?

Answer 2

No liver safety concerns have been raised in the liver disease and congestive heart failure populations, either in the preapproval clinical trials or during post marketing pharmacovigilance. There were no liver safety concerns when designing the clinical trials for tolvaptan treatment of patients with PKD and hence liver chemistry monitoring was relatively infrequent.

Question 3

What are examples of drugs that have the potential to cause ALT elevations, even high, but have very low risk of cuasing clinically important liver injury.

Answer 3

tacrine, heparins, cholestyramine, aspirin, statins

Question 4

What is eDISH?

Answer 4

An efficient way to visually summarize and analyze liver safety data in clinical trials. Peak serum ALT is shown along the x-axis and the peak serum bilirubin along the y-axis as multiples of the ULN on log scales.

Question 5

What is the ALT:ALP ratio?

Answer 5

When the ALT and ALP values are expressed as x-fold ULN, a ratio exceeding 5 represents a hepatocellular injury.

Question 6

What is Temple’s Corollary

Answer 6

The RLQ of the eDISH plot. Reflects observation that an excess in this quadrant of subjects treated with a study drug has been reliably present with drugs capable of causing serious liver injury. this is because ALT is a much more sensitive indicator of hepatocellular injury than is bilirubin, and increaes are expected to occur before or without accompanying rises in serum bilirubin.

Question 7

What is Gilbert’s syndrome?

Answer 7

Inherited deficiency in the ability to glucuronidate bilirubin.

Question 8

How does the guidance document define a Hy’s Law case?

Answer 8

In addition to being within the Hy’s law quadrant, the subject must have hepatocellular injury without cholestasis (ALP < 2x); Second, there must not be a more likely explanation for the liver injury such as passing a gall stone or viral hepatitis. finally, there should be a higher incidence of ALT > 3x (Temple’s Corollary).

Question 9

What is the extent of exposure used for this analysis? (p. 6)

Answer 9

Over 1000 people have been exposed to the drug for more than 6 months and almost 900 exposed for one year. Specifically 1017 >=6 months and 886 >= 12 months.

Question 10

How many patients received treatment and for how ong in earlier studies? (p. 7)

Answer 10

A total of 589 subjects received treatment with tolvaptan for at least 14 months in these studies, the vast majority of whom had congestive heart failure.

Question 11

What were the criteria for case selection for blinded causality assessment?

Answer 11

1. SAEs and non-serious TEAEs that led to discontinuation due to hepatic or liver function test AEs from 5 SMQs. (cholestasis and jaundice; hepatic failure, fibrosis, and cirrhosis; non-infectious hepatitis; liver-related investigations; and liver-related coagulation disturbances).
2. ALT (or AST) >3x and BT >2x even if not concurrent, but no AEs.
3. FDA set criteria of ALT or AST >5x or BT >2x.

Question 12

11. What were the components of the causality assessment scale?

Answer 12

Definite > 95%- beyond a reasonable doubt
Highly LIkely 75-95% clear but not definite
Probable 50-74% preponderance supports link
Possible 25-49% equivocal but present
Unlikely < 25% evidence that another factor caused the injury
Unassessable

Question 13

What were the scheduled liver chemistry assessments for TEMPO? (p. 9)

Answer 13

baseline, 3 weeks, at 4 monhts, and then every 4 months thereafter.

Question 14

What do the eDISH plots show?(p. 11)

Answer 14

It can be seen that there is a clear imbalance between TLV and PLC treated subjects experiencing serum ALT elevations > 3x ULN- RLB and RUQ. In the RUQ (Hy’s Law)- there are two TLV subjects and no PLC subjects. There are no PLC treated cases with serum ALT exceeding 3X that were blindly adjudicated as at least probably related to study drug.

Question 15

When does the first separation occur? (p. 15)

Answer 15

Separation is first evident at the schedule 4 month blood draw- but not at the 3 week/end of titration blood draw- and the slope of the incidence curves appears to differ until about 14 months (about 400 days). Although there were no scheduled assessments between the week 3 and month 4 treatment, the slopw of the curves suggests that changes do not typically occur before 3 months of treatment.

Question 16

What were the two subjects in the RUQ of the eDISH plot?

Answer 16

Tehy bot experienced liver injuries that were judged to be at least probably due to study drug.

Question 17

Who was the first subject?

Answer 17

4053- 34 yo F from Argentina
hpatocellular injury at Day 240- with ris that began on Day 120. Consumed 8 g of Augmentin 3 months before liver injury.
“High Heel Shoe”– with rise to about 500; Bili to about 15, then rapid drop. Event consistent with signature presentation, but resolution was a bit more rapid. Event: probable.

Question 18

Who was the second subject?

Answer 18

2738- 45 yo F from Japan. First episode 4-5 months after treatment and was resolving when she exerpienced a second episode at 7 months.Treated with predinsoe during second peak; bleedng into liver cysts. Event: probably due to drug. “Double Peak, deep valley- peak ALT 700, peak BT around 7.

Question 19

Who was the third subject?

Answer 19

4301- 44 yo F from France. Developed nausea and abdomina pain and was hospitalied with hepatocellular linjury after three monhts on tav. Liver biopsy: cytoltyic hepatitis suggestive of DILI. Event: highly likely. “Double peak, shallow valey post discontinuation.

Question 20

What is the signature presentation?

Answer 20

1. In addition to a characteristic onset between 3 and 14 months.
2. Hepatocellular injury (R>5)
3. Continued rise in serum ALT after discontinuation for weeks to months followed by gradual resolution after 1-4 months.

Question 21

Were any patients rechallenged? (p. 24)

Answer 21

Two: 2401 a 50 yo F from Japan and 0605 and 49 yo Female. Rechallenge precipitated increases and confirmed that the event was due to tolvaptan treatment.