Biomarker Evaluation and Development

Specificity and sensitivity

Proper evaluation of biomarkers requires their application by many independent investigators using hundreds of compounds causing a wide range of toxicologic pathologies in the major organs of multiple species. This level of experience reveals the strengths and weaknesses of any biomarker, and builds confidence in our use of the biomarker to make decisions about drug safety. To begin this evaluation process for DMBT1 as a biomarker of BDH, we measured DMBT1 transcript levels in liver tissues from rats treated with compounds that

DMBT1

Bile duct

Average

DMBT1

DMBT1

Compound

proliferation

Fold-

Std

Fold

(Histology

Change

Deviation

Change

score)

(Group)

(Group)

(Individual)

Control

0

0.9

0.1

1.0

0

0.8

0

1.0

1

3

2.7

0.6

3.0

1.9

3.1

2

1 1 1

2.2

0.9

2.0

3.1

1.4

3

0

1.4

0.3

1.1

0

1.7

0

1.3

4

0

1.0

0.1

0.9

0

1.0

0

1.1

5

2 2 3

3.9

0.7

4.8

3.4

3.6

6

0

1.3

0.2

1.5

0

1.1

0

1.2

Figure 6. Quantitation of DMBT1 transcript in livers of rats administered agents that cause hepatocellular necrosis and apoptosis. Compounds that also caused bile duct hyperplasia are indicated, along with histological assessment of severity of BDH.

Figure 6. Quantitation of DMBT1 transcript in livers of rats administered agents that cause hepatocellular necrosis and apoptosis. Compounds that also caused bile duct hyperplasia are indicated, along with histological assessment of severity of BDH.

caused a range of hepatotoxicities. Of six compounds from one lead optimization program, three caused hepatocellular necrosis, vacuolation, and bile duct hyperplasia, and three caused only hepatocellular necrosis and vacuolation. The severity of BDH and magnitude of transcriptional induction of DMBT1 after 4 days exposure showed that all compounds causing BDH also caused greater than 2-fold increases in the level of DMBT1 transcript, whereas those compounds that did not cause BDH did not increase transcription of DMBT1 (Fig.6). These data show not only that DMBT1 induction correlates with BDH, but also that DMBT1 is not induced by more common toxicologic pathologies of the liver, including hepatocellular necrosis and vacuolation.

DMBT1 transcript levels were determined for a second set of compounds to assess the effect of phospholipidosis (PL) on DMBT1 transcription. All six of these compounds caused PL in livers of rats exposed in vivo. Of these, the development of three of these compounds was terminated in lead optimization because they induced PL at exposures that were too similar to those that cause efficacy in rodent models. None of these compounds was known to cause BDH, and therefore were chosen as "negative controls" that should cause hepatotoxicity in the form of PL but should not induce DMBT1. We were surprised to identify a compound that increased DMBT1 transcription (Fig.7). Upon more focused histologic examination of the livers, it became apparent that of the six compounds, this compound was the only one that caused BDH. Interestingly, this compound has been studied extensively for its ability to induce PL (Rorig, Ruben, and Anderson, 1987), but to our knowledge there are no descriptions in the literature reporting

Figure 7. Quantitation of DMBT1 transcript levels in livers of rats administered agents that cause phospholipidosis. Tilorone has not previously been reported to cause bile duct hyperplasia.

Figure 7. Quantitation of DMBT1 transcript levels in livers of rats administered agents that cause phospholipidosis. Tilorone has not previously been reported to cause bile duct hyperplasia.

its ability to cause BDH. This unexpected event bolstered our confidence in the value of DMBT1 as a biomarker of BDH.

Finally, we examined hepatic DMBT1 transcript levels in rats treated with eight different kinase inhibitors. These compounds have a common pharmacological target, inhibition of which causes a dramatic increase in hepatocyte mitotic index. The key question we asked was whether DMBT1 would be induced in hepatocellular proliferation, or whether there was specificity for cholangiocytes and oval cells. None of the eight compounds caused a significant increase in DMBT1 transcript level (data not shown), consistent with the hypothesis that DMBT1 is specifically involved in proliferation and/or differentiation of biliary epithelial cells.

Streamlining biomarker development

Development of DMBT1 as a biomarker of BDH is a successful case study in the application of toxicogenomics for biomarker discovery. The objective of developing a serum-based biomarker of BDH was not achieved, however, in spite of evidence that DMBT1 is a secreted protein (Sasaki et al., 2002). As occurs in drug development itself, where the majority of candidates selected during lead optimization will ultimately fail, we can expect that the majority of candidate biomarkers will also fail to meet our validation criteria. Biomarker efforts are therefore more likely to be successful if we think in terms of developing biomarker panels, rather than individual biomarkers (Ryan, Watson, and Berridge, 2004), and thus DMBT1 is just one of a set of markers being developed to detect oval cell proliferation.

The ability to move quickly from biomarker discovery to biomarker evaluation is an important attribute of a successful biomarker development strategy. A key resource to support these biomarker evaluations are sera and tissues banked from toxicology studies that are conducted routinely by pharmaceutical companies. In the case of DMBT1, livers of rats treated with greater than 30 compounds that caused hepatotoxicity in lead optimization studies helped us to quickly evaluate the specificity and sensitivity of DMBT1 as a biomarker of BDH. The value of specimens such as the rat livers used in these studies is largely attributable to the data sets associated with them, including pharmacology, clinical pathology, histology, and animal behavior. Convenient access to both the specimens and the associated data sets is therefore a necessary component of a successful biomarker development program.

Was this article helpful?

0 0

Post a comment