Limits...
Gene expression profiling of canine osteosarcoma reveals genes associated with short and long survival times.

Selvarajah GT, Kirpensteijn J, van Wolferen ME, Rao NA, Fieten H, Mol JA - Mol. Cancer (2009)

Bottom Line: Fifty-one transcripts were found to be differentially expressed, with common upregulation of these genes in the short survivors.The overexpressed genes in short survivors are associated with possible roles in proliferation, drug resistance or metastasis.A molecular-based method for discrimination of outcome for short and long survivors is useful for future prognostic stratification at initial diagnosis, where genes and pathways associated with cell cycle/proliferation, drug resistance and metastasis could be potential targets for diagnosis and therapy.

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Clinical Sciences of Companion Animals, Utrecht University, Yalelaan 108, 3584 CM Utrecht, The Netherlands. G.T.Selvarajah@uu.nl.

ABSTRACT

Background: Gene expression profiling of spontaneous tumors in the dog offers a unique translational opportunity to identify prognostic biomarkers and signaling pathways that are common to both canine and human. Osteosarcoma (OS) accounts for approximately 80% of all malignant bone tumors in the dog. Canine OS are highly comparable with their human counterpart with respect to histology, high metastatic rate and poor long-term survival. This study investigates the prognostic gene profile among thirty-two primary canine OS using canine specific cDNA microarrays representing 20,313 genes to identify genes and cellular signaling pathways associated with survival. This, the first report of its kind in dogs with OS, also demonstrates the advantages of cross-species comparison with human OS.

Results: The 32 tumors were classified into two prognostic groups based on survival time (ST). They were defined as short survivors (dogs with poor prognosis: surviving fewer than 6 months) and long survivors (dogs with better prognosis: surviving 6 months or longer). Fifty-one transcripts were found to be differentially expressed, with common upregulation of these genes in the short survivors. The overexpressed genes in short survivors are associated with possible roles in proliferation, drug resistance or metastasis. Several deregulated pathways identified in the present study, including Wnt signaling, Integrin signaling and Chemokine/cytokine signaling are comparable to the pathway analysis conducted on human OS gene profiles, emphasizing the value of the dog as an excellent model for humans.

Conclusion: A molecular-based method for discrimination of outcome for short and long survivors is useful for future prognostic stratification at initial diagnosis, where genes and pathways associated with cell cycle/proliferation, drug resistance and metastasis could be potential targets for diagnosis and therapy. The similarities between human and canine OS makes the dog a suitable pre-clinical model for future 'novel' therapeutic approaches where the current research has provided new insights on prognostic genes, molecular pathways and mechanisms involved in OS pathogenesis and disease progression.

Show MeSH

Related in: MedlinePlus

Kaplan-Meier survival analysis comparing long survivors (LS) and short survivors (SS) of dogs with OS. (A) Survival time (ST) and (B) disease free interval (DFI) Kaplan-Meier product limit estimate revealed differences in ST with significance of P < 0.0001 (log rank test of 36.58) and DFI with P < 0.0001 (log rank test of 28.35).
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC2746177&req=5

Figure 1: Kaplan-Meier survival analysis comparing long survivors (LS) and short survivors (SS) of dogs with OS. (A) Survival time (ST) and (B) disease free interval (DFI) Kaplan-Meier product limit estimate revealed differences in ST with significance of P < 0.0001 (log rank test of 36.58) and DFI with P < 0.0001 (log rank test of 28.35).

Mentions: The 32 dogs in this study varied in clinicopathological parameters (Table 1). The present experimental design hereby defines dogs with poor prognosis (short survivors, SS) as those surviving less than 6 months and favorable prognosis (long survivors, LS) as those dogs with survival of 6 months or longer. The overall survival times (ST) of these dogs were between 6 and 1752 days with a median value of 204 days. Dogs with poor prognosis (SS) had a ST of 6-169 days (<6 months); while the better prognosis group (LS) had a ST of 239-1752 days (>6 months). Beside the distinction in survival time between these two groups, the SS group also exhibited a significantly shorter disease free interval in a Kaplan-Meier survival analysis (Figure 1). Fisher's exact test revealed no significant differences in discrete data distribution between the 2 groups of survival with respect to sex, neuter status, histological grade and postoperative chemotherapy, the exception being metastatic disease status (P < 0.018). This is because metastatic disease was seen in all dogs in the short survivor's group while 6 dogs from the long survivors' group did not develop metastasis before death (died due to other causes); this was censored for subsequent analyses (Table 2). Tumors were located mostly at appendicular sites, but other locations such as mandible, rib, scapula, metatarsus and extraskeletal were also present. Further potential confounders identified through a univariable Cox regression analysis of the whole population (n = 32) included histological grade, postoperative chemotherapy, alkaline phosphatase (ALP) measurement at diagnosis, sex, age and neuter status (Additional file 1). Alkaline phosphatase (ALP) level and age at presentation were found to have P values of less than 0.15 and were then further subjected to multivariate analysis. Elevation of serum ALP is a known negative prognosticator in dogs with OS and this is apparently also true here, multivariate analysis using the ALP data available for the dogs in this study (n = 23) showed this confounder to be significantly associated with survival (HR = 1.005; CI: 1.000-1.009 with corresponding P value of 0.035).


Gene expression profiling of canine osteosarcoma reveals genes associated with short and long survival times.

Selvarajah GT, Kirpensteijn J, van Wolferen ME, Rao NA, Fieten H, Mol JA - Mol. Cancer (2009)

Kaplan-Meier survival analysis comparing long survivors (LS) and short survivors (SS) of dogs with OS. (A) Survival time (ST) and (B) disease free interval (DFI) Kaplan-Meier product limit estimate revealed differences in ST with significance of P < 0.0001 (log rank test of 36.58) and DFI with P < 0.0001 (log rank test of 28.35).
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC2746177&req=5

Figure 1: Kaplan-Meier survival analysis comparing long survivors (LS) and short survivors (SS) of dogs with OS. (A) Survival time (ST) and (B) disease free interval (DFI) Kaplan-Meier product limit estimate revealed differences in ST with significance of P < 0.0001 (log rank test of 36.58) and DFI with P < 0.0001 (log rank test of 28.35).
Mentions: The 32 dogs in this study varied in clinicopathological parameters (Table 1). The present experimental design hereby defines dogs with poor prognosis (short survivors, SS) as those surviving less than 6 months and favorable prognosis (long survivors, LS) as those dogs with survival of 6 months or longer. The overall survival times (ST) of these dogs were between 6 and 1752 days with a median value of 204 days. Dogs with poor prognosis (SS) had a ST of 6-169 days (<6 months); while the better prognosis group (LS) had a ST of 239-1752 days (>6 months). Beside the distinction in survival time between these two groups, the SS group also exhibited a significantly shorter disease free interval in a Kaplan-Meier survival analysis (Figure 1). Fisher's exact test revealed no significant differences in discrete data distribution between the 2 groups of survival with respect to sex, neuter status, histological grade and postoperative chemotherapy, the exception being metastatic disease status (P < 0.018). This is because metastatic disease was seen in all dogs in the short survivor's group while 6 dogs from the long survivors' group did not develop metastasis before death (died due to other causes); this was censored for subsequent analyses (Table 2). Tumors were located mostly at appendicular sites, but other locations such as mandible, rib, scapula, metatarsus and extraskeletal were also present. Further potential confounders identified through a univariable Cox regression analysis of the whole population (n = 32) included histological grade, postoperative chemotherapy, alkaline phosphatase (ALP) measurement at diagnosis, sex, age and neuter status (Additional file 1). Alkaline phosphatase (ALP) level and age at presentation were found to have P values of less than 0.15 and were then further subjected to multivariate analysis. Elevation of serum ALP is a known negative prognosticator in dogs with OS and this is apparently also true here, multivariate analysis using the ALP data available for the dogs in this study (n = 23) showed this confounder to be significantly associated with survival (HR = 1.005; CI: 1.000-1.009 with corresponding P value of 0.035).

Bottom Line: Fifty-one transcripts were found to be differentially expressed, with common upregulation of these genes in the short survivors.The overexpressed genes in short survivors are associated with possible roles in proliferation, drug resistance or metastasis.A molecular-based method for discrimination of outcome for short and long survivors is useful for future prognostic stratification at initial diagnosis, where genes and pathways associated with cell cycle/proliferation, drug resistance and metastasis could be potential targets for diagnosis and therapy.

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Clinical Sciences of Companion Animals, Utrecht University, Yalelaan 108, 3584 CM Utrecht, The Netherlands. G.T.Selvarajah@uu.nl.

ABSTRACT

Background: Gene expression profiling of spontaneous tumors in the dog offers a unique translational opportunity to identify prognostic biomarkers and signaling pathways that are common to both canine and human. Osteosarcoma (OS) accounts for approximately 80% of all malignant bone tumors in the dog. Canine OS are highly comparable with their human counterpart with respect to histology, high metastatic rate and poor long-term survival. This study investigates the prognostic gene profile among thirty-two primary canine OS using canine specific cDNA microarrays representing 20,313 genes to identify genes and cellular signaling pathways associated with survival. This, the first report of its kind in dogs with OS, also demonstrates the advantages of cross-species comparison with human OS.

Results: The 32 tumors were classified into two prognostic groups based on survival time (ST). They were defined as short survivors (dogs with poor prognosis: surviving fewer than 6 months) and long survivors (dogs with better prognosis: surviving 6 months or longer). Fifty-one transcripts were found to be differentially expressed, with common upregulation of these genes in the short survivors. The overexpressed genes in short survivors are associated with possible roles in proliferation, drug resistance or metastasis. Several deregulated pathways identified in the present study, including Wnt signaling, Integrin signaling and Chemokine/cytokine signaling are comparable to the pathway analysis conducted on human OS gene profiles, emphasizing the value of the dog as an excellent model for humans.

Conclusion: A molecular-based method for discrimination of outcome for short and long survivors is useful for future prognostic stratification at initial diagnosis, where genes and pathways associated with cell cycle/proliferation, drug resistance and metastasis could be potential targets for diagnosis and therapy. The similarities between human and canine OS makes the dog a suitable pre-clinical model for future 'novel' therapeutic approaches where the current research has provided new insights on prognostic genes, molecular pathways and mechanisms involved in OS pathogenesis and disease progression.

Show MeSH
Related in: MedlinePlus