Periostin levels linked to osteoarthritis prevalence, progression

PHILADELPHIA – Women with prevalent knee osteoarthritis had significantly lower levels of periostin, particularly those with progressive disease, than did women without the condition in an analysis of the OFELY study.

Although the mechanism remains to be investigated, the vitamin-K dependent protein "may be a new biological marker of prevalence and progression of knee OA in women," Jean-Charles Rousseau, Ph.D., said at the meeting, sponsored by Osteoarthritis Research Society International.

Periostin is secreted by osteoblasts and chondrocytes, and is one of only a handful of gamma-carboxyglutamic acid-containing proteins so far identified in humans. It was recently shown to be overexpressed in human OA cartilage and to regulate cartilage metabolism via matrix metallopeptidase-13 activation (Osteoarthritis Cartilage 2011;19 [Suppl. 1]:S34).

OFELY (Os des Femmes de Lyon) is the first study to report an association between periostin OA prevalence and progression, Dr. Rousseau said at the meeting.

To evaluate its role in OA, the investigators used a new commercially available ELISA sandwich assay to evaluate blood samples from 594 women in the ongoing, prospective OFELY study. Of these, 511 had no knee OA and 83 had radiographic knee OA, Kellgren-Lawrence grade of 2 or more. The average age was 60 years in those free of OA and 70 in those with OA; 78% and 98%, respectively, were menopausal.

Baseline serum periostin averaged 1,101 ng/mL in women with OA vs. 1,181 ng/mL in those without OA, after researchers adjusted for age (P = .011), said Dr. Rousseau of INSERM Research Unit 1033, University of Lyon (France).

Collagen type II cross-linked C-telopeptide levels (CTX-II) were significantly higher in women with OA vs. those without (278 ng/mL vs. 173 ng/mL; P less than .001), whereas serum levels were similar for type I CTX (0.43 ng/mL vs. 0.39 ng/mL), intact N-terminal propeptide of type I procollagen (38.1 mcg/L vs. 37.2 mcg/L) and osteocalcin (28.4 ng/mL vs. 25.7 ng/mL).

During the 4-year follow-up period, 82 women had radiological progression of their knee OA, defined by an increase of 1 or more Kellgren-Lawrence grade regardless of the baseline grade.

Baseline periostin levels were significantly lower in progressors than nonprogressors, after adjustment for age (1,112 ng/mL vs. 1,179 ng/mL; P = .046), Dr. Rousseau said.

For each increase of one periostin quartile, the risk of progression decreased by 21% after adjusting for age, KL score at baseline, and OA at other anatomical sites, including hand, hip, and lumbar spine (odds ratio 0.79; P = .038), he said.

During a discussion of the paper, an attendee questioned whether periostin is exclusive to bone metabolites, highlighting a recent randomized trial indicating that it is highly relevant in lung function. In that trial (N. Engl. J. Med. 2011;365:1088-98), anti-interleukin-13 therapy was more effective in asthmatic adults who had high serum periostin levels.

Dr. Rousseau acknowledged that periostin is distributed in nonarticular tissues including periodontal ligament and cardiac valves, and that the specificity of the periostin assay for circulating forms is unknown.

Session moderator Ali Mobasheri, D.Phil., of the University of Nottingham, England, said in an interview that it can be difficult to differentiate between a biomarker emanating from bone, articular cartilage, and tendon and other tissues such as heart and lung, "and in this particular case, I think it’s going to be extremely difficult to differentiate where it’s coming from."

He went on to say that periostin may be differentially processed in different tissues, which could be the differentiator in the future, but that the Chinese assay (USCNK, China) used in the study doesn’t make that distinction.

"So a lot more validation is required in this case," Dr. Mobasheri said.

Dr. Rousseau said in an interview that overlap exists with osteoarthritis biomarkers, but that they remain useful, even when OA is limited to a single knee.

OFELY was sponsored by Prévention des Maladies Osseuses. Dr. Rousseau reported no relevant disclosures.

pwendling@frontlinemedcom.com

PHILADELPHIA – Women with prevalent knee osteoarthritis had significantly lower levels of periostin, particularly those with progressive disease, than did women without the condition in an analysis of the OFELY study.

Although the mechanism remains to be investigated, the vitamin-K dependent protein "may be a new biological marker of prevalence and progression of knee OA in women," Jean-Charles Rousseau, Ph.D., said at the meeting, sponsored by Osteoarthritis Research Society International.

Periostin is secreted by osteoblasts and chondrocytes, and is one of only a handful of gamma-carboxyglutamic acid-containing proteins so far identified in humans. It was recently shown to be overexpressed in human OA cartilage and to regulate cartilage metabolism via matrix metallopeptidase-13 activation (Osteoarthritis Cartilage 2011;19 [Suppl. 1]:S34).

OFELY (Os des Femmes de Lyon) is the first study to report an association between periostin OA prevalence and progression, Dr. Rousseau said at the meeting.

To evaluate its role in OA, the investigators used a new commercially available ELISA sandwich assay to evaluate blood samples from 594 women in the ongoing, prospective OFELY study. Of these, 511 had no knee OA and 83 had radiographic knee OA, Kellgren-Lawrence grade of 2 or more. The average age was 60 years in those free of OA and 70 in those with OA; 78% and 98%, respectively, were menopausal.

Baseline serum periostin averaged 1,101 ng/mL in women with OA vs. 1,181 ng/mL in those without OA, after researchers adjusted for age (P = .011), said Dr. Rousseau of INSERM Research Unit 1033, University of Lyon (France).

Collagen type II cross-linked C-telopeptide levels (CTX-II) were significantly higher in women with OA vs. those without (278 ng/mL vs. 173 ng/mL; P less than .001), whereas serum levels were similar for type I CTX (0.43 ng/mL vs. 0.39 ng/mL), intact N-terminal propeptide of type I procollagen (38.1 mcg/L vs. 37.2 mcg/L) and osteocalcin (28.4 ng/mL vs. 25.7 ng/mL).

During the 4-year follow-up period, 82 women had radiological progression of their knee OA, defined by an increase of 1 or more Kellgren-Lawrence grade regardless of the baseline grade.

Baseline periostin levels were significantly lower in progressors than nonprogressors, after adjustment for age (1,112 ng/mL vs. 1,179 ng/mL; P = .046), Dr. Rousseau said.

For each increase of one periostin quartile, the risk of progression decreased by 21% after adjusting for age, KL score at baseline, and OA at other anatomical sites, including hand, hip, and lumbar spine (odds ratio 0.79; P = .038), he said.

During a discussion of the paper, an attendee questioned whether periostin is exclusive to bone metabolites, highlighting a recent randomized trial indicating that it is highly relevant in lung function. In that trial (N. Engl. J. Med. 2011;365:1088-98), anti-interleukin-13 therapy was more effective in asthmatic adults who had high serum periostin levels.

Dr. Rousseau acknowledged that periostin is distributed in nonarticular tissues including periodontal ligament and cardiac valves, and that the specificity of the periostin assay for circulating forms is unknown.

Session moderator Ali Mobasheri, D.Phil., of the University of Nottingham, England, said in an interview that it can be difficult to differentiate between a biomarker emanating from bone, articular cartilage, and tendon and other tissues such as heart and lung, "and in this particular case, I think it’s going to be extremely difficult to differentiate where it’s coming from."

He went on to say that periostin may be differentially processed in different tissues, which could be the differentiator in the future, but that the Chinese assay (USCNK, China) used in the study doesn’t make that distinction.

"So a lot more validation is required in this case," Dr. Mobasheri said.

Dr. Rousseau said in an interview that overlap exists with osteoarthritis biomarkers, but that they remain useful, even when OA is limited to a single knee.

OFELY was sponsored by Prévention des Maladies Osseuses. Dr. Rousseau reported no relevant disclosures.

pwendling@frontlinemedcom.com

PHILADELPHIA – Women with prevalent knee osteoarthritis had significantly lower levels of periostin, particularly those with progressive disease, than did women without the condition in an analysis of the OFELY study.

Although the mechanism remains to be investigated, the vitamin-K dependent protein "may be a new biological marker of prevalence and progression of knee OA in women," Jean-Charles Rousseau, Ph.D., said at the meeting, sponsored by Osteoarthritis Research Society International.

Periostin is secreted by osteoblasts and chondrocytes, and is one of only a handful of gamma-carboxyglutamic acid-containing proteins so far identified in humans. It was recently shown to be overexpressed in human OA cartilage and to regulate cartilage metabolism via matrix metallopeptidase-13 activation (Osteoarthritis Cartilage 2011;19 [Suppl. 1]:S34).

OFELY (Os des Femmes de Lyon) is the first study to report an association between periostin OA prevalence and progression, Dr. Rousseau said at the meeting.

To evaluate its role in OA, the investigators used a new commercially available ELISA sandwich assay to evaluate blood samples from 594 women in the ongoing, prospective OFELY study. Of these, 511 had no knee OA and 83 had radiographic knee OA, Kellgren-Lawrence grade of 2 or more. The average age was 60 years in those free of OA and 70 in those with OA; 78% and 98%, respectively, were menopausal.

Baseline serum periostin averaged 1,101 ng/mL in women with OA vs. 1,181 ng/mL in those without OA, after researchers adjusted for age (P = .011), said Dr. Rousseau of INSERM Research Unit 1033, University of Lyon (France).

Collagen type II cross-linked C-telopeptide levels (CTX-II) were significantly higher in women with OA vs. those without (278 ng/mL vs. 173 ng/mL; P less than .001), whereas serum levels were similar for type I CTX (0.43 ng/mL vs. 0.39 ng/mL), intact N-terminal propeptide of type I procollagen (38.1 mcg/L vs. 37.2 mcg/L) and osteocalcin (28.4 ng/mL vs. 25.7 ng/mL).

During the 4-year follow-up period, 82 women had radiological progression of their knee OA, defined by an increase of 1 or more Kellgren-Lawrence grade regardless of the baseline grade.

Baseline periostin levels were significantly lower in progressors than nonprogressors, after adjustment for age (1,112 ng/mL vs. 1,179 ng/mL; P = .046), Dr. Rousseau said.

For each increase of one periostin quartile, the risk of progression decreased by 21% after adjusting for age, KL score at baseline, and OA at other anatomical sites, including hand, hip, and lumbar spine (odds ratio 0.79; P = .038), he said.

During a discussion of the paper, an attendee questioned whether periostin is exclusive to bone metabolites, highlighting a recent randomized trial indicating that it is highly relevant in lung function. In that trial (N. Engl. J. Med. 2011;365:1088-98), anti-interleukin-13 therapy was more effective in asthmatic adults who had high serum periostin levels.

Dr. Rousseau acknowledged that periostin is distributed in nonarticular tissues including periodontal ligament and cardiac valves, and that the specificity of the periostin assay for circulating forms is unknown.

Session moderator Ali Mobasheri, D.Phil., of the University of Nottingham, England, said in an interview that it can be difficult to differentiate between a biomarker emanating from bone, articular cartilage, and tendon and other tissues such as heart and lung, "and in this particular case, I think it’s going to be extremely difficult to differentiate where it’s coming from."

He went on to say that periostin may be differentially processed in different tissues, which could be the differentiator in the future, but that the Chinese assay (USCNK, China) used in the study doesn’t make that distinction.

"So a lot more validation is required in this case," Dr. Mobasheri said.

Dr. Rousseau said in an interview that overlap exists with osteoarthritis biomarkers, but that they remain useful, even when OA is limited to a single knee.

OFELY was sponsored by Prévention des Maladies Osseuses. Dr. Rousseau reported no relevant disclosures.

pwendling@frontlinemedcom.com