Chronic inflammation in myeloproliferative neoplasms (MPNs) is characterized by persistent connective tissue remodeling, which leads to organ dysfunction and ultimately, organ failure, due to excessive accumulation of extracellular matrix (ECM). The connective tissue responds uniformly to injuries of any kind by distinctive sequential changes in the ECM expression, including oedema formation, angiogenesis and finally, fibrosis, with the deposition of type III collagen in the early phase, mainly as fine fibers, and type I collagen as coarse fibers in the later phase of the lesion). This injury–repair process is qualitatively similar in all organs and is accompanied by the release of various matrix components into the circulation during the synthesis and breakdown of connective tissue constituents at the site of injury.
Biomarker | No. Patients | Summary of Results | Conclusions/Comments | Refs. |
---|---|---|---|---|
S-PIIINP | 441 | S-PIIINP values were increased in PV with even more elevated levels in post-PV-MF and strikingly elevated in patients with severe myelofibrosis [81][1]. | S-PIIINP associates significantly with the extent of reticulin fibrosis. S-PIIINP is a quantitative marker for myelofibrosis [81][1]; |
[81,83,84,85,86,[1][3][4][5][6] 96,98,261,[10][48][66] 262,268,272,276][11][17][25][34] |
S-PIIINP values were normal in patients without reticulin fibrosis; increased in PV and MF. S-PIIINP values above 25 ng/mL associated with MF of recent onset (less than or equal to 2 years) and values below 25 ng/mL with MF of more than 4 years’ duration [83][3]. |
S-PIIINP is a non-invasive method for accurate assessment of bone marrow fibrosis. S-PIIIINP may be used to evaluate the efficacy of antifibrosing agents [83][3]. |
|||
S-PIIINP values were increased in PV and related to degree of reticulin fibrosis. S-PIIINP values were increased in spent phase of PV only treated with phlebotomy. S-PIIINP values were increased in patients transforming into post-PV-MF and increased in PMF [86][6]. |
S-PIIINP is higher in myelofibrosis of recent onset (less than 2 years) than in myelofibrosis longer than 2 years. S-PIIINP is stable in PV patients treated with 32P or hydroxyurea [86][6]. |
|||
S-PIIINP values were increased in PV and even more in patients with TMD and MF; S-PIIIINP values were virtually normal in OMS (deposition of type I collagen) [84][4]. | S-PIIINP is a useful indicator of disease activity in MPNs S-PIIINP positively correlates to the degree of reticulin fibrosis. Near normal S-PIIINP values in OMS likely reflect stable disease without concurrent type III collagen synthesis. [84][4]. |
|||
S-PIIINP values were normal or elevated in PV and TMD. S-PIIINP values were normal or even low levels in OMS. S-PIIINP values were increased in MF and in CML associated with bone marrow MF. S-PIIINP and S-Type IV collagen correlated significantly with each other and with the leucocyte count [85][5]. |
S-PIIINP is a useful indicator of disease activity in MPNs. Normal and even low S-PIIINP values in OMS may reflect stable disease without ongoing type III collagen synthesis Interstitial type III collagen and basement membrane metabolism are closely related [85][5]. |
|||
S-PIIINP values were strongly raised in MPNs. All three biomarkers (S-PIIINP, S-PICP and S-Laminin ) were significantly elevated in patients with active/transforming disease [98][66]. |
S-PIIINP is a useful indicator of disease activity in MPNs. S-Laminin and S-PICP do not offer offer any advantage over S-PIIIP. Interstitial type III collagen and basement membrane metabolism are closely related [98][66]. |
|||
S-PIIINP values were slightly elevated in patients with stable disease and highly elevated in patients transforming into myelofibrosis. S-PIIINP values covariated closely with S-laminin, the leucocyte count and LDH [276][34]. |
S-PIIINP is a useful indicator of disease activity in MPNs. Interstitial type III collagen and basement membrane metabolism are closely related [276][34]. |
|||
S-PIIINP values correlated significantly with the leukocyte count and with S-HU. S-PIIINP values decreased during cytotoxic treatment in concert with declining leukocyte counts and S-HU [96][48]. |
S-PIIINP is a useful indicator of disease activity in MPNs. S-PIIINP may be useful in monitoring the efficacy of cytotoxic treatment in terms of inhibiting development and progression of bone marrow fibrosis [96][48]. |
|||
S-PIIINP values were increased in PMF. S-PIIINP values were only increased slightly in patients with stable disease (n = 3) compared to the single patient with more active disease. S-PIIINP values declined during treatment with acetylsalicylic acid (ASA), although normalization did not occur. Using gel filtration analysis the antigens related to S-PIIIINP were found to be heterogenous with at least two peaks, exhibiting molecular masses equal to and smaller than PIIINP [261][10]. |
This study includes only four patients and accordingly does not allow robust conclusions [261][10]. Studies on the impact of ASA and anti-inflammatory treatment upon type III collagen metabolism are needed [261][10]. |
|||
S-PIIIP values were elevated in PMF. S-PIIIP values were normal in younger patients, having higher Hb- and platelet counts and lower S-ferritin values platelet count. S-PIIIP values were significantly higher in patients with active disease (fever, sweating, weight loss) than in patients with non-active disease S-PIIIP values correlated with decreasing Hb-concentration and platelet count and increasing WBC, serum ferritin and number of transfusions (univariate analysis). S-PIIINP values correlated independently with increasing WBC, serum ferritin and age (multivariate analysis). S-PIIINP values did not associate with morphometric grading of bone marrow fibrosis, megakaryocyte number, or lymphoid infiltration [262][11] |
S-PIIIP values in PMF correlates more with overall disease activity than with the extent of bone marrow fibrosis [262][11]. The association between normal S-PIIINP and lower S-ferritin values in younger patients with higher HB-concentrations may likely reflect that S-PIIINP also is a biomarker of the chronic inflammatory state in PMF. Studies on the associations between S-PIIINP, biomarkers of chronic inflammation (e.g., CRP, ferritin, inflammatory cytokines), bone marrow megakaryocyte morphology and bone marrow fibrosis and the impact of cytotoxic (HU) or stem-cell targeting therapy ( pegylated interferon-alpha2 ) as monotherapies or in combination (e.g., with JAK1-2 inhibitor) are needed. |
|||
S-PIIINP values were elevated in PMF, S-PIIINP values decreased during treatment with anthracycline, which was given due to accelerated phase disease [268][17]. |
S-PIIINP is a valuable biomarker in PMF. Cytotoxic treatment lowers elevated S-PIIINP values [268][17]. |
|||
S-PIIINP values were elevated in MPNs. S-PIIINP values were highest in patients with MF. S-PIIINP and S-ICTP correlated significantly. S-PIIIINP and P-suPAR correlated significantly. S-PIIINP values did not correlate with P-MMP-2 and MMP-9 [272][25]. |
S-PIIINP is a useful indicator of disease activity in MPNs. Type III and type I collagen metabolism are closely associated, reflecting concurrent type III synthesis (PIIINP) and type I degradation (ICTP). Elevated S-ICTP values in MPN may not only reflect type I collagen degradation in the bone marrow but also increased bone resorption. Enzymes of the uPA system might participate in the bone marrow remodelling processes characteristic of MPN [272][25]. |
|||
S-PICP | 26 | S-PICP values were slightly elevated in MPNs, reflecting increased type I collagen synthesis. S-PICP values were significantly elevated in patients with active/transforming disease. S-PICP and S-laminin P1 values showed a strong correlation [98][66]. |
S-PICP values do not offer any advantage over S-PIIIP for monitoring of disease activity. Increased type I collagen synthesis associates with progressive disease. Interstitial type I collagen and basement membrane metabolism are closely related [98][66]. |
[98][66] |
S-ICTP | 50 | S-ICTP values were elevated in MPN. S-ICTP values were only significantly higher among MF and PV patients. S-ICTP and S-PIIINP values correlated significantly. S-ICTP and P-suPAR values correlated significantly. S-ICTP and P-MMP-2/MMP-9 values did not correlate significantly. |
Elevated S-ICTP values in MPNs reflect ongoing type I collagen degradation. Elevated S-ICTP values may not only reflect enhanced type I collagen degradation in the bone marrow but also increased type I collagen degradation in bone tissue (increased bone resorption). Increased bone resorption with the development of osteopenia/osteoporosis may be driven by chronic inflammation in MPNs. In this regard, the significant correlation between S-ICTP and P-suPAR may also reflect the chronic inflammatory state and not only the involvement of these biomarkers in the bone marrow remodeling processes in MPNs. |
[272][25] |
S-TIVC | 41 | S-TIVC values were normal or elevated in PV and TMD. S-TIVC values were elevated in MF and in CML associated with bone marrow MF. S-TIVC and S-PIIINP correlated significantly and with the leucocyte count. |
Measurement of type IV collagen provides a noninvasive means for following the accumulation of basement membrane collagen in the bone marrow in patients with MPN. S-TIVC associates with disease activity as assessed by the leukocyte count. Interstitial (type III collagen ) and basement membrane metabolism (type IV collagen ) are tightly associated processes in MPNs. |
[85][5] |
S-Laminin1 | 58 | S-Laminin1 values were slightly elevated in MPNs S-Laminin1, S-PIIINP and S-PICP values were significantly elevated in patients with active/transforming disease. S-Laminin P1 and S-PICP levels showed a strong correlation [98][66]. |
S-Laminin did not appear to offer any advantage over S-PIIIP for the monitoring of disease activity. Basement membrane (laminin) and interstitial collagen ( PIIINP, PICP) metabolism are closely related in MPNs [98][66]. |
[98,276][34][66] |
S-Laminin1 values were slightly elevated in patients with stable disease. S-Laminin1 values were highly elevated in patients with progressive disease transforming into myelofibrosis. S-Laminin1 covariated closely with S-PIIINP, the leucocyte count and LDH [276][34]. |
S-Laminin1 values were significantly lower in patients with a huge spleen as compared with patients, having a normal spleen size or previously being splenectomized.The above observation may reflect that the aminin uptake/degradation is increased in the enlarged spleen in MPNs. | |||
S-HYA | 59 | S-HYA values were normal in patients with stable disease and increased in patients with active disease. S-HYA values correlated significantly with the leukocyte count and with S-PIIINP. S-HYA values decreased during cytotoxic treatment in concert with declining leukocyte counts and S-PIIINP. |
S-HYA values displayed only slight changes in MPNs with with frequent overlaps between patient categories and HC. The clinical utility of S-HYA may be restrained, although sequential measurements of S-HYA may provide a reflection of the MPN process in individual patients. | [96][48] |
P-Fibronectin | 69 | P-Fibronectin values were normal in ET. P-Fibronectin values were significantly reduced in PV and MF. P-Fibronectin values were lowest in in patients with marked splenomegaly. P-Fibronectin values rose from less than 100 mg/L to 177 mg/L after splenectomy in a patients with MF [82][2]. |
Low P-fibronectin values in MPNs may be attributed to increased consumption of P-Fibronectin in the expanded mononuclear phagocyte system in the liver and spleen, reduced hepatic synthesis, and/or fibronectin taking part in the clearance of circulating immune complexes. Low P-Fibronectin values in patients with MPNs may contribute to an increased risk of infections. |
[82,88,90][2][37][58] |
P-Fibronectin values were significantly lower in patients with PMF. P-Fibronectin values in MF patients differed significantly, when compared with patients with PV, TMD or CML. P-Fibronectin values were lowest in patients with large spleens [88][37]. |
Low P-fibronectin concentrations in splenomegalic patients may be due to enhanced consumption of the opsonin in the expanded splenic mononuclear-macrophage system [88][37]. | |||
P-Fibronectin values correlated inversely with CIC, which were highly elevated in 11 of 20 with MF secondary to MPN. The CIC contained fibronectin, IgG and C3. P-Fibronectin levels increased after therapeutic plasmapheresis, which efficiently removed CIC [90][58]. |
The findings suggest that fibronectin as a major non-specific opsonin is important for the normal clearance of immune complexes [90][58]. | |||
P-Fibronectin (EDA) |
122 | P-EDA FN values were significantly elevated in PMF as compared to HCs. P-EDA F values did not differ between PV/ET patients and HCs. P-EDA FN values differed among patients with different degrees of BM fibrosis with a trend towards increasing P-EDA FN levels with increasing BM fibrosis grades (not significant). P-EDA FN differed significantly between patients with pre-fibrotic myelofibrosis (BM fibrosis grade 0) + those with early myelofibrosis (BM fibrosis grade 1) as compared to those with BM fibrosis grade 2 + those with BM fibrosis grade 3 (advanced fibrosis) [297][61]. |
Patients with PMF exhihited higher levels of the EDA FN isoform as compared to HCs. | [297,298][61][62] |
P-EDA FN values were higher in patients with a homozygous JAK2V617F genotype Increased P-EDA-FN values were associated with anemia, elevated high-sensitivity C-reactive protein, bone marrow fibrosis and splanchnic vein thrombosis at diagnosis. Elevated P-EDA-FN at diagnosis was a predictor of large splenomegaly [298][62]. |
P-EDA-FN in primary myelofibrosis may represent a marker of disease progression, and a novel target to treat splenomegaly [298][62]. | |||
P-YKL-40 | 48 | P-YKL-40 values were significantly elevated in PMF vs. control subjects. P-YKL-40 values were increased from ET over PV to PMF [302][67]. |
P-YKL-40 may be a novel biomarker of chronic inflammation, tissue remodelling and atherosclerotic inflammation in MPN [302,303][67][68]. |
[302,303][67][68] |
P-YKL-40 values were significantly elevated in PMF vs. control subjects. P-YKL-40 values were increased from ET over PV to PMF [302][67]. |
P-YKL-40 might be a novel marker of disease burden and progression in MPN [303][68]. |
|||
S-YKL-40 | 111 | S-YKL-40 values were significantly higher in post-ET MF, PV, post-PV MF and PMF patients, when compared to HC. S-YKL-40 values were associated with biomarkers of an increased inflammatory state (higher C-reactive protein, poor performance status, presence of constitutional symptoms and cardiovascular risk factors).Higher S-YKL-40 values in MF patients were also associated with blast phase disease, lower hemoglobin and higher Dynamic International Prognostic Scoring System score. Higher S-YKL-40 values were independently associated with an increased risk of thrombosis and impaired survival in MF patients [304][69]. |
Higher S-YKL-40 might have a pathophysiological role in disease progression and thrombosis development. Assessing S-YKL-40 could help in identification of ET and PV patients at a high risk of future cardiovascular events and has a good potential for improving prognostication of MF patients [304][69]. |
[304][69] |
S-CHIT1 | 91 | S-CHIT1 values were significantly higher in PV and post-PV myelofibrosis transformation (post-PV MF). S-CHIT1 values were not significantly higher in ET, post-ET MF transformation, and PMF patients, when compared to healthy controls. S-CHIT1 values in PV were positively correlated with hemoglobin, hematocrit, absolute basophil count and the presence of reticulin fibrosis in the bone marrow. |
A positive correlation between S-CHIT1 and the hemoglobin, hematocrit, and absolute basophil count in PV might imply macrophages closely related to clonal erythropoiesis as cells of CHIT1 origin. A positive association between S-CHIT1 and reticulin fibrosis might indicate its potential role in PV progression. S-CHIT1 might a circulating biomarker of macrophage activation with an important role in inflammation-induced tissue remodeling and fibrosis in PV. |
[305][70] |
P-Pentraxin-3 (P-PTX3) |
244/477/140 | P-PTX3 and P-hs-CRP were measured in 244 consecutive ET and PV patients. After a median follow up of 5.3 years (range 0–24), 68 CV events were diagnosed. Major thrombosis rate was higher in the highest hsCRP and lower at the highest PTX3 levels. These associations remained significant in multivariate analyses [306][71]. |
P-hs-CRP and P-PTXT3 independently and in opposite ways modulate the intrinsic risk of CV events in patients with MPN [306][71]. | [306,307][71][72] [308][73] |
P-PTX3 levels in 477 ET and PV patients were significantly increased in carriers of homozygous JAK2V617F mutation compared to all the other genotypes and triple negative ET patients, while hs-CRP levels were independent of the mutational profile. The risk of hematological evolution and death from any cause was about 2- and 1.5-fold increased in individuals with high PTX-3 levels, while the thrombosis rate tended to be lower. High hs-CRP levels were associated with risk of haematological evolution, death and also major thrombosis. After sequential adjustment for potential confounders (age, gender, diagnosis and treatments) and the presence of JAK2V617F homozygous status, high hs-CRP levels remained significant for all outcomes, while JAK2V617F homozygous status as well as treatments were the factors independently accounting for adverse outcomes among patients with high PTX3 levels [307][72]. |
The JAK2V617F mutation influences MPN-associated inflammation with a strong correlation between allele burden and PTX3 levels. P-hs-CRP and P-PTX3 might be of prognostic value for patients with ET and PV, but their validation in future prospective studies is needed [307][72]. | |||
P-PTX3 values were significantly higher in PMF patients than in HC. High PTX3 values (≥70 ng/mL) associated with an unfavourable overall survival. P-PTX3 values independently predicted PMF patients’ overall survival. P-PTX3 values correlated with parameters of tumor burden, including total leucocyte count, mutated JAK2 allele burden, lactate dehydrogenase levels, and spleen size [308][73]. |
PTX3is released from macrophages and endothelial cells, and promotes the transition of monocytes to fibrocytes. P-PTX3 levels constitute an independent indicator of disease burden, clonal expansion and overall survival in patients with PMF. Monitoring of PTX3 plasma levels might be a useful tool in clinical decision making [308][73]. |
|||
P-suPAR | 50 | P-suPAR correlated significantly with serum markers of collagen metabolism (S-PIIINP and S-ICTP) [272][25] | Enzymes of the uPA system might participate in the bone marrow remodelling processes characteristic of MPN [272][25]. | [272,273][25][26] |
P-suPAR values were significantly higher in MPN patients. P-suPAR values differed significantly between MPN-subgroups, the highest levels being found in patients with MF and PV. P-suPAR values were only significantly increased in PV and MF patients. P-suPAR significantly correlated to P-LDH. P-uPA did not differ between patients and controls [273][26]. |
Increased P-suPAR levels in patients with MPN may reflect increased uPAR production in the bone marrow, leading to enhanced bone marrow remodeling [273][26]. | |||
P-TIMP-1 P-MMP-1 P-MMP-2 P-MMP-3 P-MMP-9 |
67 | P-MMP-3 levels were decreased in patients with advanced MF. P-MMP-1, P-MMP-2, and P-MMP-9 levels were not significantly different from HC. P-TIMP-1 levels were elevated in ET, PV and MF patients and in particular in advanced MF. P-MMPs levels were not elevated in ET and PV patients |
The abnormal accumulation of ECM is dependent upon the balance between matrix metalloproteinases (MMPs) and tisse inhibitors of metalloproteinases (TIMPs). Accumulation of connective tissue in the bone marrow is associated with reduced MMP activity together with increased TIMP-1 activity, which may be important in fibrosis formation in the bone marrow in MPNs. |
[299,300][63][64] |
P-TIMP-1 | 50 | Plasma levels of total-, free- and complexed TIMP-1, TIMP-2, MMP-2 and MMP-9 were measured in 50 patients with MPN. P-TIMP-1 levels were significantly higher in MPN patients. P-TIMP-1 levels significantly correlated with P-suPAR and P-uPAR. P-TIMP-2 and P-MMP-2 levels did not differ beween patients and controls. P-TIMP-1 and P-TIMP-2 levels correlated significantly. P-MMP-9 levels significantly higher among PV patients. P-TIMP-1/MMP-9 ratio was significantly higher in patients with MF. |
The family of MMPs and TIMPs facilitate and inhibit matrix degradation processes, respectively. A disturbed TIMP-1/MMP ratio may reflect an imbalance of the extracellular homeostasis towards an increased matrix deposition promoting fibrosis. |
[275][30] |
U-HYPRL | 47 | U-HYPRL was normal in 16 patients with PMF and in 5 out of 6 patients with acute myelofibrosis. In patients with OMS (n = 8) values for U-HYPRL were insignificantly higher than those PMF. U-HYPRL increased in 10 patients (1 AMF patient, 3 OMS patients and 6 patients with CML in the accelerated phase of the disease). All but 1 of these patients had been treated, or were being treated, with cytotoxic agents at the time of investigation [89][74]. |
The findings of normal U-HYPRL may be explained by impaired degradation of bone marrow collagen which, together with enhanced collagen synthesis from bone marrow fibroblasts, accounts for progressive accumulation of connective tissue in the bone marrow in myelofibrosis patients. This process is influenced by cytotoxic treatment as reflected in increased urinary hydroxyproline excretion in those patients receiving cytotoxic agents [89][74]. |
[80,89][74][75] |
U-HYPRL was normal in PMF patients. U-HYPRL was increased in patients with metastasis, the highest levels being recorded in those with concomitant bone marrow fibrosis [80][75]. |
The result suggests differences in the pathogenesis of “MPN-myelofibrosis” (normal U-HYPRL) as compared to myelofibrosis consequent to bone marrow metastasis (increased U-HYPRL) [80][75]. |