Role of receptor activator of nuclear factor-kappa B ligand (RANKL), osteoprotegerin and macrophage protein 1-alpha (MIP-1a) in monoclonal gammopathy of undetermined significance (MGUS)


Evangelos Terpos, Department of Haematology, Faculty of Medicine Imperial College London, Hammersmith Hospital, Du Cane Road, W12 0NN, London, UK. E-mail:


The aim of this study was to evaluate the role of markers of bone remodelling, and osteoclast activation/function in patients with monoclonal gammopathy of undetermined significance (MGUS). We have measured serum levels of soluble RANKL (sRANKL), osteoprotegerin (OPG), macrophage inflammatory protein-1alpha (MIP-1α), markers of bone resorption [N-telopeptide of collagen type-I (NTX), and tartrate-resistant acid phosphatase isoform-5b (TRACP-5b)] and bone formation [bone-alkaline phosphatase (bALP)] in 40 MGUS patients. These parameters were compared with those of 42 newly diagnosed myeloma patients, and 45 healthy, gender- and age-matched controls. MGUS patients had elevated levels of NTX, sRANKL, and sRANKL/OPG ratio compared with controls (P < 0·0001). Furthermore, TRACP-5b, MIP-1α and NTX were decreased in patients with MGUS compared with myeloma patients (P < 0·001), while OPG and bALP were increased (P < 0·001). Serum levels of MIP-1α, as well as TRACP-5b, and sRANKL/OPG ratio were reduced, while bALP was increased in MGUS patients, even when compared with myeloma patients who had stage I/II disease. These results demonstrate that increased osteoclastogenesis leading to increased bone resorption is present in MGUS but seems to be compensated for by normal bone formation, which is absent in MM. Furthermore MIP-1α, bALP, and sRANKL/OPG may be useful tools for distinguishing between cases of MGUS and early myeloma.

Bone disease is a major clinical feature of multiple myeloma (MM). Recent characterization of novel molecules such as the receptor activator of nuclear factor kappa-B ligand (RANKL), osteoprotegerin (OPG) and macrophage inflammatory protein-1 alpha (MIP-1α) has provided new insight into the pathophysiology of bone disease in MM (Oyajobi et al, 2003; Sezer et al, 2003). Our group has already shown that RANKL/OPG and MIP-1α are increased in MM patients, correlate with the extent of bone disease and have an impact on survival (Terpos et al, 2003a,b). Patients with monoclonal gammopathy of undetermined significance (MGUS) have increased bone resorption, as evaluated by the increased levels of collagen type-I degradation products (Pecherstorfer et al, 1997; Vejlgaard et al, 1997; Jakob et al, 2002). However, the mechanism for this is unclear. In this study we have evaluated the role of RANKL, OPG, MIP-1α, and markers of bone remodelling in patients with MGUS and compared them with MM patients, in an attempt to better understand the imbalance of bone turnover in MGUS.

Patients and methods

We have prospectively studied 40 MGUS patients (25M/15F; median age 65 years, range: 39–86 years), diagnosed according to the International Myeloma Working Group criteria, and compared them with 42 newly diagnosed MM patients (18M/24F; median age 72 years, range: 36–91 years), and 45 healthy, gender- and age-matched, controls. Levels of soluble RANKL (sRANKL), OPG, MIP-1α and bone-remodelling markers [N-telopeptide of collagen type-I (NTX), tartrate resistant acid phosphatase isoform-5b (TRACP-5b) and bone-alkaline phosphatase (bALP)] were measured in serum samples using ELISA (as described elsewhere; Terpos et al, 2003a). Markers of disease activity (beta2-microglobulin, and CRP) were also evaluated. Skeletal survey was performed in all patients. Differences between MGUS, MM patients and controls were analysed using the Mann–Whitney test. Associations between bone disease stage and biochemical markers were examined by the Kruskal–Wallis test, whilst the Spearman Rank correlation test was employed to examine relationships between various parameters and the clinical characteristics of the patients. The study was conducted with Ethical Committee approval and under the guidelines of the Declaration of Helsinki.

Results and discussion

Soluble RANKL/OPG ratio, NTX, TRACP-5b, and MIP-1α were significantly decreased in MGUS compared with MM (P < 0·01), while OPG and bALP were increased (P < 0·001). In myeloma patients serum levels of RANKL/OPG ratio, and MIP-1α were increased, while bALP was decreased in parallel with the stage of the disease (according to Salmon–Durie criteria; P < 0·001); and all three parameters differ significantly between MGUS and stage I/II myeloma (P < 0·05; Fig 1). When the above parameters of MGUS patients were compared with the ones of controls, only sRANKL/OPG ratio and NTX levels were significantly increased (P < 0·0001; Fig 1A and D). These results are summarized in Table I. No MGUS patient had radiographic evidence of osteopenia/osteoporosis.

Figure 1.

Markers of bone remodelling and osteoclast activation and function in MGUS patients, myeloma patients and normal controls. In MGUS patients the increased bone resorption [increased NTX levels compared with controls; P < 0·0001, (A) and Table I] seems to be accompanied by normal osteoblastic activity [no difference in bALP and OPG between MGUS and controls (B) and Table I]. On the contrary, patients with MM have increased osteoclast activity when compared with MGUS patients as measured by the levels of NTX (A), TRACP-5b (C), sRANKL/OPG ratio (D), and MIP-1α (E), while bone formation (bALP, and OPG levels) is dramatically suppressed (B); thus, leading to osteolytic bone disease. The increase of sRANKL/OPG ratio in MGUS patients compared with controls seems to be responsible for the increased bone resorption in these patients. Median values and range of the above parameters for MM stages I/II and MGUS, respectively, were as follows: bALP (U/L): 17 (4–31) vs. 33·8 (13·8–94·4), P < 0·001 (B); TRACP-5b (U/L): 3·8 (1·7–7·2) vs. 3·0 (0·4–7·3), P < 0·04 (C); sRANKL/OPG (×10−2): 4·9 (3·0–234·5) vs. 4·1 (0·3–192·5), P = 0·042 (D); MIP-1α (pg/mL): 11·5 (0·2–31) vs. 3·4 (0·2–26·6), P < 0·001 (E).

Table I.  Median values (range) of markers of bone remodelling, osteoclast function, and MIP-1α in patients and controls.
ParameterMM (n = 42)MGUS (n = 40)Controls (n = 45)Significance MM versus MGUSSignificance MGUS versus controls
  1. *MIP-1a levels were measured in 12 controls.

  2. NS, non-significant (P > 0·05).

sRANKL (pmol/l)0·27 (0·07–1·85)0·34 (0·14–3·7)0·02 (0–3·95)NSP < 0·0001
OPG (pmol/l)4·7 (0·3–16·6)8·3 (0·4–46·5)5·8 (3·2–10·4)P < 0·001P = 0·067; NS
sRANKL/OPG (×10−2)5·7 (2·4–264·2)4·1 (0·3–192·5)0·3 (0–123·4)P < 0·01P < 0·0001
TRACP-5b (U/l)6·5 (1·3–47·1)3·0 (0·4–7·3)3·1 (1–5)P < 0·0001NS
MIP-1α (pg/ml)16·2 (0·2–408·5)3·4 (0·2–26·6)13·0* (1·4–54)P < 0·001NS
NTX (nm BCE/mmol/l creatinine)132·0 (25–571)81·2 (19·1–418·6)37·9 (11–56·2)P < 0·001P < 0·0001
bALP (U/l)15·5 (2·8–39·5)33·8 (13·8–94·4)33·9 (22–41·4)P < 0·001NS

In this study, we have tried for the first time to evaluate in MGUS patients, the role of novel molecules such as RANKL, OPG and MIP-1α, which are involved in MM bone disease. It has already been shown with bone histomorphometry that increased bone resorption may be present in patients with plasma cell dyscrasias before the appearance of clinical symptoms or osteolytic lesions. It has been also proposed that a quantifiable excess of bone resorption in MGUS is significantly associated with progression and this is an early symptom of malignancy (Bataille et al, 1996).

Over the last few years, non-invasive and more sensitive methods have been introduced to assess bone resorption and various novel bone resorption markers have been identified. In previous studies, markers of bone resorption such as NTX, carboxy-terminal telopeptide of type-I collagen (ICTP), and urinary deoxypyridinoline (uDPYD) were evaluated in MGUS patients. All markers were found to be significantly decreased in MGUS patients compared with MM patients (Pecherstorfer et al, 1997; Vejlgaard et al, 1997; Corso et al, 2001; Diamond et al, 2001; Jakob et al, 2002).

RANKL/OPG pathway has been recently characterized as the final common pathway of osteoclastogenesis. RANK, which is expressed on osteoclasts precursors, competes with the decoy receptor OPG for binding to RANKL. The binding of RANKL to RANK leads to the osteoclast differentiation and activation. In MM, the RANKL/OPG ratio is increased. MM cells, through adhesive interactions in the bone marrow microenvironment, enhance the production of RANKL by stromal cells and reduce the OPG levels, favouring osteoclast activation and the subsequent bone resorption (Sezer et al, 2003; Terpos et al, 2003a). MIP-1α is a low molecular weight chemokine that belongs to the RANTES family. It has been characterized as a potent osteoclast stimulatory factor in MM. MIP-1a is produced by myeloma cells and may either act through RANKL or may act directly by stimulating osteoclast formation (Oyajobi et al, 2003; Terpos et al, 2003b).

In our study, we have shown that not only NTX, but also TRACP-5b, sRANKL/OPG ratio and MIP-1α were decreased, while OPG and bALP were increased in MGUS compared with MM. When the above parameters were evaluated in MGUS patients versus patients with stages I and II MM, RANKL/OPG, MIP-1a and bALP remained statistically different implying that these parameters could differentiate MGUS from early stages of MM, and would be useful tools in identifying patients with early MM who have no evidence of osteolysis.

Serum sRANKL levels and sRANKL/OPG ratio, as well as NTX levels were increased in MGUS patients compared with controls, implying that bone resorption is an early event in monoclonal gammopathies. The presence of osteoporosis/osteopenia that could account for the increased bone resorption was excluded in our patients who had no evidence of bone loss on skeletal survey and had normal values of bALP. Bone ALP is a parameter that is increased in patients with osteoporosis, correlates with risk for fractures and has been used for the follow-up of post-menopausal osteoporosis (Garnero et al, 2000). The fact that increased osteoclastogenesis is present in MGUS patients, has already been shown in morphometric studies, although the mechanisms that lead to osteoclast activation in MGUS remain unclear. Although we have not detected increased serum levels of MIP-1α in MGUS patients, the increased ratio of sRANKL/OPG suggests a putative role of the RANK/RANKL/OPG pathway in the pathogenesis of bone resorption in MGUS. Increased bone resorption may be an early local process in the bone marrow of MGUS, which does not result in any cytokine changes in the serum. MGUS is an indolent disease and only 1% of the patients per year transform to MM. It would be of interest to study whether a more drastic alteration of RANKL/OPG ratio is present in the MGUS patients who transform to MM compared with those who do not.

The above data demonstrate that bone resorption is increased in MGUS. However, patients with MGUS do not develop diffuse osteoporosis or lytic lesions as the osteoblast function is normal with normal compensatory mechanism for increased bone loss. In MM, however the osteoblast function/bone formation does not appear to be normal and cannot compensate the increased osteoclast function, thus leading to characteristic lytic bone disease. This may be explained by the recent finding that MM cells produce DKK-1 protein, which inhibits Wnt-signalling and therefore inhibit osteoblast differentiation (Tian et al, 2003). Furthermore MIP-1α, bALP, and sRANKL/OPG, could be useful tools for distinguishing between cases of MGUS and early MM.