Priority Paper Evaluation

Whither goest the lifeless osteoclast?

Evaluation of: Weinstein RS, Roberson PK, Manolagas SC: Giant osteoclast formation and long-term oral bisphosphonate therapy. N. Engl. J. Med. 360, 53–62 (2009). Giant osteoclasts are pathognomonic in Paget’s Disease, secondary hyperparathyroidism, giant tumors and fibrous dysplasia. Weinstein et al. provided a detailed histological analysis of abnormal hypernucleated osteoclasts in half of a set of biopsies from patients who received clinical doses of alendronate, a bisphosphonate inhibitor of bone resorption. In addition, 30% of the osteoclasts were apoptotic, a feature not found in the other clinical conditions. Their report reveals new information about skeletal effects of alendronate in addition to known effects to inhibit osteoclastic bone resorption. Their report raises new questions about the lifecycle of the human osteoclast, effects of bisphosphonates on cell fusion and on efferocytosis (removal of apoptotic cells), significance, if any, of the abnormal osteoclasts, and distinguishing features of the set of alendronate-treated subjects whose biopsies showed these abnormalities.

KEYWORDS: alendronate n n bisphosphonate n efferocytosis n giant cells Julie Glowacki n osteoclasts Department of Orthopedic Surgery, Harvard Medical Weinstein et al. compared osteoclasts in bone that removes cells and is evident by pycnotic School, Department of Oral & biopsies from control and alendronate-treated nuclei, loss of ruffled border and detachment Maxillofacial Surgery, women [1]. Osteoclasts are the large, multi­ from the bone surface. Apoptosis is distinct Harvard School of Dental nucleated cells responsible for bone resorption. from cell necrosis, which occurs in groups of Medicine, Brigham and They are formed by fusion of mononuclear pre- swollen cells, often with intracellular precipi- Women’s Hospital, cursors that are derived from hematopoietic stem tation of calcium. Apoptosis occurs rapidly, 75 Francis Street, Boston, cells of the monocytic lineage and are specialized with phagocytes ingesting and eliminating the MA 02115, USA to attach to bone surfaces and to degrade the apoptotic bodies; apoptotic osteoclasts may be Tel.: +1 617 732 5397; bone matrix. Active osteoclasts adhere to bone difficult to detect in normal bone biopsies, even Fax: +1 617 732 6937; matrix with a specialization of inte- with specialized stains that detect early stages jglowacki@ grins that are organized into a ring-like sealing of DNA fragmentation. rics.bwh.harvard.edu zone. This zone surrounds the ruffled border Bisphosphonates are potent inhibitors of and defines the resorption lacuna where bone is osteoclastic bone resorption and were first intro- degraded. Acidification of the lacuna is achieved duced to retard bone turnover in patients with by the action of a membrane-bound vacuolar- Paget’s disease. Seven bisphosphonates have been type ATPase proton pump and results in solu- approved by the US FDA and new, more potent bilization of the mineral phase. Thereupon, pro- compounds continue to appear. Intravenous teolytic enzymes of two major classes, lysosomal nitrogen-containing formulations are routinely cysteine proteases (especially cathepsin K) and used for treating bone metastases and hypercal- matrix metalloproteinases (MMP-9), degrade cemia of malignancy in patients with cancer. The collagen and other bone matrix proteins. most common indications for bisphosphonates Degradation products are endocytosed into are the treatment and prevention of osteoporo- vesicles and large vacuoles and are carried to sis. Off-label uses are expanding, particularly in the secretory surface of the cell. During devel- pediatric conditions associated with low bone opment and growth, osteoclasts resorb calcified mass, such as osteogenesis imperfecta. cartilage and contour the shapes of enlarging Bisphosphonates are chemically stable ana- bones; thereafter, they act in coordination with logs of pyrophosphate, and all have two phos- bone-forming osteoblasts to remodel and renew phate groups that bind to the calcium phosphate bone tissue. It is likely that osteoclasts undergo mineral of bone’s extracellular matrix [2]. The cycles of resorption and rest, but in vivo, details fact that they accumulate in bone is the reason part of of kinetics are lacking. The osteoclast lifespan for their potent effects on the skeleton. They is limited by apoptosis – the controlled process selectively target the osteoclast because they

10.2217/IJR.09.45 © 2009 Future Medicine Ltd Int. J. Clin. Rheumatol. (2009) 4(5), 519–522 ISSN 1758-4280 519 Priority Paper Evaluation Glowacki Whither goest the lifeless osteoclast? Priority Paper Evaluation

preferentially bind to bone mineral exposed at taking bisphosphonates may show few osteoclasts. resorbing surfaces, as the osteoclast acidifies the Normally, the number of osteoclasts is used as bone surface, which dissolves the bisphospho- an index of bone resorption [5]. Nonetheless, nate, and because the osteoclast can uptake the Weinstein et al. reported that there were signifi- released bisphosphonates along with the matrix cantly more osteoclasts in samples from subjects products of resorption. Like pyrophosphate, treated with 10 mg/day alendronate than placebo bisphosphonates can inhibit mineralization [1]. It must be concluded that those osteoclasts did in vitro and in vivo, but the newer ones, especially not resorb bone well. Morphological abnormalities the nitrogen-containing ones, such as alendro- suggesting this was found in the treated groups: nate and zoledronate, can inhibit bone resorp- the osteoclasts were detached from the bone, often tion at doses far lower than those that inhibit with intervening cells; resorption lacunae were mineralization. Several mechanism account for shallow or absent. In addition, they had more than the effects of bisphosphonates on osteoclasts. eight nuclear profiles, often as many as 40, and Etidronate and clodronate are metabolized into they had pycnotic nuclei. The number of detached ATP analogs that induce osteoclast apoptosis. It osteoclasts was statistically elevated in the group is now understood that the nitrogen-containing receiving 10 mg/day alendronate for 3 years. The bisphosphonates bind to the enzyme farnesyl number in the group receiving 20 mg/day for diphosphate (FPP) synthase, thereby inhibiting 2 years followed by 1 year of placebo was almost prenylation of essential GTPase proteins and normal; this suggests some measure of recovery inducing osteoclast apoptosis. These effects of in the drug-free period. Apoptotic nuclei were bisphosphonates would occur in all cell types, present in all treatment groups. The authors but their accumulation on bone surfaces targets speculated that alendronate may prolong apopto- their actions to osteoclasts. sis. We offered another possible mechanism, that Weinstein et al. had access to bone biopsies phagocytes that should be recruited to dispose of from a 3-year trial of oral alendronate for pre- the apoptotic osteoclasts are themselves inhibited venting bone loss in healthy postmenopausal by alendronate in the bone microenvironment [6]. women of 40–59 years of age [1]. The original In vitro evidence demonstrates that bisphospho- study described an increase in bone mineral den- nates inhibit development [7] and sity and a decrease in the biochemical markers of induce apoptosis in macrophage-like cells [8]. bone turnover in the groups of subjects treated The report by Weinstein et al. is the first with alendronate at 5 or 10 mg/day for 3 years, detailed publication on the numbers and features or in those given 20 mg/day for 2 years followed of osteoclasts in humans receiving alendronate, by placebo for 1 year [3]. Some of the subjects although it cites similar observations reported volunteered for transiliac bone biopsy at the by others in abstract form. Some animal studies end of the 3-year period; the specimens were have noted, in passing, that administration of used in the original report to demonstrate that bisphosphonates increased osteoclast numbers. alendronate reduced bone turnover, as expected. For example, in a study demonstrating that For this study, which was published more than nitrogen-containing bisphosphonates reduced 10 years after the original one, 51 specimens immunoreactivity for a product downstream of were processed for the specialized examination FPP synthase, it was also noted that osteoclast of osteoclast morphology, staining for tartrate- numbers were significantly elevated in bones resistant acid phosphatase and staining for DNA from the groups of rats treated with the nitro- fragmentation. These are not routine assays and gen-containing biphosphates but not with the require fastidious technique to avoid distortions bisphosphonates lacking nitrogen [9]. and misinterpretation. Professor Weinstein Although osteoclast apoptosis may be a real (University of Arkansas for Medical Sciences, effect of alendronate, biopsies from ten of the AR, USA) disclosed the many virtuoso measures 17 subjects receiving the highest doses of alen- he took to ensure the highest quality of histol- dronate exhibited no abnormal osteoclasts. ogy needed for these analyses of nondecalcified The mean prevalence of apoptotic osteoclasts bone biopsies. was 30% in the group receiving 10 mg/day. It It is clear from the biochemical markers of would be very useful for future studies to delve bone turnover that bone resorption was decreased into whether the occurrence of these abnormal in subjects receiving alendronate [3]. Because cells is related to, for example, subjects’ clinical bisphosphonates inhibit bone resorption and features (e.g., lowest body weight) or subjects’ promote apoptosis of osteoclasts in vitro [4], it responses to alendronate (e.g., greatest suppres- could be expected that biopsies from subjects sion of bone turnover). Alendronate treatment of

520 Int. J. Clin. Rheumatol. (2009) 4(5) future science group Priority Paper Evaluation Glowacki Whither goest the lifeless osteoclast? Priority Paper Evaluation

2 or 3 years duration is not considered long-term not lost all functionality. Efferocytosis is a term for osteoporosis. Growing concerns regarding used to describe the process by which phagocytes skeletal complications of oral bisphosphonates, dispose of apoptotic cells. It is notable that Rho such as presentation with unusual subtrochan- GTPases in phagocytes play crucial roles in acti- teric or shaft fracture through thick cortices [10] vating efferocytosis and [12]. Another and osteonecrosis of the jaw [101], highlight the important regulator of cell survival is osteopro- need for more information about these agents. tegerin (OPG). OPG is well known as a decoy The greatest clinical significance of this prior- receptor, produced by marrow stromal cells and ity paper evaluation is the alert to pathologists osteoblasts, that blocks the binding of RANK concerning the occurrence of these cells in the ligand (RANKL) to RANK in osteoclast pro- setting of oral bisphosphonate use for osteo­ genitors [13]. In addition, OPG can stimulate cell porosis. These cells should not be confused with survival by binding to TNF-related apoptosis- other types of hypernucleated cells. Giant osteo- inducing ligand (TRAIL), thereby allowing cells clasts are features of bone in Paget’s disease and to escape cell death [14]. We recently reported that in hyperparathyroidism, but those conditions alendronate upregulated OPG in human marrow are associated with elevated osteoid and osteo- stromal cells in vitro, and that OPG was consti- blast levels. The hypernucleated cells of giant tutively elevated (by 1.53-fold) in marrow from cell tumors and fibrous dysplasia are located in women treated with alendronate, compared with fibrous stroma, distant from the bone. Only in age-matched untreated women [15]. the instance of bisphosphonate therapy are there Efferocytosis has been demonstrated to be high numbers of apoptotic osteoclasts. A case impaired in a number of disorders, such as rheu- study recently published by Jain and Weinstein matoid arthritis [16], advanced athero­sclerosis exemplifies the issues in histological differential [17] and chronic lung diseases [12]. In those set- diagnoses of giant cells in bone [11]. tings, it is believed that impaired efferocytosis Several mysteries need to be solved. Apoptosis perpetuates . Although Weinstein is a mechanism for the controlled removal of described no evidence of inflammation or fibrosis injured or worn-out cells that usually occurs in the bone biopsies (possibly attributable to the quickly. Cell death, removal, and replenishment anti-inflammatory actions of bisphosphonates), are essential for homeostasis in healthy organisms. the persistence of apoptotic cells may ultimately Finding large numbers of apoptotic osteoclasts in have untoward effects when, and if, the bisphos- the alendronate groups may signify increased for- phonates are eliminated from the bone matrix. mation and/or decreased removal. The increased It has been calculated that following discontinu- numbers of nuclei may indicate continued fusion ation of 10 years of treatment with 10 mg/day of mononuclear precursors with osteoclasts in the alendronate, skeletal release of stored alendronate presence of alendronate. It also indicates that the approximates the equivalent of an oral daily dose of the apoptotic osteoclast has of 2.5 mg with a very gradual increase in bone

Executive summary Findings from the study ƒƒ Biopsies available from a previous controlled trial showed elevated numbers of osteoclasts in samples from subjects receiving 10 mg/day alendronate for 3 years. ƒƒ The biopsies showed that, in a dose-related manner, alendronate therapy was associated with abnormal-appearing osteoclasts (detached from bone surfaces, greater number of nuclei or prevalence of nuclear apoptosis). ƒƒ Only half of the specimens revealed abnormalities and only 30% of the giant osteoclasts demonstrated evidence of apoptosis. The findings need to be correlated with clinical parameters and measures of bone turnover. Impact of the study ƒƒ The use of bisphosphonates should be added to the differential diagnosis of finding giant cells in bone biopsies. ƒƒ Questions are raised regarding the lifecycle of the human osteoclast, the process of cell fusion in the bisphosphonate-containing milieu, and the possible impact of abnormal cells. ƒƒ Bisphosphonate treatment is added to the list of disorders of apoptosis/efferocytosis. Conclusion ƒƒ This histopathological study revealed skeletal effects of alendronate in addition to the known effects inhibiting osteoclastic bone resorption. Future perspective ƒƒ More information is needed about the relative effects of the various bisphosphonates on the osteoclast and macrophage lifecycle and whether those effects are related to antiresorptive potency. ƒƒ The potential pathophysiological impact of these abnormal osteoclasts should be investigated.

future science group www.futuremedicine.com 521 Priority Paper Evaluation Glowacki

turnover [18]. Unlike with discontinuation of Financial & competing interests disclosure conjugated estrogen, bone mineral density was The author has no relevant affiliations or financial maintained in the year following discontinuation involvement with any organization or entity with a finan- after 2 years of 10 mg/day alendronate [19]. cial interest in or financial conflict with the subject matter Weinstein’s observations raise many new or materials discussed in the manuscript. This includes questions about the life and death of human employment, consultancies, honoraria, stock ownership or osteoclasts that can be addressed experimentally. options, expert testimony, grants or patents received or Without doubt, more information is needed pending, or royalties. regarding all of the effects of bisphosphonates No writing assistance was utilized in the production of on bone and bone marrow cells. this manuscript

8 Roers MJ, Chilton KM, Coxon FP et al.: 16 Firestein GS, Yeo M, Zvaifler NJ: Apoptosis Bibliography Bisphosphonates induce apoptosis in mouse in synovium. J. Clin. 1 Weinstein RS, Roberson PK, Manolagas SC: macrophage-like cells in vitro by a nitric Invest. 96, 1631–1638 (1995). Giant osteoclast formation and long-term oxide-independent mechanism. J. Bone Miner. 17 Schrijvers DM, De Meyer GR, Kockx MM, oral bisphosphonate therapy. N. Engl. Res. 11, 1482–1491 (1996). Herman AG, Martinet W: of J. Med. 360, 53–62 (2009). 9 Fisher JE, Rodan GA, Reszka AA: In vivo apoptotic cells by is impaired in 2 Russell RG: Bisphosphonates: mode of effects of bisphosphonates on the osteoclast . Arterioscler. Thromb. Vasc. action and pharmacology. Pediatrics 119, mevalonate pathway. Endocrinology Biol. 25, 1256–1261 (2005). S150–S162 (2007). 141, 4793–4796 (2000). 18 Rodan G, Reszka A, Golub E, Rizzoli R: 3 McClung M, Clemmesen B, Daifotis A 10 Lenart BA, Neviaser AS, Lyman S et al.: Bone safety of long-term bisphosphonate et al.: Alendronate prevents postmenopausal Association of low-energy femoral fractures treatment. Curr. Med. Res. Opin. 20, bone loss in women without osteoporosis. with prolonged bisphosphonate use: a case 1291–1300 (2004). Ann. Intern. Med. 128, 253–261 (1998). control study. Osteoporos. Int. 20(8), 19 Greenspan SL, Emkey RD, Bone HG et al.: 4 Hughes DE, Wright KR, Uy HL et al.: 1353–1362 (2008). Significant differential effects of alendronate, Bisphosphonates promote apoptosis in 11 Jain N, Weinstein RS: Giant osteoclasts after estrogen, or combination therapy on the rate murine osteoclasts in vitro and in vivo. long-term bisphosphonate therapy: diagnostic of bone loss after discontinuation of treatment J. Bone Miner. Res. 10, 1478–1487 (1995). challenges. Nat. Rev. Rheumatol. 5, 341–346 of postmenopausal osteoporosis. Ann. Intern. 5 Parfitt AM, Drezner MK, Glorieux FH (2009). Med. 137, 875–883 (2002). et al.: Bone histomorphometry: 12 Vandivier RW, Henson PM, Douglas IS: standardization of nomenclature, symbols, Burying the dead: the impact of failed and units. Report of the ASBMR apoptotic cell removal (efferocytosis) on „ „Website Histomorphometry Nomenclature chronic inflammatory lung disease. Chest 101 FDA site containing memorandum for the Committee. J. Bone Miner. Res. 2, 595–610 129, 1673–1682 (2006). matter of osteonecrosis of the jaw (1987). 13 Teitelbaum SL: Bone resorption by www.fda.gov/ohrms/dockets/ac/05/ 6 Glowacki J: The deceiving appearances of osteoclasts. Science 289, 1504–1508 (2000). briefing/2005–4095B2_03_04-FDA-TAB3 osteoclasts. N. Engl. J. Med. 360, 80–82 14 Reid P, Holen I: Pathophysiological roles of Accessed 8 July 2009 (2009). osteoprotegerin (OPG). Eur. J. Cell Biol. 88, 7 Cecchini MG, Fleisch H: 1–17 (2009). Bisphosphonates in vitro specifically inhibit, 15 Eslami B, Zhou S, LeBoff MS, Glowacki J: among the hematopoietic series, the Reduced osteoclastogenesis and RANKL development of the mouse mononuclear expression in marrow from women taking phagocyte lineage. . 5, J. Bone Miner. Res alendronate. J. Bone Miner. Res. 23, S344 1019–1027 (1990). (2008).

522 Int. J. Clin. Rheumatol. (2009) 4(5) future science group