Management of Tenosynovial Giant Cell Tumor: A Neoplastic and Inflammatory Disease

Background: Patients with diffuse tenosynovial giant cell tumor (TGCT) face a high risk of recurrence, progression, and disability. This systematic review assesses the recent evidence of surgical, adjuvant, and systemic treatments for TGCT. Methods: We searched PubMed and Ovid with the terms “Giant cell tumor of tendon sheath” OR “pigmented villonodular synovitis” OR “tenosynovial giant cell” AND “treatment” OR “surgery.” Inclusion criteria: published 2013 to present; prospective or retrospective design; English language; > 20 patients with histopathological confirmed diagnosis of TGCT; and ≥ 1 efficacy and/or safety outcome from surgery, systemic drug therapy, or adjuvant 90yttrium radiosynoviorthesis. Results: Of the 434 studies identified, 25 met the inclusion criteria. Of 11 studies in patients with disease in the knee, nine examined surgical treatment approaches, and two evaluated adjuvant 90yttrium radiosynoviorthesis. Of 11 studies in patients with mixed sites of disease, six assessed surgical treatment approaches, and five evaluated systemic drug therapies. Three studies assessed surgery in patients with TGCT in the hand, hip, and ankle or foot. Discussion: The high rates of recurrence and risks associated with surgery emphasize the need for novel treatments in patients with symptomatic, advanced TGCT. Systemic therapy may be valuable as part of a multidisciplinary approach.

T enosynovial giant cell tumors (TGCTs) are rare, locally aggressive, typically benign neoplasms of joints, bursae, and tendon sheaths. [1][2][3] Symptoms of TGCT include pain, stiffness, swelling, and limitation in range of motion. TGCTs have a wide clinical spectrum and affect patients of all ages. Until the World Health Organization re-classified them in 2013, 1,2 TGCTs were classified according to their site of origin (ie, bone, soft tissue, synovium, or tendon sheath) as giant cell tumor of tendon sheath (GCTTS) or nodular tenosynovitis and diffuse type giant cell tumors or pigmented villonodular synovitis (PVNS). 4  term localized TGCT encompasses GCTTS and nodular tenosynovitis, whereas diffuse TGCT encompasses diffuse-type giant cell tumor and PVNS. 1,2 This reclassification emphasizes that the driving force in the pathology of TGCT is a tumor and that the symptoms are often defined by a secondary inflammatory joint response.
Compared with localized TGCT, diffuse TGCT tends to affect a younger population and is more predominant in women than men. 2 Development of TGCT has been associated with a clonal neoplastic process, often involving a specific chromosomal translocation, t(1;2) (CSF-1;COL6A3), resulting in the overexpression of colony-stimulating factor 1 and recruitment of CSF1 receptor (CSF1R) macrophages, giant cells, and osteoclasts. 3 A subset of tumors exists (2% to 16%) where there seems to be another effector downstream that results in excess CSF1R expression/response without the mutation. 5,6 Updated data on TGCT epidemiology exist. A 2017 Danish registry analysis reported a population prevalence of 44 per 100,000 for localized TGCT and 12 per 100,000 for diffuse TGCT, and a 10-year risk of recurrence of 9.8% for localized TGCT and 19.1% for diffuse TGCT. 7 A 2017 Dutch registry study reported standardized worldwide incidence rates of 29 cases per million for TGCT affecting digits, 10 for localized TGCT, and four for diffuse TGCT. In the study, the recurrence rate was 2.6 times higher for diffuse TGCT compared with localized TGCT. 8 There is currently no consensus on the optimal standard of care for TGCT, especially for patients with widely diffuse or recurrent disease. Three European groups have previously published recommendations. 3,4,9 In 2012, a combined UK and Dutch group proposed an integrated, multidisciplinary treatment protocol for TGCT. 4 A 2016 UK guideline stated that patients with TGCT are usually treated by surgery alone and that, rarely, radiotherapy or imatinib may be used. 9 In addition, in 2016, an Italian group remarked on the lack of high-quality studies of TGCT treatment and identified open surgical excision as the benchmark treatment of diffuse disease. 3 Most patients with diffuse TGCT are treated by arthroscopic or open surgical excision with partial or extensive synovectomy. 10 The most appropriate form of surgery has not been identified and probably differs by tumor site and location. 11 Radiosynoviorthesis (RSO) and external beam radiotherapy (EBRT) are sometimes used alone or in combination with surgery. Their contribution is unclear because of limited low-level of evidence (LOE), poorquality data, and contradictory results. The authors of a 2015 metaanalysis of individual patient data from 35 observational studies reported what they deemed to be very low-quality evidence that perioperative EBRT might reduce the rate of recurrence in patients with diffuse PVNS. 12 The main issues with RSO are morbidity to the joint, including early onset arthritis, osteonecrosis, increased postradiotherapy perioperative risks for complications, and concerns with malignant transformation or secondary radiationassociated sarcomas. 12 In managing patients with advanced TGCT, surgeons must balance the potential benefits and harms of different surgical and nonsurgical approaches with the risk of recurrence, disease progression, and long-term disability. Given the paucity of proven, effective local or systemic therapies, clinicians have struggled in this regard to manage refractory disease. Notable advances in treatment options for diffuse TGCT have recently been made, which suggest a potential role for systemic therapies as part of the overall management of the disease process. This systematic review assesses the recent evidence of treatment strategies for patients with TGCT, providing a critical, up-todate assessment of trends while highlighting continued unmet needs.

Search Strategy
On January 30, 2020, we searched MEDLINE (via PubMed) and EM-BASE using the following search terms: ([GCTTS] OR [PVNS] OR [tenosynovial giant cell]) AND (treatment OR surgery). We set the publication date search limits from January 1, 2013, to January 31, 2020, for PubMed and 2013 to 2020 for EMBASE. We limited the searches to English-language publications. The EMBASE searches excluded conference abstracts/articles, case reports, preclinical studies, reviews, editorials, letters, notes, chapters, and surveys.

Eligibility Criteria
We included prospective or retrospective studies that reported . 1 efficacy and/or safety outcome from surgery, systemic drug therapy, or adjuvant 90 yttrium RSO in . 20 patients with a histopathologically confirmed diagnosis of TGCT, GCTTS, or PVNS. Reasons for study exclusion were having # 20 patients with TGCT, EBRT as the primary treatment modality, lack of confirmed histopathologic diagnosis, mixed population with no separate reporting of data for patients with TGCT, patient-reported outcomes as the only efficacy measures, and lack of an abstract.

Study Selection
After removal of duplicates from PubMed and Ovid searches, the title and abstract were screened to exclude records not meeting all the inclusion criteria. Studies meeting the initial screening underwent repeat screening through full-text review to verify all the information necessary for complete application of inclusion criteria.

Data Extraction and Analysis
From the included studies, we extracted and recorded the following data: publication year, study design, treatment period, number of patients, type and anatomic site of TGCT, type of treatment of TGCT, length of follow-up, key efficacy end points, and main complications or adverse events (AEs). Wide heterogeneity in study design, end points, and types of treatment prevented quantitative statistical analyses. We applied the criteria outlined in the J Am Acad Orthop Surg author instructions to assign an LOE for each included study.
Three studies (1 study each) reported the results of surgical outcomes in patients with TGCT in the hand, 35 hip, 36 and foot or ankle 37 (Table 3). Except for four systemic drug therapy studies with prospective clinical trial design, all other studies had a retrospective cohort design (LOE III). The prospective studies included 2 phase 1 trials (LOE II), 30,32 1 phase 2 trial (LOE II), 31 and 1 randomized controlled trial (RCT; LOE I). 33 Surgery Only 11 of the 24 included studies reported comparative data regarding the treatment of patients with TGCT. Seven retrospective cohort studies compared open synovectomy with arthroscopic synovectomy. 11,14,16,20,21,27,28 Note that some studies reported results in terms of recurrence rates (where lower percentages indicate better outcomes), and other studies reported results in terms of failure-free or relapse-free survival (where higher percentages indicate better outcomes).
Among the knee studies (Table 1)     Superficial wound infection (n = 6); foot drop (n = 3); hemarthrosis (n = 3); stiffness requiring manipulation under anesthesia (n = 2); complex regional pain syndrome (n = 2); blistering from tourniquet (n = 1); DVT (n = 1)     11 A retrospective cohort study reported no statistically significant differences in mean knee-function measures (Lysholm and Ogilvie-Harris scores) and lesion recurrence rate, with a higher incidence of chronic regional pain syndrome, but not of other complications, after arthroscopically assisted mini-open partial synovectomy versus arthroscopic excision of the lesion in 44 patients with localized PVNS (Table 1). 15

Radiosynoviorthesis
Both studies of adjuvant 90 yttrium RSO 23,24 reported similar rates of recurrence with and without RSO treatment in patients with diffuse disease in the knee ( Table 1). One of these studies also found no significant differences in overall physical health and mental health scores, perception of pain, and patient satisfaction after a mean follow-up of 7.3 years. 24

Systemic Therapy
The five systemic drug therapy studies included 1 study of intravenous emactuzumab (an anti-CSF1R    (Table 2). In a phase 1 study, patients with locally advanced, diffuse TGCT ach-ieved an 86% objective response rate (ORR), including 7% complete responses (CRs) and 79% partial responses (PRs), with emactuzumab biweekly infusions. 30 The most commonly reported AEs were facial edema (64%), asthenia (56%), and pruritus (56%). A phase 2 study of twice-daily oral nilotinib yielded no CRs or PRs, but a 90% rate of stable disease (SD) among patients with inoperable progressive or relapsing PVNS, or PVNS only resectable with mutilating surgery. 31 Many patients (41%) had AEs requiring treatment modification and 11% of patients had various grade 3 or higher AEs deemed to be related to nilotinib treatment.
Encouraging results of a phase 1 trial of the CSF1R inhibitor pexidartinib 32 led to the ENLIVEN 24-week double-blind, placebocontrolled RCT of pexidartinib in patients with symptomatic advanced TGCT at various sites (primarily knee and ankle). 33 Compared with the placebo group (n = 59), the twice- Postop: wound revisions due to hematomas (n = 2); necrosis of femoral condyle (n = 1); peroneal nerve palsy (n = 1); deep infection (n = 1); instability of collateral ligament at knee (n = 1)      three patients (5%) with aspartate aminotransferase and alanine aminotransferase $ 3 · upper limit of normal, along with total bilirubin and alkaline phosphatase $ 2 · upper limit of normal. The only death reported in the study (in a patient with a history of cardiovascular disease) was unrelated to pexidartinib  treatment. The cases of mixed or cholestatic hepatotoxicity seen in the pexidartinib arm and other cases reported in non-TGCT studies led to early closing of study recruitment after 120 (95%) of the planned 126 patients had enrolled and led to stoppage of crossover from placebo to the pexidartinib arm in the nonrandomized extension phase of the trial. A retrospective cohort study (n = 62) reported a 31% ORR, including 4% CR and 27% PR, and 65% SD with once-daily oral imatinib in patients with advanced or recurrent diffuse TGCT. 34 This study included longterm follow-up of 29 patients previously reported in a 2012 publication 38 headache, dizziness, hepatic disorders (n = 1); pruritus and toxidermia (n = 1); diarrhea (n = 1); increased g-glutamyl transferase concentrations (n = 1); anorexia (n = 1); increased headache (n = 1) Tap et al 32      no statistically significant differences in recurrence rate (8.6% versus 9.1%) 20 or 5-year LRFS (72% versus 84%) 28 and 1 study found that open surgery was associated with higher LRFS (87% versus 80%). 27 The two largest TGCT cohort studies 11,27 included in our review (and not included in previous systematic reviews) noted that the statistically significant improvements in LRFS associated with open versus arthroscopic surgery observed by univariate analyses disappeared following multivariate analyses. Although one might assume that complications of open surgery would be higher than complications of arthroscopic surgery, there are no data from head-to-head RCTs to support or refute this supposition. Existing published data are circumstantial and sparse. Thus far, we have not found clear differences in complication rates between open or arthroscopic one-or two-staged surgery.
Thus, regarding surgical treatments, the results of our systematic review align with previous reports of some-what contradictory results and high risk of disease progression, especially in patients with diffuse TGCT. In a 2017 systematic review of long-term clinical outcomes and rates of recurrence for open or arthroscopic excision in patients with GCTTS involving four types of joints (ie, shoulder, hip, knee, and ankle), Gouin's group concluded that although arthroscopy showed effectiveness for localized disease for all four joints, the data supported arthroscopy only for the knee joint because of the risk of osteoarthritis degradation. 39 Unfortunately, the findings from our study do not bring more clarity to the open versus arthroscopic debate. At present, more and more centers use a hybrid arthroscopic anterior and open posterior approach in patients with diffuse TGCT of the knee, including two studies 14,20 that met the inclusion criteria for our analysis and 1 study that had too few patients. 40 Interpretation of the combined surgical studies is restricted by study design limitations, particularly the retrospective design and compar-isons of noncontemporaneous cohorts (with study periods spanning more than a decade in most cases). Lack of control for potential sources of bias such as differences in surgeon's experience and learning curve, patient clinical characteristics, outcome measures, and length of follow-up are also limiting factors. Better quality studies-especially RCTs or case-matched analysesare needed to determine which surgical techniques are optimal for TGCT. RCTs are expensive and difficult to perform in a surgical setting, especially for a rare disease such as TGCT.
Among the four agents evaluated in clinical trials identified in our literature review, emactuzumab, 30 imatinib, 34    no ongoing studies. Since its initial US approval in 2001, imatinib had gained several indications for various hematologic malignancies, dermatofibrosarcoma protuberans, and gastrointestinal stromal tumors 41 ; its administration to patients with TGCT would be off-label. The ENLIVEN study met its primary end point with a higher ORR for pexidartinib compared with the placebo. 33 These results supported the August 2019 United States FDA approval of pexidartinib for the treatment of adults with symptomatic TGCT associated with severe morbidity or functional limitations and not amenable to improvement with surgery. Because the drug was approved with a boxed warning regarding the risk of serious and potentially fatal liver injury, it is prescribed and dispensed solely via a manufacturer-supported Risk Evaluation and Mitigation Strategy safety program. 42 As of August 2019, the National Comprehensive Cancer Network clinical practice guidelines for soft tissue sarcoma list only two systemic therapy agents for TGCT: pexidartinib as a category 1 recommendation and imatinib as a category 2A recommendation. 10 In the context of pexidartinib, additional research and clinical experience are warranted to better understand optimal patient selection, treatment course, patient adherence to treatment, and prevention and management of toxicities. Our systematic review was limited by the restriction of PubMed and Ovid searches to studies published from 2013 to the present to reflect the most current treatment practices for patients with TGCT. Other limitations included exclusion of unpublished studies, studies of EBRT as the primary treatment modality, and studies with fewer than 20 patients. By limiting studies to those with histologically confirmed patients, we were able to more accurately reflect the current standard of care.
Investigation of other systemic therapies for TGCT is currently limited to a few ongoing studies. These include an open-label phase 1 study (NCT03069469) of the oral CSF1R inhibitor DCC-3014 in patients with solid tumors, including tumor types with high colony-stimulating factor 1 expression such as TGCT and an ongoing Japanese randomized placebo-controlled double-blind phase 2 trial that is evaluating the nonsteroidal anti-inflammatory drug zaltoprofen in patients with diffuse TGCT and unresectable localized TGCT. 43 One important question that needs to be addressed is what is the best outcome measure to assess treatment in patients with TGCT. The studies of surgical treatments assessed local progression or local recurrence (which is often not a recurrence but a local progression). Would different measures be more appropriate for different treatment modalities (eg, surgery versus systemic therapy)? Because cure is rare, should we be assessing SD without complaints?
In 2012, van der Heijden et al 4 suggested that a multidisciplinary approach is required to improve outcomes of patients with recurrent and refractory diffuse disease. This approach should include dedicated magnetic resonance imaging, histologic assessment, and planned surgery with adjuvant radiotherapy or systemic targeted therapy. The subsequent collective experience supports the notion that multidisciplinary treatment is key to better outcomes for patients with TGCT, particularly given the tumor's inflammatory response and ensuing symptoms. The US FDA approval of pexidartinib, the first systemic drug therapy for selected adults with symptomatic TGCT, augments the armamentarium available for multidisciplinary treatment. Questions remain regarding the optimal combination and sequencing of available therapies. Surgeons, medical oncologists, and other clinicians have a unique opportunity to work together to improve health and quality of life outcomes for this patient population.

Conclusions
The significant rates of recurrence and risks associated with surgery point to the need for novel systemic treatments for patients with symptomatic advanced TGCT. The recent approval of a systemic therapy for selected adults with symptomatic TGCT underscores the need for improved and more coordinated multidisciplinary care.