Background:
Tenosynovial giant cell tumor (TGCT) may be misdiagnosed as osteoarthritis (OA), or the chronic course of TGCT may lead to development of secondary OA. However, little is known about the effect of comorbid OA on long-term surgical patterns and costs among TGCT patients.
Methods:
This cohort study used claims data from the Merative MarketScan Research Databases. The study included adults diagnosed with TGCT from January 1, 2014, to June 30, 2019, who have at least 3 years of continuous enrollment before and after the first TGCT diagnosis (date of the first TGCT diagnosis = index date) and no other cancer diagnosis during the study period. Patients were stratified by the presence of an OA diagnosis relative to the index date. Outcomes included surgical procedure patterns, healthcare resource utilization, and costs in the 3-year pre- and postindex periods. Multivariable models were used to assess the effect of OA on the study outcomes, controlling for baseline characteristics.
Results:
The study included 2856 TGCT patients: 1153 (40%) had no OA before or after index (OA[−/−]), 207 (7%) had OA before index but not after (OA[+/−]), 644 (23%) had OA after index but not before (OA[−/+]), and 852 (30%) had OA before and after index (OA[+/+]). The mean age was 51.6 years, and 61.7% were female. During the postperiod, joint surgery was more common among OA(−/+) and OA(+/+) patients compared with OA(−/−) and OA(+/−) patients (55.7% vs 33.2%). The mean all-cause total costs in the 3-year postperiod were $19,476 per patient per year. Compared with OA(−/−) patients, OA(−/+) and OA(+/+) patients had a higher risk of undergoing recurrent surgery and higher total healthcare costs postindex.
Discussion:
Higher rates of surgery and increased healthcare cost observed in TGCT patients with postindex OA underscore the need for effective treatment options to reduce joint damage, especially among patients with comorbid OA.
Introduction
Tenosynovial giant cell tumor (TGCT) is a rare, nonmalignant tumor that involves synovium found in joints, bursae, or tendon sheaths.1 TGCT occurs most commonly at 20 to 50 years of age and presents in either a localized or diffuse form.2 In the United States, the annual incidence of TGCT is estimated at 9.8 and 1.2 cases per million for localized and diffuse TGCT, respectively.3 Localized TGCT typically presents as single, well-demarcated tumor, whereas diffuse TGCT presents as multiple nodules throughout the synovium with poorly defined borders, most commonly affecting larger joints, such as the knee, hip, shoulder, and ankle.
The most common symptoms of TGCT include pain, stiffness, swelling, and reduced range of motion,4,5 which resemble the symptoms of other joint conditions, such as osteoarthritis (OA). Recurrent atraumatic hemarthroses are often a part of the clinical presentation. Because of the rarity and nonspecific clinical presentation, diagnosis of TGCT is often delayed by an average of 2 to 3 years.6,7 In some cases, patients were initially diagnosed with and treated for OA and subsequently diagnosed with TGCT during total knee arthroplasty.8 Although surgical resection is the mainstay of treatment for TGCT, recurrence rates remain high, with reports up to 20% in the localized type and 50% in the diffuse type. Recurrent disease will likely require multiple surgeries leading to substantial morbidity of the joint, which may accelerate a TGCT-mediated degenerative process toward secondary OA.9 TGCT, especially the diffuse type affecting the knee, hip, and ankle, presents a higher risk of OA, resulting from both the joint destruction from the disease itself and from the multiple operations on recurrences.9
Burden of illness measured by healthcare resource utilization and healthcare costs has been reported independently for TGCT and OA.10-13 Previous studies have shown that healthcare costs among TGCT patients increase substantially in the first year after TGCT diagnosis, primarily driven by outpatient visits.10 TGCT is also associated with higher healthcare resource utilization and decreased work productivity compared with patients without TGCT.11 Medical care for OA, the most prevalent joint disorder in the United States, was estimated to cost $12,505 per patient per year (PPPY),12 resulting in an estimated healthcare cost of over $318 billion in 2015.12 The annual all-cause healthcare cost was estimated as $24,550 for patients with knee OA in 2017 compared with $16,843 for matched controls without knee OA.13
The objective of this study is to assess the effect of comorbid OA on surgical patterns, healthcare resource utilization, and healthcare costs of patients with TGCT. To account for the chronic nature of both OA and TGCT, the occurrence of joint surgeries, healthcare resource utilization, and healthcare costs, stratified by the presence and timing of an OA diagnosis in relation to the index TGCT diagnosis, were analyzed during the 3 years before and after the initial TGCT diagnosis. The findings of this study provide insights on how comorbid OA affect the disease burden of patients with TGCT.
Methods
Data Source
This study used deidentified US administrative claims data from the Merative MarketScan Commercial Claims and Encounters (Commercial) and Medicare databases from January 1, 2011, to June 30, 2019. The database captures medical care (inpatient and outpatient services, office-administered medications, and use of medical equipment) and pharmacy services. The commercial database includes employees and their dependents covered under a variety of fee-for-service and managed care health plans, and the Medicare database includes Medicare-eligible workers and retirees with employer-sponsored Medicare coverage. In 2019, approximately 21.2 million lives are represented in the two databases. Because this study used only deidentified patient records and did not involve the collection, use, or transmittal of individually identifiable data, it was exempted by the institutions from institutional review board approval.
Patient Selection
Patients with at least one inpatient or two nondiagnostic outpatient claims at least 30 days apart with a diagnosis code for TGCT (International Classification of Diseases, 9th and 10th Revision, Clinical Modification diagnosis codes 727.02 and 719.2x or International Classification of Diseases, 10th Revision, Clinical Modification D48.1, D21.0-D21.9, and M12.2) between January 1, 2014, and June 30, 2019, were identified. The index date was set as the date of the earliest TGCT diagnosis. Patients were required to be at least 18 years old on the index date and have continuous enrollment with medical and pharmacy benefits for at least 3 years before the index date (baseline period) and at least 3 years from the index date (follow-up period). Patients with evidence of TGCT during the baseline period or evidence of any cancer during the baseline or follow-up periods were excluded. Patients were categorized into four subgroups based on the presence and timing of an OA diagnosis in relation to the index TGCT diagnosis: no OA claims before or after index, hereafter reported as OA(−/−), OA claims before index but not after, hereafter reported as OA(+/−), OA claims after index but not before, hereafter reported as OA(−/+), and OA claims before and after index, hereafter reported as OA(+/+).
Outcomes
Demographic characteristics, measured on the index date, included age, sex, geographic region, and health plan type. Clinical characteristics, measured in a 1-year preindex period, included comorbidities and the Charlson Comorbidity Index (CCI).
Surgical patterns were reported in each subgroup during the 3-year baseline and follow-up periods. The number of patients with a joint surgical procedure (arthroscopic excision, open excision, arthrodesis, arthroplasty, or amputation) and the number of surgeries were reported. Procedures that occurred within 30 days of the initial procedure claim were counted as one surgical episode.
All-cause healthcare resource utilization and costs included inpatient admissions, emergency room visits, outpatient services, and outpatient pharmacy and were reported PPPY for each subgroup during the 3-year baseline and follow-up periods. Outpatient services of interest included imaging procedures (CT, MRI, positron emission tomography, radiograph, and ultrasonography), radiation procedures, and supplemental care (chiropractic, acupuncture, and physical therapy services).
Statistical Analysis
Descriptive statistics were reported using mean and SD for continuous variables and frequency and percentage for categorical variables. Cox proportional hazards models were used to assess the effect of having OA on the time from the index date to first surgery and on the time from the date of first surgery to second surgery during the follow-up period, controlling for demographics and clinical characteristics, including age, sex, geographic region, comorbidities, payer type, and surgery history preindex. Healthcare costs during the follow-up period were compared with the baseline period using Student t-tests. A generalized linear model was used to assess the effect of having OA on the PPPY total healthcare costs during the follow-up period, controlling for demographics and clinical characteristics (age, sex, geographic location, comorbidities, payer type, and PPPY cost preindex). All data analyses were conducted using WPS Analytics version 4.1 (World Programming, United Kingdom).
Results
Patient Characteristics
A total of 2856 TGCT patients met the selection criteria (Figure 1). The mean (SD) age was 51.6 (12.8) years, 61.7% were female, and 87% were covered by commercial insurance (Table 1). Approximately 40.4% had no OA claims before or after index (OA[−/−] subgroup), 22.5% had OA claims before index but not after (OA[+/−] subgroup), 7.2% had OA claims after index but not before (OA[−/+] subgroup), and 29.8% had OA claims before and after index (OA[+/+] group). The mean age was lowest in the OA(−/−) subgroup and highest in the OA(+/+) subgroup. In general, the subgroup with OA(+/+) was older (mean [SD] age = 58.1 [10.9] years) and had higher comorbidity burden (mean [SD] CCI score = 0.7 [1.1]).
Figure 1.
Patient selection. OA = osteoarthritis, TGCT = tenosynovial giant cell tumor
Table 1.
Patient Demographics and Baseline Clinical Characteristics
| Demographic/Characteristic | All Patients N = 2856 |
Without OA Postindex | With OA Postindex | ||
| OA(−/−) n = 1153 |
OA(+/−) n = 207 |
OA(−/+) n = 644 |
OA(+/+) n = 852 |
||
| Agea, mean (SD) | 51.6 (12.8) | 46.0 (12.7) | 52.6 (12.3) | 52.9 (11.1) | 58.1 (10.9) |
| Age categorya, n (%) | |||||
| 18–34 | 244 (8.6) | 190 (16.5) | 14 (6.8) | 34 (5.3) | 6 (0.7) |
| 35–44 | 510 (17.9) | 302 (26.2) | 31 (15.0) | 93 (14.4) | 84 (9.9) |
| 45–54 | 893 (31.3) | 368 (31.9) | 76 (36.7) | 236 (36.7) | 213 (25.0) |
| 55–64 | 859 (30.1) | 242 (21.0) | 61 (29.5) | 207 (32.1) | 349 (41.0) |
| 65+ | 350 (12.3) | 51 (4.4) | 25 (12.1) | 74 (11.5) | 200 (23.6) |
| Female sexa, n (%) | 1761 (61.7) | 715 (62.0) | 122 (58.9) | 393 (61.0) | 531 (62.3) |
| Geographic regiona, n (%) | |||||
| Northeast | 1068 (37.4) | 397 (34.4) | 87 (42.0) | 242 (37.6) | 342 (40.1) |
| North central | 402 (14.1) | 155 (13.4) | 21 (10.1) | 81 (12.6) | 145 (17.0) |
| South | 1087 (38.1) | 451 (39.1) | 78 (37.7) | 255 (39.6) | 303 (35.6) |
| West | 292 (10.2) | 146 (12.7) | 20 (9.7) | 64 (9.9) | 62 (7.3) |
| Unknown | 7 (0.3) | 4 (0.4) | 1 (0.5) | 2 (0.3) | 0 (0.0) |
| Payera, n (%) | |||||
| Commercial | 2493 (87.3) | 1100 (95.4) | 179 (86.5) | 567 (88.0) | 647 (75.9) |
| Medicare supplemental | 363 (12.7) | 53 (4.6) | 28 (13.5) | 77 (12.0) | 205 (24.1) |
| CCIb, mean (SD) | 0.5 (1.0) | 0.3 (0.8) | 0.6 (1.2) | 0.4 (0.8) | 0.7 (1.1) |
| Selected comorbiditiesb, n (%) | |||||
| Gout | 51 (1.8) | 17 (1.5) | 2 (1.0) | 12 (1.9) | 20 (2.3) |
| Hypertension | 1033 (36.2) | 275 (23.9) | 90 (43.5) | 228 (35.4) | 440 (51.6) |
| Obesity | 441 (15.4) | 135 (11.7) | 40 (19.3) | 91 (14.1) | 175 (20.5) |
| Rheumatoid arthritis | 78 (2.7) | 8 (0.7) | 6 (2.9) | 12 (1.9) | 52 (6.1) |
| Type 2 diabetes | 361 (12.6) | 86 (7.5) | 37 (17.9) | 88 (13.7) | 150 (17.6) |
CCI = Charlson Comorbidity Index, OA = osteoarthritis
Demographic characteristics were measured on the index date.
Clinical characteristics were measured during the 1-year preindex period.
Surgical Patterns
During the 3-year baseline period, 14.7% of patients underwent joint surgery in any location, whereas 45.0% had surgery in any location during the follow-up period (Tables 2 and 3). More patients with OA postindex had surgery than those without OA claims postindex (55.7% vs 33.2%). The mean (SD) time from the index date to the first surgery was shortest in the OA(−/−) subgroup at 126.3 (220.2) days and longest in the OA(−/+) subgroup at 210.7 (305.5) days. The mean (SD) time from first surgery to second surgery was shortest in the OA(+/+) subgroup (384.6 [261.7] days).
Table 2.
Surgical Patterns of First Surgery for TGCT Patients With and Without OA Over 3 Years of Follow-up
| All Patients N = 2856 |
Without OA Postindex | With OA Postindex | |||
| OA(−/−) n = 1153 |
OA(+/−) n = 207 |
OA(−/+) n = 644 |
OA(+/+) n = 852 |
||
| Patients with preindex surgery, n (%) | 419 (14.7) | 69 (6.0) | 50 (24.2) | 57 (8.9) | 57 (8.9) |
| Patients with postindex surgery, n (%) | 1285 (45.0) | 380 (33.0) | 71 (34.3) | 318 (49.4) | 516 (60.6) |
| Time from index to first surgery postindex (days), mean (SD), median | 170.6 (274.9) 41.0 |
126.3 (220.2) 41.0 |
169.9 (288.2) 40.0 |
210.7 (305.5) 56.0 |
178.7 (285.4) 30.0 |
| Type of first surgery postindex, N (%) | |||||
| Arthroscopic excision | 564 (43.9) | 126 (33.2) | 26 (36.6) | 177 (55.7) | 235 (45.5) |
| Open excision | 472 (36.7) | 248 (65.3) | 39 (54.9) | 96 (30.2) | 89 (17.2) |
| Arthrodesis | 50 (3.9) | 4 (1.1) | 5 (7.0) | 14 (4.4) | 27 (5.2) |
| Arthroplasty | 197 (15.3) | 2 (0.5) | 1 (1.4) | 30 (9.4) | 164 (31.8) |
| Amputation | 2 (0.2) | 0 (0.0) | 0 (0.0) | 1 (0.3) | 1 (0.2) |
OA = osteoarthritis, TGCT = tenosynovial giant cell tumor
Table 3.
Surgical Patterns of Second Surgery for TGCT Patients With and Without OA Over 3 Years of Follow-up
| All Patients N = 2856 |
Without OA Postindex | With OA Postindex | |||
| OA(−/−) n = 1153 |
OA(+/−) n = 207 |
OA(−/+) n = 644 |
OA(+/+) n = 852 |
||
| Patients with at least two surgeries postindex among patients with surgery, n (%) | 378 (29.4) | 70 (18.4) | 11 (15.5) | 104 (32.7) | 193 (37.4) |
| Time from first surgery to second surgery postindex (days), mean (SD), median | 414.2 (277.6), 363.0 | 437.2 (298.2), 385.0 | 485.5 (338.2), 394.0 | 445.9 (283.1), 360.5 | 384.6 (261.7), 331.0 |
| Type of second surgery postindex, N (%) | |||||
| Arthroscopic excision | 160 (12.5) | 32 (8.4) | 4 (5.6) | 53 (16.7) | 71 (13.8) |
| Open excision | 98 (7.6) | 37 (9.7) | 7 (9.9) | 25 (7.9) | 29 (5.6) |
| Arthrodesis | 26 (2.0) | 1 (0.3) | 0 (0.0) | 8 (2.5) | 17 (3.3) |
| Arthroplasty | 94 (7.3) | 0 (0.0) | 0 (0.0) | 18 (5.7) | 76 (14.7) |
| Amputation | 0 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) |
OA = osteoarthritis, TGCT = tenosynovial giant cell tumor
After controlling for baseline characteristics, adjusted hazard of joint surgery in OA(+/+) patients was more than twice that of the OA(−/−) patients {adjusted hazard ratio (HR) (95% confidence interval [CI]) = 2.48 (2.13 to 2.88)} (Figure 2). Adjusted hazard of joint surgery in OA(−/+) patients was almost twice that of the OA(−/−) patients (adjusted HR [95% CI] = 1.74 [1.49 to 2.03]) (Figure 2). Compared with OA(−/−) patients, the hazard of revision surgery (second surgery) was higher in OA(+/+) patients (adjusted HR [95% CI] = 2.28 [1.68 to 3.11]) and in OA(−/+) patients (adjusted HR [95% CI] = 2.18 [1.60–2.97]) (Figure 3).
Figure 2.
Cox proportional hazards regression model results evaluating the time to first joint-related surgery of TGCT patients with and without OA over 3 years of follow-up. The sample population was 2856 TGCT patients (1285 with a first surgery; 1571 without a first surgery). CI = confidence interval, HR = hazard ratio, OA = osteoarthritis, TGCT = tenosynovial giant cell tumor
Figure 3.
Cox proportional hazards regression model results evaluating the time to second surgery of TGCT patients with and without OA over 3 years of follow-up. The sample population was 1285 TGCT patients with a first surgery (378 with a second surgery; 907 without a second surgery). CCI = Charlson Comorbidity Index, CI = confidence interval, HR = hazard ratio, OA = osteoarthritis, TGCT = tenosynovial giant cell tumor
Healthcare Resource Utilization
During the follow-up period, inpatient admission was infrequent in all four subgroups. Nevertheless, patients with comorbid OA, especially those with OA postindex, were more likely to have an inpatient admission (Table 4). The median number of outpatient office visits was higher among patients with an OA diagnosis (5.7 OA[−/−] compared with 7.7 OA[+/−], 8.3 OA[−/+], and 11.0 OA[+/+]). Orthopaedic surgeon visit was the most common type of specialist visit. Similarly, the median number of visits for supplemental care services, which included chiropractic, physical therapy, and acupuncture services, was also higher in patients with an OA diagnosis (0 OA[−/−] compared with 3.0 OA[+/−], 6.3 OA[−/+], and 11.3 OA[+/+]).
Table 4.
PPPY All-Cause Healthcare Resource Utilization of TGCT Patients With and Without OA Over 3 Years of Follow-up
| All Patients N = 2856 |
Without OA Postindex | With OA Postindex | |||
| OA(−/−) n = 1153 |
OA(+/−) n = 207 |
OA(−/+) n = 644 |
OA(+/+) n = 852 |
||
| Inpatient | |||||
| Patients with an admission, n (%) | 721 (25.2) | 179 (15.5) | 35 (16.9) | 155 (24.1) | 352 (41.3) |
| No. of inpatient admissions, mean (SD), median | 0.1 (0.3), 0.0 | 0.1 (0.2), 0.0 | 0.1 (0.2), 0.0 | 0.1 (0.2), 0.0 | 0.3 (0.5), 0.0 |
| Length of stay (d) per admission, mean (SD), median | 3.5 (3.6), 3.0 | 3.5 (3.5), 2.0 | 3.6 (3.0), 3.0 | 3.4 (3.3), 2.5 | 3.6 (3.9), 3.0 |
| ED visits | |||||
| Patients with an ED visit, n (%) | 1276 (44.7) | 451 (39.1) | 98 (47.3) | 304 (47.2) | 423 (49.6) |
| No. of ED visits, mean (SD), median | 0.4 (0.9), 0.0 | 0.3 (0.6), 0.0 | 0.5 (1.6), 0.0 | 0.4 (0.7), 0.0 | 0.5 (1.0), 0.0 |
| Supplemental care visitsa | |||||
| Patients with a supplemental care visit, n (%) | 1841 (64.5) | 567 (49.2) | 127 (61.4) | 481 (74.4) | 666 (78.2) |
| No. of supplemental care visits, mean (SD), median | 13.4 (23.9), 4.0 | 7.5 (15.8), 0.0 | 11.5 (22.9), 3.0 | 15.1 (23.6), 6.3 | 20.6 (30.5), 11.3 |
| Imaging procedureb | |||||
| Patients with an imaging procedure (n, %) | 2586 (90.5) | 940 (81.5) | 624 (96.9) | 187 (90.3) | 835 (98.0) |
| No. of imaging procedures, mean (SD), median | 2.7 (2.8), 2.0 | 1.5 (1.7), 1.0 | 3.0 (2.4), 2.3 | 2.1 (2.2), 1.7 | 4.1 (3.5), 3.3 |
| Patients with a CT scan, n (%) | 911 (31.9) | 274 (23.8) | 224 (34.8) | 76 (36.7) | 337 (39.6) |
| Patients with an MRI, n (%) | 1171 (41.0) | 350 (30.4) | 333 (51.7) | 63 (30.4) | 425 (49.9) |
| Patients with a PET scan, n (%) | 7 (0.2) | 3 (0.3) | 1 (0.2) | 0 (0.0) | 3 (0.4) |
| Patients with a radiographc, n (%) | 2133 (74.7) | 658 (57.1) | 563 (87.4) | 136 (65.7) | 776 (91.1) |
| Patients with an ultrasonography, n (%) | 919 (32.2) | 255 (22.1) | 256 (39.8) | 72 (34.8) | 336 (39.4) |
| Outpatient office visitsd | |||||
| Patients with an office visit, n (%) | 2852 (99.9) | 1150 (99.7) | 207 (100) | 644 (100) | 851 (99.9) |
| Patients with office visit with rheumatologiste, n (%) | 322 (11.3) | 61 (5.3) | 12 (5.8) | 93 (14.4) | 156 (18.3) |
| Patients with office visit with orthopaedic surgeone, n (%) | 1880 (65.8) | 547 (47.4) | 105 (50.7) | 496 (77.0) | 732 (85.9) |
| Patients with office visit with oncologiste, n (%) | 71 (2.5) | 29 (2.5) | 4 (1.9) | 21 (3.3) | 17 (2.0) |
| No. of office visits, mean (SD), median | 9.4 (7.2), 7.7 | 6.7 (5.0), 5.7 | 9.5 (7.5), 7.7 | 10.0 (6.8), 8.3 | 12.7 (8.4), 11.0 |
| Outpatient pharmacy | |||||
| Patients with a prescription, n (%) | 2804 (98.2) | 1119 (97.1) | 203 (98.1) | 639 (99.2) | 843 (98.9) |
| No. of prescriptions, mean (SD), median | 23.3 (22.8), 16.7 | 15.1 (15.9), 10.0 | 24.0 (21.6), 16.7 | 24.2 (21.7), 18.5 | 33.5 (27.2), 27.8 |
ED = emergency department, OA = osteoarthritis, PET = positron emission tomography, PPPY = per patient per year, TGCT = tenosynovial giant cell tumor
Supplemental care visits included chiropractic, physical therapy, and acupuncture services.
Imaging procedures include CT, MRI, and PET scans, radiographs, and ultrasonography.
Radiograph encompassed a range of procedures for radiography restricted to bones, joints, and body areas affected by TGCT.
Specific provider types reported will not sum to the total number of patients with an outpatient office visit. Patients could have visits in multiple categories.
Specialist office visits were identified by a medical claim with an outpatient office visit with a corresponding provider code. This code may not always be used, or the field may not be populated. Patients could have visits in multiple categories.
Healthcare Costs
Among all patients, the total mean cost difference between the 3-year baseline and follow-up periods increased by $6378 from $13,098 to $19,476 PPPY (Figure 4). Although no statistically significant changes in mean total costs were observed in the subgroup of patients with OA(+/−), the mean PPPY cost difference between the pre- and postperiods increased by $9248 in the OA(−/+) subgroup, $9524 in the OA(+/+) subgroup, and $3392 in the OA(−/−) subgroup. In all subgroups of patients with OA postindex, inpatient, outpatient, and outpatient pharmacy costs were all significantly greater during follow-up compared with the baseline period.
Figure 4.
Mean PPPY all-cause healthcare costs of TGCT patients with and without OA over 3 years of follow-up. OA = osteoarthritis, PPPY = per patient per year, TGCT = tenosynovial giant cell tumor
Compared with OA(−/−) patients, adjusted all-cause total healthcare costs postindex were significantly higher in OA(−/+) patients (adjusted cost ratio [95% CI]) = 1.49 [1.37–1.62]) and in OA(+/+) patients (adjusted cost ratio [95% CI] = 1.68 [1.54–1.83]), but similar in OA(+/−) patients (Figure 5).
Figure 5.
Adjusted PPPY all-cause total costs of TGCT patients with and without OA over 3 years of follow-up. GLMs evaluating the outcome PPPY all-cause total costs over 3 years of follow-up. The sample population was 2856 TGCT patients. CI = confidence interval, GLM = generalized linear model, OA = osteoarthritis, PPPY = per patient per year, TGCT = tenosynovial giant cell tumor
Discussion
The symptoms of TGCT, such as pain, stiffness, swelling, and reduced range of motion, resemble symptoms of OA,4,5 and the literature has reported cases in which joint revision surgery might be required to alleviate symptoms of OA secondary to TGCT.9 To investigate how comorbid OA impacts the surgical treatment and burden of illness associated TGCT, we characterized the impact of comorbid OA on these outcomes. In this study, we found that 59.3% of TGCT patients had a comorbid OA diagnosis at some point of time during the study, higher than the estimated prevalence of doctor-diagnosed arthritis of approximately 24% in the general population in the United States.14,15 A third of patients maintained an OA diagnosis after confirmation of TGCT diagnosis, suggesting that these patients might not be initially misdiagnosed as OA. A small group of TGCT patients had a medical claim for OA before but not after TGCT diagnosis;9 these individuals may have been misdiagnosed with OA initially, although the claims data did not allow confirmation of this hypothesis.
Although the mainstay treatment for TGCT is surgical excision,7,16 less than half of patients in our study underwent joint surgery. Our findings are comparable to a retrospective claims study with 12-month follow-up postindex TGCT diagnosis in the United States (which reported 50.9%),10 but lower than the surgery proportion reported from an international multicenter chart review study of diffuse TGCT with a median follow-up time of 54 months (which reported 81%).17 The differences in the proportion of patients undergoing surgery might be impacted by clinical practice across various geographic regions, type of TGCT patients included, and length of follow-up.10 A greater proportion of patients with OA both before and after TGCT diagnosis (60%) underwent surgery compared with those without OA (33%). The adjusted risk of joint surgery postindex, controlling for baseline characteristics (eg, demographics and CCI) and evidence of joint surgery before TGCT diagnosis, remained significantly higher in patients with OA after TGCT diagnosis (HR, 2.48) regardless of whether OA was diagnosed before index or not.
Almost 30% of patients had at least two surgeries during follow-up, and the risk of recurrent surgery was twice as high in patients with OA compared with those without. Time from initial surgery to revision surgery was shorter in those with OA. Because local recurrence might develop years after initial surgery,18 and our follow-up period was 3 years after initial diagnosis, it is plausible that our estimates are an underestimation of the true recurrence rate. Factors associated with TGCT recurrence include tumor size, extent of the tumor, and initial treatment.16 A systematic review and a more recent meta-analysis on TGCT treatment concluded that open resection was associated with a lower risk of recurrence compared with arthroscopy,16,19 whereas arthroplasty was indicated in the case of joint destruction or development of OA.16,20 In our study, arthroplasty and arthroscopic excisions were the most common types of surgery among those with OA, whereas open excisions followed by arthroscopic excisions were most common in those without OA. These findings suggest that patients with comorbid OA may have significant joint dysfunction and more complex disease progression, necessitating arthroplasty as the first surgery in some cases.
Overall, healthcare resource utilization in the post-TGCT diagnosis period was higher in TGCT patients with OA, with increased utilization of inpatient admissions and outpatient services, including supplemental care (chiropractic, physical therapy, and acupuncture) and specialist visits. The increased healthcare resource utilization contributed to higher all-cause healthcare costs. Overall mean healthcare costs increased from $13,098 to $19,476 PPPY (a factor of 1.5), comparing the 3-year baseline and follow-up periods, which was consistent with findings from other healthcare cost studies.10,11,21 The retrospective analysis by Burton et al.10 estimated an increase in healthcare costs from $12,960 to $17,555 (a factor of 1.4) comparing the 12-month pre- and post-TGCT diagnosis periods. This is comparable to findings from the retrospective analysis by Lin et al11 that showed an increase in mean total healthcare costs from $10,394 at baseline to $15,382 (a factor of 1.5) in the postindex period among TGCT patients.11 Compared with patients without OA, healthcare costs were 68% higher in patients with OA(+/+) and 49% higher in OA(−/+) patients in the current study. The additional costs associated with OA among TGCT patients indicate that OA contributes to a substantial additional burden in healthcare costs among TGCT patients, and these costs may be tied to the increased healthcare resource utilization associated with OA management.
This study was subject to limitations that are inherent to retrospective administrative claims analysis. Healthcare resource utilization and cost analyses were limited to only those individuals with commercial health coverage or private Medicare supplemental coverage. Consequently, results of this analysis may not be generalizable to TGCT patients with other insurance or without health insurance coverage. In addition, information on patient clinical profiles and study outcomes was obtained from diagnosis codes, procedure codes, and pharmacy claims and may be subject to missed or incorrect coding, which may introduce bias or measurement errors. Furthermore, there are no specific diagnosis codes for TGCT, but the codes selected were confirmed by a previous analysis. Because OA can occur before or after surgery, the multivariable analyses could not establish a causal link between OA and TGCT-related surgery or the temporal relationship between OA diagnosis and timing of joint surgery. Finally, the underlying reason for conducting procedures, such as tumor location, size and volume, and patient preference, is not available within claims data, so it is possible that the joint procedures were performed for reasons other than TGCT.
Conclusion
OA was a common comorbid condition in TGCT patients. In this study, patients who had a comorbidity of OA generally fared worse, regardless of whether it was diagnosed before TGCT diagnosis or after. The high rates of surgery, revision surgeries, and increased healthcare costs observed in TGCT patients with postindex OA underscore the need to control the disease before the development of secondary OA. In this context, the difficulty of controlling the disease with surgery alone is highly concerning. This reinforces the need for effective nonsurgical treatment options for TGCT, particularly in patients with comorbid OA.
Acknowledgments
Medical writing support was provided by Jessamine-Winer Jones, PhD, and analytic services were provided by Xue Song, PhD, and Matthew Brouillette, MPH, who were employees of Merative, Inc. at the time the study was conducted. Programming services were provided by Laurie Costa, and statistical analysis services were provided by David Smith, PhD, who are employees of Merative, Inc. These services were paid for by Daiichi Sankyo, Inc.
Footnotes
This study was funded by Daiichi Sankyo, Inc.
Dr. Lin or an immediate family member is an employee of Daiichi Sankyo, Inc., and has stock or stock options held in Daiichi Sankyo, Inc. Dr. Kwong or an immediate family member is an employee of Daiichi Sankyo, Inc., and has stock or stock options held in Daiichi Sankyo, Inc. Wilson or an immediate family member is an employee of Merative, Inc., which received consulting fees from Daiichi Sankyo, Inc., to conduct this study. Dr. Abraham has received consulting fees from Daiichi Sankyo Inc.
Dr. Lin, Wilson, Dr. Kwong, and Dr. Abraham contributed to the study design and interpretation of the data. Wilson and Dr. Lin contributed to manuscript writing and revisions. All authors read and approved the final manuscript.
Contributor Information
Kathleen Wilson, Email: kwilson@merative.com.
Winghan J. Kwong, Email: wkwong@dsi.com.
John A. Abraham, Email: abrahaja@gmail.com.
References
- 1.Verspoor FG, van der Geest IC, Vegt E, Veth RP, van der Graaf WT, Schreuder HB: Pigmented villonodular synovitis: Current concepts about diagnosis and management. Future Oncol 2013;9:1515-1531. [DOI] [PubMed] [Google Scholar]
- 2.de Saint Aubain Somerhausen N: Soft Tissue and Bone Tumours wCoT, ed 5 Cancer. LIIAfRo, 133-136. [Google Scholar]
- 3.Myers BW, Masi AT, Feigenbaum SL: Pigmented villonodular synovitis and tenosynovitis: A clinical epidemiologic study of 166 cases and literature review. Medicine 1980;59:223-238. [PubMed] [Google Scholar]
- 4.Mastboom MJ, Planje R, van de Sande MA: The patient perspective on the impact of tenosynovial giant cell tumors on daily living: Crowdsourcing study on physical function and quality of life. Interact J Med Res 2018;7:e4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Gelhorn HL, Tong S, McQuarrie K, et al. : Patient-reported symptoms of tenosynovial giant cell tumors. Clin Ther 2016;38:778-793 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Xie GP, Jiang N, Liang CX, et al. : Pigmented villonodular synovitis: A retrospective multicenter study of 237 cases. PLoS One 2015;10:e0121451. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Bernthal NM, Spierenburg G, Healey JH, et al. : The diffuse-type tenosynovial giant cell tumor (dt-TGCT) patient journey: A prospective multicenter study. Orphanet J Rare Dis 2021;16:191. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Lei P, Sun R, Liu H, Zhu J, Wen T, Hu Y: Prognosis of advanced tenosynovial giant cell tumor of the knee diagnosed during total knee arthroplasty. J Arthroplasty 2017;32:1850-1855. [DOI] [PubMed] [Google Scholar]
- 9.Machado V, San-Julián M: Risk factors for early osteoarthritis in tenosynovial giant cell tumour. Rev Esp Cir Ortop Traumatol (Engl Ed) 2020;64:199-205. [DOI] [PubMed] [Google Scholar]
- 10.Burton TM, Ye X, Parker ED, Bancroft T, Healey J: Burden of illness associated with tenosynovial giant cell tumors. Clin Ther 2018;40:593-602.e1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Lin F, Ionescu-Ittu R, Pivneva I, et al. : The economic burden of tenosynovial giant cell tumors among employed workforce in the United States. J Occup Environ Med 2021;63:e197-e202. [DOI] [PubMed] [Google Scholar]
- 12.Zhao X, Shah D, Gandhi K, et al. : Clinical, humanistic, and economic burden of osteoarthritis among noninstitutionalized adults in the United States. Osteoarthritis Cartilage 2019;27:1618-1626. [DOI] [PubMed] [Google Scholar]
- 13.Bedenbaugh AV, Bonafede M, Marchlewicz EH, Lee V, Tambiah J: Real-world health care resource utilization and costs among US patients with knee osteoarthritis compared with controls. Clinicoecon Outcomes Res 2021;13:421-435. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Barbour KE, Helmick CG, Boring M, Brady TJ: Vital signs: Prevalence of doctor-diagnosed arthritis and arthritis-attributable activity limitation—United States, 2013-2015. MMWR Morb Mortal Wkly Rep 2017;66:246-253. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Hootman JM, Helmick CG, Barbour KE, Theis KA, Boring MA: Updated projected prevalence of self-reported doctor-diagnosed arthritis and arthritis-attributable activity limitation among US adults, 2015-2040. Arthritis Rheumatol 2016;68:1582-1587. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.van der Heijden L, Gibbons CMH, Hassan AB, et al. : A multidisciplinary approach to giant cell tumors of tendon sheath and synovium–a critical appraisal of literature and treatment proposal. J Surg Oncol 2013;107:433-445. [DOI] [PubMed] [Google Scholar]
- 17.Mastboom MJL, Palmerini E, Verspoor FGM, et al. : Surgical outcomes of patients with diffuse-type tenosynovial giant-cell tumours: An international, retrospective, cohort study. Lancet Oncol 2019;20:877-886. [DOI] [PubMed] [Google Scholar]
- 18.Verspoor FGM, Zee AAG, Hannink G, van der Geest ICM, Veth RPH, Schreuder HWB: Long-term follow-up results of primary and recurrent pigmented villonodular synovitis. Rheumatology (Oxford) 2014;53:2063-2070. [DOI] [PubMed] [Google Scholar]
- 19.Chandra AA, Agarwal S, Donahue A, Handorf E, Abraham JA: Arthroscopic versus open management of diffuse-type tenosynovial giant cell tumor of the knee: A meta-analysis of retrospective cohort studies. J Am Acad Orthop Surg Glob Res Rev 2021;4:e21.00217. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Vastel L, Lambert P, De Pinieux G, Charrois O, Kerboull M, Courpied JP: Surgical treatment of pigmented villonodular synovitis of the hip. J Bone Joint Surg 2005;87:1019-1024. [DOI] [PubMed] [Google Scholar]
- 21.Lopez-Bastida J, Aranda-Reneo I, Rodríguez-Sánchez B, et al. : Economic burden and health-related quality of life in tenosynovial giant-cell tumour patients in Europe: An observational disease registry. Orphanet J Rare Dis 2021;16:294. [DOI] [PMC free article] [PubMed] [Google Scholar]