Young patients with hematologic and lymphatic malignancies have an increased risk of hip and knee arthroplasty

Abstract

Background Skeletal complications such as osteonecrosis (ON) are potential adverse events in patients treated for cancer, especially in those treated for hematologic and lymphatic malignancies (HLMs). ON may damage the hip or knee joints and may lead to arthrosis requiring total joint arthroplasty (TJA). The aim of this study was to address the risk of TJA in patients with cancer, especially those treated for HLM, in a nationwide population-based setting.

Material and methods All patients who had undergone TJA after cancer diagnosis between the years 2000 and 2012 were identified by linking the Arthroplasty Register and the Cancer Registry. Standardized incidence ratios (SIRs) of TJAs were calculated to assess whether patients with any cancer, but especially HLM, have increased risk for TJA when compared with the general population.

Results In patients with HLM or other cancer, the overall SIRs were similar compared with the general population. However, in HLM patients under 50 years of age, the SIR was 7.6, and in patients under 35 years of age, it was 45.5. The corresponding SIRs in patients with other cancers were 3.6 and 6.6, respectively. The highest SIRs, including all age groups, were among patients with acute lymphoblastic leukemia (SIR = 4.5) and acute myeloid leukemia (SIR = 1.9).

Discussion HLMs imply an increased risk for TJA compared with the general population. The risk is especially high in patients younger than 50 years, regardless of the type of HLM. Young patients with HLM, as well as their healthcare providers, should be aware of the highly increased risk of skeletal complications requiring TJA.

The relationship between total joint arthroplasty (TJA) and the risk of developing cancer after TJA has been widely studied over the last decade, especially focusing on total hip replacements and metal-on-metal bearings [1–3]. However, the risk of cancer causing TJA has not been fully established. The cancer may originate in the bone or spread to the bone, the conditions of which might necessitate a subsequent joint replacement surgery [4,5]. In addition, the cancer or its therapy may indirectly affect the quality of the bone or cause skeletal side effects [6–8].

Hematologic and lymphatic malignancies (HLMs) imply a risk of skeletal complications, such as osteonecrosis (ON) (Figure 1A and B) [9–15]. Usually, ON is related to the treatments for the disease, especially the use of corticosteroids or hematopoietic stem cell transplantation (HSCT), rather than the disease itself [16]. ON lesions are most frequently located in the lower extremities and around the weight bearing joints, where they may cause the collapse of the articular surface and the destruction of the joint, leading to a subsequent TJA [17]. Severe ON increases the risk of TJA in young patients with HLM [18].

Figure 1. Radiograph of hip joints in 54-year-old female patient with acute lymphoblastic leukemia showing osteonecrosis of the left femoral head (arrow) (A). Cross-section of the femoral head during the total hip arthroplasty procedure showing osteonecrotic lesion and partial collapse of the femoral head (arrow) (B).

The purpose of this study was to address the impact of HLM on the risk of undergoing TJA in a nationwide population-based register study.

Materials and methods

Data sources

We identified all primary hip and knee TJAs, that had been performed on patients with cancer by linking the data from the Finnish Arthroplasty Register and the Finnish Cancer Registry. The connection between TJA and a cancer diagnosis was established by using patient-specific social security codes. The Arthroplasty Register search was limited to between January 2000 and December 2012 because the implants and surgical techniques have developed quickly, and potential changes to the indications for the TJA procedure have occurred [19]. TJAs performed before cancer diagnoses were excluded. The general population information for between the years of 2000 and 2012 was received from the Statistics Finland database.

The cancer patients with TJA were divided into two groups: HLM and other cancers. The classification of cancers was done using the morphological codes (ICD-O-3) from the International Classification of Diseases for Oncology, third edition.

HLMs were identified by using ICD-O-3 codes 9590–9985, and were categorized further into seven sub-groups: acute lymphoblastic leukemia (ALL), acute myeloid leukemia (AML), chronic myeloid leukemia (CML), chronic lymphoblastic leukemia (CLL), non-Hodgkin’s lymphoma, Hodgkin’s lymphoma, and myeloma. The other group included all of the remaining cancers identified in the cancer registry, except for osteosarcoma (9180–9115), chondrosarcoma (9220–9243), giant cell tumor (9250–9251), and Ewing sarcoma (9260). Exclusions were performed because the aim of the study was not to investigate the consequences of cancers originating from bone or cartilage. In addition, patients with non-melanoma skin cancers (8070–8110) were excluded because their treatment was not expected to influence the risk for TJA.

Based on the inclusion and exclusion criteria, we identified 18 981 TJAs performed after cancer diagnosis, of which 1261 (6.6%) were performed on patients with HLM (Figure 2). The mean age of all cancer patients at the time of undergoing TJA was 70 years (range 11–96), and 62.0% (8052) of TJAs were performed on women. Over the same period, 170 854 TJAs were performed on patients without cancer (Table I).

Figure 2. Flowchart of included patients with total joint arthroplasty performed between 2000 and 2012 after cancer diagnosis.

Table I. Patients with total joint arthroplasty performed between 2000 and 2012 after cancer diagnosis.

	Patients with cancer
Variables	Total		Hematologic and lymphatic malignancies		Other cancers		Patients with no cancer
No. of TJA	18 981		1261		17 720		170 854
No. of patients	15 034		1029		14 005		130 515
– Females	9301	(61.9%)	595	(57.8%)	8706	(62.2%)	82 815	(63.5%)
– Person years	2 323 512		193 973		2 129 539		68 752 631
– Mean age*	69.7	(SD 9.2)	68.3	(SD 10.6)	69.9	(SD 8.9)	66.7	(SD 10.8)
– Hip joint	8888	(46.8%)	724	(57.4%)	8164	(46.1%)	79 094	(46.3%)
– Knee joint	10 093	(53.2%)	537	(42.6%)	9556	(53.9%)	91 760	(53.7%)

SD, standard deviation; TJA, total joint arthroplasty.

*Age at the time of total joint arthoplasty.

Statistics

To assess whether patients with HLM undergo more TJAs compared with the general population, standardized incidence ratios (SIRs) and corresponding 95% confidence intervals (CIs) were calculated. SIRs were calculated as a ratio of the observed number of TJAs performed in patients with cancer to the expected number of TJAs performed in the general population. The observed numbers of TJAs were stratified by age, year, and cancer type. The expected numbers of TJAs were calculated by applying age- and year-specific national TJA incidence rates to the relevant categories of person years. SIRs were calculated separately for all different types of HLM and six different age groups: 0–19, 20–34, 35–49, 50–64, 65–79, and ≥80 years. Patients were no longer followed up with if they died or migrated out of the country.

Results

In the patients with HLM, the overall SIR was similar compared with the general population (SIR = 1.03) (Table II). The corresponding overall SIR for other cancers was 0.99 (Table III). In the HLM patients under 50 years of age, SIR was 7.6, and in patients under 35 years of age, SIR was 45.5. The corresponding SIRs in patients with other cancers in the same age groups were 3.6 and 6.6, respectively. In patients with HLM, 6.5% of all TJAs were performed in patients under 50 years in age. In patients without cancer, the corresponding proportion was 4.4%. The highest SIRs were among patients with ALL (SIR = 4.50) and AML (SIR = 1.86) (Table IV). In patients with other cancers or without any cancer, the proportion of total hip arthroplasties among the TJAs were 46.3% and 46.1%, respectively. In patients with HLM, the corresponding hip arthroplasty proportion was 57.4%.

Table II. Standardized incidence ratios for all hematologic and lymphatic malignancies by age group.

Age group	Person years	Observed # of TJA	Expected # of TJA	SIR (obs/exp)	95% CI
0–19	10 411	5	0.0	112	36.2–260
20–34	19 570	31	1.1	27.7	18.8–39.2
35–49	25 619	50	14.7	3.4	2.5–4.5
50–64	56 144	281	238	1.2	1.1–1.3
65–79	60 113	712	770	0.9	0.9–1.0
80–	22 114	182	197	0.9	0.8–1.1
Total	193 972	1261	1221	1.0	1.0–1.1

CI, confidence interval; SIR, standardized incidence ratio; TJA, total joint arthroplasty.

Table III. Standardized incidence ratios for cancers other than hematologic or lymphatic malignancies by age group.

Age group	Person years	Observed # of TJA	Expected # of TJA	SIR (obs/exp)	95% CI
0–19	12 514	0	0.1	0	n/a
20–34	38 986	9	2.1	4.2	1.9–8.1
35–49	159 479	146	90.4	1.6	1.4–1.9
50–64	614 839	2989	2586	1.2	1.1–1.2
65–79	889 760	11 218	11 430	1.0	0.96–1.00
80–	413 961	3358	3750	0.9	0.87–0.93
Total	2 129 539	17 720	17 858	1.0	0.98–1.01

CI, confidence interval; n/a, not applicable; SIR, standardized incidence ratio; TJA, total joint arthroplasty.

Table IV. Standardized incidence ratios for hematologic and lymphatic malignancies by malignancy type.

	ALL
Age group	Per. Years	Obs.#	Exp.#	SIR	95% (CI)
0–19	6537	4	0.0	142	38.7–364
20–34	5732	13	0.3	39.6	21.1–67.7
35–49	1984	7	1.0	7.3	2.9–14.9
50–64	695	10	2.9	3.4	1.6–6.3
65–79	336	5	4.3	1.16	0.4–2.7
80–	10	0	0.1	0	n/a
Total	14 994	39	8.7	4.5	3.2–6.2
	AML					CML					CLL
Age group	Per. Years	Obs.#	Exp.#	SIR	95% (CI)	Per. Years	Obs.#	Exp.#	SIR	95% (CI)	Per. Years	Obs.#	Exp.#	SIR	95% (CI)
0–19	793	0	0.0	0	n/a	101	0	0.0	0	n/a	3	0	0.0	0	n/a
20–34	951	6	0.1	110	40.4–240	390	3	0.0	134	27.7–392	53	0	0.0	0	n/a
35–49	1549	14	0.9	15.8	8.6–26.5	979	2	0.6	3.6	0.4–12.9	652	2	0.4	5.4	0.65–19.3
50–64	2254	21	9.6	2.2	1.4–3.4	1518	7	6.4	1.1	0.44–2.24	4634	17	19.6	0.9	0.5–1.4
65–79	1538	19	19.7	1.0	0.6–1.5	922	9	11.8	0.8	0.35–1.45	8662	93	111	0.8	0.7–1.0
80–	293	1	2.6	0.4	0.1–2.1	291	1	2.6	0.4	0.01–2.15	3695	31	32.9	0.9	0.6–1.3
Total	7378	61	32.8	1.9	1.4–2.4	4201	22	21.4	1.03	0.64–1.6	17 699	143	164	0.9	0.7–1.0
	Non-Hodgkin’s lymphoma					Hodgkin’s lymphoma					Myeloma
Age group	Per. Years	Obs.#	Exp.#	SIR	95% (CI)	Per. Years	Obs.#	Exp.#	SIR	95% (CI)	Per. Years	Obs.#	Exp.#	SIR	95% (CI)
0–19	1628	1	0.0	143	3.6–795	1047	0	0.0	0	n/a	1	0	0.0	0	n/a
20–34	4102	3	0.2	12.8	2.6–37.3	8057	5	0.5	10.8	3.5–25.3	43	1	0.0	406	10.3–2261
35–49	10 418	18	6.0	3.0	1.8–4.8	8805	5	5.0	1.0	0.3–2.3	799	2	0.5	4.37	0.5–15.8
50–64	31 040	138	132	1.1	0.9–1.2	9213	28	39.0	0.7	0.5–1.0	5075	45	21.5	2.09	1.5–2.8
65–79	35 301	449	452	1.0	0.9–1.1	4141	38	53.1	0.7	0.5–1.0	7383	79	94.6	0.84	0.7–1.0
80–	13 667	121	122	1.0	0.8–1.2	939	4	8.4	0.5	0.1–1.2	2635	18	23.5	0.77	0.5–1.2
Total	96 156	730	712	1.0	1.0–1.1	32 202	80	106	0.8	0.6–0.9	15 936	145	140	1.04	0.9–1.2

ALL, acute lymphoblastic leukemia; AML, acute myeloid leukemia; CI, confidence interval; CLL, chronic lymphoblastic leukemia; CML, chronic myeloid leukemia; Exp.#, number of expected total joint arthroplasties in general population; n/a, not applicable; Obs.#, number of observed total joint arthroplasties in study group; Per. Years, observer person years; SIR, standardized incidence ratio.

Discussion

In this population-based study, a significant increase in the risk of receiving TJA was found in young patients with HLM. Patients younger than 50 years had an eight-fold risk, and patients younger than 35 years had up to a 45-fold risk of undergoing TJA compared with the general population. The risk of TJA was highest in patients with ALL and AML, and lowest in patients with Hodgkin’s lymphoma.

A TJA is a rare procedure in the general population among young patients, and is most commonly attributed to rheumatoid arthritis or congenital hip disease [20]. In the current study, the incidence of TJA in patients younger 35 years of age was only 3/100 000 in the general population. Even though young patients with HLM had a significantly increased risk of having TJA, it is still a rare procedure. Over the study period, a total of 2471 HLMs were diagnosed in patients less than 35 years (incidence 1.8/100 000) and only 36 TJAs were performed on those patients.

The high risk of TJA among young patients is most likely related to the increased risk for ON, which is caused by corticosteroids and HSCT, and is typical in patients with ALL [18,21–24]. However, it may be possible that HLMs also increase the risk of undergoing TJA in elderly patients, but because of the large number of TJAs performed on elderly patients for primary osteoarthritis, the small increase in the number of TJAs cannot be seen in the SIRs [16]. The patients with HLM did undergo more hip than knee arthroplasties, which emphasizes the clinical significance of ON for the femoral head. Glucocorticoids play an essential part in the treatment of ALL, but their optimum doses have not been established. The efficacy of prednisone and dexamethasone is dose dependent and needs to be carefully assessed against the toxic effects, especially in patients receiving dexamethasone, which causes higher risk for skeletal side effects [25]. In patients with leukemia or lymphoma, the other medications than glucocorticoids should be considered in the treatment of chemotherapy-induced nausea and vomiting.

A previously published population-based study on the risk of undergoing a total hip arthroplasty after cancer stated that cancer survivors, mainly those with pelvic cancers and HLM, have a small increase in their risk of receiving this procedure [26]. In the current study, the overall SIR result was consistent with those results, which had reported an SIR of 1.30 in patients with lymphoma and an SIR of 1.17 in patients with leukemia. In this study, we focused in more detail on HLMs by separately observing different types of HLM and different age groups. We succeeded in demonstrating that the risk of undergoing TJA is highly dependent on the type of HLM and, especially, on the age of the patient.

The weaknesses of the current study are common in all register-based studies. The registers may have data errors and the register data do not include any information about the patient’s medical history, including symptoms, medication, radiation therapy, or HSCT. In addition, the Finnish Arthroplasty Register do not contain information on hemiarthroplasties, which may be an option in patients with short life expectancy. Also, the small total number of TJAs limited to perform separate analysis of hip and knee arthroplasties. However, our study has many advantages, including its nationwide coverage of patients with cancer, which allows a large sample size and sufficient follow-up, and is thus a good method to research rare situations, such as TJA after cancer diagnosis. The used registers, the Finnish Arthroplasty Register and the Finnish Cancer Registry, have demonstrated good accuracy and coverage [27,28].

HLMs lead to an increased risk of a TJA procedure, especially in younger patients under 50 years of age compared with the general population. The risk was highest in patients with ALL and AML. Young patients with HLM, as well as their healthcare providers, should be aware of the highly increased risk of skeletal complications requiring TJA.

Funding

This study was financially supported by grants from the Nona and Kullervo Väre Foundation, the Alma and K.A. Snellman Foundation, and the Finnish Medical Foundation. These foundations had no role in the study design, data collection, analysis, or writing of this work, or in the decision to submit the manuscript.