The implications of the Sequestosome 1 mutation P392L in patients with Paget's disease in a United States cohort

Abstract

Background

Paget's disease of bone (PDB) is associated with a germline mutation in Sequestosome1 /p62 (SQSTM1) found in ≤ 16% of sporadic cases worldwide, and in 19-46% of those studied with familial PDB. The P392L is the most prevalent mutation identified to date. This mutation by itself does not confer PDB or define the phenotype of PDB in a given person. Environmental determinants remain elusive, although increasing age of the individual, other gene polymorphisms in the context of SQSTM1 mutations, and measles virus have been implicated. Measles exposure has been unexamined in this context.

Objectives

The goal of this study is to compare the background history and phenotype of patients with PDB carrying the SQSTM1 P392L mutation to those patients without. Focusing on age, ancestry, P329L mutation, family history, measles exposure, distribution of PDB and age of onset, we examined outcomes at 10 years. We postulated that aging may play a role in defining phenotype, and that this may become more visible in a well characterized cohort.

Methods

This is an observational study focused on a cohort of patients with PDB drawn from the New England Registry in whom environmental and family history has been catalogued, linked to radiographic data. Of the 217 persons who were enrolled in the Registry, 42 (19%) responded to a letter inviting them to participate in testing for the presence of the measles antibody, and in genetic testing for the P392L mutation.

Results

The mean age of the cohort in 2001 was 70 years (range 55-79); 27 were men (64%). The measles antibody was found in all cases tested. Nine patients had the P392L mutation (21%), 2 with familial PDB. In these persons, early diagnosis of disease and spinal stenosis marked the male phenotype only. European ancestry was noted in the minority of those with P392L mutation. Most deaths recorded occurred in the 9^th decade of life or later.

Conclusions

Spinal stenosis emerges as a prominent phenotype in SQSTM1 P392L + men with aging. In these 42 patients with PDB from the New England Registry, most do not carry the SQSTM1 P392L mutation, and many do not have European ancestry. Exposure to measles was confirmed in the majority.

INTRODUCTION

Paget's disease of bone (PDB) is a focal disorder of bone remodeling that may affect one or more bones in an aging skeleton. It is often thought of as a British disease, characterized most definitively by Sir James Paget, identified in the bones buried in old graveyards in Great Britain, and found in other countries where European migration was integral to their populations.

The burden of Paget's disease reflects the abnormal overgrowth of bone, and its consequences to musculoskeletal health through nerve impingement (deafness, spinal stenosis), fracture, deformity and early osteoarthritis leading to joint replacement.¹ Because of this progressive physical impairment caused by PDB, it has often been described as a crippling disease. In Europe, it has been defined as one marked by co-morbidities and early mortality.² A single retrospective epidemiological study in the US was the first to suggest that patients with PDB, particularly women, may have a slightly longer life expectancy than the Caucasian population in general. ³

In the last 10 years, some of the genetic determinants of familial and sporadic cases of Paget's disease have been described, with mutations clustering around the ubiquitin-associated domain of SQSTM1.⁴ The most prevalent mutation P392L results from a T to C transition in position 1215 in SQSTM1.⁵ This mutation is carried on two rather conserved haplotypes, presumed European in origin, ⁶ but its role in onset and skeletal distribution of PDB in an individual remains elusive. Although measles virus has been posited as contributory in PDB in the last century, evidence of the paramyxovirus in pagetic tissue has not been confirmed in all laboratories. ^7,8

This study describes ancestry, measles exposure, the presence or absence of the SQSTM1 mutation (P392L) and the musculoskeletal correlates in a remarkably diverse population of people with PDB from the New England Registry for PDB, Boston, MA.

METHODS

Study Population

In 2001, the New England Registry for PDB (NE Registry) was founded in an effort to understand the demographics of this disease in the United States. Enrollment was voluntary. Recruitment depended on responses to information about the study mailed to members of the Paget Foundation (New York, New York); on referrals from physicians in New England; and on patients willingness to participate who were seen at the Massachusetts General Hospital (MGH). Medical record searches through the Research Patient Data Registry at Partners (Boston, MA) were used to identify patients as well, and letters requesting participation were sent to their physicians. Recruitment closed in early 2005 as numbers of interested patients dwindled. We were able to capture 254 persons with confirmed PDB who completed the study questionnaire; in 217 of these imaging was available documenting the skeletal distribution of disease. The Partners Institutional Review Board (Boston, MA) approved the study.

Analyses

In 2004, 42 patients enrolled in the NE Registry responded to a letter inviting them to participate in this study, which involved blood drawn for the genetic analysis (Sequenom) of the SQSTM1 P392L mutation, and the enzyme-linked immunosorbent assay (ELISA) for measles antibody. The primer for the SQSTM1 P392L mutation has been previously described. ⁹ The patient DNA was isolated and the sequences analyzed at Harvard Partners Center for Genetics and Genomics High Throughput Sequenom Genotyping Facility, Cambridge, MA. The samples were de-identified prior to genetic analysis. Measles antibody testing was performed by the MGH Clinical Laboratory Services (VIDAS Measles IgG assay, BIOMERIEUX SA, France). We compared the SQSTM1 P392L positive patients to the SQSTM1 P392L negative patients. Formal statistics were not pursued because of the small sample size. Living status was documented when that information was available.

RESULTS

Forty-two patients from the NE Registry agreed to have blood drawn for genetic analysis of the SQSTM1 P392L mutation, and for measles antibody testing; 27 were men (64%). The mean age of the cohort at the time of enrollment was 72 (range 30-87 years). This was comparable to the mean age in the NE Registry in general, 73.2 years, but reflected a slightly higher proportion of male participants. Most participants in this study were born in New England towns, with parents or grandparents who immigrated to the US during the early 20^th century.

Nine of the 42 patients (21%) tested positive for the SQSTM1 P392L mutation; 7 were men, 2 of whom (28%) had familial PDB. (Table I) The ancestry of the SQSTM1 P392L + group was striking in that 6 of the 9 patients (67%) were from eastern Mediterranean countries, including Greece, Albania, Turkey and Lebanon. Age at diagnosis <50 years of age (67%), polyostotic disease and the evolution of spinal stenosis (56%) appeared more commonly in the men with this mutation. (Image 1) The initial diagnosis of PDB tended to be on the basis of radiographic findings in the SQSTM1 P392L + cohort (55%), rather than on the basis of pain or elevated serum alkaline phosphatase.

Patient	Gender	DOB	Age (2001)	Heritage	Measles Antibody	Age at Diagnosis (years)	FH PDB*	Diagnosis made by	Pagetic bones	Sites	Living 2011	Musculoskeletal Morbidity
1	M	1936	65	Portugal	UNK	<40	Yes	P, X, B	14	R/L pelvis, C4, LS spine, T9, sacrum, R/L pelvis, L femur, L/R humerus, L/R scapula, L tibia	Yes	Spinal stenosis, fracture spine
2	M	1925	76	Lebanon	Yes	41-50	Yes	X	8	Skull, R/L pelvis, R femur, L 4th rib T spine, L 4 and L5, sacrum	Yes	Spinal stenosis
3	M	1946	55	Italy Scotland	Yes	<40	No	X	3	Sacrum, L 4 and R pelvis	Yes	R THR, gout
4	M	1928	73	Greece Turkey	Yes	41-50	No	B	5	R tibia, L/R pelvis, R/L femur	Yes	Spinal stenosis, fracture tibia, R THR
5	F	1928	73	Greece	Yes	61-70	No	P-X-B	2	Skull, L tibia	Yes	Hearing loss
6	M	1922	79	Albania	Yes	41-50	No	B	8	L/R pelvis, L1 and L3, T11 and T12, R/L femur, R scapula, skull	Yes	Spinal stenosis
7	F	1939	62	Portugal	Yes	61-70	No	X	1	L pelvis	Yes	Crouzon's in grandchild
8	M	1927	74	Greece	Yes	51-60	No	X	5	5 R/L pelvis, L femur and T spine, LS spine	Yes	Spinal stenosis, fracture L femur
9	M	1930	71	Turkey	Yes	41-50	No	X	1	L tibia	Yes	L TKR

Family history of PDB in 1^st degree relative; P=pain / X= x-ray / B= blood lest. Three samples were improperly delivered to the lab for measles antibody titer. Patients labeled unknown are presumed alive given last visit in medical record, but not confirmed by visit since 2010.

SQSTM1 P392L- Patient with back pain R hip and spine.

In the SQSTM1 P392L negative cohort, 9 (27%) reported a family history of PDB. (Table II) The majority (67%) were from countries in which PDB is readily diagnosed, including Quebec (Canada), Italy, Ireland, Scotland, the United States and France. Diagnosis after age 50 was almost equal to those diagnosed <50 years of age. There was a tendency for less skeletal involvement (2-3 bones average). End-stage degenerative disease of a pagetic joint and fracture, rather than spinal stenosis, characterized the SQSTM1 P392L - cohort. Low back pain was reported in 9 SQSTM1 P392L -persons (27%), a number comparable with that previously reported from the NE Registry.¹⁰ Low back pain attributed to sciatica, degenerative disease of the spine, scoliosis were more prevalent than spinal stenosis. (Image 2) Pain alone or in combination with abnormal elevations in serum alkaline phosphatase or radiographic findings led to the diagnosis in 40% of SQSTM1 P392L - cohort.

Patient	Gender	DOB	Age (2001)	Heritage	Measles Antibody	Age at Diagnosis (years)	FH PDB*	Diagnosis made by	Pagetic bones	Sites	Living (2011)	Musculoskeletal Morbidity
1	F	1918	73	Italy	Yes	71-80	Yes	P-X	4	LS spine, sacrum, L/R femurs	↓80	Spinal stenosis
2	F	1933	68	Italy	Yes	51-60	Yes	P-X-B	3	R/L pelvis, L3 and L5	Yes
3	M	1935	66	Quebec	Yes	41-50	No	X	2	Sacrum, R pelvis
4	F	1926	75	Turkey	Yes	<40	No	P-B	2	Skull, R pelvis	Yes	Endstage DJD R hip, hearing loss
5	M	1971	30	USA	Yes	<40	No	B	5	R scapula, R humerus, L3 and 5, sternum, skull	Yes	Gout
6	M	1918	83	Russia	Yes	51-60	No	X	2	Skull, R pelvis	↓87	DJD R hip
7	M	1950	51	Italy	Yes	41-50	No	X	1	L pelvis
8	F	1915	86	Russia Poland	Yes	61-70	No	X-B	2	Skull, L femur	↓95	Fracture L femur
9	M	1933	68	Albania	Yes	<40	No	B	1	Skull	unknown	Hearing loss
10	M	1932	69	Italy	Yes	41-50	No	B	11	R/L pelvis, sacrum, R/L femur, LS spine, skull, T spine, L tibia, R humerus, R ribs	Yes	Spinal stenosis
11	F	1947	54	Quebec	Yes	41-50	No	B	2	Skull, R tibia	Yes
12	F	1928	73	Italy	Yes	41-50	No	P	1	R tibia	unknown	R TKR
13	M	1930	71	Ireland	Yes	41-50	No	X	4	Sacrum, L5, L pelvis, R femur	↓76	Low back pain, gout giant cell arteritis
14	F	1961	40	USA	Unknown	41-50	No	X	1	L pelvis	Yes
15	M	1932	69	France	Yes	41-50	No	X	3	Sacrum, R femur, L tibia	Yes	Fracture L tibia
16	M	1920	81	Russia Lithuania	Unknown	51-60	No	P-B	7	Sacrum, R/L pelvis, scapula, R clavicle, R/L femurs	↓90	Fracture femur
17	M	1943	58	Africa	Yes	51-60	No	P	1	L2	Yes	Low back pain
18	F	1928	73	Armenian	Yes	61-70	No	P-X	4	Sacrum, L3, L femur, R pelvis	Yes	Low back pain
19	M	1925	76	USA	Yes	61-70	Yes	B	2	R pelvis, R femur	↓86	Endstage DJD R hip
20	F	1924	77	Russia Austria	Yes	41-50	Yes	P-X-B	1	L pelvis	Yes	THR (bilateral)
21	M	1948	53	Quebec, Poland, Czech	Yes	41-50	No	X-B	1	R pelvis	unknown	Fracture femur
22	F	1927	74	Nepal	Yes	71-80	No	P-X-B	3	R/L pelvis, L3 and L4	unknown	Fracture spine
23	M	1931	70	Ireland Scotland	Yes	61-70	Yes	X	2	R scapula, L femur	unknown
24	F	1942	59	Italy Ireland	Yes	51-60	Yes	P	2	C2, L pelvis	Yes	Low back pain/disc herniation with srugery, child with HME
25	M	1923	78	Italy	Yes	41-50	Yes	X-B	7	Skull, L pelvis, R femur, R humerus, C-spine, tibia, fibula	Yes	Hyperparathyroidism, low back pain, neck pain
26	M	1933	68	Quebec	Unknown	61-70	No	P-X	3	Sacrum, R pelvis, R tibia	unknown	Gout
27	M	1919	82	Sicily	Yes	51-60	Yes	X-B	2	Skull, L3	↓87	Low back pain, gout
28	M	1918	83	Russia	Yes	71-80	No	P	3	R tibia, L femur, L 4	unknown	Fracture L femur
29	F	1924	77	Italy	Yes	51-60	No	P-X-B	3	Skull, L pelvis and L foot	↓88	Fracture L hip
30	M	1939	62	Italy	Yes	41-50	No	X-B	1	L femur	Yes	Fracture L femur
31	M	1921	80	Sicily	Yes	61-70	No	X	1	R femur	↓90	Low back pain, gout
32	M	1914	87	Lithuania	Yes	61-70	Yes	B	3	L humerus, R pelvis, R femur	↓96	Fracture R femur, gout
33	F	1935	66	Ireland	Yes	51-60	No	X-B	2	R/L pelvis	unknown

Lumbar spine lesions are counted as one, although the distribution of involvement varied; > 3 vertebral bodies = LS Spine w/o further definition FH indicates a first degree relative with PDB

Diagnosis made by x=x-ray, p= pain and b= blood test (elevated serum alkaline phosphatase)

SQSTM1 P392L+ Patient with back pain MRI lateral spine.

All persons tested were positive for the presence of measles antibody. Some causes of death were known, and in these cases cardiovascular complications seemed to prevail, including stroke (age 88), ischemic heart disease, atrial fibrillation and congestive heart failure (ages 80, 87,88, 90, 90, 95, and 96). Cancer was reported in both SQSTM1 P392L + and SQSTM1 P392L - patients, 3 men with prostate cancer, 2 women with breast cancer and 1 woman with renal cancer. Metastatic disease was not recorded in those patients followed through the Partners HealthCare System during the period of observation.

All patients with PDB received multiple cycles of therapy with drugs such as calcitonin, etidronate, tiludronate, pamidronate, alendronate, risedronate and zoledronate throughout the course of their lives.

DISCUSSION

In this study, we describe the diverse ancestry of SQSTM1 P392L + and SQSTM1 P392L - persons, the presence of these mutations in both sporadic and familial forms of PDB in the US, and the occurrence of spinal stenosis as a prevalent phenotype in men carrying the P392L mutation. This study highlights the finding that most people with PDB live well into their 80's and some into their 90's. This longevity is associated with significant physical impairments attributable to PDB that accrue over time, as shown in the accompanying tables. Fracture is a common complication of PDB. There were cases of breast, renal and prostate cancer reported in several patients both SQSTM1 P392L +/− but no skeletal complications of malignancy were noted during years of follow up. ^11,12 No one suffered from a pagetic osteosarcoma.

PDB is prevalent in Europe, Canada, the US, Australia and New Zealand. Within these countries, geographic and familial clusters have been reported, particularly in England. ¹³ The SQSTM1 mutation has been reported in 19% – 46% of familial PDB-cohorts studied in these countries, and in 2.6% - 16% of sporadic cases. ^14,15 It is considered rare on the Asian and African continents. In one US study of families with PDB, the SQSTM1 mutation was identified in 20.5% of those with familial PDB, and in 0 in those with sporadic PDB. ¹⁶ In a subsequent study of somatic mutations as causal in PDB, 3 patients with sporadic disease were identified as having a germline SQSTM1 P392L mutation.⁹ Boston, a city of immigrants from the European continent, seemed an ideal site to establish a registry to examine the determinants of PDB, and to search for the prevalence of the most common SQSTM1 P392L mutation, in a cohort of these patients. The NE Registry gave us a unique opportunity to examine the outcomes of these patients.

We found deformity was a consistent finding in PDB, usually present radiographically as a consequence of bone remodeling, and invariably present if there was a fracture through pagetic bone. Twelve fractures occurred in these 42 patients (29%), either as a presenting feature of PDB or as a late complication. All of the patients had received bisphosphonate therapy over the years. As demonstrated in the PRISM trial, ¹⁷ treatment may not alter outcome in these fractures occurring in elderly patients with established disease and extant deformity of bone.

There have been different efforts to link the phenotype of patients with PDB with the genetic mutations in SQSTM1, the most convincing being that truncating mutations affecting SQSTM1 may result in more severe disease. ¹⁵ The SQSTM1 P392L mutation alone has not been associated with a distinct phenotype. ¹⁸ A recent paper studying an international cohort of persons with PDB, tried to assess skeletal extent and severity of PBD as a consequence of certain risk alleles identified in a genome-wide association study. ¹⁹ Severity was derived from a composite score of deformity, fracture, osteosarcoma, orthopaedic procedures and hearing loss that were variably documented in this population; along with family history, prior bisphosphonate treatment, and age of onset. Despite discrepancies in data collection and the rather low numbers of bones involved in all patients recruited, there was a striking observation that in the presence of a SQSTM1 mutation, these alleles conveyed a risk for PDB, particularly in men. Without the detailed clinical analysis, spinal stenosis as a distinct outcome was not recognized.

The radiographic evidence of spinal stenosis in the pagetic spine in SQSTM1 P392L + patients, the compression fractures and physical impairments attendant with aging are distinct findings. (Image 1) These were not as prevalent in the low back pain of SQSTM1 P392L – patients, in whom scoliosis imposed by a pagetic hip (Image 2) or disc herniation in a degenerative spine were more common. The American Academy of Orthopaedic Surgeons estimates spinal stenosis occurs in 8-11% of aging Americans, so it is hard to know if the findings are chance alone.

We did not look for other SQSTM1 mutations. Enrollment occurred during 2001, was initiated by a letter, and relied on patients volunteering their time. The report of the findings may reflect bias inherent in a small observational study. Documentation of the last years of their life may prove incomplete, and co-morbidities may have been missed. Some of the challenges in studying PDB are evident in the NE Registry, as the collection of data depends on screening measures to ascertain prevalence in a disease that is largely asymptomatic, and on voluntary enrollment. Fifty-five to 60% of the patients in this study were diagnosed with PDB incidentally, by radiographs or blood tests ordered for other indications. The absence of a national health care registry in the United States means that the prevalence of PDB across New England relies on population estimates from limited screening. ²⁰

The strengths of this study are in the family history, its detail and the correlation of genetic markers with measles virus exposure, which proved universal in those with and those without the SQSTM1 P392L mutation. Measles was prevalent at the turn of the 20^th century, and remains prevalent in other parts of the world where PDB is rare, e.g. India and other parts of Asia. Does the measles virus infection play a role in this disorder against a distinct genetic background? This is unknown, as are other environmental determinants. In this small study, other musculoskeletal mutations were identified in the off-spring of two families, Crouzon's syndrome and hereditary multiple exostoses. No patient in the NE Registry has reported a child affected with PDB, implying either a decrease in PDB or diminished expression in this next generation. The declining prevalence of PDB reported in most countries worldwide suggests fewer cases will be diagnosed in the future. ^21,22 It is a curious predicament for a disease that is quite ancient. ²³

This observational study reports on cases of PDB in the United States with and without the SQSTM1 P392L mutation, documents their remarkable ancestry, confirms the presence of measles exposure and describes what is known of the long-term outcomes in these persons. Debilitating spinal stenosis was a prevalent phenotype in males with the SQSTM1 P392L + mutation, as was early-onset, polyostotic disease. In all persons with PDB, fractures were prevalent. The negative consequences of the disease on musculoskeletal health are well-defined in this cohort, but the possibility that there are positive consequences are suggested by the long life span of these patients; the absence of early dementia that is seen in rare, inherited forms of PDB; and the absence of skeletal complications of malignancy captured in this New England Registry cohort. The role of the SQSTM1 P392L mutation in mediating PDB remains unclear, as the incidence of this disorder of bone seems diminishing in most countries and the majority of patients with PDB test negative for a SQSTM1 germline mutation.