Natural history of Ollier disease and Maffucci syndrome: Patient survey and review of clinical literature

Abstract

Ollier disease (OD) and Maffucci syndrome (MS) are characterized by multiple enchondromas. Patients with MS also have benign vascular overgrowths that become malignant in 8.5% of cases. OD is characterized by multiple enchondromas, typically unilateral in distribution with a predilection for the appendicular skeleton. MS is characterized by multiple enchondromas bilaterally distributed in most of the cases. Both disorders feature multiple swellings on the extremity, deformity around the joints, limitations in joint mobility, scoliosis, bone shortening, leg-length discrepancy, gait disturbances, pain, loss of function, and pathological fractures. About 50% of patients with OD or MS develop a malignancy, such as chondrosarcoma, glioma, and ovarian juvenile granulosa cell tumor. To better understand the natural history of OD and MS, we reviewed 287 papers describing patients with OD and MS. We also created a survey that was distributed directly to 162 patients through Facebook. Here, we compare the review of the cases described in the literature to the survey’s responses. The review of the literature showed that: the patients with OD are diagnosed at a younger age; the prevalence of chondrosarcomas among patients with OD or MS was ∼30%; in four patients, vascular anomalies were identified in internal organs only; and, the prevalence of cancer among patients with OD or MS was ∼50%. With these data, health care providers will better understand the natural history, severity, and prognosis of these diseases and the prevalence of malignancies in these patients. Here, we recommend new guidelines for the care of patients with OD and MS.

1 |. INTRODUCTION

Enchondromas are benign cartilaginous tumors that arise from the medulla of bones, typically in the metaphysis (Morris, Lee, & Gebhardt, 2001; Silve & Jüppner, 2006). They can arise as solitary lesions or as multiple lesions, such as in the context of Ollier disease (OD, OMIM 166000) or Maffucci syndrome (MS, OMIM 6145692). OD is characterized by multiple (3+) enchondromas, which usually begin to appear in the first decade of life, but can appear in early adolescence and adulthood (Schwarz, Hardes, & Schulte, 2002; Silve & Jüppner, 2006). These lesions affect, mostly, the appendicular skeleton and can be variable in terms of number, location, and size (Ranger & Szymczak, 2009; Sun, Swee, Shives, & Unni, 1985). MS is also characterized by multiple (3+) enchondromas with vascular anomalies (Prokopchuk et al., 2016). In MS, enchondromas have been described as bilateral in most of the cases. In approximately 13% of cases, MS is noticeable in the first year of age. Both disorders can cause multiple swellings on the extremities, deformity around the joints, limitations in joint mobility, scoliosis, bone shortening, leglength discrepancy, gait disturbances, pain, and loss of function (Kumar, Jain, Bharadwaj, & Arya, 2015). Pathological fractures, facial asymmetry, and cranial nerve palsies can also occur (Schaison et al., 1999). Surgery is the only treatment currently available for complications such as limb-length discrepancy, oversized enchondromas, or vascular overgrowths in patients with OD and MS.

OD and MS occur as isolated cases. No familial cases have been reported to date, and no germline variants have been associated with these diseases thus far. Somatic mutations in IDH1 (p.Arg132His, p.Arg132Cys, and p.Arg132Ser) and IDH2 (p.Arg172Ser) have been identified in the tumors (enchondromas, chondrosarcomas, and vascular anomalies) of !80% of the individuals with OD and MS but not in unaffected tissue such as blood, muscle or saliva (Amary et al., 2011; Pansuriya et al., 2011). Variants in PTHR1, which encodes parathyroid hormone receptor 1, have been found in a small subset (20%, 5/31) of enchondromas or leukocyte DNA from patients with OD (Couvineau et al., 2008).

In the literature, the rate of malignant transformation of enchondromas to chondrosarcomas is highly variable in patients with OD and MS. The malignancy rate in OD varies from 20 to 45.8% (Fletcher, Unni, & Mertens, 2002; Schaison et al., 1999; Schwartz et al., 1987) and in MS from 52 to 57.1% (Schaison et al., 1999; Verdegaal et al., 2011). The prognosis of the secondary low-grade chondrosarcomas in both disorders is good, since these tumors rarely metastasize; the 5-year survival is estimated to be 90% (Beltrami, Ristori, Scoccianti, Tamburini, & Capanna, 2016). Enchondromas located in long bones or axial skeleton, especially the pelvis, have an increased risk of becoming chondrosarcomas. Therefore, these tumors need more frequent screening for early detection of malignant transformation (Verdegaal et al., 2011). Other cancers have also been described in both disorders: gliomas, juvenile granulosa cell tumors, and acute myeloid leukemia in OD; and pancreatic and hepatic adenocarcinoma, mesenchymal ovarian tumors, glioma, astrocytoma, and various kinds of sarcomas in MS (Amary et al., 2011; Lewis & Ketcham, 1973; Pansuriya et al., 2011; Ranger & Szymczak, 2009; Schwartz et al., 1987; Sun et al., 1985). However, the prevalence of these and other cancers in individuals with OD and MS is not known.

Here, we performed a literature review of 287 publications and used aggregate data to describe the characteristics of the OD and MS patient population and to determine the location and age-of-onset of tumors and frequency of malignancies in OD and MS. Although the current literature suggests that individuals with OD and MS have a higher risk of developing malignancies other than chondrosarcoma than the general population, the cases described in the literature most likely represent the most severe and unusual presentations of these disorders. Therefore, we directly, surveyed 162 individuals diagnosed with OD or MS in an effort to more accurately describe the natural history of OD and MS. With our results, we aim to improve patient care by providing a more accurate picture of the course of these rare diseases, including malignancy prevalence estimates, and by suggesting appropriate screening measures.

2 |. MATERIALS AND METHODS

2.1 |. Data collection and literature review

We selected 460 publications using the following PubMed query on August 28, 2017: ollier[Title/Abstract] OR ollier’s[Title/Abstract] OR maffucci[Title/Abstract] OR maffucci’s[Title/Abstract] AND “English”[Language] (Supporting Information). We excluded 173 papers from the analysis because they either did not report patient clinical information or described surgical approaches without further clinical description of the patients. We collected data from 287 publications describing 160 individuals with OD and 199 individuals with MS. We reviewed patient demographics, clinical descriptions, radiological test descriptions, surgical and histological reports. From these case reports, we collected information on the number, location, and distribution of enchondromas and vascular anomalies and occurrence of malignancies.

2.2 |. Patient survey

This study was approved by Johns Hopkins University’s IRB. A questionnaire consisting of 90 questions was designed (by N.L.M.S.) to collect demographic data, clinical and family history, and genetic testing results from patients with OD and MS. The clinical history questions asked individuals with OD or MS to describe the number and location of enchondromas, vascular anomalies, and exostoses; the occurrence of chondrosarcomas, spindle cell hemangiomas, and other benign and malignant tumors; and other comorbidities (Supporting Information). We made the survey digitally accessible via private Facebook groups for individuals with OD or MS and their family members from May 2017 to May 2018. It was completed by 126 patients with OD and 36 with MS from 27 countries in Europe, North and South America, Asia, and Africa (Table S4). We ensured that all patients in our survey were not previously reported in the clinical literature; we excluded one patient with Maffucci Syndrome from our survey results due to prior publication. The data from this survey are available from the corresponding author upon reasonable request.

2.3 |. Statistical analysis

Descriptive analyses were performed separately on individuals with OD and individuals with MS, identifying the median age at diagnosis of enchondromas, vascular anomalies, and malignancies in both groups of patients. Median time from OD or MS diagnosis to the development of malignancy was calculated. In MS, the median time from diagnosis of enchondromas to appearance of vascular anomalies was also calculated. The median was used in this study because of the extreme outliers that were identified. Distribution of enchondromas was dichotomized as either unilateral or bilateral and was stratified by location (upper extremity, lower extremity, and trunk). An additional category of visceral organ involvement was added for vascular anomalies. Overall prevalence of malignancy in OD and MS was calculated to identify the most common cancer types.

Mann–Whitney U (two tailed) and χ² tests were used to compare continuous (age, time) and categorical (gender, distribution, location) variables, respectively. We performed univariate logistic-regression to compare the prevalence of overall malignancy in OD and MS, in addition to the prevalence of specific cancer types. Statistical significance was set at p < .05.

3 |. RESULTS

3.1 |. Literature review

3.1.1 |. Demographics

In total, 359 cases were reviewed (160—OD, 199—MS). Both sexes were approximately evenly represented (50.6% female—OD, 50.1% female—MS). The median age at the time of diagnosis of an enchondroma in OD cases in this review was 7 years (range: 0–58 years, mean 11.6 years). For individuals with MS, the median age at diagnosis of enchondromas was 11 years (range: 0–68 years, mean 15.6 years). Patients with OD were diagnosed with enchondromas significantly earlier than patients with MS (p = .029; Table 1).

TABLE 1.

Population characteristics of individuals with OD or MS from the literature review

Population characteristics	OD	MS	p Value
Number of patients	160	199
Male gender	77 (49.4%)	81 (40.9%)	.112
Female gender	79 (50.6%)	117 (50.1%)	.112
Median age (years) at the time of enchondroma diagnosis (range)	7(0–58)	11 (0–68)	.029
Distribution of enchondromas
Unilateral	78 (57.8%)	81 (46.3%)	.045
Bilateral	57 (42.2%)	94 (53.7%)	.045
Location of enchondromas
Upper extremity	102 (75.6%)	148 (84.6%)	.046
Lower extremity	96(71.1%)	120 (68.6%)	.630
Axial skeleton	31 (23.0%)	37 (21.1%)	.701
Median age (years) at the time of ascular anomaly diagnosis (range)		13(0–68)
Distribution of vascular anomalies
Unilateral		99 (57.9%)
Bilateral		72 (42.1%)
Location of vascular anomalies
Upper extremity		120 (70.2%)
Lower extremity		88 (51.5%)
Axial skeleton		58 (33.9%)
Visceral		27 (15.8%)
Median time (years) from enchondroma diagnosis to vascular anomaly diagnosis (range)		0 (22–35)

Note: Not all characteristics were reported in the literature: Gender was based on data from 156/160 patients with OD and 198/199 patients with MS; Median age at time of enchondroma diagnosis was based on data from 111/162 patients with OD and 149/199 patients with MS; Distribution and location of enchondromas were based on data from 135/160 patients with OD and 175/199 patients with MS; Median age at time of vascular anomaly diagnosis was based on data from 153/199 patients with MS; Distribution and location of vascular anomalies were based on data from 171/199 patients with MS; Median time from enchondroma diagnosis to vascular anomaly diagnosis was based on data from 146/199 patients with MS. Mann–Whitney U (two tailed) and χ² tests were used to compare continuous (age, time) and categorical (gender, distribution, location) variables, respectively. Statistical significance was set at p < .05 (values in bold).

Abbreviations: MS, Maffucci syndrome; OD, Ollier disease.

A majority of cases were diagnosed with an enchondroma before the age of 20 years (83.8%—OD, 71.8%—MS). Interestingly, approximately 8% of patients were diagnosed with an enchondroma before the age of 1 year in both disorders (8.1%—OD, 7.5%—MS). Twenty individuals with MS (13.1%) were diagnosed with vascular anomalies before the age of 1 year but, the median age at the time of the diagnosis of a vascular anomaly was 13 years (range: 0–68 years). Thirteen individuals with MS (8.7%) were diagnosed with vascular anomalies before enchondromas were found (Table 1).

3.1.2 |. Enchondromas

Seventy-eight individuals with OD (57.8%) and 81 individuals with MS (46.3%) had enchondromas on one side of the body only (p = .045). Although enchondromas affected both upper and lower extremities in more than 50% of the cases in both OD and MS, a significantly higher proportion of patients with MS had upper extremity involvement (75.68%—OD, 84.6%—MS; p = .046; Table 1).

In individuals with OD, the precise location of the enchondromas was noted in 135 of the 160 individuals described in these reports. Enchondromas were predominantly located in the femur (46.6%), tibia (30.4%), humerus (27.2%), pelvis (20.8%), fibula and radius (14.1% each), ulna (13.2%), and scapula (11.1%). The small tubular bones of the hands were more often affected with enchondromas (43.7%) than the small tubular bones of the feet (18.5%). In the axial skeleton, the skull was the most common site affected by enchondromas (11.9%) followed by the ribs (7.4%) and the spine (1.5%).

Among individuals with MS, the precise location of the enchondromas was noted in 175 of the 199 individuals described in these reports. Enchondromas were predominantly found in the femur (29.1%), tibia (25.7%), humerus and ulna (19.4% each), radius (15.4%), fibula (14.3%), and pelvis and scapula (13.7% each). The small tubular bones of the hands were more often affected with enchondromas (62.9%) than the small tubular bones of the feet (28.6%). In the axial skeleton, the ribs were the most common site affected by enchondromas (9.7%) followed by the skull (6.3%) and the spine (1.1%).

3.1.3 |. Vascular anomalies

Among individuals with MS, the precise location of the vascular anomaly was noted in 175 of the 199 individuals described in these reports. Among individuals with MS, 57.9% of vascular anomalies were unilateral, 70.2% were found in the upper extremity, and 51.5% of them were found in the lower extremity. Twenty-seven individuals (15.8%) had visceral vascular anomalies, of which 4 (2%) had visceral involvement only; these four individuals were diagnosed by magnetic resonance imaging (MRI) or on autopsy (Table 1).

3.1.4 |. Cancer

The total prevalence of malignancies was not statistically different in OD compared to MS (53.1%—OD, 48.7%—MS; p = .409). However, individuals with MS were significantly older than individuals with OD at cancer diagnosis (p = .033). Individuals with OD were diagnosed with cancer at a median age of 25 years (range, <1–69), while individuals with MS were diagnosed with cancer at a median age of 30 years (range, <1–77; Table 2).

TABLE 2.

Malignancy characteristics of individuals with OD or MS from the literature review

Malignancy characteristics	OD	MS	χ2	OR (95%CI)
Patients with malignancies	85 (53.1%)	97 (48.7%)	0.409	1.192 (0.786–1.808)
Median age (years) at the diagnosis of a malignancy (range)	25 (0–69)	30 (0– 77)	<0.033
Median time (years) from enchondroma to malignancy diagnosis (range)	11 (4–62)	15 (24–58)	0.464
Malignancy type
Bone	49 (30.6%)	63 (31.7%)	0.834	0.953(0.608–1.494)
Brain	26 (16.3%)	9 (4.5%)	<0.001	4.096 (1.860–9.022)
Gonad	13 (8.1%)	7 (3.5%)	0.059	2.426 (0.944–6.232)
Others	5 (3.1%)	11 (5.6%)	0.274	0.551 (0.188–1.621)
Vascular	1 (0.06%)	17 (8.5%)	0.001	0.067 (0.009–0.512)
Hematological	1 (1.06%)	5 (2.5%)	0.166	0.244 (0.28–2.110)

Note: Not all characteristics were reported in the literature: Median age at time of malignancy diagnosis was based on data from 84/85 OD patients with malignancy, and 87/97 MS patients with malignancy; Median time from enchondroma to malignancy diagnosis was based on data from 55/85 OD patients with malignancy, and 73/97 MS patients with malignancy; Malignancy type information was based on data from 85/85 OD patients with malignancy, and 97/97 MS patients with malignancy. Mann–Whitney U (two tailed) and χ² tests were used to compare continuous (age, time) and categorical (gender, distribution, location) variables, respectively. We performed univariate logistic-regression to compare the prevalence of overall malignancy in OD and MS, in addition to the prevalence of specific cancer types. Statistical significance was set at p < .05 (values in bold).

Abbreviations: MS, Maffucci syndrome; OD, Ollier disease.

The most common malignancy in both groups was chondrosarcoma, affecting about a third of individuals (30.6%—OD, 31.7%—MS; Table 2). Among cases with chondrosarcoma, a significantly higher percent of OD cases had lower extremity involvement (55.1%—OD, 35.6%—MS; p = .042; Table 3). A higher, but not statistically different proportion of MS cases had axial skeletal involvement (30.1%—OD, 47.5%—MS; p = .075; Table 3, Table S2).

TABLE 3.

Location of chondrosarcomas in the literature review

Chondrosarcoma location	OD	MS	χ2
Lower extremity	27 (55.1%)	21 (35.6%)	0.042
Upper extremity	12 (24.5%)	20 (33.9%)	0.286
Axial	15 (30.1%)	28 (47.5%)	0.075
Visceral	3 (6.1%)	3 (5.1%)	0.815

Note: Distribution and location of chondrosarcoma was based on data from 49/49 OD patients with chondrosarcoma and 59/63 MS patients with chondrosarcoma. Mann–Whitney U (two tailed) and χ² tests were used to compare continuous (age, time) and categorical (gender, distribution, location) variables, respectively. We performed univariate logistic-regression to compare the prevalence of overall malignancy in OD and MS, in addition to the prevalence of specific cancer types. Statistical significance was set at p < .05 (values in bold).

Abbreviations: MS, Maffucci syndrome; OD, Ollier disease.

A significantly higher proportion of patients with OD developed a brain malignancy compared to patients with MS (16.3%—OD, 4.5%—MS; p < .001). The odds for patients with OD to develop a brain malignancy were 4.1 times more than patients with MS (95% CI: 1.86–9.0, p < .001). A small percentage of cases in each disorder developed a cancer that affected their gonads (8%—OD, 3.5%—MS; p = .059). A significantly higher proportion of patients with MS were diagnosed with vascular malignancies compared to patients with OD (0.06%—OD, 8.5%—MS; p = .001). Patients with OD were significantly less likely to develop a vascular malignancy when compared to patients with MS (OR: 0.07, 95% CI: 0.01–0.51). There was no significant difference between individuals with OD and MS in the development of hematological and other malignancies (Table 2). The different malignancies described in patients with OD or MS are listed on Table S1.

3.2 |. Patient survey

3.2.1 |. Demographics

In total, 162 individuals answered the survey (126—OD, 36—MS). About half of respondents were affected with either OD or MS, while the other half were the parent or guardian of an affected individual (54.3%—affected, 45.7%—relative). The majority of respondents were female (63.5%—OD, 77.8%—MS). At the time they answered the survey, participants with OD were significantly younger than participants with MS (median: 17.6 years, range: 2–62 years—OD; median: 39.7 years, range: 2–60 years—MS; p < .001; Table 1).

The median age at diagnosis of enchondromas was younger, but not statistically different in individuals with OD compared to individuals with MS (median: 4.8 years, range: 0.2–61.4 years—OD; median: 6.0 years, range: 0.67–54 years—MS; p = .139). In both disorders, most individuals were diagnosed before the age of 10 years (86%—OD, 71%—MS). A small number of individuals were diagnosed before the age of 1 year (3.3%—OD, 10.7%—MS). In individuals with MS, the median age at the time of the diagnosis of a vascular anomaly was 9 years (range: 0–37 years). Four individuals (14.2%) were diagnosed with vascular anomalies before the age of 1 year. Vascular anomalies in seven patients (26.9%) were identified before any enchondromas were diagnosed (Table 4).

TABLE 4.

Population characteristics of individuals with OD or MS from the survey

Population characteristics	OD	MS	P Value
Number of patients	126	36
Median age (years) at survey (range)	17.6 (2–62)	39.7 (2–60)	˂.001
Male gender	46 (36.5%)	8 (22.2%)	.109
Female gender	78 (63.5%)	26 (77.8%)
Median age (years) at enchondroma diagnosis (range)	4.8 (0.17–61.42)	6.0 (0.67–54)	.623
Distribution of enchondromas
Unilateral	54 (42.8%)	10 (27.7%)	.103
Bilateral	67 (53.1%)	25 (69.4%)
Location of enchondromas
Upper extremity	95 (75.4%)	31 (86.1%)	.173
Lower extremity	94 (74.6%)	23 (63.9%)	.206
Axial skeleton	30 (23.8%)	11 (30.6%)	.412
Median age (years) at vascular anomaly diagnosis (range)		9 (0–37)
Distribution of vascular anomalies
Unilateral		17 (47.2%)
Bilateral		13 (36.1%)
Location of vascular anomalies
Upper extremity		23 (63.8%)
Lower extremity		19 (52.7%)
Axial		5 (13.8%)
Visceral		2 (5.5%)
Time (years) from enchondroma— vascular anomaly (range)		4 (-17–28)

Note: Not all respondents answered all questions: Median age at time of the survey was based on data from 123/126 patients with OD and 36/36 patients with MS; Median age at time of enchondroma diagnosis was based on data from 123/126 patients with OD and 28/36 patients with MS; Distribution and location of enchondromas were based on data from 121/126 patients with OD and 35/36 patients with MS; Median age at time of vascular anomaly diagnosis was based on data from 26/36 patients with MS; Distribution and location of vascular anomalies were based on data from 30/36 patients with MS; Median time from enchondroma diagnosis to vascular anomaly diagnosis was based on data from 19/36 patients with MS. Mann–Whitney U (two tailed) and χ² tests were used to compare continuous (age, time) and categorical (gender, distribution, location) variables, respectively. Statistical significance was set at p < .05 (values in bold).

Abbreviations: MS, Maffucci syndrome; OD, Ollier disease.

3.2.2 |. Enchondromas

Fifty-four individuals with OD (42.8%) and 10 individuals with MS (27.7%) had enchondromas on only one side of the body in our survey. In both disorders, the enchondromas affected upper and lower extremities in more than 50% of the cases (Table 4). Although only three enchondromas are required to be diagnosed with OD or MS, many individuals reported many more enchondromas. Eighty individuals with OD had 5 or more enchondromas (63.4%), and 51 had 10 or more enchondromas (40.4%). Thirteen individuals with MS had 5 or more enchondromas (36.1%), and 15 had more than 10 enchondromas (41.7%; Table 4).

Among individuals with OD, enchondromas were predominantly found in the femur (46%), tibia (44%), humerus (40%), pelvis (33%), radius (31%), ulna (28%), and fibula (25%). The small tubular bones of the hands were more often affected with enchondromas (76%) than the small tubular bones of the feet (46%). In the axial skeleton, ribs were the most common site affected by enchondromas (18%) followed by the spine (6%) and skull (5%).

Among individuals with MS, enchondromas were predominantly found in the humerus (57%), pelvis and tibia (51% each), femur (49%), ulna (34%), and radius and fibula (40% each). The small tubular bones of the hands were more often affected with enchondromas (77%) than the small tubular bones of the feet (57%). In the axial skeleton, ribs were the most common site affected by enchondromas (37%) followed by the spine (9%). No individual with MS reported skull involvement.

3.2.3 |. Vascular malformations

Among individuals with MS, about half had unilateral vascular anomalies (47.2%). Sixty-three percent (63.8%) had vascular anomalies in the upper extremity, while 52.7% had them in the lower extremity. Two individuals (5.5%) had visceral involvement (Table 4).

3.2.4 |. Cancer

The prevalence of malignancies among survey participants was about 30% (25.4%—OD, 36.1%—MS; p = .206). Individuals with MS were diagnosed with cancer at a younger age than individuals with OD (median: 28.5, range: 6 to 61—OD; median: 22, range: 1.5 to 55—MS; p = .139), although the difference was not statistically significant.

The most common malignancy in both groups was chondrosarcoma (22.2%—OD, 30.5%—MS; p = .206; Table 5). Among individuals with chondrosarcoma, a significantly higher proportion of individuals with OD had lower extremity involvement (48.1%—OD, 9.1%—MS; p = .042). A significantly higher proportion of individuals with MS had upper extremity involvement (37.0%—OD, 72.7%—MS; p = .046), and a higher percentage of individuals with MS had axial involvement (29.6%—OD, 45.5%—MS; p = .351; Table 6; Table S3).

TABLE 5.

Malignancy characteristics of individuals with OD or MS from the survey

Malignancy characteristics	OD	MS	Chi2	OR (95%CI)
Number of patients with malignancies	32 (25.4%)	13 (36.1%)	0.206	1.58 (0.273–1.327)
Median age (years) at the diagnosis of a malignancy (range)	28.5 (6–61)	22 (1.5–55)	0.139
Median time (years) from enchondroma to malignancy diagnosis (range)
Malignancy type
Bone	28 (22.2%)	11 (30.5%)	0.302 1.54	(0.675–3.511)
Brain	3 (2.4%)	0 (0%)	0.350	-
Gonads	0 (0%)	2 (5.6%)	0.008	-
Other	2 (1.6%)	0 (0)	0.447	-

Note: Not all respondents answered all questions: Median age at the time of malignancy diagnosis was based on data from 21/25 OD patients with malignancy, and 13/13 MS patients with malignancy; Median time from enchondroma to malignancy diagnosis was based on data from 30/32 OD patients with malignancy, and 8/13 MS patients with malignancy; Malignancy type information was based on data from 32/32 OD patients with malignancy, and 13/13 MS patients with malignancy. Mann–Whitney U (two tailed) and χ² tests were used to compare continuous (age, time) and categorical (gender, distribution, location) variables, respectively. We performed univariate logistic-regression to compare the prevalence of overall malignancy in OD and MS, in addition to the prevalence of specific cancer types. Statistical significance was set at p < .05 (values in bold).

Abbreviations: MS, Maffucci syndrome; OD, Ollier disease.

TABLE 6.

Survey—Chondrosarcoma location

Chondrosarcoma location	OD	MS	Chi2
Lower extremity	13 (48.1%)	1 (9.1%)	0.024
Upper extremity	10 (37.0%)	8 (72.7%)	0.046
Axial	8 (29.6%)	5 (45.5%)	0.351

Note: Location of chondrosarcomas from the survey. Not all respondents answered all questions: Distribution and location of chondrosarcoma was based on data from 27/28 OD patients with chondrosarcoma and 11/11 MS patients with chondrosarcoma. Mann–Whitney U (two tailed) and χ² tests were used to compare continuous (age, time) and categorical (gender, distribution, location) variables, respectively. We performed univariate logistic-regression to compare the prevalence of overall malignancy in OD and MS, in addition to the prevalence of specific cancer types. Statistical significance was set at p < .05 (values in bold).

Abbreviations: MS, Maffucci syndrome; OD, Ollier disease.

Interestingly, very few individuals reported brain malignancies (2.4%—OD, 0%—MS, p = .350). Only a small percentage of individuals with MS reported cancer of the gonads (0%—OD, 5.6%—MS; p = .008) No individuals with MS reported vascular malignancies, and two individuals with OD (1.6%) developed other malignancies (Table 5).

3.2.5 |. Molecular test results

Of the 162 individuals that answered the survey, only 17 patients had their enchondroma or chondrosarcoma tested for variants in IDH1 (± IDH2). Four of them reported a negative result for the IDH1 sequencing, and seven of them reported a positive result but did not know what the variant was. Three patients reported a negative result for the IDH2 sequencing and only one reported a positive result but again without the variant information. In total, of the 17 patients only 7 patients were positive for variants in IDH1 or IDH2.

4 |. DISCUSSION

OD and MS are rare diseases with a prevalence of less than 1 in 100,000. As of August 28, 2017, with our search strategy we identified only 160 individuals with OD and 199 individuals with MS clinically described in the literature. Published articles about these disorders often do not include a detailed clinical description of the patients. To define the full spectrum of these diseases, including the frequency and type of associated malignancies, we took two approaches. First, we performed a review of 287 existing publications that clinically describe individuals with OD or MS. However, we hypothesized that these published cases represent the most severe and unusual presentations of OD and MS. Therefore, we also directly surveyed 126 individuals with OD and 36 individuals with MS about their demographic information, clinical and family history, and molecular test results.

Although it has been previously described that OD is diagnosed twice as often in males than in females (Schwarz et al., 2002), we found in the literature review that the number of males and females described was similar. In our survey, a preponderance of affected individuals (63.5%) were females; however, since disease presentation is similar between the sexes, this difference should not bias our results.

A retrospective cohort study by Verdegaal et al., 2011 found that patients with OD were diagnosed on average 1 year later than patients with MS (12 vs. 11 years of age). Our literature review results showed that patients with OD were diagnosed with enchondromas a median of 4 years earlier than patients with MS (7 vs. 11 years of age [median]; 11 vs. 15 years of age [mean]). Patients with MS were diagnosed with vascular anomalies at a median of 13 years (Table 1). We observed a similar trend in our survey, where patients with OD were diagnosed with enchondromas a median of 1.2 years earlier than patients with MS (4.8 vs. 6 years), while patients with MS were diagnosed with vascular anomalies at a median of 9 years of age (Table 4). Comparing the median age of onset of enchondromas and vascular anomalies in patients with MS, our literature review found that vascular anomalies were diagnosed a median of 2 years after the enchondromas. A similar trend was observed in our survey, where the difference was 3 years. This delay in vascular anomaly development in patients with MS may lead to patients being initially, incorrectly, diagnosed with OD at the onset of multiple enchondromas. Interestingly, 13 patients (8.7%) in our literature review and 7 patients (26.9%) in our survey were diagnosed with vascular malformations prior to enchondroma diagnosis. This may be due to the visible appearance of vascular anomalies, compared to enchondromas, which may go unnoticed until the onset of other symptoms and raises the need for an imaging test such as x-ray for investigation of enchondromas when a vascular anomaly is diagnosed. Although it has been previously reported that 25% of the patients with MS are diagnosed at birth (Verdegaal et al., 2011), we found that fewer than 15% of patients with MS were diagnosed with vascular anomalies before the age of 1 year (13.1%—literature review, 14.2%—survey).

There is high variability in the clinical presentation of enchondromas in OD and MS. The current literature suggests that in OD, enchondromas are most commonly observed bilaterally, typically asymmetrically with predilection for the appendicular skeleton (Ranger & Szymczak, 2009; Sun et al., 1985). Axial involvement is rare, typically of the ribs, sternum, and skull in severe cases (Murakami et al., 1993). Similarly, in MS, enchondromas are described as bilateral and asymmetric. Contrary to the existing literature, our literature review found that enchondromas in OD most commonly occurred unilaterally (57.8%). In MS, not OD, they commonly occur bilaterally (53.7%). A similar trend was observed by Verdegaal et al. (2011), who found that 59% of patients with OD (85 of 144 patients) had unilateral enchondromas, while 62.5% of patients with MS (11 of 17 patients) had bilateral involvement (Prokopchuk et al., 2016).

We also analyzed the distribution of enchondromas in patients with OD and MS. Our literature review found that in patients with OD, enchondromas predominantly affected the appendicular skeleton, with a similar distribution in the upper and lower extremity. This trend was also observed in cases with MS; however, the upper extremity was more commonly affected in individuals with MS when compared to individuals with OD. Our survey showed that, in both disorders, the appendicular skeleton was more commonly affected than the axial skeleton. Our survey also showed that about a third of individuals with MS, a much higher percentage than previously reported, had axial skeleton involvement. The most common locations of enchondromas in our literature review and survey in both OD and MS were the femur, tibia, and humerus. Furthermore, both also showed that the smaller tubular bones of the hands were more commonly affected than the small tubular bones of the feet. The reason(s) why the largest bones are more commonly affected is not known. One possible reason is simple: the large bones represent the largest population of cells that could change somatically. Small tubular bones of the hands are more often be affected than similar bones of the feet; the reason for this difference is also unknown. The full understanding of the molecular bases of these disorders may inform us if differential timing of the development of these tissues is the cause for such distribution of enchondromas in these patients.

According to the National Organization for Rare Disorders, vascular anomalies in patients with MS usually appear in children around 4–5 years of age, most commonly in the hands (Couvineau et al., 2008). Contrary to the distribution of enchondromas, vascular anomalies were mostly unilateral in both our literature review and survey. Vascular anomalies were most commonly found in the upper extremity, followed by lower extremity and axial skeleton. While vascular anomalies in MS are known to affect the meninges, tongue and oral mucosa, few case reports in the literature describe visceral organ involvement (Beltrami et al., 2016; Burgetova et al., 2017). In our literature review, we found that 15.8% of vascular anomalies were in visceral organs, and, surprisingly, the visceral vascular anomalies were found despite the absence of any visible cutaneous lesions in four patients (2.3%). In our survey, 5.5% of vascular anomalies were in the visceral organs. The lack of visually identifiable vascular anomalies in some patients highlights the importance of imaging screening of vascular anomalies in patients presenting with multiple enchondromatosis. Full body imaging, including visceral examination, would further prevent misdiagnosis of MS cases with OD.

The overall risk of malignancy has been described as higher in MS than in OD (Prokopchuk et al., 2016; Silve & Jüppner, 2006). Our literature review and survey found no significant difference in the overall prevalence of malignancy between individuals with OD and individuals with MS. However, in our literature review, individuals with MS were significantly older at malignancy diagnosis when compared to individuals with OD. Malignant transformation of enchondromas to chondrosarcoma has been estimated between 20–45.8% in OD and 52–57.1% in MS (Liu, Hudkins, Swee, & Unni, 1987; Prokopchuk et al., 2016; Rozeman, Hogendoorn, & Bovée, 2002; Schaison et al., 1999). These variable occurrences in previous studies may be due to their small sample size. The largest case series by Verdegaal et al. (2011) reported the prevalence of chondrosarcoma among 144 patients with OD and 17 patients with MS. They found that 40% of patients with OD and 53% of patients with MS developed chondrosarcoma at a median of 33 and 30 years of age, respectively. These estimated prevalences are higher than we found in both our literature review (30.6%—OD, 31.7%—MS) and in our survey (22.2%—OD, 30.5%—MS). The current literature may over-estimate the prevalence of chondrosarcoma due to selection bias for extreme cases. Alternatively, the lower prevalence in our survey, specifically among patients with OD, may be due to the young age of patients at the time of the survey. While individuals with MS were a median age of 39.7 years at the time they answered our survey, individuals with OD were only a median age of 17.6 years of age at the time they answered. Several studies, including a retrospective cohort study by Liu et al. (1987), estimated that the mean age at malignant transformation in patients with OD was 40.5 years old. As such, our survey may underestimate the true lifetime risk of chondrosarcoma occurrence, warranting further investigation of our cohort over time. More importantly, both our literature review and survey showed a lower prevalence of chondrosarcomas in both disorders than what has been suggested by the current literature. While some studies estimate a higher prevalence of chondrosarcoma among patients with MS, we found no significant difference in chondrosarcoma occurrence in individuals with OD or MS (Prokopchuk et al., 2016).

Chondrosarcomas are most commonly located in the long bones in OD and MS. Liu et al. (1987) reported that OD chondrosarcomas are most often found in the femur, followed by the tibia, humerus, and ulna. Verdegaal et al. (2011) reported 89 patients with OD or MS diagnosed with chondrosarcoma; in that report, the most common locations of chondrosarcomas were the femur (20%), phalanges of the feet (15.8%), pelvis (12.4%), humerus and tibia (11% each), and phalanges of the hand (10%). In our study, we also investigated the locations of chondrosarcoma in patients with OD and MS separately to determine if chondrosarcoma location differed according to diagnosis. We found that in both our literature review and survey, a significantly higher proportion of patients with OD had chondrosarcomas of their lower extremity (literature review: 55.1%—OD, 35.6%—MS, p = .042; survey: 48.1%—OD, 9.1%—MS, p = .024). Our survey showed that a significantly higher proportion of patients with MS had chondrosarcomas in their upper extremities, but this may be due to the small number of individuals in this group (literature review: 24.5%—OD, 33.9%—MS, p = .286; survey: 37.0%—OD, 72.7%—MS, p = .046). The most common locations of chondrosarcomas in our literature review in both OD and MS were the skull, femur, tibia, and humerus. Alternatively, in our survey, the most common locations were the femur, humerus, pelvis, and hands. This widespread distribution of chondrosarcoma locations highlights the importance of whole-body MRI in individuals diagnosed with OD or MS in order to facilitate timely diagnoses of possible chondrosarcomas.

OD has been associated with an increased risk of intracranial malignancies, specifically brain tumors of glial origin (Ranger and Szymczak, 2009). A literature review by Ranger and Szymczak (2009) found that among patients with intracranial neoplasms, patients with OD were more likely to have glial malignancies compared to patients with MS (Schaison et al., 1999). In our study, we estimated the prevalence of brain malignancies of glial origin and compared the risk in OD and MS. In our literature review, we found that a significantly higher proportion of individuals with OD were diagnosed with brain malignancies (16.3%—OD, 4.5%—MS; p < .001), and that patients with OD were four times more likely to be diagnosed with a brain malignancy (95% CI: 1.86–9.0). Likely due to our small sample size for patients with MS, we did not observe the same significant difference in the prevalence of brain malignancies in our survey (2.4%—OD, 0%—MS; p = .350).

Gonadal neoplasms, specifically juvenile granulosa cell tumor, have also been associated with OD and MS. A literature review by Pansuriya et al. (2010) found seven cases of juvenile granulosa cell tumor in patients with OD, and six cases in patients with MS. In 2017, Burgetova et al. (2017) described 10 cases of juvenile granulosa cell tumor in patients with OD. Our literature review found a slightly higher prevalence of gonadal malignancies in OD compared to MS (8.1%—OD, 3.5%—MS; p = .059). In contrast, our survey found a significantly higher prevalence of gonadal malignancies in patients with MS (0%—OD, 5.6%—MS; p = .008). Due to the young age of individuals with OD, our survey may underestimate the true prevalence of all cancers in OD, including gonadal neoplasms.

Although malignant transformation to angiosarcoma has been previously described, the prevalence of malignant transformation of the vascular anomalies in patients with MS is not known (Prokopchuk et al., 2016). In our literature review, as expected, the prevalence of vascular malignancies was significantly higher in MS compared to OD (0.006%—OD, 8.5%—MS; p = .001), and patients with MS were 14.9 times more likely to be diagnosed with a vascular malignancy compared to patients with OD (95% CI: 0.009–0.5). In theory, patients with OD should not be prone to developing vascular malignancies, as vascular anomalies are the hallmark of MS. Interestingly, one case with OD in our literature review was described with a vascular malignancy, which may reflect an error in diagnosis (Murakami et al., 1993). No individuals with MS in our survey developed a vascular malignancy, which may be due to our small sample size (n = 36).

Based on our review of the literature and our survey, individuals with OD will be diagnosed earlier with enchondromas compared with the individuals with MS. In OD, enchondromas are most commonly found in the femur, small tubular bones of the hands, and tibia. Enchondromas also affect skull, ribs, and spine and predominantly grow on only one side of the body. About half of individuals with OD will develop a malignancy, and these malignancies will be diagnosed at a significantly younger age than in individuals with MS (Table 2). Approximately a third of patients will develop chondrosarcoma, and approximately a fifth will develop a brain cancer. Based on our review and survey, we suggest the following diagnosis and surveillance guideline for patients diagnosed with OD: (a) Patients should have a whole-body MRI/MRA at the time of their diagnosis for a full assessment of all the bones affected, for the assessment of the kind of bone and/or cartilage tumors they have, and for the assessment of vascular anomalies in the visceral organs; (b) Depending on the type of bone and cartilage lesions that they have, on the other features that they present, and on their family history, sequencing of PTPN11, EXT1, EXT2, ACP5, and/or NKX3–2 is indicated. Sequencing of these genes is recommended to rule out similar disorders characterized by enchondromas and/or exostoses, including metachondromatosis (OMIM #156250), multiple exostoses Types 1 and 2 (OMIM #133700 and #133701), spondyloenchondrodysplasia with immune dysregulation or SPENCDI (OMIM #607944), and spondylomegaepiphyseal-metaphyseal dysplasia (OMIM #613330); (c) Deletions and duplications should be investigated by an array; (d) A urine organic acid test should be performed for investigation of hydroxyglutaric aciduria; (e) In children, periodic clinical screening every 6–12 months and plain radiographs of the known lesions every 2–3 years should be performed for early detection of growth abnormalities that need surgical treatment; (f) In adults, regular clinical examination every 12–24 months and plain radiographs of the known lesions every 2–3 years depending on the enchondroma location should be performed for early detection and adequate treatment of malignant transformation; (g) If chondrosarcoma is suspected, MRI is recommended since it is the imaging of choice to distinguish between enchondroma and low-grade chondrosarcoma; (h) Needle biopsy may be necessary for the diagnosis of chondrosarcoma; (i) If an individual have an enchondroma located in the pelvis, or scapula, a yearly clinical examination together with a whole-body MRI should be performed; (j) If an individual have an enchondroma bigger than 5–6 cm, a yearly clinical examination together with a whole-body MRI should be performed; and (k) A thorough clinical examination along with a full body MRI should be offered starting at 25 years of age (median age of malignancy diagnosis in OD) and reoffered every other year because of the risk of chondrosarcoma, brain tumors, and other malignancies (Table S1). Clinicians should investigate possible malignancy occurrence at a younger age if symptoms are suggestive of malignant transformation. In case of new pain, immediate clarification including clinical and radiological examination is indicated. If MRI is not possible for any reason, a bone-scintigraphy is recommended instead. Computed tomography imaging could be considered in anatomically difficult regions such as pelvis and scapula (Beltrami et al., 2016; Herget et al., 2014; Sonne-Holm, Wong, & Sonne-Holm, 2014; Table 7).

TABLE 7.

Recommended diagnosis and surveillance protocol for patients with Ollier disease (OD) or Maffucci syndrome (MS)

1.	Patients should be submitted to a whole-body MRI/MRA at the time of their diagnosis for a full assessment of all the bones affected, for the assessment of the kind of bone and/or cartilage tumors they have, and for the assessment of vascular anomalies in the visceral organs— At the time of diagnosis.
2.	Depending on the type of bone and cartilage lesions that they have, on the other features that they present, and on their family history, sequencing of PTPN11, EXT1, EXT2, EXT3, ACP5, and/or NKX3–2 is indicated— At the time of diagnosis.
3.	Deletions and duplications should be investigated by an array-At the time of diagnosis.
4.	A urine organic acid test should be performed for investigation of hydroxyglutaric aciduria-At the time of diagnosis.
5.	In children, periodic clinical screening every 6–12 months and plain radiographs of the known lesions every 2–3 years should be performed for early detection of growth abnormalities that need surgical treatment.
6.	In adults, regular clinical examination every 12–24 months and plain radiographs of the known lesions every 2–3 years depending on the enchondroma location should be performed for early detection and adequate treatment of malignant transformation.
7.	If chondrosarcoma is suspected, MRI is recommended since it is the imaging of choice to distinguish between enchondroma and low-grade chondrosarcoma.
8.	Needle biopsy may be necessary for the diagnosis of chondrosarcoma.
9.	If an individual have an enchondroma located in the pelvis, or scapula, a yearly clinical examination together with a whole-body MRI should be performed.
10.	If an individual have an enchondroma bigger than 5–6 cm, a yearly clinical examination together with a whole-body MRI should be performed.
11.	A thorough clinical examination along with a full body MRI should be offered starting at 25 years of age (median age of malignancy diagnosis in OD) and reoffered every other year because of the risk of chondrosarcoma, brain tumors, and other malignancies. Clinicians should investigate possible malignancy occurrence at a younger age if symptoms are suggestive of malignant transformation. In case of new pain, immediate clarification including clinical and radiological examination is indicated. If MRI is not possible for any reason, a bone-scintigraphy is recommended instead. Computed tomography imaging could be considered in anatomically difficult regions such as pelvis and scapula.

Based on our review of the literature and our survey, we found that individuals with MS have a later diagnosis of enchondromas and vascular anomalies when compared to the individuals with OD. In MS, enchondromas are most commonly found in the small tubular bones of the hands, femur, and tibia, but the ribs, skull, and spine can also be affected. Half of individuals with MS will have both sides of the body affected by enchondromas, but most vascular anomalies will be unilateral. Malignancies are diagnosed at an older age (∼30 years of age) than in patients with OD. Half of individuals with MS will develop a malignancy; ∼32% will develop a chondrosarcoma, and 8.5% will develop a vascular malignancy. Based on our review and survey, we suggest the same diagnosis and surveillance guideline as described above for the patients with MS with one modification only: A thorough clinical examination along with a full body MRI should be offered starting at 30 years of age (median age of malignancy diagnosis in MS) and reoffered every other year because of the risk of chondrosarcoma and brain tumors (Table 7).

Here, for the surveillance of chondrosarcomas, brain tumors and other malignancies, we choose to use the median age of malignancy diagnosis identified by the literature review since only 36 patients with MS answered the survey while 199 patients with MS were identified by the literature review.

The lack of diagnostic and surveillance guidelines for patients with OD and MS has led to mistaken diagnosis and reproductive counseling or to the late diagnosis of malignancies for many individuals with these disorders. Here, we adapted the guidelines suggested by Herget et al. (2014) to include genetic tests that will rule out other conditions during the diagnosis process and to include surveillance for other malignancies besides chondrosarcomas. We chose to recommend that screening should begin at the median age of cancer diagnoses, since measures of mean and standard deviation are limited by small sample sizes.

Our study is not without limitations. In both our literature review and survey, no review of radiographs and histology was performed. As such, the validity of our data depends on the validity of case reports in the literature and patient self-reported information in the survey. Furthermore, the large difference in sample size in our survey between patients with OD and MS influences the power of our statistical analysis. In addition, the age of the cohorts reported here are not similar, which may introduce lead time bias to our study. Despite these limitations, our review of the literature and survey together represent the largest case study of individuals with OD and MS ascertained to date and provide the most comprehensive, current description of each disease’s natural history. It shows that OD and MS are both characterized by a wide phenotypic variability and that while both are characterized by the occurrence of enchondromas and chondrosarcomas they are clinically different with different age of onset, different distribution of the enchondromas, different distribution of the chondrosarcomas, and by different prevalence of other malignancies, mainly gliomas.