0
We're unable to sign you in at this time. Please try again in a few minutes.
Retry
We were able to sign you in, but your subscription(s) could not be found. Please try again in a few minutes.
Retry
There may be a problem with your account. Please contact the AMA Service Center to resolve this issue.
Contact the AMA Service Center:
Telephone: 1 (800) 262-2350 or 1 (312) 670-7827  *   Email: subscriptions@jamanetwork.com
Error Message ......
Observation |

Familial Acanthosis Nigricans Due to K650T FGFR3 Mutation FREE

David R. Berk, MD; Elaine B. Spector, PhD; Susan J. Bayliss, MD
[+] Author Affiliations

Author Affiliations: Department of Internal Medicine, Division of Dermatology, and Department of Pediatrics, Washington University School of Medicine, St Louis, Missouri (Drs Berk and Bayliss); and DNA Diagnostic Laboratory and Department of Pediatrics, University of Colorado at Denver and Health and Sciences Center, Aurora (Dr Spector).


Arch Dermatol. 2007;143(9):1153-1156. doi:10.1001/archderm.143.9.1153.
Text Size: A A A
Published online

Background  Acanthosis nigricans is a feature of several syndromes caused by activating mutations of the fibroblast growth factor receptor 3 gene (FGFR3), including Crouzon syndrome with acanthosis nigricans, thanatophoric dysplasia, and severe achondroplasia with developmental delay and acanthosis nigricans (SADDAN syndrome).

Observations  We describe a healthy 4-year-old African American girl with generalized acanthosis nigricans since infancy. Her father had a history of acanthosis nigricans since childhood, in addition to Crohn disease, obesity, and adult-onset diabetes mellitus. A pedigree with numerous affected family members was constructed. Other than slightly short stature, no associated anomalies were found, including dysmorphic features or skeletal or neurologic defects. Genetic testing revealed a previously undescribed, heterozygous lysine to threonine mutation at codon 650 of the FGFR3 gene in the 4 affected family members who were tested.

Conclusion  Extensive acanthosis nigricans in early childhood, especially with a family history of acanthosis nigricans, may warrant testing for FGFR3 mutations.

Figures in this Article

Numerous genetic syndromes are associated with acanthosis nigricans (AN), particularly syndromes characterized by obesity, hyperinsulinemia, and/or craniosynostosis.1 In a recent review, Torley et al1 divided these genetic syndromes into insulin resistance syndromes and fibroblast growth factor (FGF) defects. Insulin resistance syndromes are caused by mutations in the insulin receptor (leprechaunism and Rabson-Mendenhall syndrome), peroxisome proliferator-activated receptor-γ (insulin-resistant diabetes mellitus with AN and hypertension), 1-acylglycerol-3-phophate O-acyltransferase-2 (Berardinelli-Seip syndrome), seipin (Berardinelli-Seip syndrome), lamin A/C (Dunnigan syndrome), and Alstrom syndrome genes. Fibroblast growth factor defects associated with AN include specific activating mutations of FGFR2 (Beare-Stevenson syndrome) and FGFR3 (Crouzon syndrome with AN and thanatophoric dysplasia [TD], and severe achondroplasia with developmental delay and AN [SADDAN syndrome]).

We describe herein a family with AN associated with a previously undescribed lysine to threonine mutation at codon 650 of the FGFR3 gene. All 4 affected family members who were tested demonstrated the mutation. The inheritance pattern was autosomal dominant, and there were no obvious associated skeletal or neurological abnormalities other than short stature.

A 4-year-old African American girl was referred for evaluation of generalized, minimally pruritic skin thickening since infancy. She was diagnosed as having atopic dermatitis by her pediatrician years earlier but had not responded to treatment with topical corticosteroids. She was otherwise healthy without diabetes mellitus and was taking no medications. Neurological development was normal. Her father (age, 25 years) described having a similar skin condition since early childhood. He was slightly obese and had Crohn disease and mild adult-onset diabetes mellitus (diet controlled). There was no history of consanguinity.

On physical examination, the patient had generalized, thick, velvety, hyperpigmented plaques that were most prominent on her neck, back, and axillae (Figure 1). There was no erythema or lichenification. Her oral mucosa, palms, and external genitalia were normal. Height and weight at age 5.5 years were 105.5 cm (eighth percentile) and 17.3 kg (20th percentile), respectively. Results from laboratory evaluations (complete blood cell count, urinalysis, and random glucose and hemoglobin A1c measurements) were normal. Examination of the patient's father revealed similarly thick, velvety, hyperpigmented plaques localized to the neck, back, and axillae. His height was 156 cm (< fifth percentile). Over the past few years, his weight had fluctuated from 48 to 106 kg due to his Crohn disease. Neither the patient nor her father had facial dysmorphism (eg, frontal bossing) or tibial bowing. Familial AN was diagnosed.

Place holder to copy figure label and caption
Figure 1.

Patient at age 4 years with thick, velvety, hyperpigmented plaque involving the neck and chest.

Graphic Jump Location

Because activating FGFR3 mutations have been identified in several forms of syndromic AN, this gene was evaluated for mutations. Genetic testing was performed on DNA isolated from saliva (Oragene; DNA Genotek Inc, Ottawa, Ontario, Canada) followed by sequencing of exons 10, 13, and 15 of the FGFR3 gene. A single nucleotide change was observed in exon 15 (Figure 2). The patient and her father both demonstrated a previously undescribed lysine to threonine mutation at codon 650 (K650T) in one FGFR3 gene, suggesting autosomal dominant inheritance. No mutation was identified in the patient's mother or the patient's 3-year-old sister, both of whom were clinically unaffected. Genetic counseling was performed. There was no change with topical pimecrolimus and steroids.

Place holder to copy figure label and caption
Figure 2.

Sequencing of exon 15 of the FGFR3 gene was performed using the following primers: forward, 5′ GTA AAA CGA CGG CCA GT AGG TGT GGG TGG AGT AGG; and reverse, 5′ CAG GAA ACA GCT ATG ACC TC AGG CGC CAT CCA CTT 5′ CTG TCA CCG TAG CCG TGA AG. Sequencing was performed using an ABI 3130 sequencer (Applied Biosystems, Foster City, California). A, The arrow indicates K650T mutation; A→C change at nucleotide 1950, amino acid 650. B, Normal sequence.

Graphic Jump Location

The father's 29-year-old cousin (Figure 3) and her 11-year-old daughter (Figure 4) were subsequently evaluated for the same condition and diagnosed as having familial AN, resulting in a pedigree with multiple affected relatives (Figure 5). The daughter's height and weight at age 11 years were 139 cm (26th percentile) and 40.9 kg (68th percentile), respectively. Her mother's height and weight were 145 cm (< fifth percentile) and 64.2 kg (25th-50th percentile), respectively. No further associated anomalies were identified in these 2 patients or reported in their relatives, including dysmorphic features, diabetes mellitus, Crohn disease, or skeletal or neurologic defects. Both this child and her mother demonstrated heterozygous K650T mutations on genetic testing, identical to their 2 relatives who had been previously tested.

Place holder to copy figure label and caption
Figure 3.

Cousin (age, 29 years) demonstrating thick, velvety, hyperpigmented plaques on the neck and posterior auricular region.

Graphic Jump Location
Place holder to copy figure label and caption
Figure 4.

Cousin (age, 11 years) demonstrating thick, velvety, hyperpigmented plaques on the neck and forehead.

Graphic Jump Location
Place holder to copy figure label and caption
Figure 5.

Pedigree of family. Shaded boxes indicate affected family members. The asterisks indicate family members tested for FGFR3 mutations.

Graphic Jump Location

Acanthosis nigricans is characterized by velvety, hyperpigmented plaques often accentuated in the flexures. Numerous forms of AN have been described including benign (nonsyndromic, insulin resistance-associated AN), malignant (paraneoplastic), obesity-associated (pseudo-AN), acral, medication-induced (corticosteroids, estrogens, oral contraceptives, niacin, triazinate, somatotrophin, and diethylstilbestrol), and syndromic AN (including isolated, “pure” familial AN).2,3

Generalized AN is rare and often paraneoplastic,4 even in very early childhood. However, generalized AN in early childhood may also suggest isolated (nonsyndromic, noninsulin resistance–associated) familial AN.48 Furthermore, there are reports of benign, generalized AN in early childhood without a positive family history, malignancy, associated syndrome, causative medication, or comorbid condition.915

Only a few families with isolated (nonsyndromic, noninsulin resistance–associated) familial AN have been reported in the literature.48 Inheritance tends to be autosomal dominant with variable penetrance. This “pure” familial AN usually appears in infancy, stabilizes at puberty, and is not associated with obesity or diabetes mellitus. The term familial AN is confusing because AN associated with insulin resistance also tends to be hereditary and the term familial AN has been used in the literature in such cases.16,17 Comorbid conditions have rarely been reported with isolated familial AN. Chuang et al18 reported a family with AN and ectodermal defects including madarosis. In addition, 2 cases of benign AN (no family history) in early childhood have been associated with pyramidal tract degeneration.11,15 Familial AN may be confused with ichthyosis hystrix,5 generalized epidermal nevi, or lichenification in atopic dermatitis.

The pathogenesis of AN is poorly understood. In patients with hyperinsulinemia, excess insulin may directly or indirectly stimulate epidermal proliferation through the insulin-like growth factor 1 receptor.19 Growth factors (particularly transforming growth factor-α) produced by tumors may cause malignant AN through the epidermal growth factor receptor.20

Fibroblast growth factors are involved in angiogenesis, embryogenesis, mitogenesis, and wound healing. There are more than 20 human FGF ligands and 4 human FGF receptors. The FGFR3 gene is located at 4p16.3 and encodes a transmembrane receptor tyrosine kinase, which down-regulates long-bone growth. FGFR3 mutations have been reported in several dermatologic conditions including seborrheic keratoses,21 epidermal nevi,22 and syndromic AN. Several autosomal dominant syndromes are caused by specific FGFR3 mutations, including achondroplasia, Muenke syndrome, hypochondroplasia, TD I and II, Crouzon syndrome with AN, and SADDAN syndrome. The latter 3 disorders demonstrate AN.

Codon 650 of FGFR3 is located in its tyrosine kinase domain II. Mutations of this codon have been reported in skeletal disorders including hypochondroplasia (K650N and K650Q), SADDAN syndrome (K650M), TD I (K650M), and TD II (K650E), as well as malignant conditions including bladder cancer and multiple myeloma (K650M, K650E, and K650T). Germline K650T mutations have never been reported. Bellus et al23 studied the effects of various amino acid changes on the activation of FGFR3 in vitro. Constructs were made that contained all of the possible FGFR3 mutations at amino acid 650. In a [32P]-ATP autophosphorylation assay, the relative activity of the mutations compared with that of a wild-type construct was as follows: K650R, 1.5x; K650N, 3.7x; K650Q, 4.9x; K650T, 3.1x; K650E, 9.6x; and K650M, 18.1x. These data suggest that since the K650T mutation was a relatively weak activator of FGFR3, it would have a less severe effect on phenotype than the mutations found in hypochondroplasia, SADDAN syndrome, TD I, and TD II, as was observed in the patients described herein.

Activating FGFR3 mutations and/or ectopic overexpression have been reported in a number of tumors, such as bladder,24 cervical,25 multiple myeloma,26 and colorectal cancers.27 Tyrosine kinase inhibitors of FGFR3 are being tested in several hematologic malignancies with positive results.26 Topical preparations of these inhibitors may prove beneficial for seborrheic keratoses, epidermal nevi, and/or AN.22

In summary, we describe a family with AN caused by an autosomal dominant K650T mutation in the FGFR3 gene. Extensive AN in early childhood, especially with a family history of AN, may warrant testing for FGFR3 mutations.

Correspondence: David R. Berk, MD, Division of Dermatology, Washington University School of Medicine, 660 S Euclid, Campus Box 8123, St Louis, MO 63110 (DBerk@im.wustl.edu).

Financial Disclosure: None reported.

Accepted for Publication: March 27, 2007.

Author Contributions: Dr Berk had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Berk and Bayliss. Acquisition of data: Berk and Spector. Analysis and interpretation of data: Berk and Spector. Drafting of the manuscript: Berk and Spector. Critical revision of the manuscript for important intellectual content: Berk, Spector, and Bayliss. Statistical analysis: Berk. Administrative, technical, and material support: Berk, Spector, and Bayliss. Study supervision: Berk and Spector.

Torley  DBellus  GAMunro  CS Genes, growth factors and acanthosis nigricans. Br J Dermatol 2002;147 (6) 1096- 1101
PubMed
Curth  HO Classification of acanthosis nigricans. Int J Dermatol 1976;15 (8) 592- 593
PubMed
Curth  HO Acanthosis nigricans. Birth Defects Orig Artic Ser 1971;7 (8) 31- 39
PubMed
Inamadar  ACPalit  A Generalized acanthosis nigricans in childhood. Pediatr Dermatol 2004;21 (3) 277- 279
PubMed
Curth  HOAschner  BM Genetic studies on acanthosis nigricans. AMA Arch Derm 1959;79 (1) 55- 66
PubMed
Curth  HO Benign acanthosis nigricans. Arch Dermatol 1967;96 (5) 603- 604
PubMed
Tasjian  DJarratt  M Familial acanthosis nigricans. Arch Dermatol 1984;120 (10) 1351- 1354
PubMed
Dhar  SDawn  GKanwar  AJNada  R Familial acanthosis nigricans. Int J Dermatol 1996;35 (2) 126- 127
PubMed
Curth  HO Benign type of acanthosis nigricans. Arch Derm Syphilol 1936;34 (3) 353- 366
Pindborg  JJGorlin  RJ Oral changes in acanthosis nigricans (juvenile type): survey of the literature and report of a case. Acta Derm Venereol 1962;4263- 71
PubMed
Maranon  GAlvarez Cascos  M Benign juvenile acanthosis nigricans; its relation to congenital abnormalities. Acta Derm Venereol 1957;37 (3) 249- 258
PubMed
Skiljevic  DSNikolic  MMJakovljevic  ADobrosavljevic  DD Generalized acanthosis nigricans in early childhood. Pediatr Dermatol 2001;18 (3) 213- 216
PubMed
Uyttendaele  HKoss  TBagheri  B  et al.  Generalized acanthosis nigricans in an otherwise healthy young child. Pediatr Dermatol 2003;20 (3) 254- 256
PubMed
Rai  VMBalachandran  C Generalized acanthosis nigricans in childhood. Dermatol Online J 2006;12 (6) 14
PubMed
Curth  HO Significance of acanthosis nigricans. AMA Arch Derm Syphilol 1952;66 (1) 80- 100
PubMed
Taylor  GJames  MPSimpson  H Familial acanthosis nigricans. J R Soc Med 1994;87 (3) 169
PubMed
Friedman  CIRichards  SKim  MH Familial acanthosis nigricans: a longitudinal study. J Reprod Med 1987;32 (7) 531- 536
PubMed
Chuang  SDJee  SHChiu  HCChen  JSLin  JT Familial acanthosis nigricans with madarosis. Br J Dermatol 1995;133 (1) 104- 108
PubMed
Cruz  PD  JrHud  JA  Jr Excess insulin binding to insulin-like growth factor receptors: proposed mechanism for acanthosis nigricans. J Invest Dermatol 1992;98 (6) ((suppl)) 82S- 85S
PubMed
Haase  IHunzelmann  N Activation of epidermal growth factor receptor/ERK signaling correlates with suppressed differentiation in malignant acanthosis nigricans. J Invest Dermatol 2002;118 (5) 891- 893
PubMed
Logié  ADunois-Larde  CRosty  C  et al.  Activating mutations of the tyrosine kinase receptor FGFR3 are associated with benign skin tumors in mice and humans. Hum Mol Genet 2005;14 (9) 1153- 1160
PubMed
Hafner  Cvan Oers  JMVogt  T  et al.  Mosaicism of activating FGFR3 mutations in human skin causes epidermal nevi. J Clin Invest 2006;116 (8) 2201- 2207
PubMed
Bellus  GASpector  EBSpeiser  PW  et al.  Distinct missense mutations of the FGFR3 lys650 codon modulate receptor kinase activation and the severity of the skeletal dysplasia phenotype. Am J Hum Genet 2000;67 (6) 1411- 1421
PubMed
van Rhijn  BWvan Tilborg  AALurkin  I  et al.  Novel fibroblast growth factor receptor 3 (FGFR3) mutations in bladder cancer previously identified in non-lethal skeletal disorders. Eur J Hum Genet 2002;10 (12) 819- 824
PubMed
Cappellen  DDe Oliveira  CRicol  D  et al.  Frequent activating mutations of FGFR3 in human bladder and cervix carcinomas. Nat Genet 1999;23 (1) 18- 20
PubMed
Grand  EKChase  AJHeath  CRahemtulla  ACross  NC Targeting FGFR3 in multiple myeloma: inhibition of t(4;14)-positive cells by SU5402 and PD173074. Leukemia 2004;18 (5) 962- 966
PubMed
Jang  JHShin  KHPark  JG Mutations in fibroblast growth factor receptor 2 and fibroblast growth factor receptor 3 genes associated with human gastric and colorectal cancers. Cancer Res 2001;61 (9) 3541- 3543
PubMed

Figures

Place holder to copy figure label and caption
Figure 1.

Patient at age 4 years with thick, velvety, hyperpigmented plaque involving the neck and chest.

Graphic Jump Location
Place holder to copy figure label and caption
Figure 2.

Sequencing of exon 15 of the FGFR3 gene was performed using the following primers: forward, 5′ GTA AAA CGA CGG CCA GT AGG TGT GGG TGG AGT AGG; and reverse, 5′ CAG GAA ACA GCT ATG ACC TC AGG CGC CAT CCA CTT 5′ CTG TCA CCG TAG CCG TGA AG. Sequencing was performed using an ABI 3130 sequencer (Applied Biosystems, Foster City, California). A, The arrow indicates K650T mutation; A→C change at nucleotide 1950, amino acid 650. B, Normal sequence.

Graphic Jump Location
Place holder to copy figure label and caption
Figure 3.

Cousin (age, 29 years) demonstrating thick, velvety, hyperpigmented plaques on the neck and posterior auricular region.

Graphic Jump Location
Place holder to copy figure label and caption
Figure 4.

Cousin (age, 11 years) demonstrating thick, velvety, hyperpigmented plaques on the neck and forehead.

Graphic Jump Location
Place holder to copy figure label and caption
Figure 5.

Pedigree of family. Shaded boxes indicate affected family members. The asterisks indicate family members tested for FGFR3 mutations.

Graphic Jump Location

Tables

References

Torley  DBellus  GAMunro  CS Genes, growth factors and acanthosis nigricans. Br J Dermatol 2002;147 (6) 1096- 1101
PubMed
Curth  HO Classification of acanthosis nigricans. Int J Dermatol 1976;15 (8) 592- 593
PubMed
Curth  HO Acanthosis nigricans. Birth Defects Orig Artic Ser 1971;7 (8) 31- 39
PubMed
Inamadar  ACPalit  A Generalized acanthosis nigricans in childhood. Pediatr Dermatol 2004;21 (3) 277- 279
PubMed
Curth  HOAschner  BM Genetic studies on acanthosis nigricans. AMA Arch Derm 1959;79 (1) 55- 66
PubMed
Curth  HO Benign acanthosis nigricans. Arch Dermatol 1967;96 (5) 603- 604
PubMed
Tasjian  DJarratt  M Familial acanthosis nigricans. Arch Dermatol 1984;120 (10) 1351- 1354
PubMed
Dhar  SDawn  GKanwar  AJNada  R Familial acanthosis nigricans. Int J Dermatol 1996;35 (2) 126- 127
PubMed
Curth  HO Benign type of acanthosis nigricans. Arch Derm Syphilol 1936;34 (3) 353- 366
Pindborg  JJGorlin  RJ Oral changes in acanthosis nigricans (juvenile type): survey of the literature and report of a case. Acta Derm Venereol 1962;4263- 71
PubMed
Maranon  GAlvarez Cascos  M Benign juvenile acanthosis nigricans; its relation to congenital abnormalities. Acta Derm Venereol 1957;37 (3) 249- 258
PubMed
Skiljevic  DSNikolic  MMJakovljevic  ADobrosavljevic  DD Generalized acanthosis nigricans in early childhood. Pediatr Dermatol 2001;18 (3) 213- 216
PubMed
Uyttendaele  HKoss  TBagheri  B  et al.  Generalized acanthosis nigricans in an otherwise healthy young child. Pediatr Dermatol 2003;20 (3) 254- 256
PubMed
Rai  VMBalachandran  C Generalized acanthosis nigricans in childhood. Dermatol Online J 2006;12 (6) 14
PubMed
Curth  HO Significance of acanthosis nigricans. AMA Arch Derm Syphilol 1952;66 (1) 80- 100
PubMed
Taylor  GJames  MPSimpson  H Familial acanthosis nigricans. J R Soc Med 1994;87 (3) 169
PubMed
Friedman  CIRichards  SKim  MH Familial acanthosis nigricans: a longitudinal study. J Reprod Med 1987;32 (7) 531- 536
PubMed
Chuang  SDJee  SHChiu  HCChen  JSLin  JT Familial acanthosis nigricans with madarosis. Br J Dermatol 1995;133 (1) 104- 108
PubMed
Cruz  PD  JrHud  JA  Jr Excess insulin binding to insulin-like growth factor receptors: proposed mechanism for acanthosis nigricans. J Invest Dermatol 1992;98 (6) ((suppl)) 82S- 85S
PubMed
Haase  IHunzelmann  N Activation of epidermal growth factor receptor/ERK signaling correlates with suppressed differentiation in malignant acanthosis nigricans. J Invest Dermatol 2002;118 (5) 891- 893
PubMed
Logié  ADunois-Larde  CRosty  C  et al.  Activating mutations of the tyrosine kinase receptor FGFR3 are associated with benign skin tumors in mice and humans. Hum Mol Genet 2005;14 (9) 1153- 1160
PubMed
Hafner  Cvan Oers  JMVogt  T  et al.  Mosaicism of activating FGFR3 mutations in human skin causes epidermal nevi. J Clin Invest 2006;116 (8) 2201- 2207
PubMed
Bellus  GASpector  EBSpeiser  PW  et al.  Distinct missense mutations of the FGFR3 lys650 codon modulate receptor kinase activation and the severity of the skeletal dysplasia phenotype. Am J Hum Genet 2000;67 (6) 1411- 1421
PubMed
van Rhijn  BWvan Tilborg  AALurkin  I  et al.  Novel fibroblast growth factor receptor 3 (FGFR3) mutations in bladder cancer previously identified in non-lethal skeletal disorders. Eur J Hum Genet 2002;10 (12) 819- 824
PubMed
Cappellen  DDe Oliveira  CRicol  D  et al.  Frequent activating mutations of FGFR3 in human bladder and cervix carcinomas. Nat Genet 1999;23 (1) 18- 20
PubMed
Grand  EKChase  AJHeath  CRahemtulla  ACross  NC Targeting FGFR3 in multiple myeloma: inhibition of t(4;14)-positive cells by SU5402 and PD173074. Leukemia 2004;18 (5) 962- 966
PubMed
Jang  JHShin  KHPark  JG Mutations in fibroblast growth factor receptor 2 and fibroblast growth factor receptor 3 genes associated with human gastric and colorectal cancers. Cancer Res 2001;61 (9) 3541- 3543
PubMed

Correspondence

CME
Meets CME requirements for:
Browse CME for all U.S. States
Accreditation Information
The American Medical Association is accredited by the Accreditation Council for Continuing Medical Education to provide continuing medical education for physicians. The AMA designates this journal-based CME activity for a maximum of 1 AMA PRA Category 1 CreditTM per course. Physicians should claim only the credit commensurate with the extent of their participation in the activity. Physicians who complete the CME course and score at least 80% correct on the quiz are eligible for AMA PRA Category 1 CreditTM.
Note: You must get at least of the answers correct to pass this quiz.
You have not filled in all the answers to complete this quiz
The following questions were not answered:
Sorry, you have unsuccessfully completed this CME quiz with a score of
The following questions were not answered correctly:
Commitment to Change (optional):
Indicate what change(s) you will implement in your practice, if any, based on this CME course.
Your quiz results:
The filled radio buttons indicate your responses. The preferred responses are highlighted
For CME Course: A Proposed Model for Initial Assessment and Management of Acute Heart Failure Syndromes
Indicate what changes(s) you will implement in your practice, if any, based on this CME course.
NOTE:
Citing articles are presented as examples only. In non-demo SCM6 implementation, integration with CrossRef’s "Cited By" API will populate this tab (http://www.crossref.org/citedby.html).
Submit a Comment

Multimedia

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging & repositioning the boxes below.

See Also...
Articles Related By Topic
Related Topics
PubMed Articles