Published online May 1, 2007
PEDIATRICS Vol. 119 No. 5 May 2007, pp. e1088-e1093 (doi:10.1542/peds.2006-2315)
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow P3Rs: Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when P3Rs are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow E-mail this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My File Cabinet
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via CrossRef
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Johnson, K. J.
Right arrow Articles by Ross, J. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Johnson, K. J.
Right arrow Articles by Ross, J. A.
Related Collections
Right arrow Tumors

ARTICLE

Childhood Cancer and Birthmarks in the Collaborative Perinatal Project

Kimberly J. Johnson, MPHa, Logan G. Spector, PhDa, Mark A. Klebanoff, MD, MPHb and Julie A. Ross, PhDa

a Department of Pediatrics, University of Minnesota, Minneapolis, Minnesota
b National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, Maryland


   ABSTRACT
TOP
 ABSTRACT
METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
OBJECTIVE. Three previous retrospective studies noted a positive association between birthmarks and childhood cancer. The objective of this study was to determine whether the incidence of cancer is increased in children with birthmarks relative to those without birthmarks using data from the Collaborative Perinatal Project cohort, a large, prospective study.

METHODS. Our study population comprised 49503 US children who were born between 1959 and 1968. Birthmarks were documented as definite or suspected during the first year through history or medical examinations and included hemangiomas (port-wine, strawberry, or cavernous), pigmented nevi, lymphangiomas, and café-au-lait spots. The association between birthmarks and childhood cancer was determined using Cox proportional hazards regression.

RESULTS. In the Collaborative Perinatal Project, 2505 individuals had a documented definite or suspected birthmark, including 7 of 47 children who developed cancer. Birthmarks were associated with a significant increase in the risk for cancer. There was a slight attenuation of the risk estimate when cases that were diagnosed in the first year of life were excluded. No specific childhood malignancies were notably affected by birthmarks.

CONCLUSIONS. Although this study was based on a small number of cases, we found birthmarks to be in excess in children who received a diagnosis of cancer using prospective data. These findings provide additional support for the possibility of a shared etiology between birthmarks and childhood cancer that could offer insight into the pathogenesis of pediatric malignancy.

Key Words: birthmarks • childhood cancer • etiology • risk factors

Abbreviations: BWS—Beckwith-Wiedemann syndrome • NF1—neurofibromatosis type I • ALL—acute lymphoblastic leukemia • CPP—Collaborative Perinatal Project • HR—hazard ratio • CI—confidence interval • SES—socioeconomic status • AML—acute myelogenous leukemia • IGF—insulin-like growth factor

It is widely recognized that genetic syndromes and chromosomal anomalies such as Down syndrome,1,2 Beckwith-Wiedemann syndrome (BWS),3,4 and neurofibromatosis type I (NF1)5,6 are associated with a markedly increased risk for childhood malignancies. Studies also have suggested that children with minor malformations or variants, including "birthmarks," may have an increased risk,7–11 and these could be a marker of "altered prenatal development."12 A significant increased frequency of birthmarks in children with acute lymphoblastic leukemia (ALL) and rhabdomyosarcoma has been reported in 3 studies.9–11 To provide additional data on the association between childhood cancer and the presence of skin lesions that are detected during infancy, we explored whether birthmarks were associated with childhood cancer in the Collaborative Perinatal Project (CPP), a large, prospective cohort study.


   METHODS
TOP
 ABSTRACT
 METHODS
RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Study Population
The CPP cohort was assembled between 1959 and 1966 with the main objective of investigating the causes of neurologic defects. Complete details on the study have been described elsewhere.13 Briefly, ~45000 women were enrolled during their pregnancies at 12 academic institutions throughout the United States. A total of 59843 pregnancies that resulted in 54795 live births occurred. Data were collected at numerous time points by trained interviewers and examiners using standardized methods. During the prenatal phase, extensive information was collected on the mother's health history, pregnancy-related events, social and demographic factors, and family medical histories. During and after delivery, trained observers were present to collect standard information and to document any significant events. Children were examined several times during the newborn period and at 4, 8, and 12 months during the first year. For most individuals in the cohort, a summary form was filled out to record all events, illnesses, and conditions that occurred between discharge from the birth hospital and the first birthday. Investigators continued to collect data at several time points through the age of 7 or 8 years (depending on the institution).

Our study population for the analysis of the association between birthmarks and childhood cancers was composed of 49503 children who were born alive and had documented information on birthmarks that were observed during the first year of life. Fifty-one cancers that were identified during 8 years of follow-up were described previously by Klebanoff et al.14 Four of the 51 children who received a diagnosis of cancer had no documentation on the existence of birthmarks and were excluded from this analysis. Children were examined for the presence or absence of infantile hemangiomas (strawberry and cavernous), port-wine stains, pigmented nevi, lymphangiomas, and café-au-lait spots at examinations that occurred during the first year of life. Examination forms allowed the examiners to note birthmarks that they considered to be present beyond doubt (definite birthmarks) and those for which they had some degree of doubt (suspected birthmarks). Examiners were instructed to exclude nevus simplex (also known as stork bites or salmon patch), the most common skin lesion that is observed during infancy,15 and not to record pigmented nevi or café-au-lait spots that were <3 cm or fewer than 6.

Statistical Analysis
All statistical analyses were performed using SAS 9.1 (SAS Institute, Cary, NC). The Kaplan-Meier method was used to estimate the cumulative hazard of childhood cancer with respect to the presence of birthmarks.16 We used Cox proportional hazards regression modeling to calculate hazard ratios (HRs), 95% confidence intervals (CIs), and likelihood ratio {chi}2 statistics.16 Person-time was calculated as the number of months from birth to cancer diagnosis or the last date of follow-up as documented in electronic records, whichever came first. The maximum follow-up was 96 months. Because birthmarks were noted throughout the first year of life and not necessarily at birth, analyses were repeated in which person-time started at age 1 year (excluding those who had person-times of <12 months) to disallow the possibility that birthmarks were discovered coincident with cancer diagnosis.

We considered whether the following potential confounders that have previously been linked to childhood cancer17 were associated with cancers that were diagnosed in the CPP in Cox proportional hazards regression models: socioeconomic status (SES; 3 categories), birth weight (≤2500 g, 2501–4000 g, or ≥4000 g), maternal race (white, black, or other), gender, maternal education (less than high school graduate, high school graduate, more than high school graduate), maternal age (≥35 years), and gestational age (<37 weeks). For analyses that examined the association between childhood cancer and birthmarks, we created binary variables to describe the presence or absence of 1 or more of these characteristics using 2 birthmark definitions, 1 that included only definite birthmarks as present and 1 that also included suspected birthmarks. We used the change in estimate method18 using a cutoff of 10% to evaluate whether the association between childhood cancer and birthmarks was confounded by each of the variables described.


   RESULTS
TOP
 ABSTRACT
 METHODS
 RESULTS
DISCUSSION
 CONCLUSIONS
 REFERENCES
 
A total of 47 and 38 incident cancer cases were diagnosed between 0 and 8 years and 1 and 8 years, respectively, in children who had documentation about birthmarks (Table 1). The 3 most common cancers were leukemia, central nervous system tumors, and Wilms’ tumor, as is the normal pattern of childhood cancers in this age group.19


View this table:
[in this window]
[in a new window]

  		TABLE 1 Cancers Diagnosed in the CPP Cohort (N = 49 503)

 
No significant associations (as determined by the likelihood ratio {chi}2 test comparing models with and without the covariate) were observed for any of the following risk factors that are known to be associated with some childhood cancers17 (Table 2): SES, birth weight, maternal race, gender, maternal education, or maternal or gestational age. However, a general tendency was apparent for the HR of childhood cancer to decrease with lower SES and maternal education and for an increased hazard of childhood cancer associated with maternal white ethnicity.


View this table:
[in this window]
[in a new window]

  		TABLE 2 HRs of Childhood Cancer by Potential Confounding Factors in the CPP (N = 49 503)

 
Approximately 15% of cases (n = 7) and 5% of noncases (n = 2498) had a definite or suspected birthmark documented during the first year of life (Table 3), with strawberry hemangiomas and port-wine stains being the most common in both groups. A similar percentage of birthmarks in cases (2 [29%] of 7) and noncases (665 [24%] of 2721) were documented as suspected. The median follow-up time in children with and without documented birthmarks was identical at 96 months.


View this table:
[in this window]
[in a new window]

  		TABLE 3 Frequency of Birthmarks by Type Reported in the CPP (N = 49 503)

 
Figure 1 shows the cumulative hazard of childhood cancer over time with respect to birthmark group. The log-rank test of equality over birthmark strata showed that children with birthmarks were significantly more likely to have a subsequent diagnosis of cancer than those without birthmarks during the follow-up period (P = .0028).


Figure 1
View larger version (9K):
[in this window]
[in a new window]

  		FIGURE 1 Kaplan-Meier estimates of the cumulative hazard of childhood cancer in those with birthmarks relative to those without.

 
Table 4 summarizes the analysis of the association between childhood cancer and birthmarks. Having a documented definite or suspected birthmark was associated with a significantly increased hazard of cancer (HR: 3.19; 95% CI: 1.43–7.12) when all cases (n = 47) were included. When individuals with suspected birthmarks were excluded (n = 485), the HR decreased to 2.81 but was still significant (95% CI: 1.11–7.13). The HR for the association between childhood cancer and birthmarks when individuals with follow-up times of <1 year were excluded and both definite and suspected birthmarks were included did not materially change the results (HR: 2.74; 95% CI: 1.07–7.02). Excluding individuals with follow-up times of <1 year resulted in an HR of 2.0 that was no longer significant (95% CI: 0.62–6.62) for definite birthmarks only. Adjustment for potential confounders (SES, birth weight, maternal race, gender, maternal education, maternal age, and gestational age) did not materially change the results.


View this table:
[in this window]
[in a new window]

  		TABLE 4 Cox Proportional Hazards Regression Modeling of the Association Between Birthmarks and Childhood Cancer in the CPP

 
Table 5 shows the types of birthmarks (definite and suspected) that were reported in individuals who received a diagnosed of cancer during the follow-up period. Three cases of acute leukemia had documented definite or suspected birthmarks (1 of each of the types: port-wine stain, strawberry hemangioma, and café-au-lait spots). Two patients with Wilms’ tumor were reported to have had a port-wine stain and a strawberry hemangioma, respectively. A port-wine stain also was reported for a retinoblastoma case, and a hairy pigmented nevi was reported for an individual who received a diagnosis of glioma.


View this table:
[in this window]
[in a new window]

  		TABLE 5 Types of Birthmarks Found in Individuals With Cancer

 
No particular type of birthmark occurred more frequently in children with cancer than without (P = .1208, Fisher's exact test). Because of the small number of cases with birthmarks, we were not able to determine whether any individual type of birthmark was more common in individual malignancies or to examine whether there was an interaction between birthmarks and other variables, such as gender. However, it is interesting to note that of the 7 individuals who had a birthmark documented either as definite or suspected during the first year, 6 were female (data not shown). In addition, no genetic syndromes that are known to be associated with childhood cancer20 were recorded in individuals with birthmarks during the follow-up time, including neurofibromatosis, which is linked to both leukemia and café-au-lait spots5 (data not shown). Of note, however, is that an umbilical hernia was reported in the patient who had Wilms’ tumor and also had a strawberry hemangioma (Table 5). Children with BWS have an increased frequency of umbilical hernias and an elevated risk for embryonal tumors.3,21 This child may have had unrecognized BWS, which was first described in 1963 and was not part of data collection.3


   DISCUSSION
TOP
 ABSTRACT
 METHODS
 RESULTS
 DISCUSSION
CONCLUSIONS
 REFERENCES
 
The results of this prospective study support those of previous studies9–11 that reported an association between childhood malignancy and birthmarks. Children who received a diagnosis of cancer in the CPP were ~3 times more likely to have a documented birthmark than those without this diagnosis.

To our knowledge, no other analyses have been conducted to investigate specifically the association between birthmarks and childhood cancer. As mentioned, 3 studies have reported a significantly increased prevalence of birthmarks in children with cancer compared with control subjects. A study conducted by Mertens et al10 examined the association between congenital abnormalities and childhood leukemia in 2117 cases of ALL and 605 cases of acute myelogenous leukemia (AML). Unexpectedly, more mothers of patients with ALL and AML than of control subjects reported their child's having large or multiple birthmarks (ALL odds ratio: 1.35 [95% CI: 1.35–1.71]; AML odds ratio 1.89 [95% CI: 1.15–3.11]). The authors noted that no patients were reported to have had NF1. The authors did not note which types of birthmarks were ascertained but did mention that they were defined as "any one larger than a quarter; or 6 or more about the size of a dime." In a study by Roganovic et al,9 a significantly increased frequency of pigmented nevi and café-au-lait spots was found in patients with childhood leukemia compared with control subjects. These authors also note that NF1 was not found in any of their study participants.9 Finally, a study of rhabdomyosarcoma was conducted to look at the association between parental use of marijuana and cocaine and rhabdomyosarcoma. It is interesting that birthmarks were reported to be significantly associated with rhabdomyosarcoma by the investigators, who were exploring potential confounders for inclusion in their primary analysis.11

Other than the known associations between certain types of birthmarks and cancer (eg, congenital pigmented nevi and melanoma22; café-au-lait spots in patients with NF1 and certain types of cancers, including AML5), there has been little research to provide a biological basis for the connection between the two. A small body of evidence suggests 2 possible links. First, insulin-like growth factor (IGF) system members IGF-1 and IGF-2 have been reported to be involved in the pathogenesis of infantile hemangiomas and also in Wilms’ tumor, ALL, and AML.23–26 It also is of interest to note that the IGF system has been implicated in the cause of several adult cancers, including breast, prostate, colon, and lung cancer.27 In addition, high circulating levels of IGF-1 have been proposed to be a possible explanation for the association between high birth weight and childhood and adult cancers.28 It remains to be determined whether certain types of birthmarks and cancer could stem from aberrations in the IGF system. Second, both solid tumors and vascular anomalies (eg, hemangiomas, port-wine stains) are known to have increased expression of angiogenic proteins, including vascular endothelial growth factor and basic fibroblast growth factor.29–31 In addition, limited evidence suggests that angiogenic proteins are involved in the pathogenesis of hematopoietic cancers.32,33 Whether angiogenic factors that are produced by vascular anomalies could act in a paracrine manner to promote tumor growth is unknown.

The major strength of this study was the rigorous and thorough prospective collection of data in contrast to previous studies that used retrospective designs9–11 and collected data through questionnaire-based methods.10,11 This analysis also had several limitations. First, we had low statistical power because of the small number of cases and the low prevalence of birthmarks in the CPP, making our results particularly sensitive to any misclassification of the exposure. However, any misclassification was likely nondifferential in that it did not depend on disease status; in this specialized case in which the exposure is binary, the resulting bias would favor the null.34 Our results were robust to any misclassification of birthmarks because exclusion of individuals with suspected birthmarks did not materially change the results. Second, we were not able to determine the temporal relationship between ascertainment of the birthmark and the diagnosis for cancers that occurred during the first year of life. Analyses that included cases that were diagnosed during the first year of life could have been subject to bias if cases were scrutinized more carefully for other abnormalities than noncases. To determine whether this type of bias occurred, we conducted analyses that excluded individuals with follow-up times of <1 year. Although the HRs decreased in magnitude, they were still indicative of an increased risk. Finally, because of the small number of childhood cancer cases, it is difficult to discount chance completely as an explanation for these results.


   CONCLUSIONS
TOP
 ABSTRACT
 METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
REFERENCES
 
We have reported an association between childhood cancer and birthmarks in the CPP. These results could offer new insight into the potential causes of childhood cancer and suggest a possibly fruitful avenue for additional exploration.


   ACKNOWLEDGMENTS
 
This work was supported by the Children's Cancer Research Fund.


   FOOTNOTES
 
Accepted Nov 8, 2006.

Address correspondence to Julie A. Ross, PhD, Department of Pediatrics, University of Minnesota, 420 Delaware St SE, MMC 422, Minneapolis, MN 55455. E-mail: ross{at}epi.umn.edu

The authors have indicated they have no financial relationships relevant to this article to disclose.


   REFERENCES
TOP
 ABSTRACT
 METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 

  1. Hasle H. Pattern of malignant disorders in individuals with Down's syndrome. Lancet Oncol. 2001;2 :429 –436[CrossRef][Medline]
  2. Ross JA, Spector LG, Robison LL, Olshan AF. Epidemiology of leukemia in children with Down syndrome. Pediatr Blood Cancer. 2005;44 :8 –12[CrossRef][ISI][Medline]
  3. Weksberg R, Shuman C, Smith AC. Beckwith-Wiedemann syndrome. Am J Med Genet C Semin Med Genet. 2005;137 :12 –23[Medline]
  4. Rump P, Zeegers MP, van Essen AJ. Tumor risk in Beckwith-Wiedemann syndrome: a review and meta-analysis. Am J Med Genet A. 2005;136 :95 –104[Medline]
  5. Lynch TM, Gutmann DH. Neurofibromatosis 1. Neurol Clin. 2002;20 :841 –865[CrossRef][ISI][Medline]
  6. Cichowski K, Jacks T. NF1 tumor suppressor gene function: narrowing the GAP. Cell. 2001;104 :593 –604[CrossRef][ISI][Medline]
  7. Mehes K, Kajtar P, Sandor G, Scheel-Walter M, Niethammer D. Excess of mild errors of morphogenesis in childhood lymphoblastic leukemia. Am J Med Genet. 1998;75 :22 –27[CrossRef][ISI][Medline]
  8. Mehes K, Signer E, Pluss HJ, Muller HJ, Stalder G. Increased prevalence of minor anomalies in childhood malignancy. Eur J Pediatr. 1985;144 :243 –254[CrossRef][ISI][Medline]
  9. Roganovic J, Radojcic-Badovinac A, Ahel V. Increased prevalence of minor anomalies in children with hematologic malignancies. Med Pediatr Oncol. 2002;38 :128 –130[CrossRef][ISI][Medline]
  10. Mertens AC, Wen W, Davies SM, et al. Congenital abnormalities in children with acute leukemia: a report from the Children's Cancer Group. J Pediatr. 1998;133 :617 –623[CrossRef][ISI][Medline]
  11. Grufferman S, Schwartz AG, Ruymann FB, Maurer HM. Parents’ use of cocaine and marijuana and increased risk of rhabdomyosarcoma in their children. Cancer Causes Control. 1993;4 :217 –224[ISI][Medline]
  12. Mehes K. Mild errors of morphogenesis in malignancy: macroscopic manifestation of genetic instability? Med Pediatr Oncol. 2000;34 :111 –112[CrossRef][ISI][Medline]
  13. Hardy JB. The Collaborative Perinatal Project: lessons and legacy. Ann Epidemiol. 2003;13 :303 –311[CrossRef][ISI][Medline]
  14. Klebanoff MA, Read JS, Mills JL, Shiono PH. The risk of childhood cancer after neonatal exposure to vitamin K. N Engl J Med. 1993;329 :905 –908[Abstract/Free Full Text]
  15. Paller A, Mancini AJ, Hurwitz S. Hurwitz's Clinical Pediatric Dermatology: A Textbook of Skin Disorders of Childhood and Adolescence. 3rd ed. Philadelphia, PA: Elsevier Saunders; 2006
  16. Allison PD, SAS Institute. Survival Analysis using the SAS System: A Practical Guide. Cary, NC: SAS Institute; 1995
  17. Linet MS, Wacholder S, Zahm SH. Interpreting epidemiologic research: lessons from studies of childhood cancer. Pediatrics. 2003;112 :218 –232[Free Full Text]
  18. Maldonado G, Greenland S. Simulation study of confounder-selection strategies. Am J Epidemiol. 1993;138 :923 –936[Abstract/Free Full Text]
  19. Ries LAG, Smith MA, Gurney JG, et al, eds. Cancer Incidence and Survival Among Children and Adolescents: United States SEER Program 1975–1995. Bethesda, MD: National Cancer Institute, SEER Program; 1999
  20. Stiller CA. Epidemiology and genetics of childhood cancer. Oncogene. 2004;23 :6429 –6444[CrossRef][ISI][Medline]
  21. DeBaun MR, Tucker MA. Risk of cancer during the first four years of life in children from the Beckwith-Wiedemann syndrome registry. J Pediatr. 1998;132 :398 –400[CrossRef][ISI][Medline]
  22. Tannous ZS, Mihm MC Jr, Sober AJ, Duncan LM. Congenital melanocytic nevi: clinical and histopathologic features, risk of melanoma, and clinical management. J Am Acad Dermatol. 2005;52 :197 –203[CrossRef][ISI][Medline]
  23. Ritter MR, Dorrell MI, Edmonds J, Friedlander SF, Friedlander M. Insulin-like growth factor 2 and potential regulators of hemangioma growth and involution identified by large-scale expression analysis. Proc Natl Acad Sci U S A. 2002;99 :7455 –7460[Abstract/Free Full Text]
  24. Rivera MN, Haber DA. Wilms’ tumour: connecting tumorigenesis and organ development in the kidney. Nat Rev Cancer. 2005;5 :699 –712[CrossRef][ISI][Medline]
  25. Vorwerk P, Wex H, Bessert C, Hohmann B, Schmidt U, Mittler U. Loss of imprinting of IGF-II gene in children with acute lymphoblastic leukemia. Leuk Res. 2003;27 :807 –812[CrossRef][ISI][Medline]
  26. Wu HK, Weksberg R, Minden MD, Squire JA. Loss of imprinting of human insulin-like growth factor II gene, IGF2, in acute myeloid leukemia. Biochem Biophys Res Commun. 1997;231 :466 –472[CrossRef][ISI][Medline]
  27. LeRoith D, Roberts CT Jr. The insulin-like growth factor system and cancer. Cancer Lett. 2003;195 :127 –137[ISI][Medline]
  28. Ross JA. High birthweight and cancer: evidence and implications. Cancer Epidemiol Biomarkers Prev. 2006;15 :1 –2[Free Full Text]
  29. Tonini T, Rossi F, Claudio PP. Molecular basis of angiogenesis and cancer. Oncogene. 2003;22 :6549 –6556[CrossRef][ISI][Medline]
  30. Cohen MM Jr. Vascular update: morphogenesis, tumors, malformations, and molecular dimensions. Am J Med Genet A. 2006;140 :2013 –2038[Medline]
  31. Marler JJ, Fishman SJ, Kilroy SM, et al. Increased expression of urinary matrix metalloproteinases parallels the extent and activity of vascular anomalies. Pediatrics. 2005;116 :38 –45[Abstract/Free Full Text]
  32. Keyhani A, Jendiroba DB, Freireich EJ. Angiogenesis and leukemia. Leuk Res. 2001;25 :639 –645[CrossRef][ISI][Medline]
  33. Ribatti D, Scavelli C, Roccaro AM, Crivellato E, Nico B, Vacca A. Hematopoietic cancer and angiogenesis. Stem Cells Dev. 2004;13 :484 –495[CrossRef][ISI][Medline]
  34. Rothman KJ, Greenland S. Modern Epidemiology. 2nd ed. Philadelphia, PA: Lippincott-Raven Publishers; 1998
PEDIATRICS (ISSN 1098-4275). ©2007 by the American Academy of Pediatrics



This article has been cited by other articles:


Home page
 Cancer Epidemiol. Biomarkers Prev.Home page
T. Bjorge, S. Cnattingius, R. T. Lie, S. Tretli, and A. Engeland
Cancer Risk in Children with Birth Defects and in Their Families: A Population Based Cohort Study of 5.2 Million Children from Norway and Sweden
Cancer Epidemiol. Biomarkers Prev., March 1, 2008; 17(3): 500 - 506.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow P3Rs: Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when P3Rs are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow E-mail this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My File Cabinet
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via CrossRef
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Johnson, K. J.
Right arrow Articles by Ross, J. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Johnson, K. J.
Right arrow Articles by Ross, J. A.
Related Collections
Right arrow Tumors