Skip to content
Licensed Unlicensed Requires Authentication Published by De Gruyter February 28, 2015

Neural modelling of growth hormone therapy for the prediction of therapy results

  • Urszula Smyczyńska EMAIL logo , Joanna Smyczyńska and Ryszard Tadeusiewicz

Abstract

In this paper, we presented the problem of predicting response to recombinant human growth hormone (GH) treatment in GH-deficient children. Such a prediction can be done by techniques of mathematical modelling and is important because the therapy consists of daily injections and is expensive; thus, it should be administered only to those patients who will, with high probability, benefit from it. Until now, the leading methodological approach to this problem was multiple regression analysis. Several authors demonstrated that it is possible to derive useful models by this method; however, it has some obvious limitations that can be avoided with the use of the proposed neural network approach.


Corresponding author: Urszula Smyczyńska, Department of Automatics and Biomedical Engineering, AGH University of Science and Technology, Cracow 30-059, Poland, E-mail:

  1. Author contributions: All the authors have accepted responsibility for the entire content of this submitted manuscript and approved submission.

  2. Research funding: None declared.

  3. Employment or leadership: None declared.

  4. Honorarium: None declared.

  5. Competing interests: The funding organization(s) played no role in the study design; in the collection, analysis, and interpretation of data; in the writing of the report; or in the decision to submit the report for publication.

Appendix

Human growth and GH treatment

Human growth is regulated and influenced by many different factors, among them genetic, hormonal, and environmental ones. The growth of a child is genetically determined and it is possible to calculate the TH of a child, that is, to estimate what should be his/her FH after the completion of linear growth, based on the heights of parents. Different unfavorable factors, such as malnutrition, social deprivation, and chronic diseases, including hormonal disorders and genetic defects, may decrease growth rate and lead to short stature.

The main hormone regulating linear growth during childhood is somatotropin (GH), synthesized in the anterior lobe of pituitary gland and released to the bloodstream in a pulsatile manner. The main regulators of GH secretion are two hypothalamic hormones: somatoliberin (GH-releasing hormone) that stimulates GH secretion and somatostatin that inhibits this process. Under physiological conditions, GH secretion is stimulated by such factors as sleep, exercise, acute stress, fasting, and hypoglycemia but reduced in terms of chronic stress or hyperglycemia. Moreover, numerous pharmacological substances may either stimulate or inhibit GH secretion in healthy subjects.

The main peripheral mediator of GH action is IGF-I. The main pool of IGF-I is produced in the liver and secreted into the bloodstream, but IGF-I is also synthesized locally in many tissues, including growth plates. The effect of GH on growth plates occurs mainly through IGF-I as well as directly by binding with GH receptors on chondrocytes. Other important hormones are thyroxin (hypothyroidism decreases GH secretion and action as well as IGF-I synthesis by several mechanisms) and sex steroids (responsible for the increase of GH secretion during puberty, leading to the pubertal growth spurt).

The Polish criteria of qualification GH-deficient children to rhGH treatment are as follows ([1]):

  1. Patient height below the third centile for age and gender according to current reference data (centile charts) for a given population; in Poland, the most commonly used are centile charts of Palczewska and Niedzwiecka ([2]);

  2. Decreased growth rate (HV) during at least 6 months of observation;

  3. Delayed bone maturation (BA), assessed on the ground of the radiogram of left (nonwriting) hand and wrist, according to Greulich-Pyle standards [3];

  4. Decreased GH peak <10.0 ng/mL in two stimulation tests (at present, also decreased GH peak after falling asleep);

  5. Excluded other causes of impaired growth (chronic diseases, malignancy, malnutrition, and genetic syndromes, especially Turner syndrome in girls);

  6. Other conditions, such as hypoglycemia in neonates and infants, organic abnormalities of the hypothalamic-pituitary region, observed in some patients.

The assessment of IGF-I secretion before rhGH therapy administration is an obligatory procedure, but IGF-I deficiency is not the criterion that must be met for inclusion of treatment. Serum concentrations of both GH and IGF-I may be measured by different methods, of which the most frequently used are the chemiluminescent enzyme immunometric assays.

Although the criteria of qualification to rhGH treatment are strictly defined, it still happens that, in some patients, the efficacy of treatment is not fully satisfactory. Moreover, there are some doubts concerning certain parts of diagnostic process. For example, due to the pulsatile rhythm of GH release and variability of its concentration, the only approved methods of assessment of GH secretion are stimulation tests. Unfortunately, these tests have relatively low sensitivity and specificity, so it is recommended to perform at least two tests in each patient. Besides, it has been reported that at least 34 different provocative tests with 189 combinations are used, whereas the cutoff value was arbitrarily established on the same level (10 ng/mL) for all test variants, with no differentiation made to account for the strength of particular stimuli. The most commonly used pharmacological agents are insulin, glucagon, clonidine, arginine, and L-DOPA. In Poland, a few years ago, an assessment of GH peak during 3 h after falling asleep was introduced as an obligatory screening procedure in diagnosing GHD in children; however, it was previously reported that even overnight assessment of GH release may not identify all patients with GHD [4, 5]. Finally, there are disorders that are potentially treatable with rhGH, in which GH peak in stimulation tests is normal, but IGF-I secretion is decreased. They are as follows:

  • Neurosecretory dysfunction, defined as decreased spontaneous GH secretion despite normal response to pharmacological stimulation (diagnosed on the ground of decreased spontaneous GH secretion after falling asleep but normal in stimulation tests) and

  • GH bioinactivity, which is the secretion of GH molecule of decreased bioactivity (unable to bind with GH receptor and stimulate IGF-I synthesis) but of normal immunoreactivity (detectable by immunometric assays).

Appendix

  1. Romer TE, Walczak M, Wisniewski A, Roszkowska-Blaim M, Korman E, Graliński JS, et al. Children with growth disorders qualified in Poland for growth hormone therapy. Endokrynol Diabetol Chor Przemiany Materii Wieku Rozw 2001;9:41–54.

  2. Palczewska I, Niedzwiecka Z. Somatic development indices in children and youth of Warsaw. Med Wieku Rozwoj 2001;5:18–118.

  3. Greulich WW, Pyle SI. Radiographic atlas of skeletal development of the hand and wrist, 2nd ed. Stanford University Press, 1993.

  4. Rosenfeld RG, Albertsson-Wikland K, Cassorla F, Frasier SD, Hasegawa Y, Hintz RL, et al. Diagnostic controversy – the diagnosis of childhood growth-hormone deficiency revisited. J Clin Endocrinol Metab 1995;80:1532–40.

  5. Webb EA, Dattani MT. Diagnosis of growth hormone deficiency. Endocr Dev 2010;18:55–66.

References

1. Wei C, Gregory JW. Physiology of normal growth. Paediatr Child Health 2009;19:236–40.10.1016/j.paed.2009.02.007Search in Google Scholar

2. Wit JM, Ranke MB, Albertsson-Wikland K, Carrascosa A, Rosenfeld RG, Van Buuren S, et al. Personalized approach to growth hormone treatment: clinical use of growth prediction models. Horm Res Paediatr 2013;79:257–70.10.1159/000351025Search in Google Scholar PubMed

3. Ranke MB, Lindberg A. Predicting growth in response to growth hormone treatment. Growth Horm IGF Res 2009;19:1–11.10.1016/j.ghir.2008.08.001Search in Google Scholar PubMed

4. Ranke MB, Lindberg A, Chatelain P, Wilton P, Cutfield W, Albertsson-Wikland K, et al. Derivation and validation of a mathematical model for predicting the response to exogenous recombinant human growth hormone (GH) in prepubertal children with idiopathic GH deficiency. J Clin Endocrinol Metab 1999;84:1174–83.10.1210/jcem.84.4.5634Search in Google Scholar PubMed

5. Wikland KA, Kristrom B, Rosberg S, Svensson B, Nierop AF. Validated multivariate models predicting the growth response to GH treatment in individual short children with a broad range in GH secretion capacities. Pediatr Res 2000;48:475–84.10.1203/00006450-200010000-00010Search in Google Scholar PubMed

6. Schonau E, Westermann F, Rauch F, Stabrey A, Wassmer G, Keller E, et al. A new and accurate prediction model for growth response to growth hormone treatment in children with growth hormone deficiency. Eur J Endocrinol 2001;144:13–20.10.1530/eje.0.1440013Search in Google Scholar PubMed

7. Carel JC, Ecosse E, Nicolino M, Tauber M, Leger J, Cabrol S, et al. Adult height after long term treatment with recombinant growth hormone for idiopathic isolated growth hormone deficiency: observational follow up study of the French population based registry. Br Med J 2002;325:70.10.1136/bmj.325.7355.70Search in Google Scholar PubMed PubMed Central

8. Ranke MB, Lindberg A, Martin DD, Bakker B, Wilton P, Albertsson-Wikland K, et al. The mathematical model for total pubertal growth in idiopathic growth hormone (GH) deficiency suggests a moderate role of GH dose. J Clin Endocrinol Metab 2003;88:4748–53.10.1210/jc.2003-030600Search in Google Scholar PubMed

9. de Ridder MA, Stijnen T, Hokken-Koelega AC. Prediction of adult height in growth-hormone-treated children with growth hormone deficiency. J Clin Endocrinol Metab 2007;92:925–31.10.1210/jc.2006-1259Search in Google Scholar PubMed

10. Dahlgren J, Kristrom B, Niklasson A, Nierop AF, Rosberg S, Albertsson-Wikland K. Models predicting the growth response to growth hormone treatment in short children independent of GH status, birth size and gestational age. BMC Med Inform Decis Making 2007;7:40.10.1186/1472-6947-7-40Search in Google Scholar PubMed PubMed Central

11. Lee PA, Germak J, Gut R, Khutoryansky N, Ross J. Identification of factors associated with good response to growth hormone therapy in children with short stature: results from the ANSWER Program. Int J Pediatr Endocrinol 2011;2011:6.10.1186/1687-9856-2011-6Search in Google Scholar

12. Swiac A, Bilski J. Metoda wstecznej propagacji błędów i jej modyfikacje. In: Duch W, Korbicz J, Rutkowski L, Tadeusiewicz R, editors. Sieci neuronowe. Biocybernetyka i inżynieria biomedyczna 2000, 1st ed. Warszawa: Akademicka Oficyna Wydawnicza Exit, 2000.Search in Google Scholar

13. Rumelhart DE, Hinton GE, Williams RJ. Learning internal representations by error propagation. In: Rumelhart DE, McClelland JL, Group CP, editors. Parallel distributed processing: explorations in the microstructure of cognition. Cambridge, USA: MIT Press, 1986:318–62.Search in Google Scholar

14. Dreyfus G, editor. Neural networks: methodology and applications. Berlin, Heidelberg: Springer, 2005.Search in Google Scholar

15. Tadeusiewicz R, Chaki R, Chaki N. Exploring neural networks with C#. Boca Raton: CRC Press, Taylor & Francis Group, 2014:298.Search in Google Scholar

16. Prank K, Kloppstech M, Brabant G. Neural networks in the analysis of episodic growth hormone release. Hum Reprod Update 1997;3:215–34.10.1093/humupd/3.3.215Search in Google Scholar

17. Temurtas F. A comparative study on thyroid disease diagnosis using neural networks. Expert Syst Appl 2009;36:944–9.10.1016/j.eswa.2007.10.010Search in Google Scholar

18. Omiotek Z, Burda A, Wojcik W. The use of decision tree induction and artificial neural networks for automatic diagnosis of Hashimoto’s disease. Expert Syst Appl 2013;40:6684–9.10.1016/j.eswa.2013.03.022Search in Google Scholar

19. Mantzaris DH, Anastassopoulos GC, Lymberopoulos DK. Medical disease prediction using artificial neural networks. 8th IEEE International Conference on Bioinformatics and Bioengineering, vols 1 and 2, 2008:793–8.10.1109/BIBE.2008.4696782Search in Google Scholar

20. Tomida S, Hanai T, Koma N, Suzuki Y, Kobayashi T, Honda H. Artificial neural network predictive model for allergic disease using single nucleotide polymorphisms data. J Biosci Bioeng 2002;93:470–8.10.1016/S1389-1723(02)80094-9Search in Google Scholar

21. Klencki M, Slowinska-Klencka D, Lewinski A. Multifarious system for quantitative analysis of histologic compartments. Comput Biomed Res 1997;30:165–9.10.1006/cbmr.1997.1443Search in Google Scholar PubMed

22. Palczewska I, Niedzwiecka Z. Somatic development indices in children and youth of Warsaw. Med Wieku Rozwoj 2001;5:18–118.Search in Google Scholar

23. Smyczynska J, Lewinski A, Hilczer M. Wskaźniki auksologiczne przydatne w diagnostyce dzieci z niedoborem wzrostu i monitorowaniu skuteczności ich leczenia. Endokrynol Pediatr 2013;12:51–6.Search in Google Scholar

24. Blum WR, Schweitzer R. Insulin-like growth factors and their binding globulins. In: Ranke MB, editor. Diagnostic of endocrine function in children and adolescents, 3rd revised and extended edition. Basel: Karger, 2003:166–99.Search in Google Scholar

Received: 2014-12-2
Accepted: 2014-12-22
Published Online: 2015-2-28
Published in Print: 2015-3-31

©2015 by De Gruyter

Downloaded on 26.4.2024 from https://www.degruyter.com/document/doi/10.1515/bams-2014-0021/html
Scroll to top button