Descriptive epidemiology and survival analysis of acromegaly in Korea

DOI: https://doi.org/10.21203/rs.3.rs-197371/v1

Abstract

Acromegaly is a rare, slowly progressive disease. Its mechanism is not fully understood, and epidemiological research on Korean patients with acromegaly is scarce. To determine the incidence and prevalence of acromegaly and assess the comorbidities and survival benefits based on treatment options. This nationwide population-based cohort study was conducted using data of the Korean Health Insurance Review and Assessment claims database to evaluate the incidence of newly diagnosed acromegaly cases during 2013–2017. During the 5-year period, 1,093 patients were newly diagnosed with acromegaly. The average annual incidence was 4.2 cases per million per year, and the prevalence was 32.1 cases per million during this period. The incidence of hypertension was low after medical treatment (hazard ratio 0.257, 95% confidence interval: 0.082–0.808, P = 0.0201), but the incidence of diabetes showed no significant difference across treatment modalities. Over a period of 6 years since diagnosis, we found that patients treated for acromegaly had a significantly higher survival rate than those untreated (P = 0.0003). The annual incidence rate of Korean patients with acromegaly was similar to that reported in previous studies. Using nationwide population data, our study emphasized the importance of treatment in acromegaly patients.

Introduction

Acromegaly is a chronic disease resulting from secretion of excess growth hormone (GH) and insulin-like growth factor (IGF-1) by a GH-producing pituitary adenoma, leading to bone and soft tissue overgrowth 1. As postoperative GH level is inversely associated with mortality 2, the purpose of treatment in these patients is to normalize GH and IGF-1 levels for symptom relief and to reduce mortality 3,4. Surgery through the trans-sphenoidal approach is generally the treatment of choice, and a first-generation long-acting somatostatin receptor ligand (SRl) may be used as the primary medical treatment for inoperable or incompletely removed cases. In case of primary treatment failure, second-generation SRIs such as pasireotide or GH receptor antagonists such as pegvisomant may be indicated as second-line therapy. Sterotactic radiosurgery may also be considered when the clinical response is insufficient 5.

Acromegaly is rare, with a reported prevalence of 2.8–13.7 cases per million and an annual incidence rate of 0.2–1.1 cases per million per year 6. In a mortality outcome study on acromegaly, cardiovascular disease was one of the leading causes of death (60%), followed by respiratory disease (25%) and malignancy (15%) 7. A significantly higher mortality rate was also observed when accompanied with diabetes.8 It is known that the survival rates of patients with well-controlled GH or IGF-1 after treatment are similar to those of the general population 9,10. However, large-scale epidemiological research on the survival and complication rates of acromegaly patients in Korea is scarce due to its rarity.

Since 1989, the nationwide health insurance system has been implemented in Korea, and a variety of medical information necessary to process the insurance claims is being securely stored and managed by the Health Insurance Review and Assessment (HIRA) database. The data are open to researchers with the aim of contributing to the development of healthcare and medical knowledge under the supervision or consignment of the National Health Insurance Act and other statutes 11. This database has enabled many researchers to perform large-scaled clinical or epidemiological data analyses in Korea. Hence, our study was designed to utilize the HIRA claims dataset to determine the incidence and prevalence of acromegaly in Korea and to assess the comorbidities and survival benefits of the treatment modalities.

Materials And Methods

Data collection

This was a nationwide population-based cohort study based on the HIRA claims dataset. According to the insurance system in Korea, the claims data with the appropriate diagnostic codes are submitted by service providers to the HIRA for reimbursement afterward. As mentioned previsouly, this claims data are released to investigators after deidentification for large-scaled analysis in Korea. It is of note that those diagnostic codes often fail to reflect the actual clinical conditions of patients due to diagnostic discrepancies or disease input errors 12. Because the HIRA database is deidentificated which does not allow investigators to retrieve a patients’ individual laboratory test results, it is necessary to screen and obtain the group of patients satisfying the researcher’s purpose of analysis within the dataset by applying multiple variables such as diagnostic codes, laboratory test and procedure codes, and drug prescriptions and so on. This combination of variables in the data is called the “operational definition” 13. The reliability of data used is determined by how appropriately the operational definition identifies the actual target patient and extracts the proper data.

To apply more accurate operational definitions in our study, we crossmatched the benefit extension policy (BEP) application codes with the HIRA codes. The BEP in Korea was established to support medical expenses of patients with rare and incurable diseases such as cancers and rare genetic disorders accompanying high economic burden 14. The BEP code is assigned to diseases satisfying criteria of rarity and necessity to support the cost during diagnosis and treatment, and is mandatory to exempt certain portion of medical expenses needed to manage these diseases. Double comparison with the HIRA dataset and the BEP codes can assure the reliability of data extracted in our study.

This study was conducted in accordance with the guidelines laid down in the Declaration of Helsinki. Kyung Hee University Hospital Institutional Review Board (KHUH 2018-11-001) approved this study and also decided to waive the informed consent from the participants according to the retrospective design of this study.

Prevalence and annual incidence rate

Acromegaly was defined as a case in a patient who had a history of outpatient care or hospitalization based on both the International Classification of Diseases (ICD), the 10th Revision code (E22.0), and the BEP code (V112). We analyzed the prevalence and annual incidence of acromegaly during 2013–2017 with a washout period between 2009 and 2012 to calculate the incidence rate, which showed 1,093 patients with newly diagnosed cases of acromegaly.

Treatment effect

An effective treatment has been reported to normalize GH secretion and prevent metabolic complications in patients with acromegaly 3. Because the HIRA database includes information necessary to process the incurance claims and does not provide detailed data regarding clinical parameters such as individual laboratory test results or radiologic examinations, it is difficult to directly evaluate the treatment effects. To solve this limitation, we investigated the incidence of metabolic complications such as diabetes and hypertension after diagnosed as acromegaly between those with diagnosed but untreated acromegaly (defined as no treatment record after diagnosis) and those with treated acromegaly, which would enable us to indirectly observe the treatment effects in patients with acromegaly.

Among those newly diagnosed with acromegaly during 2011–2012 (N = 377), data of those with either hypertension (I10) or diabetes mellitus (E10–14) newly diagnosed during the follow-up period (2013–2017) were additionally extracted (hypertension N = 147; diabetes mellitus N = 109). We compared their incidence rates between those diagnosed and treated for acromegaly (treatment group) and those diagnosed but not treated for acromegaly (non-treatment group). In addition, after sub-dividing the treatment group into (1) medical, (2) surgical, and (3) medical and surgical treatment, each sub-group was compared with the non-treatment group for the incidence of hypertension and diabetes. This grouping was based on whether appropriate codes for either medical or surgical treatment was found during the follow-up period. This comparison was based on the notion that those with diagnosed and properly treated acromegaly would be less likely to experience metabolic complications when compared with non-treated patients (Supplementary Fig. 1).

Mortality

The mortality between those diagnosed but untreated acromegaly and treated acromegaly was compared by using the identical dataset used for analyzing the incidence of hypertension and diabetes. For 377 patients with newly diagnosed acromegaly during 2011–2012, their mortality rates during 2013–2017 were compared between the treatment and non-treatment groups (Supplementary Fig. 2).

Statistical analysis

The hazard ratio (HR) and 95% confidence interval (CI) for hypertension and diabetes incidence and for mortality were estimated using the log-rank test and Cox's proportional hazard regression analysis. The multivariate model analysis was performed after adjusting for gender, age, and income level. The survival period was set between the date when acromegaly was first treated as the start date of observation, to either the date of a new clinical event, or on December 31, 2017, as the end date of observation. Patients who had been treated before the initial diagnosis of acromegaly were excluded from the analysis. The result was considered statistically significant if P < 0.05. Statistical analyses were performed using SAS software, version 9.4 (SAS Institute, North Carolina, USA).

Results

Epidemiologic data

A total of 1,093 patients (M:F = 594:497) were newly diagnosed with acromegaly between January 2013 and December 2017 (Table 1). More than 200 people were diagnosed each year, and 195 were newly diagnosed in 2017 (Fig. 1a). The average annual incidence rate was 4.2 cases per million, and the prevalence rate increased in 2014 before reaching 32 cases per million (Fig. 1b).

Development of medical comorbidity according to treatment modality

It was found that the risk of diabetes (P = 0.0384) and hypertension (P = 0.0249) was significantly lower in the treatment group than in the non-treatment group. Further analyses after subdividing the treatment group according to the treatment modality demonstrated that the incidence of diabetes in each group showed no significant difference when compared to the non-treatment group. However, the risk of hypertension in the medically treated group was significantly lower than that in the non-treatment group (HR 0.259, 95% CI: 0.082–0.814, P = 0.0208) (Table 2a) and remained significantly lower after adjustment for confounders (HR 0.257, 95% CI: 0.082–0.808, P = 0.0201) (Table 2b).

Survival analysis of acromegaly patients

According to the survival analysis conducted on patients newly diagnosed with acromegaly during 2011–2012, patients in the non-treatment group showed a significantly lower probability of survival (curve difference at P = 0.0003 by log-rank) than those in the non-treatment group (Fig. 2), which remained significant after adjustment for confounders (HR 3.668, 95% CI 1.644–8.183, P = 0.0015)

Discussion

This study was designed to investigate the annual incidence and prevalence of acromegaly, which further compared the occurrence of complications by treatment modality in Koreans from January 2013 to December 2017. Survival rate analysis was also conducted between treatment and non-treatment groups. From 2013 to 2017, the annual incidence of acromegaly in Korea was 4.2 cases per million per year, not much different from the 3.57 cases per million reported in a previous study conducted during 2010–2013 15. This was also similar to the annual occurrence rate of 2–11 cases per million people per year from a systematical review 6. There were no significant differences in the development of diabetes accrding to treatment modalities. However, hypertension showed significantly lower incidence in the medical treatment group. Over a period of 6 years since diagnosis, patients treated for acromegaly had a significantly higher survival rate than the untreated patients.

A cardiovascular disease is one of the main causes of death for acromegaly 7, and it is known that the prevalence of diabetes and hypertension is higher than that in the general population 16,17, but no studies have analyzed the occurrence of comorbidities in acromegaly patients in Korea. This study is the first report to evaluate the risk of diabetes and hypertension according to treatment modality in Korean patients with acromegaly. Patients with acromegaly are more likely to have concurrent hypertension; 1.9 times higher than that in the general population. This could be due to the direct anti-natriuretic effects of exess GH 18. Several animal studies have suggested that increased GH causes overactivation of the renin-angiotensin-aldosterone-system 19. Hyperinsulinemia also causes hyperactivity of the renin-angiotensin-aldosterone system, which promotes renal sodium reabsorption and thus increases plasma volume 20. It is also known that elevated insulin and GH levels contribute to hypertension by activation of the sympathetic nervous system 21. To indirectly estimate the therapeutic effects of acromegaly, we divided the acromegaly-treated patients into three groups: 1) medical, 2) surgical, and 3) medical plus surgical treatment group. The incidence of diabetes in the treatment group showed no significant difference when compared to the non-therapeutic group. We also investigated the incidence rate of hypertension after newly diagnosed acromegaly. This was similar to a previous study that showed significantly reduced development of hypertension after five years of medical treatment 22.

As for diabetes, there was no significant difference between the treated and non-treated groups. This was unexpected and inconsistent with previous studies showing the improvement of glucose metabolism regardless of the outcome of surgical treatment 23. The study conducted by Kinoshita et al. on patients with acromegaly in Japan demonstrated that glucose metabolism did not normalize even after acromegaly treatment due to impaired beta cell function. In addition, Shekhawat et al. showed no significant difference between pre- and postoperative beta cell function in acromegaly patients with diabetes, and this was due to increased glucose-dependent insulinotropic polypeptide resistance and reduced beta cell function due to hyperglucagonemia. Similar to Japanese, Koreans are known to have reduced insulin secretion and compensatory insulin responses before the onset of diabetes 24; thus, decreased beta-cell function may fail to recover and lead to diabetes development even after treatment of acromegaly.

In our study, the treatment group had significant survival benefits compared to the non-treatment group. A previously reported survival analysis on acromegaly patients in New Zealand showed that the life expectancy was the same as that of the general population when the GH level was less than 1 g/L after treatment9. A Finnish study, instead, showed that the life expectancies were the same as that of the general population if the GH levels remained below 2.5 g/L after treatment10. Along with these results, our analysis confirmes and emphazises the importance of effective treatment to normalize GH secretion and prevent metabolic complications in patients with acromegaly, which could further assure the survival benefit. To our best knowledge, this is the first study to report the results of survival analysis in Korean patients with acromegaly.

In this study, patients with acromegaly were defined according to the operational definition based on both the claims data codes and the BEP codes in the HIRA database. This operational definition was considered to be appropriate because the incidence and prevalence of acromegaly were observed to be similar to those of previous studies. However, due to the inevitable nature of health insurance claims data, some disease codes of patients with acromegaly might have been omitted, which could have led to an underestimation of data. Because the HIRA database did not include individual test results such as IGF-1 or GH levels, the degree of severity could not be ascertained. To compare the complications among treatment modalities, the incidence of hypertension and diabetes was analyzed for five years in patients newly diagnosed with acromegaly during 2011–2012; however, this might not have been a sufficient period for the development of complications.

Conclusion

The annual incidence of acromegaly in Korea was similar to previously reported data. The incidence of diabetes did not significantly differ across treatment modalities, but that of hypertension was significantly lower after medical treatment. The treatment group showed significant survival benefits compared to the non-treatment group. Based on this study, we plan to expand our analysis to compare the cardiovascular outcomes between patients with acromegaly and the general population.

Declarations

Acknowledgments

All of the participating authors would like to thank the Committee of Research of the Korean Endocrine Society for provinding great support for this analysis. The analysis was performed under the permission of data utilization by the National Health Insurance Service in Korea (NHIS-2019-1-137).

Competing interests

The authors declare no competing interests.

Ethics approval

This study was conducted in accordance with the guidelines laid down in the Declaration of Helsinki and was approved by Kyung Hee University Hospital Institutional Review Board (KHUH 2018-11-001).

Consent to participate

Informed consent was waived due to the retrospective design of this study, based on the decision by Kyung Hee University Hospital Institutional Review Board (KHUH 2018-11-001).

Consent for publication

A declaration of non-objection was obtained.

References

  1. Melmed, S. Acromegaly pathogenesis and treatment. J Clin Invest. 119, 3189–3202 https://doi.org/10.1172/JCI39375 (2009).
  2. Melmed, S. Acromegaly and cancer: not a problem? J Clin Endocrinol Metab. 86, 2929–2934 https://doi.org/10.1210/jcem.86.7.7635 (2001).
  3. Holdaway, I. M., Bolland, M. J. & Gamble, G. D. A meta-analysis of the effect of lowering serum levels of GH and IGF-I on mortality in acromegaly. Eur J Endocrinol. 159, 89–95 https://doi.org/10.1530/EJE-08-0267 (2008).
  4. Dekkers, O. M., Biermasz, N. R., Pereira, A. M., Romijn, J. A. & Vandenbroucke, J. P. Mortality in Acromegaly: A Metaanalysis. The Journal of Clinical Endocrinology & Metabolism. 93, 61–67 https://doi.org/10.1210/jc.2007-1191 (2008).
  5. Melmed, S. et al. A Consensus Statement on acromegaly therapeutic outcomes. Nat Rev Endocrinol. 14, 552–561 https://doi.org/10.1038/s41574-018-0058-5 (2018).
  6. Lavrentaki, A., Paluzzi, A., Wass, J. A. & Karavitaki, N. Epidemiology of acromegaly: review of population studies. Pituitary. 20, 4–9 https://doi.org/10.1007/s11102-016-0754-x (2017).
  7. Colao, A., Ferone, D., Marzullo, P. & Lombardi, G. Systemic complications of acromegaly: epidemiology, pathogenesis, and management. Endocr Rev. 25, 102–152 https://doi.org/10.1210/er.2002-0022 (2004).
  8. Wen-Ko, C. et al. The impact of diabetes mellitus on the survival of patients with acromegaly. Endokrynol Pol. 67, 501–506 https://doi.org/10.5603/EP.a2016.0031 (2016).
  9. Holdaway, I. M., Rajasoorya, R. C. & Gamble, G. D. Factors influencing mortality in acromegaly. J Clin Endocrinol Metab. 89, 667–674 https://doi.org/10.1210/jc.2003-031199 (2004).
  10. Kauppinen-Mäkelin, R. et al. A nationwide survey of mortality in acromegaly. The Journal of Clinical Endocrinology & Metabolism. 90, 4081–4086 (2005).
  11. Cheol Seong, S. et al. Data Resource Profile: The National Health Information Database of the National Health Insurance Service in South Korea. Int J Epidemiol. 46, 799–800 https://doi.org/10.1093/ije/dyw253 (2017).
  12. Bae, S. O. & Kang, G. W. A Comparative Study of the Disease Codes between Korean National Health Insurance Claims and Korean National Hospital Discharge In-Depth Injury Survey. Health Policy and Management. 24, 322–329 https://doi.org/10.4332/kjhpa.2014.24.4.322 (2014).
  13. Bridgman, P. W., Bridgman, P. W., Bridgman, P. W., Bridgman, P. W. & Physicien, E. U. The logic of modern physics. Vol. 3(Macmillan New York, 1927).
  14. National Health Insurance System of Korea. (2015).
  15. Park, K. H., Lee, E. J., Seo, G. H. & Ku, C. R. Risk for Acromegaly-related Comorbidities by Sex in Korean Acromegaly. J Clin Endocrinol Metab. 105, https://doi.org/10.1210/clinem/dgz317 (2020).
  16. Vitale, G. et al. Hypertension in acromegaly and in the normal population: prevalence and determinants. Clin Endocrinol. 63, 470–476 (2005).
  17. Ferrau, F., Albani, A., Ciresi, A., Giordano, C. & Cannavo, S. Diabetes Secondary to Acromegaly: Physiopathology, Clinical Features and Effects of Treatment. Front Endocrinol (Lausanne). 9, 358 https://doi.org/10.3389/fendo.2018.00358 (2018).
  18. Kamenicky, P. et al. Epithelial sodium channel is a key mediator of growth hormone-induced sodium retention in acromegaly. Endocrinology. 149, 3294–3305 https://doi.org/10.1210/en.2008-0143 (2008).
  19. Bielohuby, M. et al. Chronic growth hormone excess is associated with increased aldosterone: a study in patients with acromegaly and in growth hormone transgenic mice. Exp Biol Med (Maywood). 234, 1002–1009 https://doi.org/10.3181/0901-RM-34 (2009).
  20. Powlson, A. S. & Gurnell, M. Cardiovascular Disease and Sleep-Disordered Breathing in Acromegaly. Neuroendocrinology. 103, 75–85 https://doi.org/10.1159/000438903 (2016).
  21. Bondanelli, M., Ambrosio, M. R. & Uberti, E. C. degli Pathogenesis and prevalence of hypertension in acromegaly. Pituitary 4, 239–249, doi:10.1023/a:1020798430884 (2001).
  22. Colao, A., Auriemma, R. S., Galdiero, M., Lombardi, G. & Pivonello, R. Effects of Initial Therapy for Five Years with Somatostatin Analogs for Acromegaly on Growth Hormone and Insulin-Like Growth Factor-I Levels, Tumor Shrinkage, and Cardiovascular Disease: A Prospective Study. 94,3746–3756, doi:10.1210/jc.2009-0941 (2009).
  23. Helseth, R. et al. Preoperative octreotide therapy and surgery in acromegaly: associations between glucose homeostasis and treatment response. Endocrine. 51, 298–307 https://doi.org/10.1007/s12020-015-0679-6 (2016).
  24. Kim, D. J., Lee, M. S., Kim, K. W. & Lee, M. K. Insulin secretory dysfunction and insulin resistance in the pathogenesis of korean type 2 diabetes mellitus. Metabolism. 50, 590–593 https://doi.org/10.1053/meta.2001.22558 (2001).

Tables

Table 1. Incidence of acromegaly in Korea between 2013-2017

Year

2013

2014

2015

2016

2017

Gender (n, %)

 

 

 

 

 

Male

105(45.5)

129(51.4)

83(41.1)

96(44.9)

84(43.1)

Female

125(54.1)

122(48.6)

119(58.9)

118(55.1)

110(56.4)

Unknown

1

 

 

 

1

Total (n)

231

251

202

214

195

Age (n, %)

 

 

 

 

 

0-19

6(2.6)

7(2.8)

7(3.5)

6(2.8)

6(3.1)

20-29

15(6.5)

17(6.8)

21(10.4)

13(6.1)

18(9.2)

30-39

35(15.2)

49(19.5)

33(16.3)

38(17.8)

24(12.3)

40-49

50(21.6)

61(24.3)

44(21.8)

50(23.4)

41(21)

50-59

77(33.3)

67(26.7)

52(25.7)

63(29.4)

57(29.2)

60-69

31(13.4)

35(13.9)

35(17.3)

26(12.1)

35(17.9)

70-79

15(6.5)

12(4.8)

8(4)

15(7)

12(6.2)

>79

1(0.4)

3(1.2)

2(1)

3(1.4)

1(0.5)

Unknown

1

 

 

 

1

Treatment

 

 

 

 

 

Surgery (TSA+open)

118

108

97

118

88

Drug

29

24

24

24

15

Surgery+Drug

48

52

41

44

17

Radiation

19

21

20

32

28

 

 

Table 2. Prevalence of diabetes mellitus and hypertension in patients with acromegaly

(a) Univariate analysis

Outcome

Treatment

HR

95% CI

P

Diabetes

Medication only

0.548

0.221

1.359

0.1945

Surgery only

3.397

0.969

11.912

0.0561

Both

1.316

0.473

3.656

0.5988

Hypertension

Medication only

0.259

0.082

0.814

0.0208

Surgery only

0.404

0.052

3.129

0.3856

Both

0.149

0.019

1.157

0.0687

 

(b) Multivariate analysis*

Outcome

Treatment

HR

95% CI

P

Diabetes

Medication only

0.518

0.209

1.285

0.1559

Surgery only

2.347

0.643

8.57

0.1967

Both

1.361

0.484

3.827

0.5588

Hypertension

Medication only

0.257

0.082

0.808

0.0201

Surgery only

0.381

0.047

3.089

0.366

Both

0.143

0.018

1.118

0.0637

*Adjusted for age, sex and income level