Disease progression and treatment response in data-driven subgroups of type 2 diabetes compared with models based on simple clinical features: an analysis using clinical trial data

Abstract

Background: Research using data-driven cluster analysis has proposed five subgroups of diabetes with differences in diabetes progression and risk of complications. We aimed to compare the clinical utility of this subgroup-based approach for predicting patient outcomes with an alternative strategy of developing models for each outcome using simple patient characteristics.

Methods: We identified five clusters in the ADOPT trial (n=4351) using the same data-driven cluster analysis as reported by Ahlqvist and colleagues. Differences between clusters in glycaemic and renal progression were investigated and contrasted with stratification using simple continuous clinical features (age at diagnosis for glycaemic progression and baseline renal function for renal progression). We compared the effectiveness of a strategy of selecting glucose-lowering therapy using clusters with one combining simple clinical features (sex, BMI, age at diagnosis, baseline HbA1c) in an independent trial cohort (RECORD [n=4447]).

Findings: Clusters identified in trial data were similar to those described in the original study by Ahlqvist and colleagues. Clusters showed differences in glycaemic progression, but a model using age at diagnosis alone explained a similar amount of variation in progression. We found differences in incidence of chronic kidney disease between clusters; however, estimated glomerular filtration rate at baseline was a better predictor of time to chronic kidney disease. Clusters differed in glycaemic response, with a particular benefit for thiazolidinediones in patients in the severe insulin-resistant diabetes cluster and for sulfonylureas in patients in the mild age-related diabetes cluster. However, simple clinical features outperformed clusters to select therapy for individual patients.

Interpretation: The proposed data-driven clusters differ in diabetes progression and treatment response, but models that are based on simple continuous clinical features are more useful to stratify patients. This finding suggests that precision medicine in type 2 diabetes is likely to have most clinical utility if it is based on an approach of using specific phenotypic measures to predict specific outcomes, rather than assigning patients to subgroups.

Funding: UK Medical Research Council.

Copyright © 2019 The Author(s). Published by Elsevier Ltd. This is an Open Access article under the CC BY 4.0 license. Published by Elsevier Ltd.. All rights reserved.

Figures

Figure 1

Cluster characteristics and cluster distribution in ADOPT (A) Distributions of HbA1c, BMI, age at diagnosis, HOMA2-B, and HOMA2-IR at baseline for each cluster. (B) Distribution of ADOPT participants (n=4003) according to k-means clustering. SAID=severe autoimmune diabetes. SIDD=severe insulin-deficient diabetes. SIRD=severe insulin-resistant diabetes. MOD=mild obesity-related diabetes. MARD=mild age-related diabetes. HOMA2-B=homoeostatic model assessment 2 estimates of β-cell function. HOMA2-IR=homoeostatic model assessment 2 estimates of insulin resistance.

Figure 2

Glycaemic progression by cluster in ADOPT from 1 to 5 years (A) HbA1c change by cluster (n=3016). (B) HbA1c change by age at diagnosis (10th, 50th, and 90th percentile of ADOPT participants. Data are estimates from repeated measures, mixed-effects models. SAID=severe autoimmune diabetes. SIDD=severe insulin-deficient diabetes. SIRD=severe insulin-resistant diabetes. MOD=mild obesity-related diabetes. MARD=mild age-related diabetes.

Figure 3

Renal progression by cluster in ADOPT over 5 years (A) Cumulative incidence of chronic kidney disease stage 3 (confirmed eGFR 2) in individuals with eGFR ≥60 mL/min per 1·73 m2 at baseline (n=3694). (B) Cumulative incidence of albuminuria (UACR ≥30 mg/g) in individuals with UACR <30 mg/g at baseline (n=3168). eGFR=estimated glomerular filtration rate. UACR=urinary albumin to creatinine ratio. SAID=severe autoimmune diabetes. SIDD=severe insulin-deficient diabetes. SIRD=severe insulin-resistant diabetes. MOD=mild obesity-related diabetes. MARD=mild age-related diabetes.

Figure 4

Change in HbA1c by drug for clusters 2–5 in ADOPT over 3 years (n=3607) Adjusted mean HbA1c over 3 years by drug. Shading shows 95% CIs. Data for cluster 1 (SAID; n=158) are shown in the appendix (p 16). SAID= severe autoimmune diabetes. SIDD=severe insulin-deficient diabetes. SIRD=severe insulin-resistant diabetes. MOD=mild obesity-related diabetes. MARD=mild age-related diabetes.

Figure 5

Change in HbA1c over 3 years in concordant and discordant treatment selection groups (A) ADOPT development cohort (n=3785), clusters strategy (left panel) and clinical features strategy (right panel). (B) RECORD validation cohort (n=4057), clusters strategy (left panel) and clinical features strategy (right panel).

References

1. 1. Merino J, Florez JC. Precision medicine in diabetes: an opportunity for clinical translation. Ann N Y Acad Sci. 2018;1411:140–152.
   2. J Merino, JC Florez. Precision medicine in diabetes: an opportunity for clinical translation. Ann N Y Acad Sci, 1411, 2018, 140–152
2. 1. Gloyn AL, Drucker DJ. Precision medicine in the management of type 2 diabetes. Lancet Diabetes Endocrinol. 2018;6:891–900.
   2. AL Gloyn, DJ Drucker. Precision medicine in the management of type 2 diabetes. Lancet Diabetes Endocrinol, 6, 2018, 891–900
3. 1. McCarthy MI. Painting a new picture of personalised medicine for diabetes. Diabetologia. 2017;60:793–799.
   2. MI McCarthy. Painting a new picture of personalised medicine for diabetes. Diabetologia, 60, 2017, 793–799
4. 1. Ahlqvist E, Storm P, Käräjämäki A. Novel subgroups of adult-onset diabetes and their association with outcomes: a data-driven cluster analysis of six variables. Lancet Diabetes Endocrinol. 2018;6:361–369.
   2. E Ahlqvist, P Storm, A Käräjämäki. Novel subgroups of adult-onset diabetes and their association with outcomes: a data-driven cluster analysis of six variables. Lancet Diabetes Endocrinol, 6, 2018, 361–369
5. 1. Feher MD, Munro N, Russell-Jones D, de Lusignan S, Khunti K. Novel diabetes subgroups. Lancet Diabetes Endocrinol. 2018;6:439.
   2. MD Feher, N Munro, D Russell-Jones, S de Lusignan, K Khunti. Novel diabetes subgroups. Lancet Diabetes Endocrinol, 6, 2018, 439
6. 1. Dennis JM, Henley WE, Weedon MN. Sex and BMI alter the benefits and risks of sulfonylureas and thiazolidinediones in type 2 diabetes: a framework for evaluating stratification using routine clinical and individual trial data. Diabetes Care. 2018;41:1844–1853.
   2. JM Dennis, WE Henley, MN Weedon. Sex and BMI alter the benefits and risks of sulfonylureas and thiazolidinediones in type 2 diabetes: a framework for evaluating stratification using routine clinical and individual trial data. Diabetes Care, 41, 2018, 1844–1853
7. 1. van Smeden M, Harrell FE, Jr, Dahly DL. Novel diabetes subgroups. Lancet Diabetes Endocrinol. 2018;6:439–440.
   2. M van Smeden, FE Harrell Jr DL Dahly. Novel diabetes subgroups. Lancet Diabetes Endocrinol, 6, 2018, 439–440
8. 1. Kahn SE, Haffner SM, Heise MA. Glycemic durability of rosiglitazone, metformin, or glyburide monotherapy. N Engl J Med. 2006;355:2427–2443.
   2. SE Kahn, SM Haffner, MA Heise. Glycemic durability of rosiglitazone, metformin, or glyburide monotherapy. N Engl J Med, 355, 2006, 2427–2443
9. 1. Home PD, Pocock SJ, Beck-Nielsen H. Rosiglitazone evaluated for cardiovascular outcomes in oral agent combination therapy for type 2 diabetes (RECORD): a multicentre, randomised, open-label trial. Lancet. 2009;373:2125–2135.
   2. PD Home, SJ Pocock, H Beck-Nielsen. Rosiglitazone evaluated for cardiovascular outcomes in oral agent combination therapy for type 2 diabetes (RECORD): a multicentre, randomised, open-label trial. Lancet, 373, 2009, 2125–2135
10. 1. Levy JC, Matthews DR, Hermans MP. Correct homeostasis model assessment (HOMA) evaluation uses the computer program. Diabetes Care. 1998;21:2191–2192.
    2. JC Levy, DR Matthews, MP Hermans. Correct homeostasis model assessment (HOMA) evaluation uses the computer program. Diabetes Care, 21, 1998, 2191–2192
11. 1. Zinman B, Kahn SE, Haffner SM, O'Neill MC, Heise MA, Freed MI. Phenotypic characteristics of GAD antibody-positive recently diagnosed patients with type 2 diabetes in North America and Europe. Diabetes. 2004;53:3193–3200.
    2. B Zinman, SE Kahn, SM Haffner, MC O'Neill, MA Heise, MI Freed. Phenotypic characteristics of GAD antibody-positive recently diagnosed patients with type 2 diabetes in North America and Europe. Diabetes, 53, 2004, 3193–3200
12. 1. Levey AS, Stevens LA, Schmid CH. A new equation to estimate glomerular filtration rate. Ann Intern Med. 2009;150:604–612.
    2. AS Levey, LA Stevens, CH Schmid. A new equation to estimate glomerular filtration rate. Ann Intern Med, 150, 2009, 604–612
13. 1. Levey ASL, Coresh J, Greene T. Expressing the modification of diet in renal disease study equation for estimating glomerular filtration rate with standardized serum creatinine values. Clin Chem. 2007;53:766–772.
    2. ASL Levey, J Coresh, T Greene. Expressing the modification of diet in renal disease study equation for estimating glomerular filtration rate with standardized serum creatinine values. Clin Chem, 53, 2007, 766–772
14. 1. Stratton IM, Adler AI, Neil HAW. Association of glycaemia with macrovascular and microvascular complications of type 2 diabetes (UKPDS 35): prospective observational study. BMJ. 2000;321:405–412.
    2. IM Stratton, AI Adler, HAW Neil. Association of glycaemia with macrovascular and microvascular complications of type 2 diabetes (UKPDS 35): prospective observational study. BMJ, 321, 2000, 405–412
15. 1. Hennig C. Cluster-wise assessment of cluster stability. Comput Stat Data Anal. 2007;52:258–271.
    2. C Hennig. Cluster-wise assessment of cluster stability. Comput Stat Data Anal, 52, 2007, 258–271
16. 1. Nakagawa S, Schielzeth H. A general and simple method for obtaining R2 from generalized linear mixed-effects models. Methods Ecol Evol. 2013;4:133–142.
    2. S Nakagawa, H Schielzeth. A general and simple method for obtaining R2 from generalized linear mixed-effects models. Methods Ecol Evol, 4, 2013, 133–142
17. 1. Frank E, Harrell J. Springer-Verlag; New York: 2006. Regression modeling strategies.
    2. E Frank, J Harrell. Regression modeling strategie, s 200, 6 Springer-Verla, ew York
18. 1. Bossuyt PM, Parvin T. Evaluating biomarkers for guiding treatment decisions. EJIFCC. 2015;26:63–70.
    2. PM Bossuyt, T Parvin. Evaluating biomarkers for guiding treatment decisions. EJIFCC, 26, 2015, 63–70
19. 1. Janes H, Pepe MS, Bossuyt PM, Barlow WE. Measuring the performance of markers for guiding treatment decisions. Ann Intern Med. 2011;154:253–259.
    2. H Janes, MS Pepe, PM Bossuyt, WE Barlow. Measuring the performance of markers for guiding treatment decisions. Ann Intern Med, 154, 2011, 253–259
20. 1. European Medicines Agency Guideline on clinical investigation of medicinal products in the treatment or prevention of diabetes mellitus CPMP/EWP/1080/00 Rev. 1. May 14, 2012.
    2. European Medicines Agenc, uideline on clinical investigation of medicinal products in the treatment or prevention of diabetes mellitus CPMP/EWP/1080/00 Rev. 1, , ay 14, 2012
21. 1. Zou X, Zhou X, Zhu Z, Ji L. Novel subgroups of patients with adult-onset diabetes in Chinese and US populations. Lancet Diabetes Endocrinol. 2019;7:9–11.
    2. X Zou, X Zhou, Z Zhu, L Ji. Novel subgroups of patients with adult-onset diabetes in Chinese and US populations. Lancet Diabetes Endocrinol, 7, 2019, 9–11
22. 1. Donnelly LA, Zhou K, Doney ASF, Jennison C, Franks PW, Pearson ER. Rates of glycaemic deterioration in a real-world population with type 2 diabetes. Diabetologia. 2018;61:607–615.
    2. LA Donnelly, K Zhou, ASF Doney, C Jennison, PW Franks, ER Pearson. Rates of glycaemic deterioration in a real-world population with type 2 diabetes. Diabetologia, 61, 2018, 607–615
23. 1. Hattersley AT, Patel KA. Precision diabetes: learning from monogenic diabetes. Diabetologia. 2017;60:769–777.
    2. AT Hattersley, KA Patel. Precision diabetes: learning from monogenic diabetes. Diabetologia, 60, 2017, 769–777
24. 1. Pearson ER, Starkey BJ, Powell RJ, Gribble FM, Clark PM, Hattersley AT. Genetic cause of hyperglycaemia and response to treatment in diabetes. Lancet. 2003;362:1275–1281.
    2. ER Pearson, BJ Starkey, RJ Powell, FM Gribble, PM Clark, AT Hattersley. Genetic cause of hyperglycaemia and response to treatment in diabetes. Lancet, 362, 2003, 1275–1281
25. 1. Pearson ER, Flechtner I, Njolstad PR. Switching from insulin to oral sulfonylureas in patients with diabetes due to Kir6.2 mutations. New Eng J Med. 2006;355:467–477.
    2. ER Pearson, I Flechtner, PR Njolstad. Switching from insulin to oral sulfonylureas in patients with diabetes due to Kir6.2 mutations. New Eng J Med, 355, 2006, 467–477
26. 1. Li L, Cheng W-Y, Glicksberg BS. Identification of type 2 diabetes subgroups through topological analysis of patient similarity. Sci Transl Med. 2015;7:311ra174.
    2. L Li, W-Y Cheng, BS Glicksberg. Identification of type 2 diabetes subgroups through topological analysis of patient similarity. Sci Transl Med, 7, 2015, 311ra174
27. 1. Udler MS, Kim J, von Grotthuss M. Type 2 diabetes genetic loci informed by multi-trait associations point to disease mechanisms and subtypes: a soft clustering analysis. PLoS Med. 2018;15:e1002654.
    2. MS Udler, J Kim, M von Grotthuss. Type 2 diabetes genetic loci informed by multi-trait associations point to disease mechanisms and subtypes: a soft clustering analysis. PLoS Med, 15, 2018, e1002654
28. 1. Jones AG, McDonald TJ, Shields BM. Markers of beta-cell failure predict poor glycemic response to glp-1 receptor agonist therapy in type 2 diabetes. Diabetes Care. 2016;39:250–257.
    2. AG Jones, TJ McDonald, BM Shields. Markers of beta-cell failure predict poor glycemic response to glp-1 receptor agonist therapy in type 2 diabetes. Diabetes Care, 39, 2016, 250–257
29. 1. DeFronzo RA, Ferrannini E, Schernthaner G. Slope of change in HbA1c from baseline with empagliflozin compared with sitagliptin or glimepiride in patients with type 2 diabetes. Endocrinol Diabetes Metab. 2018;1:e00016.
    2. RA DeFronzo, E Ferrannini, G Schernthaner. Slope of change in HbA1c from baseline with empagliflozin compared with sitagliptin or glimepiride in patients with type 2 diabetes. Endocrinol Diabetes Metab, 1, 2018, e00016
30. 1. Dennis JM, Shields BM, Hill AV. Precision medicine in type 2 diabetes: clinical markers of insulin resistance are associated with altered short- and long-term glycemic response to DPP-4 inhibitor therapy. Diabetes Care. 2018;41:705–712.
    2. JM Dennis, BM Shields, AV Hill. Precision medicine in type 2 diabetes: clinical markers of insulin resistance are associated with altered short- and long-term glycemic response to DPP-4 inhibitor therapy. Diabetes Care, 41, 2018, 705–712
31. 1. Strom BL, Buyse ME, Hughes J, Knoppers BM. Data sharing—is the juice worth the squeeze? N Eng J Med. 2016;375:1608–1609.
    2. BL Strom, ME Buyse, J Hughes, BM Knoppers. Data sharing—is the juice worth the squeeze?. N Eng J Med, 375, 2016, 1608–1609