Циркуляция уровней инсулиновоподобных факторов роста I (IGF-I) и риска множественной миэломы: исследование наблюдений и менделии случайности.

INTRODUCTION

The insulin‐like growth factor (IGF) system, composed of two polypeptide hormones IGF‐I, IGF‐II, their receptors (IGF‐IR and ‐IIR) and six IGF‐binding proteins (IGFBP‐1 to ‐6), is associated with cell growth, survival, apoptosis, angiogenesis, metabolism and migration. 1 , 2 Circulating IGF‐I is associated with the risk of several solid cancers (colorectal, breast and prostate) in observational studies with further support from Mendelian randomisation (MR) studies. 3 , 4 , 5 , 6 , 7 , 8 Evidence is accumulating that the IGF system may contribute to the biology of multiple myeloma (MM). 9 Borderline non‐significant positive associations were observed for IGF‐I and IGFBP‐3 levels with MM in a nested case–control study within the Multiple Myeloma Cohort Consortium (MMCC). 10 In the European Prospective Investigation into Cancer and Nutrition (EPIC) study sample, IGF‐I levels were not associated with lymphoid neoplasm (LN) or associated subtypes. 11 A study examining the risk of 30 cancers and IGF‐I concentrations conducted within the UK Biobank but reported a weak positive association with MM that failed to meet correction for multiple testing. 7 The recent update of cancer data in the UK Biobank, which shows a doubling of most types of cases, allows for more detailed subtype‐specific investigations to advance aetiological understanding.

Here, we first aimed to estimate if pre‐diagnostic serum levels of IGF‐I are related to MM risk in the UK Biobank study. Then, using a two‐sample MR approach, we obtained estimates that combine the potential effect of genetic variants associated with IGF‐I and IGFBP‐3 levels and estimate the effect of these variants on the risk of MM using genome‐wide association study (GWAS) data from the InterLymph consortium. 12 We also included additional LN subtypes for comparison. This allowed us to assess whether the observed associations are specific to MM or reflect broader patterns across related LN, thereby strengthening the interpretation and robustness of our findings. Finally, motivated by previous findings linking obesity to MM 13 and chronic lymphocytic leukaemia/small lymphocytic lymphoma (CLL/SLL) 14 we examined the role of body mass index (BMI) for these two subtypes.

METHODS

The UK Biobank study

Observational study

The UK Biobank is a prospective cohort of nearly half a million adults recruited between 2006 and 2010 and aged 40–69 years at enrolment. 15 , 16 Further information can be found in Supplementary Methodology S1 . Participants with a diagnosis of cancer before recruitment other than non‐melanoma skin cancer (ICD‐10 C44) ( n = 24 699) and those lacking an IGF‐I measurement ( n = 33 247) were excluded from the present analysis (Supplementary Methodology S1 ). Our analysis thus included 444 187 participants (Number 52433).

Assessment of outcome

Incident cancer cases within the UK Biobank cohort were identified through linkage to national cancer registries. Cancer incidence data were coded using the 10th Revision of the International Classification of Diseases (ICD‐10) or the 3rd Revision of the ICD‐Oncology ( https://biobank.ndph.ox.ac.uk/showcase/ukb/docs/CancerLinkage.pdf ). We defined the following first registered end‐points for analysis based on guidelines from the InterLymph Consortium 17 : MM, all LN, Hodgkin lymphoma (HL) and non‐Hodgkin lymphoma (NHL, excluding plasma cell neoplasms 18 ); the latter included T‐cell NHL (T‐NHL) and B‐cell NHL (B‐NHL). We also analysed selected common mature B‐cell LN subtypes separately, including diffuse large B‐cell lymphoma (DLBCL), follicular lymphoma (FL) and CLL/SLL (Table S2 for disease classifications).

Blood collection and laboratory methods

Serum concentrations of total IGF‐I were measured using chemiluminescence immunoassays (DiaSorin Liaison XL, analytic range 1.3–195 nmol/L). Measurements were conducted at a purpose‐built laboratory for UK Biobank in Stockport, England, and the average within‐laboratory coefficients of variation (ratio of the SD to the mean) were 6.03%, 5.29% and 6.18% for low, medium and high concentrations respectively. Further information can be found in Supplementary Methodology S1 .

Statistical analysis

Cox proportional hazard models with different levels of covariate adjustment based on a priori assumptions 19 , 20 , 21 , 22 , 23 were used, both of which were stratified by age (years). For information on selected variables and categorisation, see Supplementary Methodology S1 . Model 1 adjusted for sex, education, UK region and Townsend deprivation index. Model 2 further adjusted for ethnicity, height, BMI, smoking status, alcohol consumption, vigorous physical activity, diabetes history and fasting time. We analysed IGF‐I levels as categorical (sex‐specific quartiles) and continuous variables (per sex‐specific 1‐standard deviation [SD] increase; 1‐SD = 5.7 nmol/L; 5.8 nmol/L and 5.5 nmol/L for all participants, among females and among males respectively). Model 2 was additionally corrected for a regression dilution ratio for IGF‐I (0.77) obtained from the subsample of 17 697 participants with repeated IGF‐I measurement; 24 , 25 specifically, to obtain corrected HRs, the log HRs and their standard errors were divided by 0.77 and then exponentiated. 26 We also obtained results stratified by time from recruitment to the end of follow‐up (<5 years, 5–9 years, 10 or more years) and assessed multiplicative interactions between the continuous concentration of IGF‐I and the following variables by including an interaction term in model 2: sex, BMI (<25, ≥25 kg/m 2 ), smoking status (never, ever), alcohol intake (never, ever), ethnicity (white, other), height (≤, >sex‐specific median) for the end‐points (MM, LN subtypes alone and categorised as NHL and B‐NHL). As a secondary analysis, 13 , 14 we examined the interaction between the continuous concentration (per 1‐SD increase) of IGF‐I and BMI in relation to the risk of MM and CLL/SLL, and we investigated a potential three‐way interaction of IGF‐I, sex and BMI in relation to risk of those end‐points. As sensitivity analyses, we additionally adjusted model 2 for markers of inflammatory, sex hormone and glycaemic pathways that are known to inter‐relate/cross‐talk with the IGF system, namely, C‐reactive protein (CRP), testosterone, sex hormone‐binding globulin (SHBG) and glycosylated haemoglobin (HbA1c) 27 (model 3), excluding the first 2 years of follow‐up and examined the main findings (e.g. model 2 with regression dilution correction) excluding those in the highest and lowest 1% of the IGF‐I distribution ( n = 4442) (Supplementary Methodology S1 ).

Mendelian randomisation study

Genetic determinants of IGF ‐I and IGFBP ‐3 circulating levels

We used previously described genetic instruments 28 (Supplementary Methodology S1 ). For a summary of the genetic instruments and of the corresponding single nucleotide polymorphism (SNP)‐specific associations with IGF‐I and IGFBP‐3 levels and with the different LN subtypes, see Table S3 .

Data on LN subtypes

We obtained estimates of the association of IGF‐I and IGFBP‐3 genetic markers with LN subtypes from summary data from GWAS meta‐analyses conducted within the InterLymph consortium, which included 2434 MM and their 2567 controls of European ancestry as well as 3100 CLL/SLL, 2847 FL and 3857 DLBCL and their 8542 controls. 29 , 30 , 31 , 32 Table S4 lists the InterLymph studies included in this analysis.

Statistical analysis

Prior to the analysis, the alleles were harmonised by aligning the effect allele for the exposure (IGF‐I and IGFBP‐3) and the outcomes (MM and LN subtypes), thus ensuring that the effect alleles were the same across both datasets. We used the three most common models to estimate separately the potential effect of genetically predicted levels for the IGF‐I‐ and IGFBP‐3‐related instruments: Inverse variance‐weighted (IVW) models, weighted median models and MR‐Egger method. 33 We employed the latter to evaluate causal relationships between genetic instruments and outcomes, allowing for the detection of pleiotropy and providing robust causal effect estimates. 34 GWAS summary statistics from the IEU OpenGWAS and GWAS catalogue for BMI‐associated SNPs were used to test potential violation of the independence assumption. The MR‐PRESSO (Mendelian Randomisation Pleiotropy RESidual Sum and Outlier) 35 distortion test was employed to assess the presence of horizontal pleiotropy (see Supplementary Methodology S1 ).

RESULTS

Observational study: The UK biobank

The final analytical cohort included 444 187 men and women. During a mean (SD) of 10.5 (1.7) years of follow‐up, 121 incident HL and 2430 incident cases of other LN types were identified; the latter included 732 MM, 667 CLL/SLL, 680 DLBCL, 418 FL and 163 T‐NHL. At enrolment, cohort participants had a mean age of 56 years and a higher proportion of females ( n = 239 048; 54%) than males (Table 1 ). Compared to participants with circulating levels of IGF‐I within the lower quartiles, those in the highest quartile (Q4) were younger, had lower BMI, were taller, more likely to be non‐smokers, less likely to have self‐reported diabetes and less likely to be in the most deprived (third tertile) category of the Townsend index (Table 1 ). Overall, the mean IGF‐I level was 21.4 (SD: 5.7 nmol/L), with a range from 1.4 nmol/L to 126.8 nmol/L. Among LN subtypes, the highest IGF‐I values were observed among participants who subsequently developed MM and CLL/SLL.

We detected a statistically significant interaction between sex and circulating IGF‐I levels with the main aggregated endpoints (LN, NHL and B‐NHL; Table S5 ) and decided to present results for the whole cohort and stratified by sex. In multivariate models, higher levels of IGF‐I were associated with increased MM risk (corrected HR 1‐SD = 1.11; 95% confidence interval [CI]: 1.01–1.22; p ‐value = 0.03), whereas IGF‐I levels were not associated with risk of LN, HL, NHL, B‐NHL or LN subtypes (Table 2 ; for quartile‐specific findings, see Table S6 ). In the sex‐specific models, we observed an increased risk of CLL/SLL per 1‐SD increase in IGF‐I levels in males but a suggested inverse association in females (males, corrected HR 1‐SD : 1.13, 95% CI: 1.00–1.28; females, HR 1‐SD : 0.87, 95% CI: 0.73–1.04; p ‐value for interaction = 0.01). A three‐way interaction for the association between CLL/SLL and IGF‐I was detected between sex and BMI variables ( p ‐value for interaction = 0.002 and 0.004 for two‐ and three‐level BMI variable, respectively; Table S5 and Table 3 ). Specifically, we noted the strongest associations of circulating IGF‐I levels with CLL/SLL risk in males with higher BMI (corrected HR 1‐SD = 0.79, 1.12 and 1.36 among normal BMI, overweight and obese males respectively). No interaction by sex or BMI was detected for MM. However, the association between IGF‐I levels and MM was stronger among females than males (corrected HR 1‐SD for MM: 1.16, 95% CI: 1.00–1.34 in females and 1.06, 95% CI: 0.94–1.20 in males; P‐value for interaction = 0.54; Table 3 ).

The sensitivity analyses revealed no meaningful differences (Tables S6–S8 ). When examined separately by follow‐up period (Table 4 ), the strongest associations were observed for risk of MM within a shorter time period after blood draw (corrected HR 1‐SD : 1.22, 1.03 and 1.02 within <5 years, 5–9 years and 10 or more years respectively).

Mendelian randomisation

A 1‐SD increment (around 5 nmol/L) in genetically inferred circulating IGF‐I levels was associated with a 27% increase of MM risk (OR = 1.27, 95% CI: 1.05–1.54) (Table 5 ). For the other LN subtypes, no association was observed for genetically inferred IGF‐I levels, and we found no association of genetically inferred IGFBP‐3 levels with any LN end‐point. The estimates obtained from the weighted‐median approach and MR‐Egger were consistent with those obtained from IVW for all subtypes and for both IGF‐I and IGFBP‐3. As a sensitivity analysis, SNPs associated with BMI were excluded. The causal estimates remained essentially unchanged ( data not shown ). The MR‐PRESSO test identified potential horizontal pleiotropy for the associations of genetically inferred IGF‐I levels with FL and CLL/SLL (global test p‐value 0.001 and <1 × 10 −4 for FL and CLL/SLL respectively). Removing outliers did not materially change the estimates, suggesting that outlying SNPs were not influencing the original results. No evidence of horizontal pleiotropy was found for MM or DLBCL, supporting the robustness of the estimates (Supplementary Methodology S1 , Figure S9 ).

DISCUSSION

In the observational analysis, IGF‐I levels were positively associated with MM risk overall and, in a secondary analysis, with CLL/SLL risk among males (especially those with higher BMI) but not among females. The MR analysis yielded a positive association between genetically inferred circulating IGF‐I levels and MM risk, thus supporting the involvement of the IGF‐I pathway in MM development. In contrast, we did not observe an association of genetically inferred IFGBP‐3 levels with any LN subtype.

Our results are consistent with previous data from the UK Biobank with only 367 cases 7 and a pooled study nested in six cohorts comprising the MMCC 10 that included 493 MM and 978 controls but were inconsistent with a nested case–control study within the EPIC cohort, which included 237 MM and matched controls (Table S10 ). 11 Supporting our results, Birmann et al. did not report interaction by sex or BMI in the association between IGF‐I or related biomarker levels and MM. 10 Although these studies—including ours—had large sample sizes, they were limited in their ability to stratify by time (mean time from blood draw to diagnosis: 9 [SD: 2] years and 6.6 [range 0.3–18.1] years for the EPIC and the MMCC studies respectively), and the large 95% CIs precluded definitive conclusions. Of note, however, in the present analysis, we also observed the strongest association with MM diagnosis <5 years after blood draw, which might reflect disease progression from monoclonal gammopathy of undetermined significance (MGUS) to MM.

IGF is a complex system, 9 , 36 and by examining total IGF‐I, we did not disentangle free (bioavailable) and bound IGF‐I. Indeed, the bioavailability of IGF‐I is strictly controlled by IGFBPs and in particular by the abundant IGFBP‐3. 37 Free IGF‐I can bind to IGF‐1 receptor (IGF‐1R), which might produce cell proliferation as well as inhibition of apoptosis. 38 MM cells have autocrine IGF‐I production that have been associated with cell survival and disease progression 39 , 40 , 41 , 42 and MM cells display aberrant IGF‐1R expression. 43 Hence, serum total circulating IGF‐I levels might not be the best biomarker to study the IGF‐I system. However, there is currently no genetic instrument as a proxy for free IGF‐I exposure, and our findings serve as a preliminary step and underscore the need for future MR studies incorporating free IGF‐I levels. The MMCC study reported a non‐significant positive association between IGFBP‐3 levels and MM Hazard ratio‐HR‐per one standard deviation‐SD‐increase (HR 1SD : 1.1; 95% CI: 1.0–1.2; p ‐value: 0.08 and HR Q4 vs. Q1 : 1.4; 95% CI: 1.0–1.9). However, the molar ratio of IGF‐I to IGFBP‐3 (an empirical indicator of IGF‐I bioavailability) was not associated with MM risk. Additionally, the MMCC study noted a positive association between pre‐diagnosis IGFBP‐1 levels and MM risk, 10 with the strongest association within 3 years after blood draw. Further examination of additional IGF pathway signalling molecules is warranted.

Consistent with our results, the case–control study nested within the EPIC cohort 11 detected no overall association between CLL/SLL and pre‐diagnosis IGF‐I concentrations; conventional (retrospective) case–control studies have reported contradictory results. 14 , 44 , 45 , 46 The EPIC study detected borderline statistically significant separate interactions of IGF‐I levels and LN with sex and with BMI but did not present results stratified jointly by sex and BMI; in the EPIC study, circulating IGF‐I levels were inversely associated with LN risk (all subtypes grouped) in women and in individuals with normal BMI. 11 In the EpiLymph and the Spanish multi‐case–control (MCC‐Spain) studies, we examined circulating IGF‐I and IGFBP‐3 levels among 209 untreated CLL/SLL patients and their age‐ and sex‐matched controls. IGF‐I levels were positively associated with CLL/SLL risk among obese patients, but interactions of BMI with sex, a matching variable, were not assessed. 14 In the current examination of those interactions, we found a stronger positive association of circulating IGF‐I with CLL/SLL among overweight/obese men. Of interest, the inverse association observed between MM risk and pre‐diagnosis plasma adiponectin, an adipokine, varied by BMI in the MMCC study. 47 These findings warrant further investigation of these biomarkers in studies that include data on sex and BMI, in order to clarify potential stratum‐specific associations.

Observational studies are prone to residual confounding and reverse causation. MR uses genetic variants as instrumental variables to assess potential effects, mimicking randomisation and reducing residual confounding. The present MR analysis supported the observational data for MM, the plausibility of which is further supported by extensive laboratory studies reporting a role for the IGF‐I system in the growth, survival, apoptosis and migration of MM cells. 9 , 36 , 40 , 48 , 49 , 50 Based on the laboratory evidence, several clinical trials have also been designed in recent decades to test putative therapeutic molecules that target the IGF‐I receptor. 51 , 52 , 53 Overall results from our MR study of CLL/SLL did not support a role of IGF‐I, but heterogeneity of the estimates by sex and BMI could not be evaluated by MR since sex‐specific data were not available. To address the key MR assumption of no horizontal pleiotropy, we conducted sensitivity analyses, which supported the robustness of our estimates for IGF‐I and MM.

This is the largest and most comprehensive study to date, exploring the role of IGF‐I in LN subtype development using two complementary designs. However, despite the large sample size, stratified results were based on small numbers of cases. Furthermore, we studied only total IGF‐I levels in the observational study, and the IGF‐I instrument variable in the MR analyses was likewise related to total rather than free IGF‐I. Those limitations do not support inferences regarding bioavailable IGF‐I per se, and we did not observe associations for the IGFBP‐3 instrument variable. Our main hypothesis was on the association between IGF‐I levels and MM. We included other LN subtypes as comparators, allowing us to determine whether any observed associations were unique to MM or indicative of broader LN patterns. While our results would not remain statistically significant after adjusting for multiple testing, our hypothesis‐driven framework ensured that the observed findings remain scientifically relevant and worthy of further investigation. Another important point is the small sample sizes in stratified groups that result in wider confidence intervals and less precise estimates. This highlights the need to interpret findings within these subgroups with caution. If feasible, increasing the sample size in future studies could help improve the stability and reliability of these estimates. Therefore, we recommend caution in interpreting our findings until additional large, prospective studies with information on relevant covariates can confirm them. Finally, the prevalence of MGUS among the different studies was not known and might have affected the results.

In conclusion, this work is consistent with a direct relationship between MM and IGF‐I levels. However, given the modest effect sizes observed, the MR findings should be contextualised within the inherent limitations of genetic instrumental variable analyses and limited statistical power. In turn, further studies are needed to better understand the relevance of other components of the IGF‐I signalling pathway (e.g. IGF‐1R) and of bioavailable IGF‐I, to confirm the aetiologically relevant timing of IGF‐I signalling for these end‐points and to further elucidate the apparent difference in associations that we observed between circulating IGF‐I levels and CLL/SLL by sex and BMI. Future work should also include participants with precursor stages of CLL/SLL and MM (monoclonal B‐cell lymphocytosis and MGUS, respectively) to better understand the IGF‐I associations.

AUTHOR CONTRIBUTIONS

DC designed the research study. DC and YB performed the research. DC, YB and SH analysed the data. DC, YB, LA, SH and BMB involved in funding acquisition. DC, YB, SH and BMB wrote the paper. All authors involved in data acquisition, paper revision and approval of the submitted and final versions.

FUNDING INFORMATION

This work was partially supported by Instituto de Salud Carlos III—ISCIII (Spanish Government) co‐funded by FEDER funds/European Regional Development Fund (ERDF)—a way to build Europe (CIBERESP CB06/02/0073, PI20/00288, FI21/00090, PI20/01845 and PI25/00156). Additional support came from the Agency for Management of Universities and Research Grants (AGAUR; 2021SGR01354, 2014SGR756) and the Consejería de Salud y Familia de la Junta de Andalucía (PY20/01282) and FEDER. We thank the Fundacio Institució dels Centres de Recerca de Catalunya (CERCA) Program/Generalitat de Catalunya for institutional support to IDIBELL. The InterLymph data coordinating centre (DCC) was supported by the NCI/U01CA257679. Information on funding for individual studies can be found in Table S4 .

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

ETHICAL APPROVAL STATEMENT

The UK biobank study received ethical approval from relevant committees and is overseen by an independent Ethics and Governance Council ( https://www.ukbiobank.ac.uk/learn‐more‐about‐uk‐biobank/about‐us/ethics ). All studies from the InterLymph consortium were approved by their respective ethical committees.

Supporting information

Data S1.

ACKNOWLEDGEMENTS

This research has been conducted using data from UK Biobank, a major biomedical database, under Application Number 52433. We thank the investigators and participants involved in the UK Biobank for their contributions to this study. We thank Murat Güler from the Genomic Epidemiology/C055, German Cancer Research Center (DKFZ) for the help with LN classification and Dr. Christine F. Skibola for her work in the UCSF. Where authors are identified as personnel of the International Agency for Research on Cancer/World Health Organization, the authors alone are responsible for the views expressed in this article and they do not necessarily represent the decisions, policy or views of the International Agency for Research on Cancer/World Health Organization. Specific study acknowledgements can be found in Table S4 .

DATA AVAILABILITY STATEMENT

Genotype data from the NCI NHL GWAS are available on dbGaP (phs000801.v2.p1) for research purposes in accordance with dbGaP data access policies. Other data in this manuscript are available for shared research purposes through the InterLymph Consortium upon approval in accordance with the InterLymph data sharing framework, institutional review boards and general data protection regulations ( https://datacollect.iarc.fr/redcap/surveys/?s=73H8NHL8PN8XCJ4J ). The UK Biobank data can be accessed applying at https://www.ukbiobank.ac.uk/enable‐your‐research/apply‐for‐access .

REFERENCES

1 Pollak MN , Schernhammer ES , Hankinson SE . Insulin‐like growth factors and neoplasia . Nat Rev Cancer . 2004 ; 4 ( 7 ): 505 – 518 . 10.1038/nrc1387 15229476

2 Pollak M . Insulin and insulin‐like growth factor signalling in neoplasia . Nat Rev Cancer . 2008 ; 8 ( 12 ): 915 – 928 . 10.1038/nrc2536 19029956

3 Key TJ , Appleby PN , Reeves GK , Roddam AW . Insulin‐like growth factor 1 (IGF1), IGF binding protein 3 (IGFBP3), and breast cancer risk: pooled individual data analysis of 17 prospective studies . Lancet Oncol . 2010 ; 11 ( 6 ): 530 – 542 . 10.1016/S1470-2045(10)70095-4 20472501 PMC3113287

4 Rinaldi S , Cleveland R , Norat T , Biessy C , Rohrmann S , Linseisen J , et al. Serum levels of IGF‐I, IGFBP‐3 and colorectal cancer risk: results from the EPIC cohort, plus a meta‐analysis of prospective studies . Int J Cancer . 2010 ; 126 ( 7 ): 1702 – 1715 . 10.1002/IJC.24927 19810099

5 Roddam AW , Allen NE , Appleby P , Key TJ , Ferrucci L , Carter HB , et al. Insulin‐like growth factors, their binding proteins, and prostate cancer risk: analysis of individual patient data from 12 prospective studies . Ann Intern Med . 2008 ; 149 ( 7 ): 461 – 471 , W83‐8. 18838726 10.7326/0003-4819-149-7-200810070-00006 PMC2584869

6 Murphy N , Knuppel A , Papadimitriou N , Martin RM , Tsilidis KK , Smith‐Byrne K , et al. Insulin‐like growth factor‐1, insulin‐like growth factor‐binding protein‐3, and breast cancer risk: observational and mendelian randomization analyses with ∼430 000 women . Ann Oncol . 2020 ; 31 ( 5 ): 641 – 649 . 10.1016/J.ANNONC.2020.01.066 32169310 PMC7221341

7 Knuppel A , Fensom GK , Watts EL , Gunter MJ , Murphy N , Papier K , et al. Circulating insulin‐like growth factor‐I concentrations and risk of 30 cancers: prospective analyses in UK biobank . Cancer Res . 2021 ; 80 ( 18 ): 4014 – 4021 . 10.1158/0008-5472.CAN-20-1281 32709735

8 Travis RC , Appleby PN , Martin RM , Holly JMP , Albanes D , Black A , et al. A meta‐analysis of individual participant data reveals an association between circulating levels of IGF‐I and prostate cancer risk . Cancer Res . 2016 ; 76 ( 8 ): 2288 – 2300 . 10.1158/0008-5472.CAN-15-1551 26921328 PMC4873385

9 Bieghs L , Johnsen HE , Maes K , Menu E , van Valckenborgh E , Overgaard MT , et al. The insulin‐like growth factor system in multiple myeloma: diagnostic and therapeutic potential . Oncotarget . 2016 ; 7 ( 30 ): 48732 – 48752 . 10.18632/oncotarget.8982 27129151 PMC5217049

10 Birmann BM , Neuhouser ML , Rosner B , Albanes D , Buring JE , Giles GG , et al. Prediagnosis biomarkers of insulin‐like growth factor‐1, insulin, and interleukin‐6 dysregulation and multiple myeloma risk in the multiple myeloma cohort consortium . Blood . 2012 ; 120 ( 25 ): 4929 – 4937 . 10.1182/blood-2012-03-417253 23074271 PMC3525019

11 Perez‐Cornago A , Appleby PN , Tipper S , Key TJ , Allen NE , Nieters A , et al. Prediagnostic circulating concentrations of plasma insulin‐like growth factor‐I and risk of lymphoma in the European prospective investigation into cancer and nutrition . Int J Cancer . 2017 ; 140 ( 5 ): 1111 – 1118 . 10.1002/ijc.30528 27870006 PMC5299544

12 Morton LM , Sampson JN , Cerhan JR , Turner JJ , Vajdic CM , Wang SS , et al. Rationale and design of the international lymphoma epidemiology consortium (InterLymph) non‐Hodgkin lymphoma subtypes project . JNCI Monographs . 2014 ; 2014 ( 48 ): 1 – 14 . 10.1093/jncimonographs/lgu005 PMC4155460 25174022

13 Lauby‐Secretan B , Scoccianti C , Loomis D , Grosse Y , Bianchini F , Straif K . Body fatness and cancer—viewpoint of the IARC Working Group . N Engl J Med . 2016 ; 375 ( 8 ): 794 – 798 . 10.1056/NEJMsr1606602 27557308 PMC6754861

14 Casabonne D , Benavente Y , Costas L , Robles C , Gonzalez‐Barca E , de la Banda E , et al. Insulin‐like growth factor levels and chronic lymphocytic leukaemia: results from the MCC‐Spain and EpiLymph‐Spain studies . Br J Haematol . 2019 ; 185 ( 3 ): 608 – 612 . 10.1111/bjh.15583 30450673

15 Sudlow C , Gallacher J , Allen N , Beral V , Burton P , Danesh J , et al. UK biobank: an open access resource for identifying the causes of a wide range of complex diseases of middle and old age . PLoS Med . 2015 ; 12 : e1001779 . 10.1371/journal.pmed.1001779 25826379 PMC4380465

16 Collins R . What makes UK Biobank special? Lancet . 2012 ; 379 : 1173 – 1174 . 10.1016/S0140-6736(12)60404-8 22463865

17 Turner JJ , Morton LM , Linet MS , Clarke CA , Kadin ME , Vajdic CM , et al. InterLymph hierarchical classification of lymphoid neoplasms for epidemiologic research based on the WHO classification (2008): update and future directions . Blood . 2010 ; 116 : e90 – e98 . 10.1182/blood-2010-06-289561 20699439 PMC2993636

18 Alaggio R , Amador C , Anagnostopoulos I , Attygalle AD , Araujo IB d O , Berti E , et al. The 5th edition of the World Health Organization classification of haematolymphoid tumours: lymphoid neoplasms . Leukemia . 2022 ; 36 : 1720 – 1748 . 10.1038/s41375-022-01620-2 35732829 PMC9214472

19 Bertrand KA , Teras LR , Deubler EL , Chao CR , Rosner BA , Wang K , et al. Anthropometric traits and risk of multiple myeloma: a pooled prospective analysis . Br J Cancer . 2022 ; 127 : 1296 – 1303 . 10.1038/s41416-022-01907-2 35840735 PMC9519635

20 Rosenberg PS , Barker KA , Anderson WF . Future distribution of multiple myeloma in the United States by sex, age, and race/ethnicity . Blood . 2015 ; 125 : 410 – 412 . 10.1182/blood-2014-10-609461 25573972 PMC4287646

21 Rota M , Porta L , Pelucchi C , Negri E , Bagnardi V , Bellocco R , et al. Alcohol drinking and multiple myeloma risk—A systematic review and meta‐analysis of the dose‐risk relationship . Eur J Cancer Prev . 2014 ; 23 : 113 – 121 . 10.1097/CEJ.0000000000000001 24469244

22 Matthews CE , Moore SC , Arem H , Cook MB , Trabert B , Håkansson N , et al. Amount and intensity of leisure‐time physical activity and lower cancer risk . J Clin Oncol . 2020 ; 38 : 686 – 697 . 10.1200/JCO.19.02407 31877085 PMC7048166

23 Tsilidis KK , Kasimis JC , Lopez DS , Ntzani EE , Ioannidis JPA . Type 2 diabetes and cancer: umbrella review of meta‐analyses of observationlal studies . BMJ . 2015 ; 350 : g7607 . 10.1136/bmj.g7607 25555821

24 MacMahon S , Peto R , Collins R , Sorlie P , Neaton J , Abbott R , et al. Blood pressure, stroke, and coronary heart disease. Part 1, prolonged differences in blood pressure: prospective observational studies corrected for the regression dilution bias . Lancet . 1990 ; 335 ( 8692 ): 765 – 774 . 10.1016/0140-6736(90)90878-9 1969518

25 Clarke R , Shipley M , Lewington S , Youngman L , Collins R , Marmot M , et al. Underestimation of risk associations due to regression dilution in long‐term follow‐up of prospective studies . Am J Epidemiol . 1999 ; 150 ( 4 ): 341 – 353 . 10.1093/OXFORDJOURNALS.AJE.A010013 10453810

26 Clarke R , Emberson JR , Breeze E , Casas JP , Parish S , Hingorani AD , et al. Biomarkers of inflammation predict both vascular and non‐vascular mortality in older men . Eur Heart J . 2008 ; 29 ( 6 ): 800 – 809 . 10.1093/EURHEARTJ/EHN049 18303034

27 Hosgood HD , Gunter MJ , Murphy N , Rohan TE , Strickler HD . The relation of obesity‐related hormonal and cytokine levels with multiple myeloma and non‐Hodgkin lymphoma . Front Oncol . 2018 ; 8 ( APR ): 103 . 10.3389/FONC.2018.00103 29713614 PMC5911620

28 Murphy N , Carreras‐Torres R , Song M , Chan AT , Martin RM , Papadimitriou N , et al. Circulating levels of insulin‐like growth factor 1 and insulin‐like growth factor binding protein 3 associate with risk of colorectal cancer based on serologic and mendelian randomization analyses . Gastroenterology . 2020 ; 158 ( 5 ): 1300 – 1312.e20 . 10.1053/j.gastro.2019.12.020 31884074 PMC7152801

29 Berndt SI , Camp NJ , Skibola CF , Vijai J , Wang Z , Gu J , et al. Meta‐analysis of genome‐wide association studies discovers multiple loci for chronic lymphocytic leukemia . Nat Commun . 2016 ; 7 : 10933 . 10.1038/ncomms10933 26956414 PMC4786871

30 Skibola CF , Berndt SI , Vijai J , Skibola C F , Berndt S I , Conde L , et al. Genome‐wide association study identifies five susceptibility loci for follicular lymphoma outside the HLA region . Am J Hum Genet . 2014 ; 95 ( 4 ): 462 – 471 . 10.1016/j.ajhg.2014.09.004 25279986 PMC4185120

31 Cerhan JR , Berndt SI , Vijai J , Ghesquières H , McKay J , Wang SS , et al. Genome‐wide association study identifies multiple susceptibility loci for diffuse large B cell lymphoma . Nat Genet . 2014 ; 46 ( 11 ): 1233 – 1238 . 10.1038/NG.3105 25261932 PMC4213349

32 Clay‐Gilmour AI , Hildebrandt MAT , Brown EE , Hofmann JN , Spinelli JJ , Giles GG , et al. Coinherited genetics of multiple myeloma and its precursor, monoclonal gammopathy of undetermined significance . Blood Adv . 2020 ; 4 ( 12 ): 2789 – 2797 . 10.1182/BLOODADVANCES.2020001435 32569378 PMC7322948

33 Bowden J , Davey Smith G , Haycock PC , Burgess S . Consistent estimation in mendelian randomization with some invalid instruments using a weighted median estimator . Genet Epidemiol . 2016 ; 40 ( 4 ): 304 – 314 . 10.1002/GEPI.21965 27061298 PMC4849733

34 Bowden J , Smith GD , Burgess S . Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression . Int J Epidemiol . 2015 ; 44 ( 2 ): 512 – 525 . 10.1093/IJE/DYV080 26050253 PMC4469799

35 Verbanck M , Chen CY , Neale B , Do R . Detection of widespread horizontal pleiotropy in causal relationships inferred from mendelian randomization between complex traits and diseases . Nat Genet . 2018 ; 50 ( 5 ): 693 – 698 . 10.1038/s41588-018-0099-7 29686387 PMC6083837

36 Vishwamitra D , George SK , Shi P , Kaseb AO , Amin HM . Type I insulin‐like growth factor receptor signaling in hematological malignancies . Oncotarget . 2017 ; 8 ( 1 ): 1814 – 1844 . 10.18632/oncotarget.12123 27661006 PMC5352101

37 Jones JI , Clemmons DR . Insulin‐like growth factors and their binding proteins: biological actions* . Endocr Rev . 1995 ; 16 ( 1 ): 3 – 34 . 10.1210/edrv-16-1-3 7758431

38 Baserga R , Peruzzi F , Reiss K . The IGF‐1 receptor in cancer biology . Int J Cancer . 2003 ; 107 ( 6 ): 873 – 877 . 10.1002/ijc.11487 14601044

39 Georgii‐Hemming P , Jernberg Wiklund H , Ljunggren Ö , Nilsson K . Insulin‐like growth factor I is a growth and survival factor in human multiple myeloma cell lines . Blood . 1996 ; 88 ( 6 ): 2250 – 2258 . 10.1182/BLOOD.V88.6.2250.BLOODJOURNAL8862250 8822946

40 Sprynski AC , Hose D , Caillot L , Réme T , Shaughnessy JD Jr , Barlogie B , et al. The role of IGF‐1 as a major growth factor for myeloma cell lines and the prognostic relevance of the expression of its receptor . Blood . 2009 ; 113 ( 19 ): 4614 – 4626 . 10.1182/BLOOD-2008-07-170464 19228610 PMC2691749

41 Chiron D , Maïga S , Surget S , Descamps G , Gomez‐Bougie P , Traore S , et al. Autocrine insulin‐like growth factor 1 and stem cell factor but not interleukin 6 support self‐renewal of human myeloma cells . Blood Cancer J . 2013 ; 3 ( 6 ): e120 . 10.1038/BCJ.2013.18 23749045 PMC3698536

42 Hose D , Moreaux J , Meissner T , Seckinger A , Goldschmidt H , Benner A , et al. Induction of angiogenesis by normal and malignant plasma cells . Blood . 2009 ; 114 ( 1 ): 128 – 143 . 10.1182/BLOOD-2008-10-184226 19299335

43 Régis B , Nelly R , Avet‐Loiseau H , Harousseau JL , Moreau P . CD221 (IGF‐1R) is aberrantly expressed in multiple myeloma, in relation to disease severity . Haematologica . 2005 ; 90 ( 5 ): 706 – 707 . 15921396

44 Schillaci R , Galeano A , Becu‐Villalobos D , Spinelli O , Sapia S , Bezares RF . Autocrine/paracrine involvement of insulin‐like growth factor‐I and its receptor in chronic lymphocytic leukaemia . Br J Haematol . 2005 ; 130 ( 1 ): 58 – 66 . 10.1111/j.1365-2141.2005.05579.x 15982345

45 Molica S , Vitelli G , Mirabelli R , Digiesu G , Giannarelli D , Cuneo A , et al. Serum insulin‐like growth factor is not elevated in patients with early B‐cell chronic lymphocytic leukemia but is still a prognostic factor for disease progression . Eur J Haematol . 2006 ; 76 ( 1 ): 51 – 57 . 10.1111/j.1600-0609.2005.00553.x 16343271

46 Ayer M , Sakin A , Ay S , Ayer A , Sazak EG , Aktan M . Evaluation of IGF‐1 and IGFBP‐3 expression levels in patients with chronic lymphocytic leukemia . Turk J Hematol . 2016 ; 33 ( 4 ): 335 – 338 . 10.4274/tjh.2016.0075 PMC5204190 27094973

47 Hofmann JN , Birmann BM , Teras LR , Pfeiffer RM , Wang Y , Albanes D , et al. Low levels of circulating adiponectin are associated with multiple myeloma risk in overweight and obese individuals . Cancer Res . 2016 ; 76 ( 7 ): 1935 – 1941 . 10.1158/0008-5472.CAN-15-2406/652213/AM/LOW-LEVELS-OF-CIRCULATING-ADIPONECTIN-ARE 26921332 PMC4878138

48 Ferlin M , Noraz N , Hertogh C , Brochier J , Taylor N , Klein B . Insulin‐like growth factor induces the survival and proliferation of myeloma cells through an interleukin‐6‐independent transduction pathway . Br J Haematol . 2000 ; 111 ( 2 ): 626 – 634 . 10.1046/J.1365-2141.2000.02364.X 11122111

49 Ge NL , Rudikoff S . Insulin‐like growth factor I is a dual effector of multiple myeloma cell growth . Blood . 2000 ; 96 ( 8 ): 2856 – 2861 . 11023522

50 Qiang YW , Yao L , Tosato G , Rudikoff S . Insulin‐like growth factor I induces migration and invasion of human multiple myeloma cells . Blood . 2004 ; 103 ( 1 ): 301 – 308 . 10.1182/blood-2003-06-2066 14504085

51 Moreau P , Cavallo F , Leleu X , Hulin C , Amiot M , Descamps G , et al. Phase I study of the anti insulin‐like growth factor 1 receptor (IGF‐1R) monoclonal antibody, AVE1642, as single agent and in combination with bortezomib in patients with relapsed multiple myeloma . Leukemia . 2011 ; 25 ( 5 ): 872 – 874 . 10.1038/LEU.2011.4 21321571

52 Khan S , LeBlanc R , Gyger M , White D , Kaufman J , Jazubowiak A , et al. A phase‐1 trial of linsitinib (OSI‐906) in combination with bortezomib and dexamethasone for the treatment of relapsed/refractory multiple myeloma . Leuk Lymphoma . 2021 ; 62 ( 7 ): 1721 – 1729 . 10.1080/10428194.2021.1876864 33509009

53 Macaulay VM , Middleton MR , Protheroe AS , Tolcher A , Dieras V , Sessa C , et al. Phase I study of humanized monoclonal antibody AVE1642 directed against the type 1 insulin‐like growth factor receptor (IGF‐1R), administered in combination with anticancer therapies to patients with advanced solid tumors . Ann Oncol . 2013 ; 24 ( 3 ): 784 – 791 . 10.1093/ANNONC/MDS511 23104723 PMC3574548