A systematic review of literature on Insulin‐like growth factor‐2‐mediated hypoglycaemia in non‐islet cell tumours

Abstract Introduction Insulin‐like growth factor‐2 (IGF‐2)‐mediated hypoglycemia is a rare yet clinically significant entity with considerable morbidity and mortality. Existing literature is limited and fails to offer a comprehensive understanding of its clinical trajectory, management and prognostication. Methods Systematic review of English‐language articles reporting primary patient data on IMH was searched using electronic databases (PubMed, Scopus and Embase) from any date up to 21 December 2022. Data were analysed in STATA‐16. Results The systematic review contains 172 studies, including 1 Randomised controlled trial, 1 prospective observational study, 5 retrospective observational studies, 150 case reports, 11 case series and 4 conference abstracts. A total of 233 patients were analysed, averaging 60.6 ± 17.1 years in age, with comparable proportions of males and females. The commonest tumours associated with Insulin‐like Growth Factor‐2‐mediated hypoglycaemia were fibrous tumours (N = 124, 53.2%), followed by non‐fibrous tumours originating from the liver (N = 21, 9%), hemangiopericytomas (N = 20, 8.5%) and mesotheliomas (N = 11, 4.7%). Hypoglycaemia was the presenting feature of NICT in 42% of cases. Predominant clinical features included loss of consciousness (26.7%) and confusion (21%). The mean IGF‐2 and IGF‐1 levels were 882.3 ± 630.6 ng/dL and 41.8 ± 47.8, respectively, with no significant correlation between these levels and patient outcomes. Surgical removal was the most employed treatment modality (47.2%), followed by medication therapy. The recovery rate was 77%, with chronic liver disease (CLD) significantly associated with a poor outcome (OR: 7.23, P: 0.03). Tumours originating from fibrous tissues were significantly associated with recovery (p < .001). In the logistic regression model, CLD remained a significant predictor of poor outcomes. Conclusion This systematic review highlights that most non‐islet‐cell tumour‐hypoglycaemia (NICTH) is due to fibrous tumours. NICTs demonstrate a variable prognosis, which is fair if originating from fibrous tissue. Management such as octreotide, corticosteroids, diazoxide, embolization, radiotherapy and surgical resection have disparate success rates.


| INTRODUC TI ON
Hypoglycaemia is commonly defined as a plasma glucose concentration of less than 70 mg/dL, although signs and symptoms may not appear until plasma glucose concentrations fall below 55 mg/dL (3 mmol/L). 1 Since 1938, the Whipple triad has been used to confirm spontaneous hypoglycaemia before proceeding with further diagnostic workup.Whipple triad criteria include (1)   low plasma glucose concentration, (2) signs and symptoms due to hypoglycaemia and (3) remission of the hypoglycaemic symptoms after increasing plasma glucose concentration. 2,3According to the current Endocrine Society Clinical Practice Guidelines, hypoglycaemia causes can be divided into two categories based on clinical status: (1) patients who appear to be healthy (endogenous hyperinsulinemia, insulinoma, post-bariatric hypoglycaemia, autoimmune insulin hypoglycaemia, malicious hypoglycaemia, hypoglycaemia due to antibodies to insulin and insulin receptors), and (2) patients who look to be unwell or medicated (malnutrition, IGF-2-secreting non-islet cell tumour, drugs, alcohol, hypopituitarism and hypoadrenalism). 2 Hypoglycaemia occurs uncommonly in the non-diabetic population.Some causes include critical illness, adrenal insufficiency, deficiency of other hormones in glucose metabolism, certain medications (salicylates, sulfa drugs, quinine), stomach surgeries, certain enzyme deficiencies and alcohol intake. 4Neoplastic processes within the pancreas can instigate hypoglycaemic conditions by an unchecked overproduction of insulin.Besides these insulinomas or insulin-secreting tumours within the pancreas, tumours in other body regions may also precipitate hypoglycaemia.These non-pancreatic tumours induce hypoglycaemia through an unrestrained production of Insulin-like Growth Factor-2 (IGF-2). 4IGF-2 predominantly contributes to embryonic development, functioning as a pivotal regulatory element.However, in adult physiology, IGF-2 is present in modest quantities, particularly within skeletal muscle tissue, facilitating tissue proliferation.Moreover, IGF-2 can be detected in increased amounts in injured tissue, where it functions to aid the wound healing process. 5IGF-2 typically demonstrates weak affinity for insulin receptors under normal physiological conditions.However, abnormally elevated levels of IGF-2 can result in sufficient stimulation of insulin receptors, potentially inducing a state of clinically significant hypoglycaemia.This is mainly observed when IGF-2 secretion is unchecked, such as in malignancies. 6 clinical practice, non-islet cell tumour hypoglycaemia (NICTH) is considered a rare condition, with an incidence rate of one per million person-years. 7,8The true prevalence and clinical impact of hypoglycaemia induced by NICTH remains uncertain in the current medical literature.However, it is estimated to be four times less prevalent than insulinoma. 9NICTH is a paraneoplastic syndrome that manifests through excess IGF-2-led insulin receptor activation, causing hypoglycaemia. 7,10NICTH was first identified in patients with hepatocellular cancer in 1929.Since then, several different forms of neoplasms have been linked to NICTH.In approximately half of the patients with IGF-2-mediated hypoglycaemia, the tumour is identified before developing hypoglycaemia. 11In NICTH, workup typically reveals low insulin, beta-hydroxybutyrate and C-peptide levels, suggesting an insulinmimetic effect in the absence of excess insulin levels, which is an indication to evaluate IGF-1 and IGF-2 levels. 12A ratio of IGF-2 and IGF-1 in a healthy individual is 3:1.A molar ratio larger than 10 in recurrent hypoglycaemia supports NICTH. 7Given the relative infrequency of testing for IGF-1, IGF-2 and pro-IGF-2 testing, the diagnosis of NICTH poses considerable challenges and is often overlooked, which may lead to significant clinical consequences.
However, early recognition of NICTH, followed by the implementation of tumour-targeted and systemic therapeutic approaches, can aid in mitigating its poor prognosis. 13Diagnosing paraneoplastic IGF-2-induced hypoglycaemia is important and time-sensitive, as early identification of the underlying tumour permits comprehensive surgical resection.This, in turn, offers the potential for complete resolution and cure of the ensuing hypoglycaemia.A growing body of evidence supports screening for IGF-2 in nondiabetic persons with hypoglycaemia and suppressed insulin and c-peptide levels. 14n-islet cell tumour hypoglycaemia remains a rare but serious condition with high morbidity and mortality risk.Most literature on NICTH consists of case series and small retrospective studies representing variable results.Current literature is devoid of larger, systematic investigations that thoroughly elucidate the clinical trajectory and prognostic understanding of NICTH.Hence, the prevailing guidelines offer insufficient guidance regarding the optimal therapeutic strategies for managing IGF-2-mediated hypoglycaemia.We conducted a systematic review to generate cumulative evidence on the epidemiological characteristics, clinical trajectory, therapeutic strategies and resultant outcomes associated with IGF-2-mediated hypoglycaemia.

| Study selection
The articles were exported from the databases to endnote 9 and then transferred to Rayyan AI for screening.HC and AAK screened the retrieved articles independently, initially from the title and abstract, followed by full-text review of the shortlisted studies.FA reviewed the disputed articles independently to resolve conflicts.

| Inclusion criteria
The inclusion criteria were English-language original studies reporting primary data (case reports, series, observational retrospective, prospective studies and clinical trials) regarding IGF-2-mediated hypoglycaemia.

| Exclusion criteria
Studies reporting IGF-1 as a cause of hypoglycaemia or studies mentioning causes of hypoglycaemia other than IGF-2 were excluded from the review.Studies reporting high IGF-2 levels without hypoglycaemia were also excluded.Animal studies on IGF-2-mediated hypoglycaemia were also excluded from this review.Finally, studies reporting secondary patient data were also excluded.

| Quality assessment
The AAK and IK performed quality assessment of the included studies.Case reports and series were assessed using the Joanna Briggs Institute case report appraisal checklist for inclusion in systematic reviews. 15Observational studies were assessed using the methodological index for non-randomised studies (MINORS) scoring system. 16Finally, the quality of clinical trials was assessed using the Cochrane Collaboration's tool. 17FA resolved the conflicts via an independent quality assessment in case of disagreements.

| Data collection and statistical analyses
Four reviewers collected baseline demographic details, including age, gender and comorbidities of the patients.Data were double-checked for accuracy by two reviewers who were not initially involved in the data collection.Symptoms of hypoglycaemia were collected.The data of the initial hypoglycaemia workup was collected, including glucose levels with corresponding insulin (μIU/mL), C-peptide (ng/mL), beta-hydroxybutyrate (mmol/L), proinsulin, oral hypoglycaemic screen and insulin antibody results.
The lowest pre-treatment glucose level was also extracted.All data was collected and reported where available.We collected the highest pre-treatment levels of IGF-1 (ng/mL), IGF-2 (ng/mL) and IGF binding protein (IGFBP).Data on big IGF-2 molecules was also collected.Tumour-related data included the type, location and size of tumours, radiological details, biopsy and histopathology findings of the tumours.We encountered a diverse range of cancers/ tumours, necessitating a categorization approach for analysis.Due to the wide variety of tumour types, it was impractical to analyse and summarise each individually.Therefore, we classified them into two broad categories: fibrous and non-fibrous tumours.The fibrous category encompassed numerous fibromatous tumours, regardless of the region.Non-fibrous tumours included primary cancers or tumours based on the region.Treatment data included surgery, embolisation, radiotherapy, steroid use and other medical management.Outcome data included tumour recurrence, hypoglycaemia and in-hospital mortality.No imputation was made to address the missing data.We followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines for reporting systematic reviews in this study. 18Data were analysed using R Studio version 2023.03.0 (build 386).The Mann-Whitney U test and Chi-square test were used as appropriate.A logistic regression model was also developed, as discussed in the results section.We reported results as means with standard deviation, medians with interquartile ranges, or numbers with percentages as appropriate.

| RE SULTS
A total of 172 studies were included in this systematic review (Figure 1).These included one randomised controlled trial, one prospective observational study, five retrospective observational studies, 150 case reports, 11 case series and four conference abstracts (Data S1).

| Baseline characteristics
A total of 233 participants were identified, with 125 males and 108 females.The mean age was 60.6 ± 17.1 years.Hypoglycaemia was identified as the presenting feature of NICT in 42% (98) of patients.Loss of consciousness and confusion were the most common features in the patients, present in 26.7% and 21%, respectively.Weight loss, dizziness and weakness were other relatively less common features.Hypertension and diabetes were the two most common comorbidities prevalent in 11.6% and 6.9% of patients, respectively.Chronic liver disease (CLD), heart failure and chronic kidney disease were found in 2.6%, 2.1% and 1.3% of patients.

| Diagnosis of IGF-II mediated hypoglycaemia
The average IGF2 and IGF1 levels in this study were 882.3 ± 630.6 ng/dL and 41.8 ± 47.8, respectively.The IGF 2:1 ratio was 192.8 +/− 1704.9, but this was varied due to three outlier values above 500.They were 500.7,6670 and 22,150 in 3 cases.
Excluding these three outliers, the average ratio was 32.3 ± 36.3 (Table 1).Cancers originating from fibrous tissues (fibroma and fibrosarcoma) were the most associated cancers, comprising 123 cases, with gastrointestinal cancers being the second most common.

| Outcome analysis
Outcomes were available for 223 cases, and the overall recovery rate among all the participants was 77% (n = 172).There was no significant difference between recovered and non-recovered patients regarding gender, age, clinical presentations, or comorbidities except for chronic liver disease (Table 2).Chronic liver disease (CLD) was more likely to be associated with death than recovery (OR: 7.23, P: 0.03).Of the 233 patients, only six had CLD.These patients had hepatocellular carcinoma (N = 4), gastrointestinal stromal tumour (N = 1) and metastatic hemangiopericytoma (N = 1).IGF values at presentation or IGF 2:1 ratio were not associated with any outcome, as shown in Table 2. Fibrous tumours were significantly associated with recovery compared to non-fibrous ones, and the association was statistically significant (65.1% vs. 33%, p < .001).

| Comparison of IGF 2:1 ratio
Another analysis with the IGF 2:1 ratio used as a dependent variable was done to see if it was associated with our study variables (Table 3).Overall, the IGF 2:1 ratio was not significantly different between genders, presentations, comorbidities, or treatment groups.
However, patients with hypoglycaemia as the first presentation of NICT had a lower IGF 2:1 ratio than those without (Median: 18.9 vs. 23.4), and the association was statistically significant (p = .046).

| Logistic regression outcome model
A logistic regression model with the outcome as a dependent variable was developed using the IGF 2:1 ratio, tumour types and chronic liver disease as independent variables (Table 4).Tumour category and chronic liver disease were still significant predictors of outcome while adjusting each other and for IGF 2:1 ratio (p = .001and p = .03,respectively) as seen in Table 4.

| DISCUSS ION
This systematic review focuses on IGF-2-mediated hypoglycaemia, encompassing a wide range of underlying tumours.Of the 233 patients, most had an underlying fibrous tumour (53.2%).Generally, the patients were older adults (mean age 60.6 years).After excluding the outliers, the mean IGF-2:IGF-1 ratio was 34.9 ng/dL.Surgical intervention emerged as the most prevalent treatment modality, employed in 47.2% of cases, followed by steroid therapy (39.1%).
Other treatments included embolisation, radiotherapy and other drugs such as octreotide and diazoxide.Most patients successfully recovered (77%); however, a significant mortality rate was identified (23%).Multivariate logistic regression analysis identified chronic liver disease as a significant predictor of adverse outcomes, while a fibrous origin of the tumour was associated with a more favourable prognosis in patients with IGF-2-mediated hypoglycaemia.
A complex interplay of multiple mechanisms underlies the pathophysiology of hypoglycaemia in patients with NICT.The primary driver is the tumours' excessive production of immature and mature IGF-2 molecules. 7Other mechanisms that can play an essential role The diagnosis of IGF-2-mediated mechanism behind NICTH is sometimes complex and needs careful evaluation and ruling out other more common causes of hypoglycaemia.A raised IGF-2 to IGF-1 ratio from a normal 3:1 to 10:1 has been referred to confirmation for NICTH secondary to IGF-2 action. 7Our study corroborates this observation, revealing a ratio of IGF-2 to IGF-1 exceeding 10 in 87.6% of the patient cohort.In diagnostically challenging scenarios, this elevated ratio is a valuable confirmatory marker, facilitating The values represent the ones after excluding the outliers.
managing NICTH for complex cases may be of considerable significance.Interestingly, our analysis shows that higher levels of IGF2 alone might not be clinically significant when reviewed alone.Our analysis of the IGF-2 to IGF-1 ratio disclosed lower ratios in patients for whom hypoglycaemia constituted the initial clinical presentation.Moreover, a multivariate logistic regression model evaluating outcomes failed to identify IGF-2 levels as a statistically significant predictor of mortality.This highlights the imperative of a holistic evaluation of individual cases when formulating subsequent management strategies.While IGF levels may aid in the diagnosis of complex cases, our findings indicate that prognostic assessments cannot be predicated solely upon these biochemical markers.
Previous investigations involving smaller cohorts have posited that approximately half of the patients with NICT may present with hypoglycaemia as an initial clinical manifestation. 6,19The current investigation also shows that a sizable proportion (42%) of patients A multimodal management approach is sometimes required, including specialised and multidisciplinary care. 23In the management of IGF2-mediated hypoglycaemia, particularly in challenging cases of tumour-induced hypoglycaemia unresponsive to conventional therapies, Pasireotide might be a future option once more evidence in its broader efficacy is available. 24This secondgeneration somatostatin receptor analogue, with its enhanced affinity for somatostatin receptor subtype 5 (SSTR5), effectively inhibits excessive insulin secretion, one of the factors in insulinoma and NICTH. 25,26Evidence regarding its use in NICTH is extremely limited, although it is plausible due to its potential to suppress IGF-2.
Pasireotide 's dual anti-hypoglycaemic and antitumor properties may make it a potentially favourable first-line treatment option, especially when surgical interventions are not possible or other medical treatments prove ineffective. 24e literature lacks studies investigating factors associated with mortality in patients with IGF-2-mediated hypoglycaemia.
Multivariate logistic regression analysis in our study indicated that fibrous tumours are associated with a more favourable prognosis in terms of mortality outcomes.This aligns with the general understanding of fibrous tumours, which have a better prognosis than other types of tumours. 27Furthermore, our analysis identified CLD as a significant predictor of mortality.The occurrence of hypoglycaemia in the context of CLD is an established determinant of adverse outcomes. 28However, how CLD can worsen prognosis in patients with IGF-2-mediated hypoglycaemia is not studied.As gluconeogenesis and glycogenolysis can be compromised in patients with CLD, the effects of IGF-2-mediated hypoglycaemia can potentiate and might lead to severe manifestations and worse outcomes. 29 light of a high mortality rate, approximating 23%, identifying and comprehensively evaluating prognostic factors is paramount and necessitate further investigative efforts.

| Strengths and limitations
The primary strength of this investigation lies in its comprehensive, systematic appraisal of data pertaining to IGF-2-mediated hypoglycaemia.The infrequent incidence of non-islet cell tumour hypoglycaemia poses challenges to the execution of large-scale, prospective study designs, further underscoring the significance of our work.
Consequently, the synthesis of previously published data in our study enhances the understanding of this phenomenon, providing a clearer, more rigorous framework for both clinical interpretation and future research endeavours.Nevertheless, it is imperative to acknowledge the limitations inherent to this study, which predominantly stem from conducting a systematic review based on observational data.This review included only those patients with NICT who developed hypoglycaemia; hence, the prevalence or burden of NICT as a whole could not be calculated.Secondly, the potential impact of publication bias must be acknowledged, given that unpublished studies could have influenced the aggregated findings of this systematic review.Furthermore, the exclusion of certain studies from this review due to the unavailability of individual-level data constitutes an additional limitation that could potentially impact the comprehensiveness and generalizability of our findings.Nevertheless, this review highlights the clinical trajectory of IGF-2-mediated hypoglycaemia and identifies key prognostic factors.It serves as a foundational resource, paving the way for future research to identify factors that could enhance diagnostic accuracy and therapeutic management.
in the development of clinically significant hypoglycaemia include increased uptake of glucose by the tumour cells (especially if of muscle origin) and replacement of liver tissue responsible for as a presenting feature.11Irrespective of the underlying etiological mechanisms, hypoglycaemia stands as the most clinically salient manifestation of these tumours.This necessitates concentrated research efforts to elucidate optimal management guidelines for managing general and cause-specific hypoglycaemia.A diverse array of anatomical and pathological tumour types has been implicated in the aetiology of IGF-2-mediated hypoglycaemia, with frequencies varying considerably across existing literature.In a retrospective study on 78 patients with NICTH, Fukuda et al. reported the highest prevalence of liver and gastric malignancies.19The study was not included in this systematic review due to insufficient data.While the mean age of the patient cohort in our analysis aligns with that reported in the retrospective study, we observed significant disparities, most notably in the predominant malignancies causing NICTH.Our study identified fibrous tumours as the leading aetiology, accounting for 53.2% of NICTH cases.In contrast, the study by Fukuda et al. reported a markedly lower prevalence, with only 5% of the tumours being of fibrous origin.However, older reviews have documented fibromas as one of the most common aetiologies of NICTH.20 Historically, discrepancies in the literature have been attributed to small sample sizes, introducing potential bias.With an evaluation of 233 patients, our study constitutes the most extensive systematic dataset to date, thereby corroborating and solidifying prior findings regarding the most prevalent tumour types associated with non-islet cell tumour hypoglycaemia (NICTH).
had hypoglycaemia at the initial presentation.Nonetheless, the prognostic implications of hypoglycaemia as an initial clinical presentation remain ambiguous and warrant further prospective evaluation to ascertain its prognosis.The treatment of IGF-2-mediated hypoglycaemia can be divided into two phases: acute management, which can be similar to hypoglycaemia of other causes, and definitive treatment by focusing on reduction in tumour function.Bodnar et al. 21generated a narrative review focused on the management of NICTH in 2014.The treatment options reported by Bodnar et al. persist, and we did not find any new management strategy in the current study.The selection of therapeutic modalities ranging from surgical intervention to localised tumour-targeting approaches such as embolisation and radiotherapy, as well as pharmacological agents including corticosteroids, diazoxide, growth hormone and octreotide, is principally guided by the type, anatomical location and extent of the tumour.

TA B L E 1
Baseline clinical characteristics of patients with IGF-2 mediated hypoglycaemia (N = 233).
Type and frequency of various histological stains reported in patients with IGF-2-mediated hypoglycaemia (N = 88).Outcome analysis (based on mortality vs. recovery) among patients with IGF-2-mediated hypoglycaemia.Data presented as mean (SD), median (IQR) and numbers (%) as appropriate.p values in bold are statistically significant.
Abbreviations: IGF, Insulin like growth factor; IQR, Interquartile range, Std.Dev, Standard Deviation.a IGF-2-mediated hypoglycaemia is a rare manifestation of nonislet cell tumours.This elaborative systematic review highlights that most non-islet cell tumours hypoglycaemia (NICTH) is due to fibrous tumours.Furthermore, NICTH-related tumours demonstrate a variable prognosis, which is fair if they originate from fibrous tissue.A notable mortality rate, approximating one-fourth of the patients, highlights significant healthcare implications and necessitates urgent attention to clinical management strategies.Various therapeutic modalities have been employed, encompassing pharmacological interventions such as octreotide, corticosteroids and diazoxide, as well as embolization, radiotherapy and surgical resection, with variable success rates.The inherent rarity of IGF2mediated hypoglycaemia presents unique challenges for traditional prospective research methodologies needed to generate high-level evidence.Recognising this, it is more practical to utilise tailored statistical approaches tailored to rare disease research, such as case-control studies, with an in-depth investigation of fewer cases.Additionally, utilising patient registries and real-world data may provide valuable longitudinal information.