SGLT2 inhibitor use reduces progression and surgical intervention of persistent pulmonary nodules

Abstract

Objective:

In-situ and early-stage cancers are the fastest growing subset of pulmonary malignancies, often presenting as persistent nodules and ground glass opacities. These lesions are frequently surveilled, though many progress and require interventions. We hypothesized that sodium-glucose cotransporter-2 inhibitors (SGLT2i) may slow progression of early pulmonary malignancy.

Methods:

In this retrospective cohort study of 6,362 patients, 200 adults with type 2 diabetes mellitus (T2DM) and a pulmonary nodule suspicious for lung cancer were matched based on exposure to an SGLT2i or an alternative T2DM medication. Primary outcomes included nodule progression (growth > 2 mm or solid component development) and need for intervention (biopsy, surgery, radiotherapy, or chemotherapy). Multivariable Cox regression and log-rank tests were used for analysis.

Results:

The SGLT2i and non-SGLT2i groups had similar comorbidity profiles, initial nodule sizes and surveillance durations. After a median follow-up of 28 months (range: 3–125), 12% in the SGLT2i group versus 24% in the non-SGLT2i group had progression (p = 0.04). Adjusting for covariates, SGLT2i use was associated with reduced risk of nodule progression (HR = 0.38, 95% CI: 0.18–0.80) and longer mean time to progression (37.7 versus 29.2 months). Surgical intervention was less frequent in the SGLT2i group (3% versus 13%, p < 0.01), with prolonged time to intervention (HR: 0.18, 95% CI: 0.07–0.47).

Conclusion:

SGLT2i use was associated with lower risk of pulmonary nodule progression and surgical intervention among patients with T2DM, independent of comorbidities, glycemic control, and baseline nodule size. These findings suggest a potential role for SGLT2i in reducing growth of persistent pulmonary nodules.

1. Introduction

Pulmonary nodules are identified in approximately 1.6 million people in the United States per year and are present in as many as 30% of chest computed tomography (CT) scans. [1] While most of these nodules are benign granulomas or lymph nodes, many persistent nodules are premalignant and in situ lesions that eventually progress to invasive lung cancer. [2] Patients with persistent nodules are often followed expectantly, with serial imaging to evaluate for progression. [3] Despite the indolent growth of these lesions, patients with ground glass opacities often are referred for surgical interventions including biopsies and lung resections, sometimes multiple times in patients with multiple lesions, which carries added risk. At present, no medical therapy exists to address these lung cancers at their earliest stages, and this represents an important window for early management as an alternative to surgical intervention.

Emerging evidence has demonstrated that sodium-glucose cotransporter-2 (SGLT2), is a key glucose transporter in early-stage adenocarcinomas. [4,5] SGLT2 is a sodium-coupled glucose transporter that is physiologically expressed on the proximal renal tubule and functions in glucose reuptake. Recent studies have shown that SGLT2 is upregulated in multiple cancers, including lung, prostate, and pancreatic adenocarcinoma, and is particularly expressed in well-differentiated, early-stage adenocarcinoma. [5,6] Upregulation of SGLT2 in early-stage pulmonary adenocarcinoma provides a potential therapeutic target to reduce glucose uptake, possibly limiting growth.

SGLT2 inhibitors, an FDA-approved class of anti-diabetes medication that lowers blood glucose by inhibiting reuptake of urinary glucose by inhibiting SGLT2 transporters, have been suggested to have anti-cancer effects. Preclinical studies have demonstrated that SGLT2 inhibitors attenuate tumor growth in murine lung adenocarcinoma models [5,7] through various pathways including glucose deprivation and PD-L1 degradation. Luo et al. performed a large-scale retrospective study that suggests that administration of an SGLT2 inhibitor was associated with improved survival of lung cancer patients.[8] We hypothesized the administration of an SGLT2 inhibitor would slow the growth of pulmonary nodules associated with early-stage pulmonary malignancies and thereby lead to fewer overall surgical interventions. This study aimed to compare incidence of pulmonary nodule progression and surgical intervention in diabetic patients with persistent pulmonary nodules who were prescribed an SGLT2 inhibitor versus those prescribed an alternative anti-diabetic medication.

2. Methods

2.1. Data sources and design

This study was a single-institution, retrospective matched cohort study designed to evaluate the impact of SGLT2 inhibitor administration on the growth and intervention rates of persistent pulmonary nodules in patients with T2DM. The study population was identified using the electronic medical record (EMR) at the University of Pennsylvania. A study flow chart is shown in Fig. 1. Patients included in the study were adults aged 18 years or older with a diagnosis of T2DM and a pulmonary nodule between 2014 and 2024. The start date of 2014 was selected because canagliflozin, the first FDA-approved SGLT2 inhibitor, was approved in 2013, with minimal early adoption in that year.[8].

Fig. 1.

Flowchart of inclusion and exclusion of patients with T2DM and a persistent pulmonary nodule. T2DM = Type 2 diabetes mellitus; SGLT2 = sodium-glucose cotransporter-2; CT = computed tomography.

An initial sample of 6,362 patients were identified, of whom 526 (8.3%) had a documented SGLT2 inhibitor exposure, while 5,836 (91.7%) had no exposure. Based on previously reported studies examining the effect of SGLT2 inhibitors on lung cancer, a sample size calculation determined that 100 patients per group would be required to detect a 60% reduction in pulmonary nodule growth (assumed approximately 30%) with 80% statistical power and a significance level of 5%. Therefore, a cohort of 100 patients with an SGLT2 inhibitor exposure was identified and included in the study, followed by the selection of a matched control group of 100 patients without an SGLT2 inhibitor exposure.

Patients with a documented SGLT2 inhibitor exposure were evaluated sequentially for inclusion. To be eligible, patients were required to have a diagnosis of T2DM, a persistent pulmonary nodule identified on thin-cut computed tomography (CT) scan for a duration of at least three months, and a prescription for an SGLT2 inhibitor for at least three months. Patients were excluded if the period of SGLT2 inhibitor use did not overlap with the period in which the pulmonary nodule was documented on CT. Additionally, patients were excluded if the nodule was intervened on within three months of initial pulmonary nodule detection, as such cases were deemed highly suspicious for malignancy, thereby precluding watchful waiting.

Once 100 patients with an SGLT2 inhibitor exposure were identified, patients without an SGLT2 inhibitor exposure were evaluated sequentially and matched 1:1 to the SGLT2 inhibitor group based on age (<65, 65–74, 75+), sex (male/female), race (non-Hispanic White, Black/African American, Asian/Pacific Islander, Other), and smoking history (yes/no). This study was conducted in accordance with the principles outlined in the Declaration of Helsinki and was deemed exempt by University of Pennsylvania Institutional Review Board (IRB). Given the retrospective nature of the study, the requirement for written informed consent was waived by the IRB.

2.2. Data collection

Demographic data were collected for each patient including age, sex, and race/ethnicity (including non-Hispanic White, non-Hispanic Black or African American, non-White Hispanic, or Asian/Pacific Islander). Chronic conditions including diabetes, hypertension, chronic kidney disease, cardiovascular disease, congestive heart failure, and a history of malignancy were extracted from the EMR. Smoking status and blood glucose control as measured by hemoglobin A1c (HbA1c) at the time of diagnosis of the pulmonary nodule were also recorded.

Pulmonary nodule characteristics, including nodule size (as measured by diameter calipers on the longest axis) and radiographic density (categorized as pure ground glass, part-solid, or solid) were obtained from radiology reports and confirmed by two independent investigators (KOW and LC). In cases where multiple nodules were present, the largest or morphologically most suspicious nodule was selected for data collection, in accordance with the Fleischner Society recommendations for pulmonary nodule assessment.[9].

2.3. Outcomes

Primary outcomes of the study were the occurrence of pulmonary nodule progression and intervention on the pulmonary nodule following surveillance. Pulmonary nodule progression was defined as an increase in size greater than 2 mm on serial CT scans to account for expected inter-scan variability, or the growth of a solid component suggesting the development of invasive disease. An intervention on the pulmonary nodule was defined as biopsy, surgical resection, or the initiation of empiric radiotherapy, immunotherapy, or chemotherapy.

Secondary outcomes included time to pulmonary nodule progression, measured as the interval between the initial CT scan on which the pulmonary nodule was detected and the scan demonstrating nodule progression and censored at the date of the last CT scan if no nodule progression was seen. The time to intervention was defined as the interval from the date of pulmonary nodule diagnosis to either the date of the intervention or censored at the study analysis cutoff date of June 1, 2024, whichever occurred first. When available, pathology results of the resected or biopsied nodule were recorded.

2.4. Statistical analysis

Patients were matched 1:1 based on demographic characteristics including age, sex, race, and smoking history as delineated above. Covariates, including the presence of chronic conditions such as hypertension, coronary artery disease, chronic kidney disease, congestive heart failure, and a history of malignancy, as well as HbA1c and initial pulmonary nodule size were compared between the groups using Pearson’s chi-squared test for categorical variables and t-tests for continuous variables.

To evaluate the primary outcomes of nodule progression and intervention, the proportion of patients with pulmonary nodule progression and intervention were compared between patients with and without an SGLT2 inhibitor exposure using Pearson’s chi-square test. The distribution of the time to pulmonary nodule progression and time to intervention was assessed using Kaplan-Meier estimates and compared between patients with and without an SGLT2 inhibitor exposure using log-rank tests.

Cox proportional hazards regression models were used to estimate the adjusted hazard ratios for experiencing pulmonary nodule progression in relation to SGLT2 inhibitor use (yes vs. no) after controlling for potential confounders including concomitant use of metformin, the presence of chronic comorbidities, and the initial size of the pulmonary nodule. Proportional hazards assumptions were assessed graphically and did not demonstrate evidence of non-proportionality. All statistical analyses were performed using GraphPad Prism 8 (GraphPad Software, CA, USA). All tests were two-sided, and a p-value of less than 0.05 was considered statistically significant.

3. Results

A total of 6,362 patients aged 18+ with T2DM and a pulmonary nodule were identified in the EMR. Of these, 526 patients were prescribed an SGLT2 inhibitor, while 5,836 patients had no SGLT2 inhibitor exposure. 367 consecutive patients in the SGLT2 inhibitor group were assessed for study inclusion, and 100 patients were included. Patients in the non-SGLT2 inhibitor group were matched 1:1 by age, sex, race, and smoking history, resulting in 100 matched pairs. A full breakdown of exclusion reasons is shown in Fig. 1.

Table 1 summarizes the demographic and clinical characteristics of both groups. Patients using SGLT2 inhibitors were similar to non-users in their comorbidities including hypertension, chronic kidney disease, cardiovascular disease, and history of malignancy, but were more likely to have congestive heart failure (48% vs 24%, p < 0.001). Patients using SGLT2 inhibitors had a higher mean HbA1c compared to patients who did not receive SGLT2 inhibitors (mean ± SD: 7.71 ± 1.73 vs 7.13 ±1.51; p < 0.001) (Fig. 2). Mean length of SGLT2 inhibitor use in the treatment population was 28.7 (SD: 22.58 months). Both groups had comparable rates of metformin use, as well as similar initial pulmonary nodule sizes and surveillance durations.

Table 1.

Baseline Demographic and Clinical Characteristics of Study Patients. Data are presented as mean +/− standard deviation or No./Total (%).

	SGLT2i	No SGLT2i	P value
Age
<65	34/200 (17%)
65–74	62/200 (31%)
75+	104/200 (52%)
Sex
Female	136/200 (68%)
Male	64/200 (32%)
Race
Non-Hispanic White	70/200 (35%)
Black/African American	130/200 (65%)
Smoking History
Negative	48/200 (24%)
Positive	152/200 (76%)
Comorbidities
Hypertension	90/100 (90%)	85/100 (85%)	0.39
Chronic Kidney Disease	27/100 (27%)	28/100 (28%)	0.99
Coronary Artery Disease	41/100 (41%)	34/100 (34%)	0.38
Congestive Heart Failure	48/100 (48%)	24/100 (24%)	<0.001
History of Malignancy	25/100 (25%)	32/100 (32%)	0.35
Mean Duration of Surveillance (months) ± SD	39.8 ± 26.5	32.9 ± 24.7	0.06
Size of Nodule at Diagnosis (mm) ± SD	6.42 ± 5.38	6.69 ± 5.70	0.74
HbA1c at time of diagnosis ± SD	7.71 ± 1.73	7.13 ± 1.51	0.006
Metformin Use
Yes	74/100 (74%)	70/100 (70%)	0.63
No	26/100 (26%)	30/100 (30%)

Fig. 2.

Violin plots of initial nodule size, duration of surveillance, and HbA1c at time of pulmonary nodule diagnosis in patients taking SGLT2 inhibitors compared to alternative diabetes medications. SGLT2i = SGLT2 inhibitor; HbA1c = hemoglobin A1c.

The proportion of patients with pulmonary nodule progression was significantly lower in patients using SGLT2 inhibitors compared to patients who did not (12% vs 24%, p = 0.04). Log-rank testing demonstrated that the administration of an SGLT2 inhibitor was associated with a significantly greater time to pulmonary nodule progression (p = 0.004) (Fig. 3). Of the nodules in each group that progressed, the mean time to pulmonary nodule progression was 29.17 ± 18.89 and 37.67 ± 19.08 months in the non-SGLT2 inhibitor and the SGLT2 inhibitor groups, respectively. The Cox proportional hazards model was used to test the association between SGLT2 inhibitor use and pulmonary nodule progression and is summarized in Fig. 4. Adjusting for covariates, SGLT2 inhibitor use was associated with significantly reduced risk of pulmonary nodule progression (HR = 0.38, 95% CI = 0.18–0.80).

Fig. 3.

Kaplan-Meier curve for time to progression of pulmonary nodules in patients taking SGLT2 inhibitors compared to alternative diabetes medications.

Fig. 4.

Adjusted HR for pulmonary nodule progression, performed using a Cox proportional hazards model. HR = hazards ratio, HTN = hypertension, CKD = chronic kidney disease, CAD = coronary artery disease, CHF = congestive heart failure.

A significantly higher proportion of patients not receiving SGLT2 inhibitors underwent an intervention on their pulmonary nodule compared to patients receiving an SGLT2 inhibitor. Among the SGLT2 inhibitor group, 3% (3/100) of patients underwent an intervention, including 2 biopsies and 1 surgical resection. In contrast, 13% (13/100) of patients not receiving SGLT2 inhibitors underwent a surgical intervention, resulting in 6 biopsies, 5 surgical resections, and 2 cases of empiric radiotherapy. Nodule histopathology was comparable between groups, with adenocarcinoma identified on pathology in 66% of samples from patients receiving SGLT2 inhibitors and 61% of samples from the non-SGLT2 inhibitor group (p > 0.99) (Table 2). Log-rank testing demonstrated that the administration of an SGLT2 inhibitor was associated with a significantly longer time to requiring a pulmonary nodule intervention (HR: 0.18 [95% CI: 0.07–0.47]); p = 0.002) (Fig. 5).

Table 2.

Final histology of pulmonary nodules that underwent intervention, the majority of which were adenocarcinoma. SGLT2i = SGLT2 inhibitor; SBRT = stereotactic body radiation therapy.

SGLT2i (N = 3)	No SGLT2i (N = 13)
Adenocarcinoma (N = 2)	Adenocarcinoma (N = 8)
Papillary (N = 1)	Acinar (N = 3)
Acinar (N = 1)	Mucinous (N = 1)
	Solid with micropapillary (N = 2)
	Poorly Differentiated (N = 1)
	Unknown subtype (N = 1)
	Squamous Cell Carcinoma (N = 1)
Benign (N = 1)	Benign (N = 2)
acellular fluid (N = 1)	no tumor seen (N = 2)
	Unknown (N = 1)
	Empiric SBRT (N = 1)
	Other (N = 1)
	prostate adenocarcinoma (N = 1)

Fig. 5.

Kaplan-Meier curve for time to intervention on pulmonary nodules in patients taking SGLT2 inhibitors compared to alternative diabetes medications.

4. Discussion

This study demonstrates that patients with T2DM who were prescribed SGLT2 inhibitors exhibited a significantly lower incidence of pulmonary nodule growth and required fewer surgical interventions after adjusting for potential confounders including comorbidities, blood glucose control, metformin use, and initial nodule size. To our knowledge, this is the first study to specifically examine the effect of SGLT2 inhibitors on persistent pulmonary nodules.

These results contribute to a growing body of literature suggesting that SGLT2 inhibitors confer protective effects against cancer progression. [10–12] Prior clinical research has demonstrated an association between SGLT2 inhibitor use and improved survival in patients with NSCLC. [8] Furthermore, a Phase 1b prospective clinical trial demonstrated that dapagliflozin exerted tumor-suppressive effects in pancreatic ductal adenocarcinoma. [13] Epidemiological studies have linked SGLT2 inhibitor use to a decreased incidence of pancreatic adenocarcinoma [14] and prostate cancer. [15] Our study builds on these findings by demonstrating that SGLT2 inhibitors are associated with an inhibitory effect on premalignant or in situ lesions, as represented by persistent pulmonary nodules..

The mechanism by which SGLT2 inhibitors influence lung cancer progression remains incompletely understood. SGLT2 is a glucose transporter primarily expressed in the kidney, where it functions in urinary glucose reabsorption. [16,17] However, SGLT2 is aberrantly upregulated in several malignancies, including lung, prostate, pancreatic, hepatocellular, and breast cancers. [6,18–21] In lung cancer, SGLT2 expression is particularly elevated in early-stage lepidic and ground-glass lesions and is subsequently downregulated as tumors progress to more invasive stages, [5,22] suggesting that SGLT2 inhibitors may be most effective in suppressing tumor metabolism and thereby growth at early stages of disease development.

Preclinical studies have proposed several potential mechanisms through which SGLT2 inhibitors may exert anticancer effects. In NSCLC models, SGLT2 inhibitors have been shown to inhibit cell cycle progression[23,24] and suppress epidermal growth factor receptor (EGFR) kinase activity.[25] These agents may also disrupt cancer metabolism through inhibition of mitochondrial complex I[26] and AMPK-mediated cell cycle arrest and apoptosis.[4,19,23] Other research has implicated the activation of the STING pathway[14] and suppression of PD-L1 expression,[7] suggesting that SGLT2 inhibitors may enhance antitumor immune responses. Furthermore, studies have proposed that SGLT2 inhibitors exert their effects through SGLT2-independent mechanisms, such as blocking the nuclear translocation of β-catenin, which may disrupt oncogenic Wnt signaling.[22 It is possible that a combination of SGLT2-dependent glucose deprivation and SGLT2-independent pathways contribute to the observed reduction in pulmonary nodule progression, however further studies are needed to elucidate the mechanism of inhibition.

The findings of this study have important potential clinical implications. Pulmonary nodules are a common incidental finding, identified in 1.6 million individuals in the United States annually and present in up to 30% of chest CT scans.[27–29] Although many of these nodules represent benign processes, a subset progresses to NSCLC, most commonly lung adenocarcinoma. Current clinical guidelines recommend watchful waiting for nodules that do not meet criteria for immediate biopsy or resection, with periodic surveillance by serial CT scans. [9,30,31] However, a substantial proportion of these nodules eventually enlarge and require invasive interventions, including biopsy, surgical resection, or empiric radiotherapy.[2,32–35] Notably, there are currently no pharmacologic interventions available to prevent or delay the progression of pulmonary nodules. The results of this study suggest that SGLT2 inhibitors may serve as a novel therapeutic option for patients diagnosed with a pulmonary nodule. This would be particularly important for those who are poor surgical candidates or have multiple nodules requiring long-term surveillance, which poses challenges in terms of lung volume preservation. Given that our study population comprised individuals with persistent pulmonary nodules, it is likely that some nodules included in the analysis were benign.[2,32,36] Despite this, the two groups exhibited significant differences in both nodule growth and the need for intervention. By broadly including pulmonary nodules regardless of biopsy status, our findings are highly generalizable to patients who remain under serial radiologic monitoring for suspicious nodules.

Despite the promising nature of these findings, this study has several limitations. The retrospective, single-center design introduces inherent selection bias, and the relatively small sample size may limit the statistical power to detect smaller differences and the generalizability of the results. Additionally, the study population consisted exclusively of non-Hispanic White and Black/African American individuals, raising questions regarding the applicability of these findings to other racial and ethnic groups. Furthermore, while efforts were made to control for confounding variables, residual confounding cannot be entirely excluded. Cumulative tobacco exposure could not be fully quantified due to incomplete pack-year documentation. Additionally, the SGLT2 inhibitor group had higher rates of congestive heart failure and higher HbA1c levels, which may have influenced surgical candidacy. However, the effect of these comorbidities is uncertain, as recent studies have shown a reduction of cardiac inflammatory markers and arrhythmias with SGLT2 inhibitors in patients with diabetes and heart failure, suggesting an effect on morbidity.[37,38] Prospective clinical trials are needed to validate the findings in our study and further elucidate the mechanisms by which SGLT2 inhibitors modulate lung cancer progression.

In conclusion, this retrospective cohort study suggests that SGLT2 inhibitor use is associated with a decreased risk of pulmonary nodule growth and a reduced need for surgical intervention in patients with T2DM. These findings provide preliminary evidence supporting the potential role of SGLT2 inhibitors in preventing growth of persistent pulmonary nodules, which are linked with early lung cancer. Future research should focus on prospective studies to confirm these associations and explore the underlying biological mechanisms. The results of this study may help inform the design of clinical trials evaluating SGLT2 inhibitors as a targeted therapy for high-risk pulmonary nodules.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.lungcan.2026.109356.