Rethinking bone marrow biopsy: is PET/CT enough for staging low-grade follicular lymphomas?

Abstract

This study evaluates the diagnostic performance of staging 18F-fluorodeoxyglucose positron emission tomography/computed tomography (18F-FDG PET/CT) for detecting bone marrow involvement (BMI) in low-grade follicular lymphoma (FL) and its impact on the clinical necessity of bone marrow biopsy (BMB). We retrospectively analyzed patients newly diagnosed with low-grade (grade 1–2) FL who underwent both ¹⁸F-FDG PET/CT and BMB from 2010 to 2022 at two Chinese institutions. PET/CT’s diagnostic accuracy for BMI was assessed using BMB as gold standard. Clinical and imaging data were analyzed to identify risk factors for BMI. Among 171 patients, 27 had positive BMB results. PET/CT demonstrated 86.5% accuracy in detecting BMI. In patients with PET/CT-based stage I–II disease, all BMB results were negative. In contrast, among patients with PET/CT-based stage III–IV disease, BMB upstaged 13 to stage IV. In patients with advanced-stage disease as determined by PET/CT, four risk factors were significantly linked to BMB positivity: female sex, Eastern Cooperative Oncology Group performance score > 1, elevated beta2-microglobulin levels, and involvement of > 4 lymph node regions. Patients were stratified into low-, intermediate-, and high-risk groups with corresponding BMB-positive rates of 5.6%, 40.7%, and 68.8%, respectively. No survival differences were observed between BMB-positive and BMB-negative patients. PET/CT can safely and effectively substitute for BMB in the staging of early-stage, low-grade FL. However, in advanced-stage disease, BMB remains valuable for diagnosis. The rate of BMB positivity correlates with tumor burden -related risk factors.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00277-025-06487-8.

Background

Follicular lymphoma (FL) is the most common type of indolent lymphoma worldwide, comprising approximately 35% of all non-Hodgkin lymphomas [1]. Patients with FL have experienced a significant improvement in survival owing to advancements in molecular imaging, introduction of novel medications, such as rituximab, and innovative radiotherapy techniques [2–5]. The initial treatment strategy FL depends on the stage of the disease and the histologic grade of the tumor, which are important factors for predicting outcomes [6, 7]. Grades 1 and 2 follicular lymphoma (FL) are widely classified as low-grade FL, exhibiting less aggressive behavior compared to grade 3 FL [8, 9]. The slow progression, frequent relapses, and rare cure of this particular type of lymphoma [10].

Radiotherapy is a common treatment in the early stages of low-grade FL, whether used alone or in conjunction with systemic therapy [11, 12]. Approximately 50% of patients can achieve long-term complete remission and excellent disease control with this treatment approach [11]. Indeed, the initial staging is key in determining the best management strategy for FL. The typically suggested method for staging is positron emission tomography-computed tomography (PET/CT), as recommended in available guidelines [6, 7]. Additionally, bone marrow biopsy (BMB) is a standard staging procedure, largely considered the most reliable for assessing bone marrow involvement (BMI) in all lymphoma types [13–15]. However, multiple studies have suggested that PET/CT is sensitive enough to detect BMI in Hodgkin’s lymphoma and diffuse large B-cell lymphoma (DLBCL) [16–19]. In our recent study, we found that PET/CT imaging can be considered reliable for excluding BMI in early nasal-type extranodal natural killer/T-cell lymphoma [20]. However, its’ significance of detecting BMI in indolent lymphomas is still debated [19, 21–25].

A wide range of sensitivities have been reported for the assessment of BMI using PET/CT in FL [21, 26–32]. This variability may be explained by differences in BMI proportions, tumor heterogeneity, and applied methodologies [23, 33]. FL grading and staging are significant predictive factors that reveal biological behavior and different patterns of BMI [34–37]. Therefore, we conducted a comprehensive retrospective analysis to evaluate the diagnostic performance of staging ¹⁸F-fluorodeoxyglucose positron emission tomography/computerized tomography (¹⁸F-FDG PET/CT) in detecting bone marrow infiltration (BMI) in low-grade follicular lymphoma (FL) and its potential impact on the necessity of bone marrow biopsy (BMB).

Methods

From December 2010 to December 2022, data for patients aged ≥ 15 years with newly diagnosed low-grade (grade 1–2) FL at Fujian Medical University Union Hospital and Sun Yat-Sen University Cancer Center were retrospectively reviewed. Patients were included in the study if both BMB and ¹⁸F-FDG PET/CT were performed as part of the routine clinical staging and no malignancy other than FL was known to be present at the time of imaging. Patients were staged according to Ann Arbor staging system. The Follicular Lymphoma International Prognostic Index-1 (FLIPI) and FLIPI2 score, which were stratified for all patients as low, intermediate, or high risk were collected.

This study received comprehensive review and approval from the Fujian Medical University Union Hospital Ethics Committee (approval number: 2022WSJK019). Because this was a retrospective study, the requirement to obtain signed informed consent was waived.

Bone marrow biopsy and PET/CT imaging

All patients were diagnosed with low-grade FL in accordance with the criteria of the World Health Organization classification of lymphomas [38]. Prior to treatment, routine unilateral posterior iliac crest trephine biopsy was routinely performed in all patients. The BMBs were evaluated by experienced hematopathologists at each center. Results from immunohistochemistry and flow cytometry were considered with bone marrow histology to determine bone marrow positivity for lymphoma.

Patients were asked to fast for 4–6 h before ¹⁸F-fluorodeoxyglucose (¹⁸F-FDG) PET/CT scanning. Serum glucose concentrations were measured before FDG injection and were less than 200 mg/dl (11.1 mmol/l). ¹⁸F-FDG tracer was administered at a dosage of 3.7–4.44MBq/kg. Data acquisition by an integrated PET/CT system (Discovery LS; GE Medical Systems, Milwaukee, WI, USA) was performed within 45–70 min of tracer injection. A low-dose whole-body CT scan was performed followed by a PET scan from the base of the skull to the mid-upper segment of the femur. PET images were reconstructed using ordered subset expectation maximization reconstruction algorithms.

In the PET/CT evaluation, FDG uptake more than that of the liver based on the visual assessment, localized in the marrow, was considered positive (a criterion that is often used in several types of lymphoma [21, 39, 40]) (Fig. 1). These findings were not attributed to other causes, such as growth factor support, bone fractures, or increased uptake resulting from bone marrow biopsy. PET/CT-assessed BMI were characterized as isolated, multifocal, diffuse, and negative. The maximum standardized uptake value (SUVmax), normalized for body weight, was calculated from the most intense area of FDG uptake in the bone marrow. Pure diffuse FDG uptake without focal activity often represents reactive hyperplasia. PET/CT images were carefully reviewed by 2 senior nuclear medicine physicians.

Fig. 1.

Examples of FDG-PET/CT-based analyses of lymphomatous bone marrow lesions in patients newly diagnosed with FL. a PET/CT-positive for focal bone involvement, right iliac bone uptake ≥ liver, with BMB+. b PET/CT-negative for bone involvement, with BMB+. BMB, bone marrow biopsy

Statistical methods

The Mann–Whitney U test and Pearson’s χ² test were used to compare continuous and categorical variables, respectively. The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and diagnostic accuracy of PET/CT for detecting BMI were assessed while considering BMB results as the reference standard. Clopper–Pearson exact confidence intervals were calculated for sensitivity, specificity, PPV, NPV, and accuracy. We used logistic regression models to investigate the effects of clinical factors on BMI, which were significantly associated with BMI in univariate analysis. A forward entry–backward elimination method of selection was used; P values for entry in and removal from the logistic regression model were 0.05 and 0.10, respectively. we performed goodness-of-fit checks using the Hosmer-Lemeshow test. Covariates were handled as categorical. Overall survival (OS) was defined as the time from FL diagnosis until death from any cause or the last follow-up date. Progression-free survival (PFS) was defined as the time from FL diagnosis until the first occurrence of disease progression, relapse, death from any cause, or the last follow-up date. The Kaplan–Meier method and the log-rank test were used to estimate OS and PFS. Two-sided p < 0.05 was considered significant. All statistical analyses were performed using SPSS, version 26.0 (IBM Corp., Armonk, NY, USA), and R, version 4.1.2 (www.r-project.org).

Results

Baseline characteristics

This study involved 171 patients, with baseline clinical characteristics detailed in Table 1. The median age at diagnosis was 49 years (range: 30–76 years), with 44 (25.7%) patients older than 60 years. Sixty-nine (40.4%) patients had advanced-stage disease (stage III–IV) according to combined BMB and ¹⁸F-FDG PET/CT (hereafter termed “PET/CT”) results. More than four lymph node regions were involved in 49 (28.7%) patients; 27 (15.8%) patients had splenic invasion and 3 (1.8%) had liver invasion. The low-risk category of FLIPI and FLIPI2 comprised 101 patients (59.1%) and 78 patients (45.6%), respectively. BMI was confirmed in 27 patients (15.8%) through BMB, whereas PET/CT identified such involvement in 18 patients (10.5%). BMB-positive and -negative patients with BM involvement by PET/CT generally exhibited similar clinical features, including age, Eastern Cooperative Oncology Group (ECOG) performance score, Spleen involvement, B symptoms, FLIPI, etc., However, a higher proportion of BMB-positive patients were female (supplemental Table 1). Furthermore, when comparing the clinical characteristics between patients with positive and negative BMB results specifically within the stage IV FL group, the findings were statistically similar, with the exception of higher proportion of females and a greater involvement of more than four lymph node regions (supplemental Table 2).

Table 1.

Baseline demographic and disease characteristics

Characteristics	Cases no./Total no. (%)	Characteristics	Cases no./Total no. (%)
Sex		B symptoms
Male	85/171 (49.7)	Yes	18/171 (10.5)
Female	86/171 (50.3)	No	153/171 (89.5)
Age (years)		> 4 node areas involved
≤ 60	127/171 (74.3)	Yes	49/171 (28.7)
> 60	44/171 (25.7)	No	118/171 (69.0)
ECOG score		Missing	4/171 (2.3)
0–1	154/171 (90.1)	Liver involvement
≥ 2	17/171 (9.9)	Yes	3/171 (1.8)
Ann Arbor stage		No	168/171 (98.2)
I-II	102/171(59.6)	Spleen involvement
III-IV	69/171(40.4)	Yes	27/171 (15.8)
Bone marrow biopsy		No	144/171 (84.2)
Positive	27/171 (15.8)	HGB < 120 g/L
Negative	144/171 (84.2)	Yes	25/171 (14.6)
BMI on PET/CT		No	146/171 (85.4)
Yes	18/171 (10.5)	Elevated LDH
No	153/171 (89.5)	Yes	15/171 (8.8)
FLIPI-1		No	153/171 (89.5)
0–1	101/171 (59.1)	Missing	3/171 (1.8)
2	42/171 (24.6)	Elevated β−2-microglobulin
3–5	24/171 (14.0)	Yes	34/171 (19.9)
Missing	4/171 (2.3)	No	109/171 (63.7)
FLIPI-2		Missing	28/171 (16.4)
0–1	78/171 (45.6)	Long diameter of largest LN > 6 cm
2	20/171 (11.7)	Yes	28/171 (16.4)
3–5	14/171 (8.2)	No	92/171 (53.8)
Missing	59/171 (34.5)	Missing	51/171 (29.8)

Abbreviations: ECOG Eastern Cooperative Oncology Group; BMI Bone marrow involvement; LN Lymph nodes; PET/CT Positron emission tomography/computed tomography; FLIPI-1 Follicular Lymphoma International Prognostic Index-1; HGB Hemoglobin; LDH Lactate dehydrogenase; FLIPI-2 Follicular Lymphoma International Prognostic Index-2

Diagnostic performance of PET/CT to determine BM status

In this study, BMB results were positive in 27 patients and negative in 144 patients (Table 2). PET/CT identified positive BMI in 18 patients, accounting for 10.5% of the cohort, while 153 patients had negative results. Among those with positive PET/CT results, 11 (61.4% of the PET/CT-positive cases) also had positive BMB results. Conversely, in the subgroup of 153 patients who showed no abnormal skeletal fluordeoxyglucose uptake on PET/CT, 16 (10.5%) had positive BMB results.

Table 2.

Diagnostic performance of PET/CT and BMB for detecting bone marrow involvement

PET/CT		BMB		Total
		Positive	Negative
Positive		11	7	18
Negative		16	137	153
Total		27	144	171

Abbreviations: BMB Bone marrow biopsy; PET/CT Positron emissiontomography/computed tomography

When considering BMB results as the gold standard, the performance metrics for PET/CT in detecting BMI were as follows: sensitivity, 40.7% (95% confidence interval (CI), 33.4–48.1%); specificity, 95.1% (95% CI, 91.9–98.4%); PPV, 61.1% (95% CI, 53.8–68.4%); and NPV, 89.5% (95% CI, 84.9–94.1%) (supplemental Table 3). The overall diagnostic accuracy of PET/CT for identifying BMI was 86.5% (95% CI, 81.4–91.6%).

Staging capabilities of BMB and PET/CT

Among the patients initially classified as early-stage (I or II) based on PET/CT, all had negative bone marrow biopsy results, and none were upstaged to stage IV. However, for patients classified as stage III by PET/CT, positive BMB results contributed to upstaging to stage IV of 13 individuals, representing 28.9% of the 45 patients in this category (Table 3). Furthermore, 2 out of 104 (1.9%) patients in stages I–II were upstaged to stage IV due to bone marrow involvement identified via PET/CT. Regardless of bone marrow involvement (BMI), 11 patients were classified as stage IV due to extranodal organ involvement, three of whom had only extranodal involvement without BMI. Among the 37 stage IV patients, 24 were diagnosed as stage IV based on PET/CT alone, without the need for bone marrow biopsy (BMB). Thirteen patients were upstaged to stage IV solely due to BMI confirmed by BMB. Among the 27 patients with positive BMB results, 14 (51.9%) were already categorized as having stage IV disease on the basis of their PET/CT results.

Table 3.

Correlation between PET findings and BMB results

Stage by PET/CT	BMB, No. (%)
	Positive	Negative
I-II	0 (0)	102 (100)
III	13 (28.9)	32 (71.1)
IV	14 (58.3)	10 (41.7)

Abbreviations: BMB Bone marrow biopsy; PET/CT Positron emission tomography/computed tomography; BMI Bone marrow involvement

To further evaluate the factors associated with BMB positivity and to guide biopsy, we classified advanced stage (stage III–IV) patients by PET/CT into four subgroups: Group 1 comprised 32 patients staged as III by PET/CT, with negative BMB results; Group 2 comprised 13 patients staged as III by PET/CT, with positive BMB results; Group 3 comprised 10 patients staged as IV by PET/CT, with negative BMB results; and Group 4 comprised 14 patients staged as IV by PET/CT, with positive BMB results. Patients with noncontiguous extralymphatic involvement (stage IV on the basis of PET/CT, encompassing groups 3 and 4) exhibited a higher rate of BMB positivity (58.3%) compared with those in stage III (28.9%), P = 0.017.

In a subsequent multivariate logistic regression analysis, female sex, Eastern Cooperative Oncology Group (ECOG) performance score > 1, abnormal beta2 micro-globulin levels, and involvement of more than four lymph node regions were significantly associated with BMB-positivity, with stage III-IV determined by PET/CT findings (Fig. 2). We have conducted the goodness-of-fit test using the Hosmer-Lemeshow test, which yielded a p-value of 0.669 (> 0.05). Additionally, the classification table shows a classification accuracy of 74.5%. The variance inflation factors (VIF) for all variables were less than 2, and the condition index was below 10, indicating no significant multicollinearity. Using these four risk factors, patients in stages III–IV on the basis of PET/CT were stratified into three risk levels: low-risk (0–1 factor), intermediate risk (2 factors), and high risk (3–4 factors). The corresponding BMB positivity rates were 5.6% (1/18), 40.7% (11/27), and 68.8% (11/16), respectively(P < 0.01) (Fig. 3). Even when sex was excluded from the variables in the stratification model, a similar trend was still observed (supplemental Fig. 1).

Fig. 2.

Forest plots showing the independent prognostic effects of clinical parameters on bone marrow involvement in patients stage III–IV FL by PET/CT. PET/CT, positron emission tomography/computed tomography; ECOG, Eastern Cooperative Oncology Group; HGB, hemoglobin; LDH, lactate dehydrogenase

Fig. 3.

Risk levels and BMB-positive rates. BMB, bone marrow biopsy. Risk factors: female sex, Eastern Cooperative Oncology Group (ECOG) performance score > 1, abnormal beta2 micro-globulin levels, involvement of more than four lymph node regions

Prognostic performance of BM biopsy and PET/CT Staging in advanced-stage patients

With a median follow-up of 34 months, the 2-year OS and PFS rates were 97.4% and 90.2%. For stage III and IV patients, the 2-year OS were 95.5% and 97.2%, and PFS were 83.5% and 84.5%, respectively. No significant differences in OS and PFS between BMB-positive and BMB-negative patients in stage III-IV or IV (P > 0.05, Fig. 4; supplemental Fig. 2). The 2-year OS and PFS rates by group were: Group 1 (95.5%, 83.5%), Group 2 (100.0%, 90.0%), Group 3 (100.0%, 78.8%), Group 4 (92.9%, 84.4%), with no significant differences between groups. (P > 0.35, supplemental Figs. 3 and 4).

Fig. 4.

Survival comparison between BMB-positive and BMB-negative patients with advanced-stage follicular lymphoma. Overall survival (a) and progression-free survival (b). BMB, bone marrow biopsy

Discussion

Patients with follicular lymphoma represent a heterogeneous population with distinct tumor burdens and bone marrow involvement rates. This two-institutional study was the first to our knowledge to assess the value of the BMB and FDG PET/CT with an initial diagnosis of low-grade follicular lymphoma. We have demonstrated that bone marrow biopsy can be safely omitted in early-stage, low-grade follicular lymphoma (FL), as no patients initially classified as stage I or II by PET/CT were upstaged due to positive BMB findings. For advanced-stage patients, the addition of BM biopsy changed the stage of 18.8% of the patients. Furthermore, patients with PETCT-based advanced-stage FL were classified into three groups using four independent factors to predict bone marrow involvement, which may help guide clinical decision-making regarding the necessity of performing a bone marrow biopsy.

Because FL is a widespread disease with a significant risk of noncontiguous extralymphatic involvement [41], it is essential that any initial treatment strongly emphasizes accurate staging. In modern clinical practice, BMI in FL is often considered high-risk advanced disease, BMB is still performed routinely [6, 7]. In the current study, 15.8% of the patients were BM biopsy-positive, similar to the results in a previous large cohort study [42]. However, other studies found positivity rates of > 20% [26, 27, 43–47], indicating the presence of selection bias and differing proportions of patients with advanced-stage disease. Generally, BM-positivity is more commonly observed in patients with advanced-stage diseases [48, 49], while some previous studies showed that the frequency of BM involvement is higher in lower-grade lymphomas^, [50–52]. We suspect that the higher proportion of bone marrow (BM)-positive patients in grade 1 follicular lymphoma (FL) (30.4%) may, in part, be attributed to the ongoing risk of BM involvement over long-term survival [37]. Similar to a previous FL study [44], our findings demonstrate that none of the patients staged as I-II by PET/CT were upstaged to stage IV based on bone marrow biopsy results, indicating a 100% negative predictive value for early-stage patients. As a result, PET/CT can be used to select early-stage patients with low-grade FL for localized radiotherapy.

Although PET/CT may occasionally miss BMI in patients who test positive via BM biopsy, it is noteworthy that the frequency of disease upstaging owing to BM biopsy results is remarkably low for both Hodgkin’s lymphoma and DLBCL, with incidences < 10%. This statistic highlights the efficacy and reliability of PET/CT in the staging process for these lymphomas, despite its limitations in capturing every instance of BM involvement [15, 18]. Nevertheless, previous studies have yielded inconsistent conclusions regarding the detection of BM involvement using PET/CT in patients with FL. The reported sensitivities vary from 39–68% [21, 29, 31, 43, 53], reflecting the heterogeneity in metabolic activity of the disease, varying proportions of patients with advanced-stage disease, and the absence of indicators for further tailoring of BM examination. Notably, in the current study, BM biopsy resulted in the upstaging of approximately 30% of the patients initially categorized as stage III based on PET/CT findings alone. This significant rate of upstaging in low-grade FL patients highlights the critical role of BM biopsy in the comprehensive staging of the disease, suggesting that PET/CT alone may not suffice to accurately determine the extent of disease involvement in a substantial subset of patients. The shift from stage III to stage IV in FL may not directly influence management strategies; however, a higher FLIPI-2 score, which often accompanies advanced staging, is associated with a worse prognosis [47–49]. Despite these prognostic implications, the current study did not demonstrate a survival difference between individuals with or without BMI, which could be attributed to the limited number of patients and the relatively short follow-up duration. Consequently, it remains advisable to continue to perform bone marrow biopsies for patients with PETCT-based advanced-stage low-grade FL. This diagnostic step is pivotal to achieve precise staging and for garnering deeper insights into the disease’s biological characteristics, and facilitating tailored therapeutic approaches and potentially impacting patient outcomes.

Based on our thorough analysis, four parameters were significantly associated with BM involvement, most of which indicate a greater tumor burden. With an increase in tumor burden, there is a corresponding increase in the proportion of patients with BMB-positivity. Although sex was identified as a statistically significant factor for BMB positivity, the exact reasons for this association are unclear. Given the retrospective nature of this study, this finding may be due to selection bias or other confounding factors not captured in our analysis. After excluding sex as a risk factor, a similar trend was still observed. The proportion of early-stage patients who were upgraded to stage IV owing to BM biopsy was extremely small in our previous study of extranodal nasal-type natural killer/T-cell lymphoma [20]. Similar patterns in Hodgkin’s lymphoma have also been documented in previous studies [54, 55]. Therefore, it is logical to omit BM biopsy for patients with no risk factors. However, an important association between the identified risk factors and the likelihood of BMB positivity in patients with advanced-stage low-grade FL, which could potentially guide clinical decision-making and biopsy interventions. Those with high tumor burden should still undergo the procedure to ensure comprehensive staging and appropriate clinical management.

There are limitations in this retrospective study. First, as this study focused only on low-grade FL, the contribution of BM biopsy in grade 3 disease remains uncertain. However, some physicians may consider FL grade 3 A as low-grade FL and others may treat it as an aggressive NHL. So, the grade 1–2 follicular lymphoma was taken into our study to avoid the potential effects of confounders. Second, the impact of BM involvement on survival outcomes could not be assessed owing to an insufficient number of patients with low-grade FL and insufficient follow-up. Other limitations are the lack of information on certain tumor burden-associated biomarkers, such as circulating tumor deoxyribonucleic acid [56], and their association with the risk of relapse in BM. External validation of the diagnostic and prognostic value of BM examination in a larger cohort is necessary in future research.

Conclusions

In summary, BM biopsy did not significantly alter the stage distribution for early-stage patients initially staged using PET/CT. However, it appears reasonable to reserve BM biopsy for advanced-stage patients. In the future, it may be possible to tailor BM examination strategies on the basis of tumor burden factors.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Abbreviations

BM

Bone marrow

BMB

Bone marrow biopsy

PET/CT

Positron emission tomography/computed tomography

¹⁸F-FDG

¹⁸F-fluorodeoxyglucose

FL

Follicular lymphoma

BMI

Bone marrow involvement

DLBCL

Diffuse large B-cell lymphoma

PET+

Patients with positive bone marrow involvement by positron emission tomography

PPV

Positive predictive value

NPV

Negative predictive value

OS

Overall survival

PFS

Progression-free survival

FLIPI-1

Follicular Lymphoma International Prognostic Index-1

FLIPI-2

Follicular Lymphoma International Prognostic Index-2

ECOG

Eastern Cooperative Oncology Group

CI

Confidence interval

Author contributions

Conception and design: YY, TBL, and YJZ.Financial support: YY.Administrative support: YY, TBL, andYJZProvision of study material or patients: YYC, JJW, WHZ, JRG, JHC, TBL, YJZ, and YY.Data collection and organization: YYC, JJW, WHZ, JRG, JHC, TBL, YJZ, and YY.Data analysis and interpretation: YYC, JJW, WHZ, JRG, YYQ, TXL, HZ, SLC, SQL, RZZ, GQS, TLT, YPL, CH, JHC, TBL, YJZ and YY.Manuscript writing: YYC, JJW, WHZ, JRG, YYQ, TXL, HZ, SLC, SQL, RZZ, GQS, TLT, YPL, CH, JHC, TBL, YJZ and YY.Final approval of the manuscript: YYC, JJW, WHZ, JRG, YYQ, TXL, HZ, SLC, SQL, RZZ, GQS, TLT, YPL, CH, JHC, TBL, YJZ and YY.Accountable for all aspects of the work: YYC, JJW, WHZ, JRG, YYQ, TXL, HZ, SLC, SQL, RZZ, GQS, TLT, YPL, CH, JHC, TBL, YJZ and YY.

Funding

This work was supported by grants from the Major Scientific Research Program for Young and Middle-aged Health Professionals of Fujian Province, China (Grant No. 2022ZQNZD002), the Fujian Key Laboratory of Intelligent Imaging and Precision Radiotherapy for Tumors (Fujian Medical University), and Clinical Research Center for Radiology and Radiotherapy of Fujian Province (Digestive, Hematological and Breast Malignancies). The funding sources had no influence on the design, performance, or reporting of this study.

Data availability

Research data are stored in an institutional repository and will be shared upon request to the corresponding author.

Declarations

Ethics approval

Ethics approval was obtained from Fujian Medical University Union Hospital Ethics Committee (approval number: 2022WSJK019).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.