Incidence, severity, and risk factors of early postoperative complications in laryngectomy patients: a retrospective analysis

Abstract

Purpose

To determine the incidence, severity and risk factors of complications in patients undergoing laryngectomy.

Methods

433 patients with hypopharyngeal and laryngeal carcinomas who underwent laryngectomies from February 2023 to October 2023 were enrolled in this retrospective study. Clinical data of the patients were collected according to the Electronic Medical Record System, including data of demographic and clinical characteristics, surgical treatment, laboratory test, complications within 30 days following laryngectomy. Risk factors for early postoperative complications were identified with univariate and multivariate logistic regression analysis.

Results

Early postoperative complications occurred in 19.4% of patients following laryngectomy. In partial laryngectomies, subcutaneous emphysema was the predominant issue, affecting 12.0% of patients (n = 33). Among the potential postoperative complications of total laryngectomy, wound infection and pharyngocutaneous fistula stand out as the most prevalent, each affecting 6.3% of the patients in the study (n = 10). The majority of complications (93.5%) were classified as mild according to Clavien–Dindo classification. For total laryngectomy patients, alcohol consumption, a high ASA classification, and TNM stage IV were identified as significant risk factors for overall complications. In partial laryngectomy patients, low preoperative hemoglobin levels (< 12.5 g/dl) emerged as an independent risk factor for local complications. For total laryngectomy patients, both a high ASA classification and a BMI exceeding 25 kg/m² were associated with local complications, while low preoperative hemoglobin was a key risk factor for airway complications. No significant risk factors were identified for overall and airway complications in partial laryngectomy patients.

Conclusions

This study reports a typical incidence and mild severity of early postoperative complications following laryngectomy. Severe complications and mortality rates post-laryngectomy are notably infrequent. The ASA classification and preoperative nutritional status are suggested as potential predictors of early postoperative complications in laryngectomy patients.

Introduction

Hypopharyngeal and laryngeal carcinomas (HLC) are common malignancies of the head and neck, with laryngeal malignancy being the second most common malignancy in the head and neck region [1]. The standardized incidence rate of laryngeal cancer in China is 0.57 per 100,000 people, with a notably higher rate of 0.70 per 100,000 observed specifically in the northern regions [2]. In the context of laryngeal carcinoma management, while radiotherapy and transoral laser microsurgery (TLM) have emerged as predominant treatment modalities in many regions, the prevalence of total laryngectomy has experienced a decline in recent years [3–5]. However, in China, surgical resection remains the cornerstone of therapy for laryngeal malignancies. For early-stage (T1 and T2) laryngeal carcinomas, partial laryngectomy is often selected to achieve local control of the disease while maximizing the preservation of laryngeal function. On the other hand, total laryngectomy is indicated for locally advanced (T3 and T4) laryngeal and hypopharyngeal carcinomas, particularly in scenarios where the tumor extends beyond the larynx [6]. The workload for healthcare professionals is substantial and is expected to increase further due to the aging population in China. Postoperative problems in patients with laryngeal and hypopharyngeal cancers are linked to a poor short- and long-term prognosis [7, 8]. Postoperative complications correlate with more hospitalization and increased hospital readmission, thereby adding the patient’s financial burden and the burden of healthcare professionals [9, 10]. Therefore, reducing postoperative complications and enhancing perioperative management are of vital importance for patients undergoing laryngectomy. Early identification of patients with potential postoperative problems could help to improve prognosis and lessen the burden.

Due to the lack of comparability of studies, complication rates and risk factors of complications are not described uniformly in the literature. In fact, the number of reports on the complications of partial laryngectomy (PL) is still limited. Early postoperative complications in PL included surgical site infection, subcutaneous emphysema, bleeding, laryngocutaneous fistula and pneumonia complications. The reported incidence of early postoperative complications of patients with PL ranges from 16.8% to 35%. Reported risk factors for developing early postoperative complications in patients with PL are age, smoking and prior radiation [11–14]. Total laryngectomy (TL) was associated with a notable incidence of postoperative complications, estimated to be between 20% and 30% [9, 15]. In instances of salvage laryngectomy, this rate may be even more pronounced, with a potential range of 30% to 40% [16, 17]. The spectrum of common postoperative complications after TL encompasses a variety of issues, such as wound infections, impaired wound healing, hemorrhagic events and hematoma formation, pulmonary infections and septic episodes, dysphagia, and the development of pharyngocutaneous fistulas [18]. Investigators have identified various risk factors that are associated with early postoperative complications after TL, including prior smoking, prior radiation, the Charlson Comorbidity Index, reconstruction and lower preoperative hemoglobin [9, 16, 19].

The Clavien-Dindo Classification (CDC) is a standardized framework for categorizing surgical complications, facilitating the comparison of quality assessment data across different institutions. This system underscores the importance of morbidity and the therapeutic interventions employed to address complications, thereby enabling a precise determination of their severity [20]. In recent years, the CDC has garnered international recognition as the preferred method for uniformly classifying surgical complications, ensuring a consistent approach to evaluating patient outcomes and surgical quality [21].

Nevertheless, there is a notable scarcity of comprehensive complication data derived from extensive patient cohorts who have undergone laryngectomies.

tomies in China, with a particular dearth of information pertaining to patients who have received partial laryngectomies (PL). The present retrospective cohort study aimed to meticulously document and scrutinize the early postoperative complications ensuing from both partial and total laryngectomies (PL and TL, respectively), and to discern the risk factors associated with these complications. Additionally, the study endeavors to assess the severity of these early postoperative complications by employing the Clavien-Dindo Classification (CDC) as a grading system. It is anticipated that the identification of risk factors for early postoperative complications in the context of PL or TL will facilitate the prompt recognition of patients at elevated risk and inform a more targeted approach to selecting the optimal therapeutic intervention.

Materials and methods

Inclusion and exclusion criteria

At the Department of Otorhinolaryngology of Head and Neck Surgery, Eye & ENT Hospital of Fudan University, between March and October in 2023, 433 consecutive patients received laryngectomies for laryngeal and hypopharyngeal cancers. According to our Institutional policy, criteria for eligibility were: undergoing partial laryngectomy or total laryngectomy for the first time, and older than 18 years. Furthermore, patients with second primary tumor in the head and neck, severe chronic obstructive pulmonary disease, neurological impairment and accepting mechanically ventilated patients were excluded from the study. Patients who were missing the important clinical data and complication data were also excluded from the study.

Data collection

After approval from the Institutional Review Board was obtained, the clinical data of patients were collected retrospectively from hospital charts, including data of demographic and clinical characteristics, surgical treatment, laboratory tests and postoperative complications. Demographic data were collected including age at diagnosis, gender, tumor location, tumor stage, nodal stage and TNM stage according to the seventh edition of the TNM system, smoking status, alcohol use, preoperative nutritional status (Body mass Index, BMI), previous oncologic treatment (surgery or chemotherapy or radiotherapy), length of stay, previous respiratory diseases, medical comorbidity, ASA status and age-adjusted Charlson Commodity Index. Surgical data collected included operation duration, extent of surgery (PL and TL), neck dissection, and reconstruction. We collected the following laboratory tests variables associated with postoperative complications, including hemoglobin and red blood cell distribution width (RDW). All data were anonymized and participants were assured that no institution would be identifiable in any publication.

Definition of complications

In this research, we defined postoperative complications as any adverse events that manifest within the 30-day period following surgery, encompassing a spectrum of issues such as subcutaneous emphysema, wound infections at the stoma or neck region, postoperative hemorrhage, pharyngocutaneous fistulas (PCF), various fistulas including anastomotic and chyle leaks, pneumothorax, as well as pneumonic and vascular complications like deep vein thrombosis and pulmonary embolism. These complications were systematically categorized into two groups: local, which included infections at the wound site, bleeding, PCF, and other fistulas; and airway-related, which comprised subcutaneous emphysema, pneumothorax, and pneumonia. The severity of each complication was meticulously assessed utilizing the Clavien-Dindo classification system. For the calculation of composite outcome rates (namely, the overall, local, and airway complication rates) and their respective risk factors, a patient was counted only once per composite category, even if they experienced multiple specific complications within that category. In contrast, the incidence of each specific complication type (e.g., subcutaneous emphysema, PCF, etc.) was determined by analyzing all individual events separately.

Statistical analysis

Categorical data are presented as a number with a percentage. Continuous variables are summarized as median with interquartile range (IQR), or as mean and standard deviation in case of the normal distribution as indicated by the respective label. Covariates exhibiting significant association with complications were identified initially using univariate analysis. Multivariate regression models were fit using backwards stepwise method for variable entry and removal. Covariates significant on univariate analysis were then assessed by multivariate analysis using multiple logistic regression method. Results were expressed as odds ratios (OR) and 95% confidence intervals (CI). Results were considered significant if the p-value was ≤ 0.05%. All analyses were done using SPSS 24.0 (IBM Corporation, Armonk, New York, USA).

Results

A total of 433 patients undergoing PL or TL for laryngeal cancer were analyzed. The mean age was 63.43 ± 8.0 years, with a male majority (96.30%) and PL being the most common indication (63.5%). Preoperative radiation therapy and chemotherapy was given to 2.5% and 5.3% of the patients, respectively. The most common tumor grade was T2. The majority of patients were below ASA 2. A total of 229 (52.9%) patients received concomitant neck dissection. Reconstruction using pedicled flaps was performed in 38 cases. The median operation duration was 173 (106.5–276.5) minutes and the mean length of stay was 14.5 ± 5.1days. All demographics of patients are presented in Table 1.

Table 1.

Demographics of the study population

Characteristics	Total (n = 433)
Age (y)	63.4 ± 8.0
Sex, n (%)
Men	417 (96.3)
Women	16 (3.7)
Preoperative radiation, n (%)
Yes	11 (2.5)
No	422 (97.5)
Preoperative chemotherapy, n (%)
Yes	23 (5.3)
No	410 (94.7)
Smoking (tobacco), n (%)
None	103 (23.8)
Former	160 (37)
Current	170 (39.3)
Alcohol use, n (%)
None	248 (57.3)
Former	73 (16.9)
Current	112 (25.9)
Diabetes, n (%)
Yes	58 (13.4)
No	375 (86.6)
Primary tumor site, n (%)
Hypopharynx	83 (19.2)
Larynx	350 (80.8)
Tumor stage, n (%)
T1	106 (24.5)
T2	150 (34.6)
T3	106 (24.5)
T4	71 (16.4)
TNM stage, n (%)
I	60 (13.8)
II	106 (24.4)
III	107 (24.7)
IV	160 (36.9)
BMI, n (%)
<25 kg/m2	324 (74.8)
≥25 kg/m2	109 (25.2)
Preoperative hemoglobin, n (%)
<12.5 g/dl	55 (12.7)
≥12.5 g/dl	374 (86.4)
Missing	4 (0.9)
Preoperative red blood cell distribution width (RDW), n (%)
<14.5	399 (92.1)
≥14.5	29 (6.7)
Missing	5 (1.2)
ASA classification, n (%)
≤2	414 (95.6)
>2	19 (4.4)
ACCI, n (%)
<3	296 (68.4)
≥3	137 (31.6)
PL or TL, n (%)
PL	275 (63.5)
TL	158 (36.5)
Neck dissection, n (%)
No	204 (47.1)
Unilateral	143 (33)
Bilateral	86 (19.9)
Reconstruction, n (%)
Yes	38 (8.8)
No	395 (91.2)
Mean operation duration (min)	173 (106.5-276.5)
Hospital length of stay (days)	14.5 ± 5.1

ACCI Adjusted Charlson Comorbidity Index, PL Partial laryngectomy, TL Total laryngectomy

Data are expressed as the number (%) or median (interquartile range) for nonnormally distributed data. Data was expressed as the mean ± standard deviation for normally distributed data

A total of 92 complications were detected in 84 (19.40%) patients. Eight patients experienced two distinct complication events. The complications were further divided into PL- (19.3%) and TL-related complications (19.6%) (Table 2). The most common complication was subcutaneous emphysema in PL (n = 33; 12.0%), wound infection (n = 10; 6.3%) and pharyngocutaneous fistula in TL (n = 10; 6.3%). Local bleeding was reported in three patients after PL (1.1%), and no local bleeding was reported in patients after TL. Timing of the Complications and Clavien-Dindo classification (CDC) of complications is presented in Table 3. 56 (60.9%) complications occurred within 7 days of surgery. In total, there were 40 (43.5%) CDC grade I complications, 46 (50.0%) were grade II complications, and another 6 (6.5%) were grade III complications. No procedure-related deaths or Clavien–Dindo grade IV complications were observed (Table 3).

Table 2.

Postoperative complications during the first 30 days for laryngectomy patients

Type of complication	All patients (n = 433)
	PL (n = 275)	TL (n = 158)
All complications	53 (19.3)	31 (19.6)
Local complications	14 (5.1)	24 (15.2)
Wound infection	12 (4.5)	10 (6.3)
Bleeding	3 (1.1)	0 (0)
PCF	0 (0)	10 (6.3)
Anastomotic fistula	0 (0)	1 (0.6)
Chyle leak	0 (0)	4 (2.5)
Airway complications	41 (15.9)	6 (3.8)
Pneumonia	8 (2.9)	6 (3.8)
Pneumothorax	1 (0.4)	0 (0)
Subcutaneous emphysema	33 (12.0)	0 (0)
DVT/PE	1 (0.4)	3 (1.9)

PL Partial laryngectomy, TL Total laryngectomy, PCF Pharyngocutaneous fistula formation, DVT Deep venous thrombosis, PE Pulmonary embolism

Table 3.

Timing and Clavien-Dindo classification (CDC) of complications (n = 92)

Timing in relation to surgery (n; %)
Intraoperative	1 (1.1)
< 7 days postoperatively	56 (60.9)
7–14 days postoperatively	23 (25.0)
>14 days postoperatively	12 (13.0)
Classification of complications by Clavien–Dindo (n; %)
I	40 (43.5)
II	46 (50.0)
IIIa	1 (1.1)
IIIb	5 (5.4)

Tables 4 and 5 provide a summary of the univariate and multivariate logistic regression analyses assessing risk factors for postoperative overall complications in patients undergoing partial (PL) and total laryngectomy (TL), respectively. The univariate analysis for PL did not identify any significant risk factors. In contrast, for TL patients, the univariate analysis indicated several factors associated with an increased risk of postoperative complications: former alcohol use (OR 4.71, 95% CI 1.66–13.37, p < 0.05), current alcohol use (OR 2.53, 95% CI 0.98–6.53, p = 0.056), TNM stage IV (OR 0.36, 95% CI 0.13–0.98, p < 0.05), and an ASA classification (OR 8.49, 95% CI 2.55–28.26, p < 0.05). Upon further multivariate logistic regression analysis, independent risk factors for postoperative overall complications in TL patients were identified as follows: former alcohol use (OR 5.58, 95% CI 1.74–17.92, p < 0.05), current alcohol use (OR 3.19, 95% CI 1.11–9.17, p < 0.05), TNM stage IV (OR 0.21, 95% CI 0.06–0.65, p < 0.05), and an ASA classification greater than 2 (OR 8.68, 95% CI 2.40-31.35, p < 0.05)

Table 4.

Factors predictive of overall complications in PL

Characteristics	Univariate
	OR	95%CI	P-value
Age
≤ 60
>60	1.46	0.77–2.79	0.249
Preoperative radiation
No
Yes	1.05	0.11–9.57	0.967
Preoperative chemotherapy
No
Yes	2.16	0.52–8.93	0.288
Smoking (tobacco)
None
Former	1.62	0.69–3.78	0.264
Current	1.60	0.69–3.73	0.276
Alcohol use
None
Former	0.68	0.28–1.64	0.388
Current	0.81	0.38–1.72	0.590
Tumor stage
T1
T2	0.81	0.41–1.61	0.546
T3	0.98	0.41–2.35	0.972
T4	3.24	0.80-13.12	0.100
TNM stage
I
II	1.02	0.50–2.08	0.949
III	0.79	0.31–2.02	0.615
IV	1.14	0.46–2.88	0.774
BMI
<25 kg/m2
≥25 kg/m2	0.72	0.36–1.46	0.367
ASA classification
≤2
>2	2.14	0.38–11.99	0.388
ACCI
<3
≥3	1.43	0.75–2.75	0.279
Neck dissection
No
Unilateral	0.67	0.33–1.38	0.278
Bilateral	0.78	0.25–2.43	0.668
Reconstruction
No
Yes	1.20	0.24–5.97	0.820
Operation duration
<300 min
≥300 min	0.75	0.28–2.06	0.579
Preoperative hemoglobin
<12.5 g/dl
≥12.5 g/dl	0.42	0.16–1.10	0.078
Preoperative RDW
<14.5
≥14.5	3.14	0.96–10.32	0.059

PL Partial laryngectomy, ACCI Adjusted Charlson Comorbidity Index, RDW Red blood cell distribution width

Table 5.

Factors predictive of overall complications in TL

Characteristics	Univariate			Multivariate
	OR	95%CI	P-value	OR	95%CI	P-value
Age
≤60
>60	1.17	0.49–2.75	0.727
Preoperative radiation
No
Yes	2.12	0.37–12.14	0.398
Preoperative chemotherapy
No
Yes	2.52	0.78–8.15	0.122
Smoking (tobacco)
None
Former	0.76	0.26–2.18	0.607
Current	1.09	0.41–2.88	0.870
Alcohol use
None
Former	4.71	1.66–13.37	0.004	5.58	1.74–17.92	0.004
Current	2.53	0.98–6.53	0.056	3.19	1.11–9.17	0.031
Tumor stage
T2
T3	0.40	0.15–1.09	0.074
T4	0.49	0.19–1.28	0.147
TNM stage
II
III	0.49	0.16–1.44	0.192	0.44	0.13–1.43	0.171
IV	0.36	0.13–0.98	0.045	0.21	0.06–0.65	0.007
Diabetes
No
Yes	2.53	0.86–7.49	0.093
BMI
<25 kg/m2
≥25 kg/m2	2.15	0.89–5.18	0.087
ASA classification
≤2
>2	8.49	2.55–28.26	< 0.001	8.68	2.40-31.35	< 0.001
ACCI
<3
≥3	0.91	0.41–2.04	0.820
Neck dissection
No
Unilateral	1.18	0.34–4.12	0.793
Bilateral	2.40	0.73–7.90	0.149
Reconstruction
No
Yes	2.19	0.88–5.44	0.092
Operation duration
<300 min
≥300 min	1.94	0.88–4.31	0.102
Preoperative hemoglobin
≥12.5 g/dl
<12.5 g/dl	0.97	0.38–2.48	0.949
Preoperative RDW
<14.5
≥14.5	1.83	0.59–5.64	0.295

Bold values denotes p<0.05, TL Total laryngectomy, ACCI Adjusted Charlson Comorbidity Index, RDW Red blood cell distribution width

The predictive variables for local and airway complications following partial (PL) and total laryngectomy (TL) procedures are presented in Tables 6, 7 and 8. In the context of PL, both univariate and multivariate analyses identified low preoperative hemoglobin (OR 5.88, 95% CI 1.36–20.41, p < 0.05) as a significant independent risk factor for local complications. While preoperative chemotherapy and tumor stage demonstrated statistical significance in univariate analysis, they did not maintain this in multivariate analysis, thus they are not considered independent risk factors for patients undergoing PL. For TL patients, the independent risk factors for local complications were identified as the ASA classification (OR 6.37, 95% CI 1.86 to 21.81, p < 0.05) and BMI (OR 6.365, 95% CI 1.858 to 21.807, p < 0.05). In the case of airway complications for TL patients, low preoperative hemoglobin (OR 7.81, 95% CI 1.37–45.46, p < 0.05) was the only independent risk factor, as determined by both univariate and multivariate analyses. Data pertaining to airway complications in PL patients were excluded from presentation due to the absence of statistically significant predictive factors.

Table 6.

Univariate and multivariable of local complications with PL

Variables	Univariate analysis		Multivariate analysis
	OR (95%CI)	P value	OR (95%CI)	P value
Age	0.42 (0.14–1.25)	0.120
Preoperative radiation	4.94 (0.52–47.41)	0.166
Preoperative chemotherapy	6.05 (1.13–32.28)	0.035	2.74 (0.41–18.25)	0.298
Smoking	1.46 (0.70–3.07)	0.313
Alcohol use	0.72 (0.35–1.51)	0.385
Diabetes	2.42 (0.72–8.13)	0.152
BMI	0.70 (0.19–2.59)	0.596
Tumor stage	1.87 (1.02–3.44)	0.043	1.78 (0.96–3.23)	0.069
Neck dissection	1.77 (0.85–3.70)	0.117
Operation duration	1.28 (0.27–6.01)	0.752
Preoperative hemoglobin	6.02 (1.70-21.28)	0.005	5.88 (1.36–20.41)	0.016
Preoperative RDW	1.72 (0.21–14.36)	0.616

Bold values denotes p<0.05, PL Partial laryngectomy, RDW Red blood cell distribution width

Table 7.

Univariate and multivariable of local complications with TL

Variables	Univariate analysis		Multivariate analysis
	OR (95%CI)	P value	OR (95%CI)	P value
Age	0.91 (0.36–2.30)	0.847
Preoperative radiation	1.12 (0.13–10.05)	0.918
Preoperative chemotherapy	1.60 (0.41–6.21)	0.449
Smoking	1.29 (0.73–2.28)	0.372
Alcohol use	1.60 (0.98–2.60)	0.059
Diabetes	2.68 (0.85–8.45)	0.093
BMI	2.75 (1.08–7.02)	0.034	2.89 (1.09–7.70)	0.033
ASA classification	6.05 (1.83–20.02)	0.003	6.37 (1.86–21.81)	0.003
Tumor stage	0.79 (0.45–1.37)	0.398
Neck dissection	1.17 (0.86–3.05)	0.139
Operation duration	1.66 (0.69–3.99)	0.262
Preoperative hemoglobin	1.51 (0.48–4.73)	0.485
Preoperative RDW	4.53 (0.77–26.86)	0.096

Bold values denotes p<0.05, TL Total laryngectomy, RDW Red blood cell distribution width

Table 8.

Univariate and multivariable of airway complications with TL

Variables	Univariate analysis		Multivariate analysis
	OR (95%CI)	P value	OR (95%CI)	P value
Age	2.38 (0.27–20.91)	0.435
Preoperative chemotherapy	5.83 (0.97–35.18)	0.054
Preoperative tracheostomy	1.18 (0.13–10.60)	10.881
Smoking	1.01 (0.26–2.85)	0.979
Alcohol use	1.04 (0.41–2.59)	0.934
Diabetes	1.70 (0.19–15.48)	0.638
BMI	0.75 (0.09–6.65)	0.796
ASA classification	2.33 (0.25–21.62)	0.456
Tumor stage	0.75 (0.27–2.14)	0.596
Neck dissection	1.28 (0.41–4.04)	0.675
Operation duration	0.90 (0.16–5.06)	0.903
Preoperative hemoglobin	7.81 (1.37–45.46)	0.021	7.81 (1.37,45.46)	0.021
Preoperative RDW	4.53 (0.76–26.86)	0.218

Bold values denotes p<0.05, TL Total laryngectomy, RDW Red blood cell distribution width

Currently, there are no large, single-institution cohort for reporting complications in patients after laryngectomies in China. Thus, we retrospectively analyzed the clinical data of 433 patients to determine the incidence, severity and risk factors of complications in patients with laryngectomies in the early postoperative period. Given the substantial differences in complications between PL and TL, the incidence and associated independent factors were separately analyzed. The aforementioned results showed that ASA classification is independent predictor of postoperative early overall and local complications in TL patients. Nutritional status (Preoperative hemoglobin and BMI) was independent predictor in the development of postoperative local complications of PL, local and airway complications of TL. However, statistically significant risk factors for early postoperative overall and airway complications could not be recognized for PL patients.

We found the overall complication rate was 19.4%, with 19.3% and 19.6% complication rate in PL and TL respectively. The most frequent complications in our study included subcutaneous emphysema in PL patients, wound infection and pharyngocutaneous fistula in TL patients. Overall complications are difficult to compare with other studies because of heterogeneity of the patient selection and classification of complications. Radiology has been the primary technique of initial treatment for early laryngeal cancer in most centers. Different from other centers, PL is the frequent treatment modality in our center. Therefore, we report and analyze postoperative complications after PL. However, publications addressing the complications of PL are limited in recent years. Ganly et al. [13] reported postoperative short-term complications in 20.0% (n = 30) of their PL which was similar compared to our study (n = 53, 19.3%). Gallo et al. [12] reported the overall complications (n = 189, 35.3%), postoperative bleeding (n = 6, 1.1%), pneumonia (n = 71, 13.3%), neck emphysema (n = 5, 0.9%). In our series, 2.9% ( n = 8) reported pneumonia in the postoperative period, 1.1% ( n = 3) reported hematoma and neck emphysema complications (12%, n = 33) in the total PL population. The complications between these two studies are quite different apart from postoperative bleeding. This large variation could be explained by center’s expertise and methodological definition of the events. Meanwhile, we report a complication rate of 19.6% after TL, including PCF formation (6.3%), wound infection (6.3%), postoperative bleeding (0), fistula (3.1%, anastomotic fistula, chyle leak), pneumatic (3.8%) and DVT/PE (1.9%), which is less than some previous studies [16, 22, 23]. Meulemans et al. [16] reported the 34.2% overall complications with PCF formation in 25.5%, wound infection in 16.2%, postoperative hemorrhages in 5.8%. With a rate of 6.3%, clinical PCF formation is less than the reported PCF pooled incidence in a recent systematic review (21%) [24].The difference could be explained by the selection of primary total laryngectomies in our study instead of salvage total laryngectomies with a higher complication rate than primary total laryngectomies. And we took the policy of postponing peroral feeding after surgery, potentially preventing evolution to PCF formation. Interestingly, no postoperative hematoma was found in TL populations, while three PL patients experienced wound bleeding in our study. However, hematoma are well-recognized as one of the most common complications in primary and salvage TL [9, 25, 26]. This study defined postoperative bleeding as wound hematoma requiring surgery, which was similar to previous studies [27]. This finding may be attributed to the low rate of preoperative radiotherapy in our cohort. RT can contribute to pancytopenia for head and neck cancers, which may lead to bleeding [28].

Complications were mostly mild in our cohort, as 93.5% of complications were classified to Clavien-Dindo grade I-II. Six patients (6.5%) were classified to Clavien–Dindo grade III. No procedure-related deaths or Clavien–Dindo grade IV complications were observed. Our finding is inconsistent with the result of Koenen et al. with 28.9% above grade III [19]. This might be due to the adopt of conservative treatment for complications except bleeding.

In our series, ASA classification was identified as significant risk factors of the occurrence of overall and local complications during the postoperative course with TL patients. Poor general health (ASA III and IV) was associated with the poor functional outcome and worse survival rates in patients with laryngeal or hypopharyngeal Carcinoma [29, 30]. But the study by Koenen did not seem to support the hypothesis that increasing ASA grade leads to the occurrence of more postoperative complications, which is supported by our study after laryngectomies [19]. Alcohol consumption was also identified as a significant risk factor for overall complications in total laryngectomy patients. Notably, former drinkers had a higher risk (OR: 5.58) than current drinkers (OR: 3.19), possibly reflecting pre-existing health impairment that prompted cessation. While alcohol is known to impair wound healing by reducing epithelization and collagen deposition [31]– [32], its effect in our cohort was systemic rather than localized, influencing overall complications without dominating any single category. This broad impact may stem from alcohol’s multifactorial pathophysiology, including immune suppression and delayed tissue repair. Based on the result, the risk of postoperative complications was lower in patients with TNM stage IV than in those patients with TNM stage II on multivariate analysis (OR: 0.21, 0.06–0.65), which was paradox with the common sense. This paradox may stem from confounding factors, such as selection bias leading to more meticulous surgical and perioperative management for advanced-stage patients. Therefore, TNM stage was not identified as a definitive risk factor in our cohort, precluding its designation as a reliable predictor. Further research is needed to determine the association between TNM stage and postoperative complications after laryngectomy. Our results showed that preoperative hemoglobin was a significant factor in the development of postoperative local complications of PL and airway complications of TL. BMI was significant in the development of postoperative airway complications of TL. Hemoglobin levels and BMI were the common indicators of overall nutritional status. Malnutrition has long been associated with postoperative complications following laryngectomy including PCF, wound site infection and delayed wound healing [33–35].The identification of malnutrition prior to surgery with laryngectomy should be taken seriously in order to improve postoperative complications.

Many studies showed patients with elevated preoperative RDW had worse clinical outcomes in patients with laryngeal carcinoma [36–38]. Correlation between RDW and post-operative outcomes requires further evaluation among laryngectomy patients. Similar to our study with the same variable cut-off values for elevated RDW, Marcus found that an elevated RDW was not associated with complications in patients with total laryngectomy. In our cohort, preoperative RDW were not associated with postoperative overall complications for PL patients in the univariate and multivariate analysis. However, univariate analysis did show a strong trend significance with a negative effect of preoperative RDW (P = 0.059). The conflicting results have also reported on RDW among oral cavity cancer patients. In addition, preoperative radiation, flap reconstruction and lymph node invasion, which have been well-known risk factors for postoperative complications, were not found to be associated with higher complication rates [19, 22, 39]. This may be due to our inclusion of multiple types of complications rather than a single category so as to it is difficult to draw the consistent risk factors by statistical analysis.

Limitations of the study

Firstly, the retrospective nature of our data collection introduces potential biases, which preclude the establishment of causality. The reliance on previously documented medical records means that some complications might have been underreported or inconsistently recorded, lacking the rigor of a prospective protocol. Secondly, our focus was exclusively on complications occurring within a 30-day postoperative window. This short-term focus fails to capture later sequelae such as stomal stenosis, hypothyroidism, or long-term functional deficits in voice and swallow, which are significant for patient quality of life and overall recovery. Consequently, our findings should be interpreted as reflecting only the immediate postoperative period. The exclusion of swallowing difficulties from our list of postoperative complications is attributable to the extended duration of tube nasal feeding, which can last up to one month post-laryngectomy in our institution, potentially leading to an underreporting of these issues in our dataset. Moreover, our study did not include some salvage laryngectomies, such as those performed subsequent to a prior laryngectomy. This selection bias limits the applicability of our findings to these high-risk subgroups, which are known to be associated with significantly higher complication rates, particularly for pharyngocutaneous fistula. All data were sourced from a single, high-volume tertiary institution. While this ensures consistency in surgical technique and perioperative care, it may limit the generalizability (external validity) of our findings. Our clinical practices, patient demographics, and surgeon expertise may not be fully representative of the broader spectrum of healthcare settings managing laryngeal cancer cases. Lastly, our analysis was confined to the immediate clinical complications and did not explore the correlation between these complications and long-term oncological outcomes, such as overall survival or disease-free survival. Understanding whether early postoperative morbidity influences long-term survival is a critical area for future research.

Conclusion

This retrospective analysis delineates the incidence and spectrum of early complications following partial and total laryngectomy, revealing a characteristically low rate of severe morbidity. The complication profile differed between procedures, with subcutaneous emphysema being predominant in PL and wound-related issues in TL. The study conclusively identifies ASA physical status and preoperative nutritional markers as independent risk factors, highlighting a critical window of opportunity for prehabilitation. For TL patients, optimizing nutritional status and carefully evaluating comorbidities (ASA class) preoperatively is paramount. For PL patients, nutritional support emerges as a primary focus. Implementing these targeted preoperative strategies is essential for personalizing risk mitigation and enhancing recovery trajectories in laryngectomy care.

Author contributions

M.S. contributed to study drafting manuscript, review of manuscript and statistical analysis. J.W. contributed to study set-up, data quality control and review of the manuscript. C.W. and D.Z. contributed to data curation, visualization and supervision. K.D. was responsible for interpretation of data and revising the manuscript. L.T. contributed to revising the manuscript. P.S. contributed to data collection, and is the co-author of this article. All authors contributed to the article and approved the submitted version.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by Shanghai Nursing Association (No. 2023QN-B05); the Science and Technology Commission of Shanghai Municipality under grand (No. 24010703200); the Fudan University Fosun Nursing Research Fund (No. FNF202544).

Data availability

All the data will be available upon motivated request to the corresponding author of the present paper.

Declarations

Ethics approval

This study was conducted in accordance with the guidelines of the Eye & ENT Hospital of Fudan University Ethics Committee and was approved by the same committee (approval number: 2024183). The protocol was carried out in accordance with the Ethical Guidelines for Medical Research Involving Human Subjects and relevant national regulations.

Consent to participate

The need for written informed consent was waived by the Eye & ENT Hospital of Fudan University Ethics Committee.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.