CASE REPORT article

Front. Oncol., 04 November 2026

Sec. Head and Neck Cancer

Volume 16 - 2026 | https://doi.org/10.3389/fonc.2026.1979498

Case Report: Rapid recurrence and nodal metastasis of laryngeal spindle cell squamous cell carcinoma initially underestimated on superficial sampling

  • Department of Otolaryngology Head and Neck Surgery, Yibin First People’s Hospital, Yibin, China

Abstract

Spindle cell squamous cell carcinoma (SpSCC) is an uncommon and aggressive subtype of squamous cell carcinoma. These tumors are often polypoid, and superficial biopsy or frozen section may sample only the squamous cap. A 71-year-old woman presented with a polypoid glottic lesion involving the anterior commissure and both vocal cords, with normal cord mobility, staged cT1bN0M0. Transoral CO2 laser excision was performed. Permanent sections showed moderately differentiated squamous cell carcinoma with no spindle cell component, and adjuvant treatment was not indicated. Laryngoscopy at 3 months showed no evidence of recurrence. Eight months after surgery she presented with dyspnea and cervical adenopathy. Salvage total laryngectomy with bilateral neck dissection showed SpSCC with metastases in 3 of 9 lymph nodes. Re-cutting and restaining the original paraffin block identified no spindle cell component. She declined adjuvant treatment, developed a further recurrence 6 months later, and died 12 months after the salvage operation. Because no spindle cell population was demonstrable in the first specimen, the case cannot distinguish a component that never reached the slide from one that developed in the recurrent tumor, and without clonality data neither can be excluded. That ambiguity is the substance of the report, since both readings carry the same warning. A diagnosis of conventional squamous cell carcinoma made on a small, superficial, or thermally altered specimen from a polypoid glottic lesion describes the tissue examined rather than the behavior of the tumor, and follow-up should not rest on laryngoscopy alone.

1 Introduction

Spindle cell squamous cell carcinoma (SpSCC) accounts for approximately 1%–3% of laryngeal malignancies and follows an aggressive course (1, 2). The World Health Organization classifies it as a subtype of squamous cell carcinoma rather than a true biphasic neoplasm: the spindle cell population is clonally epithelial and its mesenchymal appearance is attributed to epithelial–mesenchymal transition (3). The older terms sarcomatoid carcinoma and carcinosarcoma are no longer recommended. In the largest laryngeal series, most patients were elderly men with a history of smoking, and the tumor was usually polypoid and pedunculated (2). A recent systematic review of head and neck SpSCC reported similar demographics and confirmed the larynx as the most common site (4).

The polypoid architecture creates a practical difficulty. Surface biopsy and frozen section often capture only the squamous epithelium that caps the lesion, so a tumor that is predominantly spindled at depth may be reported as conventional squamous cell carcinoma, or occasionally as a benign polyp (5, 6). We describe a patient whose laser excision specimen showed conventional squamous cell carcinoma and in whom SpSCC was identified only at a rapid submucosal recurrence with bilateral nodal metastasis; she died within 2 years.

2 Case description

A 71-year-old woman presented in January 2024 with hoarseness of 5 months’ duration. She had smoked approximately 50 g of home-grown, hand-rolled leaf tobacco a day for >40 years, roughly 160 pack-years if 12.5 g of loose tobacco is counted as one pack (7). She had no comorbidities or family history of cancer, and no genetic testing was performed. The neck was supple, the trachea midline, and no enlarged lymph nodes were palpable. Videolaryngoscopy showed a polypoid lesion arising from the anterior third of the right vocal cord and crossing the anterior commissure onto the adjacent surface of the left cord. Both cords moved normally, although glottic closure was incomplete (Figures 1A, B). Contrast-enhanced CT showed an enhancing nodule at the anterior commissure measuring 1.0 cm × 0.7 cm, without paraglottic or subglottic extension and without evidence of thyroid cartilage involvement (Figure 1C), and no cervical node was radiologically suspicious. Chest CT and abdominal ultrasonography showed no distant metastasis. Neither MRI nor fluorodeoxyglucose positron emission tomography/computed tomography (FDG-PET/CT) was performed. The differential diagnosis included vocal fold polyp, squamous papilloma, keratosis with dysplasia and invasive carcinoma; her age and smoking history, extension across the anterior commissure and contrast enhancement on CT made glottic carcinoma the working diagnosis. The tumor was staged cT1bN0M0 according to the eighth edition of the American Joint Committee on Cancer (AJCC) staging manual.

Figure 1

Transoral CO2 laser excision under suspension laryngoscopy and general anesthesia was offered as a larynx-preserving treatment and was accepted by the patient and her family. Intraoperative frozen section from the superficial part of the lesion on the left vocal fold showed squamous cell carcinoma. An extended cordectomy (European Laryngological Society type Va) was performed: starting approximately 3 mm from the visible tumor margin, the anterior part of the right ventricle, the right vocal fold, the anterior commissure and the anterior part of the left vocal fold were removed en bloc. Permanent sections showed moderately differentiated keratinizing squamous cell carcinoma with no spindle cell component (Figures 1D, E). Marked thermal artifact was present at the resection surface, with nuclear elongation and cytoplasmic streaming producing a pseudo-spindled appearance in the immediately subjacent tissue. These changes were confined to the coagulation zone, lacked the vesicular nuclei, nucleoli, and mitotic activity of a true spindle cell population, and were vimentin-negative on the sections re-cut after the salvage operation. All resection margins were negative on permanent sections. Neither lymphovascular nor perineural invasion was present. Adjuvant radiotherapy or chemotherapy was not indicated for a completely excised T1b glottic carcinoma and was not administered. Follow-up with clinical examination and videolaryngoscopy was planned every 1–3 months in the first year. Laryngoscopy in April 2024, 3 months after surgery, showed smooth mucosa across the operative field with no evidence of residual or recurrent tumor (Figure 1F). No further visit is recorded before she re-presented in September 2024, and surveillance up to that point had relied on videolaryngoscopy alone, without CT or MRI.

In September 2024, 8 months after the operation, she attended the emergency department with 10 days of progressive dyspnea. Multiple nodes were palpable at left neck level III. Laryngoscopy showed the left laryngeal ventricle bulging medially and compressing the glottis, with an intact and relatively smooth mucosal surface (Figure 2A). Contrast-enhanced CT demonstrated an irregular enhancing mass extending from the anterior commissure to the right aryepiglottic fold and paraglottic space, measuring 2.1 cm × 1.3 cm (Figure 2B). Post-laser edema or perichondritis was considered, but a submucosal mass with cervical adenopathy made recurrent carcinoma the working diagnosis. Chest CT and abdominal ultrasonography again showed no distant metastasis; FDG-PET/CT was not performed. The recurrence was staged cT3N2bM0 on clinical and radiological grounds. Total laryngectomy with bilateral neck dissection was performed.

Figure 2

The laryngectomy specimen contained a tumor with a maximum diameter of 3.0 cm and negative resection margins. The tumor was biphasic, comprising a conventional squamous cell carcinoma component largely confined to the superficial aspect of the lesion and a deeper spindle cell component with fascicular architecture, marked nuclear atypia and frequent mitoses (Figures 3A–C). A transition between the two components was identified (Figure 3D). Metastases were present in two of four left and one of five right cervical lymph nodes. Extranodal extension was absent. Final pathological staging was pT3N2cM0.

Figure 3

On immunohistochemistry, pan-cytokeratin (pan-CK), CK5/6, p63, p40, and epithelial membrane antigen (EMA) were strongly expressed in the squamous component and focally or patchily expressed in the spindle cell component (Figure 3E). Vimentin was positive in the spindle cell component (Figure 3F). p16 was negative. The Ki-67 labeling index was approximately 60%. The tumor was reported as spindle cell squamous cell carcinoma of the larynx.

Because the salvage specimen showed a subtype that had not been suspected, the original paraffin block from the laser excision was re-cut and stained again. Pan-CK, CK5/6, p63, p40, and EMA remained confined to the squamous carcinoma component; vimentin, p16, and programmed death-ligand 1 (PD-L1) were negative, and the Ki-67 index was approximately 80%. No spindle cell component was identified at any of the additional levels examined (Figure 3G).

The multidisciplinary team recommended postoperative radiotherapy with concurrent cisplatin, which in our hospital is given by the oncology department. The patient and her family declined, so no radiotherapy plan was drawn up and no adjuvant treatment was given. In March 2025, 6 months after laryngectomy, she returned with neck swelling. Hard, fixed lymph nodes were palpable at levels II and III on both sides, the largest at right level II measuring approximately 2.5 cm × 3.0 cm. Laryngoscopy showed a recurrent mass in the pharyngolaryngeal region (Supplementary Figure S1A), and contrast-enhanced CT showed a mass in the right posterior paraglottic space measuring 2.5 cm × 2.2 cm × 3.0 cm (Supplementary Figure S1B). She and her family declined further treatment, and no further anticancer therapy was given. She died in September 2025, 20 months after her first operation and 12 months after the diagnosis of SpSCC. The cause of death was not confirmed and no autopsy was performed. The clinical sequence is summarized in Table 1.

Table 1

DateEvent
August 2023Onset of hoarseness.
January 2024Videolaryngoscopy and contrast-enhanced CT; polypoid glottic lesion staged cT1bN0M0. Transoral CO2 laser excision (ELS type Va extended cordectomy, en bloc). Permanent sections: moderately differentiated squamous cell carcinoma, no spindle cell component.
April 2024Laryngoscopy at 3 months: no evidence of recurrence. No cross-sectional imaging obtained.
September 2024Dyspnea and left cervical adenopathy; recurrent laryngeal mass. Total laryngectomy with bilateral neck dissection: spindle cell squamous cell carcinoma, pT3N2cM0, three of nine nodes involved. Re-cut and additional immunohistochemistry on the original laser block: no spindle cell component identified. Postoperative radiotherapy with concurrent cisplatin recommended and declined.
March 2025Recurrent mass in the pharyngolaryngeal region and right posterior paraglottic space; hard, fixed nodes at bilateral levels II–III. Further treatment declined.
September 2025Death, 20 months after the initial operation and 12 months after the diagnosis of SpSCC.

Clinical timeline.

3 Discussion

The initial endoscopic and contrast-enhanced CT findings indicated early glottic carcinoma, for which transoral laser microsurgery is an accepted first-line option. In a recent systematic review, survival after transoral laser microsurgery and after vertical partial laryngectomy was comparable in T1 disease, and the transoral approach gave better voice outcomes, shorter admission, and fewer complications; the same review found local control and laryngeal preservation to favor open partial surgery where the anterior commissure was involved, and recurrence to be more frequent after transoral excision (8). The anterior commissure was involved in this patient, so the anatomy of the site and the histological subtype are both plausible contributors to what followed, and a single case cannot weigh one against the other. The problem was not the operation alone but the information it generated: the diagnosis derived from the resected tissue did not predict how the disease would behave.

Several factors limit what the initial specimen could show. The lesion was polypoid, and in laryngeal SpSCC the conventional squamous component often forms a thin cap over a predominantly spindled tumor body (2, 6). Frozen section, here taken from the surface of the lesion, is poorly suited to recognizing a spindle cell component. CO2 laser specimens are small, and thermal injury at the cut surface together with post-excision shrinkage further restricts what can be assessed (9). Thermal injury does more than reduce the assessable area: it can itself generate spindled nuclear profiles that mimic, and may equally conceal, a true spindle cell population, so that equivocal areas at the cut surface may need ancillary markers for interpretation. Re-cutting the original block after the salvage operation still showed no spindle cells, which argues against misinterpretation of the available material but cannot show whether a spindle cell component was present elsewhere in the tumor.

The same limitations bear on the margin assessment. Margins were reported as negative, yet a T3 recurrence with paraglottic extension appeared within 8 months. Thermal artifact, specimen shrinkage and orientation can all complicate margin assessment after laser excision (9), and orientation is particularly difficult at the anterior commissure, even in an en bloc specimen such as this one. Negative margins in the sections examined do not exclude subsequent recurrence, but recurrence alone does not establish that the initial resection was incomplete.

A second explanation should be acknowledged. The spindle cell phenotype may have developed in the recurrent tumor after the excision rather than having been present and missed. SpSCC is regarded as a monoclonal epithelial tumor in which the spindle cell population expands as the disease progresses (10, 11), and epithelial–mesenchymal transition is the mechanism usually invoked (12, 13). In the absence of clonality or molecular studies these possibilities cannot be separated in this patient. For clinical purposes, the distinction matters less than it may seem, because both lead to the same conclusion: a report of conventional squamous cell carcinoma from a small, superficial specimen does not exclude SpSCC and does not establish the biology of the lesion.

One observation is recorded as a caution rather than a recommendation. The Ki-67 labeling index in the original laser specimen was approximately 80%, higher than the approximately 60% in the salvage specimen. It was measured only after the salvage operation, on thermally damaged stored material and by counting the most active fields rather than by a standardized protocol, and the two indices come from different blocks handled at different times. The difference therefore says nothing reliable about tumor biology, and since no threshold has been established for this tumor, the higher figure cannot be treated in retrospect as a missed warning.

Surveillance after transoral laser microsurgery deserves comment. Recurrence in the operated larynx is not always visible at the mucosal surface; disease may persist beneath scar tissue or within deeper planes, where flexible laryngoscopy alone is unreliable (14). Peretti and colleagues grouped post-laser recurrences into those arising in the original area, those spreading superficially or multifocally to adjacent subsites, and those extending submucosally into the visceral spaces, cartilage or extralaryngeal tissues; none in the last group could be salvaged endoscopically (15). Laryngoscopy at 3 months in this patient was reassuring, yet 5 months later she had a submucosal recurrence with bilateral nodal disease and a mucosal surface that still appeared essentially normal. Guidelines recommend examination every 1–3 months in the first year, 2–6 months in the second and 4–8 months in years 3–5, with baseline imaging within 6 months of treatment, particularly for advanced disease, and further imaging for new symptoms or sites that are difficult to examine (16). Our planned schedule followed these intervals, but no visit is recorded between the 3-month examination and her re-presentation 5 months later, and no cross-sectional imaging had been obtained; in this patient the effects of that interval and of reliance on laryngoscopy cannot be separated. When the resected tumor is polypoid, crosses the anterior commissure, or shows any feature not fully explained by the clinical picture, contrast-enhanced MRI, or CT within the first postoperative year is a reasonable addition to endoscopic follow-up (14, 15).

Contrast-enhanced CT is an accepted staging work-up for early glottic carcinoma (16), but MRI gives better soft-tissue contrast (17) and is more sensitive, though less specific, for cartilage invasion (18), and both deep spread and cartilage invasion are at issue when a tumor involves the anterior commissure. Without MRI, a deep component not shown on CT cannot be excluded, although the later paraglottic recurrence does not in itself show that one was present at diagnosis. FDG-PET/CT would probably have added little at presentation: it is recommended mainly for stage III–IV disease (16), its resolution is limited for a 1-cm lesion, and uptake does not indicate histological subtype. At recurrence it could have supplemented whole-body restaging before salvage surgery; whether it would have changed management or outcome here cannot be determined.

After the salvage procedure, postoperative radiotherapy with concurrent cisplatin was recommended and declined. For pT3N2c disease, postoperative radiotherapy is standard, typically 60 Gy in 30 fractions over 6 weeks, with 66 Gy for extranodal extension or a positive margin (16). Concurrent cisplatin, 100 mg/m² every 3 weeks or 40 mg/m² weekly (19), has its clearest benefit when either of those features is present (20) and may be considered for other adverse features, including advanced nodal disease (16). Population-based data indicate that adjuvant radiotherapy following radical surgery is associated with improved survival in head and neck sarcomatoid carcinoma, the term used in that series, particularly in the presence of nodal disease (21). These are observational, group-level data, and whether adjuvant treatment would have altered the outcome in this patient cannot be established. After the initial excision, by contrast, adjuvant treatment was not indicated, and we know of no evidence that it would have helped even had the subtype been known.

Two practical points follow. The first concerns communication between surgeon and pathologist. When a lesion is bulky, polypoid, or ulcerated, this should be recorded on the request form together with the growth pattern and estimated depth observed at operation, so that the specimen is not assumed to be representative and the base and any morphologically distinct areas are sampled rather than the surface alone. The second concerns immunohistochemistry, which should be guided by the morphological differential diagnosis. It is not routinely required in an otherwise conventional squamous cell carcinoma simply because the lesion is polypoid, but it is useful when spindled cells are present, including equivocal profiles at a thermally damaged edge. Here it would not have altered the initial diagnosis, because no spindle cells were seen in the tissue examined, and the lessons of this case concern sampling and follow-up more than the choice of stains.

3.1 Strengths and limitations

The strengths of this report are a complete course from presentation to death, the availability of both specimens for comparison, and re-examination of the original block with an extended panel once SpSCC had been diagnosed. Its limitations are those of a single case described retrospectively. No autopsy was performed and the cause of death was not established. Molecular and clonality studies were not available, so we cannot determine whether the spindle cell population in the recurrent tumor was present at the outset or arose subsequently. Assessment of the original laser specimen was constrained by its size and by thermal artifact. No morphologically identifiable malignant spindle cell population was found in the initial sections or at the additional levels examined, and ancillary stains identified none, but they could not exclude a component that was not represented in the sampled tissue or was obscured by thermal damage. Staging used CT without MRI, and PET/CT was never performed, so deep extension at presentation and occult distant disease at recurrence cannot be excluded. Finally, the patient declined adjuvant treatment at every stage after the salvage operation, which limits any inference about the natural history of adequately treated disease.

4 Conclusion

Laryngeal SpSCC may present as a lesion that looks and stages like an early glottic carcinoma and may be reported as conventional squamous cell carcinoma when only superficial tissue is available. In this patient, laser excision of an apparently early glottic carcinoma, with negative margins, was followed by nodal metastasis and death within 2 years. Re-examination of the original block could not establish whether a spindle cell population had been missed or had not yet appeared, and in any comparable case that question is unlikely to be answerable once the lesion has been excised and thermally altered. The clinical response is the same under either reading. For polypoid or exophytic glottic tumors we suggest deep and representative sampling, immunohistochemistry whenever the morphology raises the possibility of a spindle cell component, consideration of contrast-enhanced MRI at staging when the anterior commissure is involved, and follow-up that includes cross-sectional imaging rather than laryngoscopy alone.

5 Patient perspective

The patient died before this report was prepared, and no account of her own experience of the illness had been recorded during her care. She and her family lived in a remote mountainous area far from our hospital and had limited means; for these reasons they declined postoperative chemoradiotherapy, and at the second recurrence they again chose not to pursue treatment.

Statements

Data availability statement

The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding author.

Ethics statement

The studies involving humans were approved by Ethics Committee of Yibin First People’s Hospital, Yibin First People’s Hospital, Yibin, Sichuan, China. The studies were conducted in accordance with the local legislation and institutional requirements. This was a retrospective single-patient case report based on existing clinical records. The patient had died before preparation of the report; written informed consent for publication of potentially identifiable images and data was obtained from the patient’s next of kin. The patient had died before preparation of the case report; therefore, written informed consent for publication could not be obtained from the patient herself. Written informed consent for the publication of potentially identifiable images and data was obtained from the patient’s next of kin.

Author contributions

YH: Conceptualization, Data curation, Writing – original draft. XL: Data curation, Investigation, Visualization, Writing – review & editing. YG: Investigation, Resources, Writing – original draft. LW: Investigation, Resources, Writing – original draft. KZ: Project administration, Supervision, Writing – review & editing.

Funding

The author(s) declared that financial support was not received for this work and/or its publication.

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fonc.2026.1979498/full#supplementary-material

Supplementary Figure 1

Findings at the second recurrence, March 2025. (A) Videolaryngoscopy showing a recurrent mass in the pharyngolaryngeal region. (B) Contrast-enhanced CT showing a recurrent mass in the right posterior paraglottic space.

References

Summary

Keywords

anterior commissure, case report, diagnostic pitfall, larynx, sarcomatoid carcinoma, spindle cell squamous cell carcinoma, transoral laser microsurgery

Citation

Huang Y, Luo X, Guo Y, Wang L and Zhang K (2026) Case Report: Rapid recurrence and nodal metastasis of laryngeal spindle cell squamous cell carcinoma initially underestimated on superficial sampling. Front. Oncol. 16:1979498. doi: 10.3389/fonc.2026.1979498

Received

26 August 2026

Revised

07 October 2026

Accepted

19 October 2026

Published

04 November 2026

Volume

16 - 2026

Edited by

Petr Szturz, Centre Hospitalier Universitaire Vaudois (CHUV), Switzerland

Reviewed by

Maria George Ioannou, University Hospital of Larissa, Greece

Saai Ram Thejas, RVM Institute of Medical Sciences and Research Center, India

Updates

Copyright

*Correspondence: Kai Zhang,

Disclaimer

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.

Outline

Figures

Cite article

Copy to clipboard


Export citation file


Share article

Article metrics