Lung Cancer Surgery in 2026
The Stages I–IIIA Pathway

1. The shift: from operation to integrated pathway

I have been performing lung cancer surgery for over a decade. The pathway I trained in is not the pathway I operate in today.

For most of the modern history of thoracic surgery, lung cancer surgery was a single decision: operate or do not operate, and if operating, take a lobe. Adjuvant chemotherapy was offered selectively. Immunotherapy was for advanced disease. The decision sat almost entirely with the surgeon at the point of resection.

That model has been displaced. Each of the six factors named above is determined upstream of the operating room. The operation is then chosen — open versus minimally invasive, lobectomy versus segmentectomy, with or without sleeve resection — to fit the case rather than as a default approach.

This is most visible at the two ends of the resectable spectrum. At one end, screening is generating a population of small, peripheral, often subsolid lesions where the clinical question is increasingly "is sublobar resection appropriate here?" rather than "should we do a lobectomy?" At the other, neoadjuvant chemoimmunotherapy is downstaging tumours that would historically have been called inoperable — and the systemic therapy is being delivered before the patient ever reaches the surgical clinic.

Patient fitness is no longer a yes-or-no gate at this stage of the pathway either. The modern pre-operative fitness assessment — structured against NICE NG122 thresholds, with quantitative ventilation-perfusion imaging where COPD is part of the picture — identifies patients who can be operated on with appropriate technique and prehabilitation despite borderline lung function or established cardiac disease. Where the cancer sits in destroyed emphysematous lung, the combined cancer-and-lung-volume-reduction pathway can deliver cancer cure and improved breathing in a single operation.

The remainder of this piece walks through the arc — from screening to surgery to systemic therapy — anchoring international evidence in current UK practice and, where appropriate, in my own institutional and personal experience. For patients seen elsewhere who want an independent review of imaging, pathology, MDT decision, and treatment plan before committing to a course of action, the specialist second opinion service is available within 2–3 working days.

2. Screening: low-dose computed tomography and the UK Lung Health Check

Foundation: NLST and NELSON

The clinical case for low-dose computed tomography (LDCT) screening in high-risk individuals was established by two pivotal trials. The National Lung Screening Trial (NLST), published in 2011, demonstrated a 20% relative reduction in lung cancer mortality in current and former smokers screened with LDCT compared with chest radiography [1]. The NELSON trial, published in 2020, confirmed the mortality benefit in a European population, with a 26% reduction in lung cancer mortality in men and 39–61% reduction in women [2]. These two trials underpin every modern programme.

The UK programme: implementation data through May 2026

The NHS Lung Cancer Screening Programme — formally renamed in February 2025 from the Targeted Lung Health Check Programme — invites current and former smokers aged 55 to 74 to a structured risk assessment, with low-dose CT offered to those at elevated risk based on multivariable models such as PLCOm2012 (≥1.51% over six years) or LLPv2 (≥2.5% over five years) [3].

Five-year implementation data published by Lee and colleagues in Nature Medicine in March 2026 provide the most authoritative current view of programme performance. Across 2.5 million invited individuals and 7,193 cancers diagnosed, 75.7% of screen-detected cancers are now found at Stage I or II — compared with fewer than 30% of cancers detected outside screening [4]. This represents the largest single shift in stage-at-presentation in the modern era of lung cancer care.

The corresponding pressures on the surgical service are well documented. Approximately 15 in every 100 individuals scanned have a lung nodule [5]. The vast majority are benign, but the conventional surveillance pathway involves repeat imaging at three, six, or twelve months — generating sustained anxiety and limited capacity to act. The National Lung Cancer Audit 2026 (NATCAN State of the Nation) reported that 88% of patients with early-stage lung cancer waited longer than the 49-day target from referral to surgery, and that resection volumes rose from 6,547 in 2023 to 7,878 in 2024 — a 20% single-year increase driven directly by screening [6].

AI-assisted nodule risk stratification

In January 2026, NHS England announced a national pilot integrating AI-based risk stratification (Optellum) with low-dose CT in the screening pathway, with concurrent rollout of robotic bronchoscopy for indeterminate nodules [7]. Guy's and St Thomas' NHS Foundation Trust is one of the lead participating centres.

The clinical purpose of AI-assisted reading is twofold: to reduce inter-reader variability in nodule identification, and to apply quantitative malignancy probability scoring to small nodules under 8 mm where size alone is uninformative. Where the AI flags a nodule above a malignancy probability threshold, the patient can be moved directly into expert assessment rather than routine surveillance — compressing the diagnostic timeline by months in selected cases.

Outcomes at a high-volume centre — what the audit data shows

Guy's and St Thomas' is currently the UK's largest single-site lung cancer surgical centre by Society for Cardiothoracic Surgery (SCTS) national audit volume, with 969 primary lung cancer resections in 2024–25 — approximately one in eight UK operations [8]. The National Lung Cancer Audit 2026 publishes nationally comparable performance data for every NHS trust in England. The figures for GSTT demonstrate consistently better outcomes than the national average across every major indicator [6]:

Indicator	England 2024	GSTT 2024
Patients diagnosed at Stage I or II	40%	58%
NSCLC patients having surgical resection	22%	44%
One-year survival after diagnosis	51%	72%
Curative treatment rate, Stage I–II	79%	88%
Diagnosed after emergency presentation	30%	17%
90-day post-operative mortality after lung resection	1.5%	1.2%

GSTT's surgical resection rate of 44% — double the national average — reflects a centre that is diagnosing more patients at an operable stage and offering surgery to a higher proportion of those who are candidates. A one-year survival of 72% against a national figure of 51% reflects the consequence of getting those decisions right.

As Clinical Audit Lead for thoracic surgery at GSTT, I oversee the SCTS returns process and can confirm these figures are footer-audited against the audit submission. The audit is institutional. The standard it reflects is what private patients receive — because the same surgeon, working to the same standards, performs both NHS and private operations using equivalent infrastructure across institutions.

3. Diagnostic precision: ION robotic bronchoscopy and the diagnostic gap

The diagnostic gap problem

Screening identifies more nodules. The conventional NHS surveillance pathway addresses uncertainty through repeat imaging — but for nodules with intermediate or high malignancy probability, surveillance carries clinical and psychological costs. By the time interval growth confirms malignancy, the optimal surgical window may have narrowed and the patient has lived with months of avoidable uncertainty.

The clinical case for direct tissue diagnosis at appropriate risk thresholds is now strong. The instrument that has made it scalable is robotic bronchoscopy.

Technology

The Intuitive Ion endoluminal system uses a 3.5 mm shape-sensing catheter that navigates through the airways under continuous real-time tracking. A cone-beam CT scan confirms catheter position immediately before sampling. The platform reaches all 18 anatomical lung segments — including peripheral segments inaccessible to conventional bronchoscopy — and delivers tissue samples to a pathologist who confirms specimen adequacy in the procedure room (rapid on-site evaluation, ROSE) [9, 10].

Key clinical advantages over CT-guided transthoracic needle biopsy:

• Pneumothorax rate of 2–3% versus approximately 28% for CT-guided needle biopsy in published series [10, 11].
• Diagnostic yield of 76–90% across published cohorts. The largest single-centre series — Mayo Clinic, approximately 1,900 patients — reports 85% sensitivity for malignancy and 77% strict diagnostic yield with a 2.8% complication rate [12].
• Concurrent EBUS-TBNA mediastinal staging in the same anaesthetic episode, allowing complete diagnosis and staging from one procedure.
• Fluorescent dye marker placement at biopsy for surgical localisation if resection follows.

The therapeutic counterpart of this diagnostic capability — rigid and flexible bronchoscopy with airway stenting, dilatation, and ablative therapy — is described separately in the central airway interventions guide.

UK programme experience

The GSTT navigational bronchoscopy programme performed approximately 635 ION cases in 2024–25, with diagnostic yields of 76–89% across a range of nodule sizes and a pneumothorax rate of approximately 2% [8]. London Bridge Hospital was the first private provider in Europe to offer ION outside clinical trials [13], and the integrated NHS England AI-and-Ion pilot launched in January 2026 builds directly on this institutional infrastructure [7].

Combined biopsy-and-surgery: single-anaesthetic pathway

For carefully selected fit patients with small peripheral lesions of high malignancy probability, the diagnostic and surgical episodes can be combined under a single general anaesthetic. ION bronchoscopy is performed first; ROSE confirms malignancy in real time; if confirmed, robotic anatomical resection follows immediately at the same theatre session. Where the criteria are met, this eliminates a second anaesthetic, compresses the time from diagnosis to definitive treatment to a single day, and removes the handoff between diagnostic and surgical teams.

In my own private practice at London Bridge Hospital, this is the integration that distinguishes the ION-equipped pathway from a conventional one. Full details of the combined biopsy-and-resection pathway → The institutional infrastructure that supports it — the cone-beam CT, the on-site cytopathology, the immediate availability of robotic theatre — was developed first within the NHS programme at GSTT and then extended privately. This is the relevant order. The private pathway does not exist in isolation; it draws directly on the operational scale of a centre operating at national volume.

Where conventional techniques still fit

CT-guided transthoracic needle biopsy retains a role for very peripheral lesions where bronchoscopic access is anatomically difficult, but its share of the diagnostic landscape is shrinking as robotic platforms mature. EBUS-TBNA remains the standard for mediastinal and hilar nodal staging (N2/N3) and is routinely performed in the same anaesthetic episode as ION when both peripheral and nodal sampling are required.

4. Molecular testing: front-loaded, not afterthought

The shift to upfront testing

Until recently, molecular testing in resectable disease was performed on the surgical resection specimen — that is, after the operation. The treatment implications were limited to adjuvant decisions, and biomarker turnaround did not constrain the surgical timeline.

That has reversed. In 2026, molecular testing must occur on the diagnostic biopsy, before any systemic therapy decision is made. The reason is straightforward: giving a course of chemoimmunotherapy to a patient with an EGFR-mutated tumour offers limited clinical benefit, may complicate later targeted-therapy use, and consumes time that could have gone toward an effective regimen. The evidence base discussed in section 6 makes this a non-trivial error.

The required panel

International consensus (IASLC, ESMO, NCCN, CAP) now recommends comprehensive next-generation sequencing (NGS) on tissue from the diagnostic biopsy in all resectable NSCLC where biomarker results affect management [14, 15]. The minimum panel includes:

• EGFR mutation, including exon 19 deletions, L858R, exon 20 insertions, and atypical mutations
• ALK rearrangement
• ROS1 rearrangement
• BRAF V600E mutation
• MET exon 14 skipping mutation
• RET rearrangement
• NTRK fusion
• HER2 mutation/amplification
• KRAS mutation, particularly G12C
• PD-L1 expression by immunohistochemistry

NICE references the National Genomics Test Directory for NGS panel content [16]. Liquid biopsy (circulating tumour DNA) is emerging as an adjunct where tissue is limiting but is not yet a complete substitute for tissue-based testing.

Implications for treatment selection

The panel above sorts patients into three broad treatment streams:

Result	Pre-operative strategy
EGFR mutation (sensitising)	Surgery ± adjuvant chemotherapy → adjuvant osimertinib (ADAURA pathway)
ALK rearrangement	Surgery ± adjuvant chemotherapy → adjuvant alectinib (ALINA pathway)
Driver-negative, PD-L1 ≥1%	Neoadjuvant chemoimmunotherapy → surgery → adjuvant immunotherapy
Driver-negative, PD-L1 <1%	Neoadjuvant chemoimmunotherapy still indicated based on EFS benefit

Patients with EGFR or ALK alterations were specifically excluded from the chemoimmunotherapy trials — they have their own pathway, and chemoimmunotherapy is generally avoided in this group. For a patient who has been offered neoadjuvant therapy without these test results, the most important question to ask before treatment starts is whether the testing has been done.

5. Surgery: robotic, lung-sparing, anatomically precise

The lobectomy-to-segmentectomy paradigm shift

For nearly three decades, lobectomy was the default operation for early-stage lung cancer. The lobe was the smallest unit considered to deliver oncologically adequate resection, and sublobar resection was reserved for patients with insufficient pulmonary reserve. Two large randomised trials have changed this position.

JCOG0802 / WJOG4607L

Saji and colleagues, reporting in The Lancet in 2022, randomised 1,106 Japanese patients with peripheral non-small cell lung cancer of 2 cm or less and consolidation-to-tumour ratio ≥0.5 to anatomical segmentectomy or lobectomy [17]. At five years, segmentectomy demonstrated superior overall survival (94.3% vs 91.1%, hazard ratio 0.66). Local recurrence was modestly higher with segmentectomy but did not translate into worse cancer-specific survival. The interpretation is that the preservation of pulmonary function with segmentectomy contributes to an overall survival benefit independent of cancer-specific outcomes.

CALGB 140503

Altorki and colleagues, reporting in the New England Journal of Medicine in 2023, randomised 697 North American patients with peripheral NSCLC of 2 cm or less and clinically node-negative disease to sublobar resection (predominantly segmentectomy) or lobectomy [18]. At five years, sublobar resection demonstrated non-inferior disease-free survival (63.6% vs 64.1%) and overall survival (80.3% vs 78.9%). Lung function preservation was significantly better with sublobar resection.

Together, these two trials establish segmentectomy as an oncologically equivalent — and in selected cases superior — option for small peripheral early-stage lung cancer. Full clinical detail on robotic segmentectomy →

The boundaries of the evidence

Both trials restricted eligibility to peripheral tumours of 2 cm or less with no clinical evidence of lymph node involvement. For tumours larger than 2 cm, central tumours, or tumours with PET-CT or staging suspicion of N1 or N2 nodal disease, lobectomy remains the standard of care. Segmentectomy is not a shortcut for larger tumours; it is a precision operation reserved for the patients whose disease makes it the right choice.

IASLC T-stage data — why every centimetre matters

Within Stage I, every centimetre of tumour size matters. The IASLC dataset, originally analysed by Rami-Porta and colleagues across 77,156 patients in 2015 and validated by Van Schil and colleagues across 124,581 patients in 2024, shows the following five-year survival gradient by T-stage [19, 20]:

T-stage	Tumour size	5-year survival
T1a	≤1 cm	92%
T1b	>1–2 cm	83%
T1c	>2–3 cm	76%

This is the data that makes the clinical case for screening. Finding cancer at T1a versus T1c is an absolute survival difference of 16 percentage points, before any consideration of treatment. Screening exists to move patients leftward on this gradient.

Robotic versus VATS versus open

Minimally invasive surgery (MIS) — either video-assisted thoracoscopic surgery (VATS) or robotic-assisted thoracic surgery (RATS) — is the standard for resectable Stage I–IIIA disease in fit patients. Open thoracotomy is reserved for complex resections, chest wall involvement, or major vascular reconstruction.

Within MIS, the case for robotic surgery rests on three technical advantages of clinical relevance:

• Magnified three-dimensional vision with stereoscopic depth perception, particularly relevant for the precision required in segmentectomy along the intersegmental plane.
• Wristed instrument articulation through small thoracic ports, allowing dissection in tight anatomical spaces — particularly relevant when fibrosis obliterates anatomical planes (as after neoadjuvant chemoimmunotherapy).
• Improved lymph node yield and lower conversion to open. A 2023 systematic review and meta-analysis (Tasoudis et al., Journal of Thoracic Disease) of robotic versus VATS lobectomy demonstrated higher lymph node retrieval and lower conversion-to-open rates with robotic surgery, with comparable oncological outcomes [21] — relevant because completeness of lymphadenectomy contributes to accurate staging and informs adjuvant decision-making.

Personal and institutional standards

In 2024–25 my institutional unit at GSTT performed 892 anatomical lung cancer resections, of which 71.3% were robotic — approximately three times the national average of 24% [8]. The institutional pneumonectomy rate was 1.5% in the same period, among the lowest in the UK without compromise to operative survival rate, which was 99.59%. Of segmentectomies, 88.9% were performed robotically, reflecting the technical preference for the platform when the operation requires precise definition of the intersegmental plane.

Within that unit in the same year, I personally performed 153 anatomical resections. My personal robotic segmentectomy rate was 22.7% of robotic anatomical resections — closely tracking the institutional standard. My personal wedge resection rate was 6%, against a UK national average of approximately 14%, reflecting a deliberate choice to perform anatomical resection where it is feasible rather than retreating to a non-anatomical wedge for technical reasons.

These figures matter because they describe the operating-room environment within which clinical decisions are made. The choice of segmentectomy versus lobectomy for a small peripheral tumour is not a theoretical preference. It is a daily decision against a background of nearly 900 anatomical operations per year. That volume is the precondition for delivering segmentectomy reliably — and it is the institutional standard from which my private practice operates.

Surgery after neoadjuvant chemoimmunotherapy

Surgery after three or four cycles of chemoimmunotherapy is more technically demanding than surgery on an untreated tumour. The combination of platinum chemotherapy and PD-1/PD-L1 blockade induces inflammation, fibrosis, and obliteration of the natural anatomical planes between structures that ordinarily make minimally invasive dissection feasible.

Many centres respond to this difficulty by converting to open thoracotomy. This is a clinically meaningful problem because the patients who have just completed four months of systemic therapy are precisely the patients least able to tolerate a large open operation. The robotic platform — with magnified stereoscopic vision and articulated instruments — supports dissection through fibrotic planes more reliably than VATS, and at high-volume centres surgery after chemoimmunotherapy is increasingly delivered robotically without conversion [22].

In my own experience operating on this group, the combination of platform and high case volume matters more here than at any other point in the resectable spectrum. The fibrotic planes that follow neoadjuvant therapy reward a surgical team that encounters them routinely. Centres performing two to three operations of this type per month operate in a fundamentally different environment from those performing two to three per year. This is the patient group for whom case volume is most directly translated into outcome. Detailed account of robotic surgery after chemoimmunotherapy →

Extended resections: locally advanced disease

For patients with Stage IIIA disease involving chest wall, mediastinal structures, or requiring sleeve reconstruction, the surgical pathway is more complex but resectability is often preserved with modern multidisciplinary planning. The locally advanced lung cancer pathway → sets out the staging requirements, neoadjuvant sequencing, and surgical approach for this group, alongside the broader robotic lung surgery service overview.

6. Perioperative systemic therapy: the largest shift

The trials differ in design (neoadjuvant vs perioperative), agent, and patient population. They converge on the same conclusion.

Real-world UK experience

The Guy's Cancer Centre real-world series on neoadjuvant nivolumab plus chemotherapy in operable NSCLC was published in Lung Cancer in 2025 (Toki et al.) [31]. As a co-author of that series, I can confirm that the trial outcomes seen in CheckMate 816 are reproducible in routine UK practice when delivered by an experienced multidisciplinary team. The operational challenge is not the regimen itself — it is the multidisciplinary infrastructure required to deliver biomarker testing, restaging, and surgery within the right timing windows.

The neoadjuvant-only versus perioperative debate

CheckMate 816 used neoadjuvant chemoimmunotherapy only — three cycles before surgery, no adjuvant component. KEYNOTE-671, AEGEAN, CheckMate 77T, Neotorch, and RATIONALE-315 used a perioperative design — chemoimmunotherapy before surgery, then immunotherapy continuation after.

The evidence base does not yet definitively establish whether the post-operative immunotherapy component adds incremental benefit beyond the neoadjuvant component for individual patients. A 2025 ESMO Open meta-analysis suggested that perioperative regimens have a numerical advantage in event-free survival over neoadjuvant-only, but the comparison is indirect and ongoing trials are designed to address this directly [32]. In current UK practice, the regimen used in the relevant NICE-approved indication determines the perioperative approach.

Driver-positive disease: ADAURA, NeoADAURA, and ALINA

The ADAURA trial (Tsuboi et al., NEJM 2023) randomised 682 patients with completely resected Stage IB–IIIA EGFR-mutated NSCLC to three years of adjuvant osimertinib or placebo. Disease-free survival hazard ratio was 0.27 in Stage II–IIIA — the largest single benefit in adjuvant lung cancer therapy in the modern era. Five-year overall survival was 88% with osimertinib vs 78% with placebo [33].

The NeoADAURA trial, presented at ASCO 2025 by Tsuboi and colleagues, evaluated neoadjuvant osimertinib with or without chemotherapy in resectable Stage II–IIIB EGFR-mutated NSCLC and demonstrated significantly improved major pathological response with neoadjuvant osimertinib-containing regimens [34]. Adoption of neoadjuvant osimertinib is currently progressing through national appraisal processes.

For patients with EGFR-mutated resectable lung cancer, the standard of care in 2026 is therefore: surgery (with or without adjuvant chemotherapy depending on stage and risk) followed by three years of adjuvant osimertinib.

The ALINA trial (Wu et al., NEJM 2024) randomised 257 patients with completely resected Stage IB–IIIA ALK-positive NSCLC to two years of adjuvant alectinib or platinum-based chemotherapy. Disease-free survival hazard ratio was 0.24 — comparable to ADAURA in EGFR-mutated disease [35]. Adjuvant alectinib is being adopted into UK practice through emerging NICE pathways. For patients with ALK-positive disease, immunotherapy is generally avoided in the perioperative setting.

Adjuvant immunotherapy: IMpower010 and KEYNOTE-091

Two phase 3 trials established adjuvant immunotherapy as a treatment option for completely resected Stage II–IIIA NSCLC after platinum-based adjuvant chemotherapy.

• IMpower010 (Felip et al., Lancet 2021): Adjuvant atezolizumab in patients with PD-L1 ≥1%. Disease-free survival hazard ratio 0.66 in the Stage II–IIIA PD-L1-positive subgroup [36]. Approved by NICE as TA1071 in June 2025.
• KEYNOTE-091/PEARLS (O'Brien et al., Lancet Oncology 2022): Adjuvant pembrolizumab. Disease-free survival hazard ratio 0.76, with benefit consistent across PD-L1 strata [37].

In 2026, adjuvant-only immunotherapy is largely a strategy for patients who did not receive neoadjuvant immunotherapy — for example, those who had upfront surgery without prior systemic therapy. As perioperative chemoimmunotherapy becomes the dominant strategy for resectable Stage II–IIIA disease, the role of adjuvant-only immunotherapy is contracting.

7. NICE guidance: what is approved in the UK

The National Institute for Health and Care Excellence (NICE) reviews evidence and makes formal recommendations that govern NHS access to therapies in England and Wales. As of May 2026, the following appraisals are central to the resectable NSCLC pathway:

• NG122 — Lung cancer: diagnosis and management — the framework guideline, last major update 2024, requiring multidisciplinary team management, biomarker testing, and multi-modality treatment planning [38].
• TA876 (March 2023) — Neoadjuvant nivolumab plus chemotherapy for resectable NSCLC ≥4 cm or with nodal involvement, based on CheckMate 816 [39].
• TA761 (2022) — Adjuvant osimertinib for completely resected EGFR-mutated NSCLC, based on ADAURA [40].
• TA1071 (June 2025) — Adjuvant atezolizumab for PD-L1-positive NSCLC after surgery and adjuvant chemotherapy, based on IMpower010 (Cancer Drugs Fund review of TA823) [41].
• TA1127 (4 February 2026) — Perioperative nivolumab plus chemotherapy followed by adjuvant nivolumab for resectable NSCLC ≥4 cm or with nodal involvement, based on CheckMate 77T [43].
• ID5094 (in appraisal) — Perioperative pembrolizumab, based on KEYNOTE-671 [42].

Additional appraisals covering perioperative durvalumab (AEGEAN), perioperative nivolumab (CheckMate 77T), and adjuvant alectinib (ALINA) are at various stages of submission and review. The pace of NICE appraisal activity in resectable lung cancer therapy in 2025–26 has been the highest in any oncology area.

For private patients, NICE recommendations also inform private medical insurance coverage: treatments with NICE recommendations are generally included within mainstream UK private medical insurance cancer benefits.

8. The pathway in stages I to IIIA

The general principles described above translate into stage-specific pathways. The decision algorithm depends on stage, biomarker status, and patient fitness.

Stage IA (≤2 cm, peripheral, node-negative)

• Surgery: Robotic anatomical segmentectomy with systematic mediastinal lymph node sampling. Lobectomy remains an option, particularly where margin or nodal staging concerns favour it.
• Systemic therapy: Generally surgery alone. EGFR-mutated patients in this stage do not currently receive adjuvant osimertinib outside Stage IB and above.

Stage IA (>2 cm) and Stage IB

• Surgery: Robotic lobectomy with systematic mediastinal lymph node dissection. Sublobar resection only in patients with insufficient pulmonary reserve, where structured fitness assessment supports the smaller operation.
• Systemic therapy: Adjuvant chemotherapy considered for high-risk features (visceral pleural invasion, poor differentiation, lymphovascular invasion). For EGFR-mutated patients in Stage IB and above, adjuvant osimertinib for three years (TA761).

Stage II

• Pre-operative work-up: Full staging with PET-CT, brain imaging (preferably MRI), EBUS-TBNA for mediastinal nodes if any radiological suspicion.
• Driver-negative: Neoadjuvant or perioperative chemoimmunotherapy followed by robotic anatomical resection. Adjuvant immunotherapy completion as per regimen.
• EGFR-mutated: Surgery (with or without adjuvant chemotherapy) followed by three years of adjuvant osimertinib (TA761).
• ALK-positive: Surgery (with or without adjuvant chemotherapy) followed by two years of adjuvant alectinib (emerging UK pathway).

Stage IIIA, resectable

• Pre-operative work-up: As Stage II, plus mandatory mediastinal staging (EBUS-TBNA, ± mediastinoscopy where indicated).
• Multidisciplinary team review is mandatory before treatment starts. The MDT determines whether the disease is genuinely resectable. Bulky multistation N2 disease may favour definitive chemoradiotherapy with consolidation durvalumab over surgical pathway.
• Driver-negative resectable: Neoadjuvant or perioperative chemoimmunotherapy followed by robotic anatomical resection where feasible. Surgery after neoadjuvant therapy is technically more demanding due to fibrosis; high-volume centres deliver this routinely. Detailed locally advanced pathway →
• EGFR-mutated resectable: Surgery and adjuvant osimertinib pathway as for Stage II.
• ALK-positive resectable: Surgery and adjuvant alectinib pathway as for Stage II.

Stage IIIA, borderline or unresectable

Definitive concurrent chemoradiotherapy followed by consolidation durvalumab (PACIFIC paradigm) [44]. Some patients may be reconsidered for surgery after favourable response to induction therapy.

9. Emerging directions

Circulating tumour DNA and minimal residual disease

The most active area of current investigation is the use of circulating tumour DNA (ctDNA) to identify minimal residual disease after surgery and to guide adjuvant therapy decisions. Patients who clear ctDNA after surgery may not need adjuvant therapy; patients who remain ctDNA-positive may need escalation. Multiple ongoing trials are addressing this, including biomarker-driven adjuvant escalation and de-escalation studies [45]. ctDNA is not yet embedded in routine NICE-approved pathways but is likely to become a major decision tool over the next three to five years.

Sublobar resection in subsolid disease

The screening era is generating a population of part-solid and pure ground-glass nodules with indolent clinical behaviour. The boundaries of sublobar resection — particularly wedge versus segmentectomy in ground-glass-dominant lesions — are under active investigation. The European Society of Thoracic Surgeons issued specific guidance on the surgical management of ground-glass opacities in 2023 [46].

Optimal duration of adjuvant immunotherapy

KEYNOTE-671 specified up to thirteen post-operative cycles of pembrolizumab. AEGEAN specified twelve cycles of durvalumab. Whether shorter durations achieve equivalent benefit — particularly in patients who achieved pathological complete response with neoadjuvant therapy — is an open question.

Chemotherapy-free neoadjuvant strategies

Trials of chemotherapy-free neoadjuvant immunotherapy in selected high-PD-L1 patients are ongoing. If positive, this would shift treatment for the highest-PD-L1 subgroup away from chemotherapy entirely.

Sequencing in driver-positive disease

The optimal sequencing of targeted therapy and immunotherapy in patients with both driver mutations and high PD-L1 expression remains an area of active investigation, with implications for subsequent advanced-disease management.

10. The UK pathway in practice

NHS multidisciplinary team management

NICE NG122 mandates that all lung cancer cases are reviewed by a multidisciplinary team comprising thoracic surgeons, medical oncologists, clinical oncologists, thoracic radiologists, respiratory physicians, and histopathologists. The MDT is the formal decision-making body for treatment planning. In 2026, the central role of MDT discussion is the integration of biomarker results, PET-CT staging, imaging assessment of resectability, and the perioperative therapy plan into a single coordinated pathway.

Time-to-treatment performance

NLCA 2026 reports that 88% of patients with early-stage lung cancer waited longer than the recommended 49-day target from referral to surgery [6]. This is a structural pressure created by screening-driven volume increases outpacing surgical capacity expansion. The Royal College of Surgeons of England has explicitly called for surgical capacity expansion in response.

Private and self-pay routes

Private practice can compress the diagnostic and treatment timeline meaningfully. Where NHS waits to first specialist consultation can extend beyond two weeks, private consultation is typically available within 2–3 days. Where the diagnostic work-up may extend over four to six weeks in NHS pathways, ION robotic bronchoscopy and PET-CT can be arranged within ten to fourteen days privately. Surgery can typically follow within two to four weeks of the diagnostic decision, depending on case complexity.

The clinical content of treatment is the same in both settings — the same MDT discussion, the same trial-based regimens, the same surgical techniques. The difference is in the operational compression of the timeline.

For locally advanced disease requiring multimodality treatment, the systemic therapy duration of three to four months is determined by the regimen rather than by access. Private and NHS surgical timelines after chemoimmunotherapy converge in this group.

A clinician's perspective

Summary

Lung cancer surgery in May 2026 is best understood as a single integrated pathway across screening, diagnosis, surgery, and systemic therapy. The four shifts that define current practice are:

• More patients arriving at Stage I or II through screening rather than symptomatic presentation.
• Diagnostic precision through robotic bronchoscopy, with shorter timelines and lower complication rates.
• Lung-sparing anatomical segmentectomy for selected small peripheral tumours.
• Perioperative chemoimmunotherapy as the standard of care for resectable Stage II–IIIA driver-negative disease, with targeted adjuvant therapy for EGFR-mutated and ALK-positive disease.

The trial base supporting these shifts has been built systematically across the past four years. NICE has approved key components and is reviewing further appraisals on a near-continuous basis. UK practice is well-aligned with international evidence, with structural challenges focused on time-to-treatment rather than treatment content.

The patient who presents with a Stage IIIA NSCLC in May 2026 is operating in a different therapeutic landscape from the patient who presented in May 2022.