An Evolving Understanding of Adenosquamous Carcinoma of the Lung

Rajwanth Veluswamy, MD, MSCR

Adenosquamous carcinoma (ASC) of the lung is a rare, biphasic type of non-small cell lung cancer (NSCLC) that accounts for 2% to 4% of all lung cancers.¹ According to the World Health Organization (WHO) classification, the composition of ASC includes both adenocarcinoma (AC) and squamous cell carcinoma (SCC) histologies, with each subtype comprising at least 10% of the tumor.² As with other lung cancers, the average age at ASC diagnosis is about 70 years of age, it affects more men than women, and most patients are current or former smokers.^3,4 Despite these similarities, mounting evidence suggests that the molecular and genomic features of ASC are unique and they remain poorly understood.^5-8

Perhaps owing to the distinct genomics of these tumors, ASC of the lung is reported to be relatively aggressive compared to typical AC and SCC tumors. Studies indicate that ASCs at diagnosis have higher rates of lymph node invasion, metastasize rapidly, and carry a generally poor prognosis. Accordingly, the overall survival (OS) of patients with these tumors is relatively short compared to other NSCLC subtypes.^2,3,8-10 In a 2022 population-based study of the SEER database, 5-year postsurgical survival rates for early-stage cancers were reportedly 65% for ASC vs 69% for SCC P=0.003 and 77% for AC P<0.001.³ While it is clear that underlying biology driving ASC differs from more typical NSCLC subtypes, there is a lack of effective treatment options specific to ASC and a paucity of clinical research available to support therapeutic decisions for patients with ASC histology. Current management of NSCLC is based primarily on the stage of the tumor, and clinical features of the patient. In a more personalized era of targeted treatments, tumor histology is used only to predict the presence of actionable mutations in adenocarcinomas.^7,8 However, optimal treatment strategies for ASC remain a significant unmet need in lung cancer.

Diagnosis: Complex but Critically Important

Given the mixed histologies that characterize ASC of the lung, intratumoral heterogeneity often hinders and may delay diagnosis. Studies suggest that ASC is misdiagnosed as AC or SCC in at least half of biopsies prior to surgical pathology confirming an ASC diagnosis.¹¹ In one retrospective study, nearly all ASC cases (98%) were either misdiagnosed or undiagnosed preoperatively.¹² What’s more is that different types of biopsy samples may yield different results. One case report of a patient eventually diagnosed with ASC described 3 different results on workup: SCC on bronchial lavage and bronchial biopsy, AC on immunohistochemistry, and NSCLC undifferentiated on pleural effusion cytology.¹³ While a diagnosis can be made using biopsy and cytology samples, a definitive diagnosis may require larger samples (ie, several core biopsies or complete surgical resections) to fully evaluate all components of the tumor lesion.

Comprehensively evaluating entire tumor specimens can aid in further characterization ASC of the lung. ASCs may be sub-classified according to the proportions of AC and SCC histology components present. Tumors with either AC or SCC components comprising at least 60% of the tumor are referred to as AC- or SCC-predominant ASC, respectively. Those with a more even split of AC and SCC histologies (40% to 60% of each) are referred to as structure-balanced ASC and have been reported to have a better prognosis than either of the more imbalanced subtypes.^9,14

Adding to the complexity of diagnosing ASC of the lung is its unclear histologic origin and the transitional nature of these tumors over time. Some studies have pointed to possible precursor lesions, including AC with squamous metaplasia, collision tumor, and high-grade mucoepidermoid tumors.¹⁵ Reports have also shown that the molecular and histological features of the primary tumor can differ from that of metastases/recurrences.^16,17 In one case report, a patient with a resected ASC harboring an epidermal growth factor receptor (EGFR)-sensitizing mutation recurred several months later as SCC in the brain with the same EGFR mutation. A later recurrence in the lung was diagnosed as an AC and had the same EGFR mutation.¹⁶ In this example, if only the SCC component had been diagnosed, molecular testing would likely have never been ordered and the potentially actionable EGFR mutation would have been left undetected. Therefore, careful and accurate diagnosis of ASC is critically important in guiding testing for driver mutations, as well as in informing treatment choices in ASC.

Genomics

Studies indicate that ASC of the lung exhibits genomic features of both AC and SCC, with standard immunohistochemical profiles represented in each component. As expected, TTF1 positivity is common in the AC component while p63 and CK5/6 are expressed in the SCC component.¹⁸ However, evidence also indicates that ASC of the lung is a distinct entity rather than being a simple hybrid of AC and SCC histologies. That is, despite the seemingly dichotomous nature of ASC, this type of tumor is thought to have unique molecular and genomic features that have not yet been fully identified.^5-8

While the genomics of AC and SCC of the lung have been well studied, the inherent intratumoral heterogeneity that defines ASC, together with its relative rarity, complicates its analysis. There is a paucity of data available, but several groups have conducted molecular testing to better understand the genotype of ASC and potentially discover predictors about prognosis and treatment. To date, most studies on ASC lung samples have been small, and while some groups have reported overlapping results, other findings contrast with one another. In one of the most recent and comprehensive studies published on the topic, Wang et al. used next-generation sequencing (NGS) to identify a wide range of somatic mutations in 124 Chinese patients with ASC of the lung, including TP53 (66.9%), CDKN2A (21%), TERT (21%), and LRP1B (18.5%).⁶ Importantly, they found high rates of EGFR mutations (54.8%), of which 45.6% were EGFR 19del, 38.2% were EGFR L858R and 29.4% were EGFR amplifications.

Notably, not all studies have found such a strikingly increased rate of EGFR mutations in ASC versus AC of the lung.¹⁹ Other actionable mutations were found in the analysis by Wang and colleagues, including ALK and ROS1 fusions. Regarding known predictors of immunogenicity in these tumors, a subset of patients were associated with high tumor mutational burden (TMB), which was correlated with mutations in ARID2, BRCA1, and KEAP1. Immunohistochemical analyses demonstrated half of patients were positive for PD-L1 (≥ 1% tumor proportion score [TPS]).⁶ Interestingly, another study showed that PD-L1 expression in ASC differed between SCC (30% to 40%) and AC (11% to 15%) components.²⁰

Actionable mutation rates (ie, EGFR, ALK) in AC are known to vary between Asian and White patients, a finding that seems to be similar in ASC of the lung as well, although it is less clear given the limited sample size of ASC studies. Vassella et al. performed NGS and fluorescence in situ hybridization (FISH) on ASC samples from 16 White patients and found that 30% had EGFR mutations, while Tochigi et al reported an EGFR mutation rate of 13% in a study of 23 Western patients.5,12,21 In their analysis, Vassella and colleagues also found a high rate of mutations in the PI3K pathway (25%), but no KRAS mutations, which are the most common molecular driver in typical AC (30%), and thus supporting the notion that ASC has its own molecular genomic profile, distinct from AC or SCC.^5,21 Also of interest in this study was the finding that classifier miR-205 expression was intermediate between that of classical AC and SCC, suggesting that ASC of the lung may alternatively represent a transitional stage between these tumor types rather than an unrelated entity.⁵ These findings, along with others that have been reported on the genomic landscape of ASC, have advanced our understanding of the underlying biology of this malignancy, but also highlight the unmet need for more research to improve our ability to personalize treatment for ASCs.

Treatment

Owing to the heterogeneity of ASC of the lung, as well as its complex and incompletely characterized genomic landscape, treating patients with these tumors is challenging. In general, stage-based treatment approaches are used to manage ASC. The current treatment paradigm of all NSCLC has dramatically changed in recent years, with increasing incorporation of targeted treatments and immunotherapies across all stages and histologic types. Considering ASCs are composed of glandular cell components, they can contain substantial levels of relevant actionable driver mutations as described above. Therefore, if ASC is diagnosed or if a SCC has a glandular component,molecular testing is recommended and supported by guidelines, even on surgical specimens where EGFR may be targeted as adjuvant treatment.²³ However, while targeting actionable mutations and the PD1/PDL1 axis has been studied extensively in AC and SCC in all stages, the impact of these markers in ASC is unknown because patients with this histologic subtype are frequently excluded from clinical trials.

For patients with ASC and actionable mutations, EGFR inhibitors have been perhaps the best-studied targeted therapies. EGFR inhibitors have yielded responses in ASC, but the benefit has been highly variable in small case series and generally inferior to outcomes in patients with AC alone.¹⁹ Ongoing clinical trials are aiming to better understand the effects of EGFR inhibitors in ASC. As one example, first-line almonertinib is being compared to paclitaxel/carboplatin in the phase 2 ARISE clinical trial, which is specifically enrolling patients with EGFR mutation-positive locally advanced or metastatic pulmonary ASC (Clinicaltrials.gov NCT04354961). Most other reported studies are case studies or retrospective in nature.

Given that outcomes are usually reported from single patients or a group of only a few patients, contradictory findings are not uncommon. For example, crizotinib, a multi-kinase inhibitor approved for the treatment of advanced or metastatic ALK-positive and ROS1-positive NSCLC, was reported to have a clinical response in an ASC in a patient with recurrent ALK-positive disease which lasted for just over one year.²⁴ However, the response to second-line crizotinib in a case report of female non-smoking patient with ROS1-positive ASC was only 4 months.²⁵ Newer, more specific kinase inhibitors are currently in clinical practice and trials of ALK and ROS1 NSCLCs; however, their efficacy is ASC remains unclear.

In the absence of driver mutations, the optimal choice of chemotherapy (often given with immunotherapy) for neo-/adjuvant therapy or for metastatic disease has not yet been identified. While the AC component might typically be treated with pemetrexed plus a platinum agent, the SCC component may be better treated with taxane plus a platinum agent.²³ Especially in cases where neither histologic subtype is predominant, it can be difficult to decide which combination may be suitable for an individual patient. Whether the relative proportion of AC and SCC components affect treatment outcomes is not yet known. Outcomes of pemetrexed-based chemotherapy have been reported in a case study of 2 patients with relapsed disease harboring ALK and ROS1 mutations, pemetrexed alone or as part of a combination regimen (with pembrolizumab and carboplatin) was able to maintain stable disease for at least a year.²⁶

While immune checkpoint inhibitors, either as monotherapy or in combination with chemotherapy, are currently recommended for patients with NSCLC²³, few studies have reported outcomes of patients with ASC specifically. One recent real-world analysis by Li et al. evaluated the effect of immunotherapy in 46 patients with ASC, of which 18 (39%) did not contain actionable driver mutations and 18 (39%) had unknown mutational status.²⁷ In this study, 28% of the overall cohort responded to checkpoint inhibitors, the median progression-free survival was 6 months, and the median OS was 24.7 months. Notably, similar efficacy was observed in the 20 patients receiving immunotherapy monotherapy vs 26 patients who received combination immunotherapy plus chemotherapy.²⁷ Among 4 patients with EGFR mutations, 2 received immunotherapy monotherapy and progressed immediately compared to the other 2 receiving combination immunotherapy with chemotherapy achieving disease control and improved OS (18 months).

As exemplified by the select few cases summarized above, conventional treatments used in NSCLC have achieved only modest responses in ASC, most with a shorter response duration. The lack of specific treatment strategies for ASC, based on our understanding of underlying tumor biology, limits optimal treatment outcomes for this increasingly common diagnosis. Novel therapies are sorely needed. A consensus should be developed to either study novel treatments specifically in this subtype or allow for the incorporation of ASCs into future NSCLC clinical trials.

Click to read more from 2023 Rare Diseases Report: Cancers