Unusual presentation of incidental atypical carcinoid tumor masquerading as carcinoid tumorlets arising from diffuse neuroendocrine cell hyperplasia (DIPNECH): A case report

Well-differentiated neuroendocrine tumors (WDNTs) of the lung include carcinoid tumorlets, typical carcinoid tumor (TC), and atypical carcinoid tumor (AC). Atypical carcinoids represent <1% of lung tumors in SEER database and 10–15% of carcinoid lung neoplasms [1]. Atypical carcinoids are distinguished from TC by presence of elevated mitotic rate (2–10 mitoses per 2 mm²) and or tumor necrosis [2],[3]. They behave more aggressively than TCs with increased tendency to metastasize and to recur. Carcinoid tumorlets are differentiated from TCs with a cutoff size of <5 mm, and are considered benign, but can progress to TCs in a small minority.

Well-differentiated neuroendocrine tumors are frequently encountered incidentally. They originate from neuroendocrine cells (bronchial Kulchitsky cells) present normally in respiratory epithelium [4]. Neuroendocrine cell hyperplasia (NEH) is commonly observed in proximity to WDNTs and the former is thought be a very early precursor lesion. A subset of WDNTs arises from diffuse neuroendocrine cell hyperplasia (DIPNECH) [5],[6]. Diffuse neuroendocrine cell hyperplasia is a benign and incidental finding, presenting as single, micronodular, or linear clusters of neuroendocrine cells in bronchial or bronchiolar epithelium. The etiology of DIPNECH and development of WDNT are not well understood; no risk factors are recognized, but chronic lung injury (e.g., smoking) is theorized to be a risk factor [3].

When neuroendocrine cells in DIPNECH grow through the basement membrane, carcinoid tumorlets and TCs can develop [7],[8]. However, most cases of DIPNECH do not progress beyond carcinoid tumorlets and are not visible macroscopically. The development of TCs and ACs is associated with worse prognosis: 10-year survival of 87% vs. 35% respectively for TC and AC [9]. An accurate diagnosis is, theore, important for prognostic and oncologic management.

We present the case of a 78-year-old female with a 16-pack year smoking history. She was found to have a 2.9 cm spiculated right upper lung nodule (Figure 1), concerning for malignancy, and multiple smaller ground-glass nodules (<7 mm) as part of our lung cancer screening program. Peri-operative bronchoscopy was negative for endobronchial mass. The patient underwent a lobectomy with regional lymph node sampling. Gross examination revealed a 1.2 cm spongy, ill-defined mass (Figure 2), and a smaller satellite nodule of similar histology (not shown). Regional lymph nodes were negative for tumor.

Final pathology showed two foci of invasive lepidic adenocarcinoma (pT1a(m)) with an invasive tumor size of 0.6 cm, histologically corresponding to the gross spongy mass and satellite nodule. Adjacent lung tissue sections showed 12 microscopic tumor nodules that resembled carcinoid tumorlets—all foci were <5 mm (range 0.6–2.0 mm). These showed neuroendocrine cytology (“salt and pepper” chromatin), appeared randomly distributed in the uninvolved lung, and were not grossly visible.

These small neuroendocrine proliferations had unusual histologic features not typically associated with carcinoid tumorlets: tumor abutting pleura (one focus) (Figure 3A), focal pleomorphism (Figure 3B), and nuclear pseudoinclusions (Figure 3C). Closer observation revealed characteristics incompatible/atypical for a diagnosis of carcinoid tumorlets: visceral pleural invasion (Figure 3D), punctate necrosis (Figure 3B), and elevated mitotic activity with atypical mitosis (Figure 4A). KI-67 was elevated (12%, hot-spot method) in the most mitotically active areas (Figure 4B). The tumor stained positive for synaptophysin, chromogranin, and TTF1 (Figure 4C), and showed focal lymphatic invasion (Figure 4D).

In total, six nodules (size range: 1.2–2.0 mm) met the WHO criteria for atypical carcinoid tumor based on mitotic activity (5/6) (mitotic activity range: 2–8 mitoses per 2 mm²) and/or tumor necrosis (5/6). Six nodules did not meet WHO criteria for TC or AC, but could not be ruled out due to limited tumor size for evaluation of mitotic activity (size range: 0.3–1.4 mm), and contained 1–4 HPF. One microscopic nodule abutted the visceral pleura, and an elastic stain confirmed the presence of visceral pleural invasion. The tumor necrosis was subtlety observed as punctate necrosis, and was present in five microscopic tumor nodules. Focal cellular pleomorphism was identified in 8 of 12 often with focal intranuclear inclusions.

Diffuse neuroendocrine cell hyperplasia was present in multiple lung sections (Figure 5A) and was highlighted by chromogranin stain (Figure 5B). In one tumor focus, there appeared to be a transition from bronchial epithelium with linear hyperplasia to micronodular hyperplasia to carcinoid tumor (Figure 5C), highlighted by chromogranin staining (Figure 5D). The AC nodule seemingly arose directly from DIPNECH involving invaginated bronchial mucosa.

Carcinoid tumorlets and NEH are usually discovered incidentally, and are considered reactive lesion(s) secondary to chronic lung injury. However, NEH and DIPNECH have the potential to progress to carcinoid tumors.

Our case demonstrates the preneoplastic nature of DIPNECH through an unusual presentation as multiple, microscopic tumor foci of AC. This case is more unusual because it was discovered incidentally on microscopic evaluation of background lung adjacent to a radiologically identified lepidic adenocarcinoma.

Although AC has been described in association with tumorlets and NEH [10],[11], after a thorough PubMed literature search, to our knowledge, we believe our case is the first documented report of “tumorlets” with AC morphological characteristics. Tumorlets are most commonly encountered as incidental discoveries in lung resections for carcinoid tumors, in up to 24% of cases when background lung with micronodules are well-sampled [12]. Like DIPNECH, tumorlets are more frequent in patients with chronic smoking-related lung disease, interstitial lung disease, bronchiectasis, and pulmonary vascular disease, and have been discovered in 1% of lung explants and autopsy [13],[14]. Rare metastasis of carcinoid tumorlets to regional lymph nodes have been reported, but elevated mitotic activity, necrosis, or atypia were not observed, and patients were observed to have excellent prognosis despite local nodal metastasis [15],[16].

The AC masqueraded as “tumorlets” (<5 mm), discovered incidentally, not radiologically apparent, and displayed characteristic well-differentiated neuroendocrine cytological features. Despite their small size, the micronodules had elevated mitotic activity and/or necrosis, and as well as more aggressive tumor features including lymphovascular invasion and visceral pleural invasion.

The AC micronodules likely developed synchronously from DIPNECH precursor lesion(s) and/or spread lymphatically along bronchovascular bundles. It is unlikely the ACs were gastrointestinal (GI) metastases as the tumor was TTF1+, and gastrointestinal (GI) WDNTs do not stain positive for TTF1+ [17],[18]. There was no positron emission tomography-computed tomography (PET-CT) or endoscopic evidence of metastases from a different lung lobe or a primary from central airways. The AC appeared to arise from the progression of NEH. One focus of DIPNECH was in continuity with the largest focus of AC (2.3 mm). Invaginated bronchial epithelium was heavily involved by linear and micronodular NEH and the latter appeared to grow into a carcinoid tumor. This observation of the biological progression of DIPNECH to NEH to carcinoid tumor supports the notion that DIPNECH, in a minority of cases, can be preneoplastic lesion to carcinoid tumors.

There were subtle clues for accurate tumor classification. While most tumor foci did not have 8–9 HPF for mitotic count corresponding to a 2 mm² area, the mitotic criteria for AC was supported by finding 1–2 mitoses in 2–4 HPF (limited tumor area), corresponding to a calculated elevated 2–10 mitoses per 2 mm². The observed elevated mitotic activity on H&E was also supported by elevated KI-67 index. All the neuroendocrine tumor micronodules had KI-67 > 1%, up to 5% (hot-spot). While KI-67 is not a criterion for ACs, it is commonly found to be elevated in ACs (>3%) versus TCs (<3%) [19],[20] . Another subtle histologic clue is the presence of focal cellular pleomorphism, which is more prevalent in ACs than TCs or carcinoid tumorlets [3]. One micronodule closely abutted the pleura subtlety showed pleural invasion with an elastic stain. Carcinoid tumorlets lack potential for visceral pleural invasion. Necrosis was very subtle and occurred only as punctate (single cell) necrosis, which can be easily overlooked.

The rare possibility of AC in “tumorlet”-appearing multiple micronodules should be considered in presence of elevated mitotic activity and necrosis in background of DIPNECH. Additional clues can include presence of cellular pleomorphism and tumor closely abutting pleura. Diffuse neuroendocrine cell hyperplasia can be a precursor lesion to carcinoid tumors, including AC, in a minority of cases.