Bronchogenic Carcinoma Text and photomicrographs. Nice. Human Pathology Digital Image Gallery

Lung Cancer From Chile In Spanish

Lung Cancer Images University of Washington Pictures and comments

Future pathologists: See Arch. Path. Lab. Med. 133: 1106, 2009!

The vast majority of primary malignant neoplasms of the lungs are some form of BRONCHOGENIC CARCINOMA. Easy review Lancet 355: 479, 2000.

In 2005, the US had around 184,800 new cases of bronchogenic carcinoma (102,420 men, 82,380 women) -- CA, in loc. There were around 168,140 deaths. Worldwide, supposedly there are about a million reported deaths (NEJM 356: 830, 2007), obviously too low in a world where Russia, China, and many of the poor nations have much higher rates of smoking than the US.

Bronchogenic carcinoma is the leading cancer killer of US men, although age-adjusted rates have been dropping for men since 1980 (NEJM 321: 1197, 1989), and the total absolute number is now decreasing; the total percent of people dying of lung cancer is way down.

It surpassed breast cancer as the leading cancer killer of women in 1985. The rates in women have just started to drop (2004).

In 2009, there were around 219,000 new cases in the US, and 159,000 deaths. This much-improved cure rate (almost everybody used to die) results from earlier detection.

Still, the majority ung cancers are silent until they've become inoperable. Those lucky patients whose cancers are of operable subtype and are detected very early have a reasonably good prognosis (maybe 60% long survival with stage I, 40% stage II: Ann. Thorac. Surg. 60: 466, 1995).

Bronchogenic carcinoma is a disease of older people, and is unusual under age 30. The incidence increases with age and "pack-years smoked".

Lung cancer exhibits Nowell's law in action, with mutations in the surrounding "normal" epithelium. Prognosticating lung cancer using genetic markers has been underway since the mid-1990's, and centers now look at around 100 genes (Am. J. Resp. CCM. 170: 167, 2004); clinical decisions are not yet based on the genetic profile, but probably will be within a decade.

Risk factors are well-established (Chest 103-S1: 20-S, 1993) CIGARET AND CIGAR SMOKING: This is still the overriding risk factor (so much so that never-smoker's cancer seems to be a separate entity).

The danger is proportional to the amount of smoke inhaled every day and the duration of smoking, and is measured in "pack-years." After ten pack-years the increased risk is measurable, and after fifteen pack-years, there is clear-and-present danger. After quitting, the risk decreases slowly and approaches normal after ten or fifteen years. (Cigar smokers inhale less smoke overall than cigaret smokers, but contrary to what you may hear, their risk for lung cancer is still greatly increased.)

"Passive smoking" seems to be of some importance, but much less than active smoking (Am. J. Pub. Health 82: 1525, 1992). Pipe smoking and tobacco chewing are much less risky than cigaret smoking, at least as far as the lungs are concerned.

The other risk factors are synergistic (i.e., more than just additive) with cigaret smoking.

* By the way, former smokers filling out research forms will often say they never smoked. Remember this when you're evaluating research articles (Chest 101: 19, 1992). In particular, every single article on "the relative risk of this-or-that controlling for smoking" is severely tainted for this reason alone. More about this when we come to the alleged terrible dangers of eating red meat.

RADIATION: uranium miners (powerful effect, all histologies Cancer 89: 2613, 2000), atomic bomb survivors, after radiation therapy (mild effect; Cancer 71: 3054, 1993), radon in homes ("if you say so....": NEJM 330: 159, 1994.)

* During the 1980's and early 1990's, there was a hoopla over radon from the ground building up in our energy-efficient homes. The claim, advanced by the Environmental Protection Agency, was that it caused 5-20 thousand lung cancers yearly, multiplicative with tobacco smoking. Was it just junk science? I wasn't the only one to say so at the time... Lancet 337: 1329, 1991; Arch. Int. Med. 151: 674, 1991 (AMA council report). Since "radon-proofing your home" was a mega-bucks business, relating to property values and fear of government regulation, someone more cynical than myself might have seen an even baser motivation.

* It does seem possible that smokers are at increased risk if there's radon in their homes. Most of the epidemiologic studies have come back negative, one controversial study found a negative correlation (as if radon protected you: Health Phys. 72: 623, 1997), the last big paper to find a link was Am. J. Pub. Health 89: 1042, 1999 and even the authors admit it's really weak, and even Br. J. Cancer 84: 134, 2001, which did a study of radon and lung-cancer in nonsmoking men, couldn't quite get a statistically significant correlation. There seems to be agreement that non-smokers aren't at serious risk, and that smokers can protect themselves much less expensively by stopping smoking (Am. J. Pub. Health 88: 811, 1998). The collaborative analysis in BMJ 330: 223, 2005, including old and new studies only found a modest risk and then only for smokers or recent ex-smokers.

As you would expect, the nucleic acid damage from radiation is more random than in tobacco smokers. See Lancet 339: 576, 1992.

ASBESTOS INHALATION: Notorious cause of lung cancer. According to some accounts, more than half of asbestos workers who smoke have died of lung cancer (more: Br. Med. J. 306: 1503, 1993).

"The Helsinki Criteria", supposedly a consensus document for relating a lung cancer to previous asbestos exposure, reviewed and criticized (as lawyer-advocacy): Arch. Path. Lab. Med. 131: 181, 2007.

CERTAIN METAL DUSTS (supposedly nickel -- the major work is from the late 1970's and is small-sample statistics -- and maybe silver)

HEXAVALENT CHROMIUM (the most familiar form in industry; the best study is J. Occ. Env. Med. 48: 426, 2006 which found some increased risk in the more heavily exposed German chromate workers when we control for smoking. The authors believed there was a threshold effect below which there was no measurable risk, which makes the activist-driven OSHA crackdown in 2007 senseles)

COAL TAR FUMES, carbon black (Lancet 358: 562, 2001) and other chemicals (* notably vinyl chloride, chloromethyl ethers) Occupational lung cancer: Mayo Clin. Proc. 68: 183. 1993.

URBAN POLLUTION: a minor risk factor compared with the others, with which it may be synergistic (JAMA 287: 1132, 2002)

* In one study that generated a hoopla, the investigators failed to take into account the fact that people living near the steel mill smoked much, much more than did their rural counterparts. When somebody noticed this, the "cancer risk" from the steel mill disappeared (Arch. Env. Health 48: 184, 1993).

In the poor nations, living in a coal-heated dwelling that's not ventilated clearly causes a great deal of lung cancer (Lancet 362: 849, 2003; JNCI 94: 826, 2002).

* I think that silica / sand exposure, made "official" in 1997 by the International Agency for Research on Cancer, is a crock. The numbers are very soft and there's no credible mechanism. See Am. J. Epidemiol. 153: 695, 2001.

* PIGEON KEEPING: Possible promoter, but not a mega-risk factor (Br. Med. J. 305: 986 & 989, 1992). A 1992 flap about keeping a canary or parakeet being a risk factor for lung cancer was laid to rest by a bunch of negative studies (Br. Med. J. 313: 1218, 1996).

* For a while there was a hoopla about carotenoids (there are at least five kinds) as protecting people (smokers, non-smokers) from lung cancer. Given that a high level of blood carotenoids strongly suggests an overall healthy-lifestyle, and that today's "pop wisdom" is that vegetables protect from cancer, it's hard to do reliable studies. Protection by dietary carotenoids: Am. J. Clin. Nutr. 72: 990, 2000; Am. J. Ep. 156: 536, 2002 (about a whopping 15%, and based on self-reports); no protection Canc. Causes Contr. 13: 231, 2002; more confusion Br. J. Cancer 84: 728, 2001 (something's going on but it can't all be the tomato sauce).

Familial cancer syndromes (i.e., antioncogene deletion syndromes) causing lung cancer in non-smokers are conspicuous by their absence. Li-Fraumeni and retinoblastoma-family syndromes give increased risk if you smoke. A susceptibility gene (point-mutation) discovered in 2008 turned out to be (surprise!) the nicotine receptor (Nature 452: 633, 2008; Nat. Genet. 40: 616, 2008 -- and the mechanism is ... you smoke more cigarets! Nature 452: 638, 2008).

NOTE: We know who is at risk. The traditional wisdom is that screening (routine x-rays, exfoliative cytology) is a waste of time and money (Ann. Int. Med. 111: 232 & 239, 1989). With today's improved low-radiation scanning, we may be putting all smokers into the machine yearly: Cancer 89(S-11): 2474, 2000, curing 60-80% of the early cancers we find.

Classification scheme:

The classification was updated in 1999, but most bronchogenic carcinomas still get assigned to one of the four light-microscopic categories of the old WHO classification (1967):

LUNG CANCER DIAGNOSIS FOR BEGINNERS (find one....)

Squamous Cell...keratin, squamous pearls, "bridges" (desmosomes); EM: tonofilaments, diffuse keratin; IP: high MW keratin

* Subtypes ("glycogen-loaded sugar-tumor", "adenosquamous with a little bit of mucin production". "polypoid sticking-up rather than invading much") exist but don't matter much.

* Unlike most other lung cancers, TTF-1 is usually negative

Adenocarcinoma... glands, papillae, mucin; EM: microvilli, secretory granules, gland formation; IP: low MW keratin, CEA; TTF-1 (the latter is especially helpful as an indicator that an adenocarcinoma metastasis of unknown origin came from lung)

The WHO 1999 classification rightly places bronchioloalveolar cell carcinoma in its own category.

* Other adenocarcinoma subtypes exist and except as noted below are of no known significance.

* For the more obscure histologic types of adenocarcinoma, see Arch. Path. Lab. Med. 130: 958, 2006.

Large Cell Undifferentiated... none of the above, plenty of cytoplasm; the World Health Organization has added subtypes that need not concern us

Small Cell Undifferentiated... "small blue cells" (i.e., very little cytoplasm to stain pink); EM: "neurosecretory" (dense-core, APUD) granules, fine chromatin, little cytoplasm; IP: neuron-specific enolase, bombesin, TTF-1 (Mod. Path. 19: 1117, 2006)

These categories (especially the first three--"squamous", "adeno", and "large cell undifferentiated") are difficult to distinguish unless the tumor is very well-differentiated. (* The new WHO system ignores tiny areas with "squamous" or "adeno" features taking up less than 10% of a resected tumor; so the number of "large cell undifferentiated" tumors will increase). The treatment (surgery, palliative radiation, maybe some chemotherapy) and prognosis (bad) are the same anyway.

"Subtypes exist, of course" but making the distinctions is not all that useful. Reported relative frequencies vary widely. For this course, remember that bronchogenic carcinomas of all four types are about equally common. Adenocarcinoma the most common if you are asked on somebody's exam (won't be mine); especially, these are the little ones picked up on imaging and that are getting cured surgically. Squamous cell carcinoma used to be easily the commonest, now it's adenocarcinoma. (I suspect that if these hadn't been picked up early, they would have presented with more mutations as "large-cell undifferentiated.") The distinctive small-cell variant makes up maybe 15-20%.

The trend still seems to be "men get more squamous cells, women get more adenocarcinomas" (Cancer 69: 86, 1992; Cancer 103: 2566, 2005; we very macho men cannot inhale our non-filtered cigs as deeply as the ladies inhale their filtered cigs).

* Or maybe filtered cigs produce smaller particles that get farther in the lungs. Cancer 77: 1278, 1996.

EM, immunohistochemistry, and genomics have further complicated things.

Nowadays, a lung mass is likely to be diagnosed by the pathologist performing fine-needle aspiration (AJRCCM 174: 684, 2006 -- a landmark study that shows that this does not cause tumor spread, as had been feared).

Clinicians consider it most important that the pathologist distinguish small cell undifferentiated carcinoma ("small cell" carcinoma) from the others.

"Small cell" carcinoma is not a surgical disease, and usually had a nice initial response to chemotherapy. See below.

The other types are treated surgically if feasible, radiated ("spot-welded") as needed, and occasionally respond well to chemotherapy. My own favorite article from 1998 was BMJ 317: 771, 1998: People who had been given chemotherapy for advanced non-oat-cell carcinoma of the lung would NOT have accepted it if they'd been told the truth beforehand. And a comparison of four expensive, horrific protocols that showed that none gave better than 20% transient responses ended with the frank admission that it's not right to do this (NEJM 346: 92, 2002).

During the first few years of the 21st century, treatment outcomes became a bit better using platinum-based chemotherapy ("we''ve reached a plateau": Chest 136: 1112, 2009), and the search is on for histologic and molecular markers for which cancer will respond best to what "targeted therapy agent". Adjuvant chemotherapy after surgery does seem to help some (NEJM 350: 353, 2004). The introduction of targeted therapy using the new biotech agents (notably the VEGF inhibitor bevacizumab) combined with the older (largely platinum-based) protocols seem to give a longer, better-quality life. Update Chest 130: 1211, 2006. Also keep watching the epidermal growth factor receptor antagonists (Chest 128: 3975, 2005). More biotech products are in the pipeline.

* Managed-care comes to the pathology lab. Appropriate handling of lung cancer specimens: Arch. Path. Lab. Med. 119: 695, 1995.

* My old friend Barrie R. Callileth and the rest of the integrative medicine team share an no-BS update on complementary therapies for the lung cancer patient, distinguishing the cynical frauds that pretend to affect the natural course of the disease to a host of treatments that can probably make people happier and more comfortable (Chest 132(S-3): 340-S, 2007).

* It is likely that by the time you are practicing, genetic analysis of individual lung cancers (probably all except the oat-cells) will really govern therapeutic decisions. Mutations that suggest the medications will work: Science 305: 1163, 2004. The first clinical steps: NEJM 355: 570, 2006. This generated much excitement, though it turns out that the results of the $1000 genetic test for the molecular status of the EGFR receptor don't correlate that well with outcome (results vary; Arch. Path. Lab. Med. 132: 1573, 2008). The various series-of-mutations "pathways" in a series of mostly-adenocarcinomas (JAMA 303: 535, 2010).

The most robust finding is the T790M mutation as a sign of a good response to tyrosine kinase inhibitors. The group also watched their develop resistance to "Iressa" and "Tarceva" by monitoring EGFR mutations in the circulating blood (NEJM 359: 366, 2008).

Of course, grade still means the tumor will be more aggressive. The trouble is that no one knows exactly how to assign grade, or incorporate it into current treatment protocols (J. Thor. Card. Surg. 131: 1014, 2006).

Just to confuse things further, the "T"-stage (and prognosis, it seems) depends to a great extent on whether the tumor is invading through the visceral pleura, and there's no consensus among pathologists how to determine this; requiring clear penetration of the visceral elastic layer seems reasonable. (Am. J. Clin. Path. 128: 638, 2007).

Lung cancer Large hilar tumor, emphysematous lung KU Collection

SQUAMOUS CELL CARCINOMA ("Epidermoid carcinoma")

Squamous Lung Cancer From Chile In Spanish

Squamous Cell Lung Cancer Australian Pathology Museum High-tech gross photos

Squamous cell carcinoma Lung pathology series Dr. Warnock's Collection

Well-differentiated squamous lung cancer AFIP Wikimedia Commons

Squamous cell carcinoma Big central lesion WebPath

Squamous cell carcinoma Big central lesion WebPath

Squamous cell carcinoma Big central lesion WebPath

Squamous cell carcinoma WebPath

Squamous cell carcinoma Nice desmosome prickles WebPath

Squamous cell carcinoma Nice squamous pearl WebPath

Squamous Lung Cancer Text and photomicrographs. Nice. Human Pathology Digital Image Gallery

Squamous lung cancer AFIP Wikimedia Commons

{17520} squamous cell carcinoma of lung, large mass at hilum
{20994} squamous cell carcinoma of lung, find the cavity
{20996} squamous cell carcinoma of lung, gross
{17525} squamous cell carcinoma of lung, poorly differentiated
{15401} tonofilaments
{15402} desmosome

Tobacco smoking is the major risk factor (90-99%, estimates vary).

Tumors arise anywhere in the bronchi, often near hilum. Large squamous cell carcinomas tend to cavitate. Often hemorrhage into the cavity and out the mouth is the final event (exsanguination, drowning in blood).

Generally the pathologist can find squamous metaplasia and dysplasia (and even carcinoma in situ) in the bronchi surrounding a squamous carcinoma. ("Nowell's law" triumphant.)

Squamous cell carcinomas very commonly produce a parathyroid-hormone-like substance that may cause elevated serum calcium even in the absence of bone metastases -- "humoral hypercalcemia of malignancy". (Production of other ectopic hormones is unusual.)

The protein is PTHrP, parathormone-related peptide, required for proper development of teeth and bones (Proc. Nat. Acad. Sci. 95: 11846, 1998).

The epidermal growth factor receptor, coded by the erbB proto-oncogene, is much over-expressed in this particular cancer (Nature 307: 521, 1984), but adenocarcinomas and large-cell-undifferentiated carcinomas will also stain some for this. Oat-cell is usually negative, and this is another helpful distinction.

The last few years have been dominated by the EGFR tyrosine kinase inhibitors. Gefitinib ("Irissa"), a pill that inhibits epidermal growth factor receptor, for chemotherapy-resistant non-oat-cell lung cancers that overexpress the receptor and/or bear a mutant receptor: JAMA 290: 2149, 2003; Science 304: 1497, 2004; NEJM 350: 2129, 2004. A small minority of patients have had dramatic responses, but these are only temporary. Eventually, a further mutation of the gene results in resistance (NEJM 352: 786, 2005). Update (it's pretty good if and only if EGFR is mutated): NEJM 362: 2380, 2010.

Erlotinib ("Tarceva") has replaced gefitinib in the US; unlike its predecessor, it does seem to prolong survival and perhaps quality of life. Again, resistance develops in a few months.

Overall, about 8% of patients with advanced lung cancer respond, with survival improvement averaging two months.

The most promising study, which will probably set a care standard if the test becomes cheap enough, found that patients with mutated EGRF (most often women who never smoked and had adenocarcinomas) got quite good temporary results with erlotinib (NEJM 361: 958, 2009).

* The pathology team at Harvard confirms that an activating mutation, rather than something easier to test for, is how to predict a response: Am. J. Clin. Path. 133: 922, 2010.

The lung cancers that seem to respond best are well-differentiated adenocarcnomas, especially those with bronchioloalveolar differentiation (Arch. Path. Lab. Med. 133: 470, 2009).

If the tumor is found when small, and is well-differentiated, the surgical cure rate is reasonably good. Squamous differentiation still remains an independent "good" prognositic sign in a bad disease: Ann. Thor. Surg. 77: 1173, 2004.

* Spontaneous regression of this dread cancer is very rare but happens: Chest 94: 701, 1988.

Dysplasia in bronchial epithelium Surprise finding at autopsy KCUMB Team

ADENOCARCINOMA

Adenocarcinoma From Chile In Spanish

Lung Adenocarcinoma Text and photomicrographs. Nice. Human Pathology Digital Image Gallery

Adenocarcinoma of the lung was once thought not to be related to tobacco smoking (Acta Path. Microbiol. Scand. S-157: 1, 1962.) Nowadays 75-90% of these patients in recent studies have been smokers.

Non-smokers with stage I adenocarcinoma of the lung have a much better prognosis Ann. Thorac. Surg. 81: 1193, 2006).

Given similar risk factors, women supposedly get more adenocarcinomas and men get more of the other types.

Most adenocarcinomas arise in the periphery beneath or near the pleura.

Adenocarcinoma Peripheral cancer WebPath

Poorly-differentiated lung adenocarcinoma AFIP Wikimedia Commons

Primary lung adenocarcinoma AFIP Wikimedia Commons

Lung adenocarcinoma, preinvasive H&E Wikimedia Commons

{20997} peripheral adenocarcinoma

Adenocarcinomas have been said to arise in scars in the lung, including old TB, old infarcts, and physical wounds. These are called "Yokoo tumors", and Dr. Yokoo taught me in residency. I used to think that the cancer produces the fibrosis, but I've seen several adenocarcinomas at autopsy in non-smokers and these have always been in the setting of scars of known etiology.

The histology correlates with the prognosis for small tumors (Cancer 61: 2083, 1988), while large tumors are very lethal.

Surfactant apoprotein stain is now available and should be handy in telling this tumor from metastatic disease. Other special stains: Am. J. Clin. Path. 92: 150, 1990.

Oddly, about 20% of lung adenocarcinomas arise multicentrically (Chest 95: 151, 1989).

* Molecular biologists again: Mutational activation of the K-ras proto-oncogene (at "hot spot" 12) is involved in many smoking- and occasional non-smoking-related adenocarcinomas, in contrast to other lung cancers.

About 10% (figures vary) of adenocarcinomas are of the BRONCHIOLO-ALVEOLAR CARCINOMA subtype, considered to include cancers of the bronchial goblet cells ("mucin-positive") as well as type II pneumocytes and Clara cells ("mucin-negative"; includes the "sclerosing" sub-subtype).

These malignant cells grow along the alveolar septal framework and (if "mucin positive") secrete mucus into the alveoli, producing "consolidation." This cancer does not invade, but kills by growing as a single layer of thick cells covering over the respiratory membranes.

This type of cancer is getting to be more common, or at least more recognized (Cancer 68: 1973, 1991). WHO-1999 separates it from other adenocarcinomas. Before 1980 it was thought to be non-cigaret-related; newer studies show a relationship (but less than any of the four "major" types.

* Pathologists distinguish primary bronchioloalveolar cell carcinomas from metastatic colon cancer (which can look identical) because only the latter stains for the villin antigen; this is not 100%. New technique: Cdx2 stain usually does not stain bronchiolo-alveolar carcinomas, but usually does stain colon metastases (Am. J. Clin. Path. 122: 421, 2004).

Beyond this, bronchioloalveolar carciomas are a genetically and immunologically heterogeneous lot: Arch. Path. Lab. Med. 128: 406, 2004.

The mucin-producers tend to have mutant K-ras, while the non-mucin-producers ("lung adenocarcinoma in situ") tend to have mutant EGFR (and thus response to EGFR tyrosine kinase inhibitors): Am. J. Clin. Path. 128: 100, 2007. The same study notes that the non-mucin-producers do occur with some frequency in never-smokers. There are similar results from Arch. Path. Lab. Med. 131: 1027, 2007.

Since a "pure" bronchioloalveolar carcinoma shows no evidence of stromal invasion, it's best considered "carcinoma in situ". Today, this is the common lung cancer "found on a screening CT scan and cured surgically."

* Future pathologists: the term for this pattern of growth along a surface without invasion is "lepidic".

Bronchioloalveolar carcinoma Freshly-opened at autopsy WebPath

Bronchioloalveolar carcinoma WebPath

Atypical adenomatous hyperplasia Forme fruste of bronchioloalveolar cell carcinoma?

Bronchioloalveolar cell carcinoma, great case Lung pathology series; follow the arrows Dr. Warnock's Collection

Bronchioloalveolar cell carcinoma Various cell types can give rise to it Dr. Warnock's Collection

Alveolar Cell Carcinoma Text and photomicrographs. Nice. Human Pathology Digital Image Gallery

Bronchioloalveolar carcinoma AFIP Wikimedia Commons

Bronchioloalveoliar carcinoma Mucus is red Wikimedia Commons

Bronchiolo-Alveolar Carcinoma High magnification KU Collection

{11454} bronchiolo-alveolar carcinoma; very small cells growing along the alveolar septa
{12608} bronchiolo-alveolar carcinoma; the lung is solidified by mucus in the alveoli

The tumor is slow-growing but very lethal unless it's a small resectable lesion (Ann. Thor. Surg. 71: 971, 2001). A few may be benign, but no one knows how to recognize these.

* Jaagziekte, an infectious disease of South African & other sheep, resembles this tumor histologically and clinically, but there is no good evidence of transmissibility in humans.

* Cigaret-smoking beagles, an animal model for human lung cancer, get mostly bronchioloalveolar carcinoma, and so do scleroderma patients (Arch. Pathol. Lab. Med. 108: 7, 1984.)

* Future pathologists only: Here's a popular system for subdividing the small (2 cm or less) adenocarcinomas that you will get as surgical specimens (Cancer 75 2844, 1995):
• Type A: Localized bronchioloalveolar carcinoma (claras, type II's, or goblet cells), no or minimal scarring, enough atypia to qualify as cancer; expect surgery to be curative


• Type B: As type A, but with foci of alveolar structural collapse / scar contracting; still expect surgery to be curative


• Type C: As type B, but with active fibroblasts; surgery cures about 75%


• Type D: Poorly-differentiated adenocarcinoma, but with a sharp border; about 50% surgical cures


• Type E: Tubular adenocarcinoma (fron the bronchial glands; looks like salivary gland cancer), acini, tubules, and/or back-to-back glands, sharp border


• Type F: Papillary, expansile growth; not enough cases to prognosticate


• Not placed: Micropapillary adenocarcinoma, an aggressive tumor (Am. J. Clin. Path. 131: 694, 2009)


By the time you are in practice, NEVER-SMOKING NON-SMALL CELL LUNG CANCER is likely to be a separate entity (Cancer 113: 1012, 2008). These are mostly little adenocarcinomas with mutated EGFR, normal k-ras, good response to EGFR-tyrosine-kinease inhbitors, and the patients never smoked. As smoking decreases and we pick up more early cancers, these now comprise up to 30% in some series (especially Asia -- this couldn't be industrial exposure, could it?)

LARGE CELL UNDIFFERENTIATED CARCINOMA ("hundred-day cancer")

Oat Cell / Large Cell Anaplastic From Chile In Spanish

Large cell and adenocarcinomas Lung pathology series Dr. Warnock's Collection

Large-cell lung cancer "Hundred-day cancer" KCUMB Team

Large cell lung cancer AFIP Wikimedia Commons

Large cell undifferentiated lung cancer AFIP Wikimedia Commons

Most are related to tobacco smoking (90-99%, estimates vary.) These tumors arise anywhere in lungs. Cures are rare, and death comes quickly.

Obviously these tumors are mostly occupied with growth and invasion rather than looking like they did in health. There are no useful tumor markers (hPTH-like substance may be produced) or other clinical behaviors that really show this is a distinct entity.

SMALL-CELL UNDIFFERENTIATED CARCINOMA ("OAT CELL") (Lancet 345: 1285, 1995; treating it Chest 112(4S): 251-S, 1997.)

Oat cell carcinoma Looks like sushi WebPath

Oat cell carcinoma Follows the bronchi WebPath

Oat cell carcinoma WebPath

Oat cell CA with Azzopardi phenomenon Lung pathology series; follow the arrows Dr. Warnock's Collection

Oat cell carcinoma Tom Demark's Site

Oat cell carcinoma KU Collection

Oat Cell Carcinoma Text and photomicrographs. Nice. Human Pathology Digital Image Gallery

Small cell lung cancer AFIP Wikimedia Commons

Oat cell carcinoma Lots of necrosis Wikimedia Commons

Oat cell carcinoma High magnification Wikimedia Commons

Oat cell carcinoma
These actually look like oats
Wikimedia Commons

{11428} oat cell carcinoma, spreading along the bronchi
{10475} oat cell carcinoma, growing down between the cartilage rings
{10435} oat cell carcinoma
{11693} oat cell carcinoma
{12578} oat cell carcinoma
{14076} oat cells, spinal fluid
{12581} oat cells in sputum
{38521} oat cell
{38527} oat cell
{09082} oat cell carcinoma, electron micrograph showing neurosecretory granules

The "cell of origin" is the Kulchitsky (APUD) cell of the bronchial epithelium.

Almost all (99+%) tumors are related to tobacco smoking (rarely, just uranium mining).

Small cell carcinomas arise anywhere in the lung, most often near the hilum, and quickly spread along bronchi.

This cancer is infamous for early and widespread metastases and rapid death of the patient. (Surgery is usually considered futile; a few heroes are trying it again, however.)

The tumor is composed of small (2x size of a lymphocyte) cells with very little cytoplasm ("small blue cells", "oat cells"), with many mitoses and usually a lot of necrosis.

EM shows oval nuclei, finely dispersed chromatin, scanty cytoplasm, usually a few neurosecretory (dense-core, APUD) granules, sometimes squamous and/or "adeno" features.

Future pathologists: The tumor cells stain positive for neuron-specific enolase (NSE) and bombesin/GRP. While the three other major lung cancers usually stain for erbB, oat cell doesn't.

Like the K-cells from which they derive, "small cell" carcinoma is also known for secreting a lot of different substances.

Among these are ACTH, hADH, neurophysin, neuron specific enolase, bombesin (and/or the closely-related gastrin-releasing peptide), * bradykinin, * calcitonin, * growth hormone (hGH), * histamine, * hLH, * lipotropin (finding this suggests that elevated ACTH is not of pituitary origin: J. Clin. Endo. Metab. 86: 2997, 2001), * oxytocin, * prolactin, * somatostatin, * several different enzymes, etc., etc. (Bombesin/GRP is the cancer's autocrine growth factor).

IMPORTANT: Of these, the ones most likely to be detected are hADH (hyponatremia / water intoxication) and ACTH (Cushing's syndrome). Each syndrome occurs in 5-10% of "small-cell" patients (and ectopic secretion of the hormone is much commoner); they are seldom very serious but help make the diagnosis.

PTH-like hormone secretion and hypercalcemia are not typical of small-cell tumor patients.

The molecular biology of small cell carcinoma has been under study since the early 1980's. Update NEJM 359: 1369, 2008.

This is still a confusing area, and no major type of lung cancer has a genetic signature, though not surprisingly, a majority of each histologic type has mutated p53.

Oat cell's initial response to chemotherapy is usually very good. Most cases relapse and die, but occasional cures of the early disease are now being claimed (Chest 123: 259-s, 2003; two percent of all patients alive at 5 years Cancer 89: 523, 2000). Second primaries and chemotherapy-induced leukemias may still occur and kill the patient.

* Future pathologists: (1) On biopsy, it's often tough to tell crushed oat cells from crushed lymphocytes, especially if your biopsy instrument is dull. Most helpful is a common leukocyte antigen (Semin. Onc. 20: 153, 1993). (2) On fine needle aspiration, there is a tendency to mistake oat-cell carcinoma for non-oat-cell carcinoma. and carcinoids for oat-cell carcinoma. How not to make these errors: Arch. Path. Lab. Med. 129: 614 & 619, 2005.