Dear Retroflexions,

Over two years ago, I had a colon resection for the removal of a cancerous tumor (stage 2A). Ever since, I have had on and off bleeding in my stool. My surgeon has done two sigmoidoscopies and my gastro doctor has done two colonoscopies. The findings as of November 2015 are granulation tissue at the resection site (sigmoid). The bleeding comes and goes… maybe 2 weeks to a month…no blood…then it starts again. Its been back since July 1st.

The biopsies have all shown granulation tissue (benign). 

Have you every had a patient like me? My doctors say I worry to much.

-Mr. M

(update, 5 days later)

Here is an update. I had another sigmoidoscopy yesterday.

The granulation tissue is gone and has been replaced by “vascular proliferation at the anastamosis.”

I have attached a picture.

Please do a blog about this soon if you can.

Thank You,

-Mr. M

Left: anastomosis soon after surgery with signs of active healing. Right: proliferation of blood vessels at anastomosis.
Left: Appearance of anastomosis several months after surgery with signs of active healing/granulation tissue. Right: Two years after surgery there is a proliferation of blood vessels at anastomosis.

An anastomosis is a surgical connection of two hollow organs: In this patient’s case there was a tumor in the sigmoid colon that was removed, and the surgeon attached the two cut ends of the remaining colon together to make a working colon again. Like any wound, the anastomosis undergoes a healing process after surgery. Part of this process is the growth of new blood vessels into the anastomosis, and it seems like in the above case there is an overabundance of abnormal-appearing blood vessels growing all over the anastomosis. These blood vessels appear superficial and friable, and are prone to spontaneous bleeding. I would say they could be called telangiectasias, or angioectasias, or this could be called overabundant neovascularization of the anastomosis, or as the author of the report stated “vascular proliferation at the anastomosis.”

A few questions come to my mind when seeing a case like this: 1)Why does this happen; 2)Is this condition harmful; 3)How could we treat this condition to reduce the frequent bleeding; and finally, 4)What is the natural history of this problem if left untreated?

To preface the rest of this article, I don’t necessarily have the answers to questions 1-4 above! However, I have seen several cases like this, so here are some of my (probably oversimplified) thoughts on the topic: Whenever there is an abnormal proliferation of vascular tissue in the GI tract, we gastroenterologists should ask ourselves “what forces created this lesion?” For example, when we see esophageal varices, we think about portal hypertension, when we see colonic angiodysplasia we think of a multitude of risk factors. When I see angioectasias related to an anastomosis, I think of ischemia as the inciting factor. That is, there is limited blood flow to the tissue at the site of the anastomosis, as the natural blood supply has been disrupted by surgery, and there are now staples or sutures squeezing the tissue together. However, the anastomosis needs oxygen (in the form of blood flow) to heal, so it sends signals out locally to tell the body to hurry up and grow some new blood vessels to supply the healing tissue. These signals come in the form of a substance called vascular endothelial growth factor (VEGF).

VEGF makes the blood vessels grow alright, but sometimes the way they grow ain’t pretty. The new vessels can grow erratically and in excess, and we have what you see in the pictures above. Is it harmful? Probably not, except for the occasional (usually minor) bleeding episodes. In patients that bleed frequently from these anastomotic vascular lesions, I have treated the vessels with cautery or clips to stop or prevent the bleeding. It seems to work quite well, but I am always thinking that since I did nothing to treat the underlying problem of tissue-level ischemia that new funny-looking blood vessels will probably just grow back again at some point..and they often do.

What about just leaving this alone? Maybe the healing process is ongoing and in another year or two most of these vessels will self-destruct as the anastomosis fully matures. Or maybe deeper, more effective vessels will grow in, and therefore after we ablate these superficial angioectasias there will no longer be ongoing ischemia at the anastomosis, and new vessels will not be needed to repopulate the area? Perhaps this is all just wishful thinking? In extreme cases of anastomotic bleeding, sometimes the surgery has to be revised: The old anastomosis is resected and a new one is made, and hopefully the same problem does not happen again.

I’d really like to hear other gastroenterologists and (especially) surgeons chime in and give your thoughts on this problem. What do you do with patients with frequent late anastomotic bleeding? What is your understanding of why this happens? Please send me your perspectives and I will compile them in a follow up to this article!


Bosmans JW, Jongen AC, Bouvy ND, et al. Colorectal anastomotic healing: why the biological processes that lead to anastomotic leakage should be revealed prior to conducting intervention studies. BMC Gastroenterol 2015;15:180.

Scappaticci FA, Fehrenbacher L, Cartwright T, et al. Surgical wound healing complications in metastatic colorectal cancer patients treated with bevacizumab. J. Surg Oncol 2005;91:173-180.

Tanaka IM, Sugio IT, Tanaka ST, et al. Local VEGF administration enhances healing of colonic anastomoses in a rabbit model. Eur Surg Res 2009;42:249-57.

Cologuard is a convenient, non-invasive, commercially-available screening test for colorectal cancer. The test is very easy for doctors to order and for patients to complete. There is no bowel prep necessary for this test, and no special diet to follow. The patient doesn’t even need to leave their house to complete the test! Sounds like a pretty good test so far…here’s how it’s done:

Once ordered, the company mails a kit to the patient’s house, and when the urge strikes, the patient places the collection device over the toilet and makes a deposit. There are a few simple preparation steps such as adding in a small amount of liquid that the company provides, and then the entire container is mailed back to the company in the provided packaging. The Cologuard test looks for certain DNA mutations and other abnormalities in the stool to determine if a test is positive or negative. A few weeks later the doctor gets a report indicating the result.

This polyp was found in the rectum of a patient referred due to a positive Cologuard test. It was removed during the colonoscopy: Pathology revealed a tubular adenoma.
This polyp (between red arrows) was found in the rectum of a patient referred due to a positive Cologuard test. It was removed during the colonoscopy: Pathology revealed a tubular adenoma.

The study that determined the characteristics of the Cologuard test basically performed the test on almost 10,000 patients at average-risk of colon cancer, and then had the patients undergo colonoscopy as the gold-standard test. The results of the Cologuard test were not available to the patients or the endoscopist at the time of the colonoscopy. The major results of this study showed that the Cologuard test had an excellent sensitivity (92%) for detecting colorectal cancer. It was far less sensitive for picking up advanced adenomatous polyps (42%). Remember that sensitivity is the main thing we care about in a screening test: we want the test to miss none of the patients who have the disease. A perfect screening test would have a sensitivity of near 100%, meaning that if 10 people have the disease out of the 1000 people screened, all 10 people would get a “positive” test result.

Sensitivity isn’t everything however…we also want a test that gives a negative result when someone does not have the disease in question. Since a full review of statistics is both 1)boring, and 2)not the point of this article, I will cut right to the chase!

What does a positive Cologuard test mean?

First and foremost, a positive result on the Cologuard test means that you need to have a colonoscopy. Not a virtual colonoscopy, or another stool test, or another scan of some sort…you need a real optical colonoscopy. Only about 4% of people will have cancer found on colonoscopy. 51% will have a precancerous polyp. The rest (45%) will have nothing found on colonoscopy. So to simplify even further, just a little more than half of people with positive results will have something abnormal (cancer or a polyp) found on colonoscopy.

What does a negative Cologuard test mean?

A negative test means that there is a less than one-percent (0.06% to be exact) chance of having cancer found on colonoscopy. However, about 34% of people with negative tests still have precancerous polyps found on colonoscopy, with the remainder (66%) of people with negative Cologuard results having truly negative colonoscopies.

What is immediately apparent from these numbers is that Cologuard rarely misses cancer. However, if we count polyps as a significant finding, there are plenty of false-positive results (45%) and plenty of false-negatives too (34%). So is Cologuard a good test overall?

I would argue that it is a great test, with one important caveat: Cologuard is not a replacement for colonoscopy! Colonoscopy is still superior to Cologuard in detecting both cancer and polyps. However, as an alternative to colonoscopy for people who either can’t or won’t get a colonoscopy, Cologuard is an excellent choice. In terms of population-based screening, I think it will get a lot more people screened overall (people who would have otherwise not done anything for screening). When compared to not screening at all, Cologuard is a whole lot better than nothing!


Imperiale TF, Ransohoff DF, Itzkowitz SH, et al. Multitarget stool DNA testing for colorectal-cancer screening. N Engl J Med 2014;370:1287-97.

Cologuard website:

An angiodysplasia is a type of vascular malformation that is commonly found throughout the gastrointestinal tract, and fairly often in the right colon. These lesions can be described as thin-walled, fragile, superficial abnormal “tangles” of small blood vessels that have a propensity to bleed, especially in patients who take blood thinning medications. In a patient with anemia or bleeding, we usually treat these lesions with some type of cautery, most commonly argon plasma coagulation (APC). Other terms used for angiodysplasia include arteriovenous malformation (AVM), angioectasia, telangiectasia. (These terms actually all mean slightly different things, but let’s not get into that now.)

When treating angiodysplasia in the right colon, especially in the thin-walled cecum, it is important to use a minimal amount of APC, delivered in short bursts at low flow settings. The concern is that immediate, or more often late, damage can occur to the bowel wall leading to perforation. One useful technique to minimize the risk of perforation when treating large angiodysplastic lesions (let’s arbitrarily define large as greater than 10-mm in diameter) is to first inject saline under the lesion to lift it up and away from the deeper layer of bowel wall. Yes, this is exactly like the way we treat big polyps, so it is already a skill that most gastroenterologists have.

Although this angiodysplasia was not that large (maybe just 10-mm in the largest dimension), this patient was taking dual antiplatelet therapy, and needed to go on full-dose anticoagulation right after the procedure. Since the lesion was in the cecum, it was a good candidate for lifting prior to cautery.

Treatment of cecal angiodysplasia with APC. The lesion is injected with saline (the ileocecal valve can be seen in the second picture). After injection, APC is used to ablate the lesion. A few clips were placed since the patient needed multiple blood-thinning agents.
Treatment of cecal angiodysplasia with APC: The lesion is lifted with submucosal injection of saline (the ileocecal valve can be seen in the second picture). After injection, APC is used to ablate the lesion. A few clips were placed to prevent delayed bleeding since the patient needed to remain on multiple blood-thinning agents.

For an excellent review of gastrointestinal angiodysplasia, including theories on why they form, see the following reference:

Sami SS, Al-Araji SA, et al. Review article: gastrointestinal angiodysplasia-pathogenesis, diagnosis, and management. Aliment Pharmacol Ther 2014;39:15-34.

“I’m not really a pill person.”

“I was never one for all those pills.”

“I don’t really like taking those pills.”

“I’m not really into taking pills.”

As a doctor, I regretfully hear some version of this phrase every day. It’s almost accusatory, like “Hey Doc, don’t even think about pushing all those pills on me!”

Luckily, since I am a gastroenterologist, I don’t dispense nearly as many medications as doctors in some other fields do; ninety-percent of the medications I write for are antacids, laxatives, a few antibiotics and diuretics, and a few meds for colitis. However, I recently had the experience of starting a patient on a blood pressure medication to help lower his absolutely out-of-control blood pressure. I normally refer patients with such issues back to their internists to treat their non-GI issues, but this patient was in between internists, having just moved to the area. So I reached back in my mind and tried to remember some internal medicine from residency training, and gave him a prescription for a diuretic-antihypertensive combo pill. “Start this medicine today,” I said, “and you need a new internist now that you live on Long Island, so make an appointment ASAP with Dr. Schwartz…here is his number.”

A few weeks later I see the patient for follow up regarding his gastro issues and his blood pressure is still 220/110. Maybe my memory from residency was a little fuzzy when it came to treating hypertension? After all, I was very sleep-deprived back then. So I told the patient that the meds that we started recently were not cutting it, we would have to up the dose now, and he really needs to see that internist before he ends up in the hospital with a heart attack or stroke! It was then when he told me he wasn’t really taking the blood pressure medication that I went out of my way to prescribe him. I asked him why not, was there a side-effect, could he not afford them, what happened? And that’s when he uttered the dreaded phrase: “Ehhh, I’m not really one for all those pills,” while making a dismissive rapid back-and-forth rotary movement with his open right hand that suggested that all pills in general were nonsense.

all those pills

When someone has a real disease and tells me they’re not a pill person, I can’t help but wonder what exactly does that mean? Someone with dangerously high blood pressure, who is not a pill person will soon become a stroke person. Someone with uncontrolled diabetes who is not a pill person will soon become a blind person, or a dialysis person, or both. When you have ulcerative colitis and are not a pill person, you are a bloody diarrhea person. These are usually the alternatives.

Now don’t get me wrong, I am all for “natural” methods of controlling diseases. However, when the natural remedies and diets fail, it’s time for real meds. The very natural alternative to taking pills is uncontrolled disease. Suffering and death are very natural parts of the human experience…let’s avoid those things for a while if we can.

Image: Vishnevetsky

In the last article, we covered how to retroflex the scope in the right colon. In this article, I will show you why someone would want to do such a thing. Let’s take this one step further: Should we do this routinely during every colonoscopy?

Simply put, sometimes the only way to physically get to whatever it is you need to reach, is to retroflex the scope in the right colon. Usually this is done to resect a polyp in a difficult location, such as the proximal (back) edge of a fold. In a previous article, I showed an example of using retroflexion in the right colon to remove the remaining pieces of a large polyp. The following is an example from a more recent case:  I was able to see this polyp in the regular straight view, but due to the location of the polyp on a fold, I could not see the entire thing. The first thing I tried was injecting some saline behind the polyp to try and lift it forward towards the scope. Well, the lesion lifted well, but still was really hard to see and proved impossible to get the snare around. Retroflexing in the ascending colon solved that problem easily:

Retroflexed view in the ascending colon: Red arrow shows a polyp behind a fold. The lesion was injected with saline (prior to taking this picture), and then removed with a hot snare.
Retroflexed view in the ascending colon: Red arrow shows a polyp behind a fold. The lesion was injected with saline (prior to taking this picture), and then removed with a hot snare.

You may also have to retroflex to treat bleeding, such as a bleeding angiodysplasia, or a bleeding polypectomy site. Sometimes a polyp that was easy to see is removed and the site retracts behind a fold and becomes almost impossible to see. If that site bleeds a few days later, retroflexion to get in a better position to place clips straight onto the site may be the best way to treat it.

Should retroflexion in the right colon become a routine part of screening colonoscopy? Let’s frame this question with the following facts: Colonoscopy is less-protective against right-sided cancers (which implies that colonoscopy is less-effective at finding or removing right-sided polyps.) Polyp detection in the right colon is more difficult for several reasons (flat lesions, prep quality). Another one of these reasons is that polyps may easily hide behind folds and be impossible to see in forward-view.

To combat the risk of right-sided colon cancers, some authors have advocated a routine second-examination of the right colon during screening colonoscopy. A routine second-exam from the hepatic flexure to the cecum (in regular forward view) has been shown to yield an additional adenoma about 10% of the time. To me, this is a big deal, and means that by not looking twice at the proximal colon, we are missing small polyps in 1 out of 10 patients. As it turns out, there is a statistically similar increase in polyp detection if the second exam is done in retroflexed view.

The take home message is that performing a second proximal colon exam is important, either in forward view, or retroflexed view. Doing it retroflexed has the theoretical benefit of allowing you to see things from a different angle and possibly pick up lesions that would be missed by repeat forward view. The studies don’t really support that a retroflexed view is any better than a repeat forward view, however the important thing is that you do a second look at all.

The following reference is an excellent practical read on the topic, and is free to download on the AJG website:

Rex DK. How I approach retroflexion and prevention of right-sided colon cancer following colonoscopy. Am J Gastroenterol 2016;111:9-11.

Additional reference:

Kushnir VM, Oh YS, Hollander T, et al. Impact of retroflexion vs. second forward view examination of the right colon on adenoma detection: a comparison study. Am J Gastroenterol 2015;110:415-22.