COLONIC WALL STRUCTURE

Early gross descriptions of colons bearing diverticula reported thickening of the muscle wall and shortening of the taenia, with a resulting accordion-like pleating of the folds. This appearance, called myochosis (Greek myo, “muscle,” and chosis, “a heaping up”), is corroborated by the colonoscopic appearance in which markedly thickened folds are seen. Routine histology, however, does not reveal muscle hypertrophy.

Electron microscopic studies confirm that the colonic walls in patients with diverticulosis have structurally normal muscle cells but, compared with controls, contain a greater than 200% increase in elastin deposition between the muscle cells in the taenia.²³ The elastin is laid down in a contracted form, resulting in shortening of the taenia and bunching of the circular muscle. An increase in type III collagen synthesis in patients with diverticulosis also has been described, raising the possibility that age-related changes in collagen composition play an etiologic role.²⁴

In addition to an overall increase in the collagen content, an overexpression of a tissue inhibitor of metalloproteinases has been identified in colons with diverticula.^25,26 Because matrix metalloproteinases are believed to regulate deposition of extracellular matrix proteins, an increase in their regulatory molecule, tissue inhibitor, might explain the increase in elastin and collagen deposition found in diverticular colons. The importance of intestinal wall connective tissue also is underscored by the higher rate of diverticulosis reported in patients with connective tissue disorders, such as Ehlers-Danlos syndrome, Marfan’s syndrome, and scleroderma.⁵

MOTILITY

Early investigations using colonic manometry demonstrated higher resting, postprandial, and neostigmine-stimulated luminal pressures in patients with diverticulosis compared with controls.^27,28 Based on simultaneous manometry and cineradiography, Painter proposed a theory of segmentation, postulating that contraction of the colon at haustral folds caused the colon to act not as a continuous tube but as a series of discrete “little bladders,” which led to excessively high pressures within each segment.^28,29 He further suggested that the Western diet might alter colonic motility to augment hypersegmentation, thereby increasing the tendency to form diverticula.

More recently, using flexible endoscopy to accurately place manometric catheters within the sigmoid colon, the motility abnormalities previously described have been confirmed.³⁰ Further, patients with symptomatic diverticular disease have been reported to have higher motility indices than either asymptomatic patients or persons without diverticula.³¹ In addition to increased contraction amplitude, another study found retropropagation of contractile waves in diverticular segments of colon, indicating that motility in these patients may be abnormal in magnitude and direction.³² An Asian study has confirmed elevated resting and stimulated luminal pressures in the presence of proximal colon diverticulosis.³³

Although the demonstration of abnormal motility and elevated intraluminal pressures in the diverticular colon has been consistent, the physiologic basis for these abnormalities is less clear. Ion transport across the epithelial membrane of diverticular colons is the same as in controls.³⁴ The number of myenteric and submucosal plexus neurons in diverticular colons are normal; however, the number of interstitial cells of Cajal (enteric pacemaker cells) is reduced.³⁵ Increased activity of excitatory cholinergic nerves and a decreased activity of nonadrenergic, noncholinergic inhibitory nerves, the latter in part mediated by nitric oxide, have been demonstrated in diverticular colons compared with control colons.³⁶ The magnitude of electrically stimulated contraction in diverticulosis-affected sigmoid colon is markedly reduced by antagonists of cholinergic and tachykinin neurotransmitters.³⁷ In contrast, a study of the tachykinin neurotransmitter system showed a decreased contractility of circular muscle induced by substance P in diverticular colons compared with normal ones.³⁸

The exact details of the neurochemical derangements underlying diverticulosis still need to be clarified, but there does appear to be an imbalance in the normal excitatory and inhibitory influences that results in the increased tonicity observed in colons with diverticulosis. The abnormal pressures and tone might contribute to formation of diverticula and to bowel dysfunction in patients with diverticulosis.

DIETARY FIBER

The wide geographic variation of diverticular disease, with higher prevalence in countries with a Westernized diet, has long suggested a dietary factor in its pathogenesis. Low intake of dietary fiber has been strongly suspected to be the main dietary factor behind these geographic differences. Burkitt and Painter were early proponents of this theory, labeling diverticulosis a “deficiency disease” that, like scurvy, should be avoidable with dietary changes.² In one important study, they demonstrated that persons in the United Kingdom who consumed a refined, low-fiber Western diet had stool transit times more than twice as slow and stool weights significantly less than those of rural Ugandans eating a diet very high in fiber.³⁹ The long intestinal transit time and smaller-volume stools were believed to allow an increase in intraluminal pressure, thus predisposing to diverticular herniation, whereas bulkier stools were associated with less colonic contraction and lower wall pressures.

Although more recent studies in Western cohorts have failed to confirm this finding in humans consistently, corroborative animal data do exist. Wistar rats fed low-fiber diets developed diverticula in 45% of instances, compared with only 9% of those on high-fiber diets.⁴⁰ In humans, other observational evidence exists with respect to the etiologic role of fiber in diverticulosis. In the United States, fiber intake decreased by 28% from 1909 to 1975,⁵ a period of dramatic increase in the prevalence of diverticular disease. In a British study, a group of vegetarians on a high-fiber diet had a lower prevalence of diverticulosis than did nonvegetarians (12% vs. 33%).⁴¹

Dietary influences for diverticulosis may have different effects on the right and left sides of the colon. Right-sided diverticular disease was shown in an Asian study to have no association with intake of fruits and vegetables or supplemental fiber, but it was strongly associated with meat consumption.⁴² Whether these associations apply to Westerners with right-sided diverticulosis is not known.

The potential role of fiber in the treatment of diverticular disease is addressed later in this chapter.

ASYMPTOMATIC DIVERTICULOSIS

Asymptomatic diverticulosis usually is an incidental finding in patients undergoing evaluation for other indications, such as occult blood loss or colon cancer screening. With an increasing number of people undergoing colonoscopic screening for colon cancer, it is likely that more such asymptomatic patients will be found. There is no clear indication for any specific therapy or special follow-up in these patients.

A possible prophylactic benefit of a high-fiber diet has been suggested in two publications of 47,888 male health professionals who were followed over four years and in whom 385 (0.75%) new cases of symptomatic diverticular disease were identified.^43,44 A dietary review found an inverse association between dietary fiber intake and the risk of subsequently developing clinically evident diverticular disease. The study also noted that fruit and vegetable fiber, or insoluble fiber, had a greater protective effect than did cereal fiber. Conversely, diets high in fat and red meat were associated with an increased risk of diverticular disease. Although prospective randomized trials are lacking, these findings suggest that patients with asymptomatic diverticulosis and SUDD (see later) might benefit from increasing their fruit and vegetable fiber intake while decreasing their fat and red meat consumption, a sensible lifestyle change that likely also provides other salutary health benefits.

For decades, it has been a widely held but weakly supported belief that patients with diverticulosis should avoid consumption of seeds, nuts, and popcorn to avoid possible plugging of diverticula and precipitating an episode of diverticulitis; this theory also was investigated in the aforementioned cohort of male health professionals.⁴⁵ Not only was no increase in risk of diverticular complications found as a result of consuming nuts, corn, popcorn, or seeded fruit (strawberries or blueberries), but consumption of nuts and popcorn were inversely associated with the risk of developing diverticulitis. Although there probably is not enough evidence to warrant actively encouraging patients with diverticulosis to eat large quantities of nuts, popcorn, or seeds, neither should these foods be categorically avoided.

SYMPTOMATIC UNCOMPLICATED DIVERTICULAR DISEASE

Clinical Features

Some patients come to clinical attention because of nonspecific abdominal complaints and are found to have diverticulosis coli. A causal relationship between the diverticulosis and the abdominal symptoms often is difficult to establish. If there are features that are consistent with a diverticular source, however, and there is no evidence of a serious inflammatory condition, the disease may be defined as SUDD. Most patients with SUDD present with left lower quadrant pain; the British refer to this condition as painful diverticular disease. The pain often is exacerbated by eating and diminished by defecation or the passage of flatus. Patients also may report other symptoms of colonic dysfunction, including bloating, constipation, diarrhea, or the passage of mucus per rectum. Physical examination may be normal or may reveal fullness or mild tenderness in the left lower quadrant, but frank rebound or guarding is absent. Because rates of occult bleeding in diverticulosis are similar to those in healthy controls, a positive fecal occult blood test never should be attributed to diverticulosis.⁴⁶

Because a barium enema can characterize the number, size, and location of diverticula, it previously had been a commonly used initial study in such patients. Barium enema, however, may be insufficient to rule out competing or associated diagnoses such as malignancy in patients with diverticulosis. In symptomatic patients in whom a double-contrast barium enema showed sigmoid diverticulosis, subsequent colonoscopy confirmed only 55% of neoplastic lesions that were suspected on the barium enema, whereas eight polyps and three malignancies were identified on colonoscopy that were missed on the barium enema (24% false-negative rate for barium enema).⁴⁷

Although the barium enema may remain useful in certain cases, particularly if a colonoscopy cannot be performed safely or completely, endoscopic evaluation (Fig. 117-2) has assumed a primary role in the evaluation of most patients, particularly to exclude neoplasia. It once was believed to be unsafe to perform colonoscopy in patients with diverticulosis because of an increased risk of perforation; however, it subsequently was demonstrated that manometrically measured burst pressures for diverticula far exceed the usual pressures encountered during routine sigmoidoscopy or colonoscopy, even with the endoscope pressing against the wall or with heavy air insufflation.⁴⁸ These data and many years of clinical experience have demonstrated the relative safety of using endoscopy to evaluate patients with abdominal symptoms. Caution should be exercised in patients with suspected or proved diverticulitis, however, because of a theoretical risk of perforating the wall of a diverticulum that has lost its integrity due to inflammation. In all patients, air insufflation should be minimized and excessive force in advancing the endoscope should be avoided.

Figure 117-2. Colonoscopic view of the sigmoid colon in a patient with symptomatic uncomplicated diverticular disease.

The diverticula-laden colon can be challenging for the endoscopist to navigate because of spasm, luminal narrowing, fixation from prior inflammation and fibrosis, or confusion between luminal and diverticular openings. A number of solutions have been proposed to alleviate this problem. The use of a smaller-diameter pediatric colonoscope can be useful for difficult colons. One group has reported a success rate of more than 90% with a pediatric colonoscope in cases in which an adult colonoscope could not be passed through the sigmoid; 44% of these patients had diverticulosis.⁴⁹ When colonoscopy with standard and pediatric colonoscopes were compared, the reason for failure to complete the examination was thought to be stenosing diverticular disease in 12 of 14 patients with the standard colonoscope, compared with two of eight with the pediatric colonoscope.⁵⁰ A technique involving distention of the lumen with 100 to 300 mL of water, called the sigmoid floatation maneuver, was said to have facilitated colonoscopy in six technically difficult cases of severe diverticular disease.⁵¹

Occasionally, an endoscopist encounters an inverted diverticulum, where a diverticular dome protrudes into the lumen instead of out from it. These inversions often resemble polyps endoscopically, although they may be distinguished by their normal overlying mucosa, broad base, and location within a bed or row of diverticula. They are soft-appearing when manipulated with the endoscope tip or a biopsy forceps (pillow sign) and may be reducible. On barium enema, inverted diverticula appear as broad-based sessile polyps with a characteristic central umbilication,⁵² although it is not always possible to distinguish such a diverticulum from a polyp. When inverted diverticula are encountered, their removal should be avoided, although inadvertent colonoscopic diverticulectomy has been reported,⁵³ and these patients had uneventful recoveries with conservative therapy.

The presenting symptoms of SUDD overlap considerably with those of irritable bowel syndrome (IBS). Some authorities have postulated that diverticula are, in fact, a late consequence of IBS. In a Danish cohort of IBS patients, one third of whom had diverticula, no difference in symptoms or prognosis was detected between those with diverticula and those without diverticula over more than five years of follow-up.⁵⁴ This finding led the investigators to conclude that there is no basis to consider SUDD as a separate entity from IBS. Ritchie reported that there was a similarity of pain sensation from rectal balloon distention in patients with IBS and those with diverticulosis.⁵⁵ Whether these two disorders are distinct entities is unknown and probably not clinically important, because both are treated in a similar nonspecific fashion with equally good prognoses.

UNCOMPLICATED DIVERTICULITIS

Diagnosis

Most patients with acute diverticulitis present with signs and symptoms sufficient to justify the clinical diagnosis and institution of empiric therapy. Clinical diagnosis can, however, occasionally be inaccurate, and emergency surgery for presumed diverticulitis, without the benefit of radiologic confirmation, carries a misdiagnosis rate as high as 34% to 67%.⁷⁸ Therefore, radiologic studies to confirm the diagnosis of diverticulitis should be employed, particularly if invasive intervention may be required.

Plain Films

An erect chest film, together with erect and supine abdominal films, should be performed on patients with significant abdominal pain. The erect chest film has the dual purpose of detecting pneumoperitoneum, which has been reported to be present in up to 11% of patients with acute diverticulitis,⁷⁹ and of assessing cardiopulmonary status in a generally elderly population with common comorbid illness. Plain abdominal films are abnormal in 30% to 50% of patients with acute diverticulitis,^79,80 with findings that include bowel dilatation from obstruction or ileus, or a soft tissue density suggesting an abscess.

Contrast Enema Examinations

Contrast barium enema (Fig. 117-3) had been the diagnostic standard for diverticulitis and its complications for many years. Because the use of barium in the setting of an intestinal perforation carries a risk of barium peritonitis, only water-soluble contrast enemas, such as Gastrografin, should be used in the setting of suspected diverticulitis. A gentle, single-contrast study should be performed and terminated once findings of diverticulitis are discovered, with visualization of the entire colon deferred to a later date; air (double)-contrast studies are not indicated. Findings considered diagnostic of diverticulitis include demonstration of extravasated contrast material with or without the outlining of an abscess cavity, an intramural sinus tract, or a fistula.^1,81 Extensive diverticulosis, spasm, mucosal thickening or spiking, or deformed sacs, although suggestive of diverticulitis, are not conclusive. An extraluminal mass compressing the colon is said to be the most common finding in severe diverticulitis,⁸² although this finding is not specific for this diagnosis. Obviously, in the absence of diverticula or associated findings, the diagnosis must be reconsidered. Contrast enema has been shown to have a sensitivity of 62% to 94% for detecting acute diverticulitis, with false-negative results in 2% to 15%.^68,83

Figure 117-3. Film from a barium enema in a patient with pancolonic diverticulosis. There is also marked redundancy and overlapping of the sigmoid colon, which obscures the identification of intraluminal lesions.

Computed Tomography

Because diverticulitis is mainly an extraluminal disease, luminal contrast studies may be inaccurate. Computed tomography (CT) (Fig. 117-4) has now replaced contrast enema as the diagnostic procedure of choice for acute diverticulitis and has the ability to image mural and extraluminal disease and to enable therapy with percutaneous drainage of abscesses. Abdominal and pelvic scanning ideally is performed with water-soluble contrast, given both orally and rectally, and with intravenous contrast when it is not contraindicated. CT criteria for diverticulitis include the presence of diverticula with pericolic infiltration of fatty tissue (often appearing as fat stranding), thickening of the colon wall, and formation of abscesses.

Figure 117-4. Computed tomography scan of a patient with acute diverticulitis showing colon wall thickening, and stranding of the pericolic fat.

The earliest large series of CT findings in diverticulitis reported the finding of pericolic fat inflammation in 98%, diverticula in 84%, a colonic wall thickness greater than 4 mm in 70%, and an abscess in 35%.⁸⁴ Contrast enemas in the same patients underestimated the extent of disease in 15 (41%) of 37 cases. Numerous subsequent trials comparing these two modalities in patients with suspected diverticulitis consistently have reported CT sensitivities of 93% to 98% and specificities of 75% to 100%, significantly more accurate than contrast enemas.^81,85,86 CT also has been found to be highly sensitive and specific for right-sided diverticulitis and in helping to differentiate diverticulitis from colorectal cancer of the ascending colon and cecum.^87,88

Although some reports show a lower sensitivity of CT for diverticulitis,^89,90 it is increasingly becoming clear that, when diagnosis is in doubt or clinical deterioration occurs, CT is the best primary radiologic diagnostic modality. Conversely, in patients with mild disease and in whom the diagnosis is straightforward, CT scanning might not be necessary.

Ultrasonography

Based on its relatively low cost, convenience, and noninvasive nature, ultrasonography (US) has been advocated as a potentially useful diagnostic modality in diverticulitis. Characteristic findings implying active inflammation include bowel wall thickening, presence of diverticula or abscesses, and hyperechogenicity of the bowel wall. US has a reported sensitivity of 84% to 98% and specificity of 80% to 93%.^91,92 One study of 71 patients with suspected diverticulitis who underwent US reported a negative predictive value of 100%.⁹³ A trial comparing US with CT revealed equally good accuracy.⁹⁴ US also is useful in female patients to exclude gynecologic pathology.

Despite these encouraging data, US remains highly operator dependent, especially for detecting abscesses between loops of bowel, air-filled abscesses, or disease complications posterior in the abdomen that are hard to visualize. US therefore remains a second-line diagnostic tool to be used in select circumstances.

Magnetic Resonance Imaging

With limited resolution secondary to motion artifact introduced by peristalsis and respiration, the potential role of MRI remains to be demonstrated. Whether decreased scan times and intraluminal contrast agents will overcome these limitations remains to be seen.

Endoscopy

Because of the risk of perforation, either from the instrument itself or from air insufflation, endoscopy generally is avoided in the initial evaluation of patients with suspected acute diverticulitis. A limited sigmoidoscopy with minimal air insufflation may be helpful to exclude alternative diagnoses, such as IBD, carcinoma, or ischemic colitis in unclear cases. As noted earlier, once the acute phase has passed, a colonoscopy should be electively performed one to three months later to exclude competing diagnoses, particularly neoplasia.

COMPLICATED DIVERTICULITIS

Abscess

When perforation of a colonic diverticulum occurs, the ability of the pericolic tissues to control the spread of the inflammatory process determines the subsequent clinical course of disease and its treatment. A localized phlegmon initially develops with limited spread. Further spread can lead to the formation of larger local or distant abscesses (Hinchey stages I and II, respectively). When abscess contents spread diffusely in the peritoneum, causing purulent or fecal peritonitis (Hinchey stages III and IV, respectively), it can lead to sepsis and death if the patient does not undergo urgent surgical intervention. Clinical signs suggesting abscess formation include a tender abdomen and sometimes a tender mass on physical examination, and persistent fever, or leukocytosis despite an adequate trial of appropriate intravenous antibiotics. Once an abscess is suspected, radiologic evaluation with a CT scan is the best modality for confirming the diagnosis and following its course over time.

Small pericolic abscesses (Hinchey stage I) often can be treated conservatively with broad-spectrum antibiotics and bowel rest.⁹⁸ In one series of patients with diverticulitis, seven of 10 patients with pericolic abscesses responded successfully to conservative treatment.¹⁰¹ This favorable prognosis can result from maintenance of a fistula between the abscess and the colon lumen, thus permitting spontaneous internal drainage. Continued noninterventional management of abscesses should be considered only in stable patients who demonstrate unequivocal improvements in pain, fever, tenderness, and leukocytosis over the first few days of therapy.

CT-guided percutaneous drainage of abdominal abscesses has assumed a prominent complementary role to surgery (Fig. 117-5). The immediate advantage of percutaneous catheter drainage is rapid control of sepsis and patient stabilization without the need for general anesthesia. It often eliminates the need for a multiple-stage surgical procedure with colostomy,¹⁰² instead allowing temporary palliative drainage and subsequent single-stage resection in three to four weeks. Success rates of CT-guided drainage for stabilizing patients and safely allowing subsequent single-stage procedures range from 74% to 80%.^103,104 An urgent surgical procedure is required in the 20% to 25% of patients in whom the abscess is multiloculated, anatomically inaccessible, or not resolving with percutaneous drainage.

Figure 117-5. Computed tomography (CT) scans with CT-guided percutaneous drainage of an abdominal abscess complicating acute diverticulitis. A, Arrows point to the abscess arising from the sigmoid colon. B, A drainage catheter has been inserted into the abscess. The patient also was treated with intravenous antibiotics. C, The catheter has been removed, and the abscess has resolved.

The approach to surgical intervention has evolved over the last half century. Historically, three-stage management was preferred for complicated diverticulitis of any severity: stage I involved a proximal colostomy and drainage of abscesses, stage II was for resecting the diseased bowel (colostomy maintained to protect anastomosis), and stage III was to restore bowel continuity. Although the multiple-stage surgical approach to diverticular disease was effective for managing abscesses, generally it is desirable to limit the number of laparotomies to minimize morbidity and duration of hospitalization. Additionally, only about half of patients undergoing proximal-end colostomy have the colostomy reversed because of the technical difficulties and the risks of anastomotic leakage and mortality.¹⁰⁵

In the last few decades, two-stage management (e.g., primary resection of diseased bowel with proximal end colostomy and oversewing of the distal stump, also referred to as the Hartmann procedure, followed by reanastomosis) and, increasingly, management with a single operation (resection with primary anastomosis) have become the preferred surgical approaches. The switch to fewer operations has been made without a discernable compromise in overall outcomes relative to three-stage approaches.¹⁰⁶ Single-stage management, which is becoming the standard for elective management of uncomplicated diverticulitis, also is being used increasingly in appropriate patients with complicated diverticulitis. To do so, bowel preparation is believed to be necessary.

When free perforation is ruled out, bowel preparation is performed before the operation with either a one-day polyethylene glycol (PEG)-electrolyte lavage or a traditional two- or three-day mechanical preparation; both types of preparation have been shown to have similar outcomes in a randomized trial.¹⁰⁷ In the setting of urgent or emergent surgery when free perforation is a concern, bowel lavage may be performed “on table,” thereby allowing a primary anastomosis to be performed in a single stage.^108,109

Regardless of approach, surgery for complicated diverticulitis has evolved into a relatively safe procedure with one of the highest success rates of any of the common gastrointestinal surgical procedures.¹¹⁰

RECURRENT DIVERTICULITIS

For patients who respond well to conservative therapy, the issues of likelihood of recurrence and elective prophylactic surgical resection arise. The risk of recurrent symptoms following an attack of acute diverticulitis has been reported to range from 7% to 45%,^1,5,17,68 with half of second attacks occurring within one year. Historical evidence had suggested that recurrent attacks were less likely to respond to medical therapy and had a higher mortality rate.^17,68 Predicated on the notion that diverticulitis was a progressive disease, elective resection had been traditionally recommended after two attacks of diverticulitis.^98,99,125 This approach is now being challenged.

Studies such as the 13-year experience of the Mayo Clinic have suggested that the risk of poor outcomes is not higher with recurrent diverticulitis (more than two attacks) than it is with the first one or two attacks.¹²⁶ In fact, the mortality rate from diverticulitis in this cohort was higher in those with no prior history of diverticulitis (10%) than in those who had had prior attacks (2.5%). This difference may be due to the higher rate of free perforation in initial attacks relative to subsequent attacks. Thus, multiple recurrences do not appear to predict less-favorable outcomes. Additionally, there are some emerging medical therapies for those with diverticulitis to attempt to reduce likelihood of future attacks. As noted in the discussion of treatment for SUDD, studies have begun to show a decrease in recurrence rates of diverticulitis and symptomatic diverticular disease in patients treated with mesalamine, rifaximin, probiotics, and combinations of these agents.^{58–63,65,127,128} Finally, a recent decision analysis predicted that performing colectomy after the fourth attack of diverticulitis rather than after the second attack would result in fewer deaths and colostomies while having a superior cost-effectiveness.¹²⁹ Recognizing this trend, published guidelines on the decision to perform colectomy after an attack of diverticulitis are now advocating evaluation on a case-by-case basis rather than empirically performing elective surgery after the second attack.¹³⁰

If the choice to operate is made, most patients report having a good functional outcome and low rates of recurrent disease after elective resection for recurrent diverticulitis.¹³¹ In considering an elective sigmoid resection, relevant variables include the severity and responsiveness of the attack(s), the general health of the patient, the risk to the patient of a subsequent attack, and the risk of the resection itself. The latter factor may be lessened by the increasing use of the laparoscopic approach in many patients. Additionally, up to 10% of patients have symptomatic recurrent diverticulitis after surgical resection, and reoperation may be required in 2% to 3%.^{68,98,131,132} Patients undergoing resection for diverticulitis have higher recurrence rates when the sigmoid colon is used for the distal resection margin, rather than the rectum¹³³; it is recommended to resect the entire distal colon whenever possible, forming the distal anastomosis with the proximal rectum and the proximal anastomosis with a noninflamed portion of colon.^68,98

THE YOUNG PATIENT

Diverticulitis is relatively uncommon in patients younger than 40 years (2% to 5% of all patients with diverticulitis^1,17) but the incidence in this age group may be rising.¹³⁴ Nevertheless, because diverticulitis is uncommon in younger patients, it is often missed or mistaken for other diagnoses, such as appendicitis or IBD. Like diverticulitis in older patients, the disease is mainly sigmoid in location, although one report has described a right-sided predominance in young Israelis.¹³⁵ There seems to be a significant male predominance in young patients.^1,136 Attacks often are more severe, and 40% to 88% of younger patients require urgent surgery during their initial attack; recurrence and complication rates are also higher than in older patients.^1,136,137

When patients with acute diverticulitis are managed nonoperatively, youth is an independent risk factor for poor outcome in subsequent course,¹³⁸ possibly due to delay in diagnosis. For these reasons, some authors have advocated elective segmental colectomy in a healthy young person after one well-documented episode of diverticulitis^68,125,138; others have questioned this approach.^98,135,139 In the largest series to date of young patients with diverticulitis, the authors suggested that the higher incidence of surgical management in these patients relative to their older counterparts was not due to worse outcomes but rather to a higher rate of elective procedures done to prevent the expected poor outcomes.¹³⁴ Thus, the latest surgical guidelines are advocating more of a case-by-case approach to elective resections for all patients with diverticulitis.¹³⁰

THE ELDERLY PATIENT

Because diverticular disease is more prevalent in older populations, diverticulitis in the elderly warrants special mention. Diverticulitis can manifest with more subtle symptoms and signs in the elderly, making the diagnosis more challenging. Distinguishing diverticulitis from colorectal cancer becomes a much more important issue in older patients, because the incidence of colorectal cancer also increases with age. Though the risk of a severe initial attack of diverticulitis may be lower in older patients than in younger ones, the risk of death when diverticulitis results in perforation is more than three times higher in the elderly.¹⁴⁰ However, it appears that the number of comorbidities is a stronger predictor of mortality from perforated diverticulitis than age.¹⁴¹ Because subsequent complicated disease is uncommon if the initial attack was mild,¹³⁸ and because older patients are less likely than younger patients to have recurrent disease after nonoperative management,¹⁴² surgery is often delayed or set aside in older patients. Not only is surgery more risky in the elderly, but colostomy reversal is also less often performed¹⁴³ and, when attempted, tends to have a higher rate of morbidity and mortality than in younger patients.¹⁴⁴

THE IMMUNOCOMPROMISED PATIENT

Diverticulitis can manifest more subtly in immunocompromised patients and represents a more difficult diagnostic challenge than in those with a normal immune system. Although diverticulitis in such patients does not appear to be more common, it appears to have graver consequences. One study reported that 24% of nonimmunosuppressed patients needed surgery for diverticulitis, whereas 100% of immunosuppressed ones required surgery.¹⁴⁵ Immunocompromised patients have a higher rate of free perforation (43% vs. 14%), need for surgery (58% vs. 33%), and postoperative mortality (39% vs. 2%) than do noncompromised patients.¹⁴⁶ In solid organ (e.g., heart, lung, kidney) transplant populations, mortality from diverticulitis has been found to be extremely high, ranging from 25% to 100%.^147–149 Because of this high risk, many authorities advocate elective resection after an initial episode of diverticulitis in an immunosuppressed patient.^7,125

RIGHT-SIDED DIVERTICULITIS

In Western countries, diverticulitis of the ascending colon or cecum is uncommon, because of the relatively low prevalence of diverticula in these portions of the colon. Nonetheless, right-sided diverticulitis should be part of the differential diagnosis for any patient with right-sided abdominal symptoms.

In Asia, right-sided diverticulitis is the predominant form of diverticulitis. Especially in younger patients, the diagnosis of right-sided diverticulitis is more difficult to make than is left-sided disease and is virtually indistinguishable clinically from acute appendicitis. Clinical factors that might be helpful to distinguish diverticulitis from appendicitis in a person of Asian ethnicity who might therefore have right-sided diverticulitis are that patients with diverticulitis tend to be older and have a lower frequency of nausea and vomiting than patients with appendicitis; the characteristic progression of symptoms seen with appendicitis is absent¹⁵⁰ (see Chapter 116).

Radiologically, right-sided diverticulitis and appendicitis also are easily confused, especially when a local abscess is present (Fig. 117-6). There is an estimated preoperative misdiagnosis rate in right-sided colon inflammatory conditions of 40% to 92%.^150,151 Even with excellent imaging, the diagnosis of right-sided diverticulitis often is made at laparotomy.

Figure 117-6. Computed tomography scan of a patient with a right lower quadrant abscess (arrow). The differential diagnosis of this finding includes right-sided diverticulitis and appendicitis.

When the proper diagnosis is made preoperatively, treatment of right-sided disease is the same as for left-sided diverticulitis. One study has suggested better overall responsiveness of right-sided diverticulitis to medical therapy alone, even after multiple attacks.¹⁵² The much more common complication associated with right colonic diverticula is hemorrhage, discussed later in this chapter.

SEGMENTAL COLITIS ASSOCIATED WITH DIVERTICULOSIS

A subset of patients with diverticulosis is found to have mucosal inflammation within the segment of colon containing the diverticula. Segmental colitis associated with diverticulosis (SCAD) initially was thought to be a rare form of Crohn’s colitis, but it is now increasingly recognized as a distinct, but poorly understood, manifestation of diverticular disease.^153–155 SCAD primarily affects the sigmoid colon of certain patients with diverticulosis and mimics IBD in its clinical presentation. In limited study, there does not appear to be a higher risk of diverticulitis or colon cancer in these patients.¹⁵⁵

Presenting symptoms generally are left lower quadrant cramping pain, diarrhea, and rectal bleeding. Endoscopically, in addition to diverticula, the sigmoid colonic mucosa shows erythema, friability of varying degrees, and mucosal erosions. The remaining segments of colon, including the rectum, are not visibly or histologically involved. Biopsy specimens can show chronic lymphocytic infiltration, cryptitis, crypt abscesses, and even granulomas; many cases are histologically indistinguishable from IBD.¹⁵⁶ In fact, a subset of patients when followed endoscopically through time appear to evolve into a picture similar to ulcerative proctosigmoiditis or Crohn’s colitis. This observation should prompt a low threshold for endoscopic re-evaluation if a patient with presumed SCAD develops persistent or progressive symptoms.

Patients with SCAD generally are responsive to 5-aminosalicylic acid (5-ASA) compounds, with one series reporting over 80% of patients achieving clinical remission on 5-ASA.¹⁵⁵ In most cases, the clinical course tends to be benign and self-limited,¹⁵⁷ although there are reports of cases requiring sigmoid colectomy for bleeding or stricture complications.¹⁵⁸

EPIDEMIOLOGY

Severe hemorrhage has been reported to occur in 3% to 5% of patients with diverticulosis.^20,165,166 Although most diverticula are in the left colon in Western patients, the site of bleeding diverticula has been believed to be in the proximal colon in more than one half of patients.^167–169 A large series of 180 Asian patients with diverticular hemorrhage reported a higher bleeding rate and greater need for surgery with right-sided compared with left-sided disease.¹⁶⁹ Patients with pancolonic diverticulosis appear to have a higher hemorrhage rate than those with diverticula on only one side of the colon.¹⁷⁰

PATHOPHYSIOLOGY

To study the pathogenesis of diverticular bleeding, Meyers and colleagues used sophisticated microangiographic techniques on resected colon specimens from patients with arteriographically documented bleeding diverticula.¹⁷¹ Three-dimensional histologic reconstructions demonstrated consistent findings of intimal thickening and medial thinning of the vasa recta as it coursed over the dome of the diverticulum. They proposed that these changes led to a segmental weakening of the artery, thus predisposing to its rupture. This arterial lesion was absent in diverticula that did not bleed. What predisposes to this arterial change and what precipitates its rupture are unknown. Inflammation does not appear to be a contributing factor, because it is not found histologically in bleeding diverticula that were resected. This absence might explain why bleeding rarely complicates diverticulitis.

Nonsteroidal anti-inflammatory drugs (NSAIDs) have been implicated in lower intestinal, and specifically diverticular, bleeding. A large prospective series of patients with lower intestinal bleeding reported a risk of bleeding with NSAIDs equal to that of duodenal ulcer.¹⁷² An overall increased risk of diverticular bleeding also was found at the four-year follow-up of a large study of health professionals who had been free of diverticulosis at baseline.¹⁷³ Both aspirin and nonaspirin NSAIDs increase bleeding risk, although the magnitude of this risk appears to be higher with nonaspirin NSAIDs than with aspirin.¹⁷⁰ In addition to having a higher rate of complications from diverticulosis, including bleeding and perforation, patients taking NSAIDs also appear to have more severe complications.^174–176 Hence, the necessity for NSAIDs in patients who have an episode of diverticular bleeding should be re-evaluated.

Whether patients with diverticulosis without a prior bleeding episode should be counseled as well to avoid NSAIDs is less clear. In a multicenter trial of patients randomized to naproxen (conventional NSAID) or rofecoxib (cyclooxygenase [COX]-2 selective agent), the relative risk of lower intestinal bleeding from all causes was 0.46 in patients taking rofecoxib relative to the naproxen group.¹⁷⁷ Whether hemorrhage specifically from diverticula is reduced by replacing NSAIDs with COX-2 selective agents is unknown, and the cost-effectiveness of this approach has not been established.

CLINICAL FEATURES

Diverticular hemorrhage typically manifests as abrupt, painless hematochezia. Because the bleeding is arterial, the volume of blood usually is moderate or large, an observation that can help distinguish diverticular hemorrhage from other common causes of rectal bleeding. Patients often pass red or maroon clots; melena is unusual. Because the bleeding is overt, neither a positive fecal occult blood test nor iron-deficiency anemia should be attributed to diverticular hemorrhage. Natural history studies report that bleeding ceases spontaneously in 70% to 80% of patients, and rebleeding rates range from 22% to 38%.^165,166,169 The chance of a third bleeding episode can be as high as 50%,¹⁶⁵ leading some to recommend surgical resection after a second bleeding episode.¹⁰¹

DIAGNOSIS AND TREATMENT

The diagnosis and treatment of patients with lower intestinal bleeding in general have been reviewed comprehensively elsewhere (see Chapter 19).^130,178 The algorithm in Figure 117-7 summarizes the management of patients with bleeding diverticula. In unstable patients, volume and blood product resuscitation require immediate attention. Excluding an upper gastrointestinal source by nasogastric lavage or upper endoscopy is warranted because 10% to 15% of patients with hematochezia have an upper tract cause. If bleeding is massive or if the patient remains unstable after attempted resuscitation, early angiography to attempt bleeding localization and surgical consultation should be obtained.

Figure 117-7. Algorithim for the management of patients with bleeding colonic diverticula. Angio, angiogram; EGD, esophagogastroduodenoscopy; GI, gastrointestinal; NSAIDs, nonsteroidal anti-inflammatory drugs; NG, nasogastric; RBC, red blood cell; UGI, upper gastrointestinal.

A stable patient with suspected active or recent diverticular bleeding should undergo bowel preparation for a colonoscopy. The ability to identify a diverticular source, to exclude alternative diagnoses, and to provide therapy of actively bleeding lesions support colonoscopy as a primary investigation in this setting. Rapid (over three to four hours) oral or nasogastric purge with a balanced electrolyte solution provides a safe and effective bowel preparation; a dose of metoclopramide before initiation can improve tolerance to the lavage.¹⁷⁹ If diverticulosis is not found on colonoscopy, alternative diagnoses should be entertained. If diverticula are found but bleeding has stopped and no other colonic causes are found, a presumptive diagnosis of diverticular hemorrhage is made and the patient should be instructed to avoid NSAIDs and anticoagulants, if possible. As noted, most patients with diverticular hemorrhage do not rebleed.

The endoscopic identification of active bleeding or stigmata of recent hemorrhage from a specific site (Fig. 117-8) is evident in about 10% to 20% of colonoscopic examinations for diverticular bleeding. The appearance of nonbleeding stigmata—visible vessel or adherent clot within a diverticulum—can permit one to estimate the future risk of hemorrhage from that site,¹⁸⁰ although this concept has not been validated as well in the colon as it has for peptic ulcer bleeding. Identification of a bleeding site allows endoscopic therapy to be applied. The use of epinephrine injection alone¹⁸¹ or in combination with other therapies such as heater probe coagulation,¹⁸² bipolar coagulation,^183–185 endoclips,^186–188 fibrin sealant,¹⁸⁹ and band ligation¹⁹⁰ all have been shown in small case series to achieve hemostasis safely in patients with diverticular bleeding.¹⁷⁷ If endoscopic therapy is not effective or durable, localizing the site facilitates directed therapy with angiography or segmental surgical resection.

Figure 117-8. Colonoscopy in a patient with lower intestinal bleeding from a diverticulum in whom a site of active bleeding was identified (A) and treated successfully with placement of two hemoclips (B).

(Courtesy of Janak Shah.)

With a paucity of high-quality evidence, the true effectiveness of endoscopic treatment for diverticular hemorrhage is not known. In a retrospective study, Jensen and colleagues were able to identify and treat (with epinephrine and bipolar cautery) definite bleeding sources in 10 of 48 patients with suspected diverticular bleeding.¹⁸⁵ None of the endoscopically treated patients had recurrent bleeding or required surgery. These results were compared with 17 historical controls who had bleeding stigmata but no endoscopic therapy, nine of whom rebled and six of whom required surgery, lending indirect support to the use of endoscopic therapy for diverticular hemorrhage. Stigmata of active or recent hemorrhage are not common findings at endoscopy, however, and, therefore, endoscopic therapy only is relevant to a minority of patients.

Another group at the Mayo Clinic with an approach to lower intestinal bleeding similar to Jensen’s group reported that urgent colonoscopy (performed less than 12 hours after admission) was no more effective at identifying bleeding stigmata than colonoscopy performed later in the hospitalization.¹⁹¹ Another cohort of 100 patients with acute lower intestinal bleeding was randomized to urgent colonoscopy (within eight hours of hospitalization) or standard care (radiologic bleeding studies for rapid bleeds, elective colonoscopy for slow or inactive bleeds). A definite bleeding source was found in more cases in the urgent colonoscopy group than in those who received standard care, although important outcomes such as mortality, hospital stay, rebleeding, and need for surgery were the same between groups.¹⁹² Despite having little firm evidence that expedited colonoscopy improves outcomes, it still seems reasonable to perform colonoscopy within the first 12 to 48 hours of admission in most cases.

When active bleeding is present but colonoscopy fails to allow localization or treatment of a bleeding source, further evaluation with nuclear scintigraphy (tagged red blood cell scan) or angiography can be undertaken, taking into account local availability and expertise. A third radiographic option, enhanced CT, also has shown promise for identifying active lower intestinal bleeding sources, although its role has yet to be defined.¹⁹³ Nuclear scintigraphy (Fig. 117-9) has many theoretical advantages in the evaluation of lower intestinal bleeding¹⁹⁴: It is noninvasive, technically simple, relatively inexpensive, and sensitive to bleeding rates as low as 0.1 mL/min. Scintigraphy, however, can identify only the site, not the etiology of the bleeding, and it has no therapeutic potential. Furthermore, some studies have questioned its accuracy¹⁹⁵ and have underscored a lack of proven impact on mortality, transfusion requirement, and eventual need for surgery.¹⁹⁶ Given its sensitivity and relative simplicity, however, many centers use scintigraphy before angiography to minimize the chance of a negative angiogram and to help select a specific artery for injection of contrast.¹⁹⁷

Figure 117-9. Nuclear scintigraphy scan in a patient with lower intestinal bleeding from right-sided diverticulosis. Early scan (A) shows pooling of radiolabeled blood in the right upper quadrant corresponding to a bleeding site in the hepatic flexure. Subsequent images (B, C, D) show radiolabeled blood progressing through the transverse colon over time.

Angiography has a sensitivity for lower intestinal bleeding rates of 0.5 mL/min. Advantages of angiography are identification of the site of bleeding accurately enough to direct segmental surgical resection and therapeutic capability. Although angiographic embolization had previously been felt to carry a risk of bowel infarction, superselective embolization (Fig. 117-10), or embolization of distal arterial branches, has been demonstrated to be effective (67% to 100% with lasting hemostasis) and relatively safe (ischemia rates less than 20%).^198–201 Where available, arterial embolization increasingly is becoming the nonsurgical therapy of choice when endoscopic methods are not possible.

Figure 117-10. Selective angiography of the superior mesenteric artery in a patient with lower intestinal bleeding from diverticulosis. The bleeding site is identified in the ascending colon by a “blush” of contrast material (A, arrow). The bleeding site is embolized with microcoils, which are seen in the bleeding vessel (B, arrow), and the extravasation of contrast material has stopped.

Surgery for lower intestinal bleeding usually is avoided unless endoscopic or angiographic therapies are unavailable or fail. Because diverticular bleeding stops spontaneously in most patients, surgical management is required infrequently. The primary indications for operative management are large transfusion requirements, recurrent hemorrhage that is refractory or not amenable to therapy, or hemodynamic instability unresponsive to resuscitation. When surgery is necessary, a partial colectomy is preferred to a subtotal colectomy whenever possible. Segmental resection can be performed if the bleeding site is clearly identified from a therapeutically unsuccessful angiographic or endoscopic procedure or when the extent of diverticulosis is proven to be confined to a specific segment of the colon.

The rebleeding rate was 6% in seven series of patients who underwent segmental resections for angiographically documented bleeding sites.²⁰² As noted, however, it is often difficult to identify an extravasation site angiographically. One series found that angiography resulted in a successful directed resection in only 12% of patients.²⁰³ In patients with persistent, life-threatening bleeding and no identification of a likely bleeding site, a subtotal or blind colectomy may be required as a last resort. These patients have had an extremely high morbidity and mortality,²⁰⁴ possibly because of the multiple invasive tests leading to that end and the resultant delay in definitive management. Additionally, a blind colectomy runs the risk of not resecting the bleeding lesion when the source is more proximal and has a high risk of anastomotic failure.²⁰⁵ More recent literature, however, has shown morbidity and mortality rates of a subtotal colectomy not to differ from those of a blind hemicolectomy, when the site of the bleeding is not identified.^205,206 Clearly, a close collaborative relationship between the gastroenterologist and the surgeon is paramount in managing such patients.