Incidence and risk factors for lower alimentary tract mucositis after 1529 courses of chemotherapy in a homogenous population of oncology patients

Clinical and research implications

Authors

  • Somashekar G. Krishna MD, MPH,

    1. Department of Gastroenterology and Hepatology, University of Arkansas for Medical Sciences, Little Rock, Arkansas
    Search for more papers by this author
  • Weizhi Zhao MS,

    1. Myeloma Institute for Research and Therapy, University of Arkansas for Medical Sciences, Little Rock, Arkansas
    Search for more papers by this author
  • Monica L. Grazziutti MD,

    1. Myeloma Institute for Research and Therapy, University of Arkansas for Medical Sciences, Little Rock, Arkansas
    Search for more papers by this author
  • Naveen Sanathkumar MD,

    1. Myeloma Institute for Research and Therapy, University of Arkansas for Medical Sciences, Little Rock, Arkansas
    Search for more papers by this author
  • Bart Barlogie MD, PhD,

    1. Myeloma Institute for Research and Therapy, University of Arkansas for Medical Sciences, Little Rock, Arkansas
    Search for more papers by this author
  • Elias J. Anaissie MD

    Corresponding author
    1. Myeloma Institute for Research and Therapy, University of Arkansas for Medical Sciences, Little Rock, Arkansas
    • Division of Supportive Care, Myeloma Institute for Research and Therapy, University of Arkansas for Medical Sciences, 4301 West Markham Street, Mailbox 816, Little Rock, AR 72205
    Search for more papers by this author
    • Fax: (501) 526-2273


  • This work won the award for being among the 10 best abstracts (ACG-AstraZeneca senior fellow award) and the Presidential poster award at the Annual Scientific Meeting of the American College of Gastroenterology, San Diego, California, October 23-28, 2009.

Abstract

BACKGROUND:

Lower alimentary tract mucositis is a serious complication of chemotherapy. The aim of the study was to determine the incidence, risk factors, and mortality of lower alimentary tract mucositis in a homogeneous population of patients with newly diagnosed myeloma receiving similar antineoplastic therapy and standardized supportive care.

METHODS:

Lower alimentary tract mucositis was evaluated among 303 consecutive patients with myeloma (2004-2007) enrolled in a clinical trial consisting of induction chemotherapy, tandem melphalan-based autologous stem cell transplantation (ASCT), and consolidation. Lower alimentary tract mucositis was defined as neutropenia-associated grade II-IV enteritis/colitis. Pretreatment risk factors were examined including body surface area (BSA), serum albumin (albumin), and estimated creatinine clearance (CrCl). Multiple logistic regression model was used to compute adjusted odds ratio (OR) and 95% confidence intervals (CI).

RESULTS:

Forty-seven (15.5%) patients developed lower alimentary tract mucositis during 1529 courses of chemotherapy (including 536 melphalan-based ASCT). Pre-enrollment BSA <2 m2 (OR, 2.768; 95% CI, 1.200-6.381; P = .0169) increased the risk for lower alimentary tract mucositis, whereas higher albumin was protective (OR, 0.698; 95% CI, 0.519-0.940; P = .0177). Pretransplant variables associated with lower alimentary tract mucositis were BSA <2 m2 (OR, 4.451; 95% CI, 1.459-13.58, P = .0087) and estimated CrCl <60 mL/min (OR, 3.493; 95% CI, 1.173-10.40; P = .0246). Higher albumin level conferred protection (OR, 0.500; 95% CI, 0.304-0.820; P = .0061). No lower alimentary tract mucositis-related death was observed.

CONCLUSIONS:

Lower alimentary tract mucositis is not uncommon among a homogenous population of oncology patients undergoing sequential courses of chemotherapy including melphalan-based ASCT but does not contribute to mortality. Lower BSA, renal function, and albumin are associated with increased risk for lower alimentary tract mucositis. Cancer 2011. © 2010 American Cancer Society.

Lower alimentary tract mucositis, also manifesting as neutropenic enterocolitis, typhlitis, or necrotizing enteropathy, is a common complication of cancer chemotherapy and/or radiation therapy and shares similarities with the oropharyngeal and esophageal mucosal barrier injury (mucositis/stomatitis) that develops in such settings. Lower alimentary tract mucositis is associated with increased risk for malabsorption and severe infections,1, 2 and mortality rates as high as 50% have been reported among patients with severe forms of lower alimentary tract mucositis such as neutropenic enterocolitis.3, 4

The diagnosis of neutropenic lower alimentary tract mucositis is not standardized but is usually based on a triad of fever, abdominal pain, and diarrhea, with some authors suggesting the inclusion of radiologically determined bowel wall thickness.5 In a recent systematic review of 21 studies, the authors estimated the incidence of lower alimentary tract mucositis at 5.3% (95% confidence interval [CI], 4.7%-5.9%), with a range of 0.6% to 26%, and attributed this wide variability in incidence to the lack of consensus on definition, the small and nonhomogeneous patient population studied, and low-quality methodologies. The risk factors and mortality of lower alimentary tract mucositis remain poorly defined.5

In this study, we evaluated the incidence, risk factors, and mortality of lower alimentary tract mucositis among a homogeneous and similarly treated population of 303 consecutive newly diagnosed patients with multiple myeloma undergoing 1529 courses of chemotherapy, including high-dose melphalan with autologous hematopoietic stem cell transplantation (ASCT).6 Because we have previously shown that renal dysfunction and body surface area (BSA)-based mg/kg dosing of melphalan were independent risk factors for upper alimentary tract mucositis (oral mucositis) after melphalan-based ASCT,7 we tested the hypothesis whether body measurements (such as BSA) and renal function were also risk factors for lower alimentary tract mucositis.

MATERIALS AND METHODS

The study population consisted of adults with newly diagnosed multiple myeloma cared for at the University of Arkansas for Medical Sciences between February 2004 and January 2007. All patients were enrolled in Total Therapy 3, a phase 2 prospective study consisting of 2 induction courses with multiagent chemotherapy, tandem melphalan-based ASCT, 2 consolidation courses, and maintenance (Table 1).6 Because the maintenance regimen does not cause lower alimentary tract mucositis, this treatment phase was not included in our study. Institutional review board approval was obtained for Total Therapy 3 and for the evaluation of lower alimentary tract mucositis.

Table 1. Chemotherapy Regimen (Total Therapy 3 Protocol)
InductionTransplantConsolidationMaintenance
  1. VDTPACE indicates bortezomib, dexamethasone, thalidomide, cisplatin, doxorubicin (Adriamycin), cyclophosphamide, and etoposide; TX1, transplant 1; MEL 200, melphalan 200 mg/m2 of body surface area, reduced to 140 mg/m2 for patients >70 years old or those with a creatinine level >3 mg/dL; VTD, bortezomib, thalidomide, dexamethasone; THAL + DEX, thalidomide and dexamethasone; TX2, transplant 2.

  2. Stem cell collection took place during induction. Cisplatin dose was modified for renal insufficiency. Transplants were supported by administration of >3 × 106 CD34 cells/kg of body weight.

VDTPACE1TX1, MEL 200VDTPACE3 
Year 1: VTD
THAL + DEX →THAL + DEX →THAL + DEX →
Year 2 and 3: THAL + DEX
VDTPACE2TX2, MEL 200VDTPACE4 
 
THAL + DEXTHAL + DEXTHAL + DEX 

Patients were managed according to our standards of care including antimicrobial prophylaxis (acyclovir, fluconazole, and levofloxacin) throughout the period of neutropenia, testing for circulating cytomegalovirus (CMV) by polymerase chain reaction8 and therapy with broad spectrum antibiotics (carbapenems or third/fourth generation cephalosporins) for febrile neutropenia and documented bacterial infections. With onset of significant diarrhea, ≥3 stool samples were tested for Clostridium difficile toxin,9 and computerized axial tomography of abdomen and pelvis was obtained when appropriate. A preemptive approach was taken for managing diarrhea (oral antimotility drugs, intravenous octreotide), dehydration, electrolyte imbalances and deconditioning with early initiation of physical therapy. Nasogastric suctioning and total parenteral nutrition were rarely used. Prophylactic acid suppression therapy was provided during all courses of chemotherapy (histamine 2-receptor antagonists during induction and consolidation; proton pump inhibitors during melphalan-based ASCT). All patients were cared for by clinicians with training and experience in cancer supportive care.

The evaluation period started with the first day of chemotherapy for a total of 30 days during the induction and consolidation courses and for 62 days for melphalan-based ASCT (60 days after the infusion of stem cells).6 The last patient follow-up date was January 31, 2007. Evaluation included daily examination with protocol-defined laboratory studies and other tests as clinically indicated. Adverse events were prospectively collected and graded according to the Common Terminology Criteria for Adverse Events version 3.0 of the National Cancer Institute.10

Lower alimentary tract mucositis was defined as colitis and/or enteritis (grades II-IV) that occurred during chemotherapy-induced neutropenia and in the absence of C. difficile infection. Grade II consisted of abdominal pain and mucus/blood in stool; grade III was defined as abdominal pain, fever, and change in bowel habits with ileus or presence of peritoneal signs, whereas grade IV referred to the development of life-threatening consequences (eg, perforation, bleeding, ischemia, necrosis, or toxic megacolon).10 Neutropenia was defined as absolute neutrophil count <1.0 × 109/L. The control group consisted of patients who did not develop grade II to IV lower alimentary tract mucositis or C. difficile infection.

Mortality related to lower alimentary tract mucositis was defined as any death that occurred during the evaluation period for each course of therapy unless it was caused by progressive myeloma or another well-documented etiology that was unrelated to lower alimentary tract mucositis.

Potential covariates were examined at enrollment and before each course of chemotherapy and included demographics, BSA, serum albumin, total bilirubin, estimated creatinine clearance (CrCl), CMV immunoglobulin serostatus, and others. In addition, course-specific variables included prior episode of lower alimentary tract mucositis and mg/kg of melphalan given for melphalan-based ASCT. The dose of all chemotherapeutic agents was calculated according to the Mosteller formula of BSA except for thalidomide and dexamethasone, which were given at a fixed dose.7

Statistical Analysis

Data were presented as numbers and percentages for categorical variables and as means ± standard deviation, median, and range (minimum, maximum) for continuous variables. All statistical analyses were performed using the SAS statistical package, version 9.1 (SAS Institute, Cary, NC). Race, sex, CMV immunoglobulin serostatus, and prior lower alimentary tract mucositis were treated as categorical variables. Fisher exact test was used to compare the differences between patients with lower alimentary tract mucositis and controls. Serum albumin, estimated CrCl, total bilirubin, mg/kg melphalan dose, and BSA were treated as continuous variables, and the Wilcoxon rank sum test was applied because of the skewed distribution of these measurements. BSA was also examined as a grouped variable using the cutoff point of <2.0 m2 (derived from the highest tertile of BSA at enrollment among the control population). The standard cutoff point of <60 mL/min was used for estimated CrCl11 to further examine the effect of renal insufficiency on lower alimentary tract mucositis. Multiple logistic regression was used to examine the association between pretreatment variables and occurrence of lower alimentary tract mucositis, and adjusted odds ratios (OR) and 95% CIs were estimated. Both univariate and multivariate analyses were performed at each treatment course (course-specific analysis) and for all courses combined (overall analysis). The linearity of the logit assumption was checked for continuous variables. The significant level chosen was 0.05.

RESULTS

All 303 consecutive patients with newly diagnosed myeloma enrolled in Total Therapy 3 were evaluated. Table 2 details the study flowchart including patient withdrawal from protocol; 291 patients completed both induction courses, and 285 (94%) and 251 (83%) completed the first and second melphalan-based ASCT, respectively. In total, 1529 courses of chemotherapy (including 536 melphalan-based ASCT) were given during the study period. The median age of the study population was 59 years (range, 32-75 years), and 193 (64%) were men. At study entry, the median BSA and serum albumin were 1.88 m2 and 4.0 g/dL, respectively; 49 patients (16%) had estimated CrCl <60 mL/min (Table 3).

Table 2. Patient Flowchart Including Withdrawals From Study Protocol
CycleTotalWithdrawalsWithdrawal Description
  • Administrative includes social issues, insurance coverage denial, others.

  • a

    Did not reach treatment phase by the last follow-up date of the lower alimentary tract mucositis study (January 31, 2007).

Induction 1303129 taken off protocol (7 administrative, 2 complications), 2 deaths, 1 skipped the second induction cycle because of disease progression and proceeded to transplant.
Induction 229177 taken off protocol (4 administrative, 1 patient choice, 1 progressive disease and toxicity, 1 toxicity).
Transplant 12853412 did receive transplant 2 and proceeded to consolidation (9 administrative, 3 toxicity), 21 taken off protocol (11 deaths, 4 administrative, 6 toxicity), 1 did not reach treatment phase by the last day of follow-up on the study.a
Transplant 2251486 did not receive consolidation (various causes) and proceeded to maintenance, 17 taken off protocol (3 deaths, 7 administrative, 3 patient choice, 2 disease progression, and 2 toxicity). 25 did not reach treatment phase by the last day of follow-up on the study.a
Consolidation 1215316 did not receive second consolidation, 6 taken off protocol (4 administrative, 2 disease progression), 19 did not reach treatment phase by the last day of study follow-up.a
Consolidation 2184  
Table 3. Pretreatment Characteristics of 303 Myeloma Patients
CharacteristicAll Patients, n=303Control Group,a n=213L-ATM, n=47Pb
  • L-ATM indicates lower alimentary tract mucositis; Ig, immunoglobulin; CMV, cytomegalovirus.

  • a

    The control group included patients who developed neither L-ATM (n = 47) nor Clostridium difficile infection (n = 43).

  • b

    P value comparing L-ATM versus the control group. Fisher exact test was used for categorical variables, and the Wilcoxon rank sum test was used for continuous variables.

  • c

    Inclusive of IgD, nonsecretory and light chain multiple myeloma.

  • d

    Body surface area cutoff value of 2 m2 is based on the highest tertile of the control group.

Age, y    
 Median596059.9966
 Range32-7532-7536-73 
Caucasian race, No. (%)269 (88.8)188 (88.3)45 (95.7).1855
Male sex, No. (%)193 (63.7)143 (67.1)21 (44.7).0047
Myeloma isotype    
 IgA, No. (%)71 (23.4)57 (26.8)9 (19.2).4696
 IgG, No. (%)174 (57.4)116 (54.5)30 (63.8) 
 Others,c No. (%)58 (19.1)40 (18.8)8 (17.0) 
 CMV IgG positive, No. (%)190 (62.7)139 (65.3)25 (53.2).1345
Body surface aread
 Median, m21.881.91.80.0253
 Range, m21.41-2.561.41-2.561.48-2.21 
 <2.0 m2201 (66.3)132 (62.0)39 (83.0).0063
 ≥2.0 m2102 (33.7)81 (38.0)8 (17.0) 
Serum albumin    
 Median, g/dL443.9.0207
 Range, g/dL2.1-5.32.5-5.32.1-4.9 
Estimated creatinine clearance
 Median, mL/min93.895.287.7.1463
 Range, mL/min9.0-245.79.0-245.79.9-172.4 
 <60 mL/min49 (16.2)31 (14.6)13 (27.7).0505
 ≥60 mL/min254 (83.8)182 (85.5)34 (72.3) 
Serum total bilirubin    
 Median, mg/dL0.50.50.4.2066
 Range, mg/dL0.1-1.80.1-1.80.2-1.0 

Overall, lower alimentary tract mucositis developed in 47 (15.5%) of 303 patients, including 6 who developed another episode of lower alimentary tract mucositis during subsequent courses. The rate of lower alimentary tract mucositis was higher after melphalan-based ASCT (17 of 285 patients with transplant 1 and 26 of 251 with transplant 2; 34 [13.5 %] of 251 for patients who underwent both transplants) than during the induction (9 of 303, 3.0%) or consolidation phases (7 of 215, 3.3%). Twenty-five of 47 lower alimentary tract mucositis patients were CMV seropositive, 4 of whom developed CMV reactivation (5 episodes) after melphalan-based ASCT. Endoscopic documentation of CMV enterocolitis was not obtained because of the risks involved. Management of lower alimentary tract mucositis was conservative, and no patient required surgical intervention.

Only 2 patients died after a course of therapy during which lower alimentary tract mucositis developed; 1 died 85 days after the second transplant because of rapidly progressive myeloma involving the central nervous system; the other died 54 days after her first transplant. This patient had complete resolution of lower alimentary tract mucositis-related clinical and radiological findings, achieved hematopoietic recovery, and was discharged from the hospital. She subsequently developed fatal CMV reactivation with pneumonia and multiorgan failure.

Univariate analysis indicated that patients with lower alimentary tract mucositis were more likely to have lower BSA (<2 m2) at enrollment (83% vs 62%; P = .0063), lower estimated CrCl (<60 mL/min) (27.7% vs 14.6%, P = .0505), and lower serum albumin (median of 3.9 vs 4.0 g/dL, P = .0207) than controls (Table 3). The application of multiple logistic regression analyses (Table 4) throughout treatment courses revealed that lower pre-enrollment BSA (<2.0 m2) increased the risk for lower alimentary tract mucositis (OR, 2.768; 95% CI, 1.200-6.381; P = .0169), whereas higher serum albumin was protective (30% estimated reduction in the risk of lower alimentary tract mucositis with each 0.5 g/dL increment in serum albumin [OR, 0.698; 95% CI, 0.519-0.940; P = .0177]). Higher serum albumin before induction 2 was also associated with lower risk of lower alimentary tract mucositis during this course (OR, 0.347; 95% CI, 0.129-0.934; P = .0362). Multivariate analysis of risk factors for the tandem transplant phase identified a strong association between lower alimentary tract mucositis and patients with BSA <2 m2 (4.451; 95% CI, 1.459-13.58; P = .0087) and lower estimated CrCl (<60 mL/min) (OR, 3.493; 95% CI, 1.173-10.40; P = .0246). Higher pretransplant serum albumin level conferred protection (OR, 0.500; 95% CI, 0.304-0.820; P = .0061).

Table 4. Overall and Course-Specific Risk Factors for L-ATM Versus Control Group (no L-ATM or Clostridium difficile Infectiona): Multivariate Analysis
Treatment CoursePretreatment VariablesControl Group, No. (%)L-ATM Group, No. (%)OR (95% CI)P
  • L-ATM indicates lower alimentary tract mucositis; OR, odds ratio; CI, confidence interval; ALB, serum albumin; BSA, body surface area; e-CrCl, estimated creatinine clearance; TX1, transplant 1; TX, transplant 1+2.

  • a

    Patients who developed Clostridium difficile infection were excluded from analysis (43 from overall analysis, 9 from inductions 1 and 2, respectively, 13 from transplant 1, 10 from transplant 2, and 11 from consolidation).

  • b

    OR and 95% CI were estimated per 0.5 U increment of serum albumin.

  • c

    Induction 1 is not included because only 4 patients developed L-ATM.

  • d

    One patient who developed L-ATM in induction 1 but Clostridium difficile infection in induction 2 was excluded.

  • e

    Combined analysis of transplants 1 and 2 for those 251 patients who underwent both courses; comparable findings were obtained when the transplant courses for the 285 patients who underwent 1 or both transplants were examined.

  • f

    Analysis of consolidation 1 (n = 215) and 2 (n = 184) courses is combined because of the small number of patients with L-ATM during consolidation 1 (2 patients) and consolidation 2 (5 patients).

Baseline, n = 303 21347  
 ALB, g/dLb4.00 ± 0.513.78 ± 0.600.698 (0.519-0.940).0177
 BSA <2.0 m2132 (61.97)39 (82.98)2.768 (1.200-6.381).0169
 e-CrCl <60 mL/min31 (14.55)13 (27.66)1.603 (0.732-3.512).2380
Induction 1, n = 303c 2904  
Induction 2, n = 291 2766  
 L-ATM in induction 12 (0.72)0 (0.00)No computable estimate
 ALB, g/dL b3.69 ± 0.413.30 ± 0.320.347 (0.129-0.934).0362
 BSA <2.0 m2186 (67.39)3 (50.00)0.567 (0.105-3.073).5106
 e-CrCl <60 mL/min34 (12.32)1 (16.67)1.469 (0.153-14.09).7390
Induction 1 + 2, n = 303 2789d  
 ALB, g/dLb3.97 ± 0.543.63 ± 0.530.582 (0.325-1.043).0691
 BSA <2.0 m2185 (66.55)5 (55.56)0.706 (0.175-2.838).6236
 e-CrCl <60 mL/min47 (16.91)1 (11.11)0.553 (0.060-5.090).6007
Transplant 1, n = 285 25517  
 L-ATM in induction 1+26 (2.35)1 (5.88)4.511 (0.453-44.93).1989
 ALB, g/dLb3.860 ± 4.113.553 ± 0.3160.467 (0.251-0.870).0165
 BSA <2.0 m2165 (64.71)15 (88.24)4.035 (0.853-19.08).0785
 e-CrCl <60 mL/min21 (8.24)6 (35.29)3.740 (1.110-12.61).0334
Transplant 2, n = 251 21526  
 L-ATM in induction + TX112 (5.58)5 (19.23)3.208 (0.967-10.64).0567
 ALB, g/dLb3.865 ± 4.273.696 ± 0.4450.748 (0.470-1.190).2205
 BSA <2.0 m2139 (64.65)24 (92.31)5.903 (1.339-26.01).0190
 e-CrCl <60 mL/min19 (8.84)5 (19.23)1.583 (0.510-4.918).4271
Transplants 1+2, n = 251e 19834  
 L-ATM in induction3 (1.52)1 (2.94)2.606 (0.220-30.90).4478
 ALB, g/dLb3.890 ± 4.1936.53 ± 0.2980.500 (0.304-0.820).0061
 BSA <2.0 m2125 (63.13)30 (88.24)4.451 (1.459-13.58).0087
 e-CrCl <60 mL/min11 (5.56)8 (23.53)3.493 (1.173-10.40).0246
Consolidation, n = 215f 1977  
 L-ATM in induction + TX21 (10.66)1 (14.29)1.061 (0.108-10.37).9597
 ALB, g/dLb3.81 ± 0.383.79 ± 0.410.989 (0.347-2.817).9835
 BSA <2.0 m2130 (65.99)7 (100.0)No computable estimate
 e-CrCl <60 mL/min15 (7.61)2 (28.57)4.804 (0.830-27.82).0799

DISCUSSION

This is the first study to evaluate the incidence, mortality, and pretreatment risk factors of lower alimentary tract mucositis in patients undergoing multiple courses of therapy including high-dose chemotherapy and autologous transplantation. The large sample size (303 patients, 1529 courses of therapy) and patient homogeneity (newly diagnosed patients with myeloma cared for at a single center and receiving uniform antineoplastic and supportive care therapies) permit an accurate evaluation of the characteristics of neutropenia-related lower alimentary tract mucositis.

Several novel findings emerged from this study: first that lower alimentary tract mucositis developed in 15.5% of such treated patients; second that lower alimentary tract mucositis was managed medically without attributable mortality; and third that lower pretreatment BSA, renal function, and serum albumin level predisposed patients to this complication. These risk factors were distinctly different from those that increased the risk for C. difficile infection in the same study population (data not shown).

Upper and lower alimentary tract mucositis represents treatment-induced mucosal barrier injury, and they share the same pathophysiology, although lower alimentary tract mucositis is the most clinically serious form of mucosal barrier injury.12-14 To a great extent, mucosal barrier injury results from increased drug exposure, that is, higher drug doses and/or slower drug clearance.7, 15, 16 We have previously shown that BSA-based dosing of high-dose melphalan results in wide variations in the mg/kg melphalan dose received, with significantly higher doses given to patients with lower BSA, and that the latter patients have a significantly higher risk of upper alimentary tract mucositis.7 The wide variations in mg/kg of melphalan with BSA dosing was confirmed in a study in which a 4.3-fold variation in melphalan drug exposure (determined by area under the curve values) was shown among melphalan-based ASCT patients given similar BSA-based melphalan/m2 doses.15 That higher melphalan mg/kg doses increase the risk of upper alimentary tract mucositis7 was further confirmed among patients with myeloma and lymphoma undergoing ASCT17, 18 and among allogeneic stem cell transplantation recipients with various hematological diseases.19

We have also previously shown that decreased renal function was an independent predictor of upper alimentary tract mucositis after melphalan-based ASCT.7 This study extends these findings to lower alimentary tract mucositis after melphalan-based ASCT, in keeping with the known renal excretion and pharmacokinetics of intravenous melphalan.15, 16, 20-23

We also identified lower serum albumin as a risk factor for lower alimentary tract mucositis. Serum albumin is an expression of liver function, and albumin can bind many drugs.24 Hence, patients with hypoalbuminemia may have increased drug exposure (hence higher toxicity) when treated with agents that undergo extensive hepatic metabolism and/or exhibit high protein binding.25 Most drugs applied during the induction phase of Total Therapy 3 (Table 1) undergo extensive hepatic metabolism (doxorubicin [Adriamycin], etoposide, cyclophosphamide, and bortezomib) and/or exhibit high protein binding (same agents plus cisplatin).26-30 Because intravenous melphalan is highly protein-bound (60%-90%, mainly albumin),20 hypoalbuminemia may also result in higher free melphalan exposure, which in turn may increase the risk for lower alimentary tract mucositis.

Only 1 study evaluated the risk factors for lower alimentary tract mucositis after antineoplastic chemotherapy. In that study, 83 (2.6%) of 3171 children with various malignancies who were treated between 1990 and 2001 developed typhlitis. Age >16 years was the only risk factor identified (P = .03).31 However, baseline body measurements, renal function, and serum albumin were not examined.

The 15.5% incidence of lower alimentary tract mucositis in our study is higher than the pooled incidence of 5.3% (range, 0.8%-26%) reported in a recent systematic review of adult cancer patients undergoing chemotherapy.5 However, this review included studies published over several decades (1963-2004), with heterogeneous populations (various cancers and therapies), different definitions for lower alimentary tract mucositis, and variable application of diagnostic tools. Furthermore, only 290 patients had undergone ASCT, and the period at risk for lower alimentary tract mucositis was unclear, with some studies reporting only 1 course of therapy, in contrast to the dose dense/dose intense therapy applied in Total Therapy 3, with 1529 courses given to 303 patients over a relatively short period of time (Table 1). Differences in definitions could account for some of the variations in the incidence of lower alimentary tract mucositis, which is typically defined as a triad of fever, abdominal pain, and diarrhea.5 Increased bowel wall thickness by ultrasound or computed tomography scan has been recently suggested as an additional diagnostic criterion.5, 32-34 However, monitoring bowel wall thickness has not been validated, and routine application of radiological imaging in all patients undergoing chemotherapy is expensive. Furthermore, increased bowel wall thickness can be observed in other conditions that affect cancer patients, including graft versus host disease and infections.3, 34, 35 Nevertheless, radiological imaging should be used adjunctively when there is a clinical suspicion of complication related to chemotherapy, such as typhlitis, cholecystitis, pancreatitis, intra-abdominal abscess, intestinal obstruction, or perforation, especially because classical clinical symptoms and signs might be absent in neutropenic patients.31, 36, 37

Our findings that lower alimentary tract mucositis did not cause death or require surgical intervention stand in sharp contrast to the recent reports of 30% to 50% mortality.38-42 This difference can be partly explained by the different definitions for lower alimentary tract mucositis, with some reports focusing on the most severe manifestations of lower alimentary tract mucositis only (neutropenic enterocolitis, typhlitis, or necrotizing enteropathy).5

Although all lower alimentary tract mucositis-related data were prospectively collected, our study suffers from the finding that lower alimentary tract mucositis was not the primary endpoint for Total Therapy 3 (which was myeloma outcomes), and hence additional studies focused on lower alimentary tract mucositis (such as measuring bowel wall thickness) were not performed.

Because lower pretreatment values for BSA, serum albumin, and estimated CrCl increase the risk for lower alimentary tract mucositis in patients undergoing melphalan-based ASCT, these variables should be taken into consideration when planning the dose of melphalan from a standard 200 mg/m2 to 100 mg/m2 or 140 mg/m2 of BSA, depending on the presence of 1 or more of these risk factors. Low serum albumin was the only risk factor for lower alimentary tract mucositis during the induction phase; consequently, a dose reduction of the agents used during induction should also be considered in the setting of hypoalbuminemia. Our findings also suggest that additional pharmacokinetic and pharmacogenomic studies are needed to better understand the variability of melphalan exposure and toxicity.43

Because none of our patients required surgical intervention, we recommend supportive care as the mainstay of management while reserving surgical evaluation to patients with bowel perforation or uncontrolled gastrointestinal bleeding.5, 32, 44-46 Because there is no uniform definition and grading for neutropenia-related lower alimentary tract mucositis, and because the Common Terminology Criteria for Adverse Events toxicity grading is clinician-based and remains the principal clinical and research system for reporting adverse events in oncology clinical trials,47 we recommend the application of the Common Terminology Criteria for Adverse Events grading for enteritis and/or colitis in future reports of lower alimentary tract mucositis.10

We conclude that lower alimentary tract mucositis is not an uncommon event after sequential courses of chemotherapy including ASCT but does not appear to contribute to mortality. The evaluation of pretreatment risk factors including BSA, renal function, and serum albumin level may identify patients at higher risk for lower alimentary tract mucositis, who may benefit from chemotherapy dose modification and/or the application of novel mucoprotective strategies.

CONFLICT OF INTEREST DISCLOSURES

The authors made no disclosures.

Ancillary