Colorectal carcinoma in young females.

Background: We sought to study the clinicopathologic characteristics of colorectal cancer in young female patients. We also wanted to determine the association of colorectal cancer with anemia in these female patients and, finally, to determine the effect of gender on prognosis in young patients with colorectal cancer.

Methods: We performed a retrospective analysis of all young patients diagnosed with colorectal cancer between 1982 and 1999 in two teaching hospitals in New York City.

Results: A total of 3,546 cases of colorectal cancer were diagnosed. Sixty-one (1.63%) of these patients were young patients and 32 (0.85%) were female. Young refers to all patients in the study who were younger than 40 years of age. The clinical presentation and mean age at presentation were very similar in both male and female patients. At presentation, 87.5% of female patients had anemia compared with only 69% of male patients. Males had a statistically significant higher mean hemoglobin level compared with females (12.87 versus 10.29 g) at P = 0.0001. Seventy-nine percent of female patients compared with 86% of male patients presented with left-sided tumors. Fifty-five percent of males presented with late stage disease compared with 68% of females (P = 0.27). Female sex seemed to adversely affect the prognosis, although this did not reach statistical significance (P = 0.08). Stage of disease was associated with worse prognosis and this was independent of sex. Age and hemoglobin were not independent predictors of mortality.

Conclusion: Colorectal cancer does occur in females of childbearing age who might have a tendency to present with late stage disease as evidence from this study. Young female patients with anemia should be questioned about gastrointestinal symptoms, and colorectal cancer should definitely be in the differential diagnoses. This might conceivably allow for earlier diagnosis and potential for cure in this patient group.

**********

Colorectal cancer remains a commonly diagnosed malignancy in the United States, and there are more than 50,000 deaths annually as a result of this malignancy. (1) It is generally considered a disease of older people, and the majority of cases are diagnosed in patients older than 55 years of age. (2) Therefore, it is not surprising that data regarding the clinical presentation and prognosis of colorectal cancer in young patients are scant. (3,4) The available data have been conflicting, but there are a few studies that suggest that the prognosis might be worse in this group of patients. (5,6)

Our objective was to study the clinicopathologic and prognostic characteristics of young female patients with colorectal cancer as compared with young male patients. We explored the association of colorectal cancer with anemia in these young female patients. We wanted to use this study to focus on how much family history and genetic factors might contribute to the development of cancer in this group of patients. In addition, we wanted to determine whether female sex might be more predictive of mortality when compared with male sex, after controlling for the stage of cancer in these young patients.

Methods

Patients

This is a descriptive retrospective cohort study of colorectal cancer in young females from two urban teaching hospitals in New York City. Our definition of young for this study was patients younger than 40 years. Anemia was defined as hemoglobin below 12 g for females and 13.5 g for males. We identified all young patients who were diagnosed with primary colorectal cancer at St. Vincent's Hospital Medical Centers, which consists of two different institutions. The first institution was St. Vincent's Hospital where data were collected on all young patients diagnosed with colorectal cancer between 1982 and 1999. The other hospital was the Catholic Medical Center where data was collected on all young patients diagnosed between 1986 and 1999. The minimum follow-up period was 2 years. Data collection was achieved using tumor registry, hospital medical medical records, and physician office records. The data extracted included age, sex, race, hemoglobin at presentation, family history, tumor location, stage, and histology. Data was also collected on the patient's history of familial adenomatous polyposis (FAP), hereditary nonpolyposis colorectal cancer (HNPCC), and inflammatory bowel disease (IBD).

Tumor locations were classified according to ICD-9. The right side of the colon consisted of cecum, ascending colon, hepatic flexure, and transverse colon. The left colon consisted of splenic flexure, descending colon, sigmoid colon, rectosigmoid junction, and the rectal ampulla. Appendiceal and anal cancer cases were excluded. Staging was done according to the American Joint Committee on Cancer (AJCC) tumor, node, and metastasis staging for colorectal cancer. All patients were assigned to stages according to the AJCC. Patients with Stages 0 to 2 disease were considered to have early disease, whereas patients with Stage 3 or 4 disease were considered to have late disease.

Histologic grading was recorded as well differentiated, moderately differentiated, or poorly differentiated. Also, it was noted whether tumors were mucin producing or not. Pathologists familiar with the histologic characteristics of colorectal cancer assessed the tumors. Operative findings were reviewed in all of the patients. Since the early 1990s, adjuvant chemotherapy has been recommended for those with cancer stage IIB or higher disease. In the majority of our patients, adjuvant chemotherapy was in the form of 5-fluorouracil (5-FU) and levamisole. Most patients with rectal lesions also had adjuvant radiotherapy. Treatment was categorized as curative surgery, radiotherapy, and chemotherapy. Prognosis was defined as either alive or dead at the end of the follow-up period.

Statistics

All categorical variables were expressed as proportions, and the continuous variables were expressed as means. The [chi square] statistic was calculated as a measure of association between categorical variables while a two-sample t test was used to determine differences between two proportions of means. Significance level was set at P < 0.05 for all analysis. Significance level for comparisons was set at P < 0.05.

To determine the independent role of sex in predicting mortality, logistic regression models were constructed with mortality as the dependent variable. The independent variables were sex, age, hemoglobin, and stage of disease. Initially, bivariate analysis was performed with each variable and then loaded into a multivariate logistic regression model. To determine the predictive power of each variable in the multivariate model, each was removed one at a time with a backward selection process. Significance level was set at P < 0.05.

Sex was entered into this model as a dichotomous variable, and age and hemoglobin were entered as continuous variables. Stage of disease was entered as a categorical variable. All analysis was conducted with Minitab software (release 13.1 for Windows 98; Minitab, Inc., State College, PA).

Results

Young patients accounted for 61 (1.63%) of 3,546 of all colorectal cancers diagnosed. Thirty-two (0.85%) of all patients were young females while 29 were young males. Table 1 shows the demographic characteristics of the 61 patients followed in this study. The mean age at presentation was similar for both young males and females, and the age range for all patients was 22 to 39 years. More (87.5%) females presented with anemia in contrast to 69% of male patients. Young males had a statistically significant higher mean hemoglobin level compared with young females.

Clinical presentation and tumor characteristics are illustrated in Table 2. Aside from anemia, the clinical presentation was very similar in both males and females. The most common clinical complaints in both groups were abdominal pain, rectal bleeding, and change in bowel habit. Only four patients (all female) had a predisposing factor that could have led to development of colorectal cancer. One patient had a personal and family history of FAP and was diagnosed with cancer at age 23; another had Crohn's disease from which she had suffered for 12 years; and the remaining two had an aunt and mother who developed colorectal cancer after the age of 50 which really placed these patients at only slightly increased risk of developing cancer as compared with the regular population. There were no documented cases of HNPCC by history. Most of the tumors were left-sided tumors in both males and females. Sixty-eight percent of the female patients and 55% of the male patients presented with late stage disease. Most of the tumors in both groups were either poorly differentiated or moderately differentiated, and seven females and four males had mucin-producing tumors. The treatment modality is illustrated in Table 3. At the end of the 2-year follow-up period, 9 females compared with 11 males were alive without disease while 6 females and 3 males were alive with disease. In this study, the only variable predictive of mortality in this group of patients was the stage of disease. Using bivariate and multivariate regression analysis, patients with late stage disease were more likely to die compared with those with early stage disease (odds ratio [OR] = 27.1, 95% confidence interval [CI] = 3.32-222.73, P = 0.002). Table 4 illustrates that sex, age, and hemoglobin were not independent predictors of mortality. When all variables were loaded in the multivariate model, only stage of disease remained a statistically significant predictor of mortality (OR = 64.1, CI = 4.79-826.3, P = 0.01). In this multivariate model, there was, however, a trend toward a worse prognosis for young women compared with men (OR = 6.19, CI = 0.78-49.02, P = 0.08).

Discussion

This study demonstrates that colorectal cancer is uncommon in young females. It constitutes almost 1% of all colorectal cancers in this series. This is similar to some of the other studies in medical literature. (7,8) When we compared our young female patients with the young male patients, we observed no significant differences in presenting symptoms, tumor site, and tumor differentiation. However, more of our young female patients presented with anemia. Iron deficiency is very common in young females and this is usually attributed to menstruation or the burden of pregnancy and childbirth. A study by Bini et al (9) of 186 premenopausal patients with iron deficiency anemia showed that 6 (3%) of these patients were diagnosed with colorectal cancer after endoscopic workup. Iron deficiency anemia has been shown to affect 9 to 11% of American female adolescents and women of childbearing age. (10) We think that further studies are needed to determine the best diagnostic and cost-effective approach for young female patients who present with anemia.

Only four of our patients (all female) had a reason that might have predisposed them to colorectal cancer. We did not document a history of HNPCC in any of our patients. This is in contrast to a study of young patients from Italy by Fante et al (11) in which 38.4% of all their young patients with colorectal cancer had either FAP or HNPCC. We are of the opinion that the reason why Fante et al had more patients with HNPCC and FAP is a result of the specialized nature of their cancer-screening registry, which allowed them to preferably accumulate patients with familial cancer syndromes. We think that our study is more representative of the population. The classic Amsterdam criteria (12) were the first developed to support the diagnosis of HNPCC. These criteria were found to be quite restrictive but, if present, the chance of HNPCC is quite high. The Amsterdam criteria consist of the following:

1. At least three relatives with histologically verified colorectal cancer, including a first-degree relative of the other two patients

2. At least two successive generations affected

3. In at least one of the above individuals, diagnosis of colorectal cancer before age 50

The stringency of the Amsterdam criteria has prompted some authors and experts to recommend the Bethesda criteria, which are somewhat more sensitive but less specific. (13) The Bethesda criteria include the following:

1. Individuals who meet the Amsterdam criteria

2. Individuals with two HNPCC-related cancers, including synchronous and metachronous colorectal cancers or associated extra-colonic cancers (endometrial, ovarian, gastric, hepatobiliary, small bowel, or transitional cell cancer of the renal pelvis or ureter)

3. Individuals with colorectal cancer and a first-degree relative with colorectal cancer and/or HNPCC-related extracolonic cancer and/or a colorectal adenoma. One of the cancers must have been diagnosed before age 45 and then adenoma at age less than 40 years

4. Individuals with colorectal cancer or endometrial cancer diagnosed before age 45

5. Individuals with right-sided colon cancer with an undifferentiated pattern (solid/cribriform) diagnosed before age 45

6. Individuals with signet ring cell type colorectal cancer (>50% signet ring cells) diagnosed before age 45

7. Individuals with colorectal cancer diagnosed before age 40

Testing for microsatellite instability in young patients with colorectal cancer might be a better way to screen for HNPCC. This makes sense because it is well documented that over 85% of HNPCC tumors demonstrate microsatellite instability. (14) A study by Gyrfe et al (15) of patients younger than 50 years of age demonstrated high-frequency microsatellite instability in 17% of their patients. Therefore, microsatellite instability could be performed on the tumors in these young patients as a screening test before more intensive DNA analysis is done to detect mutation in the HNPCC genes. Another approach that might be less expensive is to use immunochemistry to test for MSHLI and MSHL2 before specific genetic testing is performed. Another reason why we should probably be testing for microsatellite instability is recent studies that suggest that the presence or absence of microsatellite instability might have prognostic implications and might thereby affect the approach to patient management. (16,17)

This study again demonstrates that colorectal cancer can occur in pregnancy. Bernstein et al (18) demonstrated that no cases lower than cancer stage II were diagnosed in pregnant patients; this correlates with the poor survival rates documented in this group of patients. Many authors have attributed the delay in diagnosis in these patients to physicians attributing the signs and symptoms of colorectal cancer to those of pregnancy. (19,20) It is not clear whether hormonal effects might also be involved in the rapid spread of tumors in these patients. However, colorectal cancer does not seem to adversely affect gestation. Our two patients had normal live births. Woods et al (21) demonstrated that 25 (78%) of the 32 patients in their series had normal live births. Deaths were due to termination, intrauterine deaths, prematurity, and stillbirths. Our two patients were unfortunately diagnosed during the peripartum period with Stage 4 disease.

It is important to note that recent studies from developed countries suggest that young patients with colorectal cancer might have similar or even better prognosis than older patients. (22,23) This trend might be a reflection of advances in technology and improvements in diagnostic and chemotherapeutic regimen. The improved prognosis in young patients might also be a reflection of lack of significant comorbidities and the ability of younger patients to better tolerate an aggressive surgical and chemoradiotherapy regimen. Before assuming any of the above, however, we would like to point out that incidence, stage of presentation, and mortality have been documented to be much higher in the young compared with older patients in developing countries. (24,25) These differences between developed and developing countries might be an indication that various environmental and genetic factors might vary in their effects on the development, biology, and behavior of colorectal cancer in different population groups. The female patients in our study had a tendency to present with more advanced disease, which translated to a poorer prognosis compared with young male patients. This did not reach statistical significance and we cannot say if this is because of the relatively small number of patients in this study. We think, however, that contributory factors to worse prognosis in young females might include the burden of anemia and pregnancy, which could conceivably result in a delay in diagnosis.

One of the major limitations of this study is the retrospective nature. However, we can conclude that colorectal cancer is not an exceptional rarity in young females. The most common clinical presentation is anemia, and a high index of suspicion is needed in these patients because of other confounding factors such as menstruation and pregnancy. Premenstrual females, including pregnant patients who present with anemia coupled with symptoms like rectal bleeding, abdominal pain, and change in bowel habit, should be referred early to gastroenterologists for further workup to rule out potential colorectal cancer.

Table 1. Demographic characteristics (n = 61)

Characteristics  Males  Females  P value

Age (yrs)        34.9     34.7     0.8
Race (%)
  Whites         28       31        *
  Blacks         48       37        *
  Others         24       32        *

*[chi square] analysis of association between race and sex was not
significant.

Table 2. Clinical, staging, and pathologic characteristics of
patients (a)

                             Males  Females  P value

Clinical symptoms
  Abdominal pain             10       12     NS
  Rectal bleeding            11       12     NS
  Change in bowel habit       8        5     NS
  Nausea/vomiting             2        3     NS
  Weight loss                 4        2     NS
  Pregnant                    0        2     NS
Stage (%)
  Late                       55       68     0.27
  Anemia (%)                 69%      87.5%
  Mean hemoglobin            12.87    10.29  0.0001
Pathologic characteristics
  Left-sided tumors (%)      86       79     NS
  Poorly differentiated       9        6     NS
  Moderately differentiated  11       14     NS
  Well differentiated         5        6     NS
  Undifferentiated            3        3     NS
  Unknown                     1        3     NS
  Mucinous tumors             4        7     NS

(a) NS, not significant.

Table 3. Treatment modality

Treatment                  Males      Females

Curative surgery (%)       23 (79%)   22 (69%)
Radiation and surgery       9 (31%)    8 (25%)
Chemotherapy and surgery   13 (45%)   15 (47%)

Table 4. Multivariate odds ratio for mortality by stage, sex, age, and
hemoglobin level

                            Odds      95% CI
Variable    [beta]    P     ratio  Lower  Upper

Stage (a)
  Late       4.16    0.001  64.10   4.97  826.3
Sex (b)
  Female     1.82    0.08    6.19   0.78   49.02
Age         -0.15    0.08    0.85   0.71    1.02
Hemoglobin   0.29    0.25    1.35   0.80    2.25

(a) The referent value is early stage.
(b) The referent value is male gender.

Accepted March 3, 2003.

Copyright [c] 2004 by The Southern Medical Association

0038-4348/04/9703-0231

References

1. Parker SL, Tong T, Bolden S, Wingo PA. Cancer statistics, 1996. CA Cancer J Clin 1996;46:5-27.

2. Parkin DM, Whelan SL, Ferlay J, Raymond L, Young J (eds): Cancer Incidence in Five Continents (IARC Scientific Publications No. 143). Lyon, IARC Press, 1997, vol VII.

3. Heys SD, O'Hanrahan TJ, Brittenden J, Eremin O. Colorectal cancer in young patients: A review of the literature. Eur J Surg Oncol 1994;20:225-231.

4. Soliman AS, Bondy ML, Levin B, Hamza MR, Ismail K, Ismail S, et al. Colorectal cancer in Egyptian patients under 40 years of age. Int J Cancer 1997;71:26-30.

5. Marble K, Banerjee S, Greenwald L. Colorectal carcinoma in young patients. J Surg Oncol 1992;51:179-182.

6. Smith C, Butler JA. Colorectal cancer in patients younger than 40 years of age. Dis Colon Rectum 1989;32:843-846.

7. Isbister WH, Fraser J. Large-bowel cancer in the young: A national survival study. Dis Colon Rectum 1990;33:363-366.

8. Griffin PM, Liff JM, Greenberg RS, Clark WS. Adenocarcinomas of the colon and rectum in persons under 40 years old: A population-based study. Gastroenterology 1991;100:1033-1040.

9. Bini EJ, Rajapaksa RC, Valdes MT, Weinshel EH. Is upper gastrointestinal endoscopy indicated in asymptomatic patients with a positive fecal occult blood test and negative colonoscopy? Am J Med 1999;106:613-618.

10. Looker AC, Dallman PR, Carroll MD, Gunter EW, Johnson CL. Prevalence of iron deficiency in the United States. JAMA 1997;277:973-976.

11. Fante R, Benatti P, di Gregorio C, De Pietri S, Pedroni M, Tamassia MG, et al. Colorectal carcinoma in different age groups: A population-based investigation. Am J Gastroenterol 1997;92:1505-1509.

12. Vasen HF, Mecklin JP, Khan PM, Lynch HT. The International Collaborative Group on Hereditary Non-Polyposis Colorectal Cancer (ICG-HNPCC). Dis Colon Rectum 1991;34:424-425.

13. Syngal S, Fox EA, Eng C, Kolodner RD, Garber JE. Sensitivity and specificity of clinical criteria for hereditary non-polyposis colorectal cancer associated mutations in MSH2 and MLH1. J Med Genet 2000; 37:641-645.

14. Dietmaier W, Wallinger S, Bocker T, Kullmann F, Fishel R, Ruschoff J. Diagnostic microsatellite instability: Definition and correlation with mismatch repair protein expression. Cancer Res 1997;57:4749-4756.

15. Gryfe R, Kim H, Hsieh ET, Aronson MD, Holowaty EJ, Bull SB, et al. Tumor microsatellite instability and clinical outcome in young patients with colorectal cancer. N Engl J Med 2000;342:69-77.

16. Karran P, Hampson R. Genomic instability and tolerance to alkylating agents. Cancer Surv 1996;28:69-85.

17. Watanabe T, Wu TT, Catalano PJ, Ueki T, Satriano R, Haller DG, et al. Molecular predictors of survival after adjuvant chemotherapy for colon cancer. N Engl J Med 2001;344:1196-1206.

18. Bernstein MA, Madoff RD, Caushaj PF. Colon and rectal cancer in pregnancy. Dis Colon Rectum 1993;36:172-178.

19. Heise RH, Van Winter JT, Wilson TO, Ogburn PL Jr. Colonic cancer during pregnancy: Case report and review of the literature. Mayo Clin Proc 1992;67:1180-1184.

20. Nesbitt JC, Moise KJ, Sawyers JL. Colorectal carcinoma in pregnancy. Arch Surg 1985;120:636-640.

21. Woods JB, Martin JN Jr, Ingram FH, Odom CD, Scott-Conner CE, Rhodes RS. Pregnancy complicated by carcinoma of the colon above the rectum. Am J Perinatol 1992;9:102-110.

22. Chung YF, Eu KW, Machin D, Ho JM, Nyam DC, Leong AF, et al. Young age is not a poor prognostic marker in colorectal cancer. Br J Surg 1998;85:1255-1259.

23. Mitry E, Benhamiche AM, Jouve JL, Clinard F, Finn-Faivre C, Faivre J. Colorectal adenocarcinoma in patients under 45 years of age: Comparison with older patients in a well-defined French population. Dis Colon Rectum 2001;44:380-387.

24. Isbister WH. Colorectal cancer below age 40 in the Kingdom of Saudi Arabia. Aust N Z J Surg 1992;62:468-472.

25. Iliyasu Y, Ladipo JK, Akang EE, Adebamowo CA, Ajao OG, Aghadiuno PU. A 20-year review of malignant colorectal neoplasms at University College Hospital, Ibadan, Nigeria. Dis Colon Rectum 1996;39:536-540.

RELATED ARTICLE: Key Points

* Colorectal cancer does occur in young women and even during pregnancy.

* Unfortunately, it is rarely diagnosed early in this group of patients.

* Anemia could be a presenting sign.

* Primary care physicians and gynecologists should be aware that patients with gastrointestinal symptoms coupled with anemia should probably be referred to a gastroenterologist for evaluation for colon cancer.

Olusola Olofinlade, MD, Olusegun Adeonigbagbe, MD, Nicholas Gualtieri, MD, Hal Freiman, MD, Olugbenga Ogedegbe, MD, and James Robilotti, MD

From the Section of Gastroenterology, Department of Medicine, St. Vincent's Hospital, New York, NY.

Reprint requests to Olusola Olofinlade, East Texas Gastroenterology Associates, 701 E. Marshall, Suite 200, Longview, TX 75601. Email: [e-mail]