Prognoses

Almost everyone wants to know how serious his or her cancer is and what the prospects are for survival. This questioning is completely normal.

Limitations on accurate prognoses

Improving treatments

First, owing to robust research, treatments for NHL are evolving so rapidly that information regarding prognoses stated unequivocally today may be obsolete tomorrow. The recent FDA approval of certain monoclonal antibodies--white blood cell proteins designed to attack only tumors--is one example. The increasing use of bone marrow transplantation is another; the tremendous gains made in supportive care, such as new antifungals and antinausea drugs, are others. As always, your doctor, a well-trained and skilled person who most likely you chose carefully, is your best resource for the most current information. You may choose to follow the progress of new treatments on your own.

Do not treat what is written about prognoses in any book or article as definitive information about your condition.

Difficult classifications

The difficulty of categorizing NHL makes discussing research results equally difficult. Many major cancer centers have developed their own categories in order to tackle NHL in a consistent manner to effect a cure. This means that multiple studies from different institutions that yield conflicting results from the same treatment regimen may not compare readily to each other, and, most importantly, may not apply to you regarding fundamentals such as tumor cell type. Some NHL survivors, for example, in casual conversation might refer to having had lymphoma in the abdomen. There are several types of NHL that can occur in the abdomen, and they are sufficiently different to make comparisons about their response to treatment fruitless.

Do not become confused and worried because others with cancers that sound similar to yours appear to be doing better or worse than you.

Limitations of statistics

Those of you who studied statistics in school are aware that many different statistical methods exist to manipulate data, any two of which may in some cases give differing results. Statistical analysis is really just a method for making sense of large amounts of otherwise incomprehensible data. Consequently, sometimes the statistical model chosen represents science's closest guess regarding how to analyze the outcome of treatment; some statistical models chosen may not be a good fit for some collections of data. In spite of the best faith on the part of researchers and statisticians, these inconsistencies may creep into research papers. For more information on this topic, we recommend reading Steven Jay Gould's essay, "The Median Isn't the Message." Steven Jay Gould is a popular evolutionary biologist, and a survivor of a rare form of cancer called abdominal mesothelioma. His essay appeared in the June 1985 issue of Discover magazine, and can also be found on the Cancerguide web site.

You and your disease may not fit any statistical category.

Correlation is not causation

Complexity of the immune response

We have a great deal to learn about the immune system and are learning great amounts quickly owing to well-financed cancer research and the sharing of knowledge across scientific disciplines.

Blood cancers are unlike other cancers

Lymphomas and leukemias are hematopoietic cancers--that is, cancers of white blood cells, the disseminated body products responsible for protecting you from injury and disease, including cancer. Thus, some of the facts that apply to other cancers do not apply to lymphomas and leukemias. For example, various alternative treatments which purport to enhance the immune response against cancer may in theory make lymphomas or leukemias worse by stimulating growth of diseased white blood cells. Some alternative products such as melatonin carry printed warnings about this.

In addition, the spread (metastasis) of non-Hodgkin's lymphomas to multiple sites within the body does not necessarily mean that a successful outcome will be less assured. The idea of having cancer spread frightens many people because they have heard of people with cancer who could not be helped because their cancer had "spread to lymph nodes," "spread to the bone," or "spread to the brain." Although the involvement of certain organs may imply a worse outcome if they are found to contain NHL, and although a very large tumor burden spread to many sites is not a good sign, it is not universally true that any spreading of NHL to other organs is necessarily predictive of a poor outcome. Spread to the spleen, for instance, generally can be managed quite well by removal of the spleen, called splenectomy. Spread to bone marrow can be handled by high-dose chemotherapy and bone marrow rescue, or by the newer monoclonal antibodies. Spread to the central nervous system, including the brain, can be managed with radiation therapy or chemotherapy injected within the spine.

For reasons that are not fully understood, spread of cancers of the immune system are more manageable than the dissemination of cancers from a single organ, such as lung cancer. Thus, general statements that are true for other cancers may not apply to NHL.

Physical characteristics of patients in studies

The same treatments used on any one person may produce better results; the same treatment used on the general population of NHL patients may produce better results than were seen in clinical trials with heavily pretreated patients.

Long natural history

At the time this is being written, there are about 190 clinical trials underway for NHL funded by the National Cancer Institute, and this number does not include trials funded solely by pharmaceutical companies. For a better understanding of what this may mean for those who are diagnosed today, consider that only twelve years ago we did not have:

• Granulocyte colony stimulating factor, G-CSF, approved in 1991 for growing new white blood cells to prevent you from catching infections if chemotherapy wiped out your white blood cells.
• Erythropoietin, or Epoietin, for growing new red blood cells when bone marrow has been suppressed by chemotherapy or radiation therapy.
• Interferon alfa, a manmade copy of a blood product that can suppress NHL.
• IL-2, another manmade copy of a blood product that is active against melanoma, leukemias, kidney cancer--all cancers once considered hopeless unless caught very early.
• Monoclonal antibodies, white blood cell proteins that are grown outside the body where they are taught to travel to and attack tumors. They are almost unique in their ability to avoid damaging healthy tissue, and thus are less likely to cause serious side effects.
• Safer bone marrow or stem-cell transplants. Twelve years ago, bone marrow transplants (BMTs) were far more dangerous, done only with donor marrow because the medical community did not know for certain how to clean (purge) the cancer patient's marrow of cancer cells and reuse it. Moreover, BMTs were done only on relatively young people. Since 1993, the age of patients considered acceptable for bone marrow or stem cell rescue using their own marrow has risen from forty to sixty and, in some cases, the seventies if the patient is otherwise healthy.
• Magnetic resonance imaging (MRI) for finding very small tumors and NHL that have spread to the brain and spine, something not readily available until about eleven years ago.

Thus, bear in mind that the future holds great promise.

The aging of printed material

Remember that what you read about survival and treatment success here, and in all but the newest texts, will never be as current as the information you can receive from a well-trained oncologist active in his specialty who has access to medical journals and to other researchers.

Which factors matter least and most

Even with the following list of risk factors, nobody will be able to speak in absolute terms about your overall prognosis. You undoubtedly will have at least one risk factor for a poorer prognosis, and you will undoubtedly have several factors that point to a better prognosis.

The most important point to remember is that what are used today as reliable prognostic indicators may become meaningless when new treatments that surmount old difficulties are engaged.

The order of the sections that follow does not imply a greater or lesser effect on outcome.

High-dose therapy with bone marrow transplantation

High-dose therapy with bone marrow transplantation is common terminology used to described several related procedures, namely marrow or stem cell transplantation, or marrow or stem cell rescue.

For all of the non-Hodgkin's lymphomas, bone marrow transplantation appears to offer longer survival or cure if treatment-related side effects can be controlled, but a longer period of follow-up is necessary to be certain that the longer remissions are statistically significant for low-grade disease. Some studies indicate a better chance of relapse-free survival with donor marrow--allogeneic transplantation--than with one's own marrow. Another study has found that achieving remission during induction therapy for transplantation is a "powerful prognostic indicator" for long-term survival.¹

There are serious advantages and disadvantages to each type of transplant with respect to side effects and outcome.

Tumor grade

Intermediate- and high-grade non-Hodgkin's lymphomas are considered curable in many instances, depending on various factors.

Note that a low-grade NHL can convert to a higher grade, and that high-grade disease can retain some characteristics of low-grade disease, even as it is progressing.

Histology

The histologic naming classifications of NHL are very complex. There has been controversy about which naming classifications might correlate with overall prognosis. Confusion can be caused by the presence of two conflicting histologic features within the same tumor, for example.

The following histologic naming conventions have not been found to have a consistent bearing on long-term survival. We're explicitly listing them here because these naming conventions may be cited in an older study or book as a possible risk factor in prognosis.

• Follicular. Not all patients with follicular disease have indolent, slow-growing disease that requires no initial treatment. Up to 50 percent have abnormalities that must be addressed immediately.
• Follicular large-cell. Some follicular presentations have only small cells, some only large cells, some both. Large cells within follicular nodes sometimes are correlated with a more aggressive disease, but this is not consistent across all studies.
• Diffuse large-cell. The diagnosis and categorization of this group are so complex and so subject to overlaps with other types that predictions about the success of treatment based on differing naming conventions alone are not useful.
• Small lymphocytic/lymphoplasmacytoid (immunocytomas). The diagnosis and categorization of this group are so complex, and so subject to overlaps with each other, that predictions about the success of treatment based on differing names alone are not useful.
• Noncutaneous T-cell. Differences in survival between small- and large-cell subtypes make generalizations about noncutaneous T-cell lymphoma difficult. Those that express the CD30 cell surface antigen seem to have a better prognosis.
• Cutaneous T-cell. Other factors, such as spread of disease, are more meaningful than the naming category of this subtype.
• Adult Burkitt's. Other factors, such as LDH levels, are more meaningful in predicting outcome. LDH is discussed later in the chapter.

These histologic categories, however, appear to have a consistent correlation with outcome:

• Anaplastic large-cell. Seems to have longer survival than other noncutaneous T-cell lymphomas.
• Mucosa-associated lymphoid tissue (MALT). Most patients have limited disease outside of any lymph node and have a good prognosis. Those with gastric MALT having a diffuse large-cell component of 1 to 10 percent appear to have a worse 10-year prognosis than those without a diffuse large-cell component.
• Mantle cell. Patients with mantle cell disease may have very widely spread disease and consequently may face a poorer outcome.
• Mantle zone. True mantle zone disease appears to have a better prognosis than mantle cell.
• Childhood Burkitt's. Most children with Burkitt's will have long survival, unless bone marrow or the central nervous system is involved, which would confer a slightly worse prognosis.

Immunophenotype

The science of immunophenotyping, or immunotyping, is a rapidly advancing subfield of cancer research. Some researchers say that advances in this method of analyzing tumor cells will provide us with the most meaningful information possible for designing patient-specific anticancer products.

• B-cell versus T-cell. White blood cells can be divided into B cells, arising from the bone marrow, and T cells, which also arise from marrow but mature in the thymus gland or descend from early T cells of the thymus. Some studies show that T-cell non-Hodgkin's lymphomas are harder to treat successfully than B-cell NHLs; other studies do not support this. Keep in mind that some tumors have both B-cell and T-cell characteristics (composite lymphoma), and some tumors have neither.
• Cell surface antigens. This very new method of categorizing cancer cells may prove useful for targeting tumor cells only. All white blood cells have proteins protruding from them, called cell surface antigens, which cancer cells produce in greater abundance or in different quality from normal cells. This allows cancerous white blood cells to be targeted instead of healthy cells. Examples of cell surface antigens are CD20, CD5, CD19, CD30 (Ki-1), and so on.
• p80 antigen. In anaplastic large-cell lymphoma (ALCL), expression of the p80 antigen has been associated with a good outcome in one study.² Additional studies are necessary to confirm this.

Genetic characteristics

Some genetic abnormalities may be present in all tumor samples from the same patient, some are present in one sample but not in others, and some are present at diagnosis, but others accumulate as time progresses.

The significance of many individual genetic aberrations is not yet clear, although there is some consensus that a high number of genetic abnormalities may confer a poorer outcome. For instance, a tumor showing deletion of both chromosomes 5 and 7, extra copies of chromosomes 8 and 12, and transpositions of two or more other chromosomes would be considered to have a high number of abnormalities.

There have been some studies that indicate a statistical correlation between certain kinds of genetic damage and outcome:

• Damaged or missing material for chromosomes 7 and 17 have been correlated with a worse prognosis.
• Damage to the cell-death gene (tumor suppressor gene), p53, which resides on chromosome 17, appears to affect negatively the outcome of many cancers, including the NHLs, but at least one study has found that it does not affect prognosis in the NHLs.³
• Excessive material (three copies) of chromosomes 7 and 14 have been correlated with a worse prognosis.

In other studies, some aberrations are found in or appear to affect the outcome of only specific subtypes of NHL:

• In intermediate- and low-grade lymphomas, aberrations in chromosome 1 have been correlated with a poor outcome.
• For low-grade lymphomas, three copies of chromosome 12 (trisomy 12) have been associated with a worse outcome, and a small study has shown that trisomy 5, 6, or 8, chromosome 5 abnormalities, and damage to parts of chromosome 14 confer a worse prognosis.
• For large-cell lymphoma, a study of 205 tumors showed a correlation between damage to chromosome 1 and a shortened survival.
• For follicular disease transforming to diffuse disease, loss of chromosome 6, aberrations of chromosomes 17 and 8 and others, and swapping of chromosomal parts between chromosomes 14 and 18 have been found, implying that these chromosomal aberrations may be related to the less favorable prognosis of some diffuse tumors.
• For nonlymphoblastic T-cell lymphomas, one study has found that the presence of Epstein-Barr virus (EBV) in the tumor's DNA is a predictor of poor outcome.⁴

Patient characteristics

• Age. There is disagreement in the medical community regarding the influence of age on outcome of treatment. Several studies indicate that general health status or tumor bulk is more important than age in predicting the success of treatment. Other studies show that increasing age greater than 65 is indeed a poor prognostic factor, independent of other characteristics. Patients of any age in good health who can withstand full doses of chemotherapy with few delays fare better than those who must accept repeated reduced doses on a schedule significantly delayed by recovery from adverse side effects.
• HIV infection. Individuals whose HIV positive status has progressed to full AIDS and who are not able to maintain CD4+ white blood cell counts greater than 100/microlitre using protease inhibitors or other antiviral therapy have a poorer prognosis. Those who are able to maintain higher CD4+ counts have a prognosis similar to those without HIV, if good supportive care is available to withstand infection.

Tumor mass

• The tumor burden the body is carrying is directly related to outcome, with higher tumor burden within many organs related to poorer outcome than disease spread only to one organ.
• The number of nodal sites, however, is no longer considered relevant to outcome.
• In large-cell and follicular lymphomas, an increasing number of extranodal sites (spleen, liver, marrow, brain) is correlated to poorer outcome, but extranodal disease is in itself not riskier than nodal disease.
• Bone marrow involvement confers a worse prognosis only for Burkitt's and lymphoblastic lymphoma. The outcome for follicular cell lymphoma, immunocytoma, and mantle cell lymphoma with bone marrow involvement, however, appears no worse than these same subtypes without bone marrow involvement.
• One study has found that abnormal results from the magnetic resonance imaging (MRI) of the femoral (thigh bone) marrow "is associated with a significantly poorer survival in patients with malignant lymphoma, regardless of histologic findings in the marrow."⁵
• Pleural effusion, a collection of fluid in the lungs, is associated with a poorer outcome.

Blood-borne products of metabolism

• LDH, lactic dehydrogenase. This is an enzyme associated with the breaking down of cells for any reason, is directly correlated to tumor burden and thus to outcome, with higher LDH levels indicating more disease. This is true for almost all NHL subtypes excepting some large cell lymphomas (LCLs). It's not a foolproof indicator, though, because other things that cause cells to burst, such as recent strenuous exercise or the administration of G-CSF or GM-CSF, can temporarily raise LDH.
• Beta-2 microglobulin. This is a newer measure of tumor burden, and higher levels have been shown to be directly related to poorer outcome for large-cell and follicular lymphomas. This measure has not been tested in large studies, however.
• Other blood-borne products. This includes compounds such as tumor necrosis factor (TNF) and various cytokines such as the interleukins 6 and 10 (IL-6, IL-10) that are thought to be a measure of the body's response to the tumor as opposed to clearly meaningful prognostic tools, although some of these products appear to mirror tumor burden. Several of these blood values--sICAM-1, erythrocyte sedimentation rate (ESR), serum thymidine kinase (sTK), albumin, and orosomucoid--rise and fall along with LDH, so their value as independent predictors is questioned. At this time, these substances are not commonly used as diagnostic or follow-up tools.

Other tumor-related substances

Many new biochemical tools, such as polymerase chain reaction (PCR), fluorescence in situ hybridization (FISH), and immunoperoxidase staining, can detect tumor products that reflect tumor activity.

bcl-2 is a rearranged gene that some people with NHL are tested for. What these levels mean for prognosis is not clear. One study has found that bcl-2 levels do not correlate to prognosis.⁶ Another has found that bcl-2 does predict poor outcome, even among those who achieve full remission on treatment.⁷ A third study has found that expression of the bcl-2 protein correlates to a poorer outcome in large-cell lymphomas.⁸

Patient's response to tumor

• B symptoms. Unexplained loss of 10 percent or more of body weight, drenching night sweats, fever of 38°C (about 100°F) or higher for more than one week. Today, only unexplained weight loss appears to be related to the amount of disease.
• Performance status (Karnovsky or ECOG). Measures the patient's ability to do everyday things. The lower the performance status at diagnosis, the poorer the outcome. This measure does not apply to temporary setbacks while coping with the side effects of chemotherapy.
• Serum albumin level. Lower levels correlate with a worse prognosis.
• Hemoglobin level. Untreated anemia is correlated with a worse prognosis.
• For mediastinal (thymic) NHL, poor performance status, pericardial effusion, bulky or residual mediastinal abnormality, or inadequate response after front-line treatment predict poor response and later relapse. These patients may be candidates for high-dose treatment.

Patient's response to chemotherapy

Dose intensity of chemotherapy

The necessity to reduce dose intensity owing to patient characteristics is correlated with a poorer outcome.

Notes:

1. H. M. Prince et al., "The role of intensive therapy and autologous blood and marrow transplantation for chemotherapy-sensitive relapsed and primary refractory non-Hodgkin's lymphoma: identification of major prognostic groups," British Journal of Haematology 92, no. 4 (March 1996): 880-889.
2. M. Shiota and S. Mori, "The clinicopathological features of anaplastic large cell lymphomas expressing p80 NPM/ALK," Leukemia and Lymphoma 23, nos. 1-2 (September 1996): 25-32.
3. M. Osada et al., "Mutation of p53 does not determine prognosis in non-Hodgkin's lymphoma," Gan To Kagaku Ryoho, 24, no. 4 (February 1997): 471-475.
4. F. d'Amore et al., "Epstein-Barr virus genome in non-Hodgkin's lymphomas occurring in immunocompetent patients: highest prevalence in nonlymphoblastic T-cell lymphoma and correlation with a poor prognosis. Danish Lymphoma Study Group, LYFO" Blood 87, no. 3 (1 February 1996): 1045-1055.
5. S. Tsunoda et al., "Clinical and prognostic significance of femoral marrow magnetic resonance imaging in patients with malignant lymphoma," Blood 89, no. 1 (1 January 1997): 286-290.
6. W. H. Wilson et al., "Relationship of p53, bcl-2, and tumor proliferation to clinical drug resistance in non-Hodgkin's lymphomas," Blood 89, no. 2 (15 January 1997): 601-609.
7. M. E. Hill et al., "Prognostic significance of BCL-2 expression and bcl-2 major breakpoint region rearrangement in diffuse large-cell non-Hodgkin's lymphoma: a British National Lymphoma Investigation Study," Blood 88, no. 3 (1 August 1996): 1046-1051.
8. M. H. Kramer et al., "Clinical significance of bcl2 and p53 protein expression in diffuse large B-cell lymphoma: a population-based study," Journal of Clinical Oncology 14, no. 7 (14 July 1996): 2131 to 2138.
9. M. Davidge-Pitts, R. Dansey, and W. R. Bezwoda, "Prolonged survival in follicular non-Hodgkins lymphoma is predicted by achievement of complete remission with initial treatment: results of a long-term study with multivariate analysis of prognostic factors," Leukemia and Lymphoma 24, nos. 1-2 (December 1996): 131-140.