Catalog and Status of the Quality of Medical Evidence for the most Important Decisions in

the Management of Hematologic Malignancies in Adults:

Years: 1966 to 1996

Send comments regarding this catalog to Dr. Ben Djulbegovic at:bdjulbeg@hsc.usf.edu

Decision Intervention Best Available Evidence Reference Comment
Acute lymphoblastic leukemia (ALL) at least 27 randomized controlled trials (RCT) have been reported evaluating various forms of treatments
early pre-B, pre-T, T-cell ALL: most authors do not tailor decision according to immuno-phenotype and some do not consider age an important variable in decision making
induction Rx: age:<60,CSF negative chemoRx II Leukemia 1992;6(suppl 2):175, 178

Leukemia 1997;11 (suppl4)

several RCT performed with other regimens but best regimen not identified
induction Rx:

age:<60, CSF positive

triple i.t.Rx, followed by chemoRx II "
induction Rx:

age:>60 year old

chemoRx II Blood 1996;87:495 optimal regimen not identified; most authors would give same chemoRx as in younger patients
mature B-cell ALL:
induction Rx chemoRx II Blood 1996;87:495 ara-C based chemoRx; optimal dose of ara-C not identified (based on German protocol for Rx of mature B-cell ALL)
CNS prophylaxis (all types): brain XRT+ i.t. chemoRx II Am J Med 1994;97:176
Consolidation:
non-mature B-cell ALL chemoRx II Leukemia 1992;6(suppl 2):175 best regimen not identified; most authors suggest ara-C/etoposide based chemoRx
mature-B cell ALL No II Blood 1996;87:495
Ph+ALL (<55 old, HLA donor available) alloBMT III Bone Marrow Transplant 1996;16:663
Ph+ALL (<55 old, no donor available) MUD vs autoBMT III "
Ph+ALL (>55 old) "standard" consolidation III Ann Intern Med 195;123:428
maintenance Rx:
non-mature B-cell yes I Br J Haemtol 1996;92;665 6-MP/methotrexate/vincristine/prednisone

(GIMMMA study 0183) (no positive effect from early intensive consolidation)

mature B-cell no II Blood 1996;87:495
primary refractory ALL chemoRx III Eur J Haematol 1990;44:240 based regimen not identified (high-dose ara-C/mitoxantrone is recommended by some)
relapsed ALL:
<60 years old chemoRx followed by BMT III " based regimen not identified (high-dose ara-C/mitoxantrone is recommended by some)
>60 year old supportive care III Blood 1992;80:1813 some authors would try chemo as in patients<60
Acute myelogenous leukemia (AML) at least 165 rct have been reported evaluating various forms of treatments
induction Rx:
M3-subtype ATRA I Blood 1993;82:3241

N Engl J Med 1997;337:1021

all other FAB-subtypes chemoRx I Semin Oncol 1997;24:57 ara-C+idarubicin appears to produce highest remission rate (excellent summary in ref. listed) (meta-analysis published in the abstract form Blood 1995;86:434a)
AML following MDS high-dose ara-C II J Clin Oncol 1992;10:41 optimal "high" dose not identified; most authors recommend 3 gm/m2 (1.5 gm/m2 for patients older than 50)
consolidation Rx:
average risk AML high-dose ara-C I J Clin Oncol 1992;10:41 average risk AML would refer to normal karyotype and the absence of favorable [inv(16),t(8;21),t(15;17)] or unfavorable (chromosome 5,7 abnormalities; prior MDS) characteristics
high-riskAML:
< 55 year old (HLA donor available) alloBMT I N Engl J Med 1995;332:217
>55 year old autoBMT III J Clin Oncol 1992;10:41 many other authors favor chemoRx
primary refractory AML:
<55 year old (HLA donor available) alloBMT III Blood 1992;80:1090
>55 year old chemoRx III Blood 1990;76:473 optimal regimen not identified
relapsed AML:
after chemoRx BMT III Semin Hematol 1995;32:132
after BMT second BMT III Blood 1994;84:3605
Myelodysplastic syndrome (MDS) at least 19 rct +1 meta-analysis have been reported evaluating the use various treatment options in the management of MDS
initial treatment for stable disease observation I Ann Hematol 1992;65:162

(erratum in 1993;66:164)

several rct evaluating effect of low-dose chemoRx and hormonal therapy over supportive care only resulted in no survival benefit. One rct trial, however, did report positive effects on CBC and survival with danazol over placebo (Arch Invest Medica 1989;20:183); other authors have not found danazol useful (Cancer 1994 Jun 15;73:3073, Am J Hematol 1995;48:233)
progressive disease: numerous trials support no use of steriods, androgens

(Am J Hematol 1995;48:233), vitamin D3 (Int J Clin Pharmacol 1993;13:21, Eur J Haemtol 1990;45:255,) cis-retinoic acid (Blood 1988;71:703, Br J Haemtol 1987;66:77), etc

<55 year old (HLA matched donor available) alloBMT II Blood 1993;82:677 the goal of Rx is believed to be cure. DFS~45% at 3 years but mortality also approaches 40%
<40 year old (MUD available) alloBMT III "

Leukemia Lymphoma 1995;17:95

not much experience; best data report DFS:49% at 2 yrs, mortality at 2 years 46%
<65 year old (donor not available) AML-like treatment III Leukemia Lymphoma 1993;11:59 brief CR, median survival 9-12 monts, DFS:20-25% at 2yrs; mortality 10-40%
>65 year old supportive therapy

(with growth factors such as GM-CSF or erythropoietin)

I Leukemia Lymphoma 1995;18:457

Br J Haematol 1995;89:67

available level I evidence are of small power; however, no trial showed effect on survival; meta-analysis (Br J Haematol 1995;89:67) suggest that early use of erythropoietin (epo) in patients with low epo levels may be beneficial (reduce RBC transfusion requirement)
Myeloproliferative disorders:
P. Vera: 13 (4 big + 9 small ) randomized + pseudoRCT trials evaluating effects of various interventions were performed in the management of P. Vera
initial treatment: phlebotomy I Semin Hematol 1986;23:132

(phlebotomy shown to be superior to chlorambucil and P32)

this is one of 3 large randomized trials performed in P. Vera; other trials are: 1) P32 vs busulphan:busulphan was superior (Br J Cancer 1981;44:75; updated in Drugs Experim Clin Research 1986;12:283)

2) hydroxyurea vs pipobroman: no difference (HU better tolerated) (Presse Med 1992;21:1753; updated Blood 1997; in press); 3) P32 vs P32 + HU as maintenance treatment: life expectancy was shorter with combination Rx (except for most severe cases when P32-induced remission lasted <2 years) (Blood 1997;89:2319)

second line treatment: hydroxyurea II Semin Hematol 1997;34:17 the most commonly used myelosupressive drug in PV; appears to highly efficacious drug with no mutagenic potential (in retrospective comparison with phlebotomy; see listed ref)
-interferon II Br J Haematol 1996;92:55

(analysis of 100 cases published in the literature)

appears to be highly effective regimen; less studied than other modalities; effect on the natural history not determined; one small rct (N=22, cross-over design) concluded that IFN is superior of phlebotomy in controlling all clinical aspects of the disease (Ann Hematol 1994;68:247)
prevention of thrombosis: ASA+persantin I Semin Hematol 1986;23:172 in a dose used ( 900 mg/d) increase rate of hemorrhagic complications noted; smaller doses are currently under study in a randomized trial; ticlopidine shows beneficial effect on laboratory markers, but it was not evaluated with the respect of clinical events of interest (i.e. DVTs, etc)
use of iron in P. Vera iron III J Intern Med 1984;216:165 there is a theoretical risk for the use of iron in P. Vera (stimulation of malignant clone); this does not appear to be case by anecdotal data; also it does not appear that induced iron deficiency increases hyperviscosity (ref)
surgery in P. Vera stable disease for 3-4 months prior surgery III N Engl J Med 1963;23:1226 retrospective data indicate increase risk for surgery in uncontrolled disease
treatment of pruritus histamine blockers vs IFN vs aspirin photomochemoRx II Semin Hematol 1986;23:132

Br J Haematol 1995;89:313

Ann Hematol 1996;73:91

P. Vera Study group included H1-blockers (cyproheptadine) or H2-blockers (cimetidine) in its protocol. In one cross over study aspirin (500 mg/d) was more effective than placebo; no effect on bleeding reported in this small study (Acta Dermatovener (Stockholm) 1979;59:505-512) IFN was effective in 12 out 15 patients in the study cited. Psoralen + UV-A light controlled pruritus in 10 of 11 patients (maintenance Rx was generally required)
Essential Thrombocythemia only one rct evaluating effect of HU (vs placebo) (see below) was published in the literature in the treatment of ET
initial treatment:
young asymptomatic patient with platelets<1.5M Observation III Cancer 1991;67:2658 natural history of a disease does not suggest increased risk of mortality in this group of patients
asymptomatic elderly patient (>60 years old) or the patient with previous history of thrombosis or symptoms with platelets<1.5M hydroxyurea I N Engl J Med 1995;332:1132-6 patients with platelets>1.5M are considered to be increased risk for thrombohemorrhagic complications and also should be treated (level evidence III evidence:J Clin Oncol 1990;8:556)
initial treatment in patients with platelets>900 K, or second line treatment anagrelide II Am J Med 1992;92:69-76

Semin Hematol 1997;34:51

no comparative trial between anagrelide ans HU exists; anagrelide was also effective in those patient in whom HU or IFN failed
" -interferon II Lancet 1988;2:70 no comparative trial between IFN ,anagrelide and HU exists;
use of antiplatelet prophylaxis ASA (low dose:300 mg/d) II N Engl J Med 1995;333:802-803 (letter) most authors favor use of some antiplatelet agents; ticlopidine is also used; no comparison between these two agents exist
acute hemorrhagic complications plateletpheresis with platelet infusion III Transfusion 1979;19:147 dramatic anecdotal data support this practice
splenectomy absolute contraindication III Illand HJ and Laszlo J. Myeloproliferative disorders:Polycythemia Vera, Essential Thrombocythemia and Idiopathic Myelofibrosis. In:Hoogstraten B, ed. Hematologic Malignancies. Springer-Verlag:Berlin,1986:51-52 dramatic anecdotal data support this practice
surgery in ET heparin in prophylaxis III " surgery is considered of very high risk in ET; optimal regimen of anticoagulant prophylaxis is not established
Myelofibrosis 2 rct were published evaluating effects on androgens on IMF
asymptomatic patient Observation III Cancer 1991;67:2658 there is no evidence that any treatment modify course of natural history of Myelofibrosis
treatment of symptomatic patient: androgens vs. III N Engl J Med 1966;274:420

N Engl J Med 1961;264:104

data on efficacy not clear; it may be of some help in 50% cases. In one small randomized trial which included also other diseases characterized by marrow failure no positive effect was seen with nandrolone (Archives of Internal Medicine 1977;137:65); in other trial that also included other disease that can cause marrow-failure anemia four (29%) of 14 responded well to fluoxymesterone therapy; no effect on survival was seen (Archives of Internal Medicine 1982;142:1533)
" steroids vs. III Am J Med Sci 1962;243:697 no positive treatment effect seen in most of studies; may be beneficial if hemolytic anemia is cause of anemia
" vitamin D3 vs. III Lancet 1984;1:78 some effects seen in some studies; not confirmed in others (Br J Haematol 1986;62:399, Br J Haematol 1987;66:579)
" hydroxyurea vs. III Br J Haematol 1991;77:252 small study, with 10 patients reported; improvement in some symptoms noted in 8 patients
" busulphan vs. III Am J Med Sci 1988;295:472 3 patients studied, with some improvement

other agents reported in anecdotal cases include chlorambucil and 6-thioguanine (J Coll Physicians Lond 1981;15:17)

" low-dose XRT vs. III Cancer 1986;58:1204 responses usually transient and hematopoietic toxicity is frequently significant; one report ( listed ref) suggest combining with HU
" -interferon vs. III Eur J Haematol 1987;39:228

Eur J Haematol 1990;52(suppl):12

most other reports are also small case reports; it is difficult to establish the role of IFN in the treatment of IMF
" splenectomy vs. III Am J Med 1990;33:128 decision analysis splenectomy support splenectomy only for palliative treatment; no evidence that it prolongs survival.
BMT (allo) III Br J Haematol 1992;82:772

Br J Haematol 1989 71:158

few data exists to allow any meaningful conclusion

prognostic system (based on retrospective data, not validated in other studies) has been devised to assist in prognosis and selection of patients for BMT (Blood 1996;88:1013; Blood 1997;89:2219,letter)

" (with anemia only) eryrthropoietin vs. III Br J Haematol 1994;86:893 most data come from small case-reports series; majority of reports show no efficacy and caution possible adverse effects secondary to stimulation of extramedullary hematopoiesis (Lancet 1991;337:188). One case report, however, shows positive effect in 2 patients (Blood 76:271a)
" Supportive treatment + iron, folic acid, vitamins as indicated III Am J Med 1978;65:655 these decisions are derived from "usual" clinical reasoning; one report (ref), that was never reproduced, described positive effect of pyridoxine in 40% with anemia
Chronic Myeloid Leukemia at least 57 RCT evaluating different modalities of the treatment in CML were published in the literature
initial treatment:
<55 old (HLA matched or one antigen mismatch donor available) alloBMT II Blood 1996;87:3069

(review)

age cut-off not accepted by all authors; data on preparative regimens are of level I (see ); some retrospective data suggest that the patient should be transplanted in the center performing >5 procedures/ year (Blood 1992;79:2771)
<45 old (MUD available) alloBMT III " "
optimal timing of BMT within 1 year of diagnosis III " decision analysis suggest that timing of BMT is a function of a patient age, success of BMT procedure and a prognostic group of a disease (updated Seattle data suggest that the critical prognostic cut-off time may be 3rd year or later (Blood 1994;83:2752)
donor not available -interferon±ara C I N Engl J Med 1994;330:820

Lancet 1995;345:1392

Blood 1994;84:4064

Blood 1995;86:906

N Engl J Med 1997;337:223

in 4 randomized trial IFN was more successful than hudroxyurea or busulfan retrospectively. Optimal of dose of IFN is not established. Most data suggest that 5 MU/m2 is optimal dose. One single arm study (level II evidence) suggest that 2 MU/m2 daily for one month, then tiw could be equally effective (Ann Intern Med 1994;121:736); recent data (N Engl J Med 1997;337:223) suggest that IFN with ara-C may be superior to IFN alone
salvage therapy after BMT: donor lymphocyte infusion III Blood 1993;82:3211

N Engl J Med 1994;330:100

small single arm studies also published; second BMT is also alternative but carry high mortality risk (~65%); another strategy is to stop cyclosporine and administer IFN which appears not to have any effect long-term (level III evidence)
how to assess effect of therapy pcr for bcr/abl gene rearrangement q 2-3 months III Blood 1993;81:1089

Blood 1995;85:2632

conflicting data from different studies in terms of the possiblity of detecting disease relapse prior obvious hematological relapse
Hairy Cell Leukemia 17 (3 rct updated/reported on 9 different occassions) were published
initial treatment:
asymptomatic patient close follow-up III Semin Oncol 1987;11 (suppl 2):479 natural history of disease indicates long-term survival in 20-25% of patients
patient presenting with massive splenomegaly splenectomy III Cancer 1988;62:2420 retrospective data indicate positive effect of splenectomy in some(highly selected) patients
treatment of progressive and symptomatic disease cladribine II N Engl J Med 1990;322:1117 Randomized trials indicates effectiveness of IFN as well as pentostatin in this disease. There are 3 randomized trial in HCL: pentostatin vs inteferon (IFN) (J Clin Oncol 1995;13:974), splenectomy vs IFN (Am J Hematol 1992;41:13) and low-vs high-dose IFN (Blood 1991;78:3133) cladribine, however, was so highly efficacious that represented an important "quantum leap" in treatment innovations being accepted now as a treatment of choice for HCL
relapse, not responsive to cladribine -IFN vs pentostatine vs fludarabine III Leukemia 1995;9:929 few data exist to support the use of any agent over another in this setting, but appears that there is no cross-resistance between pentostatin and cladribine. One report (see ref) described 3 patients responding to -IFN after failing cladribine
Chronic Lymphocytic Leukemia at least 40 RCT evaluating different modalities of the treatment in CLL were published in the literature
initial treatment:
in early stages observation I Blood 1994;75:1414 survival appears to be worse in treated group
progressive and advanced disease chlorambucil (low dose)

(vs. CVP vs CHOP vs Fludarabine vs high-dose chlorambucil)

I Nouv Rev Fr Hematol 1994;30:343 Chlorambucil is still recommended because inconsistent results from randomized trials. CHOP was superior to CVP in stage C in French trial (Nouv Rev Fr Hematol 1988;30:449) but not to chlorambucil in Danish trial (Nouv Rev Fr Hematol 1988;30:433); CVP was not better than chlorambucil in stage B (Blood 1990;75:1422); finally fludarabine demonstrated a higher response rate to CHOP,CAP or chlorambucil, but longer follow-up is needed to determine if this higher CR will result in improvement in survival (Blood 1995; 86:607a ;Blood 1994;84:461a;Blood 1993;82:199a). In recent European randomized trial of patients with Binet stage B or C, fludarabine showed higher CR than CAP (60 vs 44%); in previously untreated group median duration of CR was 179 days for CAP and was not reached for fludarabine; this may translate into survival advantage for fludarabine which was 4.3 years for CAP and still has not been reached for fludarabine (p=0.087) (at 4.3 years 67% of patients treated with fludarabine are alive) (Lancet 1996;347:1422-38). However, in this trial 5% mortality was noted in fludarabine group (which was not differrent from CAP toxicity) but it is almost certainly higher than for chlorambucil. Finally, recent RCT evaluating high-dose CLB showed its superiority over CHOP (Cancer 1997;79:2107).

second line Rx: purine analogue (fludarabine vs cladribine vs pentostatin) II Blood 1993;82:1695

J Clin Oncol 1995;13:983

most data exists regarding the use of fludarabine in CLL, but no comparative data exists between fludarabine and other purine analogues
salvage after one purine analogue failed e.g. cladribine after fludarabine failed II N Engl J Med 1994;330:319 a study on 28 patients suggest that cladribine is ineffective in this setting; small study on 4 patients sugegst opposite (E Negl J Med 1992;327:1056)
<65 years old, with good performance status autoBMT III Blood 1995;86:1813 patients with mixed characteristics, some had purged marrow
<55 year old, with good performance status (HLA donor available) alloBMT III Ann Intern Med 1996;124:311 no comparison between auto and allo-BMT exists; very few data on MUD transplant exists
salvage therapy splenectomy vs splenic XRT III Am J Med 1992;93:435 retrospective data suggest that these modalities may prolong survival; one randomized trial (Nouv Rev Fr Hematol 1988;30:423) showed increased survival for irradiated patients
Multiple Myeloma There are at least 127 RCT published in the literature evaluating efficacy of various treatment modalities. There is also one meta-analysis of 18 rct evaluating combined chemoRx vs melphalan+prednisone, see J Clin Oncol 1992;10:334-342) , 2 studies dealing with treatment vs no therapy in early stages of myeloma, 3 studies researching effect of high-dose therapy with stem-cell/auto-BMT support, 17 rct studying effect of IFN in various phases of thedisease with different chemoRx, 6 rct establishing efficacy of various treatments (mainly diposphonates) on skeletal-related complications in myeloma, and 3 studies evaluating efficacy of various other treatment modalities in MM (such as IgG, epo and G-CSF in procurement of stem cell for high dose Rx)
initial Rx:
stage I, asymptomatic patients observation I Eur J Haematol 1993;50:95

Br J Cancer 1994;70:1203

no difference in survival in comparison with immediate treatment

(some preliminary report on the subject published in Blood 1992;79:113)

< 65 year old, stage II, good performance status (PS) autoBMT I N Engl J Med 1995;335:91 appears to be superior to combined chemoRx
<50 years old with HLA donor available alloBMT II J Clin Oncol 1995;13:1312 mortality up to 30%; it is not clear that alloBMT can cure MM
>65 year old or poor PS melphelan+prednisone I J Clin Oncol 1992;10:334 a meta-analysis integrated results from 18 randomized trials to show that M+P is likely equivavalent in efficacy to combined chemoRx regimens
maintenance Rx: -inteferon I Ann Intern Med 1996;124:212 conflicting results with respect to increase in survival; IFN likely increases median remission time but not survival ; there are at least 17 rct evaluating efficacy of IFN in myeloma (for editorial and reference list on these trials, see Ann Intern Med 1996;124:264) Maintenance treatment with melphalan-prednisone was shown not be effective in rct (Br J Cancer 1988;57:94)
salvage Rx:
resistant disease high dose sterioids III Ann Intern Med 1986;105:811 most authors would not use it as a first line Rx in relapsed patients
relapsed disease (< 6 months) VAD III Mayo Clin Proc 1994;69:7787 widely used regimen for relapsed disease, and sometimes as a first line Rx; one small rct (N=47, prematurely closed because of poor accrual) suggested no difference between VAD vs VAD+IFN. Also no obvious advantage over historical data with high-dose steroids suggested (Am J Clin Oncol 1985;18:475)
progressive on second line Rx high-dose melphalan

(with auto BMT/stem cell cell support)

III Blood 1996;82:838 favored currently by many authorities
other treatments: diphosphonates to prevent skeletal related complications I N Engl J Med 1996;334:488 pamidronate (IV) and clodronate (po) (Lancet 1992;340:1049) shown to reduce skeletal complications in stage II and III myeloma; no positive effects was seen for etidronate (Eur J Med 1993;2:449)
high dose IgG for prevention of infections I Lancet 1994;343:1059 treatment successful in reducing the risk of infections
erythropoietin to treat anemia I Arch Intern Med 1995;155:2069 epo effective and safe in patients with the anemia of myeloma
plasmapheresis in kidney failure I Archives of Internal Medicine 1990 ;150:863 Plasmapheresis and chemotherapy lowered the serum myeloma protein value much more rapidly than chemotherapy alone.
Waldenstrom's disease one small pseudo-randomized study (N=11) was reported evaluating two types of plasmapheresis (Clin Nephrol 1995;43:335)
asymptomatic patient observation III Blood 1994;83:1452 natural history suggest long-term survival for these patients; it is not clear that any known modality can change the natural history of this disease
symptomatic patient purine analogue

(cladribine vs fludarabine)

III Leukemia Lymphoma 1993;11:105 more experience exists with cladribine; no direct comparison between these two drugs exist
presentation with hyperviscosity syndrome plasma pheresis followed by chemotherapy III J Lab Clin Med 1992;119:69

Clin Nephrol 1995;43:335

although there are not many patients reported in the literature, physiological basis for this treatment is quite convincing ; second reference cited evaluated at random effect of conventional plasm exchange with cascade filtration (Clin Nephrol 1995;43:335)
symptomatic cryoglobulenemia or neuropathy with lymphoma plasma pheresis followed by chemotherapy III Blood 1994;83:1452 as above; goal would be to reduce IgM for at least 3 months
Hodgkin's Disease at least 132 rct were performed in evaluation of Rx of HD:

1) early stages (33 trials, including one meta-analysis: Cancer 1990;65:1155

2) advanced stages/others:95

3) relapsed/resistant disease:4

initial Rx:
early stage HD (IA-IIB, not bulky, with 1 or two symptoms, ESR<30 mm/h) XRT I J Clin Oncol 1993;11:2258 staging laparotomy can be avoided in patients with favorable prognostic characteristics; recent data favors chemoRx+XRT over XRT (level I evidence); (Proc Am Soc Clin Oncol 1997;16:13a)
early stage HD (IA-IIB, not bulky, with 3 B symptoms, ESR>30 mm/h) ABVDx3,XRT,ABVDx3 I J Clin Oncol 1993;11:2258 no staging laparotomy is necessary; ABVD+XRT superior to MOPP+XRT
IIB bulky disease MOPP,XRT,ABVD III J Clin Oncol 1991;9:227

Haematologica 1987;72:327

longer follow-up exists for MOPP+XRT, which is still favored by many
IIIA2-IVB disease ABVD I N Engl J Med 1992;337:1478 superior to MOPP and was less toxic than MOPP/ABVD
IIIA1 chemoRx (MVPP) I J Clin Oncol 1984:2:892 some level II evidence suggest excellent long-term DFS with MOPPX2 followed by XRT (J Clin Oncol 1986;6:1293)
salvage treatment:
primary treatment failure with limited extent disease wide field XRT III Ann Oncol 1994;5:101

J Clin Oncol 1987;5:550

this decision will be faced in a few highly selected patients who failed either after XRT or chemoRx (for a review see: Oncology 1996;10:233
primary treatment failure with extensive disease autoBMT/PBSC III Blood 1993;81:1137 Most data comes from retrospective studies. One rct (see below) included both primary resistant disease and patients that relapsed with 1 year of Rx
relapse after XRT chemotherapy I J Clin Oncol 1986;4:838 retrospective data suggest that doxorubicin-containing regimens are superior to other chemRx; randomized trial cited showed no advantage of doxorubicin-containg regimens over others
relapse after chemo:<12 months autoBMT/PBSC I Lancet 1993;341:1051 this is only rct that evaluated high-dose vs standard treatment in this setting. Decision analysis sugegst that autoBMT should be reserved for second relapse (J Cl Oncol 1992;10:200)
relapse after chemo>12 months non-cross resistant chemoRx III J Clin Oncol 1992;10:210 24% patients alive at 22 years
residual mediastinal mass acting upon result of Ga scan III Med Hypotheses 1992;38:166 decision analytic evidence
Cutaneous T-cell Lymphoma (CTCL) more than 57 modalities alone or in combinations were recommended in the literature as possible treatment options for CTCL. However, only 7 RCT trials (1 major rct only) have ever been performed in this disease (see below)
initial treatment:
early (skin only) disease (IA,IB) mechloretamine, followed by PUVA, followed by TSEB vs I N Engl J Med 1989;321:1784 listed RCT was performed showing no advantage of aggressive chemoRx over topical treatment in early disease
" PUVA vs II Acta Dermatol Venereol 1989;69:536 PUVA was compared with PUVA+retinoids; some improvement noted (see also review Oncology 1992;6:31)
" extracorporeal photochemotherapy vs III J Clin Oncol 1991;9:1298 studies difficult to interpret-many patient treated with other therapies while undergoing photopheresis
" topical carmustine III J Am Acad Dermatol

1990;22:802

can be used in patients allergic to mechloratamine
advanced disease (II-IV) single-agent chemoRx vs III Ann Intern Med 1994;121:592

(review article)

multiple agents described (MTX, cyclophosphamide, chlorambucil, vincristine, doxorubicine, VB-16, etc); no RCT compared efficacy of one agent vs another
" combination chemoRx vs I Cancer Treat Rep 1979;63:647

this small RCT (N=24) compared one combination over another-no difference in efficacy was noted; however, no RCT compared single agent with combined chemoRx; therefore, it is not clear that combined chemoRx is superior to single-agent Rx
" interferon vs II J Clin Oncol 1990;8:155 optimal dose not known; a rct comparing high- vs low-dose Rx prematurely halted due to accrual problem (J Am Acad Dermatol 1989;20:395); most authors believe that low dose Rx is equal to high-dose Rx (for review see Ann Intern Med 1994;121:592)
" interferon+PUVA vs II J Natl Cancer Inst 1990;82:203 highest response rate and longest CR (80%) noted so far
" IFN+retinoids vs II Ann Intern Med 1994;121:592 no rct have been done to determine if combination is superior to IFN alone (single arm studies suggest no major advantage)
" retinoids II " it appears that activity of retinoids is similar to that of single-agent cytotoxic chemotherapy and IFN. One rct (Acta Dermatol Venereol 1987;67:232) showed no difference between 13 cis-retinoic acid with etretinate. No rct compared any of retinoids with chemoRx, IFN or PUVA.
salvage Rx:
pentostatin vs III review articles:

Blood 1996;88:2385

Ann Intern Med 1994;121:592

poor response in SS; has also been combined with IFN (J Clin Oncol 1992;10:1907)
cladribine vs III " small series (Blood 1996;87:906)
fludarabine vs III " no direct comparison exists with other purine analogues (J Natl Cancer Inst 1990;82:1353); has also been combined with IFN (J Clin Oncol 1994;12:2051)
monoclonal antibodies+immonoconjugate vs III " few patients studied with some promising activity
IL-2 with or without fusion toxin vs III " theoretically IL-2 should be detrimental; however, responses were noted in small series (5 of 7 patients responded with 3 patients experiencing CR: Arch Dermatol 1995;131:574)
young patients, with advanced disease, good PS autoBMT III " small series with a short lived response and short follo-up (Bone Marrow Transplant 1991;7:133, Clin Exper Dermatol 1995;20:73) alloBMT has never been studied in this disease
NonHodgkin's Lymphoma at least 177 RCT were performed in evaluation of Rx of NHL

Low-grade Lymphoma
initial therapy:
early stage disease (I/II) XRT II Semin Oncol 1990;17:51 a review article
advanced stages, asymptomatic, stable disease observation II N Engl J Med 1984;311:1471 There is also level I evidence showing that in follicular NHL (small-cell or mixed with a low tumor burden) observation (with delayed Rx) does adversely influence survival at 5 yrs in comparison with alkylating agent or IFN (J Clin Oncol 1997;15:1110)
advanced, symptomatic disease chlorambucil vs CVP vs inteferon+doxorubicin based chemoRx I Semin Oncol 1990;17:51

N Engl J Med 1993;329:1608

N Engl J Med 1993;327:1336

addition of IFN to chemoRx seems to be superior to chemoRx alone in one trial ; however,no difference in survival existed after 5 years in another trial (n Engl J Med 1993;329:1821)
salvage therapy (relapsed disease) purine analogues

(fludarabine vs cladribine)

II J Clin Oncol 1994;12:788

Semin Hematol 1994;31:28

no direct comparison between these two purine analogues exists
relapsed, responding disease,<65 years old autoBMT II J Clin Oncol 1994;12:1177 small series reported in the literature
young patient, HLA donor compatibile donor available alloBMT II J Clin Oncol 13:1096 small number of patients had alloBMT so far; mortality ~20%
follow-up (follicular NHL) molecular markers for detection of minimal residual disease ( by detection of t(14;18) III J Clin Oncol 1993;11:1668

N Engl J Med 1991;325:1525

conflicting data on prediction of relapse; healthy people also can have this translocation detected (Blood 1995;85:2528)
Aggressive NHL:
initial treatment:
early stages (IA/E) CHOP+XRT I Proc Am Soc Oncol 1996;15:411 superior to CHOPx8 for stage I
advanced stages CHOP vs MACOP-B vs m-BACOD vs ProMACE-CytaBOM I N Engl J Med 1992;327:1342

N Engl J Med 1993;328:1002

no regimen superior than other; recent data from small RCT showed that high-dose therapy chemoRx is superior to MACOP-B (N Engl J Med 1997;336:1290)
slow-responsive disease CHOP I N Engl J Med 1995;332:1045 equally effective to autoBMT
salvage therapy (relapsed disease) autoBMT I N Engl J Med 1995;333:1540 superior to conventional treatment
systemic KI-1 ALCL F-MACHOP+XRT+

autoBMT

II Blood 1996;87:1243 DFS:100% at 3 yrs
mantle-cell lymphoma chemoRx III J Clin Oncol 1995;13:2819 no optimal regimen identified; current trials focus on high-dose treatment with autoBMT/PBSC rescue
Rapidly progressive NHL
initial therapy: short-duration, high-intensity cyclophosphamide based chemoRx II J Clin Oncol 1991:9:941 best regimen not identified; many regimens reported in the literature
consolidation therapy:
patients with poor prognostic features (such as CNS involvement) BMT III J Clin Oncol 1994;12:2415 small series both with auto or alloBMT reported
CNS prophylaxis methotrexate i.t. III J Clin Oncol 1991;9:1973 some authors use in all patients, other only in high-risk patients (high LDH, tumor mass>10 cm)
salvage treatment (relapsed disease) autoBMT III J Clin Oncol 1994;12:2415 few data exists for small, non-cleaved cell NHL
NHL in HIV+ patient
CNS disease XRT III Ann Intern Med 1993;119:1093 median survival without Rx: 1 month, with XRT:3 months
oustside CNS disease chemoRx I Proc ASCO 1995;14:288 equivalent results between reduced dose chemoRx or standard chemorx (i.e. m-BACOD) with growth factor support
CNS prophylaxis methotrexate it III Cancer Control 1995;March/April:97-103 recommended if marrow is involved or histology is small,non-cleaved cell
use of antiviral therapy during chemoRx No III " PCP prophylaxis is recommended, again based on indirect data
Extranodal NHL
gastric CHOP ± XRT III Med Ped Oncol 1991;19:48

J Clin Oncol 1988;6:1125

some authors argue that surgical resection provides a survival advantage (Oncology 1993;7:29); this decision will also be have to be made on level III data
colorectal surgery followed by chemoRx ± XRT III " "; note that small-bowel and colonic lymphoma are usually diagnosed at operation
gastric MALT NHL antibiotics II Ann Intern Med 1995;122:767
skin, B-cell NHL:localized XRT III J Clin Oncol 1996;14:549
skin, B-cell NHL:multiple lesions CHOP III "
intravascular lymphomatosis chemoRx III J Clin Oncol 1994;12:2573 various regimens used
splenic lymphoma with violus lymphocytes splenectomy III Br J Haematol 1991;78:206
posttransplant NHL reduce immunosupression+

acyclovir+chemoRx if there is no response

III Blood 1995;86:3333 ProMACE-CytaBOM used in cited report
primary CNS lymphoma high-dose methotrexate-based chemoRx III J Clin Oncol 1992;10:635 based survival reported (median survival:41 months)
ocular NHL XRT III Mayo Clinic Proc 1993;68:1003

Mayo Clinic Proc 1993;68:1081

most of NHL are of low-grade histology
primary spinal NHL XRT III Mayo Clin Proc 1995;70:256 9 patients reported in the literature
testicular NHL:IPI1 CHOP+ CNS prophylaxis followed by XRT III J Clin Oncol 1995;13:1367
testicular NHL:IPI>1 experimental chemoRx III " none of patients with IPI>1 remained in remission after 8-9 years in this retrospective study
other important decisions in the management of NHL
use of IPI to tailor treatment III N Engl J Med 1993;329:987

Blood 1994;83:1165

index derived on large set of retrospective, of similarly but not identically treated patients
use of REAL classification to tailor therapy III Blood 1994;84:1361

Blood 1997;89:3909

needs prospective verification of its clinical validity
follow-up tests and timing physical exam, vs lab and imaging studies III J Clin Oncol 1991;9:1196 physical exam, LDH and gallium scan had best operating test characteristics in detection of relapse in high-gade NHL; this has never been studied in other types of NHL; similarly how often patient need to be followed has never been systematically studied